├── README.md ├── README_EN.md ├── assets ├── Llama4-Maverick.png ├── base_eval.png ├── ceval.jpg ├── llama.jpg ├── llama.png ├── llama2-chinese-webui.jpg ├── llama3_eval.png ├── llama_eval.jpeg ├── meta_eval_13B.md ├── meta_eval_7B.md ├── tuned_eval.png ├── wechat-new.jpeg └── wechat.jpeg ├── data ├── dev_sft.csv ├── dev_sft_sharegpt.csv └── train_sft.csv ├── docker ├── Dockerfile ├── Dockerfile_train └── docker-compose.yml ├── docs ├── chat_gradio_guide.md └── inference_speed_guide.md ├── examples ├── chat_gradio.py ├── chat_gradio_no_merge.py └── llama2_for_langchain.py ├── inference-speed ├── CPU │ └── ggml │ │ └── README.md └── GPU │ ├── FasterTransformer_example │ └── README.md │ ├── JittorLLMs_example │ └── README.md │ ├── TensorRT-LLM_example │ ├── README.md │ ├── atom_inference.py │ └── utils.py │ ├── lmdeploy_example │ ├── README.md │ └── test_api_server.py │ └── vllm_example │ ├── README.md │ ├── api_server.py │ ├── client_test.py │ ├── multi_gpus_api_server.sh │ └── single_gpu_api_server.sh ├── requirements.txt ├── scripts ├── api │ ├── README.md │ ├── accelerate_client.py │ └── accelerate_server.py ├── convert2hf │ ├── README.md │ └── convert_llama_weights_to_hf.py └── test_model │ └── test_pretrain_model.ipynb └── train ├── merge_peft_model ├── merge.sh ├── merge_muilt.sh ├── merge_muilt_peft_adapter.py └── merge_peft_adapter.py ├── pretrain ├── accuracy.py ├── ds_config_zero2.json ├── ds_config_zero3.json ├── pretrain.sh └── pretrain_clm.py └── sft ├── accuracy.py ├── ds_config_zero2.json ├── finetune.sh ├── finetune_clm.py ├── finetune_clm_lora.py └── finetune_lora.sh /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/README.md -------------------------------------------------------------------------------- /README_EN.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/README_EN.md -------------------------------------------------------------------------------- /assets/Llama4-Maverick.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/Llama4-Maverick.png -------------------------------------------------------------------------------- /assets/base_eval.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/base_eval.png -------------------------------------------------------------------------------- /assets/ceval.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/ceval.jpg -------------------------------------------------------------------------------- /assets/llama.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/llama.jpg -------------------------------------------------------------------------------- /assets/llama.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/llama.png -------------------------------------------------------------------------------- /assets/llama2-chinese-webui.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/llama2-chinese-webui.jpg -------------------------------------------------------------------------------- /assets/llama3_eval.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/llama3_eval.png -------------------------------------------------------------------------------- /assets/llama_eval.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/llama_eval.jpeg -------------------------------------------------------------------------------- /assets/meta_eval_13B.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/meta_eval_13B.md -------------------------------------------------------------------------------- /assets/meta_eval_7B.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/meta_eval_7B.md -------------------------------------------------------------------------------- /assets/tuned_eval.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/tuned_eval.png -------------------------------------------------------------------------------- /assets/wechat-new.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/wechat-new.jpeg -------------------------------------------------------------------------------- /assets/wechat.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/assets/wechat.jpeg -------------------------------------------------------------------------------- /data/dev_sft.csv: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/data/dev_sft.csv -------------------------------------------------------------------------------- /data/dev_sft_sharegpt.csv: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/data/dev_sft_sharegpt.csv -------------------------------------------------------------------------------- /data/train_sft.csv: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/data/train_sft.csv -------------------------------------------------------------------------------- /docker/Dockerfile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/docker/Dockerfile -------------------------------------------------------------------------------- /docker/Dockerfile_train: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/docker/Dockerfile_train -------------------------------------------------------------------------------- /docker/docker-compose.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/docker/docker-compose.yml -------------------------------------------------------------------------------- /docs/chat_gradio_guide.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/docs/chat_gradio_guide.md -------------------------------------------------------------------------------- /docs/inference_speed_guide.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/docs/inference_speed_guide.md -------------------------------------------------------------------------------- /examples/chat_gradio.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/examples/chat_gradio.py -------------------------------------------------------------------------------- /examples/chat_gradio_no_merge.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/examples/chat_gradio_no_merge.py -------------------------------------------------------------------------------- /examples/llama2_for_langchain.