wehub/eugeneyan--open-llms
1
0
Fork 0
Code Issues Pull Requests Actions Deployments Models Agent Notes Packages Projects Releases Wiki Activity
Files
main
eugeneyan--open-llms/README.md
T
wehub-resource-sync 812578d9ae docs: make Chinese README the default
2026-07-13 10:46:49 +00:00

146 lines
42 KiB
Markdown
Raw Permalink Blame History

This file contains invisible Unicode characters
This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
<!-- WEHUB_ZH_README -->
> [!NOTE]
> 本文档由 WeHub 基于上游 README 翻译整理，属于社区翻译，非官方中文文档。
> [English](./README.en.md) · [原始项目](https://github.com/eugeneyan/open-llms) · [上游 README](https://github.com/eugeneyan/open-llms/blob/HEAD/README.md)
> 原作者、版权与许可证归属以原始项目及本仓库 LICENSE 文件为准。
# 开源大语言模型（Open LLMs）
这些 LLM（Large Language Models，大语言模型）均获准用于商业用途（例如 Apache 2.0、MIT、OpenRAIL-M）。欢迎贡献！
| 语言模型 | 发布日期 | 检查点 | 论文/博客 | 参数量 (B) | 上下文长度 | 许可证 | 试用 |
| --- | --- | --- | --- | --- | --- | --- |-----------------------------------------------------------------------------------------------------------------------|
| T5 | 2019/10 |[T5 & Flan-T5](https://github.com/google-research/t5x/blob/main/docs/models.md#flan-t5-checkpoints), [Flan-T5-xxl (HF)](https://huggingface.co/google/flan-t5-xxl) | [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://github.com/google-research/text-to-text-transfer-transformer#released-model-checkpoints) | 0.06 - 11 | [512](https://discuss.huggingface.co/t/does-t5-truncate-input-longer-than-512-internally/3602) | Apache 2.0 | [T5-Large](https://github.com/slai-labs/get-beam/tree/main/examples/t5) |
| RWKV 4 | 2021/08| [RWKV, ChatRWKV](https://github.com/BlinkDL/RWKV-LM#rwkv-parallelizable-rnn-with-transformer-level-llm-performance-pronounced-as-rwakuv-from-4-major-params-r-w-k-v) | [The RWKV Language Model (and my LM tricks)](https://github.com/BlinkDL/RWKV-LM) | 0.1 - 14 | [infinity (RNN)](https://github.com/BlinkDL/RWKV-LM#rwkv-parallelizable-rnn-with-transformer-level-llm-performance-pronounced-as-rwakuv-from-4-major-params-r-w-k-v) | Apache 2.0 | |
| GPT-NeoX-20B | 2022/04 | [GPT-NEOX-20B](https://huggingface.co/EleutherAI/gpt-neox-20b) | [GPT-NeoX-20B: An Open-Source Autoregressive Language Model](https://arxiv.org/abs/2204.06745) | 20 | [2048](https://huggingface.co/EleutherAI/gpt-neox-20b) | Apache 2.0 | |
| YaLM-100B | 2022/06 | [yalm-100b](https://huggingface.co/yandex/yalm-100b) | [Yandex publishes YaLM 100B, the largest GPT-like neural network in open source](https://yandex.com/company/press_center/press_releases/2022/2022-23-06) | 100 | [1024](https://github.com/yandex/YaLM-100B/blob/main/examples/generate_interactive.sh) | Apache 2.0 | |
| UL2 | 2022/10 | [UL2 & Flan-UL2](https://github.com/google-research/google-research/tree/master/ul2#checkpoints), [Flan-UL2 (HF)](https://huggingface.co/google/flan-ul2) | [UL2 20B: An Open Source Unified Language Learner](https://ai.googleblog.com/2022/10/ul2-20b-open-source-unified-language.html) | 20 | [512, 2048](https://huggingface.co/google/flan-ul2#tldr) | Apache 2.0 | |
| Bloom | 2022/11 | [Bloom](https://huggingface.co/bigscience/bloom) | [BLOOM: A 176B-Parameter Open-Access Multilingual Language Model](https://arxiv.org/abs/2211.05100) | 176 | [2048](https://huggingface.co/bigscience/bloom) | [OpenRAIL-M v1](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | |
