pytorch_implementation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyP15iWW8HzZq5z+eseQ1HEN",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Ryukijano/PyTorch-Differentiable-Stochastic-Differential-Equations/blob/main/pytorch_implementation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "##PIP Implementation of Differentiable SDE Solvers | GAN implementation using SDE"
      ],
      "metadata": {
        "id": "wgKpZI4f2cns"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "G42ERzBiz4-c",
        "outputId": "0f88e0a1-c71d-4aed-bbda-84d672c13528"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting torchsde\n",
            "  Downloading torchsde-0.2.5-py3-none-any.whl (59 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m59.2/59.2 KB\u001b[0m \u001b[31m2.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: torch>=1.6.0 in /usr/local/lib/python3.9/dist-packages (from torchsde) (2.0.0+cu118)\n",
            "Requirement already satisfied: scipy>=1.5 in /usr/local/lib/python3.9/dist-packages (from torchsde) (1.10.1)\n",
            "Collecting boltons>=20.2.1\n",
            "  Downloading boltons-23.0.0-py2.py3-none-any.whl (194 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m194.8/194.8 KB\u001b[0m \u001b[31m11.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.19.* in /usr/local/lib/python3.9/dist-packages (from torchsde) (1.22.4)\n",
            "Collecting trampoline>=0.1.2\n",
            "  Downloading trampoline-0.1.2-py3-none-any.whl (5.2 kB)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (3.10.7)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (3.1.2)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (4.5.0)\n",
            "Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (2.0.0)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (3.0)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (1.11.1)\n",
            "Requirement already satisfied: cmake in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.6.0->torchsde) (3.25.2)\n",
            "Requirement already satisfied: lit in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.6.0->torchsde) (16.0.0)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.9/dist-packages (from jinja2->torch>=1.6.0->torchsde) (2.1.2)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.9/dist-packages (from sympy->torch>=1.6.0->torchsde) (1.3.0)\n",
            "Installing collected packages: trampoline, boltons, torchsde\n",
            "Successfully installed boltons-23.0.0 torchsde-0.2.5 trampoline-0.1.2\n"
          ]
        }
      ],
      "source": [
        "!pip install torchsde\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "![Alt Text](https://raw.githubusercontent.com/google-research/torchsde/master/assets/latent_sde.gif)"
      ],
      "metadata": {
        "id": "WL5vqhS92HBP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchsde\n",
        "\n",
        "batch_size, state_size, brownian_size = 32, 3, 2\n",
        "t_size = 20\n",
        "\n",
        "class SDE(torch.nn.Module):\n",
        "    noise_type = 'general'\n",
        "    sde_type = 'ito'\n",
        "\n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "        self.mu = torch.nn.Linear(state_size, \n",
        "                                  state_size)\n",
        "        self.sigma = torch.nn.Linear(state_size, \n",
        "                                     state_size * brownian_size)\n",
        "\n",
        "    # Drift\n",
        "    def f(self, t, y):\n",
        "        return self.mu(y)  # shape (batch_size, state_size)\n",
        "\n",
        "    # Diffusion\n",
        "    def g(self, t, y):\n",
        "        return self.sigma(y).view(batch_size, \n",
        "                                  state_size, \n",
        "                                  brownian_size)\n",
        "\n",
        "sde = SDE()\n",
        "y0 = torch.full((batch_size, state_size), 0.1)\n",
        "ts = torch.linspace(0, 1, t_size)\n",
        "# Initial state y0, the SDE is solved over the interval [ts[0], ts[-1]].\n",
        "# ys will have shape (t_size, batch_size, state_size)\n",
        "ys = torchsde.sdeint(sde, y0, ts)"
      ],
      "metadata": {
        "id": "AbT5tMix0UBb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(ys)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5wggxZVZ3Fm7",
        "outputId": "d4801899-b66b-4c4e-94ee-77100bd250dd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([[[ 0.1000,  0.1000,  0.1000],\n",
            "         [ 0.1000,  0.1000,  0.1000],\n",
            "         [ 0.1000,  0.1000,  0.1000],\n",
            "         ...,\n",
            "         [ 0.1000,  0.1000,  0.1000],\n",
            "         [ 0.1000,  0.1000,  0.1000],\n",
            "         [ 0.1000,  0.1000,  0.1000]],\n",
            "\n",
            "        [[-0.1443,  0.1924, -0.1114],\n",
            "         [-0.2441,  0.1404,  0.0384],\n",
            "         [ 0.1805,  0.1814, -0.1288],\n",
            "         ...,\n",
            "         [ 0.0828,  0.0917,  0.0261],\n",
            "         [ 0.1004,  0.0957,  0.0209],\n",
            "         [-0.1686,  0.1380,  0.0123]],\n",
            "\n",
            "        [[ 0.0389,  0.1175, -0.0434],\n",
            "         [-0.0924,  0.0076,  0.3027],\n",
            "         [ 0.2038,  0.0988,  0.0042],\n",
            "         ...,\n",
            "         [ 0.3622, -0.0197,  0.1656],\n",
            "         [-0.0052,  0.0804,  0.0141],\n",
            "         [-0.3504,  0.1641, -0.0559]],\n",
            "\n",
            "        ...,\n",
            "\n",
            "        [[-0.1636, -0.5144,  0.3909],\n",
            "         [-0.1849, -0.5386,  0.4456],\n",
            "         [ 0.1194, -0.3123, -0.4978],\n",
            "         ...,\n",
            "         [ 0.0090, -0.4242,  0.1374],\n",
            "         [ 0.0446,  0.0332, -0.4238],\n",
            "         [-0.1282, -0.5379,  0.5305]],\n",
            "\n",
            "        [[-0.1856, -0.5607,  0.3975],\n",
            "         [-0.3789, -0.5340,  0.4589],\n",
            "         [-0.2903, -0.1932, -0.7486],\n",
            "         ...,\n",
            "         [ 0.0761, -0.4564,  0.1182],\n",
            "         [ 0.1974,  0.0311, -0.4412],\n",
            "         [-0.0848, -0.5805,  0.6372]],\n",
            "\n",
            "        [[-0.2255, -0.5787,  0.3285],\n",
            "         [-0.2859, -0.5883,  0.5540],\n",
            "         [ 0.3452, -0.0809, -0.8080],\n",
            "         ...,\n",
            "         [ 0.0434, -0.4851,  0.0588],\n",
            "         [ 0.0166, -0.0865, -0.2781],\n",
            "         [-0.1101, -0.6193,  0.6144]]], grad_fn=<StackBackward0>)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install fire\n",
        "!pip install torchcde\n",
        "!pip install torchsde"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J7qa-k-FfGkM",
        "outputId": "41c900d1-9eec-4cf8-ed22-0c14435493be"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting fire\n",
            "  Downloading fire-0.5.0.tar.gz (88 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m88.3/88.3 KB\u001b[0m \u001b[31m5.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.9/dist-packages (from fire) (1.16.0)\n",
            "Requirement already satisfied: termcolor in /usr/local/lib/python3.9/dist-packages (from fire) (2.2.0)\n",
            "Building wheels for collected packages: fire\n",
            "  Building wheel for fire (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for fire: filename=fire-0.5.0-py2.py3-none-any.whl size=116952 sha256=8f7060d6c289786cfaa623c532338bbf00d71095c55f3554ed2db5dec2da4c2e\n",
            "  Stored in directory: /root/.cache/pip/wheels/f7/f1/89/b9ea2bf8f80ec027a88fef1d354b3816b4d3d29530988972f6\n",
            "Successfully built fire\n",
            "Installing collected packages: fire\n",
            "Successfully installed fire-0.5.0\n",
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting torchcde\n",
            "  Downloading torchcde-0.2.5-py3-none-any.whl (28 kB)\n",
            "Requirement already satisfied: torch>=1.7.0 in /usr/local/lib/python3.9/dist-packages (from torchcde) (2.0.0+cu118)\n",
            "Requirement already satisfied: torchsde>=0.2.5 in /usr/local/lib/python3.9/dist-packages (from torchcde) (0.2.5)\n",
            "Collecting torchdiffeq>=0.2.0\n",
            "  Downloading torchdiffeq-0.2.3-py3-none-any.whl (31 kB)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.9/dist-packages (from torch>=1.7.0->torchcde) (1.11.1)\n",
            "Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.9/dist-packages (from torch>=1.7.0->torchcde) (2.0.0)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.9/dist-packages (from torch>=1.7.0->torchcde) (3.1.2)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.9/dist-packages (from torch>=1.7.0->torchcde) (3.10.7)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.9/dist-packages (from torch>=1.7.0->torchcde) (3.0)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.9/dist-packages (from torch>=1.7.0->torchcde) (4.5.0)\n",
            "Requirement already satisfied: cmake in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.7.0->torchcde) (3.25.2)\n",
            "Requirement already satisfied: lit in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.7.0->torchcde) (16.0.0)\n",
            "Requirement already satisfied: scipy>=1.4.0 in /usr/local/lib/python3.9/dist-packages (from torchdiffeq>=0.2.0->torchcde) (1.10.1)\n",
            "Requirement already satisfied: numpy>=1.19.* in /usr/local/lib/python3.9/dist-packages (from torchsde>=0.2.5->torchcde) (1.22.4)\n",
            "Requirement already satisfied: trampoline>=0.1.2 in /usr/local/lib/python3.9/dist-packages (from torchsde>=0.2.5->torchcde) (0.1.2)\n",
            "Requirement already satisfied: boltons>=20.2.1 in /usr/local/lib/python3.9/dist-packages (from torchsde>=0.2.5->torchcde) (23.0.0)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.9/dist-packages (from jinja2->torch>=1.7.0->torchcde) (2.1.2)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.9/dist-packages (from sympy->torch>=1.7.0->torchcde) (1.3.0)\n",
            "Installing collected packages: torchdiffeq, torchcde\n",
            "Successfully installed torchcde-0.2.5 torchdiffeq-0.2.3\n",
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: torchsde in /usr/local/lib/python3.9/dist-packages (0.2.5)\n",
            "Requirement already satisfied: torch>=1.6.0 in /usr/local/lib/python3.9/dist-packages (from torchsde) (2.0.0+cu118)\n",
            "Requirement already satisfied: trampoline>=0.1.2 in /usr/local/lib/python3.9/dist-packages (from torchsde) (0.1.2)\n",
            "Requirement already satisfied: scipy>=1.5 in /usr/local/lib/python3.9/dist-packages (from torchsde) (1.10.1)\n",
            "Requirement already satisfied: numpy>=1.19.* in /usr/local/lib/python3.9/dist-packages (from torchsde) (1.22.4)\n",
            "Requirement already satisfied: boltons>=20.2.1 in /usr/local/lib/python3.9/dist-packages (from torchsde) (23.0.0)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (4.5.0)\n",
            "Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (2.0.0)\n",
            "Requirement already satisfied: networkx in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (3.0)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (3.10.7)\n",
            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (3.1.2)\n",
            "Requirement already satisfied: sympy in /usr/local/lib/python3.9/dist-packages (from torch>=1.6.0->torchsde) (1.11.1)\n",
            "Requirement already satisfied: cmake in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.6.0->torchsde) (3.25.2)\n",
            "Requirement already satisfied: lit in /usr/local/lib/python3.9/dist-packages (from triton==2.0.0->torch>=1.6.0->torchsde) (16.0.0)\n",
            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.9/dist-packages (from jinja2->torch>=1.6.0->torchsde) (2.1.2)\n",
            "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.9/dist-packages (from sympy->torch>=1.6.0->torchsde) (1.3.0)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import fire\n",
        "import matplotlib.pyplot as plt\n",
        "import torch \n",
        "import torch.optim.swa_utils as swa_utils \n",
        "import torchcde\n",
        "import torchsde\n",
        "import tqdm"
      ],
      "metadata": {
        "id": "PbZHV1jW3Lgw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "###################\n",
        "# First some standard helper objects.\n",
        "###################\n",
        "\n",
        "class LipSwish(torch.nn.Module):\n",
        "  def forward(self, x):\n",
        "    return 0.99 * torch.nn.functional.silu(x)\n",
        "\n",
        "class MLP(torch.nn.Module):\n",
        "  def __init__(self, in_size, out_size, mlp_size, num_layers, tanh):\n",
        "    super().__init__()\n",
        "\n",
        "    model = [torch.nn.Linear(in_size, mlp_size),\n",
        "             LipSwish()]\n",
        "\n",
        "    for _ in range(num_layers -1):\n",
        "      model.append(torch.nn.Linear(in_size, mlp_size))\n",
        "            ###################\n",
        "            # LipSwish activations are useful to constrain the Lipschitz constant of the discriminator.\n",
        "            # (For simplicity we additionally use them in the generator, but that's less important.)\n",
        "            ###################\n",
        "      model.append(LipSwish())\n",
        "    model.append(torch.nn.Linear(mlp_size, out_size))\n",
        "    if tanh:\n",
        "      model.append(torch.nn.Tanh())\n",
        "    self._model = torch.nn.Sequential(*model)\n",
        "\n",
        "  def forward(self, x):\n",
        "    return self._model(x)\n",
        "\n",
        "###################\n",
        "# Now we define the SDEs.\n",
        "#\n",
        "# We begin by defining the generator SDE.\n",
        "################### \n",
        "class GeneratorFunc(torch.nn.Module):\n",
        "  sde_type = 'stratonovich'\n",
        "  noise_type = 'general'\n",
        "\n",
        "  def __init__(self, noise_size, hidden_size, mlp_size, num_layers):\n",
        "    super().__init__()\n",
        "    self._noise_size = noise_size\n",
        "    self._hidden_size = hidden_size\n",
        "\n",
        "        ###################\n",
        "        # Drift and diffusion are MLPs. They happen to be the same size.\n",
        "        # Note the final tanh nonlinearity: this is typically important for good performance, to constrain the rate of\n",
        "        # change of the hidden state.\n",
        "        # If you have problems with very high drift/diffusions then consider scaling these so that they squash to e.g.\n",
        "        # [-3, 3] rather than [-1, 1].\n",
        "        ###################\n",
        "    self._drift = MLP(1 + hidden_size, hidden_size, mlp_size, num_layers, tanh=True)\n",
        "    self._diffusion = MLP(1 + hidden_size, hidden_size * noise_size, mlp_size, num_layers, tanh=True)\n",
        "    \n",
        "  def f_and_g(self, t, x):\n",
        "    # t has shape ()\n",
        "    # x has shape (batch_size, hidden_size)\n",
        "    t = t.expand(x.size(0), 1)\n",
        "    tx = torch.cat([t, x], dim=1)\n",
        "    return self._drift(tx), self._diffusion(tx).view(x.size(0), self._hidden_size, self._noise_size)\n",
        "\n",
        "###################\n",
        "# Now we wrap it up into something that computes the SDE.\n",
        "###################\n",
        "class Generator(torch.nn.Module):\n",
        "    def __init__(self, data_size, initial_noise_size, noise_size, hidden_size, mlp_size, num_layers):\n",
        "        super().__init__()\n",
        "        self._initial_noise_size = initial_noise_size\n",
        "        self._hidden_size = hidden_size\n",
        "\n",
        "        self._initial = MLP(initial_noise_size, hidden_size, mlp_size, num_layers, tanh=False)\n",
        "        self._func = GeneratorFunc(noise_size, hidden_size, mlp_size, num_layers)\n",
        "        self._readout = torch.nn.Linear(hidden_size, data_size)\n",
        "\n",
        "    def forward(self, ts, batch_size):\n",
        "        # ts has shape (t_size,) and corresponds to the points we want to evaluate the SDE at.\n",
        "\n",
        "        ###################\n",
        "        # Actually solve the SDE.\n",
        "        ###################\n",
        "        init_noise = torch.randn(batch_size, self._initial_noise_size, device=ts.device)\n",
        "        x0 = self._initial(init_noise)\n",
        "\n",
        "        ###################\n",
        "        # We use the reversible Heun method to get accurate gradients whilst using the adjoint method.\n",
        "        ###################\n",
        "        xs = torchsde.sdeint_adjoint(self._func, x0, ts, method='reversible_heun', dt=1.0,\n",
        "                                     adjoint_method='adjoint_reversible_heun',)\n",
        "        xs = xs.transpose(0, 1)\n",
        "        ys = self._readout(xs)\n",
        "\n",
        "        ###################\n",
        "        # Normalise the data to the form that the discriminator expects, in particular including time as a channel.\n",
        "        ###################\n",
        "        ts = ts.unsqueeze(0).unsqueeze(-1).expand(batch_size, ts.size(0), 1)\n",
        "        return torchcde.linear_interpolation_coeffs(torch.cat([ts, ys], dim=2))\n",
        "\n",
        "  \n"
      ],
      "metadata": {
        "id": "ZpmH_J2afAPJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "###################\n",
        "# Next the discriminator. Here, we're going to use a neural controlled differential equation (neural CDE) as the\n",
        "# discriminator, just as in the \"Neural SDEs as Infinite-Dimensional GANs\" paper. (You could use other things as well,\n",
        "# but this is a natural choice.)\n",
        "#\n",
        "# There's actually a few different (roughly equivalent) ways of making the discriminator work. The curious reader is\n",
        "# encouraged to have a read of the comment at the bottom of this file for an in-depth explanation.\n",
        "###################\n",
        "class DiscriminatorFunc(torch.nn.Module):\n",
        "    def __init__(self, data_size, hidden_size, mlp_size, num_layers):\n",
        "        super().__init__()\n",
        "        self._data_size = data_size\n",
        "        self._hidden_size = hidden_size\n",
        "\n",
        "        # tanh is important for model performance\n",
        "        self._module = MLP(1 + hidden_size, hidden_size * (1 + data_size), mlp_size, num_layers, tanh=True)\n",
        "\n",
        "    def forward(self, t, h):\n",
        "        # t has shape ()\n",
        "        # h has shape (batch_size, hidden_size)\n",
        "        t = t.expand(h.size(0), 1)\n",
        "        th = torch.cat([t, h], dim=1)\n",
        "        return self._module(th).view(h.size(0), self._hidden_size, 1 + self._data_size)\n",
        "\n",
        "class Discriminator(torch.nn.Module):\n",
        "    def __init__(self, data_size, hidden_size, mlp_size, num_layers):\n",
        "        super().__init__()\n",
        "\n",
        "        self._initial = MLP(1 + data_size, hidden_size, mlp_size, num_layers, tanh=False)\n",
        "        self._func = DiscriminatorFunc(data_size, hidden_size, mlp_size, num_layers)\n",
        "        self._readout = torch.nn.Linear(hidden_size, 1)\n",
        "\n",
        "    def forward(self, ys_coeffs):\n",
        "        # ys_coeffs has shape (batch_size, t_size, 1 + data_size)\n",
        "        # The +1 corresponds to time. When solving CDEs, It turns out to be most natural to treat time as just another\n",
        "        # channel: in particular this makes handling irregular data quite easy, when the times may be different between\n",
        "        # different samples in the batch.\n",
        "\n",
        "        Y = torchcde.LinearInterpolation(ys_coeffs)\n",
        "        Y0 = Y.evaluate(Y.interval[0])\n",
        "        h0 = self._initial(Y0)\n",
        "        hs = torchcde.cdeint(Y, self._func, h0, Y.interval, method='reversible_heun', backend='torchsde', dt=1.0,\n",
        "                             adjoint_method='adjoint_reversible_heun',\n",
        "                             adjoint_params=(ys_coeffs,) + tuple(self._func.parameters()))\n",
        "        score = self._readout(hs[:, -1])\n",
        "        return score.mean()"
      ],
      "metadata": {
        "id": "Jsm6rN4XfWJN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "###################\n",
        "# Generate some data. For this example we generate some synthetic data from a time-dependent Ornstein-Uhlenbeck SDE.\n",
        "###################\n",
        "def get_data(batch_size, device):\n",
        "    dataset_size = 8192\n",
        "    t_size = 64\n",
        "\n",
        "    class OrnsteinUhlenbeckSDE(torch.nn.Module):\n",
        "        sde_type = 'ito'\n",
        "        noise_type = 'scalar'\n",
        "\n",
        "        def __init__(self, mu, theta, sigma):\n",
        "            super().__init__()\n",
        "            self.register_buffer('mu', torch.as_tensor(mu))\n",
        "            self.register_buffer('theta', torch.as_tensor(theta))\n",
        "            self.register_buffer('sigma', torch.as_tensor(sigma))\n",
        "\n",
        "        def f(self, t, y):\n",
        "            return self.mu * t - self.theta * y\n",
        "\n",
        "        def g(self, t, y):\n",
        "            return self.sigma.expand(y.size(0), 1, 1) * (2 * t / t_size)\n",
        "\n",
        "    ou_sde = OrnsteinUhlenbeckSDE(mu=0.02, theta=0.1, sigma=0.4).to(device)\n",
        "    y0 = torch.rand(dataset_size, device=device).unsqueeze(-1) * 2 - 1\n",
        "    ts = torch.linspace(0, t_size - 1, t_size, device=device)\n",
        "    ys = torchsde.sdeint(ou_sde, y0, ts, dt=1e-1)\n",
        "\n",
        "    ###################\n",
        "    # To demonstrate how to handle irregular data, then here we additionally drop some of the data (by setting it to\n",
        "    # NaN.)\n",
        "    ###################\n",
        "    ys_num = ys.numel()\n",
        "    to_drop = torch.randperm(ys_num)[:int(0.3 * ys_num)]\n",
        "    ys.view(-1)[to_drop] = float('nan')\n",
        "\n",
        "    ###################\n",
        "    # Typically important to normalise data. Note that the data is normalised with respect to the statistics of the\n",
        "    # initial data, _not_ the whole time series. This seems to help the learning process, presumably because if the\n",
        "    # initial condition is wrong then it's pretty hard to learn the rest of the SDE correctly.\n",
        "    ###################\n",
        "    y0_flat = ys[0].view(-1)\n",
        "    y0_not_nan = y0_flat.masked_select(~torch.isnan(y0_flat))\n",
        "    ys = (ys - y0_not_nan.mean()) / y0_not_nan.std()\n",
        "\n",
        "    ###################\n",
        "    # As discussed, time must be included as a channel for the discriminator.\n",
        "    ###################\n",
        "    ys = torch.cat([ts.unsqueeze(0).unsqueeze(-1).expand(dataset_size, t_size, 1),\n",
        "                    ys.transpose(0, 1)], dim=2)\n",
        "    # shape (dataset_size=1000, t_size=100, 1 + data_size=3)\n",
        "\n",
        "    ###################\n",
        "    # Package up.\n",
        "    ###################\n",
        "    data_size = ys.size(-1) - 1  # How many channels the data has (not including time, hence the minus one).\n",
        "    ys_coeffs = torchcde.linear_interpolation_coeffs(ys)  # as per neural CDEs.\n",
        "    dataset = torch.utils.data.TensorDataset(ys_coeffs)\n",
        "    dataloader = torch.utils.data.DataLoader(dataset, batch_size=batch_size, shuffle=True)\n",
        "\n",
        "    return ts, data_size, dataloader    \n"
      ],
      "metadata": {
        "id": "4NcUzhbNjK9w"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "###################\n",
        "# We'll plot some results at the end.\n",
        "###################\n",
        "def plot(ts, generator, dataloader, num_plot_samples, plot_locs):\n",
        "    # Get samples\n",
        "    real_samples, = next(iter(dataloader))\n",
        "    assert num_plot_samples <= real_samples.size(0)\n",
        "    real_samples = torchcde.LinearInterpolation(real_samples).evaluate(ts)\n",
        "    real_samples = real_samples[..., 1]\n",
        "\n",
        "    with torch.no_grad():\n",
        "        generated_samples = generator(ts, real_samples.size(0)).cpu()\n",
        "    generated_samples = torchcde.LinearInterpolation(generated_samples).evaluate(ts)\n",
        "    generated_samples = generated_samples[..., 1]\n",
        "\n",
        "    # Plot histograms\n",
        "    for prop in plot_locs:\n",
        "        time = int(prop * (real_samples.size(1) - 1))\n",
        "        real_samples_time = real_samples[:, time]\n",
        "        generated_samples_time = generated_samples[:, time]\n",
        "        _, bins, _ = plt.hist(real_samples_time.cpu().numpy(), bins=32, alpha=0.7, label='Real', color='dodgerblue',\n",
        "                              density=True)\n",
        "        bin_width = bins[1] - bins[0]\n",
        "        num_bins = int((generated_samples_time.max() - generated_samples_time.min()).item() // bin_width)\n",
        "        plt.hist(generated_samples_time.cpu().numpy(), bins=num_bins, alpha=0.7, label='Generated', color='crimson',\n",
        "                 density=True)\n",
        "        plt.legend()\n",
        "        plt.xlabel('Value')\n",
        "        plt.ylabel('Density')\n",
        "        plt.title(f'Marginal distribution at time {time}.')\n",
        "        plt.tight_layout()\n",
        "        plt.show()\n",
        "\n",
        "    real_samples = real_samples[:num_plot_samples]\n",
        "    generated_samples = generated_samples[:num_plot_samples]\n",
        "\n",
        "    # Plot samples\n",
        "    real_first = True\n",
        "    generated_first = True\n",
        "    for real_sample_ in real_samples:\n",
        "        kwargs = {'label': 'Real'} if real_first else {}\n",
        "        plt.plot(ts.cpu(), real_sample_.cpu(), color='dodgerblue', linewidth=0.5, alpha=0.7, **kwargs)\n",
        "        real_first = False\n",
        "    for generated_sample_ in generated_samples:\n",
        "        kwargs = {'label': 'Generated'} if generated_first else {}\n",
        "        plt.plot(ts.cpu(), generated_sample_.cpu(), color='crimson', linewidth=0.5, alpha=0.7, **kwargs)\n",
        "        generated_first = False\n",
        "    plt.legend()\n",
        "    plt.title(f\"{num_plot_samples} samples from both real and generated distributions.\")\n",
        "    plt.tight_layout()\n",
        "    plt.show()\n"
      ],
      "metadata": {
        "id": "gT7U2ya9jQ7p"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "###################\n",
        "# Now do normal GAN training, and plot the results.\n",
        "#\n",
        "# GANs are famously tricky and SDEs trained as GANs are no exception. Hopefully you can learn from our experience and\n",
        "# get these working faster than we did -- we found that several tricks were often helpful to get this working in a\n",
        "# reasonable fashion:\n",
        "# - Stochastic weight averaging (average out the oscillations in GAN training).\n",
        "# - Weight decay (reduce the oscillations in GAN training).\n",
        "# - Final tanh nonlinearities in the architectures of the vector fields, as above. (To avoid the model blowing up.)\n",
        "# - Adadelta (interestingly seems to be a lot better than either SGD or Adam).\n",
        "# - Choosing a good learning rate (always important).\n",
