# Model-Survey
Model Survey mainly includes Universal LLMs and Domain LLMs, In terms of Universal LLMs, including text generation model, image and video generation model, code generation model, music generation model, multimodal model;In terms of Domain LLMs, including law, medical, finance, environment, network security, education, Traffic and so on.
从GPT3到ChatGPT模型的发展路线图
![ChatGPT_family](https://i.postimg.cc/GtZmmjG2/chatgpt-3.jpg)
## baichuan Alternatives
| Target Model | Release Date | Source Model | Optimization | Checkpoints | Paper/Blog | Params (B) | Context Length | Code | Tokens | Tokenizer | Vocab size | Position Embedding | Layer Normalization | Activation Function | Attention |
| --- | --- | --- | --- |--- | --- | --- |--- | --- | --- | --- | --- | --- | --- | --- | --- |
| baichuan-7b | 2023/6/15 | | | [Model Scope](https://modelscope.cn/models/baichuan-inc/baichuan-7B/summary),[hugging face](https://huggingface.co/baichuan-inc/baichuan-7B) | [blog](https://mp.weixin.qq.com/s/qA_E_3dUe1sSOUM87ZgHdQ) | [7](https://github.com/ArronAI007/Awesome-AGI/tree/main/Model-List/model-params.md) | 4096 | [baichuan-7b Code](https://github.com/baichuan-inc/baichuan-7B),[baichuan-7b Demo](https://huggingface.co/baichuan-inc/baichuan-7B) |1.2T | BPE | 64000 | RoPE | Pre RMS Norm | SwiGLU | Flash-attention |
| baichuan-13b | 2023/7/11 | | | [hugging face Base](https://huggingface.co/baichuan-inc/Baichuan-13B-Base),[hugging face Chat](https://huggingface.co/baichuan-inc/Baichuan-13B-Chat),[modelscope Base](https://modelscope.cn/models/baichuan-inc/Baichuan-13B-Base),[modelscope Chat](https://modelscope.cn/models/baichuan-inc/Baichuan-13B-Chat) | [baichuan-13b blog](https://mp.weixin.qq.com/s/Px4h2r3VIAFI5vfjXxROxg),[百川大模型【Baichuan-13B】 多卡训练微调记录](https://mp.weixin.qq.com/s/EUZA6Lt-OcI170md9lXH1g) | [13](https://github.com/ArronAI007/Awesome-AGI/tree/main/Model-List/model-params.md) | 4096 | [baichuan-13b Code](https://github.com/baichuan-inc/Baichuan-13B) | 1.4T | | 64000 | ALiBi | RMSNorm | | Flash-attention |
| baichuan-53b | 2023/8/8 | | --- | --- |--- | 53(用于搜索) | --- | [baichuan Demo](https://chat.baichuan-ai.com/home) | --- | --- | --- | --- | --- | --- | --- |
| fireballoon/baichuan-vicuna-chinese-7b | | baichuan-7b | | | | | | | | | | | | | |
| fireballoon/baichuan-vicuna-7b | | baichuan-7b | | | | | | | | | | | | | |
| firefly-baichuan-7b-qlora-sft | | baichuan-7b | | | [blog](https://mp.weixin.qq.com/s/_eTkDGG5DmxyWeiQ6DIxBw),[Hugging Face model](https://huggingface.co/YeungNLP/firefly-baichuan-7b-qlora-sft),[Model Scope](https://modelscope.cn/models/baichuan-inc/baichuan-7B/summary),[C-EVAL](https://cevalbenchmark.com/static/leaderboard_zh.html) | | | [code](https://github.com/baichuan-inc/baichuan-7B) |
| baichuan-13b-Chat | | | | | [blog](https://mp.weixin.qq.com/s/wStOyHPd8c7V0ug1Qebryw) | | | [code](https://github.com/percent4/document_qa_with_llm) |