| ChatGLM | 2023/03<!--13--> | [chatglm-6b](https://huggingface.co/THUDM/chatglm-6b) | [ChatGLM](https://zhipuai.cn/en/news/61), [Github](https://github.com/THUDM/ChatGLM-6B/blob/main/README_en.md) | 6 | [2048](https://huggingface.co/THUDM/chatglm-6b/blob/main/config.json) | [Custom](https://huggingface.co/THUDM/chatglm-6b/blob/main/MODEL_LICENSE) 免费，但有一定使用限制（可能需要注册） | |
| Cerebras-GPT | 2023/03 | [Cerebras-GPT](https://huggingface.co/cerebras) | [Cerebras-GPT: A Family of Open, Compute-efficient, Large Language Models](https://www.cerebras.net/blog/cerebras-gpt-a-family-of-open-compute-efficient-large-language-models/) ([Paper](https://arxiv.org/abs/2304.03208)) | 0.111 - 13 | [2048](https://huggingface.co/cerebras/Cerebras-GPT-13B#model-details) | Apache 2.0 | [Cerebras-GPT-1.3B](https://github.com/slai-labs/get-beam/tree/main/examples/cerebras-gpt) |
| Open Assistant (Pythia family) | 2023/03 | [OA-Pythia-12B-SFT-8](https://huggingface.co/OpenAssistant/pythia-12b-sft-v8-7k-steps), [OA-Pythia-12B-SFT-4](https://huggingface.co/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5), [OA-Pythia-12B-SFT-1](https://huggingface.co/OpenAssistant/oasst-sft-1-pythia-12b) | [Democratizing Large Language Model Alignment](https://arxiv.org/abs/2304.07327) | 12 | [2048](https://huggingface.co/OpenAssistant/pythia-12b-sft-v8-7k-steps/blob/main/config.json) | Apache 2.0 | [Pythia-2.8B](https://github.com/slai-labs/get-beam/tree/main/examples/pythia) |
| Pythia | 2023/04 | [pythia 70M - 12B](https://github.com/EleutherAI/pythia) | [Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling](https://arxiv.org/abs/2304.01373) | 0.07 - 12 | [2048](https://arxiv.org/pdf/2304.01373.pdf) | Apache 2.0 | |
| Dolly | 2023/04 | [dolly-v2-12b](https://huggingface.co/databricks/dolly-v2-12b) | [Free Dolly: Introducing the World's First Truly Open Instruction-Tuned LLM](https://www.databricks.com/blog/2023/04/12/dolly-first-open-commercially-viable-instruction-tuned-llm) | 3, 7, 12 | [2048](https://github.com/databrickslabs/dolly#dolly) | MIT | |
| StableLM-Alpha | 2023/04 | [StableLM-Alpha](https://github.com/Stability-AI/StableLM#stablelm-alpha) | [Stability AI Launches the First of its StableLM Suite of Language Models](https://stability.ai/blog/stability-ai-launches-the-first-of-its-stablelm-suite-of-language-models) | 3 - 65 | [4096](https://github.com/Stability-AI/StableLM#stablelm-alpha) | CC BY-SA-4.0 | |
| FastChat-T5 | 2023/04 | [fastchat-t5-3b-v1.0](https://huggingface.co/lmsys/fastchat-t5-3b-v1.0) | [We are excited to release FastChat-T5: our compact and commercial-friendly chatbot!](https://twitter.com/lmsysorg/status/1652037026705985537?s=20) | 3 | [512](https://huggingface.co/lmsys/fastchat-t5-3b-v1.0/blob/main/config.json) | Apache 2.0 | |
| DLite | 2023/05 | [dlite-v2-1_5b](https://huggingface.co/aisquared/dlite-v2-1_5b) | [Announcing DLite V2: Lightweight, Open LLMs That Can Run Anywhere](https://medium.com/ai-squared/announcing-dlite-v2-lightweight-open-llms-that-can-run-anywhere-a852e5978c6e) | 0.124 - 1.5 | [1024](https://huggingface.co/aisquared/dlite-v2-1_5b/blob/main/config.json) | Apache 2.0 | [DLite-v2-1.5B](https://github.com/slai-labs/get-beam/tree/main/examples/dlite-v2) |
