├── .DS_Store ├── .gitignore ├── LICENSE ├── README.md ├── ai-compiler ├── .DS_Store ├── README.md ├── Treebeard │ ├── .DS_Store │ ├── 10-TableI-1.png │ ├── 10-TableII-1.png │ ├── 9-Figure6-1.png │ ├── README.md │ ├── TREEBEARD IR lowering 和 optimization 细节.png │ └── TREEBEARD编译器结构.png ├── treelit │ ├── README.md │ └── xgb.md └── triton-lang │ └── README.md ├── ai-framework ├── .DS_Store ├── README.md ├── TensorRT-Model-Optimizer.md ├── cuda │ └── README.md ├── deepspeed │ ├── 1.DeepSpeed入门.md │ ├── 2.安装DeepSpeed.md │ ├── 3.基于CIFAR-10使用DeepSpeed进行分布式训练 .md │ ├── DeepSpeed配置JSON文件.md │ ├── README.md │ ├── config-json │ │ ├── README.md │ │ └── deepspeed-nvme.md │ ├── deepspeed-slurm.md │ ├── hello_bert │ │ ├── README.md │ │ ├── train_bert.py │ │ └── train_bert_ds.py │ └── training │ │ └── pipeline_parallelism │ │ └── README.md ├── dlrover.md ├── huggingface-accelerate │ └── README.md ├── huggingface-peft │ └── README.md ├── huggingface-transformers │ ├── API.md │ ├── FSDP.md │ └── README.md ├── huggingface-trl │ └── README.md ├── jax │ ├── README.md │ └── reference.md ├── llama-cpp │ └── README.md ├── megatron-deepspeed │ └── README.md ├── megatron-lm │ └── README.md ├── mxnet │ ├── README.md │ ├── mnist.py │ ├── mxnet_cnn_mnist.py │ ├── mxnet_mlp_mnist.py │ ├── oneflow_cnn_mnist.py │ ├── oneflow_mlp_mnist.py │ └── reference.md ├── oneflow │ ├── README.md │ ├── oneflow_mlp_mnist.py │ └── reference.md ├── openai-triton │ └── README.md ├── paddlepaddle │ ├── README.md │ └── reference.md ├── pai-megatron-patch │ ├── .DS_Store │ └── README.md ├── pai-torchacc.md ├── pytorch │ ├── README.md │ ├── install.md │ └── reference.md ├── tensorflow │ ├── README.md │ └── reference.md ├── transformer-engine │ └── mnist │ │ ├── README.md │ │ ├── main.py │ │ └── main_stat.py └── unsloth-微调.md ├── ai-infra ├── .DS_Store ├── ai-cluster │ └── README.md ├── ai-hardware │ ├── .DS_Store │ ├── AI芯片软件生态.md │ ├── CUDA.md │ ├── GPU-network.md │ ├── GPU相关环节变量.md │ ├── NIXL.md │ ├── OEM-DGX.md │ ├── README.md │ ├── TSMC-台积电.md │ ├── cuda镜像.md │ ├── gpudirect.md │ └── 硬件对比.md ├── communication.md ├── 存储 │ ├── README.md │ ├── REF.md │ ├── nvme-ssd.md │ ├── 固态硬盘.md │ └── 存储.md ├── 算力 │ ├── AI芯片.md │ ├── GPU工作原理.md │ ├── NVIDIA-GPU型号.md │ ├── 推理芯片.md │ └── 昇腾NPU.md └── 网络 │ ├── .DS_Store │ ├── HPC性能测试.md │ ├── IB-docker.md │ ├── IB流量监控.md │ ├── IB软件.md │ ├── InfiniBand.md │ ├── NCCL.md │ ├── README.md │ ├── REF.md │ ├── Spine-Leaf和InfiniBand网络架构区别简述.md │ ├── nccl-test-集合通讯的性能测试.md │ ├── nvbandwidth.md │ ├── pic │ ├── .DS_Store │ ├── 8卡V100的混合网络拓扑.png │ ├── A800-H100-H800.jpeg │ ├── NVLink-generations比较.png │ ├── PCIe-Generation对比.png │ ├── nvidia-dgx-1-v100-nvlink-gpu-xeon-config.webp │ ├── nvidia-dgx-1-with-volta.webp │ ├── nvlink.png │ ├── nvlink性能.png │ ├── nvswitch.png │ ├── 支持 NVLink GPU 之间连接的 NVIDIA H100-1.png │ ├── 支持 NVLink GPU 之间连接的 NVIDIA H100-2.png │ ├── 服务器之间的nvlink与nvswitch.png │ ├── 网络之间的连接.png │ └── 英伟达A100-A800-H100-H800.jpeg │ ├── ringallreduce │ ├── All Gather 流程图.gif │ └── Scatter Reduce 流程图.gif │ ├── roce.md │ ├── 网络硬件.md │ ├── 通信软件.md │ └── 集合通信原语.md ├── blog ├── .DS_Store ├── TODO.md ├── ai-infra │ ├── AI 集群基础设施 InfiniBand 详解.md │ └── AI 集群基础设施 NVMe SSD 详解.md ├── distribution-parallelism │ ├── 大模型分布式训练并行技术（一）-概述.md │ ├── 大模型分布式训练并行技术（九）-总结.md │ └── 大模型分布式训练并行技术（六）-多维混合并行.md ├── llm-algo │ ├── moe.md │ └── 大白话Transformer架构.md ├── llm-compression │ ├── 大模型量化技术原理-ZeroQuant系列.md │ └── 大模型量化技术原理：QoQ量化及QServe推理服务系统.md ├── llm-inference │ └── 大模型推理框架概述.md ├── llm-localization │ ├── 大模型国产化适配1-华为昇腾AI全栈软硬件平台总结.md │ └── 大模型国产化适配4-基于昇腾910使用LLaMA-13B进行多机多卡训练.md ├── llm-peft │ ├── 大模型参数高效微调技术原理综述（一）-背景、参数高效微调简介.md │ └── 大模型参数高效微调技术原理综述（五）-LoRA、AdaLoRA、QLoRA.md └── reference │ └── 高性能 LLM 推理框架的设计与实现.md ├── docs ├── .DS_Store ├── README.md ├── conda.md ├── flash-attention │ └── FlashAttention.md ├── llm-base │ ├── .DS_Store │ ├── FLOPS.md │ ├── NVIDIA-Nsight-Systems性能分析.md │ ├── README.md │ ├── a800-env-install.md │ ├── ai-algo.md │ ├── autoregressive-lm-decoding-methods.md │ ├── dcgmi.md │ ├── distribution-parallelism │ │ ├── .DS_Store │ │ ├── README.md │ │ ├── auto-parallel │ │ │ ├── Alpa.md │ │ │ ├── Flexflow.md │ │ │ ├── Galvatron.md │ │ │ ├── Mesh-Tensorflow.md │ │ │ ├── README.md │ │ │ ├── Unity.md │ │ │ ├── auto-parallel.md │ │ │ ├── gspmd.md │ │ │ ├── 分布式训练自动并行概述.md │ │ │ └── 飞桨面向异构场景下的自动并行设计与实践.md │ │ ├── data-parallelism │ │ │ ├── README.md │ │ │ └── ddp │ │ │ │ ├── Gradient Bucketing 示意图.webp │ │ │ │ └── 代码架构.webp │ │ ├── moe-parallel │ │ │ ├── README.md │ │ │ ├── moe-framework.md │ │ │ ├── moe-parallel.md │ │ │ └── paddle_moe.py │ │ ├── multidimensional-hybrid-parallel │ │ │ ├── BloombergGPT模型超参数.png │ │ │ └── README.md │ │ ├── pipeline-parallelism │ │ │ └── README.md │ │ ├── tensor-parallel │ │ │ ├── README.md │ │ │ └── tensor-parallel.md │ │ └── 并行技术.drawio │ ├── distribution-training │ │ ├── .DS_Store │ │ ├── Bloom-176B训练经验.md │ │ ├── FP16-BF16.md │ │ ├── GLM-130B训练经验.md │ │ ├── OPT-175B训练经验.md │ │ ├── README.md │ │ └── 自动混合精度.md │ ├── gpu-env-var.md │ ├── h800-env-install.md │ ├── images │ │ └── slurm │ │ │ └── slurm.gif │ ├── monitor.md │ ├── multimodal │ │ └── sora.md │ ├── nvidia-smi-dmon.md │ ├── nvidia-smi.md │ ├── rlhf │ │ └── README.md │ ├── scenes │ │ ├── README.md │ │ ├── cv │ │ │ ├── README.md │ │ │ ├── paddle │ │ │ │ └── README.md │ │ │ ├── pytorch │ │ │ │ └── README.md │ │ │ └── reference.md │ │ └── multi-modal │ │ │ ├── README.md │ │ │ └── reference.md │ ├── singularity命令.md │ ├── slurm.md │ ├── 分布式训练加速技术.md │ ├── 多机RDMA性能测试.txt │ └── 机器学习中常用的数据类型.md ├── llm-experience.md ├── llm-inference │ ├── DeepSpeed-Inference.md │ ├── KV-Cache.md │ ├── LLM服务框架对比.md │ ├── README.md │ ├── blog.md │ ├── flexflow │ │ ├── spec_infer_demo.gif │ │ └── 投机采样.md │ ├── llm推理优化技术.md │ ├── llm推理框架.md │ └── vllm.md ├── llm-peft │ ├── LoRA-FA.md │ ├── MAM_Adapter.md │ ├── README.md │ └── ReLoRA.md ├── llm-summarize │ ├── README.md │ ├── distribution_dl_roadmap.md │ ├── pic │ │ ├── A800.jpeg │ │ ├── H800.jpeg │ │ ├── transformer架构.jpg │ │ └── why_RLHF.jpg │ ├── 大模型实践总结-20230930.md │ ├── 大模型实践总结.md │ ├── 文档大模型.md │ ├── 金融大模型.md │ └── 领域大模型.md └── transformer内存估算.md ├── faq └── FAQ.md ├── git-pull-push.sh ├── llm-algo ├── .DS_Store ├── FLOPs.md ├── InternLM-20B.md ├── README.md ├── baichuan2 │ └── baichuan.md ├── bert.md ├── bert │ └── 模型架构.md ├── bloom.md ├── bloom │ └── README.md ├── chatglm │ ├── GLM说明.png │ ├── GLM预训练.png │ ├── README.md │ └── 模型架构.md ├── chatglm2 │ ├── README.md │ └── 模型架构.md ├── chatglm3 │ ├── README.md │ └── reference.md ├── chatgpt │ └── README.md ├── deepseek │ ├── DeepSeek-R1.md │ ├── DeepSeek-V2.md │ ├── DeepSeek-V3.md │ └── README.md ├── glm-130b │ ├── README.md │ └── 模型架构.gif ├── glm4.md ├── gpt │ ├── README.md │ └── 模型结构.png ├── gpt2 │ ├── README.md │ ├── hf_modeling_gpt2.py │ └── 模型架构.md ├── gpt3 │ └── README.md ├── llama.md ├── llama │ ├── README.md │ └── 模型架构.md ├── mixtral │ └── README.md ├── mlp.md ├── moe │ └── README.md ├── qwen │ ├── README.md │ └── 参数说明及函数说明.md ├── qwen2.md ├── t5 │ └── README.md ├── transformer.md ├── transformer │ ├── README.md │ ├── Transformer中FFN的记忆功能.md │ ├── multi-head-attention.webp │ ├── transformer-building-blocks.webp │ ├── w-qkv.png │ └── 模型架构.md ├── 基本概念.md ├── 旋转编码RoPE.md ├── 模型架构类图.drawio └── 训练范式.md ├── llm-alignment ├── DPO.md ├── README.md ├── RLHF.md └── 基本概念.md ├── llm-application ├── .DS_Store ├── Higress.md ├── README.md ├── embbedding-model.md ├── gradio │ └── README.md ├── langchain │ ├── .DS_Store │ ├── README.md │ ├── serve.py │ └── tutorials │ │ ├── client.py │ │ └── serve.py ├── one-api.md ├── pre-post-handle │ └── README.md ├── rag │ ├── README.md │ ├── embedding.md │ ├── 存在的一些问题.md │ └── 方案.md ├── vector-db │ ├── README.md │ └── reference.md └── 应用场景.md ├── llm-compression ├── .DS_Store ├── PaddleSlim │ ├── quantization.md │ └── README.md ├── README.md ├── distillation │ ├── GKD.md │ ├── MINILLM.md │ ├── README.md │ ├── SCOTT.md │ └── 大模型蒸馏概述.md ├── gptqmodel │ └── README.md ├── llm-compressor │ ├── README.md │ ├── source-code.md │ ├── 剪枝.md │ └── 量化方案.md ├── quantization │ ├── .DS_Store │ ├── FP6-LLM.md │ ├── GPTQ.md │ ├── LLM-int8.md │ ├── PEQA.md │ ├── QQQ-W4A8.md │ ├── README.md │ ├── SmoothQuant.md │ ├── SpinQuant.md │ ├── ZeroQuant(4+2).md │ ├── ZeroQuant.md │ ├── fp4.md │ ├── fp6.md │ ├── fp8.md │ ├── kv-cache-quant.md │ ├── llm-qat │ │ ├── LLM-QAT.md │ │ ├── README.md │ │ ├── cfd70ff │ │ │ ├── README.md │ │ │ ├── generate_data.py │ │ │ ├── inference.py │ │ │ ├── merge_gen_data.py │ │ │ ├── pip.conf │ │ │ ├── run_train.sh │ │ │ ├── train.py │ │ │ └── utils.py │ │ ├── f4d873a │ │ │ ├── datautils.py │ │ │ ├── run_train.sh │ │ │ └── train.py │ │ └── log.md │ ├── moe模型量化.md │ ├── tools.md │ ├── 可视化 │ │ ├── README.md │ │ ├── qwen_activate_visual.ipynb │ │ └── qwen_visual.ipynb │ ├── 大模型量化概述.md │ └── 量化基础.md ├── sparsity │ └── README.md ├── tools.md ├── 大模型压缩综述.md └── 经验.md ├── llm-data-engineering ├── README.md ├── dataset │ ├── README.md │ ├── baichuan2.md │ ├── chinese-corpus-all.md │ └── english-corpus-all.md ├── reference.md └── sft-dataset │ ├── baichuan2_test.py │ ├── evol-instruct.md │ ├── firefly-template.py │ ├── jinja-demo.py │ ├── jinja-llm-baichuan.py │ ├── jinja-llm-baichuan2.py │ ├── jinja-llm-bloom.py │ ├── jinja-llm-chatglm3.py │ ├── jinja-llm.py │ ├── jinja.md │ ├── 数据格式设计.md │ └── 数据集格式.md ├── llm-eval ├── .DS_Store ├── EvalScope.md ├── README.md ├── eval-data │ ├── longtext_L115433-question.txt │ ├── longtext_L115433.txt │ ├── longtext_L32503_answer.txt │ ├── longtext_L32503_question.txt │ ├── longtext_L64031.txt │ └── longtext_L64031_question.txt ├── evalscope_framework.png ├── llm-performance │ ├── .DS_Store │ ├── AI芯片性能.md │ ├── README.md │ ├── hardware-performance │ │ ├── gpu-monitor-ui.py │ │ └── pynvml-stat-memory.py │ ├── llmperf.md │ ├── mindie │ │ ├── .DS_Store │ │ ├── lantency │ │ │ ├── README.md │ │ │ ├── perfermance-stat.py │ │ │ ├── performance-stream-baichuan2.py │ │ │ ├── performance-stream-chatglm3.py │ │ │ ├── performance-stream-qwen1.5.py │ │ │ ├── performance-stream-qwen1.py │ │ │ ├── performance-stream.py │ │ │ └── stat_input_token.py │ │ └── locust-lantency-throughput │ │ │ ├── .DS_Store │ │ │ ├── README.md │ │ │ ├── hello.py │ │ │ ├── llm-910b4-baichuan2-7b-2tp.py │ │ │ ├── llm-910b4-chatglm3-6b-2tp.py │ │ │ ├── llm-910b4-qwen-72b-8tp.py │ │ │ ├── llm-910b4-qwen1.5-4tp.py │ │ │ ├── qwen-14b-chart.jpg │ │ │ ├── qwen-14b-stat.jpg │ │ │ ├── qwen1.5-72b-8tp.html │ │ │ ├── qwen1.5-7b-4tp-chart.png │ │ │ ├── qwen1.5-7b-4tp-stat.png │ │ │ └── 示例.py │ ├── perfetto.md │ ├── stat_gpu_memory.py │ ├── tgi-benchmark.md │ ├── tgi-benchmark.png │ ├── vllm-benchmark.md │ ├── vllm │ │ ├── README.md │ │ ├── vllm-locust-qwen1.5-7b-long.py │ │ └── vllm-performance-stream-qwen1.5-long.py │ ├── wrk-性能测试工具.md │ ├── 大模型场景下训练和推理性能指标名词解释.md │ ├── 推理性能测试.md │ └── 训练性能测试.md ├── llm-precision │ ├── .DS_Store │ ├── C-Eval.md │ ├── README.md │ └── 模型质量评估.md ├── opencompass.md └── 大模型测评集.md ├── llm-inference ├── .DS_Store ├── DeepSpeed-Inference.md ├── Flash-Decoding.md ├── FlashInfer.md ├── FlexFlow-Serve.md ├── GuidedGeneration.md ├── KV-Cache优化.md ├── Mooncake.md ├── NanoFlow.md ├── PD分离.md ├── README.md ├── RTP-LLM.md ├── ascend │ ├── .DS_Store │ └── mindformers │ │ ├── .DS_Store │ │ ├── README.md │ │ ├── baichuan2 │ │ ├── README.md │ │ ├── baichuan-inference.py │ │ └── baichuan-stat.py │ │ ├── chatglm3 │ │ ├── README.md │ │ ├── chatglm-gen.py │ │ ├── chatglm-inference.py │ │ └── chatglm-stat.py │ │ ├── mindsporelite-inference.py │ │ ├── mindsporelite-stat.py │ │ └── text_generator_infer.py ├── chatgpt.md ├── deepspeed-mii │ └── README.md ├── faster-transformer │ ├── README.md │ ├── bloom │ │ ├── README.md │ │ └── firefly_lambada_1w_stat_token.py │ ├── gpt │ │ └── README.md │ ├── llama │ │ └── README.md │ └── megatron-gpt2 │ │ ├── gpt_summarization.py │ │ ├── gpt_summarization_stat.py │ │ └── megatron-gpt2-fp8.md ├── flexflow-serve │ └── benchmark-batch1.py ├── huggingface-tgi │ └── README.md ├── huggingface-transformer │ └── README.md ├── lightllm │ └── README.md ├── lmdeploy │ ├── README.md │ ├── 功能.md │ └── 服务启动参数.md ├── native-model │ └── chatglm3-6b │ │ └── cli_demo.py ├── offload.md ├── openai.md ├── sglang │ ├── README.md │ ├── source-code.md │ ├── 服务器启动参数.md │ └── 项目代码结构.md ├── tensorrt-llm │ ├── FP8.md │ ├── Memory Usage of TensorRT-LLM.md │ ├── README.md │ ├── TRT-LLM引擎构建参数.md │ ├── Triton服务启动参数.md │ └── 安装.md ├── tensorrt │ ├── README.md │ └── install.md ├── triton │ ├── REAEME.md │ ├── onnx │ │ └── README.md │ └── resnet50 │ │ ├── client.py │ │ ├── config.pbtxt │ │ ├── labels.txt │ │ └── resnet50_convert_torchscript.py ├── vllm │ ├── FAQ.md │ ├── FP8.md │ ├── README.md │ ├── REF.md │ ├── api_client.py │ ├── cmd.md │ ├── vllm.md │ ├── 服务启动参数.md │ ├── 源码.md │ ├── 请求处理流程.md │ └── 长文本推理.md ├── web │ ├── fastapi │ │ ├── README.md │ │ └── llm-qwen-mindspore-lite.py │ ├── flask │ │ ├── README.md │ │ └── llm-qwen-mindspore-lite.py │ └── sanic │ │ └── README.md ├── xinference │ └── README.md ├── 分离式推理架构.md ├── 大模型推理张量并行.md └── 解码策略.md ├── llm-interview ├── README.md ├── base.md ├── comprehensive.md ├── llm-algo.md ├── llm-app.md ├── llm-compress.md ├── llm-eval.md ├── llm-ft.md ├── llm-inference.md ├── llm-rlhf.md └── llm-train.md ├── llm-localization ├── .DS_Store ├── README.md ├── ascend │ ├── .DS_Store │ ├── FAQ.md │ ├── README.md │ ├── ascend-c │ │ └── README.md │ ├── ascend-infra │ │ ├── HCCL.md │ │ ├── MacOS环境.md │ │ ├── ascend-dmi.md │ │ ├── ascend-docker-runtime.md │ │ ├── ascend-docker.md │ │ ├── ascend-llm下载.md │ │ ├── ascend-npu-smi.md │ │ ├── docker环境升级cann.md │ │ ├── network.md │ │ ├── npu监控.md │ │ ├── 操作系统.md │ │ ├── 昇腾卡-soc版本.md │ │ ├── 昇腾卡注意事项.md │ │ ├── 昇腾镜像.md │ │ ├── 服务器配置.md │ │ ├── 环境安装.md │ │ └── 达芬奇架构.md │ ├── ascend910-env-install.md │ ├── fabric-insight │ │ └── README.md │ ├── firefly-ascend.md │ ├── mindformers │ │ ├── .DS_Store │ │ ├── README.md │ │ ├── baichuan2 │ │ │ ├── baichuan2训练.md │ │ │ ├── run_baichuan2_7b.yaml │ │ │ ├── run_baichuan2_7b_910b.yaml │ │ │ └── run_baichuan2_7b_lora_910b.yaml │ │ ├── chatglm │ │ │ ├── README.md │ │ │ ├── chat_glm.py │ │ │ ├── glm_6b.yaml │ │ │ ├── glm_6b_chat.yaml │ │ │ ├── merge_ckpt.py │ │ │ ├── merge_ckpt_lora.py │ │ │ ├── pt2ms.py │ │ │ ├── run_glm_6b_finetune.yaml │ │ │ ├── run_glm_6b_infer.yaml │ │ │ ├── run_glm_6b_lora.yaml │ │ │ └── run_glm_6b_lora_infer.yaml │ │ ├── env.md │ │ ├── llama │ │ │ └── README.md │ │ ├── qwen │ │ │ ├── qwen1训练.md │ │ │ ├── run_qwen_7b.yaml │ │ │ └── run_qwen_7b_910b.yaml │ │ ├── qwen1.5 │ │ │ ├── qwen1.5训练.md │ │ │ ├── run_qwen1_5_7b_finetune.yaml │ │ │ └── run_qwen1_5_7b_infer.yaml │ │ ├── trick.md │ │ └── 权重格式转换.md │ ├── mindie │ │ ├── .DS_Store │ │ ├── 2.0.RC2 │ │ │ └── qwen.md │ │ ├── README.md │ │ ├── config-1.0.RC1.json │ │ ├── config │ │ │ ├── .DS_Store │ │ │ ├── chatglm3-6b.json │ │ │ ├── qwen-72b.json │ │ │ └── run.sh │ │ ├── docker │ │ │ ├── README.md │ │ │ ├── TEST.md │ │ │ ├── baichuan2-13b.json │ │ │ ├── baichuan2-7b.json │ │ │ ├── deploy.sh │ │ │ ├── install_and_enable_cann.sh │ │ │ ├── llm-server.sh │ │ │ ├── mindie-1.0.Dockerfile │ │ │ ├── mindie-all-1.0.Dockerfile │ │ │ ├── mindie-env-1.0.Dockerfile │ │ │ ├── qwen-72b.json │ │ │ ├── qwen1.5-14b.json │ │ │ ├── qwen1.5-72b.json │ │ │ └── qwen1.5-7b.json │ │ ├── llm-server.sh │ │ ├── mindid-1.0-offical.md │ │ ├── mindid-performance.md │ │ ├── mindie-1.0.Dockerfile │ │ ├── mindie-1.0.RC2.md │ │ ├── mindie-1.0.md │ │ ├── mindie-1.0.rc2-config.json │ │ ├── mindie-1.0.rc2-llm-server.sh │ │ ├── mindie-2.0.rc2.md │ │ ├── mindie-20240411.md │ │ ├── mindie-api.md │ │ ├── model-test.md │ │ ├── script │ │ │ ├── model-test.py │ │ │ └── run.sh │ │ ├── 性能调优.md │ │ └── 日志分析.txt │ ├── mindspore │ │ ├── MindSpore-note.md │ │ ├── README.md │ │ ├── bert.md │ │ ├── reference.md │ │ └── 镜像.md │ ├── modellink │ │ ├── README.md │ │ ├── dataset.md │ │ ├── llm.md │ │ ├── qwen.md │ │ ├── 环境-20240521.md │ │ └── 环境安装.md │ ├── msmodelslim │ │ ├── README.md │ │ └── llm_quant │ │ │ ├── baichuan2-w8a8.py │ │ │ ├── calib_set.json │ │ │ └── qwen1.5-72b-w8a16.py │ ├── openmind │ │ └── README.md │ ├── peft │ │ ├── README.md │ │ └── finetune-lora.py │ ├── pytorch │ │ ├── .DS_Store │ │ ├── README.md │ │ └── llm-lora.py │ ├── standford-alpaca │ │ ├── README.md │ │ ├── ds_config_zero2.json │ │ ├── ds_config_zero3.json │ │ ├── requirements.txt │ │ ├── train.py │ │ └── utils.py │ ├── transformers │ │ └── README.md │ ├── vllm-ascend │ │ └── README.md │ ├── 优质学习资料.md │ ├── 昇腾LLM支持概览.md │ └── 昇腾卡注意事项.md ├── modelscope │ └── README.md ├── paddle │ └── PaddleNLP.md └── tianshuzhixin │ ├── README.md │ └── ixsmi.md ├── llm-maas ├── OpenAI-ChatGPT.md └── README.md ├── llm-optimizer ├── FlashAttention.md ├── README.md ├── SplitFuse.md ├── kv-cache.md ├── xformers.md └── 计算通信重叠.md ├── llm-pipeline └── REAEMD.md ├── llm-tools ├── Pytorch-Profiler.md ├── README.md ├── base-profiler.py ├── nsight.md ├── nsight │ ├── 0-GPU性能分析.png │ ├── 1-GPU利用率低.png │ └── README.md ├── nvtx.md ├── profiler-recipe.py ├── tensorboard-profiler.py └── 可视化.md ├── llm-train ├── .DS_Store ├── README.md ├── alpa │ └── train │ │ ├── pipeshard_parallelism.ipynb │ │ └── pipeshard_parallelism.py ├── alpaca-lora │ ├── README.md │ ├── export_hf_checkpoint.py │ ├── export_state_dict_checkpoint.py │ ├── finetune.py │ ├── finetune_metrics_epoch.py │ ├── generate.py │ └── inference.py ├── alpaca │ ├── README.md │ ├── ds_config.json │ ├── ds_config_zero2.json │ ├── ds_config_zero2_ddp.json │ ├── inference.py │ ├── train.py │ └── train_ddp.py ├── ascend │ └── .DS_Store ├── chatglm-lora │ ├── README.md │ ├── finetune.py │ ├── finetune_ddp.py │ └── inference.py ├── chatglm │ ├── README.md │ ├── deepspeed.json │ ├── ds_train_finetune.sh │ ├── evaluate.sh │ ├── evaluate_finetune.sh │ ├── inference.py │ ├── main.py │ ├── train.sh │ └── train_ptuningv2_dp.sh ├── chinese-llama-alpaca │ ├── README.md │ ├── inference_hf.py │ ├── merge_llama_with_chinese_lora.py │ ├── merge_tokenizers.py │ ├── run_clm_pt_with_peft.py │ ├── run_clm_sft_with_peft.py │ ├── run_pt.sh │ └── run_sft.sh ├── deepspeedchat │ ├── README.md │ ├── llama │ │ └── README.md │ └── training │ │ ├── step1_supervised_finetuning │ │ └── training_scripts │ │ │ └── single_node │ │ │ └── run_13b.sh │ │ ├── step2_reward_model_finetuning │ │ └── training_scripts │ │ │ └── single_node │ │ │ └── run_350m.sh │ │ ├── step3_rlhf_finetuning │ │ └── training_scripts │ │ │ └── single_node │ │ │ └── run_13b.sh │ │ └── utils │ │ └── data │ │ └── raw_datasets.py ├── firefly │ ├── README.md │ ├── bootstrap-s3.sh │ ├── bootstrap.sh │ ├── dockerfile.md │ └── test_bash_getopts.sh ├── fp8.md ├── galore │ └── torchrun_main.py ├── megatron-deepspeed │ ├── README.md │ ├── bigscience │ │ └── bloom-note.md │ ├── bloom-megatron-deepspeed.md │ ├── microsoft │ │ ├── H800多机多卡训练坑点.md │ │ ├── README.md │ │ ├── llama-note.md │ │ ├── pip.conf │ │ ├── pretrain_llama2_13b_distributed_fp16.sh │ │ ├── pretrain_llama2_distributed.sh │ │ ├── pretrain_llama_13b_distributed_fp16.sh │ │ ├── pretrain_llama_7b_distributed_fp16.sh │ │ ├── pretrain_llama_distributed_fp16.sh │ │ ├── slurm │ │ │ ├── README.md │ │ │ ├── llama-multinode-ib.sh │ │ │ ├── megatron-deepspeed-multinode-ib-part2-30b-fp16.slurm │ │ │ └── megatron-deepspeed-multinode-ib-part2-65b-fp16.slurm │ │ ├── 代码.md │ │ ├── 环境准备.md │ │ ├── 训练日志分析.md │ │ └── 项目结构-202312228.md │ └── source-code.md ├── megatron │ ├── README.md │ ├── codegeex │ │ ├── README.md │ │ └── pic │ │ │ ├── CodeGeeX模型架构.png │ │ │ └── CodeGeeX训练配置.png │ ├── gpt2 │ │ ├── README.md │ │ ├── data │ │ │ ├── cMinhash.cpp │ │ │ ├── download.py │ │ │ ├── file_utils.py │ │ │ └── merge_data.py │ │ ├── gpt-data-preprocess.md │ │ ├── merge_ck_and_inference │ │ │ ├── README.md │ │ │ ├── checkpoint_loader_megatron.py │ │ │ ├── checkpoint_saver_megatron.py │ │ │ ├── checkpoint_util.py │ │ │ ├── eval_gpt2_lambada.sh │ │ │ ├── run_text_generation_server.py │ │ │ ├── run_text_generation_server_345M.sh │ │ │ ├── run_text_generation_server_345M_2tp_2dp.sh │ │ │ ├── run_text_generation_server_345M_4_tensor_parallel.sh │ │ │ └── text_generation_cli.py │ │ ├── model_merge_eval_inference.md │ │ ├── model_train.md │ │ ├── requirements.txt │ │ └── train │ │ │ ├── pretrain_gpt.sh │ │ │ ├── pretrain_gpt_distributed.sh │ │ │ ├── pretrain_gpt_distributed_with_4pp.sh │ │ │ ├── pretrain_gpt_distributed_with_4tp.sh │ │ │ └── pretrain_gpt_distributed_with_mp.sh │ ├── kernel_fusion.png │ ├── megatron.drawio │ ├── pretrain.xmind │ ├── project.md │ └── source-code.md ├── paddle │ ├── README.md │ └── paddlenlp │ │ ├── README.md │ │ ├── baichuan2 │ │ └── README.md │ │ └── bloom │ │ ├── README.md │ │ └── sft_argument.json ├── peft │ ├── .DS_Store │ ├── LoRA-QLoRA.md │ ├── PEFT-API.md │ ├── Prefix-Tuning.md │ ├── Prompt-Tuning.md │ ├── README.md │ ├── clm │ │ ├── accelerate_ds_zero3_cpu_offload_config.yaml │ │ ├── peft_ia3_clm.ipynb │ │ ├── peft_lora_clm.ipynb │ │ ├── peft_lora_clm_accelerate_ds_zero3_offload.py │ │ ├── peft_p_tuning_clm.ipynb │ │ ├── peft_p_tuning_lstm_clm.ipynb │ │ ├── peft_p_tuning_v2_clm.ipynb │ │ ├── peft_prefix_tuning_clm.ipynb │ │ └── peft_prompt_tuning_clm.ipynb │ ├── conditional_generation │ │ └── README.md │ └── multimodal │ │ ├── blip2_lora_inference.py │ │ ├── blip2_lora_int8_fine_tune.py │ │ └── finetune_bloom_bnb_peft.ipynb ├── pytorch │ ├── Pytorch源码解读.md │ ├── README.md │ ├── api.md │ ├── distribution │ │ ├── README.md │ │ ├── api.md │ │ ├── data-parallel │ │ │ ├── README.md │ │ │ ├── ddp_launch.py │ │ │ ├── ddp_main.py │ │ │ ├── elastic_ddp.py │ │ │ ├── minGPT-ddp │ │ │ │ ├── README.md │ │ │ │ ├── multinode.sh │ │ │ │ ├── sbatch_run.sh │ │ │ │ ├── sbatch_run_sig.sh │ │ │ │ └── sbatch_run_sig_opt.sh │ │ │ ├── sbatch_run.sh │ │ │ └── 使用DDP训练真实世界的模型.md │ │ ├── pipeline-parallel │ │ │ ├── 1-流水线.md │ │ │ ├── 2-使用torchtext训练transformer模型.md │ │ │ ├── 3-使用流水线并行训练Transformer模型.md │ │ │ ├── 4-使用DDP与流水线并行训练Transformer模型.md │ │ │ ├── README.md │ │ │ ├── ddp_pipeline.py │ │ │ ├── pipeline_tutorial.ipynb │ │ │ └── transformer_tutorial.ipynb │ │ ├── rpc │ │ │ └── README.md │ │ ├── sequence-parallelism │ │ │ └── README.md │ │ ├── tensor-parallel │ │ │ ├── 2d_parallel_example.py │ │ │ ├── README.md │ │ │ ├── sequence_parallel_example.py │ │ │ ├── tensor_parallel_example.py │ │ │ └── utils.py │ │ ├── torchrun.md │ │ ├── 分布式通信包.md │ │ ├── 多机多卡.md │ │ └── 多机训练.md │ ├── resource.md │ └── torchrun.md ├── qlora │ ├── README.md │ ├── accuracy.py │ ├── export_hf_checkpoint.py │ ├── inference.py │ ├── inference_merge.py │ ├── inference_qlora.py │ └── qlora.py ├── slurm │ ├── README.md │ ├── deepspeed │ │ ├── pp-multinode-machine.slurm │ │ ├── pp-multinode-singularity.slurm │ │ ├── pp-mutinode-singularity-pmix.slurm │ │ ├── pp-standalone-singularity-v2.slurm │ │ └── pp-standalone-singularity.slurm │ ├── megatron-deepspeed │ │ └── megatron-deepspeed-multinode-ib-part2-65b-fp16.slurm │ └── pytorch │ │ ├── alpaca-docker.slurm │ │ ├── alpaca-machine.slurm │ │ ├── alpaca-singularity.slurm │ │ ├── mingpt-singularity-multinode-2.slurm │ │ └── mingpt-singularity-multinode.slurm └── vicuna │ └── README.md ├── llmops ├── FAQ.md ├── README.md ├── kubernetes.md ├── tq-llm │ └── train │ │ ├── FAQ.md │ │ ├── README.md │ │ ├── bootstrap-llm-zero3-offload.sh │ │ ├── bootstrap-llm.sh │ │ ├── bootstrap-llm2.sh │ │ ├── zero2-offload.json │ │ └── zero3-offload.json ├── 使用docker进行多机多卡训练.md ├── 千帆大模型平台.md └── 模型推理平台方案.md ├── mkdir-dir-file.sh ├── paper ├── A Survey on Efficient Training of Transformers.md ├── LESS-选择有影响力的数据进行目标指令精调.md ├── LLM增强LLMS.md ├── PagedAttention.md ├── README.md ├── data │ ├── LESS 实践：仅用少量的数据完成目标指令微调.md │ ├── LESS-选择有影响力的数据进行目标指令精调.md │ └── LESS.md ├── inference │ ├── llm-in-a-flash.md │ ├── orca.md │ └── 迈向高效的生成式大语言模型服务综述-从算法到系统.md ├── llm对齐综述.md ├── moe │ └── README.md ├── parameter-pruning │ ├── LLM-Pruner.md │ ├── SparseGPT.md │ ├── Wanda.md │ └── 公式.md └── training │ ├── A Survey on Efficient Training of Transformers.md │ ├── GaLore.md │ └── Reducing Activation Recomputation in Large Transformer Models.md ├── pic ├── .DS_Store ├── damodel.jpeg ├── llm-action-v3.png ├── llm-action-v4.jpg ├── llm │ ├── .DS_Store │ ├── model │ │ ├── llm-famliy.jpg │ │ └── llm-timeline-v2.png │ └── train │ │ ├── .DS_Store │ │ ├── pretrain │ │ └── llm-pretrain-pipeline-v2.png │ │ └── sft │ │ └── peft方法.jpg ├── wechat.jpeg ├── wx-gzh.png ├── wx.jpg └── 公众号.jpeg └── template └── server.md /.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/.DS_Store -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/.gitignore -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/LICENSE -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/README.md -------------------------------------------------------------------------------- /ai-compiler/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/.DS_Store -------------------------------------------------------------------------------- /ai-compiler/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/README.md -------------------------------------------------------------------------------- /ai-compiler/Treebeard/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/Treebeard/.DS_Store -------------------------------------------------------------------------------- /ai-compiler/Treebeard/10-TableI-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/Treebeard/10-TableI-1.png -------------------------------------------------------------------------------- /ai-compiler/Treebeard/10-TableII-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/Treebeard/10-TableII-1.png -------------------------------------------------------------------------------- /ai-compiler/Treebeard/9-Figure6-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/Treebeard/9-Figure6-1.png -------------------------------------------------------------------------------- /ai-compiler/Treebeard/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/Treebeard/README.md -------------------------------------------------------------------------------- /ai-compiler/Treebeard/TREEBEARD编译器结构.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/Treebeard/TREEBEARD编译器结构.png -------------------------------------------------------------------------------- /ai-compiler/treelit/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-compiler/treelit/README.md -------------------------------------------------------------------------------- /ai-compiler/treelit/xgb.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | ``` 6 | conda create -n model-server-venv python=3.9 -y 7 | ``` -------------------------------------------------------------------------------- /ai-compiler/triton-lang/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /ai-framework/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/.DS_Store -------------------------------------------------------------------------------- /ai-framework/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/README.md -------------------------------------------------------------------------------- /ai-framework/TensorRT-Model-Optimizer.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/TensorRT-Model-Optimizer.md -------------------------------------------------------------------------------- /ai-framework/cuda/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /ai-framework/deepspeed/1.DeepSpeed入门.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/1.DeepSpeed入门.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/2.安装DeepSpeed.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/2.安装DeepSpeed.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/3.基于CIFAR-10使用DeepSpeed进行分布式训练 .md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/3.基于CIFAR-10使用DeepSpeed进行分布式训练 .md -------------------------------------------------------------------------------- /ai-framework/deepspeed/DeepSpeed配置JSON文件.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/DeepSpeed配置JSON文件.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/README.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/config-json/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/config-json/README.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/config-json/deepspeed-nvme.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/config-json/deepspeed-nvme.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/deepspeed-slurm.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/deepspeed-slurm.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/hello_bert/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/hello_bert/README.md -------------------------------------------------------------------------------- /ai-framework/deepspeed/hello_bert/train_bert.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/hello_bert/train_bert.py -------------------------------------------------------------------------------- /ai-framework/deepspeed/hello_bert/train_bert_ds.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/hello_bert/train_bert_ds.py -------------------------------------------------------------------------------- /ai-framework/deepspeed/training/pipeline_parallelism/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/deepspeed/training/pipeline_parallelism/README.md -------------------------------------------------------------------------------- /ai-framework/dlrover.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/dlrover.md -------------------------------------------------------------------------------- /ai-framework/huggingface-accelerate/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/huggingface-accelerate/README.md -------------------------------------------------------------------------------- /ai-framework/huggingface-peft/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /ai-framework/huggingface-transformers/API.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/huggingface-transformers/API.md -------------------------------------------------------------------------------- /ai-framework/huggingface-transformers/FSDP.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/huggingface-transformers/FSDP.md -------------------------------------------------------------------------------- /ai-framework/huggingface-transformers/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/huggingface-transformers/README.md -------------------------------------------------------------------------------- /ai-framework/huggingface-trl/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/huggingface-trl/README.md -------------------------------------------------------------------------------- /ai-framework/jax/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/jax/README.md -------------------------------------------------------------------------------- /ai-framework/jax/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/jax/reference.md -------------------------------------------------------------------------------- /ai-framework/llama-cpp/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/llama-cpp/README.md -------------------------------------------------------------------------------- /ai-framework/megatron-deepspeed/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /ai-framework/megatron-lm/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /ai-framework/mxnet/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/README.md -------------------------------------------------------------------------------- /ai-framework/mxnet/mnist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/mnist.py -------------------------------------------------------------------------------- /ai-framework/mxnet/mxnet_cnn_mnist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/mxnet_cnn_mnist.py -------------------------------------------------------------------------------- /ai-framework/mxnet/mxnet_mlp_mnist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/mxnet_mlp_mnist.py -------------------------------------------------------------------------------- /ai-framework/mxnet/oneflow_cnn_mnist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/oneflow_cnn_mnist.py -------------------------------------------------------------------------------- /ai-framework/mxnet/oneflow_mlp_mnist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/oneflow_mlp_mnist.py -------------------------------------------------------------------------------- /ai-framework/mxnet/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/mxnet/reference.md -------------------------------------------------------------------------------- /ai-framework/oneflow/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/oneflow/README.md -------------------------------------------------------------------------------- /ai-framework/oneflow/oneflow_mlp_mnist.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/oneflow/oneflow_mlp_mnist.py -------------------------------------------------------------------------------- /ai-framework/oneflow/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/oneflow/reference.md -------------------------------------------------------------------------------- /ai-framework/openai-triton/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/openai-triton/README.md -------------------------------------------------------------------------------- /ai-framework/paddlepaddle/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/paddlepaddle/README.md -------------------------------------------------------------------------------- /ai-framework/paddlepaddle/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/paddlepaddle/reference.md -------------------------------------------------------------------------------- /ai-framework/pai-megatron-patch/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/pai-megatron-patch/.DS_Store -------------------------------------------------------------------------------- /ai-framework/pai-megatron-patch/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/pai-megatron-patch/README.md -------------------------------------------------------------------------------- /ai-framework/pai-torchacc.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/pai-torchacc.md -------------------------------------------------------------------------------- /ai-framework/pytorch/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/pytorch/README.md -------------------------------------------------------------------------------- /ai-framework/pytorch/install.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/pytorch/install.md -------------------------------------------------------------------------------- /ai-framework/pytorch/reference.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /ai-framework/tensorflow/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /ai-framework/tensorflow/reference.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /ai-framework/transformer-engine/mnist/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/transformer-engine/mnist/README.md -------------------------------------------------------------------------------- /ai-framework/transformer-engine/mnist/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/transformer-engine/mnist/main.py -------------------------------------------------------------------------------- /ai-framework/transformer-engine/mnist/main_stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/transformer-engine/mnist/main_stat.py -------------------------------------------------------------------------------- /ai-framework/unsloth-微调.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-framework/unsloth-微调.md -------------------------------------------------------------------------------- /ai-infra/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/.DS_Store -------------------------------------------------------------------------------- /ai-infra/ai-cluster/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-cluster/README.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/.DS_Store -------------------------------------------------------------------------------- /ai-infra/ai-hardware/AI芯片软件生态.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/AI芯片软件生态.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/CUDA.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/CUDA.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/GPU-network.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/GPU-network.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/GPU相关环节变量.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/GPU相关环节变量.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/NIXL.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | https://github.com/ai-dynamo/nixl 5 | -------------------------------------------------------------------------------- /ai-infra/ai-hardware/OEM-DGX.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/OEM-DGX.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/README.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/TSMC-台积电.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/TSMC-台积电.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/cuda镜像.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/cuda镜像.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/gpudirect.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/gpudirect.md -------------------------------------------------------------------------------- /ai-infra/ai-hardware/硬件对比.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/ai-hardware/硬件对比.md -------------------------------------------------------------------------------- /ai-infra/communication.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/communication.md -------------------------------------------------------------------------------- /ai-infra/存储/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/存储/README.md -------------------------------------------------------------------------------- /ai-infra/存储/REF.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | -------------------------------------------------------------------------------- /ai-infra/存储/nvme-ssd.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/存储/nvme-ssd.md -------------------------------------------------------------------------------- /ai-infra/存储/固态硬盘.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/存储/固态硬盘.md -------------------------------------------------------------------------------- /ai-infra/存储/存储.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/存储/存储.md -------------------------------------------------------------------------------- /ai-infra/算力/AI芯片.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/算力/AI芯片.md -------------------------------------------------------------------------------- /ai-infra/算力/GPU工作原理.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/算力/GPU工作原理.md -------------------------------------------------------------------------------- /ai-infra/算力/NVIDIA-GPU型号.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/算力/NVIDIA-GPU型号.md -------------------------------------------------------------------------------- /ai-infra/算力/推理芯片.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/算力/推理芯片.md -------------------------------------------------------------------------------- /ai-infra/算力/昇腾NPU.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | Atlas 800-9000A2 6 | 7 | 8 | 9 | -------------------------------------------------------------------------------- /ai-infra/网络/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/.DS_Store -------------------------------------------------------------------------------- /ai-infra/网络/HPC性能测试.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/HPC性能测试.md -------------------------------------------------------------------------------- /ai-infra/网络/IB-docker.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/IB-docker.md -------------------------------------------------------------------------------- /ai-infra/网络/IB流量监控.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/IB流量监控.md -------------------------------------------------------------------------------- /ai-infra/网络/IB软件.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/IB软件.md -------------------------------------------------------------------------------- /ai-infra/网络/InfiniBand.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/InfiniBand.md -------------------------------------------------------------------------------- /ai-infra/网络/NCCL.