| Baichuan2 | | | | [Baichuan2](https://huggingface.co/baichuan-inc/Baichuan2-7B-Intermediate-Checkpoints) | [Baichuan2技术报告](https://cdn.baichuan-ai.com/paper/Baichuan2-technical-report.pdf),[SuperCLUE评测效果](https://mp.weixin.qq.com/s/SV7COWNu9uGnpOBzVYCyog) | 7,13 | | [Baichuan2 Code](https://github.com/baichuan-inc/Baichuan2) | 2.6T | | | | | | |
| firefly-baichuan-13b | --- | baichuan-13b-base | QLoRA |--- | --- | --- |--- | --- | --- | --- | --- | --- | --- | --- | --- |
## ChatGLM Alternatives
| Model/Description| Paper | Code | Blog | Tokens | Tokenizer | Vocab size | Position Embedding | Layer Normalization | Activation Function | Attention |
| --- | --- | --- | --- |--- | --- | --- |--- | --- | --- | --- |
| ChatGLM-6B | | [code](https://github.com/THUDM/ChatGLM-6B.git) | [blog](https://chatglm.cn/blog),[ChatGLM-6B源码阅读](https://mp.weixin.qq.com/s/r7KEJmrpJZmY7KBP4veS6A),[ChatGLM模型底座细节分析](https://mp.weixin.qq.com/s/oOdD3MYtE6-sNeAmPthqLg) | 1T | SentencePiece | 130528 | | Post Deep Norm | GeLU |
| chatglm+langchain+互联网 | | [code](https://github.com/LemonQu-GIT/ChatGLM-6B-Engineering/) | [blog](https://mp.weixin.qq.com/s/lO6SrEuv4-vNbL8B3G-f8g) |
| ChatGLM_multi_gpu_zero_Tuning | | [code](https://github.com/CSHaitao/ChatGLM_mutli_gpu_tuning) | |
| ChatGLM+Fastapi | | | [blog](https://mp.weixin.qq.com/s/5J4UA4ePVZGXJGZsBXeN8Q) |
| ChatGLM2-6B-32K | | | [blog](https://mp.weixin.qq.com/s/Fkm_D26z1jrqA44B82v7Ww) | 1.4T | | 65024 | | Post RMS Norm | SwiGLU | GQA |
| ChatGLM-6b+langchain | | [code](https://github.com/yanqiangmiffy/Chinese-LangChain) | [blog](https://mp.weixin.qq.com/s/xAsZZ_LOkr9Nj-JafSbXnA) |
| one-shot微调chatglm-6b实践信息抽取 | | | [blog](https://mp.weixin.qq.com/s/l7lCbdJ9XGzLPTb3zKDAzQ) |
| Falcon | | | [blog1](https://mp.weixin.qq.com/s/jbRRjG2ferhFPWsMtCaJyg),[blog2](https://mp.weixin.qq.com/s/Vy_xWBuZU0AaaPMCIhKIyw) | 1.5T | | 65024 | | Pre LN | GeLU | MQA |
## LLaMA Alternatives
| Target Model | Source Model | Optimization | Checkpoints | Paper/Blog | Params (B) | Context Length | Code | Tokens | Tokenizer | Vocab size | Position Embedding | Layer Normalization | Activation Function | Attention |
| --- | --- | --- | --- |--- | --- | --- |--- | --- | --- | --- | --- | --- | --- | --- |
| LLaMA | | | | [LLaMA: Open and Efficient Foundation Language Models](https://research.facebook.com/publications/llama-open-and-efficient-foundation-language-models/),[blog1](https://mp.weixin.qq.com/s?__biz=Mzg3NDIyMzI0Mw==&mid=2247485822&idx=1&sn=b365d93a0a08769aef77f34069da1422&chksm=ced54a9af9a2c38cd5779284b5e9ae573846153e7dc00961dc163664a657d6a3fa5c8c14c7d2&token=447941009&lang=zh_CN#rd),[blog2](https://mp.weixin.qq.com/s/fGNuTcYE8QI9_JKS9LcQ7w),[详聊LLaMA大模型的技术细节](https://mp.weixin.qq.com/s/B9Ue0ihUGAFjT_X__R2u8Q) | 7,13,33,65 | 2048 | [LLaMA Code](https://github.com/facebookresearch/llama) | 1T/1.4T | BPE | 32000 | RoPE | Pre RMS Norm | SwiGLU |
