├── .gitignore ├── Dockerfile.nvtorch ├── LICENSE ├── MODEL_CARD.md ├── README.md ├── download.sh ├── llama ├── __init__.py ├── custom_pretrain_llama.py ├── custom_training.py ├── llama_model.py ├── pretrain_llama.py ├── pretrain_llama2_distributed.sh ├── pretrain_llama_distributed.sh ├── tokenizer │ ├── sentencepiece_model.proto │ ├── sentencepiece_model_pb2.py │ └── tokenizer.model └── tools │ ├── preprocess_data.py │ ├── transform_huggingface_to_megatron.py │ └── transform_megatron_to_huggingface.py └── requirements.txt /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/.gitignore -------------------------------------------------------------------------------- /Dockerfile.nvtorch: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/Dockerfile.nvtorch -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/LICENSE -------------------------------------------------------------------------------- /MODEL_CARD.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/MODEL_CARD.md -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/README.md -------------------------------------------------------------------------------- /download.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/download.sh -------------------------------------------------------------------------------- /llama/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/__init__.py -------------------------------------------------------------------------------- /llama/custom_pretrain_llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/custom_pretrain_llama.py -------------------------------------------------------------------------------- /llama/custom_training.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/custom_training.py -------------------------------------------------------------------------------- /llama/llama_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/llama_model.py -------------------------------------------------------------------------------- /llama/pretrain_llama.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/pretrain_llama.py -------------------------------------------------------------------------------- /llama/pretrain_llama2_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/pretrain_llama2_distributed.sh -------------------------------------------------------------------------------- /llama/pretrain_llama_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/pretrain_llama_distributed.sh -------------------------------------------------------------------------------- /llama/tokenizer/sentencepiece_model.proto: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/tokenizer/sentencepiece_model.proto -------------------------------------------------------------------------------- /llama/tokenizer/sentencepiece_model_pb2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/tokenizer/sentencepiece_model_pb2.py -------------------------------------------------------------------------------- /llama/tokenizer/tokenizer.model: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/tokenizer/tokenizer.model -------------------------------------------------------------------------------- /llama/tools/preprocess_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/tools/preprocess_data.py -------------------------------------------------------------------------------- /llama/tools/transform_huggingface_to_megatron.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/tools/transform_huggingface_to_megatron.py -------------------------------------------------------------------------------- /llama/tools/transform_megatron_to_huggingface.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/llama/tools/transform_megatron_to_huggingface.py -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/MoFHeka/LLaMA-Megatron/HEAD/requirements.txt --------------------------------------------------------------------------------