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/NCCL.md -------------------------------------------------------------------------------- /ai-infra/网络/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/README.md -------------------------------------------------------------------------------- /ai-infra/网络/REF.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | - [RoCE、IB和TCP等网络的基本知识及差异对比](https://support.huawei.com/enterprise/zh/doc/EDOC1100203347) 5 | 6 | 7 | 8 | 9 | 10 | -------------------------------------------------------------------------------- /ai-infra/网络/Spine-Leaf和InfiniBand网络架构区别简述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/Spine-Leaf和InfiniBand网络架构区别简述.md -------------------------------------------------------------------------------- /ai-infra/网络/nccl-test-集合通讯的性能测试.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/nccl-test-集合通讯的性能测试.md -------------------------------------------------------------------------------- /ai-infra/网络/nvbandwidth.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/nvbandwidth.md -------------------------------------------------------------------------------- /ai-infra/网络/pic/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/.DS_Store -------------------------------------------------------------------------------- /ai-infra/网络/pic/8卡V100的混合网络拓扑.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/8卡V100的混合网络拓扑.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/A800-H100-H800.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/A800-H100-H800.jpeg -------------------------------------------------------------------------------- /ai-infra/网络/pic/NVLink-generations比较.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/NVLink-generations比较.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/PCIe-Generation对比.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/PCIe-Generation对比.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/nvidia-dgx-1-v100-nvlink-gpu-xeon-config.webp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/nvidia-dgx-1-v100-nvlink-gpu-xeon-config.webp -------------------------------------------------------------------------------- /ai-infra/网络/pic/nvidia-dgx-1-with-volta.webp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/nvidia-dgx-1-with-volta.webp -------------------------------------------------------------------------------- /ai-infra/网络/pic/nvlink.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/nvlink.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/nvlink性能.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/nvlink性能.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/nvswitch.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/nvswitch.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/支持 NVLink GPU 之间连接的 NVIDIA H100-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/支持 NVLink GPU 之间连接的 NVIDIA H100-1.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/支持 NVLink GPU 之间连接的 NVIDIA H100-2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/支持 NVLink GPU 之间连接的 NVIDIA H100-2.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/服务器之间的nvlink与nvswitch.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/服务器之间的nvlink与nvswitch.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/网络之间的连接.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/网络之间的连接.png -------------------------------------------------------------------------------- /ai-infra/网络/pic/英伟达A100-A800-H100-H800.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/pic/英伟达A100-A800-H100-H800.jpeg -------------------------------------------------------------------------------- /ai-infra/网络/ringallreduce/All Gather 流程图.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/ringallreduce/All Gather 流程图.gif -------------------------------------------------------------------------------- /ai-infra/网络/ringallreduce/Scatter Reduce 流程图.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/ringallreduce/Scatter Reduce 流程图.gif -------------------------------------------------------------------------------- /ai-infra/网络/roce.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/roce.md -------------------------------------------------------------------------------- /ai-infra/网络/网络硬件.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/网络硬件.md -------------------------------------------------------------------------------- /ai-infra/网络/通信软件.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/ai-infra/网络/通信软件.md -------------------------------------------------------------------------------- /ai-infra/网络/集合通信原语.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 集合通信总结和 mpi4py 实践 5 | 6 | https://www.armcvai.cn/2025-06-28/mpi4py-summary.html 7 | 8 | 9 | -------------------------------------------------------------------------------- /blog/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/.DS_Store -------------------------------------------------------------------------------- /blog/TODO.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/TODO.md -------------------------------------------------------------------------------- /blog/ai-infra/AI 集群基础设施 InfiniBand 详解.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/ai-infra/AI 集群基础设施 InfiniBand 详解.md -------------------------------------------------------------------------------- /blog/ai-infra/AI 集群基础设施 NVMe SSD 详解.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/ai-infra/AI 集群基础设施 NVMe SSD 详解.md -------------------------------------------------------------------------------- /blog/distribution-parallelism/大模型分布式训练并行技术（一）-概述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/distribution-parallelism/大模型分布式训练并行技术（一）-概述.md -------------------------------------------------------------------------------- /blog/distribution-parallelism/大模型分布式训练并行技术（九）-总结.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/distribution-parallelism/大模型分布式训练并行技术（九）-总结.md -------------------------------------------------------------------------------- /blog/distribution-parallelism/大模型分布式训练并行技术（六）-多维混合并行.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/distribution-parallelism/大模型分布式训练并行技术（六）-多维混合并行.md -------------------------------------------------------------------------------- /blog/llm-algo/moe.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-algo/moe.md -------------------------------------------------------------------------------- /blog/llm-algo/大白话Transformer架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-algo/大白话Transformer架构.md -------------------------------------------------------------------------------- /blog/llm-compression/大模型量化技术原理-ZeroQuant系列.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-compression/大模型量化技术原理-ZeroQuant系列.md -------------------------------------------------------------------------------- /blog/llm-compression/大模型量化技术原理：QoQ量化及QServe推理服务系统.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-compression/大模型量化技术原理：QoQ量化及QServe推理服务系统.md -------------------------------------------------------------------------------- /blog/llm-inference/大模型推理框架概述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-inference/大模型推理框架概述.md -------------------------------------------------------------------------------- /blog/llm-localization/大模型国产化适配1-华为昇腾AI全栈软硬件平台总结.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-localization/大模型国产化适配1-华为昇腾AI全栈软硬件平台总结.md -------------------------------------------------------------------------------- /blog/llm-localization/大模型国产化适配4-基于昇腾910使用LLaMA-13B进行多机多卡训练.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-localization/大模型国产化适配4-基于昇腾910使用LLaMA-13B进行多机多卡训练.md -------------------------------------------------------------------------------- /blog/llm-peft/大模型参数高效微调技术原理综述（一）-背景、参数高效微调简介.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-peft/大模型参数高效微调技术原理综述（一）-背景、参数高效微调简介.md -------------------------------------------------------------------------------- /blog/llm-peft/大模型参数高效微调技术原理综述（五）-LoRA、AdaLoRA、QLoRA.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/blog/llm-peft/大模型参数高效微调技术原理综述（五）-LoRA、AdaLoRA、QLoRA.md -------------------------------------------------------------------------------- /blog/reference/高性能 LLM 推理框架的设计与实现.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | - 高性能 LLM 推理框架的设计与实现：https://zhuanlan.zhihu.com/p/682872971 5 | 6 | -------------------------------------------------------------------------------- /docs/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/.DS_Store -------------------------------------------------------------------------------- /docs/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/README.md -------------------------------------------------------------------------------- /docs/conda.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/conda.md -------------------------------------------------------------------------------- /docs/flash-attention/FlashAttention.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/flash-attention/FlashAttention.md -------------------------------------------------------------------------------- /docs/llm-base/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/.DS_Store -------------------------------------------------------------------------------- /docs/llm-base/FLOPS.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/FLOPS.md -------------------------------------------------------------------------------- /docs/llm-base/NVIDIA-Nsight-Systems性能分析.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/NVIDIA-Nsight-Systems性能分析.md -------------------------------------------------------------------------------- /docs/llm-base/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/README.md -------------------------------------------------------------------------------- /docs/llm-base/a800-env-install.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/a800-env-install.md -------------------------------------------------------------------------------- /docs/llm-base/ai-algo.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/ai-algo.md -------------------------------------------------------------------------------- /docs/llm-base/autoregressive-lm-decoding-methods.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/autoregressive-lm-decoding-methods.md -------------------------------------------------------------------------------- /docs/llm-base/dcgmi.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/dcgmi.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/.DS_Store -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/README.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/auto-parallel/Alpa.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/auto-parallel/Alpa.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/auto-parallel/Flexflow.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/auto-parallel/Flexflow.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/auto-parallel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/auto-parallel/README.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/auto-parallel/Unity.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/auto-parallel/Unity.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/auto-parallel/gspmd.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/auto-parallel/gspmd.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/data-parallelism/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/moe-parallel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/moe-parallel/README.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/moe-parallel/moe-parallel.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/tensor-parallel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/tensor-parallel/README.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-parallelism/并行技术.drawio: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-parallelism/并行技术.drawio -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/.DS_Store -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/Bloom-176B训练经验.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/Bloom-176B训练经验.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/FP16-BF16.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/FP16-BF16.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/GLM-130B训练经验.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/GLM-130B训练经验.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/OPT-175B训练经验.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/OPT-175B训练经验.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/README.md -------------------------------------------------------------------------------- /docs/llm-base/distribution-training/自动混合精度.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/distribution-training/自动混合精度.md -------------------------------------------------------------------------------- /docs/llm-base/gpu-env-var.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | CUDA_VISIBLE_DEVICES=1 5 | 6 | 7 | 8 | -------------------------------------------------------------------------------- /docs/llm-base/h800-env-install.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/h800-env-install.md -------------------------------------------------------------------------------- /docs/llm-base/images/slurm/slurm.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/images/slurm/slurm.gif -------------------------------------------------------------------------------- /docs/llm-base/monitor.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/monitor.md -------------------------------------------------------------------------------- /docs/llm-base/multimodal/sora.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/multimodal/sora.md -------------------------------------------------------------------------------- /docs/llm-base/nvidia-smi-dmon.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/nvidia-smi-dmon.md -------------------------------------------------------------------------------- /docs/llm-base/nvidia-smi.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/nvidia-smi.md -------------------------------------------------------------------------------- /docs/llm-base/rlhf/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/rlhf/README.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/README.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/cv/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/cv/README.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/cv/paddle/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/cv/paddle/README.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/cv/pytorch/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/cv/pytorch/README.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/cv/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/cv/reference.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/multi-modal/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/multi-modal/README.md -------------------------------------------------------------------------------- /docs/llm-base/scenes/multi-modal/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/scenes/multi-modal/reference.md -------------------------------------------------------------------------------- /docs/llm-base/singularity命令.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/singularity命令.md -------------------------------------------------------------------------------- /docs/llm-base/slurm.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/slurm.md -------------------------------------------------------------------------------- /docs/llm-base/分布式训练加速技术.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/分布式训练加速技术.md -------------------------------------------------------------------------------- /docs/llm-base/多机RDMA性能测试.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/多机RDMA性能测试.txt -------------------------------------------------------------------------------- /docs/llm-base/机器学习中常用的数据类型.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-base/机器学习中常用的数据类型.md -------------------------------------------------------------------------------- /docs/llm-experience.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-experience.md -------------------------------------------------------------------------------- /docs/llm-inference/DeepSpeed-Inference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/DeepSpeed-Inference.md -------------------------------------------------------------------------------- /docs/llm-inference/KV-Cache.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/KV-Cache.md -------------------------------------------------------------------------------- /docs/llm-inference/LLM服务框架对比.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/LLM服务框架对比.md -------------------------------------------------------------------------------- /docs/llm-inference/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/README.md -------------------------------------------------------------------------------- /docs/llm-inference/blog.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/blog.md -------------------------------------------------------------------------------- /docs/llm-inference/flexflow/spec_infer_demo.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/flexflow/spec_infer_demo.gif -------------------------------------------------------------------------------- /docs/llm-inference/flexflow/投机采样.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/flexflow/投机采样.md -------------------------------------------------------------------------------- /docs/llm-inference/llm推理优化技术.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/llm推理优化技术.md -------------------------------------------------------------------------------- /docs/llm-inference/llm推理框架.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/llm推理框架.md -------------------------------------------------------------------------------- /docs/llm-inference/vllm.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-inference/vllm.md -------------------------------------------------------------------------------- /docs/llm-peft/LoRA-FA.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-peft/LoRA-FA.md -------------------------------------------------------------------------------- /docs/llm-peft/MAM_Adapter.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-peft/MAM_Adapter.md -------------------------------------------------------------------------------- /docs/llm-peft/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-peft/README.md -------------------------------------------------------------------------------- /docs/llm-peft/ReLoRA.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | -------------------------------------------------------------------------------- /docs/llm-summarize/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/README.md -------------------------------------------------------------------------------- /docs/llm-summarize/distribution_dl_roadmap.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/distribution_dl_roadmap.md -------------------------------------------------------------------------------- /docs/llm-summarize/pic/A800.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/pic/A800.jpeg -------------------------------------------------------------------------------- /docs/llm-summarize/pic/H800.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/pic/H800.jpeg -------------------------------------------------------------------------------- /docs/llm-summarize/pic/transformer架构.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/pic/transformer架构.jpg -------------------------------------------------------------------------------- /docs/llm-summarize/pic/why_RLHF.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/pic/why_RLHF.jpg -------------------------------------------------------------------------------- /docs/llm-summarize/大模型实践总结-20230930.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/大模型实践总结-20230930.md -------------------------------------------------------------------------------- /docs/llm-summarize/大模型实践总结.