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/examples/llama2_for_langchain.py -------------------------------------------------------------------------------- /inference-speed/CPU/ggml/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/CPU/ggml/README.md -------------------------------------------------------------------------------- /inference-speed/GPU/FasterTransformer_example/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/FasterTransformer_example/README.md -------------------------------------------------------------------------------- /inference-speed/GPU/JittorLLMs_example/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/JittorLLMs_example/README.md -------------------------------------------------------------------------------- /inference-speed/GPU/TensorRT-LLM_example/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/TensorRT-LLM_example/README.md -------------------------------------------------------------------------------- /inference-speed/GPU/TensorRT-LLM_example/atom_inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/TensorRT-LLM_example/atom_inference.py -------------------------------------------------------------------------------- /inference-speed/GPU/TensorRT-LLM_example/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/TensorRT-LLM_example/utils.py -------------------------------------------------------------------------------- /inference-speed/GPU/lmdeploy_example/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/lmdeploy_example/README.md -------------------------------------------------------------------------------- /inference-speed/GPU/lmdeploy_example/test_api_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/lmdeploy_example/test_api_server.py -------------------------------------------------------------------------------- /inference-speed/GPU/vllm_example/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/vllm_example/README.md -------------------------------------------------------------------------------- /inference-speed/GPU/vllm_example/api_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/vllm_example/api_server.py -------------------------------------------------------------------------------- /inference-speed/GPU/vllm_example/client_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/vllm_example/client_test.py -------------------------------------------------------------------------------- /inference-speed/GPU/vllm_example/multi_gpus_api_server.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/vllm_example/multi_gpus_api_server.sh -------------------------------------------------------------------------------- /inference-speed/GPU/vllm_example/single_gpu_api_server.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/inference-speed/GPU/vllm_example/single_gpu_api_server.sh -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/requirements.txt -------------------------------------------------------------------------------- /scripts/api/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/scripts/api/README.md -------------------------------------------------------------------------------- /scripts/api/accelerate_client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/scripts/api/accelerate_client.py -------------------------------------------------------------------------------- /scripts/api/accelerate_server.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/scripts/api/accelerate_server.py -------------------------------------------------------------------------------- /scripts/convert2hf/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/scripts/convert2hf/README.md -------------------------------------------------------------------------------- /scripts/convert2hf/convert_llama_weights_to_hf.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/scripts/convert2hf/convert_llama_weights_to_hf.py -------------------------------------------------------------------------------- /scripts/test_model/test_pretrain_model.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/scripts/test_model/test_pretrain_model.ipynb -------------------------------------------------------------------------------- /train/merge_peft_model/merge.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/merge_peft_model/merge.sh -------------------------------------------------------------------------------- /train/merge_peft_model/merge_muilt.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/merge_peft_model/merge_muilt.sh -------------------------------------------------------------------------------- /train/merge_peft_model/merge_muilt_peft_adapter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/merge_peft_model/merge_muilt_peft_adapter.py -------------------------------------------------------------------------------- /train/merge_peft_model/merge_peft_adapter.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/merge_peft_model/merge_peft_adapter.py -------------------------------------------------------------------------------- /train/pretrain/accuracy.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/pretrain/accuracy.py -------------------------------------------------------------------------------- /train/pretrain/ds_config_zero2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/pretrain/ds_config_zero2.json -------------------------------------------------------------------------------- /train/pretrain/ds_config_zero3.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/pretrain/ds_config_zero3.json -------------------------------------------------------------------------------- /train/pretrain/pretrain.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/pretrain/pretrain.sh -------------------------------------------------------------------------------- /train/pretrain/pretrain_clm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/pretrain/pretrain_clm.py -------------------------------------------------------------------------------- /train/sft/accuracy.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/sft/accuracy.py -------------------------------------------------------------------------------- /train/sft/ds_config_zero2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/sft/ds_config_zero2.json -------------------------------------------------------------------------------- /train/sft/finetune.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/sft/finetune.sh -------------------------------------------------------------------------------- /train/sft/finetune_clm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/sft/finetune_clm.py -------------------------------------------------------------------------------- /train/sft/finetune_clm_lora.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/sft/finetune_clm_lora.py -------------------------------------------------------------------------------- /train/sft/finetune_lora.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/LlamaFamily/Llama-Chinese/HEAD/train/sft/finetune_lora.sh --------------------------------------------------------------------------------