        "# - Scaling the weights at initialisation to be roughly the right size (chosen through empirical trial-and-error).\n",
        "###################\n",
        "\n",
        "def evaluate_loss(ts, batch_size, dataloader, generator, discriminator):\n",
        "    with torch.no_grad():\n",
        "        total_samples = 0\n",
        "        total_loss = 0\n",
        "        for real_samples, in dataloader:\n",
        "            generated_samples = generator(ts, batch_size)\n",
        "            generated_score = discriminator(generated_samples)\n",
        "            real_score = discriminator(real_samples)\n",
        "            loss = generated_score - real_score\n",
        "            total_samples += batch_size\n",
        "            total_loss += loss.item() * batch_size\n",
        "    return total_loss / total_samples\n",
        "\n",
        "\n",
        "def main(\n",
        "        # Architectural hyperparameters. These are quite small for illustrative purposes.\n",
        "        initial_noise_size=5,  # How many noise dimensions to sample at the start of the SDE.\n",
        "        noise_size=3,          # How many dimensions the Brownian motion has.\n",
        "        hidden_size=16,        # How big the hidden size of the generator SDE and the discriminator CDE are.\n",
        "        mlp_size=16,           # How big the layers in the various MLPs are.\n",
        "        num_layers=1,          # How many hidden layers to have in the various MLPs.\n",
        "\n",
        "        # Training hyperparameters. Be prepared to tune these very carefully, as with any GAN.\n",
        "        generator_lr=2e-4,      # Learning rate often needs careful tuning to the problem.\n",
        "        discriminator_lr=1e-3,  # Learning rate often needs careful tuning to the problem.\n",
        "        batch_size=1024,        # Batch size.\n",
        "        steps=10000,            # How many steps to train both generator and discriminator for.\n",
        "        init_mult1=3,           # Changing the initial parameter size can help.\n",
        "        init_mult2=0.5,         #\n",
        "        weight_decay=0.01,      # Weight decay.\n",
        "        swa_step_start=5000,    # When to start using stochastic weight averaging.\n",
        "\n",
        "        # Evaluation and plotting hyperparameters\n",
        "        steps_per_print=10,                   # How often to print the loss.\n",
        "        num_plot_samples=50,                  # How many samples to use on the plots at the end.\n",
        "        plot_locs=(0.1, 0.3, 0.5, 0.7, 0.9),  # Plot some marginal distributions at this proportion of the way along.\n",
        "):\n",
        "    is_cuda = torch.cuda.is_available()\n",
        "    device = 'cuda' if is_cuda else 'cpu'\n",
        "    if not is_cuda:\n",
        "        print(\"Warning: CUDA not available; falling back to CPU but this is likely to be very slow.\")\n",
        "\n",
        "    # Data\n",
        "    ts, data_size, train_dataloader = get_data(batch_size=batch_size, device=device)\n",
        "    infinite_train_dataloader = (elem for it in iter(lambda: train_dataloader, None) for elem in it)\n",
        "\n",
        "    # Models\n",
        "    generator = Generator(data_size, initial_noise_size, noise_size, hidden_size, mlp_size, num_layers).to(device)\n",
        "    discriminator = Discriminator(data_size, hidden_size, mlp_size, num_layers).to(device)\n",
        "    # Weight averaging really helps with GAN training.\n",
        "    averaged_generator = swa_utils.AveragedModel(generator)\n",
        "    averaged_discriminator = swa_utils.AveragedModel(discriminator)\n",
        "\n",
        "    # Picking a good initialisation is important!\n",
        "    # In this case these were picked by making the parameters for the t=0 part of the generator be roughly the right\n",
        "    # size that the untrained t=0 distribution has a similar variance to the t=0 data distribution.\n",
        "    # Then the func parameters were adjusted so that the t>0 distribution looked like it had about the right variance.\n",
        "    # What we're doing here is very crude -- one can definitely imagine smarter ways of doing things.\n",
        "    # (e.g. pretraining the t=0 distribution)\n",
        "    with torch.no_grad():\n",
        "        for param in generator._initial.parameters():\n",
        "            param *= init_mult1\n",
        "        for param in generator._func.parameters():\n",
        "            param *= init_mult2\n",
        "\n",
        "    # Optimisers. Adadelta turns out to be a much better choice than SGD or Adam, interestingly.\n",
        "    generator_optimiser = torch.optim.Adadelta(generator.parameters(), lr=generator_lr, weight_decay=weight_decay)\n",
        "    discriminator_optimiser = torch.optim.Adadelta(discriminator.parameters(), lr=discriminator_lr,\n",
        "                                                   weight_decay=weight_decay)\n",
        "\n",
        "    # Train both generator and discriminator.\n",
        "    trange = tqdm.tqdm(range(steps))\n",
        "    for step in trange:\n",
        "        real_samples, = next(infinite_train_dataloader)\n",
        "\n",
        "        generated_samples = generator(ts, batch_size)\n",
        "        generated_score = discriminator(generated_samples)\n",
        "        real_score = discriminator(real_samples)\n",
        "        loss = generated_score - real_score\n",
        "        loss.backward()\n",
        "\n",
        "        for param in generator.parameters():\n",
        "            param.grad *= -1\n",
        "        generator_optimiser.step()\n",
        "        discriminator_optimiser.step()\n",
        "        generator_optimiser.zero_grad()\n",
        "        discriminator_optimiser.zero_grad()\n",
        "\n",
        "        ###################\n",
        "        # We constrain the Lipschitz constant of the discriminator using carefully-chosen clipping (and the use of\n",
        "        # LipSwish activation functions).\n",
        "        ###################\n",
        "        with torch.no_grad():\n",
        "            for module in discriminator.modules():\n",
        "                if isinstance(module, torch.nn.Linear):\n",
        "                    lim = 1 / module.out_features\n",
        "                    module.weight.clamp_(-lim, lim)\n",
        "\n",
        "        # Stochastic weight averaging typically improves performance.\n",
        "        if step > swa_step_start:\n",
        "            averaged_generator.update_parameters(generator)\n",
        "            averaged_discriminator.update_parameters(discriminator)\n",
        "\n",
        "        if (step % steps_per_print) == 0 or step == steps - 1:\n",
        "            total_unaveraged_loss = evaluate_loss(ts, batch_size, train_dataloader, generator, discriminator)\n",
        "            if step > swa_step_start:\n",
        "                total_averaged_loss = evaluate_loss(ts, batch_size, train_dataloader, averaged_generator.module,\n",
        "                                                    averaged_discriminator.module)\n",
        "                trange.write(f\"Step: {step:3} Loss (unaveraged): {total_unaveraged_loss:.4f} \"\n",
        "                             f\"Loss (averaged): {total_averaged_loss:.4f}\")\n",
        "            else:\n",
        "                trange.write(f\"Step: {step:3} Loss (unaveraged): {total_unaveraged_loss:.4f}\")\n",
        "    generator.load_state_dict(averaged_generator.module.state_dict())\n",
        "    discriminator.load_state_dict(averaged_discriminator.module.state_dict())\n",
        "\n",
        "    _, _, test_dataloader = get_data(batch_size=batch_size, device=device)\n",
        "\n",
        "    plot(ts, generator, test_dataloader, num_plot_samples, plot_locs)\n",
        "\n",
        "\n",
        "if __name__ == '__main__':\n",
        "    fire.Fire(main)\n",
        "\n",
        "###################\n",
        "# And that's (one way of doing) an SDE as a GAN. Have fun.\n",
        "###################"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "SnmOd31sjUN3",
        "outputId": "73f8a754-00fe-425a-bdd3-531a7ef5368c"
      },
      "execution_count": null,
      "outputs": [
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 1/10000 [00:02<7:37:27,  2.75s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:   0 Loss (unaveraged): -0.0330\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 11/10000 [00:11<3:29:26,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  10 Loss (unaveraged): -0.0343\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 21/10000 [00:21<4:04:11,  1.47s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  20 Loss (unaveraged): -0.0364\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 31/10000 [00:30<3:31:31,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  30 Loss (unaveraged): -0.0331\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 41/10000 [00:40<3:47:45,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  40 Loss (unaveraged): -0.0362\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 51/10000 [00:48<3:24:44,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  50 Loss (unaveraged): -0.0345\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 61/10000 [00:57<3:23:58,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  60 Loss (unaveraged): -0.0331\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 71/10000 [01:07<3:33:37,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  70 Loss (unaveraged): -0.0358\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 81/10000 [01:16<3:52:17,  1.41s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  80 Loss (unaveraged): -0.0360\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 91/10000 [01:24<3:20:53,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step:  90 Loss (unaveraged): -0.0359\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 101/10000 [01:34<3:23:47,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 100 Loss (unaveraged): -0.0369\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 111/10000 [01:43<3:32:31,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 110 Loss (unaveraged): -0.0383\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|          | 121/10000 [01:52<3:47:21,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 120 Loss (unaveraged): -0.0366\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|▏         | 131/10000 [02:01<3:19:48,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 130 Loss (unaveraged): -0.0375\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  1%|▏         | 141/10000 [02:10<3:23:35,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 140 Loss (unaveraged): -0.0366\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 151/10000 [02:19<3:36:36,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 150 Loss (unaveraged): -0.0368\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 161/10000 [02:29<3:47:20,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 160 Loss (unaveraged): -0.0378\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 171/10000 [02:37<3:18:21,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 170 Loss (unaveraged): -0.0371\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 181/10000 [02:46<3:27:44,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 180 Loss (unaveraged): -0.0390\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 191/10000 [02:56<3:30:15,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 190 Loss (unaveraged): -0.0378\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 201/10000 [03:05<3:45:53,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 200 Loss (unaveraged): -0.0374\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 211/10000 [03:13<3:17:59,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 210 Loss (unaveraged): -0.0364\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 221/10000 [03:23<3:22:00,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 220 Loss (unaveraged): -0.0343\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 231/10000 [03:32<3:30:41,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 230 Loss (unaveraged): -0.0391\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  2%|▏         | 241/10000 [03:42<3:50:45,  1.42s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 240 Loss (unaveraged): -0.0382\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 251/10000 [03:51<3:26:13,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 250 Loss (unaveraged): -0.0350\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 261/10000 [04:00<3:23:17,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 260 Loss (unaveraged): -0.0350\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 271/10000 [04:09<3:22:05,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 270 Loss (unaveraged): -0.0352\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 281/10000 [04:19<3:45:14,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 280 Loss (unaveraged): -0.0362\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 291/10000 [04:27<3:18:08,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 290 Loss (unaveraged): -0.0317\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 301/10000 [04:37<3:17:31,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 300 Loss (unaveraged): -0.0332\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 311/10000 [04:46<3:23:44,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 310 Loss (unaveraged): -0.0315\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 321/10000 [04:55<3:47:45,  1.41s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 320 Loss (unaveraged): -0.0300\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 331/10000 [05:04<3:17:42,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 330 Loss (unaveraged): -0.0316\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  3%|▎         | 341/10000 [05:13<3:16:45,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 340 Loss (unaveraged): -0.0306\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▎         | 351/10000 [05:22<3:21:23,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 350 Loss (unaveraged): -0.0318\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▎         | 361/10000 [05:32<3:42:46,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 360 Loss (unaveraged): -0.0307\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▎         | 371/10000 [05:40<3:14:33,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 370 Loss (unaveraged): -0.0302\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 381/10000 [05:49<3:16:40,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 380 Loss (unaveraged): -0.0283\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 391/10000 [05:59<3:26:32,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 390 Loss (unaveraged): -0.0274\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 401/10000 [06:08<3:39:41,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 400 Loss (unaveraged): -0.0271\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 411/10000 [06:16<3:14:46,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 410 Loss (unaveraged): -0.0276\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 421/10000 [06:26<3:16:13,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 420 Loss (unaveraged): -0.0271\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 431/10000 [06:35<3:22:57,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 430 Loss (unaveraged): -0.0288\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  4%|▍         | 441/10000 [06:44<3:40:03,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 440 Loss (unaveraged): -0.0274\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▍         | 451/10000 [06:53<3:14:53,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 450 Loss (unaveraged): -0.0264\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▍         | 461/10000 [07:02<3:20:04,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 460 Loss (unaveraged): -0.0274\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▍         | 471/10000 [07:11<3:18:05,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 470 Loss (unaveraged): -0.0271\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▍         | 481/10000 [07:20<3:40:49,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 480 Loss (unaveraged): -0.0280\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▍         | 491/10000 [07:29<3:13:20,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 490 Loss (unaveraged): -0.0275\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▌         | 501/10000 [07:38<3:11:32,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 500 Loss (unaveraged): -0.0280\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▌         | 511/10000 [07:48<3:20:26,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 510 Loss (unaveraged): -0.0292\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▌         | 521/10000 [07:57<3:36:29,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 520 Loss (unaveraged): -0.0302\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▌         | 531/10000 [08:05<3:14:26,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 530 Loss (unaveraged): -0.0319\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  5%|▌         | 541/10000 [08:15<3:12:47,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 540 Loss (unaveraged): -0.0319\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 551/10000 [08:24<3:21:03,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 550 Loss (unaveraged): -0.0320\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 561/10000 [08:33<3:41:21,  1.41s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 560 Loss (unaveraged): -0.0341\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 571/10000 [08:42<3:09:10,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 570 Loss (unaveraged): -0.0364\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 581/10000 [08:51<3:13:12,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 580 Loss (unaveraged): -0.0391\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 591/10000 [09:00<3:19:16,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 590 Loss (unaveraged): -0.0389\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 601/10000 [09:09<3:38:29,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 600 Loss (unaveraged): -0.0410\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 611/10000 [09:18<3:10:07,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 610 Loss (unaveraged): -0.0423\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▌         | 621/10000 [09:27<3:10:48,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 620 Loss (unaveraged): -0.0441\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▋         | 631/10000 [09:36<3:26:06,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 630 Loss (unaveraged): -0.0437\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  6%|▋         | 641/10000 [09:46<3:37:42,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 640 Loss (unaveraged): -0.0450\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 651/10000 [09:54<3:11:43,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 650 Loss (unaveraged): -0.0462\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 661/10000 [10:04<3:10:59,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 660 Loss (unaveraged): -0.0466\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 671/10000 [10:13<3:16:07,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 670 Loss (unaveraged): -0.0489\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 681/10000 [10:22<3:42:50,  1.43s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 680 Loss (unaveraged): -0.0487\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 691/10000 [10:31<3:08:57,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 690 Loss (unaveraged): -0.0494\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 701/10000 [10:40<3:12:51,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 700 Loss (unaveraged): -0.0486\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 711/10000 [10:49<3:15:06,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 710 Loss (unaveraged): -0.0494\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 721/10000 [10:59<3:33:57,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 720 Loss (unaveraged): -0.0512\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 731/10000 [11:07<3:12:57,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 730 Loss (unaveraged): -0.0481\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  7%|▋         | 741/10000 [11:16<3:07:43,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 740 Loss (unaveraged): -0.0499\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 751/10000 [11:26<3:14:12,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 750 Loss (unaveraged): -0.0495\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 761/10000 [11:35<3:31:56,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 760 Loss (unaveraged): -0.0492\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 771/10000 [11:43<3:05:44,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 770 Loss (unaveraged): -0.0488\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 781/10000 [11:53<3:10:49,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 780 Loss (unaveraged): -0.0501\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 791/10000 [12:02<3:14:43,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 790 Loss (unaveraged): -0.0487\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 801/10000 [12:11<3:36:56,  1.42s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 800 Loss (unaveraged): -0.0492\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 811/10000 [12:20<3:05:53,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 810 Loss (unaveraged): -0.0511\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 821/10000 [12:29<3:07:29,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 820 Loss (unaveraged): -0.0484\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 831/10000 [12:38<3:16:01,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 830 Loss (unaveraged): -0.0500\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  8%|▊         | 841/10000 [12:47<3:31:49,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 840 Loss (unaveraged): -0.0508\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▊         | 851/10000 [12:56<3:07:07,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 850 Loss (unaveraged): -0.0531\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▊         | 861/10000 [13:05<3:07:14,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 860 Loss (unaveraged): -0.0526\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▊         | 871/10000 [13:15<3:18:43,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 870 Loss (unaveraged): -0.0548\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 881/10000 [13:24<3:28:57,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 880 Loss (unaveraged): -0.0544\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 891/10000 [13:32<3:05:06,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 890 Loss (unaveraged): -0.0571\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 901/10000 [13:41<3:04:05,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 900 Loss (unaveraged): -0.0564\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 911/10000 [13:51<3:17:17,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 910 Loss (unaveraged): -0.0597\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 921/10000 [14:00<3:30:43,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 920 Loss (unaveraged): -0.0589\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 931/10000 [14:09<3:04:42,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 930 Loss (unaveraged): -0.0555\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  9%|▉         | 941/10000 [14:18<3:06:55,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 940 Loss (unaveraged): -0.0582\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|▉         | 951/10000 [14:27<3:10:35,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 950 Loss (unaveraged): -0.0581\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|▉         | 961/10000 [14:36<3:29:55,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 960 Loss (unaveraged): -0.0587\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|▉         | 971/10000 [14:45<3:03:42,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 970 Loss (unaveraged): -0.0612\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|▉         | 981/10000 [14:54<3:04:21,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 980 Loss (unaveraged): -0.0597\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|▉         | 991/10000 [15:03<3:10:27,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 990 Loss (unaveraged): -0.0595\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|█         | 1001/10000 [15:13<3:26:27,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1000 Loss (unaveraged): -0.0617\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|█         | 1011/10000 [15:21<3:06:24,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1010 Loss (unaveraged): -0.0626\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|█         | 1021/10000 [15:30<3:04:57,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1020 Loss (unaveraged): -0.0633\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|█         | 1031/10000 [15:40<3:09:40,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1030 Loss (unaveraged): -0.0637\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 10%|█         | 1041/10000 [15:49<3:28:51,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1040 Loss (unaveraged): -0.0665\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1051/10000 [15:58<3:04:37,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1050 Loss (unaveraged): -0.0669\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1061/10000 [16:07<3:02:57,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1060 Loss (unaveraged): -0.0691\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1071/10000 [16:16<3:06:24,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1070 Loss (unaveraged): -0.0701\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1081/10000 [16:25<3:28:45,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1080 Loss (unaveraged): -0.0711\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1091/10000 [16:34<3:01:25,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1090 Loss (unaveraged): -0.0724\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1101/10000 [16:43<3:01:50,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1100 Loss (unaveraged): -0.0727\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1111/10000 [16:52<3:11:28,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1110 Loss (unaveraged): -0.0748\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█         | 1121/10000 [17:01<3:27:45,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1120 Loss (unaveraged): -0.0756\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█▏        | 1131/10000 [17:10<3:00:36,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1130 Loss (unaveraged): -0.0789\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 11%|█▏        | 1141/10000 [17:19<2:59:18,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1140 Loss (unaveraged): -0.0803\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1151/10000 [17:29<3:13:07,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1150 Loss (unaveraged): -0.0822\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1161/10000 [17:38<3:25:16,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1160 Loss (unaveraged): -0.0834\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1171/10000 [17:47<3:06:39,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1170 Loss (unaveraged): -0.0850\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1181/10000 [17:56<3:02:03,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1180 Loss (unaveraged): -0.0884\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1191/10000 [18:06<3:11:48,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1190 Loss (unaveraged): -0.0896\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1201/10000 [18:15<3:21:08,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1200 Loss (unaveraged): -0.0922\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1211/10000 [18:24<2:58:35,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1210 Loss (unaveraged): -0.0924\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1221/10000 [18:33<3:04:16,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1220 Loss (unaveraged): -0.0931\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1231/10000 [18:42<3:04:52,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1230 Loss (unaveraged): -0.0976\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 12%|█▏        | 1241/10000 [18:51<3:22:13,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1240 Loss (unaveraged): -0.1013\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1251/10000 [19:00<2:58:00,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1250 Loss (unaveraged): -0.1029\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1261/10000 [19:09<2:57:39,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1260 Loss (unaveraged): -0.1044\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1271/10000 [19:19<3:03:47,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1270 Loss (unaveraged): -0.1079\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1281/10000 [19:28<3:17:50,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1280 Loss (unaveraged): -0.1088\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1291/10000 [19:37<3:06:54,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1290 Loss (unaveraged): -0.1125\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1301/10000 [19:46<2:57:37,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1300 Loss (unaveraged): -0.1144\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1311/10000 [19:55<3:02:10,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1310 Loss (unaveraged): -0.1170\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1321/10000 [20:04<3:18:54,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1320 Loss (unaveraged): -0.1215\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1331/10000 [20:13<2:53:31,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1330 Loss (unaveraged): -0.1203\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 13%|█▎        | 1341/10000 [20:22<2:54:37,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1340 Loss (unaveraged): -0.1261\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▎        | 1351/10000 [20:31<2:59:28,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1350 Loss (unaveraged): -0.1257\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▎        | 1361/10000 [20:40<3:20:48,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1360 Loss (unaveraged): -0.1291\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▎        | 1371/10000 [20:49<2:54:30,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1370 Loss (unaveraged): -0.1296\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1381/10000 [20:58<2:53:41,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1380 Loss (unaveraged): -0.1337\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1391/10000 [21:07<3:07:31,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1390 Loss (unaveraged): -0.1358\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1401/10000 [21:17<3:18:53,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1400 Loss (unaveraged): -0.1342\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1411/10000 [21:25<2:53:53,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1410 Loss (unaveraged): -0.1365\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1421/10000 [21:34<2:56:02,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1420 Loss (unaveraged): -0.1389\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1431/10000 [21:44<3:04:04,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1430 Loss (unaveraged): -0.1407\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 14%|█▍        | 1441/10000 [21:53<3:14:46,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1440 Loss (unaveraged): -0.1449\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▍        | 1451/10000 [22:01<2:55:11,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1450 Loss (unaveraged): -0.1378\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▍        | 1461/10000 [22:11<3:00:24,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1460 Loss (unaveraged): -0.1407\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▍        | 1471/10000 [22:20<3:00:08,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1470 Loss (unaveraged): -0.1433\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▍        | 1481/10000 [22:29<3:18:06,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1480 Loss (unaveraged): -0.1421\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▍        | 1491/10000 [22:38<2:58:00,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1490 Loss (unaveraged): -0.1426\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▌        | 1501/10000 [22:47<2:54:40,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1500 Loss (unaveraged): -0.1360\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▌        | 1511/10000 [22:56<2:58:59,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1510 Loss (unaveraged): -0.1325\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▌        | 1521/10000 [23:06<3:18:06,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1520 Loss (unaveraged): -0.1363\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▌        | 1531/10000 [23:14<2:51:52,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1530 Loss (unaveraged): -0.1297\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 15%|█▌        | 1541/10000 [23:23<2:52:26,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1540 Loss (unaveraged): -0.1338\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1551/10000 [23:33<2:59:43,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1550 Loss (unaveraged): -0.1332\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1561/10000 [23:42<3:19:55,  1.42s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1560 Loss (unaveraged): -0.1293\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1571/10000 [23:51<2:50:06,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1570 Loss (unaveraged): -0.1296\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1581/10000 [24:00<2:55:14,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1580 Loss (unaveraged): -0.1234\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1591/10000 [24:09<3:00:06,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1590 Loss (unaveraged): -0.1194\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1601/10000 [24:19<3:15:47,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1600 Loss (unaveraged): -0.1219\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1611/10000 [24:27<2:50:45,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1610 Loss (unaveraged): -0.1190\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▌        | 1621/10000 [24:37<2:50:38,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1620 Loss (unaveraged): -0.1191\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▋        | 1631/10000 [24:46<3:00:26,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1630 Loss (unaveraged): -0.1176\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 16%|█▋        | 1641/10000 [24:55<3:11:29,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1640 Loss (unaveraged): -0.1140\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1651/10000 [25:04<2:49:59,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1650 Loss (unaveraged): -0.1176\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1661/10000 [25:13<2:56:02,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1660 Loss (unaveraged): -0.1150\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1671/10000 [25:22<2:56:03,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1670 Loss (unaveraged): -0.1151\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1681/10000 [25:32<3:12:25,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1680 Loss (unaveraged): -0.1216\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1691/10000 [25:40<2:49:40,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1690 Loss (unaveraged): -0.1191\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1701/10000 [25:50<2:50:46,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1700 Loss (unaveraged): -0.1236\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1711/10000 [25:59<2:53:52,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1710 Loss (unaveraged): -0.1228\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1721/10000 [26:08<3:10:05,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1720 Loss (unaveraged): -0.1248\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1731/10000 [26:17<2:51:19,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1730 Loss (unaveraged): -0.1292\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 17%|█▋        | 1741/10000 [26:26<2:49:06,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1740 Loss (unaveraged): -0.1324\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1751/10000 [26:35<2:55:13,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1750 Loss (unaveraged): -0.1328\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1761/10000 [26:44<3:11:50,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1760 Loss (unaveraged): -0.1336\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1771/10000 [26:53<2:48:44,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1770 Loss (unaveraged): -0.1330\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1781/10000 [27:02<2:48:18,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1780 Loss (unaveraged): -0.1403\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1791/10000 [27:11<2:54:32,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1790 Loss (unaveraged): -0.1356\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1801/10000 [27:21<3:12:01,  1.41s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1800 Loss (unaveraged): -0.1370\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1811/10000 [27:29<2:45:19,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1810 Loss (unaveraged): -0.1358\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1821/10000 [27:39<2:46:12,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1820 Loss (unaveraged): -0.1279\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1831/10000 [27:48<2:50:57,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1830 Loss (unaveraged): -0.1333\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 18%|█▊        | 1841/10000 [27:57<3:08:15,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1840 Loss (unaveraged): -0.1236\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▊        | 1851/10000 [28:06<2:47:00,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1850 Loss (unaveraged): -0.1239\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▊        | 1861/10000 [28:15<2:46:23,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1860 Loss (unaveraged): -0.1112\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▊        | 1871/10000 [28:24<2:55:10,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1870 Loss (unaveraged): -0.1113\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1881/10000 [28:33<3:06:56,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1880 Loss (unaveraged): -0.0921\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1891/10000 [28:42<2:46:03,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1890 Loss (unaveraged): -0.0905\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1901/10000 [28:51<2:45:09,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1900 Loss (unaveraged): -0.0764\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1911/10000 [29:00<2:49:29,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1910 Loss (unaveraged): -0.0732\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1921/10000 [29:09<3:06:02,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1920 Loss (unaveraged): -0.0584\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1931/10000 [29:18<2:43:58,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1930 Loss (unaveraged): -0.0517\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 19%|█▉        | 1941/10000 [29:27<2:46:07,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1940 Loss (unaveraged): -0.0461\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|█▉        | 1951/10000 [29:37<2:48:51,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1950 Loss (unaveraged): -0.0362\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|█▉        | 1961/10000 [29:46<3:05:42,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1960 Loss (unaveraged): -0.0273\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|█▉        | 1971/10000 [29:54<2:46:44,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1970 Loss (unaveraged): -0.0226\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|█▉        | 1981/10000 [30:03<2:44:12,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1980 Loss (unaveraged): -0.0146\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|█▉        | 1991/10000 [30:13<2:49:50,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 1990 Loss (unaveraged): -0.0139\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|██        | 2001/10000 [30:22<3:02:22,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2000 Loss (unaveraged): -0.0079\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|██        | 2011/10000 [30:30<2:42:32,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2010 Loss (unaveraged): -0.0107\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|██        | 2021/10000 [30:40<2:44:09,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2020 Loss (unaveraged): -0.0047\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|██        | 2031/10000 [30:49<2:51:13,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2030 Loss (unaveraged): -0.0104\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 20%|██        | 2041/10000 [30:58<3:07:08,  1.41s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2040 Loss (unaveraged): -0.0039\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2051/10000 [31:07<2:40:28,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2050 Loss (unaveraged): -0.0087\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2061/10000 [31:16<2:43:34,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2060 Loss (unaveraged): -0.0045\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2071/10000 [31:25<2:50:06,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2070 Loss (unaveraged): -0.0000\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2081/10000 [31:34<3:00:55,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2080 Loss (unaveraged): 0.0033\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2091/10000 [31:43<2:39:58,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2090 Loss (unaveraged): 0.0015\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2101/10000 [31:52<2:40:08,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2100 Loss (unaveraged): 0.0006\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2111/10000 [32:01<2:52:05,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2110 Loss (unaveraged): 0.0063\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██        | 2121/10000 [32:10<2:54:20,  1.33s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2120 Loss (unaveraged): 0.0143\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██▏       | 2131/10000 [32:19<2:39:03,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2130 Loss (unaveraged): 0.0153\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 21%|██▏       | 2141/10000 [32:28<2:43:35,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2140 Loss (unaveraged): 0.0175\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2151/10000 [32:37<2:50:40,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2150 Loss (unaveraged): 0.0223\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2161/10000 [32:46<2:49:09,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2160 Loss (unaveraged): 0.0272\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2171/10000 [32:55<2:37:41,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2170 Loss (unaveraged): 0.0414\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2181/10000 [33:04<2:42:42,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2180 Loss (unaveraged): 0.0415\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2191/10000 [33:13<2:54:01,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2190 Loss (unaveraged): 0.0440\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2201/10000 [33:22<2:47:57,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2200 Loss (unaveraged): 0.0566\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2211/10000 [33:31<2:42:36,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2210 Loss (unaveraged): 0.0621\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2221/10000 [33:41<2:39:40,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2220 Loss (unaveraged): 0.0691\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2231/10000 [33:50<2:54:25,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2230 Loss (unaveraged): 0.0771\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 22%|██▏       | 2241/10000 [33:58<2:42:23,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2240 Loss (unaveraged): 0.0784\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2251/10000 [34:07<2:37:16,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2250 Loss (unaveraged): 0.0819\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2261/10000 [34:17<2:38:23,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2260 Loss (unaveraged): 0.0859\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2271/10000 [34:26<2:56:09,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2270 Loss (unaveraged): 0.0876\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2281/10000 [34:34<2:42:04,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2280 Loss (unaveraged): 0.0939\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2291/10000 [34:43<2:37:34,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2290 Loss (unaveraged): 0.0868\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2301/10000 [34:53<2:41:09,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2300 Loss (unaveraged): 0.0890\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2311/10000 [35:02<2:55:55,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2310 Loss (unaveraged): 0.0869\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2321/10000 [35:11<2:33:37,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2320 Loss (unaveraged): 0.0818\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2331/10000 [35:20<2:35:50,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2330 Loss (unaveraged): 0.0755\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 23%|██▎       | 2341/10000 [35:29<2:41:18,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2340 Loss (unaveraged): 0.0723\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▎       | 2351/10000 [35:38<2:57:41,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2350 Loss (unaveraged): 0.0651\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▎       | 2361/10000 [35:47<2:43:47,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2360 Loss (unaveraged): 0.0592\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▎       | 2371/10000 [35:57<2:37:42,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2370 Loss (unaveraged): 0.0527\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2381/10000 [36:06<2:41:09,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2380 Loss (unaveraged): 0.0444\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2391/10000 [36:15<2:52:55,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2390 Loss (unaveraged): 0.0350\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2401/10000 [36:24<2:37:43,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2400 Loss (unaveraged): 0.0285\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2411/10000 [36:33<2:31:20,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2410 Loss (unaveraged): 0.0230\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2421/10000 [36:42<2:39:58,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2420 Loss (unaveraged): 0.0148\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2431/10000 [36:51<2:51:45,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2430 Loss (unaveraged): 0.0078\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 24%|██▍       | 2441/10000 [37:00<2:31:54,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2440 Loss (unaveraged): 0.0070\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▍       | 2451/10000 [37:09<2:37:55,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2450 Loss (unaveraged): 0.0007\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▍       | 2461/10000 [37:18<2:38:16,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2460 Loss (unaveraged): -0.0030\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▍       | 2471/10000 [37:28<2:55:34,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2470 Loss (unaveraged): -0.0019\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▍       | 2481/10000 [37:36<2:32:07,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2480 Loss (unaveraged): -0.0044\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▍       | 2491/10000 [37:45<2:32:45,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2490 Loss (unaveraged): -0.0057\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▌       | 2501/10000 [37:55<2:36:38,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2500 Loss (unaveraged): -0.0064\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▌       | 2511/10000 [38:04<2:51:59,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2510 Loss (unaveraged): -0.0040\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▌       | 2521/10000 [38:12<2:33:00,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2520 Loss (unaveraged): -0.0035\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▌       | 2531/10000 [38:22<2:32:30,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2530 Loss (unaveraged): -0.0037\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 25%|██▌       | 2541/10000 [38:31<2:36:24,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2540 Loss (unaveraged): -0.0020\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2551/10000 [38:40<2:56:19,  1.42s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2550 Loss (unaveraged): -0.0006\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2561/10000 [38:49<2:32:42,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2560 Loss (unaveraged): -0.0039\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2571/10000 [38:58<2:31:06,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2570 Loss (unaveraged): -0.0054\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2581/10000 [39:07<2:35:27,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2580 Loss (unaveraged): -0.0077\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2591/10000 [39:16<2:49:27,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2590 Loss (unaveraged): -0.0090\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2601/10000 [39:25<2:30:44,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2600 Loss (unaveraged): -0.0141\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2611/10000 [39:34<2:29:31,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2610 Loss (unaveraged): -0.0189\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▌       | 2621/10000 [39:43<2:39:50,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2620 Loss (unaveraged): -0.0268\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▋       | 2631/10000 [39:52<2:49:57,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2630 Loss (unaveraged): -0.0335\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 26%|██▋       | 2641/10000 [40:01<2:30:06,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2640 Loss (unaveraged): -0.0402\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2651/10000 [40:10<2:31:27,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2650 Loss (unaveraged): -0.0496\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2661/10000 [40:20<2:36:03,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2660 Loss (unaveraged): -0.0605\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2671/10000 [40:29<2:49:13,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2670 Loss (unaveraged): -0.0673\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2681/10000 [40:37<2:27:40,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2680 Loss (unaveraged): -0.0776\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2691/10000 [40:47<2:32:10,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2690 Loss (unaveraged): -0.0876\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2701/10000 [40:56<2:34:36,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2700 Loss (unaveraged): -0.0967\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2711/10000 [41:05<2:47:16,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2710 Loss (unaveraged): -0.1061\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2721/10000 [41:13<2:25:49,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2720 Loss (unaveraged): -0.1141\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2731/10000 [41:23<2:28:02,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2730 Loss (unaveraged): -0.1200\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 27%|██▋       | 2741/10000 [41:32<2:36:01,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2740 Loss (unaveraged): -0.1255\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2751/10000 [41:41<2:43:20,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2750 Loss (unaveraged): -0.1353\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2761/10000 [41:50<2:29:14,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2760 Loss (unaveraged): -0.1382\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2771/10000 [41:59<2:29:28,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2770 Loss (unaveraged): -0.1427\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2781/10000 [42:08<2:35:29,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2780 Loss (unaveraged): -0.1437\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2791/10000 [42:17<2:42:07,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2790 Loss (unaveraged): -0.1472\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2801/10000 [42:26<2:25:22,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2800 Loss (unaveraged): -0.1523\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2811/10000 [42:35<2:25:17,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2810 Loss (unaveraged): -0.1545\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2821/10000 [42:44<2:39:36,  1.33s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2820 Loss (unaveraged): -0.1513\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2831/10000 [42:53<2:36:41,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2830 Loss (unaveraged): -0.1571\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 28%|██▊       | 2841/10000 [43:02<2:24:22,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2840 Loss (unaveraged): -0.1572\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▊       | 2851/10000 [43:11<2:25:44,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2850 Loss (unaveraged): -0.1564\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▊       | 2861/10000 [43:20<2:44:05,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2860 