| h2oGPT | 2023/05 | [h2oGPT](https://github.com/h2oai/h2ogpt) | [Building the World’s Best Open-Source Large Language Model: H2O.ai’s Journey](https://h2o.ai/blog/building-the-worlds-best-open-source-large-language-model-h2o-ais-journey/) | 12 - 20 | [256 - 2048](https://huggingface.co/h2oai) | Apache 2.0 | |
| MPT-7B | 2023/05 | [MPT-7B](https://huggingface.co/mosaicml/mpt-7b), [MPT-7B-Instruct](https://huggingface.co/mosaicml/mpt-7b-instruct) | [Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs](https://www.mosaicml.com/blog/mpt-7b) | 7 | [84k (ALiBi)](https://huggingface.co/mosaicml/mpt-7b#how-is-this-model-different) | Apache 2.0, CC BY-SA-3.0 | |
| RedPajama-INCITE | 2023/05 | [RedPajama-INCITE](https://huggingface.co/togethercomputer) | [Releasing 3B and 7B RedPajama-INCITE family of models including base, instruction-tuned & chat models](https://www.together.xyz/blog/redpajama-models-v1) | 3 - 7 | [2048](https://huggingface.co/togethercomputer/RedPajama-INCITE-Instruct-7B-v0.1/blob/157bf3174feebb67f37e131ea68f84dee007c687/config.json#L13) | Apache 2.0 | [RedPajama-INCITE-Instruct-3B-v1](https://github.com/slai-labs/get-beam/tree/main/examples/redpajama-incite-instruct) |
| OpenLLaMA | 2023/05 | [open_llama_3b](https://huggingface.co/openlm-research/open_llama_3b), [open_llama_7b](https://huggingface.co/openlm-research/open_llama_7b), [open_llama_13b](https://huggingface.co/openlm-research/open_llama_13b) | [OpenLLaMA: An Open Reproduction of LLaMA](https://github.com/openlm-research/open_llama) | 3, 7 | [2048](https://huggingface.co/h2oai) | Apache 2.0 | [OpenLLaMA-7B-Preview_200bt](https://github.com/slai-labs/get-beam/tree/main/examples/openllama) |
| Falcon | 2023/05 | [Falcon-180B](https://huggingface.co/tiiuae/falcon-180B), [Falcon-40B](https://huggingface.co/tiiuae/falcon-40b), [Falcon-7B](https://huggingface.co/tiiuae/falcon-7b) | [The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only](https://arxiv.org/abs/2306.01116) | 180, 40, 7 | [2048](https://huggingface.co/tiiuae/falcon-7b/blob/main/config.json) | Apache 2.0 |
| GPT-J-6B | 2023/06 | [GPT-J-6B](https://github.com/kingoflolz/mesh-transformer-jax/#gpt-j-6b), [GPT4All-J](https://github.com/nomic-ai/gpt4all#raw-model) | [GPT-J-6B: 6B JAX-Based Transformer](https://arankomatsuzaki.wordpress.com/2021/06/04/gpt-j/) | 6 | [2048](https://github.com/kingoflolz/mesh-transformer-jax/#gpt-j-6b) | Apache 2.0 | |
| MPT-30B | 2023/06 | [MPT-30B](https://huggingface.co/mosaicml/mpt-30b), [MPT-30B-instruct](https://huggingface.co/mosaicml/mpt-30b-instruct) | [MPT-30B: Raising the bar for open-source foundation models](https://www.mosaicml.com/blog/mpt-30b) | 30 | [8192](https://huggingface.co/mosaicml/mpt-30b/blob/main/config.json) | Apache 2.0, CC BY-SA-3.0 | [MPT 30B inference code using CPU](https://github.com/abacaj/mpt-30B-inference) |
| LLaMA 2 | 2023/06<!--18--> | [LLaMA 2 Weights](https://ai.meta.com/resources/models-and-libraries/llama-downloads/) | [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://scontent-ham3-1.xx.fbcdn.net/v/t39.2365-6/10000000_662098952474184_2584067087619170692_n.pdf?_nc_cat=105&ccb=1-7&_nc_sid=3c67a6&_nc_ohc=qhK-ahCbkBMAX94XV2X&_nc_ht=scontent-ham3-1.xx&oh=00_AfDB7dN8momft9nkv8X0gqrZdEnKltVjPOxhKBm0XLRinA&oe=64BE66FF) | 7 - 70 | [4096](https://scontent-ham3-1.xx.fbcdn.net/v/t39.2365-6/10000000_662098952474184_2584067087619170692_n.pdf?_nc_cat=105&ccb=1-7&_nc_sid=3c67a6&_nc_ohc=qhK-ahCbkBMAX94XV2X&_nc_ht=scontent-ham3-1.xx&oh=00_AfDB7dN8momft9nkv8X0gqrZdEnKltVjPOxhKBm0XLRinA&oe=64BE66FF) | [Custom](https://github.com/facebookresearch/llama/blob/main/LICENSE) 若用户数低于 7 亿则可免费使用，且不得使用 LLaMA 输出训练除 LLaMA 及其衍生模型以外的其他 LLM | [HuggingChat](https://huggingface.co/blog/llama2#demo) |
