From eaf111a15755c874ca20e6f43e633b5b7077c9c8 Mon Sep 17 00:00:00 2001 From: Yiren Lu Date: Tue, 26 Nov 2024 15:35:49 -0500 Subject: [PATCH 1/3] Update examples titles for SEO --- 05_scheduling/hackernews_alerts.py | 2 +- 06_gpu_and_ml/llm-serving/vllm_inference.py | 2 +- 06_gpu_and_ml/llm-structured/jsonformer_generate.py | 2 +- 06_gpu_and_ml/obj_detection_webcam/webcam.py | 2 +- 06_gpu_and_ml/stable_diffusion/flux.py | 2 +- 06_gpu_and_ml/text-to-video/mochi.py | 2 +- 06_gpu_and_ml/yolo/finetune_yolo.py | 2 +- 7 files changed, 7 insertions(+), 7 deletions(-) diff --git a/05_scheduling/hackernews_alerts.py b/05_scheduling/hackernews_alerts.py index 357bc8c3d..88e79d6d8 100644 --- a/05_scheduling/hackernews_alerts.py +++ b/05_scheduling/hackernews_alerts.py @@ -2,7 +2,7 @@ # lambda-test: false # tags: ["featured"] # --- -# # Hacker News Slackbot +# # Cron job in the Cloud to search Hacker News # In this example, we use Modal to deploy a cron job that periodically queries Hacker News for # new posts matching a given search term, and posts the results to Slack. diff --git a/06_gpu_and_ml/llm-serving/vllm_inference.py b/06_gpu_and_ml/llm-serving/vllm_inference.py index 2023281ca..30729ba71 100644 --- a/06_gpu_and_ml/llm-serving/vllm_inference.py +++ b/06_gpu_and_ml/llm-serving/vllm_inference.py @@ -5,7 +5,7 @@ # tags: ["use-case-lm-inference", "featured"] # --- -# # Run an OpenAI-Compatible vLLM Server +# # Fast LLM inference with Llama 3.1-8B and vLLM # LLMs do more than just model language: they chat, they produce JSON and XML, they run code, and more. # This has complicated their interface far beyond "text-in, text-out". diff --git a/06_gpu_and_ml/llm-structured/jsonformer_generate.py b/06_gpu_and_ml/llm-structured/jsonformer_generate.py index 1295d5fa4..79eb4cc4a 100644 --- a/06_gpu_and_ml/llm-structured/jsonformer_generate.py +++ b/06_gpu_and_ml/llm-structured/jsonformer_generate.py @@ -1,7 +1,7 @@ # --- # lambda-test: false # --- -# # Generate synthetic data with Jsonformer +# # Structured output generation with JSONFormer # # [Jsonformer](https://github.com/1rgs/jsonformer) is a tool that generates structured synthetic data using LLMs. # You provide a JSON spec and it generates a JSON object following the spec. It's a diff --git a/06_gpu_and_ml/obj_detection_webcam/webcam.py b/06_gpu_and_ml/obj_detection_webcam/webcam.py index e94c3f4ed..b2296570a 100644 --- a/06_gpu_and_ml/obj_detection_webcam/webcam.py +++ b/06_gpu_and_ml/obj_detection_webcam/webcam.py @@ -3,7 +3,7 @@ # deploy: true # tags: ["use-case-image-video-3d", "featured"] # --- -# # Machine learning model inference endpoint that uses the webcam +# # Real-time object detection via webcam # # This example creates a web endpoint that uses a Huggingface model for object detection. # diff --git a/06_gpu_and_ml/stable_diffusion/flux.py b/06_gpu_and_ml/stable_diffusion/flux.py index 8c0dc4e92..22ac3f546 100644 --- a/06_gpu_and_ml/stable_diffusion/flux.py +++ b/06_gpu_and_ml/stable_diffusion/flux.py @@ -4,7 +4,7 @@ # tags: ["use-case-image-video-3d", "featured"] # --- -# # Run Flux fast with `torch.compile` on Hopper GPUs +# # Run Flux fast on H100s # In this guide, we'll run Flux as fast as possible on Modal using open source tools. # We'll use `torch.compile` and NVIDIA H100 GPUs. diff --git a/06_gpu_and_ml/text-to-video/mochi.py b/06_gpu_and_ml/text-to-video/mochi.py index bc0365371..e59f8a260 100644 --- a/06_gpu_and_ml/text-to-video/mochi.py +++ b/06_gpu_and_ml/text-to-video/mochi.py @@ -2,7 +2,7 @@ # cmd: ["modal", "run", "--detach", "06_gpu_and_ml/text-to-video/mochi.py", "--num-inference-steps", "64"] # --- -# # Generate videos from text prompts with Mochi +# # Text-to-video AI with Mochi # This example demonstrates how to run the [Mochi 1](https://github.com/genmoai/models) # video generation model by [Genmo](https://www.genmo.ai/) on Modal. diff --git a/06_gpu_and_ml/yolo/finetune_yolo.py b/06_gpu_and_ml/yolo/finetune_yolo.py index 985b9ada4..4a23d4795 100644 --- a/06_gpu_and_ml/yolo/finetune_yolo.py +++ b/06_gpu_and_ml/yolo/finetune_yolo.py @@ -2,7 +2,7 @@ # args: ["--no-quick-check"] # tags: ["use-case-image-video-3d", "use-case-finetuning"] # --- -# # Fine-Tuning and Inference for Computer Vision with YOLO +# # Object