Loss (unaveraged): -0.1594\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▊       | 2871/10000 [43:29<2:35:19,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2870 Loss (unaveraged): -0.1570\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2881/10000 [43:38<2:25:38,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2880 Loss (unaveraged): -0.1621\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2891/10000 [43:47<2:28:40,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2890 Loss (unaveraged): -0.1634\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2901/10000 [43:57<2:39:17,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2900 Loss (unaveraged): -0.1669\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2911/10000 [44:06<2:34:10,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2910 Loss (unaveraged): -0.1703\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2921/10000 [44:14<2:22:25,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2920 Loss (unaveraged): -0.1651\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2931/10000 [44:24<2:30:57,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2930 Loss (unaveraged): -0.1717\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 29%|██▉       | 2941/10000 [44:33<2:36:14,  1.33s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2940 Loss (unaveraged): -0.1693\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|██▉       | 2951/10000 [44:42<2:29:03,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2950 Loss (unaveraged): -0.1783\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|██▉       | 2961/10000 [44:51<2:22:00,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2960 Loss (unaveraged): -0.1769\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|██▉       | 2971/10000 [45:00<2:26:29,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2970 Loss (unaveraged): -0.1828\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|██▉       | 2981/10000 [45:09<2:37:06,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2980 Loss (unaveraged): -0.1847\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|██▉       | 2991/10000 [45:18<2:24:41,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 2990 Loss (unaveraged): -0.1856\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|███       | 3001/10000 [45:27<2:24:27,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3000 Loss (unaveraged): -0.1890\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|███       | 3011/10000 [45:36<2:24:54,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3010 Loss (unaveraged): -0.1944\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|███       | 3021/10000 [45:45<2:39:00,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3020 Loss (unaveraged): -0.1929\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|███       | 3031/10000 [45:54<2:27:17,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3030 Loss (unaveraged): -0.1962\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 30%|███       | 3041/10000 [46:03<2:22:49,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3040 Loss (unaveraged): -0.1969\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3051/10000 [46:12<2:24:00,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3050 Loss (unaveraged): -0.1954\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3061/10000 [46:21<2:38:05,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3060 Loss (unaveraged): -0.1993\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3071/10000 [46:30<2:21:43,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3070 Loss (unaveraged): -0.1974\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3081/10000 [46:39<2:19:49,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3080 Loss (unaveraged): -0.1870\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3091/10000 [46:48<2:24:12,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3090 Loss (unaveraged): -0.1886\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3101/10000 [46:57<2:42:24,  1.41s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3100 Loss (unaveraged): -0.1813\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3111/10000 [47:06<2:19:18,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3110 Loss (unaveraged): -0.1780\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███       | 3121/10000 [47:15<2:19:49,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3120 Loss (unaveraged): -0.1699\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███▏      | 3131/10000 [47:24<2:23:41,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3130 Loss (unaveraged): -0.1715\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 31%|███▏      | 3141/10000 [47:33<2:38:46,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3140 Loss (unaveraged): -0.1649\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3151/10000 [47:42<2:17:07,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3150 Loss (unaveraged): -0.1549\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3161/10000 [47:51<2:20:37,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3160 Loss (unaveraged): -0.1543\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3171/10000 [48:00<2:24:13,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3170 Loss (unaveraged): -0.1451\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3181/10000 [48:10<2:37:50,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3180 Loss (unaveraged): -0.1339\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3191/10000 [48:18<2:16:59,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3190 Loss (unaveraged): -0.1361\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3201/10000 [48:27<2:17:58,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3200 Loss (unaveraged): -0.1260\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3211/10000 [48:37<2:31:29,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3210 Loss (unaveraged): -0.1223\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3221/10000 [48:46<2:40:09,  1.42s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3220 Loss (unaveraged): -0.1176\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3231/10000 [48:55<2:17:27,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3230 Loss (unaveraged): -0.1155\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 32%|███▏      | 3241/10000 [49:04<2:16:14,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3240 Loss (unaveraged): -0.1131\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3251/10000 [49:13<2:22:11,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3250 Loss (unaveraged): -0.1085\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3261/10000 [49:22<2:32:03,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3260 Loss (unaveraged): -0.1071\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3271/10000 [49:31<2:16:21,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3270 Loss (unaveraged): -0.1068\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3281/10000 [49:40<2:19:43,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3280 Loss (unaveraged): -0.1043\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3291/10000 [49:49<2:21:24,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3290 Loss (unaveraged): -0.1055\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3301/10000 [49:58<2:34:44,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3300 Loss (unaveraged): -0.1017\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3311/10000 [50:07<2:14:25,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3310 Loss (unaveraged): -0.0973\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3321/10000 [50:16<2:16:43,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3320 Loss (unaveraged): -0.0886\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3331/10000 [50:25<2:25:16,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3330 Loss (unaveraged): -0.0939\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 33%|███▎      | 3341/10000 [50:34<2:27:00,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3340 Loss (unaveraged): -0.0893\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▎      | 3351/10000 [50:43<2:16:57,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3350 Loss (unaveraged): -0.0909\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▎      | 3361/10000 [50:52<2:16:21,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3360 Loss (unaveraged): -0.0869\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▎      | 3371/10000 [51:01<2:23:01,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3370 Loss (unaveraged): -0.0825\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3381/10000 [51:10<2:28:50,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3380 Loss (unaveraged): -0.0744\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3391/10000 [51:19<2:11:11,  1.19s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3390 Loss (unaveraged): -0.0707\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3401/10000 [51:28<2:15:57,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3400 Loss (unaveraged): -0.0687\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3411/10000 [51:37<2:24:18,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3410 Loss (unaveraged): -0.0705\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3421/10000 [51:46<2:17:25,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3420 Loss (unaveraged): -0.0631\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3431/10000 [51:55<2:12:18,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3430 Loss (unaveraged): -0.0661\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 34%|███▍      | 3441/10000 [52:04<2:13:53,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3440 Loss (unaveraged): -0.0658\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▍      | 3451/10000 [52:13<2:32:25,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3450 Loss (unaveraged): -0.0617\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▍      | 3461/10000 [52:22<2:16:24,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3460 Loss (unaveraged): -0.0668\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▍      | 3471/10000 [52:31<2:12:48,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3470 Loss (unaveraged): -0.0702\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▍      | 3481/10000 [52:40<2:15:53,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3480 Loss (unaveraged): -0.0740\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▍      | 3491/10000 [52:49<2:27:57,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3490 Loss (unaveraged): -0.0860\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▌      | 3501/10000 [52:58<2:10:28,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3500 Loss (unaveraged): -0.0838\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▌      | 3511/10000 [53:07<2:10:22,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3510 Loss (unaveraged): -0.0915\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▌      | 3521/10000 [53:16<2:17:50,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3520 Loss (unaveraged): -0.1046\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▌      | 3531/10000 [53:25<2:27:09,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3530 Loss (unaveraged): -0.1094\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 35%|███▌      | 3541/10000 [53:34<2:11:39,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3540 Loss (unaveraged): -0.1250\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3551/10000 [53:43<2:13:32,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3550 Loss (unaveraged): -0.1405\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3561/10000 [53:52<2:14:15,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3560 Loss (unaveraged): -0.1495\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3571/10000 [54:02<2:29:01,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3570 Loss (unaveraged): -0.1754\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3581/10000 [54:10<2:10:22,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3580 Loss (unaveraged): -0.1890\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3591/10000 [54:19<2:12:23,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3590 Loss (unaveraged): -0.2129\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3601/10000 [54:29<2:15:07,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3600 Loss (unaveraged): -0.2178\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3611/10000 [54:38<2:25:58,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3610 Loss (unaveraged): -0.2410\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▌      | 3621/10000 [54:46<2:07:33,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3620 Loss (unaveraged): -0.2760\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▋      | 3631/10000 [54:55<2:10:26,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3630 Loss (unaveraged): -0.2966\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 36%|███▋      | 3641/10000 [55:05<2:15:32,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3640 Loss (unaveraged): -0.3224\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3651/10000 [55:14<2:26:31,  1.38s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3650 Loss (unaveraged): -0.3639\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3661/10000 [55:23<2:08:45,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3660 Loss (unaveraged): -0.3900\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3671/10000 [55:32<2:09:11,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3670 Loss (unaveraged): -0.4306\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3681/10000 [55:41<2:12:29,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3680 Loss (unaveraged): -0.4668\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3691/10000 [55:50<2:26:38,  1.39s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3690 Loss (unaveraged): -0.4971\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3701/10000 [55:58<2:06:16,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3700 Loss (unaveraged): -0.5186\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3711/10000 [56:08<2:06:51,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3710 Loss (unaveraged): -0.5071\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3721/10000 [56:17<2:15:38,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3720 Loss (unaveraged): -0.5673\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3731/10000 [56:26<2:19:35,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3730 Loss (unaveraged): -0.5840\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 37%|███▋      | 3741/10000 [56:34<2:06:02,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3740 Loss (unaveraged): -0.5798\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3751/10000 [56:44<2:08:10,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3750 Loss (unaveraged): -0.6069\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3761/10000 [56:53<2:18:23,  1.33s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3760 Loss (unaveraged): -0.6127\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3771/10000 [57:02<2:17:10,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3770 Loss (unaveraged): -0.6922\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3781/10000 [57:10<2:04:35,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3780 Loss (unaveraged): -0.7218\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3791/10000 [57:20<2:07:06,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3790 Loss (unaveraged): -0.7630\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3801/10000 [57:29<2:18:38,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3800 Loss (unaveraged): -0.9385\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3811/10000 [57:38<2:12:36,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3810 Loss (unaveraged): -1.1316\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3821/10000 [57:47<2:04:24,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3820 Loss (unaveraged): -1.3679\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3831/10000 [57:56<2:09:29,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3830 Loss (unaveraged): -1.4579\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 38%|███▊      | 3841/10000 [58:05<2:19:07,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3840 Loss (unaveraged): -1.4666\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▊      | 3851/10000 [58:13<2:06:33,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3850 Loss (unaveraged): -1.3458\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▊      | 3861/10000 [58:23<2:05:19,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3860 Loss (unaveraged): -1.7248\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▊      | 3871/10000 [58:32<2:08:36,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3870 Loss (unaveraged): -1.9185\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3881/10000 [58:41<2:17:49,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3880 Loss (unaveraged): -1.8460\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3891/10000 [58:50<2:06:11,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3890 Loss (unaveraged): -0.9661\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3901/10000 [58:59<2:07:19,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3900 Loss (unaveraged): -1.8161\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3911/10000 [59:08<2:07:43,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3910 Loss (unaveraged): -1.0877\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3921/10000 [59:17<2:18:32,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3920 Loss (unaveraged): -1.2288\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3931/10000 [59:26<2:01:25,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3930 Loss (unaveraged): -0.7535\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 39%|███▉      | 3941/10000 [59:35<2:01:10,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3940 Loss (unaveraged): -0.5053\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|███▉      | 3951/10000 [59:44<2:05:23,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3950 Loss (unaveraged): -0.2595\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|███▉      | 3961/10000 [59:53<2:17:21,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3960 Loss (unaveraged): -0.3963\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|███▉      | 3971/10000 [1:00:02<2:04:19,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3970 Loss (unaveraged): -0.2585\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|███▉      | 3981/10000 [1:00:11<2:01:56,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3980 Loss (unaveraged): 0.0768\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|███▉      | 3991/10000 [1:00:20<2:07:48,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 3990 Loss (unaveraged): -0.2140\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|████      | 4001/10000 [1:00:29<2:19:35,  1.40s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4000 Loss (unaveraged): -0.1002\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|████      | 4011/10000 [1:00:38<1:59:30,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4010 Loss (unaveraged): 0.0305\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|████      | 4021/10000 [1:00:47<1:59:28,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4020 Loss (unaveraged): 0.2046\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|████      | 4031/10000 [1:00:56<2:05:05,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4030 Loss (unaveraged): 0.3479\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 40%|████      | 4041/10000 [1:01:05<2:15:38,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4040 Loss (unaveraged): 0.3564\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4051/10000 [1:01:13<1:58:00,  1.19s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4050 Loss (unaveraged): 0.4871\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4061/10000 [1:01:23<2:01:30,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4060 Loss (unaveraged): 0.4723\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4071/10000 [1:01:32<2:08:30,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4070 Loss (unaveraged): 0.4683\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4081/10000 [1:01:41<2:13:24,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4080 Loss (unaveraged): 0.4608\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4091/10000 [1:01:50<2:00:00,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4090 Loss (unaveraged): 0.4763\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4101/10000 [1:01:59<1:59:10,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4100 Loss (unaveraged): 0.3617\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4111/10000 [1:02:08<2:08:53,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4110 Loss (unaveraged): 0.2564\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████      | 4121/10000 [1:02:17<2:12:02,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4120 Loss (unaveraged): 0.1637\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████▏     | 4131/10000 [1:02:26<1:57:51,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4130 Loss (unaveraged): 0.0545\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 41%|████▏     | 4141/10000 [1:02:35<2:02:58,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4140 Loss (unaveraged): -0.0643\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4151/10000 [1:02:44<2:08:19,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4150 Loss (unaveraged): -0.1646\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4161/10000 [1:02:53<2:04:36,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4160 Loss (unaveraged): -0.2852\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4171/10000 [1:03:02<1:58:36,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4170 Loss (unaveraged): -0.3631\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4181/10000 [1:03:11<1:57:04,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4180 Loss (unaveraged): -0.4585\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4191/10000 [1:03:20<2:02:19,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4190 Loss (unaveraged): -0.4890\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4201/10000 [1:03:29<2:08:23,  1.33s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4200 Loss (unaveraged): -0.5060\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4211/10000 [1:03:37<1:55:41,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4210 Loss (unaveraged): -0.4914\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4221/10000 [1:03:46<1:57:53,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4220 Loss (unaveraged): -0.4493\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4231/10000 [1:03:55<2:00:16,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4230 Loss (unaveraged): -0.3645\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 42%|████▏     | 4241/10000 [1:04:04<2:10:25,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4240 Loss (unaveraged): -0.2730\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4251/10000 [1:04:13<1:55:02,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4250 Loss (unaveraged): -0.1554\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4261/10000 [1:04:22<1:56:56,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4260 Loss (unaveraged): -0.0450\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4271/10000 [1:04:31<1:56:03,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4270 Loss (unaveraged): 0.0554\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4281/10000 [1:04:40<2:04:27,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4280 Loss (unaveraged): 0.1440\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4291/10000 [1:04:49<2:00:15,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4290 Loss (unaveraged): 0.1904\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4301/10000 [1:04:58<1:56:49,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4300 Loss (unaveraged): 0.2049\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4311/10000 [1:05:07<1:56:53,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4310 Loss (unaveraged): 0.1794\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4321/10000 [1:05:15<1:58:12,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4320 Loss (unaveraged): 0.1299\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4331/10000 [1:05:24<2:03:42,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4330 Loss (unaveraged): 0.0520\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 43%|████▎     | 4341/10000 [1:05:33<1:53:02,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4340 Loss (unaveraged): -0.0385\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▎     | 4351/10000 [1:05:42<1:53:14,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4350 Loss (unaveraged): -0.1424\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▎     | 4361/10000 [1:05:51<1:54:28,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4360 Loss (unaveraged): -0.2580\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▎     | 4371/10000 [1:06:00<2:03:41,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4370 Loss (unaveraged): -0.3535\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4381/10000 [1:06:08<1:54:24,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4380 Loss (unaveraged): -0.4620\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4391/10000 [1:06:17<1:52:07,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4390 Loss (unaveraged): -0.5433\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4401/10000 [1:06:26<1:56:16,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4400 Loss (unaveraged): -0.8460\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4411/10000 [1:06:35<1:59:04,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4410 Loss (unaveraged): -1.0269\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4421/10000 [1:06:44<1:57:02,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4420 Loss (unaveraged): -0.4475\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4431/10000 [1:06:52<1:50:26,  1.19s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4430 Loss (unaveraged): 0.2200\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 44%|████▍     | 4441/10000 [1:07:01<1:53:12,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4440 Loss (unaveraged): 0.3741\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▍     | 4451/10000 [1:07:10<1:59:28,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4450 Loss (unaveraged): 0.1980\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▍     | 4461/10000 [1:07:19<2:01:23,  1.32s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4460 Loss (unaveraged): 0.0831\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▍     | 4471/10000 [1:07:28<1:50:34,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4470 Loss (unaveraged): 0.0463\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▍     | 4481/10000 [1:07:36<1:49:38,  1.19s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4480 Loss (unaveraged): 0.0329\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▍     | 4491/10000 [1:07:45<1:53:51,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4490 Loss (unaveraged): 0.0461\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▌     | 4501/10000 [1:07:54<2:01:51,  1.33s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4500 Loss (unaveraged): 0.0848\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▌     | 4511/10000 [1:08:03<1:49:32,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4510 Loss (unaveraged): 0.1423\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▌     | 4521/10000 [1:08:12<1:54:44,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4520 Loss (unaveraged): 0.1791\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▌     | 4531/10000 [1:08:21<1:52:15,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4530 Loss (unaveraged): 