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/大模型实践总结.md -------------------------------------------------------------------------------- /docs/llm-summarize/文档大模型.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/文档大模型.md -------------------------------------------------------------------------------- /docs/llm-summarize/金融大模型.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/金融大模型.md -------------------------------------------------------------------------------- /docs/llm-summarize/领域大模型.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/llm-summarize/领域大模型.md -------------------------------------------------------------------------------- /docs/transformer内存估算.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/docs/transformer内存估算.md -------------------------------------------------------------------------------- /faq/FAQ.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/faq/FAQ.md -------------------------------------------------------------------------------- /git-pull-push.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/git-pull-push.sh -------------------------------------------------------------------------------- /llm-algo/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/.DS_Store -------------------------------------------------------------------------------- /llm-algo/FLOPs.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/FLOPs.md -------------------------------------------------------------------------------- /llm-algo/InternLM-20B.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/InternLM-20B.md -------------------------------------------------------------------------------- /llm-algo/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/README.md -------------------------------------------------------------------------------- /llm-algo/baichuan2/baichuan.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/baichuan2/baichuan.md -------------------------------------------------------------------------------- /llm-algo/bert.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/bert.md -------------------------------------------------------------------------------- /llm-algo/bert/模型架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/bert/模型架构.md -------------------------------------------------------------------------------- /llm-algo/bloom.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/bloom.md -------------------------------------------------------------------------------- /llm-algo/bloom/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/bloom/README.md -------------------------------------------------------------------------------- /llm-algo/chatglm/GLM说明.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/chatglm/GLM说明.png -------------------------------------------------------------------------------- /llm-algo/chatglm/GLM预训练.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/chatglm/GLM预训练.png -------------------------------------------------------------------------------- /llm-algo/chatglm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/chatglm/README.md -------------------------------------------------------------------------------- /llm-algo/chatglm/模型架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/chatglm/模型架构.md -------------------------------------------------------------------------------- /llm-algo/chatglm2/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/chatglm2/README.md -------------------------------------------------------------------------------- /llm-algo/chatglm2/模型架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/chatglm2/模型架构.md -------------------------------------------------------------------------------- /llm-algo/chatglm3/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | -------------------------------------------------------------------------------- /llm-algo/chatglm3/reference.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-algo/chatgpt/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | - ChatGPT技术原理解析：从RL之PPO算法、RLHF到GPT4、instructGPT: https://blog.csdn.net/v_JULY_v/article/details/128579457 6 | 7 | 8 | 9 | -------------------------------------------------------------------------------- /llm-algo/deepseek/DeepSeek-R1.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/deepseek/DeepSeek-R1.md -------------------------------------------------------------------------------- /llm-algo/deepseek/DeepSeek-V2.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/deepseek/DeepSeek-V2.md -------------------------------------------------------------------------------- /llm-algo/deepseek/DeepSeek-V3.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/deepseek/DeepSeek-V3.md -------------------------------------------------------------------------------- /llm-algo/deepseek/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/deepseek/README.md -------------------------------------------------------------------------------- /llm-algo/glm-130b/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/glm-130b/README.md -------------------------------------------------------------------------------- /llm-algo/glm-130b/模型架构.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/glm-130b/模型架构.gif -------------------------------------------------------------------------------- /llm-algo/glm4.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | -------------------------------------------------------------------------------- /llm-algo/gpt/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/gpt/README.md -------------------------------------------------------------------------------- /llm-algo/gpt/模型结构.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/gpt/模型结构.png -------------------------------------------------------------------------------- /llm-algo/gpt2/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/gpt2/README.md -------------------------------------------------------------------------------- /llm-algo/gpt2/hf_modeling_gpt2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/gpt2/hf_modeling_gpt2.py -------------------------------------------------------------------------------- /llm-algo/gpt2/模型架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/gpt2/模型架构.md -------------------------------------------------------------------------------- /llm-algo/gpt3/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | GPT（五）GPT3原理讲解：https://zhuanlan.zhihu.com/p/642745932 6 | 7 | 8 | 9 | -------------------------------------------------------------------------------- /llm-algo/llama.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/llama.md -------------------------------------------------------------------------------- /llm-algo/llama/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/llama/README.md -------------------------------------------------------------------------------- /llm-algo/llama/模型架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/llama/模型架构.md -------------------------------------------------------------------------------- /llm-algo/mixtral/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/mixtral/README.md -------------------------------------------------------------------------------- /llm-algo/mlp.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/mlp.md -------------------------------------------------------------------------------- /llm-algo/moe/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | https://www.promptingguide.ai/models/mixtral 6 | 7 | -------------------------------------------------------------------------------- /llm-algo/qwen/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/qwen/README.md -------------------------------------------------------------------------------- /llm-algo/qwen/参数说明及函数说明.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/qwen/参数说明及函数说明.md -------------------------------------------------------------------------------- /llm-algo/qwen2.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/qwen2.md -------------------------------------------------------------------------------- /llm-algo/t5/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/t5/README.md -------------------------------------------------------------------------------- /llm-algo/transformer.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer.md -------------------------------------------------------------------------------- /llm-algo/transformer/README.md : -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer/README.md -------------------------------------------------------------------------------- /llm-algo/transformer/Transformer中FFN的记忆功能.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer/Transformer中FFN的记忆功能.md -------------------------------------------------------------------------------- /llm-algo/transformer/multi-head-attention.webp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer/multi-head-attention.webp -------------------------------------------------------------------------------- /llm-algo/transformer/transformer-building-blocks.webp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer/transformer-building-blocks.webp -------------------------------------------------------------------------------- /llm-algo/transformer/w-qkv.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer/w-qkv.png -------------------------------------------------------------------------------- /llm-algo/transformer/模型架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/transformer/模型架构.md -------------------------------------------------------------------------------- /llm-algo/基本概念.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/基本概念.md -------------------------------------------------------------------------------- /llm-algo/旋转编码RoPE.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/旋转编码RoPE.md -------------------------------------------------------------------------------- /llm-algo/模型架构类图.drawio: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/模型架构类图.drawio -------------------------------------------------------------------------------- /llm-algo/训练范式.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-algo/训练范式.md -------------------------------------------------------------------------------- /llm-alignment/DPO.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-alignment/DPO.md -------------------------------------------------------------------------------- /llm-alignment/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-alignment/README.md -------------------------------------------------------------------------------- /llm-alignment/RLHF.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-alignment/RLHF.md -------------------------------------------------------------------------------- /llm-alignment/基本概念.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-alignment/基本概念.md -------------------------------------------------------------------------------- /llm-application/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/.DS_Store -------------------------------------------------------------------------------- /llm-application/Higress.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | -------------------------------------------------------------------------------- /llm-application/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/README.md -------------------------------------------------------------------------------- /llm-application/embbedding-model.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 智源的工作bge，llm-embedder 6 | 7 | 段落文本用sentence bert类 8 | 9 | 10 | 11 | 12 | 13 | 14 | -------------------------------------------------------------------------------- /llm-application/gradio/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/gradio/README.md -------------------------------------------------------------------------------- /llm-application/langchain/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/langchain/.DS_Store -------------------------------------------------------------------------------- /llm-application/langchain/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/langchain/README.md -------------------------------------------------------------------------------- /llm-application/langchain/serve.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/langchain/serve.py -------------------------------------------------------------------------------- /llm-application/langchain/tutorials/client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/langchain/tutorials/client.py -------------------------------------------------------------------------------- /llm-application/langchain/tutorials/serve.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/langchain/tutorials/serve.py -------------------------------------------------------------------------------- /llm-application/one-api.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/one-api.md -------------------------------------------------------------------------------- /llm-application/pre-post-handle/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/pre-post-handle/README.md -------------------------------------------------------------------------------- /llm-application/rag/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/rag/README.md -------------------------------------------------------------------------------- /llm-application/rag/embedding.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/rag/embedding.md -------------------------------------------------------------------------------- /llm-application/rag/存在的一些问题.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/rag/存在的一些问题.md -------------------------------------------------------------------------------- /llm-application/rag/方案.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/rag/方案.md -------------------------------------------------------------------------------- /llm-application/vector-db/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/vector-db/README.md -------------------------------------------------------------------------------- /llm-application/vector-db/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/vector-db/reference.md -------------------------------------------------------------------------------- /llm-application/应用场景.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-application/应用场景.md -------------------------------------------------------------------------------- /llm-compression/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/.DS_Store -------------------------------------------------------------------------------- /llm-compression/PaddleSlim/ quantization.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/PaddleSlim/ quantization.md -------------------------------------------------------------------------------- /llm-compression/PaddleSlim/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/PaddleSlim/README.md -------------------------------------------------------------------------------- /llm-compression/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/README.md -------------------------------------------------------------------------------- /llm-compression/distillation/GKD.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/distillation/GKD.md -------------------------------------------------------------------------------- /llm-compression/distillation/MINILLM.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/distillation/MINILLM.md -------------------------------------------------------------------------------- /llm-compression/distillation/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/distillation/README.md -------------------------------------------------------------------------------- /llm-compression/distillation/SCOTT.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/distillation/SCOTT.md -------------------------------------------------------------------------------- /llm-compression/distillation/大模型蒸馏概述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/distillation/大模型蒸馏概述.md -------------------------------------------------------------------------------- /llm-compression/gptqmodel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/gptqmodel/README.md -------------------------------------------------------------------------------- /llm-compression/llm-compressor/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/llm-compressor/README.md -------------------------------------------------------------------------------- /llm-compression/llm-compressor/source-code.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/llm-compressor/source-code.md -------------------------------------------------------------------------------- /llm-compression/llm-compressor/剪枝.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/llm-compressor/剪枝.md -------------------------------------------------------------------------------- /llm-compression/llm-compressor/量化方案.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/llm-compressor/量化方案.md -------------------------------------------------------------------------------- /llm-compression/quantization/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/.DS_Store -------------------------------------------------------------------------------- /llm-compression/quantization/FP6-LLM.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/FP6-LLM.md -------------------------------------------------------------------------------- /llm-compression/quantization/GPTQ.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/GPTQ.md -------------------------------------------------------------------------------- /llm-compression/quantization/LLM-int8.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/LLM-int8.md -------------------------------------------------------------------------------- /llm-compression/quantization/PEQA.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/PEQA.md -------------------------------------------------------------------------------- /llm-compression/quantization/QQQ-W4A8.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/QQQ-W4A8.md -------------------------------------------------------------------------------- /llm-compression/quantization/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/README.md -------------------------------------------------------------------------------- /llm-compression/quantization/SmoothQuant.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/SmoothQuant.md -------------------------------------------------------------------------------- /llm-compression/quantization/SpinQuant.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | -------------------------------------------------------------------------------- /llm-compression/quantization/ZeroQuant(4+2).md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/ZeroQuant(4+2).md -------------------------------------------------------------------------------- /llm-compression/quantization/ZeroQuant.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/ZeroQuant.md -------------------------------------------------------------------------------- /llm-compression/quantization/fp4.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/fp4.md -------------------------------------------------------------------------------- /llm-compression/quantization/fp6.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/fp6.md -------------------------------------------------------------------------------- /llm-compression/quantization/fp8.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/fp8.md -------------------------------------------------------------------------------- /llm-compression/quantization/kv-cache-quant.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/kv-cache-quant.md -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/LLM-QAT.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/LLM-QAT.md -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/README.md -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/README.md -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/generate_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/generate_data.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/inference.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/merge_gen_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/merge_gen_data.