| ChatGLM2 | 2023/06<!--25--> | [chatglm2-6b](https://huggingface.co/THUDM/chatglm2-6b) | [ChatGLM2-6B](https://zhipuai.cn/en/news/72), [Github](https://github.com/THUDM/ChatGLM2-6B) | 6 | [32k](https://huggingface.co/THUDM/chatglm2-6b/blob/main/config.json) | [Custom](https://huggingface.co/THUDM/chatglm2-6b/blob/main/MODEL_LICENSE) 免费，但有一定使用限制（可能需要注册） | |
| XGen-7B | 2023/06<!--28--> | [xgen-7b-4k-base](https://huggingface.co/Salesforce/xgen-7b-4k-base), [xgen-7b-8k-base](https://huggingface.co/Salesforce/xgen-7b-8k-base) | [Long Sequence Modeling with XGen](https://blog.salesforceairesearch.com/xgen/) | 7 | [4096](https://huggingface.co/Salesforce/xgen-7b-4k-base), [8192](https://huggingface.co/Salesforce/xgen-7b-8k-base) | Apache 2.0 | |
| Jais-13b | 2023/08<!--17--> | [jais-13b](https://huggingface.co/core42/jais-13b), [jais-13b-chat](https://huggingface.co/core42/jais-13b-chat) | [Jais and Jais-chat: Arabic-Centric Foundation and Instruction-Tuned Open Generative Large Language Models](https://arxiv.org/abs/2308.16149) | 13 | [2048](https://huggingface.co/core42/jais-13b/blob/main/config.json) | Apache 2.0 | |
| OpenHermes | 2023/09<!--14--> | [OpenHermes-7B](https://huggingface.co/teknium/OpenHermes-7B), [OpenHermes-13B](https://huggingface.co/teknium/OpenHermes-13B) | [Nous Research](https://nousresearch.com/) | 7, 13 | [4096](https://huggingface.co/teknium/OpenHermes-13B/blob/main/config.json)| MIT | [OpenHermes-V2 Finetuned on Mistral 7B](https://huggingface.co/spaces/artificialguybr/OPENHERMES-2)
| OpenLM | 2023/09<!--26--> | [OpenLM 1B](https://huggingface.co/mlfoundations/open_lm_1B), [OpenLM 7B](https://huggingface.co/mlfoundations/open_lm_7B_1.25T) | [Open LM: a minimal but performative language modeling (LM) repository](https://github.com/mlfoundations/open_lm#pretrained-models) | 1, 7 | [2048](https://github.com/mlfoundations/open_lm/blob/main/open_lm/model_configs/open_lm_7b.json) | MIT | |
| Mistral 7B | 2023/09<!--27--> | [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1), [Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) | [Mistral 7B](https://mistral.ai/news/announcing-mistral-7b/) | 7 | [4096-16K with Sliding Windows](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1/blob/main/config.json)| Apache 2.0 | [Mistral Transformer](https://github.com/mistralai/mistral-src)
| ChatGLM3 | 2023/10<!--27--> | [chatglm3-6b](https://huggingface.co/THUDM/chatglm3-6b), [chatglm3-6b-base](https://huggingface.co/THUDM/chatglm3-6b-base), [chatglm3-6b-32k](https://huggingface.co/THUDM/chatglm3-6b-32k), [chatglm3-6b-128k](https://huggingface.co/THUDM/chatglm3-6b-128k) | [ChatGLM3](https://github.com/THUDM/ChatGLM3/blob/main/README_en.md) | 6 | [8192](https://huggingface.co/THUDM/chatglm3-6b/blob/main/config.json), [32k](https://huggingface.co/THUDM/chatglm3-6b-32k/blob/main/config.json), [128k](https://huggingface.co/THUDM/chatglm3-6b-128k/blob/main/config.json) | [Custom](https://huggingface.co/THUDM/chatglm3-6b/blob/main/MODEL_LICENSE) 免费，但有一定使用限制（可能需要注册） | |
