From 0d0c65eda74c93b36cd515f22a0a5d5461309a75 Mon Sep 17 00:00:00 2001 From: Ubuntu Date: Tue, 2 Dec 2025 17:52:13 +0000 Subject: [PATCH 1/2] Add neuronx-cc installation cell before compilation Added pip install command for neuronx-cc==2.21.* to ensure the correct compiler version is available before the model compilation step. --- .../01_finetuning/Finetune-Qwen3-1.7B.ipynb | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/labs/FineTuning/HuggingFaceExample/01_finetuning/Finetune-Qwen3-1.7B.ipynb b/labs/FineTuning/HuggingFaceExample/01_finetuning/Finetune-Qwen3-1.7B.ipynb index cd2ada0..c62bf6f 100644 --- a/labs/FineTuning/HuggingFaceExample/01_finetuning/Finetune-Qwen3-1.7B.ipynb +++ b/labs/FineTuning/HuggingFaceExample/01_finetuning/Finetune-Qwen3-1.7B.ipynb @@ -105,6 +105,13 @@ "Neuron compilation optimizes the model graph and converts it into a Neuron Executable File Format (NEFF), enabling efficient execution on NeuronCores." ] }, + { + "cell_type": "code", + "source": "!pip install --upgrade neuronx-cc==2.21.* --extra-index-url=https://pip.repos.neuron.amazonaws.com", + "metadata": {}, + "execution_count": null, + "outputs": [] + }, { "cell_type": "code", "execution_count": null, @@ -219,4 +226,4 @@ }, "nbformat": 4, "nbformat_minor": 2 -} +} \ No newline at end of file From 67e1f7bbb5b3bb6b482f4ac8bc2a0e3e45759b9d Mon Sep 17 00:00:00 2001 From: Ubuntu Date: Tue, 2 Dec 2025 18:09:34 +0000 Subject: [PATCH 2/2] Add instructions for testing fine-tuned model in Chess-Deployment Added Step 4.5 with instructions on how to configure env.example to test participants' own fine-tuned models against Stockfish. Includes clear guidance on when to make this change and step-by-step instructions. --- labs/vLLM/Chess/Chess-Deployment.ipynb | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/labs/vLLM/Chess/Chess-Deployment.ipynb b/labs/vLLM/Chess/Chess-Deployment.ipynb index 183b986..9635d38 100644 --- a/labs/vLLM/Chess/Chess-Deployment.ipynb +++ b/labs/vLLM/Chess/Chess-Deployment.ipynb @@ -146,6 +146,11 @@ "MODEL_PATH=PRECOMPILED_MODEL_PATH" ] }, + { + "cell_type": "markdown", + "source": "## Step 4.5: Configure Environment for Testing Your Fine-Tuned Model (Optional)\n\nIf you want to test your own fine-tuned model against Stockfish later in this notebook, you'll need to update the environment configuration file.\n\n**When to do this:**\n- After you've confirmed your fine-tuned model exists at `/home/ubuntu/environment/ml/qwen-chess/compiled_model`\n- Before running the Stockfish games in Step 8\n\n**How to configure:**\n\n1. Open the environment configuration file:\n ```bash\n # From the terminal or a new cell:\n nano /home/ubuntu/environment/vLLM/Chess/assets/env.example\n ```\n\n2. Update the `VLLM_MODEL` line to point to your compiled model:\n ```bash\n # Change this line:\n VLLM_MODEL=kunhunjon/ChessLM_Qwen3_Trainium_AWS_Format\n \n # To this:\n VLLM_MODEL=/home/ubuntu/environment/ml/qwen-chess/compiled_model\n ```\n\n3. Save the file (in nano: Ctrl+O, Enter, Ctrl+X)\n\n**Note:** You can skip this step for now if you want to use the pre-compiled model first. You can always come back and update the configuration later to test your own fine-tuned model.", + "metadata": {} + }, { "cell_type": "markdown", "metadata": {}, @@ -553,4 +558,4 @@ }, "nbformat": 4, "nbformat_minor": 4 -} +} \ No newline at end of file