TTNN relu_bw sweep migration (#10729)

* #10147: Migrated backward relu

* #10147: Corrected ttnn ops

tests/tt_eager/python_api_testing/sweep_tests/op_map.py

@@ -1186,10 +1186,6 @@
         "tt_op": tt_lib_ops.abs_bw,
         "pytorch_op": pytorch_ops.abs_bw,
     },
-    "relu-bw": {
-        "tt_op": tt_lib_ops.relu_bw,
-        "pytorch_op": pytorch_ops.relu_bw,
-    },
     "gt-bw": {
         "tt_op": tt_lib_ops.gt_bw,
         "pytorch_op": pytorch_ops.gt_bw,

tests/tt_eager/python_api_testing/sweep_tests/tt_lib_ops.py

@@ -3067,26 +3067,6 @@ def abs_bw(
     return tt2torch_tensor(t2)
 
 
-@setup_host_and_device
-def relu_bw(
-    x,
-    y,
-    *args,
-    device,
-    dtype,
-    layout,
-    input_mem_config,
-    output_mem_config,
-    **kwargs,
-):
-    t0 = setup_tt_tensor(x, device, layout[0], input_mem_config[0], dtype[0])
-    t1 = setup_tt_tensor(y, device, layout[1], input_mem_config[1], dtype[1])
-
-    t2 = ttl.tensor.relu_bw(t0, t1, output_mem_config)[0]
-
-    return tt2torch_tensor(t2)
-
-
 @setup_host_and_device
 def gt_bw(
     x,

tests/ttnn/python_api_testing/sweep_tests/op_map.py

@@ -773,4 +773,8 @@
         "tt_op": ttnn_ops.addalpha_bw,
         "pytorch_op": pytorch_ops.addalpha_bw,
     },
+    "relu-bw": {
+        "tt_op": ttnn_ops.relu_bw,
+        "pytorch_op": pytorch_ops.relu_bw,
+    },
 }

tests/ttnn/python_api_testing/sweep_tests/test_configs/ci_sweep_tests_working/grayskull/ttnn_backward_relu_test.yaml

@@ -0,0 +1,27 @@
+---
+test-list:
+  - relu-bw:
+      shape:
+        start-shape: [1, 1, 32, 32]
+        end-shape: [6, 12, 256, 256]
+        interval: [1, 1, 32, 32]
+        num-shapes: 2
+        num-samples: 128
+        num-dims: [2, 3, 4]
+        args-sampling-strategy: "all"
+      datagen:
+        function: gen_rand
+        args:
+          low: -100
+          high: 100
+      comparison:
+        function: comp_pcc
+      args-gen: gen_scalar_args
+      args:
+        data-layout: ["TILE"]
+        data-type: ["BFLOAT16"]
+        buffer-type: ["DRAM", "L1"]
+        out-buffer-type: ["DRAM", "L1"]
+      output-file: backward_relu_sweep.csv
+      env:
+        # TT_PCI_DMA_BUF_SIZE: "1048576"

tests/ttnn/python_api_testing/sweep_tests/test_configs/ci_sweep_tests_working/wormhole/ttnn_backward_relu_test.yaml

@@ -0,0 +1,27 @@
+---
+test-list:
+  - relu-bw:
+      shape:
+        start-shape: [1, 1, 32, 32]
+        end-shape: [6, 12, 256, 256]
+        interval: [1, 1, 32, 32]
+        num-shapes: 2
+        num-samples: 128
+        num-dims: [2, 3, 4]
+        args-sampling-strategy: "all"
+      datagen:
+        function: gen_rand
+        args:
+          low: -100
+          high: 100
+      comparison:
+        function: comp_pcc
+      args-gen: gen_scalar_args
+      args:
+        data-layout: ["TILE"]
+        data-type: ["BFLOAT16"]
+        buffer-type: ["DRAM", "L1"]
+        out-buffer-type: ["DRAM", "L1"]
+      output-file: backward_relu_sweep.csv
+      env:
+        # TT_PCI_DMA_BUF_SIZE: "1048576"

tests/ttnn/python_api_testing/sweep_tests/ttnn_ops.py

@@ -3707,3 +3707,22 @@ def rsqrt_bw(
     t2 = ttnn.rsqrt_bw(t0, t1, memory_config=output_mem_config)[0]
 
     return ttnn_tensor_to_torch(t2)
+
+
+def relu_bw(
+    x,
+    y,
+    *args,
+    device,
+    dtype,
+    layout,
+    input_mem_config,
+    output_mem_config,
+    **kwargs,
+):
+    t0 = setup_ttnn_tensor(x, device, layout[0], input_mem_config[0], dtype[0])
+    t1 = setup_ttnn_tensor(y, device, layout[1], input_mem_config[1], dtype[1])
+
+    t2 = ttnn.relu_bw(t0, t1, memory_config=output_mem_config)[0]
+
+    return ttnn_tensor_to_torch(t2)