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/pip.conf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/pip.conf -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/run_train.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/run_train.sh -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/train.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/cfd70ff/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/cfd70ff/utils.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/f4d873a/datautils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/f4d873a/datautils.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/f4d873a/run_train.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/f4d873a/run_train.sh -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/f4d873a/train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/f4d873a/train.py -------------------------------------------------------------------------------- /llm-compression/quantization/llm-qat/log.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/llm-qat/log.md -------------------------------------------------------------------------------- /llm-compression/quantization/moe模型量化.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/moe模型量化.md -------------------------------------------------------------------------------- /llm-compression/quantization/tools.md: -------------------------------------------------------------------------------- 1 | tools.md -------------------------------------------------------------------------------- /llm-compression/quantization/可视化/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/可视化/README.md -------------------------------------------------------------------------------- /llm-compression/quantization/可视化/qwen_activate_visual.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/可视化/qwen_activate_visual.ipynb -------------------------------------------------------------------------------- /llm-compression/quantization/可视化/qwen_visual.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/可视化/qwen_visual.ipynb -------------------------------------------------------------------------------- /llm-compression/quantization/大模型量化概述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/大模型量化概述.md -------------------------------------------------------------------------------- /llm-compression/quantization/量化基础.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/quantization/量化基础.md -------------------------------------------------------------------------------- /llm-compression/sparsity/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/sparsity/README.md -------------------------------------------------------------------------------- /llm-compression/tools.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/tools.md -------------------------------------------------------------------------------- /llm-compression/大模型压缩综述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-compression/大模型压缩综述.md -------------------------------------------------------------------------------- /llm-compression/经验.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | MoE gate 层对量化敏感，通常不进行量化以便保持完全精度。 6 | 7 | 8 | -------------------------------------------------------------------------------- /llm-data-engineering/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/README.md -------------------------------------------------------------------------------- /llm-data-engineering/dataset/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/dataset/README.md -------------------------------------------------------------------------------- /llm-data-engineering/dataset/baichuan2.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/dataset/baichuan2.md -------------------------------------------------------------------------------- /llm-data-engineering/dataset/chinese-corpus-all.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-data-engineering/dataset/english-corpus-all.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/dataset/english-corpus-all.md -------------------------------------------------------------------------------- /llm-data-engineering/reference.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/baichuan2_test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/baichuan2_test.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/evol-instruct.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | ## Evol-Instruct 7 | 8 | 使用 LLM 而不是人类，自动大规模生产各种难度级别的开放域指令的新方法，以提高 LLM 的性能。 9 | 10 | 11 | 12 | -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/firefly-template.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/firefly-template.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja-demo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/jinja-demo.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja-llm-baichuan.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/jinja-llm-baichuan.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja-llm-baichuan2.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/jinja-llm-baichuan2.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja-llm-bloom.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/jinja-llm-bloom.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja-llm-chatglm3.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/jinja-llm-chatglm3.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja-llm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/jinja-llm.py -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/jinja.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/数据格式设计.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/数据格式设计.md -------------------------------------------------------------------------------- /llm-data-engineering/sft-dataset/数据集格式.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-data-engineering/sft-dataset/数据集格式.md -------------------------------------------------------------------------------- /llm-eval/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/.DS_Store -------------------------------------------------------------------------------- /llm-eval/EvalScope.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/EvalScope.md -------------------------------------------------------------------------------- /llm-eval/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/README.md -------------------------------------------------------------------------------- /llm-eval/eval-data/longtext_L115433-question.txt: -------------------------------------------------------------------------------- 1 | 第58届全球帅哥竞选活动有什么特别之处？ -------------------------------------------------------------------------------- /llm-eval/eval-data/longtext_L115433.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/eval-data/longtext_L115433.txt -------------------------------------------------------------------------------- /llm-eval/eval-data/longtext_L32503_answer.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/eval-data/longtext_L32503_answer.txt -------------------------------------------------------------------------------- /llm-eval/eval-data/longtext_L32503_question.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/eval-data/longtext_L32503_question.txt -------------------------------------------------------------------------------- /llm-eval/eval-data/longtext_L64031.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/eval-data/longtext_L64031.txt -------------------------------------------------------------------------------- /llm-eval/eval-data/longtext_L64031_question.txt: -------------------------------------------------------------------------------- 1 | 第58届全球帅哥竞选活动有什么特别之处？ -------------------------------------------------------------------------------- /llm-eval/evalscope_framework.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/evalscope_framework.png -------------------------------------------------------------------------------- /llm-eval/llm-performance/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/.DS_Store -------------------------------------------------------------------------------- /llm-eval/llm-performance/AI芯片性能.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/AI芯片性能.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/README.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/hardware-performance/gpu-monitor-ui.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/hardware-performance/gpu-monitor-ui.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/llmperf.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/llmperf.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/mindie/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/mindie/.DS_Store -------------------------------------------------------------------------------- /llm-eval/llm-performance/mindie/lantency/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/mindie/lantency/README.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/mindie/lantency/perfermance-stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/mindie/lantency/perfermance-stat.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/mindie/lantency/performance-stream.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/mindie/lantency/performance-stream.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/mindie/lantency/stat_input_token.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/mindie/lantency/stat_input_token.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/mindie/locust-lantency-throughput/示例.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/mindie/locust-lantency-throughput/示例.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/perfetto.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | - https://perfetto.dev/ 5 | 6 | 7 | -------------------------------------------------------------------------------- /llm-eval/llm-performance/stat_gpu_memory.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/stat_gpu_memory.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/tgi-benchmark.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/tgi-benchmark.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/tgi-benchmark.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/tgi-benchmark.png -------------------------------------------------------------------------------- /llm-eval/llm-performance/vllm-benchmark.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/vllm-benchmark.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/vllm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/vllm/README.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/vllm/vllm-locust-qwen1.5-7b-long.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/vllm/vllm-locust-qwen1.5-7b-long.py -------------------------------------------------------------------------------- /llm-eval/llm-performance/wrk-性能测试工具.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/wrk-性能测试工具.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/大模型场景下训练和推理性能指标名词解释.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/大模型场景下训练和推理性能指标名词解释.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/推理性能测试.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/推理性能测试.md -------------------------------------------------------------------------------- /llm-eval/llm-performance/训练性能测试.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-performance/训练性能测试.md -------------------------------------------------------------------------------- /llm-eval/llm-precision/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-precision/.DS_Store -------------------------------------------------------------------------------- /llm-eval/llm-precision/C-Eval.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-precision/C-Eval.md -------------------------------------------------------------------------------- /llm-eval/llm-precision/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-precision/README.md -------------------------------------------------------------------------------- /llm-eval/llm-precision/模型质量评估.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/llm-precision/模型质量评估.md -------------------------------------------------------------------------------- /llm-eval/opencompass.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/opencompass.md -------------------------------------------------------------------------------- /llm-eval/大模型测评集.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-eval/大模型测评集.md -------------------------------------------------------------------------------- /llm-inference/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/.DS_Store -------------------------------------------------------------------------------- /llm-inference/DeepSpeed-Inference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/DeepSpeed-Inference.md -------------------------------------------------------------------------------- /llm-inference/Flash-Decoding.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | Flash-Decoding for long-context inference 4 | 5 | 6 | 7 | -------------------------------------------------------------------------------- /llm-inference/FlashInfer.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/FlashInfer.md -------------------------------------------------------------------------------- /llm-inference/FlexFlow-Serve.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/FlexFlow-Serve.md -------------------------------------------------------------------------------- /llm-inference/GuidedGeneration.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/GuidedGeneration.md -------------------------------------------------------------------------------- /llm-inference/KV-Cache优化.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | https://github.com/NVIDIA/kvpress 5 | 6 | 7 | -------------------------------------------------------------------------------- /llm-inference/Mooncake.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/Mooncake.md -------------------------------------------------------------------------------- /llm-inference/NanoFlow.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | - https://github.com/efeslab/Nanoflow 5 | 6 | 7 | -------------------------------------------------------------------------------- /llm-inference/PD分离.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/PD分离.md -------------------------------------------------------------------------------- /llm-inference/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/README.md -------------------------------------------------------------------------------- /llm-inference/RTP-LLM.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/RTP-LLM.md -------------------------------------------------------------------------------- /llm-inference/ascend/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/.DS_Store -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/.DS_Store -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/README.md -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/baichuan2/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/baichuan2/README.md -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/baichuan2/baichuan-inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/baichuan2/baichuan-inference.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/baichuan2/baichuan-stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/baichuan2/baichuan-stat.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/chatglm3/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/chatglm3/chatglm-gen.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/chatglm3/chatglm-gen.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/chatglm3/chatglm-inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/chatglm3/chatglm-inference.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/chatglm3/chatglm-stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/chatglm3/chatglm-stat.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/mindsporelite-inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/mindsporelite-inference.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/mindsporelite-stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/mindsporelite-stat.py -------------------------------------------------------------------------------- /llm-inference/ascend/mindformers/text_generator_infer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/ascend/mindformers/text_generator_infer.py -------------------------------------------------------------------------------- /llm-inference/chatgpt.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/chatgpt.md -------------------------------------------------------------------------------- /llm-inference/deepspeed-mii/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/deepspeed-mii/README.md -------------------------------------------------------------------------------- /llm-inference/faster-transformer/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/faster-transformer/README.md -------------------------------------------------------------------------------- /llm-inference/faster-transformer/bloom/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/faster-transformer/bloom/README.md -------------------------------------------------------------------------------- /llm-inference/faster-transformer/gpt/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-inference/faster-transformer/llama/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/faster-transformer/llama/README.md -------------------------------------------------------------------------------- /llm-inference/flexflow-serve/benchmark-batch1.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/flexflow-serve/benchmark-batch1.py -------------------------------------------------------------------------------- /llm-inference/huggingface-tgi/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/huggingface-tgi/README.md -------------------------------------------------------------------------------- /llm-inference/huggingface-transformer/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/huggingface-transformer/README.md -------------------------------------------------------------------------------- /llm-inference/lightllm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/lightllm/README.md -------------------------------------------------------------------------------- /llm-inference/lmdeploy/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/lmdeploy/README.md -------------------------------------------------------------------------------- /llm-inference/lmdeploy/功能.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/lmdeploy/功能.md -------------------------------------------------------------------------------- /llm-inference/lmdeploy/服务启动参数.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/lmdeploy/服务启动参数.md -------------------------------------------------------------------------------- /llm-inference/native-model/chatglm3-6b/cli_demo.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/native-model/chatglm3-6b/cli_demo.py -------------------------------------------------------------------------------- /llm-inference/offload.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/offload.md -------------------------------------------------------------------------------- /llm-inference/openai.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/openai.md -------------------------------------------------------------------------------- /llm-inference/sglang/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/sglang/README.md -------------------------------------------------------------------------------- /llm-inference/sglang/source-code.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/sglang/source-code.md -------------------------------------------------------------------------------- /llm-inference/sglang/服务器启动参数.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/sglang/服务器启动参数.md -------------------------------------------------------------------------------- /llm-inference/sglang/项目代码结构.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /llm-inference/tensorrt-llm/FP8.