0.2640\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 45%|████▌     | 4541/10000 [1:08:30<1:57:44,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4540 Loss (unaveraged): 0.3014\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4551/10000 [1:08:39<1:56:00,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4550 Loss (unaveraged): 0.3422\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4561/10000 [1:08:48<1:49:06,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4560 Loss (unaveraged): 0.3470\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4571/10000 [1:08:57<1:49:12,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4570 Loss (unaveraged): 0.3368\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4581/10000 [1:09:06<1:51:55,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4580 Loss (unaveraged): 0.3426\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4591/10000 [1:09:15<2:03:20,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4590 Loss (unaveraged): 0.3021\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4601/10000 [1:09:23<1:50:33,  1.23s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4600 Loss (unaveraged): 0.2790\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4611/10000 [1:09:32<1:49:58,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4610 Loss (unaveraged): 0.2254\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▌     | 4621/10000 [1:09:41<1:49:27,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4620 Loss (unaveraged): 0.1754\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▋     | 4631/10000 [1:09:50<1:57:13,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4630 Loss (unaveraged): 0.1449\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 46%|████▋     | 4641/10000 [1:09:59<1:51:48,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4640 Loss (unaveraged): 0.1016\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4651/10000 [1:10:08<1:46:37,  1.20s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4650 Loss (unaveraged): 0.0689\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4661/10000 [1:10:17<1:51:53,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4660 Loss (unaveraged): 0.0371\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4671/10000 [1:10:26<1:53:00,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4670 Loss (unaveraged): 0.0200\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4681/10000 [1:10:35<2:01:45,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4680 Loss (unaveraged): 0.0082\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4691/10000 [1:10:44<1:50:08,  1.24s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4690 Loss (unaveraged): 0.0085\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4701/10000 [1:10:53<1:46:50,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4700 Loss (unaveraged): -0.0058\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4711/10000 [1:11:02<1:51:12,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4710 Loss (unaveraged): 0.0054\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4721/10000 [1:11:11<1:55:11,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4720 Loss (unaveraged): 0.0010\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4731/10000 [1:11:20<1:52:21,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4730 Loss (unaveraged): 0.0137\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 47%|████▋     | 4741/10000 [1:11:29<1:46:09,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4740 Loss (unaveraged): 0.0231\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4751/10000 [1:11:38<1:46:38,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4750 Loss (unaveraged): 0.0475\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4761/10000 [1:11:47<1:52:50,  1.29s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4760 Loss (unaveraged): 0.0621\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4771/10000 [1:11:56<1:53:42,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4770 Loss (unaveraged): 0.0889\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4781/10000 [1:12:04<1:45:30,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4780 Loss (unaveraged): 0.1059\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4791/10000 [1:12:13<1:44:46,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4790 Loss (unaveraged): 0.1183\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4801/10000 [1:12:22<1:49:22,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4800 Loss (unaveraged): 0.1322\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4811/10000 [1:12:31<1:56:51,  1.35s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4810 Loss (unaveraged): 0.1361\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4821/10000 [1:12:40<1:47:38,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4820 Loss (unaveraged): 0.1278\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4831/10000 [1:12:49<1:50:21,  1.28s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4830 Loss (unaveraged): 0.1260\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 48%|████▊     | 4841/10000 [1:12:59<1:47:24,  1.25s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4840 Loss (unaveraged): 0.0960\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▊     | 4851/10000 [1:13:08<1:49:03,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4850 Loss (unaveraged): 0.0746\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▊     | 4861/10000 [1:13:17<1:54:33,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4860 Loss (unaveraged): 0.0436\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▊     | 4871/10000 [1:13:26<1:47:17,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4870 Loss (unaveraged): 0.0072\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4881/10000 [1:13:35<1:47:44,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4880 Loss (unaveraged): -0.0361\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4891/10000 [1:13:45<1:47:25,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4890 Loss (unaveraged): -0.0732\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4901/10000 [1:13:54<1:51:42,  1.31s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4900 Loss (unaveraged): -0.1083\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4911/10000 [1:14:03<1:56:04,  1.37s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4910 Loss (unaveraged): -0.1414\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4921/10000 [1:14:12<1:43:20,  1.22s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4920 Loss (unaveraged): -0.1645\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4931/10000 [1:14:21<1:42:24,  1.21s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4930 Loss (unaveraged): -0.1751\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 49%|████▉     | 4941/10000 [1:14:30<1:46:20,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4940 Loss (unaveraged): -0.1802\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|████▉     | 4951/10000 [1:14:39<1:54:15,  1.36s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4950 Loss (unaveraged): -0.1882\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|████▉     | 4961/10000 [1:14:48<1:49:12,  1.30s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4960 Loss (unaveraged): -0.1665\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|████▉     | 4971/10000 [1:14:57<1:46:23,  1.27s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4970 Loss (unaveraged): -0.1351\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|████▉     | 4981/10000 [1:15:06<1:45:26,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4980 Loss (unaveraged): -0.1129\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|████▉     | 4991/10000 [1:15:16<1:45:14,  1.26s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 4990 Loss (unaveraged): -0.0745\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|█████     | 5001/10000 [1:15:25<1:51:17,  1.34s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5000 Loss (unaveraged): -0.0355\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|█████     | 5011/10000 [1:15:36<2:34:19,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5010 Loss (unaveraged): 0.0011 Loss (averaged): -0.0212\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|█████     | 5021/10000 [1:15:46<2:34:27,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5020 Loss (unaveraged): 0.0292 Loss (averaged): -0.0010\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|█████     | 5031/10000 [1:15:57<2:26:23,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5030 Loss (unaveraged): 0.0745 Loss (averaged): 0.0138\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 50%|█████     | 5041/10000 [1:16:08<2:26:40,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5040 Loss (unaveraged): 0.0947 Loss (averaged): 0.0401\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5051/10000 [1:16:19<2:24:39,  1.75s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5050 Loss (unaveraged): 0.1035 Loss (averaged): 0.0604\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5061/10000 [1:16:30<2:28:04,  1.80s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5060 Loss (unaveraged): 0.1056 Loss (averaged): 0.0619\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5071/10000 [1:16:40<2:23:44,  1.75s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5070 Loss (unaveraged): 0.0996 Loss (averaged): 0.0801\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5081/10000 [1:16:51<2:22:36,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5080 Loss (unaveraged): 0.0820 Loss (averaged): 0.0880\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5091/10000 [1:17:02<2:25:53,  1.78s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5090 Loss (unaveraged): 0.0586 Loss (averaged): 0.0831\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5101/10000 [1:17:12<2:23:03,  1.75s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5100 Loss (unaveraged): 0.0316 Loss (averaged): 0.0808\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5111/10000 [1:17:23<2:24:21,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5110 Loss (unaveraged): 0.0036 Loss (averaged): 0.0858\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████     | 5121/10000 [1:17:34<2:29:26,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5120 Loss (unaveraged): -0.0372 Loss (averaged): 0.0787\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████▏    | 5131/10000 [1:17:45<2:30:42,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5130 Loss (unaveraged): -0.0670 Loss (averaged): 0.0788\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 51%|█████▏    | 5141/10000 [1:17:55<2:29:45,  1.85s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5140 Loss (unaveraged): -0.0993 Loss (averaged): 0.0714\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5151/10000 [1:18:06<2:29:04,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5150 Loss (unaveraged): -0.1275 Loss (averaged): 0.0637\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5161/10000 [1:18:17<2:26:25,  1.82s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5160 Loss (unaveraged): -0.1549 Loss (averaged): 0.0513\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5171/10000 [1:18:27<2:26:05,  1.82s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5170 Loss (unaveraged): -0.1674 Loss (averaged): 0.0485\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5181/10000 [1:18:37<2:17:40,  1.71s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5180 Loss (unaveraged): -0.1742 Loss (averaged): 0.0317\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5191/10000 [1:18:48<2:17:08,  1.71s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5190 Loss (unaveraged): -0.1701 Loss (averaged): 0.0198\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5201/10000 [1:18:59<2:18:58,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5200 Loss (unaveraged): -0.1658 Loss (averaged): 0.0055\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5211/10000 [1:19:09<2:18:13,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5210 Loss (unaveraged): -0.1428 Loss (averaged): -0.0048\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5221/10000 [1:19:20<2:17:38,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5220 Loss (unaveraged): -0.1185 Loss (averaged): -0.0086\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5231/10000 [1:19:31<2:18:20,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5230 Loss (unaveraged): -0.0848 Loss (averaged): -0.0241\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 52%|█████▏    | 5241/10000 [1:19:41<2:20:27,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5240 Loss (unaveraged): -0.0470 Loss (averaged): -0.0381\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5251/10000 [1:19:52<2:25:31,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5250 Loss (unaveraged): -0.0065 Loss (averaged): -0.0439\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5261/10000 [1:20:03<2:26:39,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5260 Loss (unaveraged): 0.0214 Loss (averaged): -0.0462\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5271/10000 [1:20:13<2:23:41,  1.82s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5270 Loss (unaveraged): 0.0504 Loss (averaged): -0.0518\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5281/10000 [1:20:24<2:25:13,  1.85s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5280 Loss (unaveraged): 0.0858 Loss (averaged): -0.0511\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5291/10000 [1:20:35<2:25:35,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5290 Loss (unaveraged): 0.1160 Loss (averaged): -0.0535\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5301/10000 [1:20:45<2:19:19,  1.78s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5300 Loss (unaveraged): 0.1440 Loss (averaged): -0.0648\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5311/10000 [1:20:55<2:13:05,  1.70s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5310 Loss (unaveraged): 0.1509 Loss (averaged): -0.0557\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5321/10000 [1:21:06<2:13:35,  1.71s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5320 Loss (unaveraged): 0.1540 Loss (averaged): -0.0528\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5331/10000 [1:21:17<2:15:13,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5330 Loss (unaveraged): 0.1604 Loss (averaged): -0.0469\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 53%|█████▎    | 5341/10000 [1:21:27<2:13:13,  1.72s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5340 Loss (unaveraged): 0.1437 Loss (averaged): -0.0542\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▎    | 5351/10000 [1:21:38<2:14:21,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5350 Loss (unaveraged): 0.1267 Loss (averaged): -0.0459\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▎    | 5361/10000 [1:21:49<2:18:46,  1.79s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5360 Loss (unaveraged): 0.0945 Loss (averaged): -0.0392\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▎    | 5371/10000 [1:21:59<2:17:17,  1.78s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5370 Loss (unaveraged): 0.0705 Loss (averaged): -0.0351\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5381/10000 [1:22:10<2:18:38,  1.80s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5380 Loss (unaveraged): 0.0374 Loss (averaged): -0.0333\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5391/10000 [1:22:21<2:20:31,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5390 Loss (unaveraged): 0.0043 Loss (averaged): -0.0323\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5401/10000 [1:22:31<2:18:51,  1.81s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5400 Loss (unaveraged): -0.0170 Loss (averaged): -0.0305\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5411/10000 [1:22:42<2:19:36,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5410 Loss (unaveraged): -0.0557 Loss (averaged): -0.0250\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5421/10000 [1:22:52<2:18:26,  1.81s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5420 Loss (unaveraged): -0.0722 Loss (averaged): -0.0266\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5431/10000 [1:23:03<2:14:27,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5430 Loss (unaveraged): -0.0798 Loss (averaged): -0.0137\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 54%|█████▍    | 5441/10000 [1:23:13<2:13:55,  1.76s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5440 Loss (unaveraged): -0.0847 Loss (averaged): -0.0114\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▍    | 5451/10000 [1:23:24<2:08:55,  1.70s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5450 Loss (unaveraged): -0.0830 Loss (averaged): -0.0073\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▍    | 5461/10000 [1:23:34<2:09:56,  1.72s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5460 Loss (unaveraged): -0.0693 Loss (averaged): -0.0120\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▍    | 5471/10000 [1:23:45<2:13:44,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5470 Loss (unaveraged): -0.0477 Loss (averaged): 0.0005\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▍    | 5481/10000 [1:23:56<2:10:31,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5480 Loss (unaveraged): -0.0320 Loss (averaged): 0.0003\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▍    | 5491/10000 [1:24:06<2:10:37,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5490 Loss (unaveraged): 0.0069 Loss (averaged): 0.0005\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▌    | 5501/10000 [1:24:17<2:12:43,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5500 Loss (unaveraged): 0.0292 Loss (averaged): -0.0065\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▌    | 5511/10000 [1:24:28<2:16:51,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5510 Loss (unaveraged): 0.0572 Loss (averaged): 0.0043\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▌    | 5521/10000 [1:24:38<2:19:26,  1.87s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5520 Loss (unaveraged): 0.0915 Loss (averaged): 0.0053\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▌    | 5531/10000 [1:24:49<2:17:29,  1.85s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5530 Loss (unaveraged): 0.1150 Loss (averaged): 0.0037\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 55%|█████▌    | 5541/10000 [1:25:00<2:16:32,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5540 Loss (unaveraged): 0.1278 Loss (averaged): 0.0140\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5551/10000 [1:25:10<2:18:08,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5550 Loss (unaveraged): 0.1384 Loss (averaged): 0.0053\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5561/10000 [1:25:21<2:10:02,  1.76s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5560 Loss (unaveraged): 0.1372 Loss (averaged): 0.0063\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5571/10000 [1:25:31<2:06:44,  1.72s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5570 Loss (unaveraged): 0.1272 Loss (averaged): 0.0167\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5581/10000 [1:25:42<2:09:38,  1.76s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5580 Loss (unaveraged): 0.1013 Loss (averaged): 0.0117\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5591/10000 [1:25:53<2:09:08,  1.76s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5590 Loss (unaveraged): 0.0706 Loss (averaged): 0.0188\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5601/10000 [1:26:03<2:06:44,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5600 Loss (unaveraged): 0.0062 Loss (averaged): 0.0138\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5611/10000 [1:26:14<2:07:21,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5610 Loss (unaveraged): -0.0832 Loss (averaged): 0.0217\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▌    | 5621/10000 [1:26:25<2:07:21,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5620 Loss (unaveraged): -0.3229 Loss (averaged): 0.0174\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▋    | 5631/10000 [1:26:36<2:10:49,  1.80s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5630 Loss (unaveraged): -1.3524 Loss (averaged): 0.0166\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 56%|█████▋    | 5641/10000 [1:26:46<2:10:59,  1.80s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5640 Loss (unaveraged): -1.0923 Loss (averaged): 0.0187\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5651/10000 [1:26:57<2:13:07,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5650 Loss (unaveraged): -0.4093 Loss (averaged): 0.0255\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5661/10000 [1:27:07<2:14:31,  1.86s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5660 Loss (unaveraged): 0.5567 Loss (averaged): 0.0201\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5671/10000 [1:27:18<2:13:02,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5670 Loss (unaveraged): 0.2437 Loss (averaged): 0.0250\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5681/10000 [1:27:29<2:11:57,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5680 Loss (unaveraged): 0.1491 Loss (averaged): 0.0183\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5691/10000 [1:27:39<2:07:10,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5690 Loss (unaveraged): 0.0984 Loss (averaged): 0.0191\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5701/10000 [1:27:49<2:02:00,  1.70s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5700 Loss (unaveraged): 0.0848 Loss (averaged): 0.0168\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5711/10000 [1:28:00<2:00:39,  1.69s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5710 Loss (unaveraged): 0.0502 Loss (averaged): 0.0143\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5721/10000 [1:28:11<2:01:59,  1.71s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5720 Loss (unaveraged): 0.0345 Loss (averaged): 0.0175\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5731/10000 [1:28:21<2:03:41,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5730 Loss (unaveraged): 0.0206 Loss (averaged): 0.0191\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 57%|█████▋    | 5741/10000 [1:28:32<2:05:56,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5740 Loss (unaveraged): 0.0026 Loss (averaged): 0.0125\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5751/10000 [1:28:43<2:04:59,  1.77s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5750 Loss (unaveraged): -0.0021 Loss (averaged): 0.0206\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5761/10000 [1:28:53<2:05:31,  1.78s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5760 Loss (unaveraged): -0.0098 Loss (averaged): 0.0164\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5771/10000 [1:29:04<2:10:13,  1.85s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5770 Loss (unaveraged): -0.0167 Loss (averaged): 0.0130\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5781/10000 [1:29:14<2:08:44,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5780 Loss (unaveraged): -0.0195 Loss (averaged): 0.0144\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5791/10000 [1:29:25<2:07:41,  1.82s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5790 Loss (unaveraged): -0.0182 Loss (averaged): 0.0150\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5801/10000 [1:29:36<2:09:00,  1.84s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5800 Loss (unaveraged): -0.0206 Loss (averaged): 0.0102\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5811/10000 [1:29:46<2:07:47,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5810 Loss (unaveraged): -0.0165 Loss (averaged): 0.0117\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5821/10000 [1:29:57<2:01:29,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5820 Loss (unaveraged): -0.0167 Loss (averaged): 0.0116\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5831/10000 [1:30:07<2:00:18,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5830 Loss (unaveraged): -0.0017 Loss (averaged): 0.0131\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 58%|█████▊    | 5841/10000 [1:30:18<1:59:14,  1.72s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5840 Loss (unaveraged): 0.0002 Loss (averaged): 0.0132\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▊    | 5851/10000 [1:30:29<2:01:59,  1.76s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5850 Loss (unaveraged): 0.0135 Loss (averaged): 0.0103\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▊    | 5861/10000 [1:30:39<1:59:47,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5860 Loss (unaveraged): 0.0268 Loss (averaged): 0.0108\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▊    | 5871/10000 [1:30:50<2:00:17,  1.75s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5870 Loss (unaveraged): 0.0366 Loss (averaged): 0.0036\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5881/10000 [1:31:01<2:03:29,  1.80s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5880 Loss (unaveraged): 0.0395 Loss (averaged): 0.0100\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5891/10000 [1:31:12<2:03:14,  1.80s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5890 Loss (unaveraged): 0.0514 Loss (averaged): 0.0126\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5901/10000 [1:31:23<2:04:20,  1.82s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5900 Loss (unaveraged): 0.0647 Loss (averaged): 0.0137\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5911/10000 [1:31:33<2:06:17,  1.85s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5910 Loss (unaveraged): 0.0807 Loss (averaged): 0.0104\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5921/10000 [1:31:44<2:04:35,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5920 Loss (unaveraged): 0.0834 Loss (averaged): 0.0102\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5931/10000 [1:31:54<2:03:54,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5930 Loss (unaveraged): 0.0840 Loss (averaged): 0.0112\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 59%|█████▉    | 5941/10000 [1:32:05<2:03:56,  1.83s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5940 Loss (unaveraged): 0.0928 Loss (averaged): 0.0131\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|█████▉    | 5951/10000 [1:32:15<1:57:41,  1.74s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5950 Loss (unaveraged): 0.0904 Loss (averaged): 0.0135\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|█████▉    | 5961/10000 [1:32:26<1:57:41,  1.75s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5960 Loss (unaveraged): 0.0914 Loss (averaged): 0.0087\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|█████▉    | 5971/10000 [1:32:36<1:55:42,  1.72s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5970 Loss (unaveraged): 0.0816 Loss (averaged): 0.0105\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|█████▉    | 5981/10000 [1:32:47<1:55:01,  1.72s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5980 Loss (unaveraged): 0.0789 Loss (averaged): 0.0128\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|█████▉    | 5991/10000 [1:32:58<1:57:15,  1.76s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 5990 Loss (unaveraged): 0.0712 Loss (averaged): 0.0103\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|██████    | 6001/10000 [1:33:08<1:54:07,  1.71s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 6000 Loss (unaveraged): 0.0591 Loss (averaged): 0.0137\n"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stderr",
          "output_type": "stream",
          "text": [
            " 60%|██████    | 6011/10000 [1:33:19<1:55:09,  1.73s/it]"
          ]
        },
        {
          "metadata": {
            "tags": null
          },
