sarman/dgx-spark-playbooks
1
0
Fork 0
mirror of https://github.com/NVIDIA/dgx-spark-playbooks.git synced 2026-04-22 18:13:52 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

chore: Regenerate all playbooks

Browse Source
This commit is contained in:
GitLab CI 2025-10-07 23:31:50 +00:00
parent 7874a7b269
commit e07330f8dc
2 changed files with 16 additions and 40 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
nvidia/llama-factory/README.md
View File

@ -85,7 +85,7 @@ Start the NVIDIA PyTorch container with GPU access and mount your workspace dire
> **Note:** This NVIDIA PyTorch container supports CUDA 13
```bash
docker run --gpus all --ipc=host --ulimit memlock=-1 -it --ulimit stack=67108864 --rm -v "$PWD":/workspace nvcr.io/nvidia/pytorch:25.08-py3 bash
docker run --gpus all --ipc=host --ulimit memlock=-1 -it --ulimit stack=67108864 --rm -v "$PWD":/workspace nvcr.io/nvidia/pytorch:25.09-py3 bash
```
## Step 3. Clone LLaMA Factory repository
@ -105,16 +105,7 @@ Install the package in editable mode with metrics support for training evaluatio
pip install -e ".[metrics]"
```
## Step 5. Configure PyTorch for CUDA 12.9 (skip if using Docker container from Step 2)
In a python virtual environment, uninstall existing PyTorch and reinstall with CUDA 12.9 support for ARM64 architecture.
```bash
pip uninstall torch torchvision torchaudio
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu129
```
*If using Docker container*
## Step 5. Verify Pytorch CUDA support.
PyTorch is pre-installed with CUDA support. Verify installation:
@ -158,7 +149,6 @@ Verify that training completed successfully and checkpoints were saved.
```bash
ls -la saves/llama3-8b/lora/sft/
cat saves/llama3-8b/lora/sft/training_loss.png
```
@ -170,13 +160,20 @@ Expected output should show:
## Step 9. Test inference with fine-tuned model
Run a simple inference test to verify the fine-tuned model loads correctly.
Test your fine-tuned model with custom prompts:
```bash
llamafactory-cli chat examples/inference/llama3_lora_sft.yaml
## Type: "Hello, how can you help me today?"
## Expect: Response showing fine-tuned behavior
```
## Step 10. Troubleshooting
## Step 10. For production deployment, export your model
```bash
llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml
```
## Step 11. Troubleshooting
| Symptom | Cause | Fix |
|---------|--------|-----|
@ -184,7 +181,7 @@ llamafactory-cli chat examples/inference/llama3_lora_sft.yaml
| Model download fails or is slow | Network connectivity or Hugging Face Hub issues | Check internet connection, try using `HF_HUB_OFFLINE=1` for cached models |
| Training loss not decreasing | Learning rate too high/low or insufficient data | Adjust `learning_rate` parameter or check dataset quality |
## Step 11. Cleanup and rollback
## Step 12. Cleanup and rollback
> **Warning:** This will delete all training progress and checkpoints.
@ -201,18 +198,3 @@ To rollback Docker container changes:
exit # Exit container
docker container prune -f
```
## Step 12. Next steps
Test your fine-tuned model with custom prompts:
```bash
llamafactory-cli chat examples/inference/llama3_lora_sft.yaml
## Type: "Hello, how can you help me today?"
## Expect: Response showing fine-tuned behavior
```
For production deployment, export your model:
```bash
llamafactory-cli export examples/merge_lora/llama3_lora_sft.yaml
```

14
nvidia/unsloth/README.md
View File

@ -76,12 +76,12 @@ The output should show a summary of GPU information.
## Step 2. Get the container image
```bash
docker pull nvcr.io/nvidia/pytorch:25.08-py3
docker pull nvcr.io/nvidia/pytorch:25.09-py3
```
## Step 3. Launch Docker
```bash
docker run --gpus all --ulimit memlock=-1 -it --ulimit stack=67108864 --entrypoint /usr/bin/bash --rm nvcr.io/nvidia/pytorch:25.08-py3
docker run --gpus all --ulimit memlock=-1 -it --ulimit stack=67108864 --entrypoint /usr/bin/bash --rm nvcr.io/nvidia/pytorch:25.09-py3
```
## Step 4. Install dependencies inside Docker
@ -93,13 +93,7 @@ pip install --no-deps unsloth unsloth_zoo
## Step 5. Build and install bitsandbytes inside Docker
```bash
git clone https://github.com/bitsandbytes-foundation/bitsandbytes.git
cd bitsandbytes
cmake -S . -B build -DCOMPUTE_BACKEND=cuda -DCOMPUTE_CAPABILITY="80;86;87;89;90"
cd build
make -j
cd ..
pip install .
pip install --no-deps bitsandbytes
```
## Step 6. Create Python test script
@ -107,8 +101,8 @@ pip install .
Curl the test script [here](https://gitlab.com/nvidia/dgx-spark/temp-external-playbook-assets/dgx-spark-playbook-assets/-/blob/main/${MODEL}/assets/test_unsloth.py) into the container.
```bash
curl -O https://gitlab.com/nvidia/dgx-spark/temp-external-playbook-assets/dgx-spark-playbook-assets/-/blob/main/${MODEL}/assets/test_unsloth.py
```
We will use this test script to validate the installation with a simple fine-tuning task.