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt-llm/FP8.md -------------------------------------------------------------------------------- /llm-inference/tensorrt-llm/Memory Usage of TensorRT-LLM.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt-llm/Memory Usage of TensorRT-LLM.md -------------------------------------------------------------------------------- /llm-inference/tensorrt-llm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt-llm/README.md -------------------------------------------------------------------------------- /llm-inference/tensorrt-llm/TRT-LLM引擎构建参数.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt-llm/TRT-LLM引擎构建参数.md -------------------------------------------------------------------------------- /llm-inference/tensorrt-llm/Triton服务启动参数.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /llm-inference/tensorrt-llm/安装.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt-llm/安装.md -------------------------------------------------------------------------------- /llm-inference/tensorrt/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt/README.md -------------------------------------------------------------------------------- /llm-inference/tensorrt/install.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/tensorrt/install.md -------------------------------------------------------------------------------- /llm-inference/triton/REAEME.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/triton/REAEME.md -------------------------------------------------------------------------------- /llm-inference/triton/onnx/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/triton/onnx/README.md -------------------------------------------------------------------------------- /llm-inference/triton/resnet50/client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/triton/resnet50/client.py -------------------------------------------------------------------------------- /llm-inference/triton/resnet50/config.pbtxt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/triton/resnet50/config.pbtxt -------------------------------------------------------------------------------- /llm-inference/triton/resnet50/labels.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/triton/resnet50/labels.txt -------------------------------------------------------------------------------- /llm-inference/triton/resnet50/resnet50_convert_torchscript.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/triton/resnet50/resnet50_convert_torchscript.py -------------------------------------------------------------------------------- /llm-inference/vllm/FAQ.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/FAQ.md -------------------------------------------------------------------------------- /llm-inference/vllm/FP8.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/FP8.md -------------------------------------------------------------------------------- /llm-inference/vllm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/README.md -------------------------------------------------------------------------------- /llm-inference/vllm/REF.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | -------------------------------------------------------------------------------- /llm-inference/vllm/api_client.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/api_client.py -------------------------------------------------------------------------------- /llm-inference/vllm/cmd.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/cmd.md -------------------------------------------------------------------------------- /llm-inference/vllm/vllm.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/vllm.md -------------------------------------------------------------------------------- /llm-inference/vllm/服务启动参数.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/服务启动参数.md -------------------------------------------------------------------------------- /llm-inference/vllm/源码.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/源码.md -------------------------------------------------------------------------------- /llm-inference/vllm/请求处理流程.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/请求处理流程.md -------------------------------------------------------------------------------- /llm-inference/vllm/长文本推理.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/vllm/长文本推理.md -------------------------------------------------------------------------------- /llm-inference/web/fastapi/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/web/fastapi/README.md -------------------------------------------------------------------------------- /llm-inference/web/fastapi/llm-qwen-mindspore-lite.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/web/fastapi/llm-qwen-mindspore-lite.py -------------------------------------------------------------------------------- /llm-inference/web/flask/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/web/flask/README.md -------------------------------------------------------------------------------- /llm-inference/web/flask/llm-qwen-mindspore-lite.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/web/flask/llm-qwen-mindspore-lite.py -------------------------------------------------------------------------------- /llm-inference/web/sanic/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/web/sanic/README.md -------------------------------------------------------------------------------- /llm-inference/xinference/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/xinference/README.md -------------------------------------------------------------------------------- /llm-inference/分离式推理架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/分离式推理架构.md -------------------------------------------------------------------------------- /llm-inference/大模型推理张量并行.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/大模型推理张量并行.md -------------------------------------------------------------------------------- /llm-inference/解码策略.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-inference/解码策略.md -------------------------------------------------------------------------------- /llm-interview/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/README.md -------------------------------------------------------------------------------- /llm-interview/base.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/base.md -------------------------------------------------------------------------------- /llm-interview/comprehensive.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | - 解决显存不足的方法有哪些？ 5 | 6 | 训练： 7 | 8 | 推理： 9 | 10 | - 11 | 12 | 13 | 14 | -------------------------------------------------------------------------------- /llm-interview/llm-algo.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-algo.md -------------------------------------------------------------------------------- /llm-interview/llm-app.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /llm-interview/llm-compress.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-compress.md -------------------------------------------------------------------------------- /llm-interview/llm-eval.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-eval.md -------------------------------------------------------------------------------- /llm-interview/llm-ft.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-ft.md -------------------------------------------------------------------------------- /llm-interview/llm-inference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-inference.md -------------------------------------------------------------------------------- /llm-interview/llm-rlhf.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-rlhf.md -------------------------------------------------------------------------------- /llm-interview/llm-train.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-interview/llm-train.md -------------------------------------------------------------------------------- /llm-localization/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/.DS_Store -------------------------------------------------------------------------------- /llm-localization/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/.DS_Store -------------------------------------------------------------------------------- /llm-localization/ascend/FAQ.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/FAQ.md -------------------------------------------------------------------------------- /llm-localization/ascend/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-c/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-c/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/HCCL.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/HCCL.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/MacOS环境.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/MacOS环境.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/ascend-dmi.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/ascend-dmi.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/ascend-docker-runtime.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/ascend-docker-runtime.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/ascend-docker.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/ascend-docker.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/ascend-llm下载.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/ascend-llm下载.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/ascend-npu-smi.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/ascend-npu-smi.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/docker环境升级cann.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/docker环境升级cann.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/network.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/network.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/npu监控.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/npu监控.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/操作系统.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/操作系统.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/昇腾卡-soc版本.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/昇腾卡-soc版本.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/昇腾卡注意事项.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/昇腾卡注意事项.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/昇腾镜像.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/昇腾镜像.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/服务器配置.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/服务器配置.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/环境安装.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/环境安装.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend-infra/达芬奇架构.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend-infra/达芬奇架构.md -------------------------------------------------------------------------------- /llm-localization/ascend/ascend910-env-install.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/ascend910-env-install.md -------------------------------------------------------------------------------- /llm-localization/ascend/fabric-insight/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /llm-localization/ascend/firefly-ascend.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/firefly-ascend.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/.DS_Store -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/baichuan2/baichuan2训练.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/baichuan2/baichuan2训练.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/chat_glm.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/chat_glm.py -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/glm_6b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/glm_6b.yaml -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/glm_6b_chat.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/glm_6b_chat.yaml -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/merge_ckpt.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/merge_ckpt.py -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/merge_ckpt_lora.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/merge_ckpt_lora.py -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/chatglm/pt2ms.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/chatglm/pt2ms.py -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/env.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/env.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/llama/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/llama/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/qwen/qwen1训练.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/qwen/qwen1训练.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/qwen/run_qwen_7b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/qwen/run_qwen_7b.yaml -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/qwen/run_qwen_7b_910b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/qwen/run_qwen_7b_910b.yaml -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/qwen1.5/qwen1.5训练.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/qwen1.5/qwen1.5训练.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/trick.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/trick.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindformers/权重格式转换.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindformers/权重格式转换.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/.DS_Store -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/2.0.RC2/qwen.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/2.0.RC2/qwen.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/config-1.0.RC1.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/config-1.0.RC1.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/config/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/config/.DS_Store -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/config/chatglm3-6b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/config/chatglm3-6b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/config/qwen-72b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/config/qwen-72b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/config/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/config/run.sh -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/TEST.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/TEST.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/baichuan2-13b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/baichuan2-13b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/baichuan2-7b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/baichuan2-7b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/deploy.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/deploy.sh -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/llm-server.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/llm-server.sh -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/mindie-1.0.Dockerfile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/mindie-1.0.Dockerfile -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/qwen-72b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/qwen-72b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/qwen1.5-14b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/qwen1.5-14b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/qwen1.5-72b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/qwen1.5-72b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/docker/qwen1.5-7b.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/docker/qwen1.5-7b.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/llm-server.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/llm-server.sh -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindid-1.0-offical.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindid-1.0-offical.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindid-performance.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindid-performance.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-1.0.Dockerfile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-1.0.Dockerfile -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-1.0.RC2.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-1.0.RC2.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-1.0.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-1.0.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-1.0.rc2-config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-1.0.rc2-config.json -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-1.0.rc2-llm-server.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-1.0.rc2-llm-server.sh -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-2.0.rc2.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-2.0.rc2.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-20240411.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-20240411.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/mindie-api.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/mindie-api.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/model-test.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/model-test.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/script/model-test.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/script/model-test.py -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/script/run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/script/run.sh -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/性能调优.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/性能调优.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindie/日志分析.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindie/日志分析.txt -------------------------------------------------------------------------------- /llm-localization/ascend/mindspore/MindSpore-note.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindspore/MindSpore-note.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindspore/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindspore/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindspore/bert.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindspore/bert.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindspore/reference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindspore/reference.md -------------------------------------------------------------------------------- /llm-localization/ascend/mindspore/镜像.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/mindspore/镜像.md -------------------------------------------------------------------------------- /llm-localization/ascend/modellink/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/modellink/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/modellink/dataset.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/modellink/dataset.md -------------------------------------------------------------------------------- /llm-localization/ascend/modellink/llm.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/modellink/llm.md -------------------------------------------------------------------------------- /llm-localization/ascend/modellink/qwen.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/modellink/qwen.md -------------------------------------------------------------------------------- /llm-localization/ascend/modellink/环境-20240521.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/modellink/环境-20240521.md -------------------------------------------------------------------------------- /llm-localization/ascend/modellink/环境安装.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/modellink/环境安装.md -------------------------------------------------------------------------------- /llm-localization/ascend/msmodelslim/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/msmodelslim/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/msmodelslim/llm_quant/calib_set.json: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-localization/ascend/openmind/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/openmind/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/peft/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/peft/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/peft/finetune-lora.