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Step: 6010 Loss (unaveraged): 0.0479 Loss (averaged): 0.0148\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 60%|██████    | 6021/10000 [1:33:30<1:58:12,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6020 Loss (unaveraged): 0.0358 Loss (averaged): 0.0121\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 60%|██████    | 6031/10000 [1:33:40<1:59:05,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6030 Loss (unaveraged): 0.0214 Loss (averaged): 0.0164\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 60%|██████    | 6041/10000 [1:33:51<1:59:50,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6040 Loss (unaveraged): 0.0137 Loss (averaged): 0.0076\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6051/10000 [1:34:01<2:01:39,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6050 Loss (unaveraged): -0.0017 Loss (averaged): 0.0165\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6061/10000 [1:34:12<1:59:43,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6060 Loss (unaveraged): -0.0151 Loss (averaged): 0.0122\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6071/10000 [1:34:22<1:58:43,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6070 Loss (unaveraged): -0.0156 Loss (averaged): 0.0145\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6081/10000 [1:34:33<1:52:33,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6080 Loss (unaveraged): -0.0254 Loss (averaged): 0.0189\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6091/10000 [1:34:43<1:51:38,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6090 Loss (unaveraged): -0.0299 Loss (averaged): 0.0154\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6101/10000 [1:34:54<1:53:06,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6100 Loss (unaveraged): -0.0333 Loss (averaged): 0.0102\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6111/10000 [1:35:04<1:50:45,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6110 Loss (unaveraged): -0.0303 Loss (averaged): 0.0152\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████    | 6121/10000 [1:35:15<1:52:45,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6120 Loss (unaveraged): -0.0267 Loss (averaged): 0.0132\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████▏   | 6131/10000 [1:35:26<1:54:16,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6130 Loss (unaveraged): -0.0144 Loss (averaged): 0.0183\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 61%|██████▏   | 6141/10000 [1:35:36<1:53:26,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6140 Loss (unaveraged): -0.0106 Loss (averaged): 0.0154\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6151/10000 [1:35:47<1:56:28,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6150 Loss (unaveraged): 0.0031 Loss (averaged): 0.0151\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6161/10000 [1:35:58<1:56:48,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6160 Loss (unaveraged): 0.0178 Loss (averaged): 0.0165\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6171/10000 [1:36:08<1:55:59,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6170 Loss (unaveraged): 0.0275 Loss (averaged): 0.0156\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6181/10000 [1:36:19<1:57:34,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6180 Loss (unaveraged): 0.0364 Loss (averaged): 0.0112\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6191/10000 [1:36:29<1:56:48,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6190 Loss (unaveraged): 0.0535 Loss (averaged): 0.0129\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6201/10000 [1:36:40<1:52:57,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6200 Loss (unaveraged): 0.0614 Loss (averaged): 0.0139\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6211/10000 [1:36:50<1:50:27,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6210 Loss (unaveraged): 0.0685 Loss (averaged): 0.0135\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6221/10000 [1:37:01<1:49:55,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6220 Loss (unaveraged): 0.0673 Loss (averaged): 0.0149\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6231/10000 [1:37:12<1:47:57,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6230 Loss (unaveraged): 0.0670 Loss (averaged): 0.0152\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 62%|██████▏   | 6241/10000 [1:37:23<1:48:08,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6240 Loss (unaveraged): 0.0644 Loss (averaged): 0.0106\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6251/10000 [1:37:33<1:49:44,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6250 Loss (unaveraged): 0.0611 Loss (averaged): 0.0158\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6261/10000 [1:37:44<1:49:16,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6260 Loss (unaveraged): 0.0527 Loss (averaged): 0.0118\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6271/10000 [1:37:55<1:51:17,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6270 Loss (unaveraged): 0.0447 Loss (averaged): 0.0094\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6281/10000 [1:38:05<1:52:32,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6280 Loss (unaveraged): 0.0267 Loss (averaged): 0.0101\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6291/10000 [1:38:16<2:00:03,  1.94s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6290 Loss (unaveraged): 0.0177 Loss (averaged): 0.0157\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6301/10000 [1:38:27<1:53:45,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6300 Loss (unaveraged): 0.0086 Loss (averaged): 0.0158\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6311/10000 [1:38:38<1:53:20,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6310 Loss (unaveraged): 0.0002 Loss (averaged): 0.0136\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6321/10000 [1:38:49<1:56:03,  1.89s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6320 Loss (unaveraged): -0.0066 Loss (averaged): 0.0115\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6331/10000 [1:38:59<1:52:21,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6330 Loss (unaveraged): -0.0059 Loss (averaged): 0.0145\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 63%|██████▎   | 6341/10000 [1:39:10<1:48:47,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6340 Loss (unaveraged): -0.0029 Loss (averaged): 0.0111\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▎   | 6351/10000 [1:39:20<1:46:39,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6350 Loss (unaveraged): 0.0030 Loss (averaged): 0.0133\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▎   | 6361/10000 [1:39:31<1:44:22,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6360 Loss (unaveraged): 0.0067 Loss (averaged): 0.0155\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▎   | 6371/10000 [1:39:42<1:43:52,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6370 Loss (unaveraged): 0.0226 Loss (averaged): 0.0171\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6381/10000 [1:39:52<1:44:44,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6380 Loss (unaveraged): 0.0283 Loss (averaged): 0.0163\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6391/10000 [1:40:03<1:44:14,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6390 Loss (unaveraged): 0.0450 Loss (averaged): 0.0122\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6401/10000 [1:40:14<1:47:21,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6400 Loss (unaveraged): 0.0551 Loss (averaged): 0.0145\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6411/10000 [1:40:25<1:46:52,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6410 Loss (unaveraged): 0.0688 Loss (averaged): 0.0162\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6421/10000 [1:40:35<1:48:11,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6420 Loss (unaveraged): 0.0705 Loss (averaged): 0.0179\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6431/10000 [1:40:46<1:50:48,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6430 Loss (unaveraged): 0.0775 Loss (averaged): 0.0166\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 64%|██████▍   | 6441/10000 [1:40:57<1:48:52,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6440 Loss (unaveraged): 0.0825 Loss (averaged): 0.0136\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▍   | 6451/10000 [1:41:07<1:49:03,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6450 Loss (unaveraged): 0.0789 Loss (averaged): 0.0187\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▍   | 6461/10000 [1:41:18<1:48:47,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6460 Loss (unaveraged): 0.0743 Loss (averaged): 0.0157\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▍   | 6471/10000 [1:41:29<1:47:15,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6470 Loss (unaveraged): 0.0643 Loss (averaged): 0.0172\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▍   | 6481/10000 [1:41:39<1:45:14,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6480 Loss (unaveraged): 0.0539 Loss (averaged): 0.0200\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▍   | 6491/10000 [1:41:50<1:39:29,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6490 Loss (unaveraged): 0.0392 Loss (averaged): 0.0185\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▌   | 6501/10000 [1:42:00<1:40:30,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6500 Loss (unaveraged): 0.0274 Loss (averaged): 0.0187\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▌   | 6511/10000 [1:42:11<1:42:54,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6510 Loss (unaveraged): 0.0182 Loss (averaged): 0.0223\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▌   | 6521/10000 [1:42:22<1:40:07,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6520 Loss (unaveraged): 0.0064 Loss (averaged): 0.0170\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▌   | 6531/10000 [1:42:32<1:41:15,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6530 Loss (unaveraged): -0.0009 Loss (averaged): 0.0207\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 65%|██████▌   | 6541/10000 [1:42:43<1:43:14,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6540 Loss (unaveraged): -0.0025 Loss (averaged): 0.0225\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6551/10000 [1:42:54<1:41:26,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6550 Loss (unaveraged): -0.0154 Loss (averaged): 0.0176\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6561/10000 [1:43:05<1:44:26,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6560 Loss (unaveraged): -0.0125 Loss (averaged): 0.0172\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6571/10000 [1:43:15<1:46:11,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6570 Loss (unaveraged): -0.0070 Loss (averaged): 0.0180\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6581/10000 [1:43:26<1:45:53,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6580 Loss (unaveraged): -0.0053 Loss (averaged): 0.0170\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6591/10000 [1:43:37<1:44:52,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6590 Loss (unaveraged): 0.0083 Loss (averaged): 0.0211\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6601/10000 [1:43:48<1:43:59,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6600 Loss (unaveraged): 0.0209 Loss (averaged): 0.0213\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6611/10000 [1:43:58<1:43:26,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6610 Loss (unaveraged): 0.0232 Loss (averaged): 0.0227\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▌   | 6621/10000 [1:44:09<1:40:41,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6620 Loss (unaveraged): 0.0318 Loss (averaged): 0.0212\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▋   | 6631/10000 [1:44:19<1:37:43,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6630 Loss (unaveraged): 0.0345 Loss (averaged): 0.0197\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 66%|██████▋   | 6641/10000 [1:44:30<1:37:00,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6640 Loss (unaveraged): 0.0433 Loss (averaged): 0.0206\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6651/10000 [1:44:41<1:38:22,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6650 Loss (unaveraged): 0.0465 Loss (averaged): 0.0215\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6661/10000 [1:44:51<1:35:52,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6660 Loss (unaveraged): 0.0433 Loss (averaged): 0.0240\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6671/10000 [1:45:02<1:36:57,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6670 Loss (unaveraged): 0.0315 Loss (averaged): 0.0219\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6681/10000 [1:45:13<1:38:46,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6680 Loss (unaveraged): 0.0225 Loss (averaged): 0.0233\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6691/10000 [1:45:24<1:36:56,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6690 Loss (unaveraged): 0.0094 Loss (averaged): 0.0213\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6701/10000 [1:45:34<1:40:40,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6700 Loss (unaveraged): -0.0059 Loss (averaged): 0.0245\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6711/10000 [1:45:45<1:41:59,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6710 Loss (unaveraged): -0.0200 Loss (averaged): 0.0230\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6721/10000 [1:45:56<1:40:23,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6720 Loss (unaveraged): -0.0321 Loss (averaged): 0.0219\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6731/10000 [1:46:07<1:42:36,  1.88s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6730 Loss (unaveraged): -0.0437 Loss (averaged): 0.0196\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 67%|██████▋   | 6741/10000 [1:46:17<1:40:59,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6740 Loss (unaveraged): -0.0503 Loss (averaged): 0.0226\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6751/10000 [1:46:28<1:40:20,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6750 Loss (unaveraged): -0.0614 Loss (averaged): 0.0233\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6761/10000 [1:46:39<1:37:52,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6760 Loss (unaveraged): -0.0603 Loss (averaged): 0.0242\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6771/10000 [1:46:49<1:32:41,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6770 Loss (unaveraged): -0.0530 Loss (averaged): 0.0243\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6781/10000 [1:47:00<1:32:25,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6780 Loss (unaveraged): -0.0472 Loss (averaged): 0.0194\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6791/10000 [1:47:11<1:33:48,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6790 Loss (unaveraged): -0.0385 Loss (averaged): 0.0232\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6801/10000 [1:47:21<1:33:13,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6800 Loss (unaveraged): -0.0247 Loss (averaged): 0.0251\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6811/10000 [1:47:32<1:33:06,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6810 Loss (unaveraged): -0.0125 Loss (averaged): 0.0223\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6821/10000 [1:47:43<1:34:25,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6820 Loss (unaveraged): -0.0092 Loss (averaged): 0.0238\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6831/10000 [1:47:54<1:32:13,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6830 Loss (unaveraged): 0.0024 Loss (averaged): 0.0224\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 68%|██████▊   | 6841/10000 [1:48:05<1:35:18,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6840 Loss (unaveraged): 0.0106 Loss (averaged): 0.0237\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▊   | 6851/10000 [1:48:15<1:36:53,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6850 Loss (unaveraged): 0.0073 Loss (averaged): 0.0206\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▊   | 6861/10000 [1:48:26<1:35:52,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6860 Loss (unaveraged): 0.0053 Loss (averaged): 0.0191\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▊   | 6871/10000 [1:48:37<1:38:23,  1.89s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6870 Loss (unaveraged): -0.0085 Loss (averaged): 0.0202\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6881/10000 [1:48:47<1:36:58,  1.87s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6880 Loss (unaveraged): -0.0266 Loss (averaged): 0.0224\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6891/10000 [1:48:58<1:35:30,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6890 Loss (unaveraged): -0.0383 Loss (averaged): 0.0249\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6901/10000 [1:49:09<1:35:40,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6900 Loss (unaveraged): -0.0556 Loss (averaged): 0.0215\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6911/10000 [1:49:19<1:30:40,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6910 Loss (unaveraged): -0.0649 Loss (averaged): 0.0255\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6921/10000 [1:49:30<1:29:39,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6920 Loss (unaveraged): -0.0701 Loss (averaged): 0.0219\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6931/10000 [1:49:41<1:28:49,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6930 Loss (unaveraged): -0.0743 Loss (averaged): 0.0250\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 69%|██████▉   | 6941/10000 [1:49:51<1:26:54,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6940 Loss (unaveraged): -0.0548 Loss (averaged): 0.0237\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|██████▉   | 6951/10000 [1:50:02<1:29:24,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6950 Loss (unaveraged): -0.0338 Loss (averaged): 0.0237\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|██████▉   | 6961/10000 [1:50:13<1:28:15,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6960 Loss (unaveraged): -0.0091 Loss (averaged): 0.0205\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|██████▉   | 6971/10000 [1:50:23<1:28:34,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6970 Loss (unaveraged): 0.0082 Loss (averaged): 0.0231\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|██████▉   | 6981/10000 [1:50:34<1:29:57,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6980 Loss (unaveraged): 0.0172 Loss (averaged): 0.0240\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|██████▉   | 6991/10000 [1:50:45<1:32:12,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 6990 Loss (unaveraged): 0.0142 Loss (averaged): 0.0207\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|███████   | 7001/10000 [1:50:55<1:32:36,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7000 Loss (unaveraged): 0.0013 Loss (averaged): 0.0222\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|███████   | 7011/10000 [1:51:06<1:31:12,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7010 Loss (unaveraged): -0.0137 Loss (averaged): 0.0204\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|███████   | 7021/10000 [1:51:17<1:31:53,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7020 Loss (unaveraged): -0.0319 Loss (averaged): 0.0216\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|███████   | 7031/10000 [1:51:27<1:32:15,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7030 Loss (unaveraged): -0.0452 Loss (averaged): 0.0230\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 70%|███████   | 7041/10000 [1:51:38<1:27:00,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7040 Loss (unaveraged): -0.0596 Loss (averaged): 0.0200\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7051/10000 [1:51:48<1:23:53,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7050 Loss (unaveraged): -0.0675 Loss (averaged): 0.0215\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7061/10000 [1:51:59<1:25:51,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7060 Loss (unaveraged): -0.0728 Loss (averaged): 0.0235\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7071/10000 [1:52:10<1:24:24,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7070 Loss (unaveraged): -0.0729 Loss (averaged): 0.0233\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7081/10000 [1:52:20<1:24:22,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7080 Loss (unaveraged): -0.0821 Loss (averaged): 0.0195\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7091/10000 [1:52:31<1:24:16,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7090 Loss (unaveraged): -0.0722 Loss (averaged): 0.0202\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7101/10000 [1:52:42<1:24:33,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7100 Loss (unaveraged): -0.0631 Loss (averaged): 0.0181\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7111/10000 [1:52:52<1:26:01,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7110 Loss (unaveraged): -0.0659 Loss (averaged): 0.0169\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████   | 7121/10000 [1:53:03<1:27:35,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7120 Loss (unaveraged): -0.0520 Loss (averaged): 0.0179\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████▏  | 7131/10000 [1:53:13<1:27:20,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7130 Loss (unaveraged): -0.0544 Loss (averaged): 0.0175\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 71%|███████▏  | 7141/10000 [1:53:24<1:28:46,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7140 Loss (unaveraged): -0.0531 Loss (averaged): 0.0176\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7151/10000 [1:53:35<1:27:19,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7150 Loss (unaveraged): -0.0529 Loss (averaged): 0.0214\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7161/10000 [1:53:45<1:27:23,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7160 Loss (unaveraged): -0.0664 Loss (averaged): 0.0216\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7171/10000 [1:53:56<1:24:29,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7170 Loss (unaveraged): -0.0781 Loss (averaged): 0.0195\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7181/10000 [1:54:06<1:20:56,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7180 Loss (unaveraged): -0.0987 Loss (averaged): 0.0221\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7191/10000 [1:54:17<1:20:35,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7190 Loss (unaveraged): -0.1219 Loss (averaged): 0.0185\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7201/10000 [1:54:28<1:21:02,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7200 Loss (unaveraged): -0.1509 Loss (averaged): 0.0204\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7211/10000 [1:54:38<1:20:09,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7210 Loss (unaveraged): -0.1905 Loss (averaged): 0.0186\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7221/10000 [1:54:49<1:20:23,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7220 Loss (unaveraged): -0.2513 Loss (averaged): 0.0185\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7231/10000 [1:55:00<1:21:39,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7230 Loss (unaveraged): -0.3698 Loss (averaged): 0.0176\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 72%|███████▏  | 7241/10000 [1:55:10<1:20:49,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7240 Loss (unaveraged): -0.2754 Loss (averaged): 0.0197\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7251/10000 [1:55:21<1:25:04,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7250 Loss (unaveraged): -0.3188 Loss (averaged): 0.0200\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7261/10000 [1:55:32<1:23:42,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7260 Loss (unaveraged): -0.4426 Loss (averaged): 0.0172\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7271/10000 [1:55:42<1:23:05,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7270 Loss (unaveraged): -0.6816 Loss (averaged): 0.0140\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7281/10000 [1:55:53<1:24:53,  1.87s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7280 Loss (unaveraged): -0.8059 Loss (averaged): 0.0175\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7291/10000 [1:56:04<1:23:00,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7290 Loss (unaveraged): -0.5566 Loss (averaged): 0.0155\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7301/10000 [1:56:14<1:18:49,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7300 Loss (unaveraged): -0.2337 Loss (averaged): 0.0160\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7311/10000 [1:56:25<1:17:55,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7310 Loss (unaveraged): -0.0503 Loss (averaged): 0.0137\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7321/10000 [1:56:35<1:16:35,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7320 Loss (unaveraged): 0.0863 Loss (averaged): 0.0164\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7331/10000 [1:56:46<1:18:25,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7330 Loss (unaveraged): 0.1275 Loss (averaged): 0.0160\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 73%|███████▎  | 7341/10000 [1:56:57<1:17:32,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7340 Loss (unaveraged): -0.1581 Loss (averaged): 0.0151\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▎  | 7351/10000 [1:57:08<1:16:49,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7350 Loss (unaveraged): -0.3549 Loss (averaged): 0.0125\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▎  | 7361/10000 [1:57:18<1:18:45,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7360 Loss (unaveraged): -0.4924 Loss (averaged): 0.0161\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▎  | 7371/10000 [1:57:29<1:17:29,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7370 Loss (unaveraged): -0.2849 Loss (averaged): 0.0143\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7381/10000 [1:57:40<1:17:49,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7380 Loss (unaveraged): 0.2291 Loss (averaged): 0.0133\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7391/10000 [1:57:50<1:21:32,  1.88s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7390 Loss (unaveraged): 0.4119 Loss (averaged): 0.0137\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7401/10000 [1:58:01<1:19:58,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7400 Loss (unaveraged): 0.3811 Loss (averaged): 0.0135\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7411/10000 [1:58:12<1:20:28,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7410 Loss (unaveraged): 0.2563 Loss (averaged): 0.0103\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7421/10000 [1:58:23<1:19:51,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7420 Loss (unaveraged): 0.1104 Loss (averaged): 0.0091\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7431/10000 [1:58:33<1:18:17,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7430 Loss (unaveraged): -0.0779 Loss (averaged): 0.0097\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 74%|███████▍  | 7441/10000 [1:58:44<1:14:26,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7440 Loss (unaveraged): -0.2650 Loss (averaged): 0.0056\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▍  | 7451/10000 [1:58:54<1:12:42,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7450 Loss (unaveraged): -0.4408 Loss (averaged): 0.0077\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▍  | 7461/10000 [1:59:05<1:12:47,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7460 Loss (unaveraged): -0.5949 Loss (averaged): 0.0078\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▍  | 7471/10000 [1:59:15<1:13:48,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7470 Loss (unaveraged): -0.7001 Loss (averaged): 0.0059\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▍  | 7481/10000 [1:59:26<1:13:31,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7480 Loss (unaveraged): -0.7624 Loss (averaged): 0.0059\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▍  | 7491/10000 [1:59:37<1:13:14,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7490 