| LLaMA 2 | | | [[在 Hugging Face 上玩转LLaMA 2](https://mp.weixin.qq.com/s/UnzhBJjZfPXsaSu8gNnosw)] ,[[在Colab笔记本中微调自己的Llama 2模型](https://mp.weixin.qq.com/s/pnDJaOUh_xdNdqSBl53Arw)],[[三步上手 LLaMA2](https://mp.weixin.qq.com/s/lkRg8-rw57wDNr7FrjOSOQ)],[[使用 Transformers 量化 Meta AI LLaMA2 中文版大模型](https://mp.weixin.qq.com/s/DEgFNAB4gwWDlQOj7-2CEg)] | [[blog](https://mp.weixin.qq.com/s?__biz=Mzg3NDIyMzI0Mw==&mid=2247486800&idx=1&sn=9b629ca41b9f6b4feedad94363a17253&chksm=ced54eb4f9a2c7a2a5b20c182981b4323b18509f2ca8f482c2a8cdbb29bf570488bdcd280eb6&token=882149695&lang=zh_CN#rd)],[[伯克利AI博士详解Llama 2的技术细节](https://mp.weixin.qq.com/s/Mee7sMq_bxLpIOOr91li9A)],[[NLP社区对LLaMA2论文上半部分的讨论](https://mp.weixin.qq.com/s/SJNqjSCBX-k80_r3nmTiuA)],[[NLP中文社区顶尖研究员们对LLaMA2论文下半部分的讨论](https://mp.weixin.qq.com/s/6k5ML3HtmvBTTCgHBZGycQ)],[[3个最值得了解llama2开发库,助你快速搭建LLM应用](https://mp.weixin.qq.com/s/_3H6Y_NolUuxYxOo8Pl7fg)],[[使用 Docker 快速上手中文版 LLaMA2 开源大模型](https://mp.weixin.qq.com/s/9cTNa_oya2Zj9YdDYodCvw)],[[ Llama 2资料汇总](https://mp.weixin.qq.com/s/-01Dg9ZVfPYM4mZ4iKt8Cw) | 7,13,34,70 | 4096 | [LLaMA2 Code](https://github.com/facebookresearch/llama-recipes) | 2T | | | | | SwiGLU | GQA |
| Airoboros | LLaMA | | | | 13B | | | | | | | | | |
| Alpaca | LLaMA 7B | | | [Alpaca blog](https://crfm.stanford.edu/2023/03/13/alpaca.html),[Alpaca homepage](https://crfm.stanford.edu/alpaca) | 7,13 | 2048 | [Alpaca Code](https://github.com/tatsu-lab/stanford_alpaca) |
| Alpaca-Lora | LLaMA 7B | | | | 7 | | [Alpaca-Lora Code](https://github.com/tloen/alpaca-lora) |
| AlpaGasus | Alpaca | | | [AlpaGasus: Training A Better Alpaca with Fewer Data](https://arx
没有合适的资源?快使用搜索试试~ 我知道了~
资源推荐
资源详情
资源评论
收起资源包目录
AGI资料汇总学习(主要包括LLM和AIGC),持续更新.......zip (66个子文件)
8543543jhkhnjkhijtfhgftrdtdgg
.DS_Store 10KB
Model List
.DS_Store 6KB
model-params.md 555B
README.md 39KB
LLM之Generate中常见解码策略解读.ipynb 73KB
星火API.ipynb 4KB
Complete Guide to LLM Fine Tuning for Beginners.ipynb 7KB
LLM之Generate中参数解读.ipynb 29KB
SparkApi.py 4KB
fine tuning LLaMA2.ipynb 64KB
LangChain
Agent 02_ create Agent from scratch.ipynb 89KB
Agent-01_ custom_tools.ipynb 64KB
README.md 78B
baichuan-13B.ipynb 8KB
Prompt Engineering
Keyword_Extraction_with_Mistral_7B.ipynb 263KB
DataSet
.DS_Store 6KB
README.md 7KB
Agent
.DS_Store 6KB
README.md 2KB
baichuan API.ipynb 3KB
Open Tool
.DS_Store 8KB
Deployment
.DS_Store 14KB
TGI
README.md 355B
Huggingface
.DS_Store 6KB
huggingface.md 252B
CTranslate2
.DS_Store 6KB
ctranslate2.md 41B
JittorLLM
.DS_Store 6KB
jittorllms.md 110B
DeepSpeed-MII
.DS_Store 6KB
deepspeed-mll.md 41B
MLC LLM
.DS_Store 6KB
mlc-llm.md 41B
FastChat
baichuan-7B.ipynb 528KB
README.md 355B
vLLM
README.md 41B
OpenLLM
.DS_Store 6KB
openllm.md 41B
BentoML
.DS_Store 8KB
blog.md 11KB
vLLM
README.md 41B
Transformer-NLP-Service
README.md 41B
bentoml.md 41B
Text generation inference
README.md 41B
CLIP-API-service
.DS_Store 6KB
clip-api-service.md 41B
LightLLM
.DS_Store 6KB
lightllm.md 41B
Ray Serve
.DS_Store 6KB
ray-serve.md 41B
TensorRT-LLM
README.md 187B
GPT4All
README.md 374B
LMDeploy
.DS_Store 6KB
lmdeploy.md 41B
FastLLM
.DS_Store 6KB
fastllm.md 317B
OneDiffusion
.DS_Store 6KB
onediffusion.md 41B
README.md 355B
Fine-Tune
.DS_Store 6KB
fine-tune.md 1KB
README.md 2KB
Evaluation
.DS_Store 6KB
evaluation.md 613B
.gitignore 2KB
README.md 22KB
共 66 条
- 1
资源评论
极致人生-010
- 粉丝: 3201
- 资源: 3077
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功