detection with YOLO open source model # # Example by [@Erik-Dunteman](https://github.com/erik-dunteman) and [@AnirudhRahul](https://github.com/AnirudhRahul/). From 5877c06f8d0afbdcd2fc556887d4a733b6da85b5 Mon Sep 17 00:00:00 2001 From: Charles Frye Date: Tue, 26 Nov 2024 14:02:36 -0800 Subject: [PATCH 2/3] Apply suggestions from review --- 05_scheduling/hackernews_alerts.py | 2 +- 06_gpu_and_ml/llm-structured/jsonformer_generate.py | 2 +- 06_gpu_and_ml/stable_diffusion/flux.py | 2 +- 06_gpu_and_ml/text-to-video/mochi.py | 2 +- 06_gpu_and_ml/yolo/finetune_yolo.py | 2 +- 5 files changed, 5 insertions(+), 5 deletions(-) diff --git a/05_scheduling/hackernews_alerts.py b/05_scheduling/hackernews_alerts.py index 88e79d6d8..f6e7bdee6 100644 --- a/05_scheduling/hackernews_alerts.py +++ b/05_scheduling/hackernews_alerts.py @@ -2,7 +2,7 @@ # lambda-test: false # tags: ["featured"] # --- -# # Cron job in the Cloud to search Hacker News +# # Run cron jobs in the cloud to search Hacker News # In this example, we use Modal to deploy a cron job that periodically queries Hacker News for # new posts matching a given search term, and posts the results to Slack. diff --git a/06_gpu_and_ml/llm-structured/jsonformer_generate.py b/06_gpu_and_ml/llm-structured/jsonformer_generate.py index 79eb4cc4a..39ff324e2 100644 --- a/06_gpu_and_ml/llm-structured/jsonformer_generate.py +++ b/06_gpu_and_ml/llm-structured/jsonformer_generate.py @@ -1,7 +1,7 @@ # --- # lambda-test: false # --- -# # Structured output generation with JSONFormer +# # Structured output generation with Jsonformer # # [Jsonformer](https://github.com/1rgs/jsonformer) is a tool that generates structured synthetic data using LLMs. # You provide a JSON spec and it generates a JSON object following the spec. It's a diff --git a/06_gpu_and_ml/stable_diffusion/flux.py b/06_gpu_and_ml/stable_diffusion/flux.py index 22ac3f546..649584c0a 100644 --- a/06_gpu_and_ml/stable_diffusion/flux.py +++ b/06_gpu_and_ml/stable_diffusion/flux.py @@ -4,7 +4,7 @@ # tags: ["use-case-image-video-3d", "featured"] # --- -# # Run Flux fast on H100s +# # Run Flux fast on H100s with `torch.compile` # In this guide, we'll run Flux as fast as possible on Modal using open source tools. # We'll use `torch.compile` and NVIDIA H100 GPUs. diff --git a/06_gpu_and_ml/text-to-video/mochi.py b/06_gpu_and_ml/text-to-video/mochi.py index e59f8a260..a8dda2a56 100644 --- a/06_gpu_and_ml/text-to-video/mochi.py +++ b/06_gpu_and_ml/text-to-video/mochi.py @@ -2,7 +2,7 @@ # cmd: ["modal", "run", "--detach", "06_gpu_and_ml/text-to-video/mochi.py", "--num-inference-steps", "64"] # --- -# # Text-to-video AI with Mochi +# # Text-to-video generation with Mochi # This example demonstrates how to run the [Mochi 1](https://github.com/genmoai/models) # video generation model by [Genmo](https://www.genmo.ai/) on Modal. diff --git a/06_gpu_and_ml/yolo/finetune_yolo.py b/06_gpu_and_ml/yolo/finetune_yolo.py index 4a23d4795..473d9d7bf 100644 --- a/06_gpu_and_ml/yolo/finetune_yolo.py +++ b/06_gpu_and_ml/yolo/finetune_yolo.py @@ -2,7 +2,7 @@ # args: ["--no-quick-check"] # tags: ["use-case-image-video-3d", "use-case-finetuning"] # --- -# # Object detection with YOLO open source model +# # Fine-tune open source YOLO models for object detection # # Example by [@Erik-Dunteman](https://github.com/erik-dunteman) and [@AnirudhRahul](https://github.com/AnirudhRahul/). From c80dd78ae1293d74e913af25da6dc0205a2a53bf Mon Sep 17 00:00:00 2001 From: Charles Frye Date: Tue, 26 Nov 2024 14:03:31 -0800 Subject: [PATCH 3/3] add another suggestion from review --- 06_gpu_and_ml/llm-serving/vllm_inference.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/06_gpu_and_ml/llm-serving/vllm_inference.py b/06_gpu_and_ml/llm-serving/vllm_inference.py index 30729ba71..c3b4ebe08 100644 --- a/06_gpu_and_ml/llm-serving/vllm_inference.py +++ b/06_gpu_and_ml/llm-serving/vllm_inference.py @@ -5,7 +5,7 @@ # tags: ["use-case-lm-inference", "featured"] # --- -# # Fast LLM inference with Llama 3.1-8B and vLLM +# # Run OpenAI-compatible LLM inference with LLaMA 3.1-8B and vLLM # LLMs do more than just model language: they chat, they produce JSON and XML, they run code, and more. # This has complicated their interface far beyond "text-in, text-out".