Loss (unaveraged): -0.8592 Loss (averaged): 0.0064\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▌  | 7501/10000 [1:59:48<1:14:00,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7500 Loss (unaveraged): -0.7931 Loss (averaged): 0.0068\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▌  | 7511/10000 [1:59:58<1:14:08,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7510 Loss (unaveraged): -0.7540 Loss (averaged): 0.0060\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▌  | 7521/10000 [2:00:09<1:15:04,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7520 Loss (unaveraged): -0.6717 Loss (averaged): 0.0060\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▌  | 7531/10000 [2:00:19<1:16:26,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7530 Loss (unaveraged): -0.5483 Loss (averaged): 0.0036\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 75%|███████▌  | 7541/10000 [2:00:30<1:15:01,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7540 Loss (unaveraged): -0.4349 Loss (averaged): 0.0030\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7551/10000 [2:00:41<1:14:39,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7550 Loss (unaveraged): -0.2024 Loss (averaged): 0.0021\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7561/10000 [2:00:51<1:15:14,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7560 Loss (unaveraged): -0.0856 Loss (averaged): 0.0022\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7571/10000 [2:01:02<1:10:57,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7570 Loss (unaveraged): 0.0942 Loss (averaged): 0.0026\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7581/10000 [2:01:12<1:10:49,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7580 Loss (unaveraged): 0.3043 Loss (averaged): 0.0004\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7591/10000 [2:01:23<1:09:31,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7590 Loss (unaveraged): 0.3262 Loss (averaged): 0.0015\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7601/10000 [2:01:34<1:08:59,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7600 Loss (unaveraged): 0.4673 Loss (averaged): 0.0002\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7611/10000 [2:01:45<1:09:45,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7610 Loss (unaveraged): 0.5772 Loss (averaged): 0.0021\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▌  | 7621/10000 [2:01:55<1:08:44,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7620 Loss (unaveraged): 0.5535 Loss (averaged): -0.0021\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▋  | 7631/10000 [2:02:06<1:09:11,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7630 Loss (unaveraged): 0.5539 Loss (averaged): -0.0006\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 76%|███████▋  | 7641/10000 [2:02:17<1:10:47,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7640 Loss (unaveraged): 0.4840 Loss (averaged): -0.0012\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7651/10000 [2:02:27<1:11:27,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7650 Loss (unaveraged): 0.3644 Loss (averaged): -0.0028\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7661/10000 [2:02:38<1:11:30,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7660 Loss (unaveraged): 0.1668 Loss (averaged): -0.0002\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7671/10000 [2:02:49<1:11:32,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7670 Loss (unaveraged): -1.4852 Loss (averaged): 0.0001\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7681/10000 [2:02:59<1:10:47,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7680 Loss (unaveraged): -1.6851 Loss (averaged): -0.0014\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7691/10000 [2:03:10<1:12:37,  1.89s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7690 Loss (unaveraged): -1.2419 Loss (averaged): -0.0022\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7701/10000 [2:03:20<1:09:25,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7700 Loss (unaveraged): -0.8871 Loss (averaged): -0.0021\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7711/10000 [2:03:31<1:05:43,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7710 Loss (unaveraged): -0.5430 Loss (averaged): -0.0017\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7721/10000 [2:03:42<1:06:29,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7720 Loss (unaveraged): -0.7604 Loss (averaged): -0.0025\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7731/10000 [2:03:52<1:04:59,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7730 Loss (unaveraged): -0.7867 Loss (averaged): -0.0025\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 77%|███████▋  | 7741/10000 [2:04:03<1:04:38,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7740 Loss (unaveraged): -0.6926 Loss (averaged): -0.0054\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7751/10000 [2:04:14<1:05:10,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7750 Loss (unaveraged): -0.5118 Loss (averaged): -0.0059\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7761/10000 [2:04:24<1:05:23,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7760 Loss (unaveraged): -0.3755 Loss (averaged): -0.0029\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7771/10000 [2:04:35<1:05:42,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7770 Loss (unaveraged): -0.3842 Loss (averaged): -0.0044\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7781/10000 [2:04:46<1:06:24,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7780 Loss (unaveraged): -0.5904 Loss (averaged): -0.0066\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7791/10000 [2:04:56<1:07:56,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7790 Loss (unaveraged): -1.0919 Loss (averaged): -0.0087\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7801/10000 [2:05:07<1:08:12,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7800 Loss (unaveraged): -1.9223 Loss (averaged): -0.0087\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7811/10000 [2:05:18<1:06:51,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7810 Loss (unaveraged): -2.4600 Loss (averaged): -0.0071\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7821/10000 [2:05:28<1:06:23,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7820 Loss (unaveraged): -2.6262 Loss (averaged): -0.0105\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7831/10000 [2:05:39<1:07:39,  1.87s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7830 Loss (unaveraged): -2.8944 Loss (averaged): -0.0104\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 78%|███████▊  | 7841/10000 [2:05:49<1:03:53,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7840 Loss (unaveraged): -2.6213 Loss (averaged): -0.0095\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▊  | 7851/10000 [2:06:00<1:01:04,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7850 Loss (unaveraged): -2.4265 Loss (averaged): -0.0077\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▊  | 7861/10000 [2:06:11<1:02:34,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7860 Loss (unaveraged): -1.9547 Loss (averaged): -0.0103\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▊  | 7871/10000 [2:06:21<1:02:14,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7870 Loss (unaveraged): -1.8503 Loss (averaged): -0.0109\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7881/10000 [2:06:32<1:01:20,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7880 Loss (unaveraged): -1.4778 Loss (averaged): -0.0137\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7891/10000 [2:06:43<1:00:49,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7890 Loss (unaveraged): -0.9873 Loss (averaged): -0.0118\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7901/10000 [2:06:53<1:01:22,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7900 Loss (unaveraged): -0.9585 Loss (averaged): -0.0119\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7911/10000 [2:07:04<1:03:12,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7910 Loss (unaveraged): -0.3959 Loss (averaged): -0.0120\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7921/10000 [2:07:15<1:02:50,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7920 Loss (unaveraged): -0.1201 Loss (averaged): -0.0131\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7931/10000 [2:07:25<1:03:26,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7930 Loss (unaveraged): -0.0031 Loss (averaged): -0.0125\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 79%|███████▉  | 7941/10000 [2:07:36<1:04:35,  1.88s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7940 Loss (unaveraged): 0.1957 Loss (averaged): -0.0112\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|███████▉  | 7951/10000 [2:07:47<1:02:50,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7950 Loss (unaveraged): -0.1779 Loss (averaged): -0.0140\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|███████▉  | 7961/10000 [2:07:57<1:02:15,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7960 Loss (unaveraged): -0.5094 Loss (averaged): -0.0130\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|███████▉  | 7971/10000 [2:08:08<1:00:21,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7970 Loss (unaveraged): -0.7019 Loss (averaged): -0.0126\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|███████▉  | 7981/10000 [2:08:18<57:42,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7980 Loss (unaveraged): -1.3279 Loss (averaged): -0.0146\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|███████▉  | 7991/10000 [2:08:29<57:41,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 7990 Loss (unaveraged): -1.9529 Loss (averaged): -0.0129\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|████████  | 8001/10000 [2:08:40<58:12,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8000 Loss (unaveraged): -2.5726 Loss (averaged): -0.0139\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|████████  | 8011/10000 [2:08:50<57:18,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8010 Loss (unaveraged): -3.3177 Loss (averaged): -0.0126\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|████████  | 8021/10000 [2:09:01<58:24,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8020 Loss (unaveraged): -3.6976 Loss (averaged): -0.0088\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|████████  | 8031/10000 [2:09:12<57:36,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8030 Loss (unaveraged): -3.5237 Loss (averaged): -0.0139\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 80%|████████  | 8041/10000 [2:09:23<57:50,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8040 Loss (unaveraged): -3.0854 Loss (averaged): -0.0121\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8051/10000 [2:09:33<59:31,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8050 Loss (unaveraged): -3.1434 Loss (averaged): -0.0146\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8061/10000 [2:09:44<58:59,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8060 Loss (unaveraged): -2.7324 Loss (averaged): -0.0142\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8071/10000 [2:09:54<58:31,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8070 Loss (unaveraged): -1.8216 Loss (averaged): -0.0189\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8081/10000 [2:10:05<58:57,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8080 Loss (unaveraged): -1.9084 Loss (averaged): -0.0172\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8091/10000 [2:10:16<58:42,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8090 Loss (unaveraged): -1.3870 Loss (averaged): -0.0135\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8101/10000 [2:10:26<57:18,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8100 Loss (unaveraged): -0.7205 Loss (averaged): -0.0155\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8111/10000 [2:10:37<54:10,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8110 Loss (unaveraged): -0.3163 Loss (averaged): -0.0163\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████  | 8121/10000 [2:10:47<53:38,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8120 Loss (unaveraged): -0.0161 Loss (averaged): -0.0179\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████▏ | 8131/10000 [2:10:58<54:01,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8130 Loss (unaveraged): 0.2628 Loss (averaged): -0.0159\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 81%|████████▏ | 8141/10000 [2:11:09<53:43,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8140 Loss (unaveraged): 0.4943 Loss (averaged): -0.0179\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8151/10000 [2:11:19<53:27,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8150 Loss (unaveraged): 0.8555 Loss (averaged): -0.0188\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8161/10000 [2:11:30<55:04,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8160 Loss (unaveraged): 0.8635 Loss (averaged): -0.0183\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8171/10000 [2:11:41<54:01,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8170 Loss (unaveraged): 1.0477 Loss (averaged): -0.0178\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8181/10000 [2:11:52<53:38,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8180 Loss (unaveraged): 1.0942 Loss (averaged): -0.0189\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8191/10000 [2:12:02<54:34,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8190 Loss (unaveraged): 1.0954 Loss (averaged): -0.0226\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8201/10000 [2:12:13<55:04,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8200 Loss (unaveraged): 1.0198 Loss (averaged): -0.0201\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8211/10000 [2:12:24<56:26,  1.89s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8210 Loss (unaveraged): 0.9835 Loss (averaged): -0.0239\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8221/10000 [2:12:34<54:33,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8220 Loss (unaveraged): 0.9303 Loss (averaged): -0.0229\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8231/10000 [2:12:45<54:23,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8230 Loss (unaveraged): 0.8609 Loss (averaged): -0.0218\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 82%|████████▏ | 8241/10000 [2:12:56<53:30,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8240 Loss (unaveraged): 0.8597 Loss (averaged): -0.0232\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8251/10000 [2:13:06<50:36,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8250 Loss (unaveraged): 0.8331 Loss (averaged): -0.0253\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8261/10000 [2:13:17<49:35,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8260 Loss (unaveraged): 0.8106 Loss (averaged): -0.0230\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8271/10000 [2:13:28<50:06,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8270 Loss (unaveraged): 0.8046 Loss (averaged): -0.0270\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8281/10000 [2:13:38<49:44,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8280 Loss (unaveraged): 0.7774 Loss (averaged): -0.0247\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8291/10000 [2:13:49<49:08,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8290 Loss (unaveraged): 0.7278 Loss (averaged): -0.0250\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8301/10000 [2:14:00<49:09,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8300 Loss (unaveraged): 0.6982 Loss (averaged): -0.0282\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8311/10000 [2:14:10<50:03,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8310 Loss (unaveraged): 0.6918 Loss (averaged): -0.0274\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8321/10000 [2:14:21<51:16,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8320 Loss (unaveraged): 0.6727 Loss (averaged): -0.0291\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8331/10000 [2:14:32<50:59,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8330 Loss (unaveraged): 0.6712 Loss (averaged): -0.0289\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 83%|████████▎ | 8341/10000 [2:14:42<51:38,  1.87s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8340 Loss (unaveraged): 0.6368 Loss (averaged): -0.0282\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▎ | 8351/10000 [2:14:53<51:32,  1.88s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8350 Loss (unaveraged): 0.6209 Loss (averaged): -0.0300\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▎ | 8361/10000 [2:15:04<49:41,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8360 Loss (unaveraged): 0.6182 Loss (averaged): -0.0279\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▎ | 8371/10000 [2:15:14<48:10,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8370 Loss (unaveraged): 0.6021 Loss (averaged): -0.0300\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8381/10000 [2:15:25<47:32,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8380 Loss (unaveraged): 0.5856 Loss (averaged): -0.0325\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8391/10000 [2:15:35<45:57,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8390 Loss (unaveraged): 0.5772 Loss (averaged): -0.0343\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8401/10000 [2:15:46<46:07,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8400 Loss (unaveraged): 0.5441 Loss (averaged): -0.0327\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8411/10000 [2:15:57<45:01,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8410 Loss (unaveraged): 0.5477 Loss (averaged): -0.0338\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8421/10000 [2:16:07<45:53,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8420 Loss (unaveraged): 0.5525 Loss (averaged): -0.0326\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8431/10000 [2:16:18<46:56,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8430 Loss (unaveraged): 0.5399 Loss (averaged): -0.0319\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 84%|████████▍ | 8441/10000 [2:16:29<46:02,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8440 Loss (unaveraged): 0.5283 Loss (averaged): -0.0327\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▍ | 8451/10000 [2:16:40<46:07,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8450 Loss (unaveraged): 0.5304 Loss (averaged): -0.0348\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▍ | 8461/10000 [2:16:50<47:47,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8460 Loss (unaveraged): 0.5210 Loss (averaged): -0.0363\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▍ | 8471/10000 [2:17:01<47:00,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8470 Loss (unaveraged): 0.5247 Loss (averaged): -0.0380\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▍ | 8481/10000 [2:17:12<46:13,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8480 Loss (unaveraged): 0.5410 Loss (averaged): -0.0351\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▍ | 8491/10000 [2:17:22<46:28,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8490 Loss (unaveraged): 0.5120 Loss (averaged): -0.0347\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▌ | 8501/10000 [2:17:33<45:29,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8500 Loss (unaveraged): 0.5175 Loss (averaged): -0.0371\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▌ | 8511/10000 [2:17:43<42:19,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8510 Loss (unaveraged): 0.5468 Loss (averaged): -0.0379\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▌ | 8521/10000 [2:17:54<41:51,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8520 Loss (unaveraged): 0.5187 Loss (averaged): -0.0402\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▌ | 8531/10000 [2:18:04<41:45,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8530 Loss (unaveraged): 0.5430 Loss (averaged): -0.0371\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 85%|████████▌ | 8541/10000 [2:18:15<42:41,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8540 Loss (unaveraged): 0.5248 Loss (averaged): -0.0378\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8551/10000 [2:18:26<42:01,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8550 Loss (unaveraged): 0.5296 Loss (averaged): -0.0370\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8561/10000 [2:18:36<41:48,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8560 Loss (unaveraged): 0.5134 Loss (averaged): -0.0367\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8571/10000 [2:18:47<42:25,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8570 Loss (unaveraged): 0.5234 Loss (averaged): -0.0381\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8581/10000 [2:18:58<42:12,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8580 Loss (unaveraged): 0.5165 Loss (averaged): -0.0358\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8591/10000 [2:19:08<43:12,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8590 Loss (unaveraged): 0.5002 Loss (averaged): -0.0385\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8601/10000 [2:19:19<42:46,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8600 Loss (unaveraged): 0.4904 Loss (averaged): -0.0384\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8611/10000 [2:19:30<42:23,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8610 Loss (unaveraged): 0.4637 Loss (averaged): -0.0399\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▌ | 8621/10000 [2:19:40<42:42,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8620 Loss (unaveraged): 0.4352 Loss (averaged): -0.0400\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▋ | 8631/10000 [2:19:51<41:47,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8630 Loss (unaveraged): 0.3972 Loss (averaged): -0.0380\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 86%|████████▋ | 8641/10000 [2:20:01<38:26,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8640 Loss (unaveraged): 0.3795 Loss (averaged): -0.0394\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8651/10000 [2:20:12<39:08,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8650 Loss (unaveraged): 0.2768 Loss (averaged): -0.0388\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8661/10000 [2:20:22<38:20,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8660 Loss (unaveraged): -0.2528 Loss (averaged): -0.0393\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8671/10000 [2:20:33<38:11,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8670 Loss (unaveraged): -2.8558 Loss (averaged): -0.0411\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8681/10000 [2:20:44<38:01,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8680 Loss (unaveraged): -3.3102 Loss (averaged): -0.0419\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8691/10000 [2:20:54<37:46,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8690 Loss (unaveraged): -3.3737 Loss (averaged): -0.0409\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8701/10000 [2:21:05<38:32,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8700 Loss (unaveraged): -3.4560 Loss (averaged): -0.0415\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8711/10000 [2:21:16<38:41,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8710 Loss (unaveraged): -3.0327 Loss (averaged): -0.0435\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8721/10000 [2:21:26<39:23,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8720 Loss (unaveraged): -2.8947 Loss (averaged): -0.0415\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8731/10000 [2:21:37<40:00,  1.89s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8730 Loss (unaveraged): -2.0560 Loss (averaged): -0.0396\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 87%|████████▋ | 8741/10000 [2:21:48<38:31,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8740 Loss (unaveraged): -1.5895 Loss (averaged): -0.0432\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8751/10000 [2:21:58<38:16,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8750 Loss (unaveraged): -1.2900 Loss (averaged): -0.0458\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8761/10000 [2:22:09<37:36,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8760 Loss (unaveraged): -0.7647 Loss (averaged): -0.0444\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8771/10000 [2:22:19<35:15,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8770 Loss (unaveraged): -0.0437 Loss (averaged): -0.0409\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8781/10000 [2:22:30<34:51,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8780 Loss (unaveraged): 0.5680 Loss (averaged): -0.0414\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8791/10000 [2:22:41<34:41,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8790 Loss (unaveraged): 0.7869 Loss (averaged): -0.0430\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8801/10000 [2:22:51<34:32,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8800 Loss (unaveraged): 1.0527 Loss (averaged): -0.0421\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8811/10000 [2:23:02<34:05,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8810 Loss (unaveraged): 1.1419 Loss (averaged): -0.0442\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8821/10000 [2:23:12<34:12,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8820 Loss (unaveraged): 0.3812 Loss (averaged): -0.0422\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8831/10000 [2:23:23<35:08,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8830 Loss (unaveraged): -0.0418 Loss (averaged): -0.0431\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 88%|████████▊ | 8841/10000 [2:23:34<35:25,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8840 Loss (unaveraged): -0.3539 Loss (averaged): -0.0406\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▊ | 8851/10000 [2:23:45<35:09,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8850 Loss (unaveraged): -0.7134 Loss (averaged): -0.0407\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▊ | 8861/10000 [2:23:55<34:59,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8860 Loss (unaveraged): -0.9847 Loss (averaged): -0.0439\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▊ | 8871/10000 [2:24:06<34:36,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8870 Loss (unaveraged): -1.2490 Loss (averaged): -0.0437\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8881/10000 [2:24:16<34:08,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8880 Loss (unaveraged): -1.4807 Loss (averaged): -0.0405\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8891/10000 [2:24:27<34:32,  1.87s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8890 Loss (unaveraged): -1.6271 Loss (averaged): -0.0439\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8901/10000 [2:24:38<31:40,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8900 Loss (unaveraged): -1.8145 Loss (averaged): -0.0417\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8911/10000 [2:24:48<31:12,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8910 Loss (unaveraged): -1.9508 Loss (averaged): -0.0427\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8921/10000 [2:24:59<31:14,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8920 Loss (unaveraged): -2.1414 Loss (averaged): -0.0432\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8931/10000 [2:25:10<30:55,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8930 Loss (unaveraged): -2.2758 Loss (averaged): -0.0450\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 89%|████████▉ | 8941/10000 [2:25:20<30:43,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8940 Loss (unaveraged): -2.2885 Loss (averaged): -0.0447\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|████████▉ | 8951/10000 [2:25:31<31:18,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8950 Loss (unaveraged): -2.4386 Loss (averaged): -0.0464\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|████████▉ | 8961/10000 [2:25:42<30:20,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8960 Loss (unaveraged): -2.4503 