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/peft/finetune-lora.py -------------------------------------------------------------------------------- /llm-localization/ascend/pytorch/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/pytorch/.DS_Store -------------------------------------------------------------------------------- /llm-localization/ascend/pytorch/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/pytorch/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/pytorch/llm-lora.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/pytorch/llm-lora.py -------------------------------------------------------------------------------- /llm-localization/ascend/standford-alpaca/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/standford-alpaca/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/standford-alpaca/ds_config_zero2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/standford-alpaca/ds_config_zero2.json -------------------------------------------------------------------------------- /llm-localization/ascend/standford-alpaca/ds_config_zero3.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/standford-alpaca/ds_config_zero3.json -------------------------------------------------------------------------------- /llm-localization/ascend/standford-alpaca/requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/standford-alpaca/requirements.txt -------------------------------------------------------------------------------- /llm-localization/ascend/standford-alpaca/train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/standford-alpaca/train.py -------------------------------------------------------------------------------- /llm-localization/ascend/standford-alpaca/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/standford-alpaca/utils.py -------------------------------------------------------------------------------- /llm-localization/ascend/transformers/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/transformers/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/vllm-ascend/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/vllm-ascend/README.md -------------------------------------------------------------------------------- /llm-localization/ascend/优质学习资料.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/优质学习资料.md -------------------------------------------------------------------------------- /llm-localization/ascend/昇腾LLM支持概览.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/昇腾LLM支持概览.md -------------------------------------------------------------------------------- /llm-localization/ascend/昇腾卡注意事项.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/ascend/昇腾卡注意事项.md -------------------------------------------------------------------------------- /llm-localization/modelscope/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/modelscope/README.md -------------------------------------------------------------------------------- /llm-localization/paddle/PaddleNLP.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/paddle/PaddleNLP.md -------------------------------------------------------------------------------- /llm-localization/tianshuzhixin/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-localization/tianshuzhixin/ixsmi.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-localization/tianshuzhixin/ixsmi.md -------------------------------------------------------------------------------- /llm-maas/OpenAI-ChatGPT.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-maas/OpenAI-ChatGPT.md -------------------------------------------------------------------------------- /llm-maas/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-maas/README.md -------------------------------------------------------------------------------- /llm-optimizer/FlashAttention.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-optimizer/FlashAttention.md -------------------------------------------------------------------------------- /llm-optimizer/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-optimizer/README.md -------------------------------------------------------------------------------- /llm-optimizer/SplitFuse.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-optimizer/SplitFuse.md -------------------------------------------------------------------------------- /llm-optimizer/kv-cache.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-optimizer/kv-cache.md -------------------------------------------------------------------------------- /llm-optimizer/xformers.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | https://github.com/facebookresearch/xformers 4 | 5 | 6 | -------------------------------------------------------------------------------- /llm-optimizer/计算通信重叠.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-optimizer/计算通信重叠.md -------------------------------------------------------------------------------- /llm-pipeline/REAEMD.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-pipeline/REAEMD.md -------------------------------------------------------------------------------- /llm-tools/Pytorch-Profiler.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/Pytorch-Profiler.md -------------------------------------------------------------------------------- /llm-tools/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/README.md -------------------------------------------------------------------------------- /llm-tools/base-profiler.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/base-profiler.py -------------------------------------------------------------------------------- /llm-tools/nsight.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/nsight.md -------------------------------------------------------------------------------- /llm-tools/nsight/0-GPU性能分析.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/nsight/0-GPU性能分析.png -------------------------------------------------------------------------------- /llm-tools/nsight/1-GPU利用率低.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/nsight/1-GPU利用率低.png -------------------------------------------------------------------------------- /llm-tools/nsight/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/nsight/README.md -------------------------------------------------------------------------------- /llm-tools/nvtx.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/nvtx.md -------------------------------------------------------------------------------- /llm-tools/profiler-recipe.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/profiler-recipe.py -------------------------------------------------------------------------------- /llm-tools/tensorboard-profiler.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/tensorboard-profiler.py -------------------------------------------------------------------------------- /llm-tools/可视化.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-tools/可视化.md -------------------------------------------------------------------------------- /llm-train/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/.DS_Store -------------------------------------------------------------------------------- /llm-train/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/README.md -------------------------------------------------------------------------------- /llm-train/alpa/train/pipeshard_parallelism.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpa/train/pipeshard_parallelism.ipynb -------------------------------------------------------------------------------- /llm-train/alpa/train/pipeshard_parallelism.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpa/train/pipeshard_parallelism.py -------------------------------------------------------------------------------- /llm-train/alpaca-lora/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/README.md -------------------------------------------------------------------------------- /llm-train/alpaca-lora/export_hf_checkpoint.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/export_hf_checkpoint.py -------------------------------------------------------------------------------- /llm-train/alpaca-lora/export_state_dict_checkpoint.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/export_state_dict_checkpoint.py -------------------------------------------------------------------------------- /llm-train/alpaca-lora/finetune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/finetune.py -------------------------------------------------------------------------------- /llm-train/alpaca-lora/finetune_metrics_epoch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/finetune_metrics_epoch.py -------------------------------------------------------------------------------- /llm-train/alpaca-lora/generate.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/generate.py -------------------------------------------------------------------------------- /llm-train/alpaca-lora/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca-lora/inference.py -------------------------------------------------------------------------------- /llm-train/alpaca/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/README.md -------------------------------------------------------------------------------- /llm-train/alpaca/ds_config.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/ds_config.json -------------------------------------------------------------------------------- /llm-train/alpaca/ds_config_zero2.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/ds_config_zero2.json -------------------------------------------------------------------------------- /llm-train/alpaca/ds_config_zero2_ddp.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/ds_config_zero2_ddp.json -------------------------------------------------------------------------------- /llm-train/alpaca/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/inference.py -------------------------------------------------------------------------------- /llm-train/alpaca/train.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/train.py -------------------------------------------------------------------------------- /llm-train/alpaca/train_ddp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/alpaca/train_ddp.py -------------------------------------------------------------------------------- /llm-train/ascend/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/ascend/.DS_Store -------------------------------------------------------------------------------- /llm-train/chatglm-lora/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm-lora/README.md -------------------------------------------------------------------------------- /llm-train/chatglm-lora/finetune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm-lora/finetune.py -------------------------------------------------------------------------------- /llm-train/chatglm-lora/finetune_ddp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm-lora/finetune_ddp.py -------------------------------------------------------------------------------- /llm-train/chatglm-lora/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm-lora/inference.py -------------------------------------------------------------------------------- /llm-train/chatglm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/README.md -------------------------------------------------------------------------------- /llm-train/chatglm/deepspeed.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/deepspeed.json -------------------------------------------------------------------------------- /llm-train/chatglm/ds_train_finetune.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/ds_train_finetune.sh -------------------------------------------------------------------------------- /llm-train/chatglm/evaluate.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/evaluate.sh -------------------------------------------------------------------------------- /llm-train/chatglm/evaluate_finetune.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/evaluate_finetune.sh -------------------------------------------------------------------------------- /llm-train/chatglm/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/inference.py -------------------------------------------------------------------------------- /llm-train/chatglm/main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/main.py -------------------------------------------------------------------------------- /llm-train/chatglm/train.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/train.sh -------------------------------------------------------------------------------- /llm-train/chatglm/train_ptuningv2_dp.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chatglm/train_ptuningv2_dp.sh -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/README.md -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/inference_hf.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/inference_hf.py -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/merge_tokenizers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/merge_tokenizers.py -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/run_clm_pt_with_peft.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/run_clm_pt_with_peft.py -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/run_clm_sft_with_peft.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/run_clm_sft_with_peft.py -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/run_pt.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/run_pt.sh -------------------------------------------------------------------------------- /llm-train/chinese-llama-alpaca/run_sft.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/chinese-llama-alpaca/run_sft.sh -------------------------------------------------------------------------------- /llm-train/deepspeedchat/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/deepspeedchat/README.md -------------------------------------------------------------------------------- /llm-train/deepspeedchat/llama/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/deepspeedchat/llama/README.md -------------------------------------------------------------------------------- /llm-train/deepspeedchat/training/utils/data/raw_datasets.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/deepspeedchat/training/utils/data/raw_datasets.py -------------------------------------------------------------------------------- /llm-train/firefly/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/firefly/README.md -------------------------------------------------------------------------------- /llm-train/firefly/bootstrap-s3.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/firefly/bootstrap-s3.sh -------------------------------------------------------------------------------- /llm-train/firefly/bootstrap.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/firefly/bootstrap.sh -------------------------------------------------------------------------------- /llm-train/firefly/dockerfile.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/firefly/dockerfile.md -------------------------------------------------------------------------------- /llm-train/firefly/test_bash_getopts.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/firefly/test_bash_getopts.sh -------------------------------------------------------------------------------- /llm-train/fp8.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/fp8.md -------------------------------------------------------------------------------- /llm-train/galore/torchrun_main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/galore/torchrun_main.py -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/README.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/bigscience/bloom-note.md: -------------------------------------------------------------------------------- 1 | 2 | -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/bloom-megatron-deepspeed.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/bloom-megatron-deepspeed.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/H800多机多卡训练坑点.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/H800多机多卡训练坑点.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/README.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/llama-note.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/llama-note.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/pip.conf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/pip.conf -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/slurm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/slurm/README.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/代码.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/代码.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/环境准备.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/环境准备.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/训练日志分析.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/训练日志分析.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/microsoft/项目结构-202312228.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/microsoft/项目结构-202312228.md -------------------------------------------------------------------------------- /llm-train/megatron-deepspeed/source-code.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron-deepspeed/source-code.md -------------------------------------------------------------------------------- /llm-train/megatron/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/README.md -------------------------------------------------------------------------------- /llm-train/megatron/codegeex/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/codegeex/README.md -------------------------------------------------------------------------------- /llm-train/megatron/codegeex/pic/CodeGeeX模型架构.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/codegeex/pic/CodeGeeX模型架构.png -------------------------------------------------------------------------------- /llm-train/megatron/codegeex/pic/CodeGeeX训练配置.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/codegeex/pic/CodeGeeX训练配置.png -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/README.md -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/data/cMinhash.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/data/cMinhash.cpp -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/data/download.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/data/download.py -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/data/file_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/data/file_utils.py -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/data/merge_data.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/data/merge_data.py -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/gpt-data-preprocess.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/gpt-data-preprocess.md -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/merge_ck_and_inference/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/merge_ck_and_inference/README.md -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/model_merge_eval_inference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/model_merge_eval_inference.md -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/model_train.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/model_train.md -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/requirements.txt -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/train/pretrain_gpt.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/train/pretrain_gpt.sh -------------------------------------------------------------------------------- /llm-train/megatron/gpt2/train/pretrain_gpt_distributed.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/gpt2/train/pretrain_gpt_distributed.sh -------------------------------------------------------------------------------- /llm-train/megatron/kernel_fusion.