| Skywork | 2023/10<!--30--> | [Skywork-13B-Base](https://huggingface.co/Skywork/Skywork-13B-Base), [Skywork-13B-Math](https://huggingface.co/Skywork/Skywork-13B-Math) | [Skywork](https://github.com/SkyworkAI/Skywork/blob/main/README_EN.md) | 13 | [4096](https://huggingface.co/Skywork/Skywork-13B-base/blob/main/config.json) | [Custom](https://github.com/SkyworkAI/Skywork/blob/main/Skywork%20Community%20License.pdf) 免费，但有使用限制；基于 Skywork 输出训练的模型将成为 Skywork 衍生模型，并受本许可证约束。 | |
| Jais-30b | 2023/11<!--08--> | [jais-30b-v1](https://huggingface.co/core42/jais-30b-v1), [jais-30b-chat-v1](https://huggingface.co/core42/jais-30b-chat-v1) | [Jais-30B: Expanding the Horizon in Open-Source Arabic NLP](https://g42.ai/resources/publications/Jais-30B) | 30 | [2048](https://huggingface.co/core42/jais-30b-v1/blob/main/config.json) | Apache 2.0 | |
| Zephyr | 2023/11<!--10--> | [Zephyr 7B](https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1) | [Website](https://arxiv.org/abs/2310.16944) | 7 | [8192](https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1)| Apache 2.0 | |
| DeepSeek | 2023/11<!--30--> | [deepseek-llm-7b-base](https://huggingface.co/deepseek-ai/deepseek-llm-7b-base), [deepseek-llm-7b-chat](https://huggingface.co/deepseek-ai/deepseek-llm-7b-chat), [deepseek-llm-67b-base](https://huggingface.co/deepseek-ai/deepseek-llm-67b-base), [deepseek-llm-67b-chat](https://huggingface.co/deepseek-ai/deepseek-llm-67b-chat) | [Introducing DeepSeek LLM](https://github.com/deepseek-ai/deepseek-LLM), | 7, 67 | [4096](https://github.com/deepseek-ai/deepseek-LLM) | [Custom](https://github.com/deepseek-ai/DeepSeek-LLM/blob/HEAD/LICENSE-MODEL) 免费，但有使用限制；基于 DeepSeek 输出训练的模型将成为 DeepSeek 衍生模型，并受本许可证约束。
| Mistral 7B v0.2 | 2023/12<!--11--> | [Mistral-7B-v0.2](https://huggingface.co/mistral-community/Mistral-7B-v0.2), [Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) | [La Plateforme](https://mistral.ai/news/la-plateforme/) | 7 | [32k](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) | Apache 2.0 | |
| Mixtral 8x7B v0.1 | 2023/12<!--11--> | [Mixtral-8x7B-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-v0.1), [Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1) | [Mixtral of experts](https://mistral.ai/news/mixtral-of-experts/) | 46.7 | [32k](https://mistral.ai/news/mixtral-of-experts/) | Apache 2.0 | |
| LLM360 Amber | 2023/12<!--11--> | [Amber](https://huggingface.co/LLM360/Amber), [AmberChat](https://huggingface.co/LLM360/AmberChat), [AmberSafe](https://huggingface.co/LLM360/AmberSafe) | [Introducing LLM360: Fully Transparent Open-Source LLMs](https://www.llm360.ai/blog/introducing-llm360-fully-transparent-open-source-llms.html) | 6.7 | [2048](https://huggingface.co/LLM360/Amber#%F0%9F%9F%A0-model-description) | Apache 2.0 | |
| SOLAR | 2023/12<!--12--> | [Solar-10.7B](https://huggingface.co/upstage/SOLAR-10.7B-v1.0) | [Upstage](https://arxiv.org/abs/2312.15166) | 10.7 | [4096](https://huggingface.co/upstage/SOLAR-10.7B-v1.0/blob/main/config.json)| apache-2.0 | |
| phi-2 | 2023/12<!--12--> | [phi-2 2.7B](https://huggingface.co/microsoft/phi-2) | [Microsoft](https://www.microsoft.com/en-us/research/blog/phi-2-the-surprising-power-of-small-language-models/) | 2.7 | [2048](https://huggingface.co/microsoft/phi-2/blob/main/config.json)| MIT | |