Loss (averaged): -0.0484\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|████████▉ | 8971/10000 [2:25:52<30:34,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8970 Loss (unaveraged): -2.5352 Loss (averaged): -0.0468\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|████████▉ | 8981/10000 [2:26:03<30:52,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8980 Loss (unaveraged): -2.5647 Loss (averaged): -0.0427\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|████████▉ | 8991/10000 [2:26:14<30:38,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 8990 Loss (unaveraged): -2.6081 Loss (averaged): -0.0486\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|█████████ | 9001/10000 [2:26:24<31:21,  1.88s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9000 Loss (unaveraged): -2.6447 Loss (averaged): -0.0473\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|█████████ | 9011/10000 [2:26:35<30:28,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9010 Loss (unaveraged): -2.6284 Loss (averaged): -0.0498\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|█████████ | 9021/10000 [2:26:46<29:57,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9020 Loss (unaveraged): -2.6032 Loss (averaged): -0.0505\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|█████████ | 9031/10000 [2:26:56<29:22,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9030 Loss (unaveraged): -2.5368 Loss (averaged): -0.0483\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 90%|█████████ | 9041/10000 [2:27:07<27:31,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9040 Loss (unaveraged): -2.4882 Loss (averaged): -0.0510\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9051/10000 [2:27:17<27:00,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9050 Loss (unaveraged): -2.4807 Loss (averaged): -0.0547\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9061/10000 [2:27:28<27:28,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9060 Loss (unaveraged): -2.3426 Loss (averaged): -0.0525\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9071/10000 [2:27:39<26:50,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9070 Loss (unaveraged): -2.2652 Loss (averaged): -0.0543\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9081/10000 [2:27:50<26:40,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9080 Loss (unaveraged): -2.1853 Loss (averaged): -0.0579\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9091/10000 [2:28:00<26:31,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9090 Loss (unaveraged): -2.0081 Loss (averaged): -0.0573\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9101/10000 [2:28:11<26:19,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9100 Loss (unaveraged): -1.8747 Loss (averaged): -0.0583\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9111/10000 [2:28:22<27:26,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9110 Loss (unaveraged): -1.7018 Loss (averaged): -0.0594\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████ | 9121/10000 [2:28:32<26:58,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9120 Loss (unaveraged): -1.6039 Loss (averaged): -0.0642\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████▏| 9131/10000 [2:28:43<26:42,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9130 Loss (unaveraged): -1.4014 Loss (averaged): -0.0656\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 91%|█████████▏| 9141/10000 [2:28:53<26:27,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9140 Loss (unaveraged): -1.2095 Loss (averaged): -0.0607\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9151/10000 [2:29:04<26:01,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9150 Loss (unaveraged): -1.1100 Loss (averaged): -0.0663\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9161/10000 [2:29:15<25:05,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9160 Loss (unaveraged): -0.8000 Loss (averaged): -0.0629\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9171/10000 [2:29:25<23:40,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9170 Loss (unaveraged): -0.6462 Loss (averaged): -0.0690\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9181/10000 [2:29:36<23:08,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9180 Loss (unaveraged): -0.3945 Loss (averaged): -0.0684\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9191/10000 [2:29:46<22:57,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9190 Loss (unaveraged): -0.2297 Loss (averaged): -0.0702\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9201/10000 [2:29:57<22:50,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9200 Loss (unaveraged): -0.0473 Loss (averaged): -0.0716\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9211/10000 [2:30:07<22:39,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9210 Loss (unaveraged): 0.2389 Loss (averaged): -0.0699\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9221/10000 [2:30:18<23:07,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9220 Loss (unaveraged): 0.3790 Loss (averaged): -0.0767\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9231/10000 [2:30:29<23:01,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9230 Loss (unaveraged): 0.5095 Loss (averaged): -0.0771\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 92%|█████████▏| 9241/10000 [2:30:39<23:11,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9240 Loss (unaveraged): 0.6624 Loss (averaged): -0.0809\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9251/10000 [2:30:50<23:11,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9250 Loss (unaveraged): 0.8101 Loss (averaged): -0.0742\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9261/10000 [2:31:01<22:41,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9260 Loss (unaveraged): 0.8808 Loss (averaged): -0.0784\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9271/10000 [2:31:11<22:23,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9270 Loss (unaveraged): 1.0030 Loss (averaged): -0.0749\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9281/10000 [2:31:22<22:01,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9280 Loss (unaveraged): 0.8216 Loss (averaged): -0.0828\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9291/10000 [2:31:33<21:16,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9290 Loss (unaveraged): -1.5049 Loss (averaged): -0.0816\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9301/10000 [2:31:43<20:15,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9300 Loss (unaveraged): -1.6382 Loss (averaged): -0.0853\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9311/10000 [2:31:54<19:42,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9310 Loss (unaveraged): -1.4585 Loss (averaged): -0.0866\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9321/10000 [2:32:05<19:28,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9320 Loss (unaveraged): -1.3601 Loss (averaged): -0.0883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9331/10000 [2:32:15<19:48,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9330 Loss (unaveraged): -1.4831 Loss (averaged): -0.0850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 93%|█████████▎| 9341/10000 [2:32:26<18:55,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9340 Loss (unaveraged): -1.3719 Loss (averaged): -0.0872\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▎| 9351/10000 [2:32:37<18:57,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9350 Loss (unaveraged): -0.8762 Loss (averaged): -0.0925\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▎| 9361/10000 [2:32:47<18:58,  1.78s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9360 Loss (unaveraged): -0.9283 Loss (averaged): -0.0939\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▎| 9371/10000 [2:32:58<18:58,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9370 Loss (unaveraged): -0.4584 Loss (averaged): -0.0907\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9381/10000 [2:33:09<18:43,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9380 Loss (unaveraged): -0.1251 Loss (averaged): -0.0908\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9391/10000 [2:33:19<18:40,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9390 Loss (unaveraged): 0.1273 Loss (averaged): -0.0953\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9401/10000 [2:33:30<18:15,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9400 Loss (unaveraged): 0.6279 Loss (averaged): -0.0936\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9411/10000 [2:33:41<17:55,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9410 Loss (unaveraged): 0.7169 Loss (averaged): -0.0894\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9421/10000 [2:33:51<17:46,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9420 Loss (unaveraged): 0.9747 Loss (averaged): -0.0943\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9431/10000 [2:34:02<16:34,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9430 Loss (unaveraged): 1.2814 Loss (averaged): -0.0950\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 94%|█████████▍| 9441/10000 [2:34:12<16:17,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9440 Loss (unaveraged): 1.0241 Loss (averaged): -0.1017\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▍| 9451/10000 [2:34:23<15:39,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9450 Loss (unaveraged): 0.9818 Loss (averaged): -0.0997\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▍| 9461/10000 [2:34:34<15:37,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9460 Loss (unaveraged): 0.9293 Loss (averaged): -0.1024\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▍| 9471/10000 [2:34:44<15:29,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9470 Loss (unaveraged): 1.0243 Loss (averaged): -0.0971\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▍| 9481/10000 [2:34:55<15:04,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9480 Loss (unaveraged): 1.0803 Loss (averaged): -0.0984\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▍| 9491/10000 [2:35:06<14:52,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9490 Loss (unaveraged): 1.1095 Loss (averaged): -0.1046\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▌| 9501/10000 [2:35:16<14:56,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9500 Loss (unaveraged): 1.1461 Loss (averaged): -0.0988\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▌| 9511/10000 [2:35:27<14:59,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9510 Loss (unaveraged): 1.2246 Loss (averaged): -0.1000\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▌| 9521/10000 [2:35:38<14:45,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9520 Loss (unaveraged): 1.2789 Loss (averaged): -0.1005\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▌| 9531/10000 [2:35:48<14:21,  1.84s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9530 Loss (unaveraged): 1.3556 Loss (averaged): -0.1049\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 95%|█████████▌| 9541/10000 [2:35:59<14:19,  1.87s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9540 Loss (unaveraged): 1.3930 Loss (averaged): -0.1043\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9551/10000 [2:36:10<14:17,  1.91s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9550 Loss (unaveraged): 1.4642 Loss (averaged): -0.1025\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9561/10000 [2:36:20<12:42,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9560 Loss (unaveraged): 1.5179 Loss (averaged): -0.1024\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9571/10000 [2:36:31<12:17,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9570 Loss (unaveraged): 1.5694 Loss (averaged): -0.0991\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9581/10000 [2:36:41<12:16,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9580 Loss (unaveraged): 1.5941 Loss (averaged): -0.1016\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9591/10000 [2:36:52<11:41,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9590 Loss (unaveraged): 1.6122 Loss (averaged): -0.1002\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9601/10000 [2:37:03<11:27,  1.72s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9600 Loss (unaveraged): 1.6761 Loss (averaged): -0.1067\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9611/10000 [2:37:13<11:21,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9610 Loss (unaveraged): 1.6640 Loss (averaged): -0.1050\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▌| 9621/10000 [2:37:24<11:10,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9620 Loss (unaveraged): 1.6596 Loss (averaged): -0.1027\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▋| 9631/10000 [2:37:35<11:03,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9630 Loss (unaveraged): 1.6842 Loss (averaged): -0.1098\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 96%|█████████▋| 9641/10000 [2:37:45<10:47,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9640 Loss (unaveraged): 1.6493 Loss (averaged): -0.1111\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9651/10000 [2:37:56<10:31,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9650 Loss (unaveraged): 1.6780 Loss (averaged): -0.1060\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9661/10000 [2:38:07<10:28,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9660 Loss (unaveraged): 1.6338 Loss (averaged): -0.1005\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9671/10000 [2:38:17<10:03,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9670 Loss (unaveraged): 1.5999 Loss (averaged): -0.1076\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9681/10000 [2:38:28<09:36,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9680 Loss (unaveraged): 1.5534 Loss (averaged): -0.1055\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9691/10000 [2:38:38<09:12,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9690 Loss (unaveraged): 1.5272 Loss (averaged): -0.1003\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9701/10000 [2:38:49<08:32,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9700 Loss (unaveraged): 1.4964 Loss (averaged): -0.1100\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9711/10000 [2:38:59<08:14,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9710 Loss (unaveraged): 1.4310 Loss (averaged): -0.1082\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9721/10000 [2:39:10<08:02,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9720 Loss (unaveraged): 1.3876 Loss (averaged): -0.1104\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9731/10000 [2:39:20<07:40,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9730 Loss (unaveraged): 1.3307 Loss (averaged): -0.1072\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 97%|█████████▋| 9741/10000 [2:39:31<07:30,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9740 Loss (unaveraged): 1.2617 Loss (averaged): -0.0957\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9751/10000 [2:39:42<07:21,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9750 Loss (unaveraged): 1.2156 Loss (averaged): -0.1054\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9761/10000 [2:39:52<07:06,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9760 Loss (unaveraged): 1.1984 Loss (averaged): -0.1057\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9771/10000 [2:40:03<07:04,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9770 Loss (unaveraged): 1.1048 Loss (averaged): -0.1004\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9781/10000 [2:40:14<06:41,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9780 Loss (unaveraged): 1.0435 Loss (averaged): -0.0965\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9791/10000 [2:40:24<06:25,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9790 Loss (unaveraged): 0.9529 Loss (averaged): -0.1044\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9801/10000 [2:40:35<06:04,  1.83s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9800 Loss (unaveraged): 0.8852 Loss (averaged): -0.1018\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9811/10000 [2:40:45<05:32,  1.76s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9810 Loss (unaveraged): 0.7912 Loss (averaged): -0.0895\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9821/10000 [2:40:56<05:04,  1.70s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9820 Loss (unaveraged): 0.6859 Loss (averaged): -0.0915\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9831/10000 [2:41:06<04:48,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9830 Loss (unaveraged): 0.6330 Loss (averaged): -0.0955\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 98%|█████████▊| 9841/10000 [2:41:17<04:34,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9840 Loss (unaveraged): 0.5607 Loss (averaged): -0.0937\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▊| 9851/10000 [2:41:28<04:24,  1.77s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9850 Loss (unaveraged): 0.5079 Loss (averaged): -0.1028\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▊| 9861/10000 [2:41:38<04:01,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9860 Loss (unaveraged): 0.4262 Loss (averaged): -0.0984\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▊| 9871/10000 [2:41:49<03:44,  1.74s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9870 Loss (unaveraged): 0.3934 Loss (averaged): -0.0974\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9881/10000 [2:42:00<03:34,  1.81s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9880 Loss (unaveraged): 0.3458 Loss (averaged): -0.0984\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9891/10000 [2:42:10<03:15,  1.79s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9890 Loss (unaveraged): 0.3356 Loss (averaged): -0.0933\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9901/10000 [2:42:21<03:00,  1.82s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9900 Loss (unaveraged): 0.3184 Loss (averaged): -0.0918\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9911/10000 [2:42:32<02:44,  1.85s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9910 Loss (unaveraged): 0.3009 Loss (averaged): -0.0916\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9921/10000 [2:42:42<02:28,  1.88s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9920 Loss (unaveraged): 0.3201 Loss (averaged): -0.0954\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9931/10000 [2:42:53<02:08,  1.86s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9930 Loss (unaveraged): 0.3175 Loss (averaged): -0.0850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            " 99%|█████████▉| 9941/10000 [2:43:03<01:45,  1.80s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9940 Loss (unaveraged): 0.3521 Loss (averaged): -0.0891\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|█████████▉| 9951/10000 [2:43:14<01:24,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9950 Loss (unaveraged): 0.3665 Loss (averaged): -0.0849\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|█████████▉| 9961/10000 [2:43:25<01:08,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9960 Loss (unaveraged): 0.4081 Loss (averaged): -0.0949\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|█████████▉| 9971/10000 [2:43:35<00:49,  1.71s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9970 Loss (unaveraged): 0.4485 Loss (averaged): -0.0887\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|█████████▉| 9981/10000 [2:43:46<00:32,  1.73s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9980 Loss (unaveraged): 0.4862 Loss (averaged): -0.0978\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|█████████▉| 9991/10000 [2:43:57<00:15,  1.75s/it]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9990 Loss (unaveraged): 0.5392 Loss (averaged): -0.0858\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 10000/10000 [2:44:07<00:00,  1.02it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Step: 9999 Loss (unaveraged): 0.5508 Loss (averaged): -0.0891\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "ERROR: Could not consume arg: -f\n",
            "Usage: ipykernel_launcher.py -\n",
            "\n",
            "For detailed information on this command, run:\n",
            "  ipykernel_launcher.py - --help\n"
          ]
        },
        {
          "output_type": "error",
          "ename": "FireExit",
          "evalue": "ignored",
          "traceback": [
            "An exception has occurred, use %tb to see the full traceback.\n",
            "\u001b[0;31mFireExit\u001b[0m\u001b[0;31m:\u001b[0m 2\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/IPython/core/interactiveshell.py:3561: UserWarning: To exit: use 'exit', 'quit', or Ctrl-D.\n",
            "  warn(\"To exit: use 'exit', 'quit', or Ctrl-D.\", stacklevel=1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "###################\n",
        "# Appendix: discriminators for a neural SDE\n",
        "#\n",
        "# This is a little long, but should all be quite straightforward. By the end of this you should have a comprehensive\n",
        "# knowledge of how these things fit together.\n",
        "#\n",
        "# Let Y be the real/generated sample, and let H be the hidden state of the discriminator.\n",
        "# For real data, then Y is some interpolation of an (irregular) time series. (As with neural CDEs, if you're familiar -\n",
        "# for a nice exposition on this see https://github.com/patrick-kidger/torchcde/blob/master/example/irregular_data.py.)\n",
        "# In the case of generated data, then Y is _either_ the continuous-time sample produced by sdeint, _or_ it is an\n",
        "# interpolation (probably linear interpolation) of the generated sample between particular evaluation points, We'll\n",
        "# refer to these as cases (*) and (**) respectively.\n",
        "#\n",
        "# In terms of the mathematics, our options for the discriminator are:\n",
        "# (a1) Solve dH(t) = f(t, H(t)) dt + g(t, H(t)) dY(t),\n",
        "# (a2) Solve dH(t) = (f, g)(t, H(t)) d(t, Y(t))\n",
        "# (b) Solve dH(t) = f(t, H(t), Y(t)) dt.\n",
        "# Option (a1) is what is stated in the paper \"Neural SDE as Infinite-Dimensional GANs\".\n",
        "# Option (a2) is theoretically the same as (a1), but the drift and diffusion have been merged into a single function,\n",
        "# and the sample Y has been augmented with time. This can sometimes be a more helpful way to think about things.\n",
        "# Option (b) is a special case of the first two, by Appendix C of arXiv:2005.08926.\n",
        "# [Note that just dH(t) = g(t, H(t)) dY(t) would _not_ be enough, by what's known as the tree-like equivalence property.\n",
        "#  It's a bit technical, but the basic idea is that the discriminator wouldn't be able to tell how fast we traverse Y.\n",
        "#  This is a really easy mistake to make; make sure you don't fall into it.]\n",
        "#\n",
        "# Whether we use (*) or (**), and (a1) or (a2) or (b), doesn't really affect the quality of the discriminator, as far as\n",
        "# we know. However, these distinctions do affect how we solve them in terms of code. Depending on each combination, our\n",
        "# options are to use a solver of the following types:\n",
        "#\n",
        "#      | (a1)   (a2)   (b)\n",
        "# -----+----------------------\n",
        "#  (*) | SDE           SDE\n",
        "# (**) |        CDE    ODE\n",
        "#\n",
        "# So, (*) implies using an SDE solver: the continuous-time sample is only really available inside sdeint, so if we're\n",
        "# going to use the continuous-time sample then we need to solve generator and discriminator together inside a single SDE\n",
        "# solve. In this case, as our generator takes the form\n",
        "# Y(t) = l(X(t)) with dX(t) = μ(t, X(t)) dt + σ(t, X(t)) dW(t),\n",
        "# then\n",
        "# dY(t) = l(X(t)) dX(t) = l(X(t))μ(t, X(t)) dt + l(X(t))σ(t, X(t)) dW(t).\n",
        "# Then for (a1) we get\n",
        "# dH(t) = ( f(t, H(t)) + g(t, H(t))l(X(t))μ(t, X(t)) ) dt + g(t, H(t))l(X(t))σ(t, X(t)) dW(t),\n",
        "# which we can now put together into one big SDE solve:\n",
        "#  ( X(t) )   ( μ(t, X(t)                                )      ( σ(t, X(t))                  )\n",
        "# d( Y(t) ) = ( l(X(t))μ(t, X(t)                         ) dt + ( l(X(t))σ(t, X(t))           ) dW(t)\n",
        "#  ( H(t) )   ( f(t, H(t)) + g(t, H(t))l(X(t))μ(t, X(t)) )      ( g(t, H(t))l(X(t))σ(t, X(t)) ),\n",
        "# whilst for (b) we can put things together into one big SDE solve:\n",
        "#  ( X(t) )   ( μ(t, X(t))       )      ( σ(t, X(t))        )\n",
        "# d( Y(t) ) = ( l(X(t))μ(t, X(t) ) dt + ( l(X(t))σ(t, X(t)) ) dW(t)\n",
        "#  ( H(t) )   ( f(t, H(t), Y(t)) )      ( 0                 )\n",
        "#\n",
        "# Phew, what a lot of stuff to write down. Don't be put off by this: there's no complicated algebra, it's literally just\n",
        "# substituting one equation into another. Also, note that all of this is for the _generated_ data. If using real data,\n",
        "# then Y(t) is as previously described always an interpolation of the data. If you're able to evaluate the derivative of\n",
        "# the interpolation then you can then apply (a1) by rewriting it as dY(t) = (dY/dt)(t) dt and substituting in. If you're\n",
        "# able to evaluate the interpolation itself then you can apply (b) directly.\n",
        "#\n",
        "# The benefit of using (*) is that everything can be done inside a single SDE solve, which is important if you're\n",
        "# thinking about using adjoint methods and the like, for memory efficiency. The downside is that the code gets a bit\n",
        "# more complicated: you need to be able to solve just the generator on its own (to produce samples at inference time),\n",
        "# just the discriminator on its own (to evaluate the discriminator on the real data), and the combined\n",
        "# generator-discriminator system (to evaluate the discriminator on the generated data).\n",
        "#\n",
        "# Right, let's move on to (**). In comparison, this is much simpler. We don't need to substitute in anything. We're just\n",
        "# taking our generated data, sampling it at a bunch of points, and then doing some kind of interpolation (probably\n",
        "# linear interpolation). Then we either solve (a2) directly with a CDE solver (regardless of whether we're using real or\n",
        "# generated data), or solve (b) directly with an ODE solver (regardless of whether we're using real or generated data).\n",
        "#\n",
        "# The benefit of this is that it's much simpler to code: unlike (*) we can separate the generator and discriminator, and\n",
        "# don't ever need to combine them. Also, real and generated data is treated the same in the discriminator. (Which is\n",
        "# arguably a good thing anyway.) The downside is that we can't really take advantage of things like adjoint methods to\n",
        "# backpropagate efficiently through the generator, because we need to produce (and thus store) our generated sample at\n",
        "# lots of time points, which reduces the memory efficiency.\n",
        "#\n",
        "# Note that the use of ODE solvers for (**) is only valid because we're using _interpolated_ real or generated data,\n",
        "# and we're assuming that we're using some kind of interpolation that is at least piecewise smooth. (For example, linear\n",
        "# interpolation is piecewise smooth.) It wouldn't make sense to apply ODE solvers to some rough signal like Brownian\n",
        "# motion - that's what case (*) and SDE solvers are about.\n",
        "#\n",
        "# Right, let's wrap up this wall of text. Here, we use option (**), (a2). This is arguably the simplest option, and\n",
        "# is chosen as we'd like to keep the code readable in this example. To solve the CDEs we use the CDE solvers available\n",
        "# through torchcde: https://github.com/patrick-kidger/torchcde.\n",
        "###################"
      ],
      "metadata": {
        "id": "BKhDPnlqjtH3"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}