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/kernel_fusion.png -------------------------------------------------------------------------------- /llm-train/megatron/megatron.drawio: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/megatron.drawio -------------------------------------------------------------------------------- /llm-train/megatron/pretrain.xmind: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/pretrain.xmind -------------------------------------------------------------------------------- /llm-train/megatron/project.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/project.md -------------------------------------------------------------------------------- /llm-train/megatron/source-code.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/megatron/source-code.md -------------------------------------------------------------------------------- /llm-train/paddle/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/paddle/README.md -------------------------------------------------------------------------------- /llm-train/paddle/paddlenlp/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/paddle/paddlenlp/README.md -------------------------------------------------------------------------------- /llm-train/paddle/paddlenlp/baichuan2/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/paddle/paddlenlp/baichuan2/README.md -------------------------------------------------------------------------------- /llm-train/paddle/paddlenlp/bloom/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/paddle/paddlenlp/bloom/README.md -------------------------------------------------------------------------------- /llm-train/paddle/paddlenlp/bloom/sft_argument.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/paddle/paddlenlp/bloom/sft_argument.json -------------------------------------------------------------------------------- /llm-train/peft/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/.DS_Store -------------------------------------------------------------------------------- /llm-train/peft/LoRA-QLoRA.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/LoRA-QLoRA.md -------------------------------------------------------------------------------- /llm-train/peft/PEFT-API.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/PEFT-API.md -------------------------------------------------------------------------------- /llm-train/peft/Prefix-Tuning.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/Prefix-Tuning.md -------------------------------------------------------------------------------- /llm-train/peft/Prompt-Tuning.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/Prompt-Tuning.md -------------------------------------------------------------------------------- /llm-train/peft/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/README.md -------------------------------------------------------------------------------- /llm-train/peft/clm/accelerate_ds_zero3_cpu_offload_config.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/accelerate_ds_zero3_cpu_offload_config.yaml -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_ia3_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_ia3_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_lora_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_lora_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_p_tuning_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_p_tuning_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_p_tuning_lstm_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_p_tuning_lstm_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_p_tuning_v2_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_p_tuning_v2_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_prefix_tuning_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_prefix_tuning_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/clm/peft_prompt_tuning_clm.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/clm/peft_prompt_tuning_clm.ipynb -------------------------------------------------------------------------------- /llm-train/peft/conditional_generation/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-train/peft/multimodal/blip2_lora_inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/multimodal/blip2_lora_inference.py -------------------------------------------------------------------------------- /llm-train/peft/multimodal/blip2_lora_int8_fine_tune.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/multimodal/blip2_lora_int8_fine_tune.py -------------------------------------------------------------------------------- /llm-train/peft/multimodal/finetune_bloom_bnb_peft.ipynb: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/peft/multimodal/finetune_bloom_bnb_peft.ipynb -------------------------------------------------------------------------------- /llm-train/pytorch/Pytorch源码解读.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/Pytorch源码解读.md -------------------------------------------------------------------------------- /llm-train/pytorch/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/README.md -------------------------------------------------------------------------------- /llm-train/pytorch/api.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/api.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/README.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/api.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/data-parallel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/data-parallel/README.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/data-parallel/ddp_launch.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/data-parallel/ddp_launch.py -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/data-parallel/ddp_main.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/data-parallel/ddp_main.py -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/data-parallel/elastic_ddp.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/data-parallel/elastic_ddp.py -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/data-parallel/sbatch_run.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/data-parallel/sbatch_run.sh -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/data-parallel/使用DDP训练真实世界的模型.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/data-parallel/使用DDP训练真实世界的模型.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/pipeline-parallel/1-流水线.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/pipeline-parallel/1-流水线.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/pipeline-parallel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/pipeline-parallel/README.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/rpc/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/rpc/README.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/sequence-parallelism/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/tensor-parallel/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/tensor-parallel/README.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/tensor-parallel/utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/tensor-parallel/utils.py -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/torchrun.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/torchrun.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/分布式通信包.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/分布式通信包.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/多机多卡.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/多机多卡.md -------------------------------------------------------------------------------- /llm-train/pytorch/distribution/多机训练.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/distribution/多机训练.md -------------------------------------------------------------------------------- /llm-train/pytorch/resource.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/resource.md -------------------------------------------------------------------------------- /llm-train/pytorch/torchrun.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/pytorch/torchrun.md -------------------------------------------------------------------------------- /llm-train/qlora/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/README.md -------------------------------------------------------------------------------- /llm-train/qlora/accuracy.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/accuracy.py -------------------------------------------------------------------------------- /llm-train/qlora/export_hf_checkpoint.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/export_hf_checkpoint.py -------------------------------------------------------------------------------- /llm-train/qlora/inference.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/inference.py -------------------------------------------------------------------------------- /llm-train/qlora/inference_merge.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/inference_merge.py -------------------------------------------------------------------------------- /llm-train/qlora/inference_qlora.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/inference_qlora.py -------------------------------------------------------------------------------- /llm-train/qlora/qlora.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/qlora/qlora.py -------------------------------------------------------------------------------- /llm-train/slurm/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/README.md -------------------------------------------------------------------------------- /llm-train/slurm/deepspeed/pp-multinode-machine.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/deepspeed/pp-multinode-machine.slurm -------------------------------------------------------------------------------- /llm-train/slurm/deepspeed/pp-multinode-singularity.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/deepspeed/pp-multinode-singularity.slurm -------------------------------------------------------------------------------- /llm-train/slurm/deepspeed/pp-mutinode-singularity-pmix.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/deepspeed/pp-mutinode-singularity-pmix.slurm -------------------------------------------------------------------------------- /llm-train/slurm/deepspeed/pp-standalone-singularity-v2.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/deepspeed/pp-standalone-singularity-v2.slurm -------------------------------------------------------------------------------- /llm-train/slurm/deepspeed/pp-standalone-singularity.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/deepspeed/pp-standalone-singularity.slurm -------------------------------------------------------------------------------- /llm-train/slurm/pytorch/alpaca-docker.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/pytorch/alpaca-docker.slurm -------------------------------------------------------------------------------- /llm-train/slurm/pytorch/alpaca-machine.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/pytorch/alpaca-machine.slurm -------------------------------------------------------------------------------- /llm-train/slurm/pytorch/alpaca-singularity.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/pytorch/alpaca-singularity.slurm -------------------------------------------------------------------------------- /llm-train/slurm/pytorch/mingpt-singularity-multinode-2.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/pytorch/mingpt-singularity-multinode-2.slurm -------------------------------------------------------------------------------- /llm-train/slurm/pytorch/mingpt-singularity-multinode.slurm: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llm-train/slurm/pytorch/mingpt-singularity-multinode.slurm -------------------------------------------------------------------------------- /llm-train/vicuna/README.md: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /llmops/FAQ.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/FAQ.md -------------------------------------------------------------------------------- /llmops/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/README.md -------------------------------------------------------------------------------- /llmops/kubernetes.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/kubernetes.md -------------------------------------------------------------------------------- /llmops/tq-llm/train/FAQ.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/FAQ.md -------------------------------------------------------------------------------- /llmops/tq-llm/train/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/README.md -------------------------------------------------------------------------------- /llmops/tq-llm/train/bootstrap-llm-zero3-offload.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/bootstrap-llm-zero3-offload.sh -------------------------------------------------------------------------------- /llmops/tq-llm/train/bootstrap-llm.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/bootstrap-llm.sh -------------------------------------------------------------------------------- /llmops/tq-llm/train/bootstrap-llm2.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/bootstrap-llm2.sh -------------------------------------------------------------------------------- /llmops/tq-llm/train/zero2-offload.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/zero2-offload.json -------------------------------------------------------------------------------- /llmops/tq-llm/train/zero3-offload.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/tq-llm/train/zero3-offload.json -------------------------------------------------------------------------------- /llmops/使用docker进行多机多卡训练.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | - Docker容器中DeepSpeed多机多卡集群分布式训练大模型实践：https://cloud.baidu.com/article/3273769 -------------------------------------------------------------------------------- /llmops/千帆大模型平台.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/千帆大模型平台.md -------------------------------------------------------------------------------- /llmops/模型推理平台方案.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/llmops/模型推理平台方案.md -------------------------------------------------------------------------------- /mkdir-dir-file.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/mkdir-dir-file.sh -------------------------------------------------------------------------------- /paper/A Survey on Efficient Training of Transformers.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/A Survey on Efficient Training of Transformers.md -------------------------------------------------------------------------------- /paper/LESS-选择有影响力的数据进行目标指令精调.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/LESS-选择有影响力的数据进行目标指令精调.md -------------------------------------------------------------------------------- /paper/LLM增强LLMS.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/LLM增强LLMS.md -------------------------------------------------------------------------------- /paper/PagedAttention.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/PagedAttention.md -------------------------------------------------------------------------------- /paper/README.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | -------------------------------------------------------------------------------- /paper/data/LESS 实践：仅用少量的数据完成目标指令微调.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/data/LESS 实践：仅用少量的数据完成目标指令微调.md -------------------------------------------------------------------------------- /paper/data/LESS-选择有影响力的数据进行目标指令精调.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/data/LESS-选择有影响力的数据进行目标指令精调.md -------------------------------------------------------------------------------- /paper/data/LESS.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/data/LESS.md -------------------------------------------------------------------------------- /paper/inference/llm-in-a-flash.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/inference/llm-in-a-flash.md -------------------------------------------------------------------------------- /paper/inference/orca.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/inference/orca.md -------------------------------------------------------------------------------- /paper/inference/迈向高效的生成式大语言模型服务综述-从算法到系统.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/inference/迈向高效的生成式大语言模型服务综述-从算法到系统.md -------------------------------------------------------------------------------- /paper/llm对齐综述.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/llm对齐综述.md -------------------------------------------------------------------------------- /paper/moe/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/moe/README.md -------------------------------------------------------------------------------- /paper/parameter-pruning/LLM-Pruner.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/parameter-pruning/LLM-Pruner.md -------------------------------------------------------------------------------- /paper/parameter-pruning/SparseGPT.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/parameter-pruning/SparseGPT.md -------------------------------------------------------------------------------- /paper/parameter-pruning/Wanda.md: -------------------------------------------------------------------------------- 1 | 2 | 3 | -------------------------------------------------------------------------------- /paper/parameter-pruning/公式.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/parameter-pruning/公式.md -------------------------------------------------------------------------------- /paper/training/GaLore.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/paper/training/GaLore.md -------------------------------------------------------------------------------- /pic/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/.DS_Store -------------------------------------------------------------------------------- /pic/damodel.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/damodel.jpeg -------------------------------------------------------------------------------- /pic/llm-action-v3.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm-action-v3.png -------------------------------------------------------------------------------- /pic/llm-action-v4.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm-action-v4.jpg -------------------------------------------------------------------------------- /pic/llm/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm/.DS_Store -------------------------------------------------------------------------------- /pic/llm/model/llm-famliy.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm/model/llm-famliy.jpg -------------------------------------------------------------------------------- /pic/llm/model/llm-timeline-v2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm/model/llm-timeline-v2.png -------------------------------------------------------------------------------- /pic/llm/train/.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm/train/.DS_Store -------------------------------------------------------------------------------- /pic/llm/train/pretrain/llm-pretrain-pipeline-v2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm/train/pretrain/llm-pretrain-pipeline-v2.png -------------------------------------------------------------------------------- /pic/llm/train/sft/peft方法.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/llm/train/sft/peft方法.jpg -------------------------------------------------------------------------------- /pic/wechat.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/wechat.jpeg -------------------------------------------------------------------------------- /pic/wx-gzh.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/wx-gzh.png -------------------------------------------------------------------------------- /pic/wx.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/wx.jpg -------------------------------------------------------------------------------- /pic/公众号.jpeg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/pic/公众号.jpeg -------------------------------------------------------------------------------- /template/server.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/liguodongiot/llm-action/HEAD/template/server.md --------------------------------------------------------------------------------