| FLOR | 2023/12<!--22--> | [FLOR-760M](https://huggingface.co/projecte-aina/FLOR-760M), [FLOR-1.3B](https://huggingface.co/projecte-aina/FLOR-1.3B), [FLOR-1.3B-Instructed](https://huggingface.co/projecte-aina/FLOR-1.3B-Instructed), [FLOR-6.3B](https://huggingface.co/projecte-aina/FLOR-6.3B), [FLOR-6.3B-Instructed](https://huggingface.co/projecte-aina/FLOR-6.3B-Instructed) | [FLOR-6.3B: a chinchilla-compliant model for Catalan, Spanish and English](https://medium.com/@mpamies247/flor-6-3b-a-chinchilla-compliant-model-for-catalan-spanish-and-english-7cdb389a9aac) | 0.76, 1.3, 6.3 | [2048](https://huggingface.co/bigscience/bloom-1b1#technical-specifications) | Apache 2.0，继承自 BLOOM 的使用限制 | |
| RWKV 5 v2 | 2024/01<!--28--> | [rwkv-5-world-0.4b-2, rwkv-5-world-1.5b-2, rwkv-5-world-3b-2, rwkv-5-world-3b-2(16k), rwkv-5-world-7b-2](https://huggingface.co/BlinkDL/rwkv-5-world) | [RWKV 5](https://www.rwkv.com/) | 0.4, 1.5, 3, 7 | [unlimited(RNN), trained on 4096 (and 16k for 3b)](https://huggingface.co/BlinkDL/rwkv-5-world/tree/main) | Apache 2.0 | |
| OLMo | 2024/02<!--01--> | [OLMo 1B](https://huggingface.co/allenai/OLMo-1B), [OLMo 7B](https://huggingface.co/allenai/OLMo-7B), [OLMo 7B Twin 2T](https://huggingface.co/allenai/OLMo-7B-Twin-2T) | [AI2](https://blog.allenai.org/hello-olmo-a-truly-open-llm-43f7e7359222) | 1,7 | [2048](https://huggingface.co/allenai/OLMo-7B-Twin-2T/blob/main/config.json) | Apache 2.0 | |
| Qwen1.5 | 2024/02<!--04--> | [Qwen1.5-7B](https://huggingface.co/Qwen/Qwen1.5-7B), [Qwen1.5-7B-Chat](https://huggingface.co/Qwen/Qwen1.5-7B-Chat), [Qwen1.5-14B](https://huggingface.co/Qwen/Qwen1.5-14B), [Qwen1.5-14B-Chat](https://huggingface.co/Qwen/Qwen1.5-14B-Chat), [Qwen1.5-72B](https://huggingface.co/Qwen/Qwen1.5-72B), [Qwen1.5-72B-Chat](https://huggingface.co/Qwen/Qwen1.5-72B-Chat) | [Introducing Qwen1.5](https://qwenlm.github.io/blog/qwen1.5/) | 7, 14, 72 | [32k](https://huggingface.co/Qwen/Qwen1.5-7B-Chat/blob/main/config.json) | [Custom](https://huggingface.co/Qwen/Qwen1.5-7B-Chat/blob/main/LICENSE) 若用户数低于 1 亿则可免费使用，且不得使用 Qwen 输出训练除 Qwen 及其衍生模型以外的其他 LLM | |
| LWM | 2024/02<!--07--> | [LWM-Text-Chat-128K](https://huggingface.co/LargeWorldModel/LWM-Text-Chat-128K), [LWM-Text-Chat-256K](https://huggingface.co/LargeWorldModel/LWM-Text-Chat-256K), [LWM-Text-Chat-512K](https://huggingface.co/LargeWorldModel/LWM-Text-Chat-512K), [LWM-Text-Chat-1M](https://huggingface.co/LargeWorldModel/LWM-Text-Chat-1M), [LWM-Text-128K](https://huggingface.co/LargeWorldModel/LWM-Text-128K), [LWM-Text-256K](https://huggingface.co/LargeWorldModel/LWM-Text-256K), [LWM-Text-512K](https://huggingface.co/LargeWorldModel/LWM-Text-512K), [LWM-Text-1M](https://huggingface.co/LargeWorldModel/LWM-Text-1M) | [Large World Model (LWM)](https://github.com/LargeWorldModel/LWM) | 7 | [128k, 256k, 512k, 1M](https://github.com/LargeWorldModel/LWM#available-models) | LLaMA 2 许可证 | |
| Jais-30b v3 | 2024/03<!--08--> | [jais-30b-v3](https://huggingface.co/core42/jais-30b-v3), [jais-30b-chat-v3](https://huggingface.co/core42/jais-30b-chat-v3) | [Jais 30b v3](https://huggingface.co/core42/jais-30b-v3) | 30 | [8192](https://huggingface.co/core42/jais-30b-v3/blob/main/config.json) | Apache 2.0 | |
| Gemma | 2024/02<!--21--> | [Gemma 7B](https://huggingface.co/google/gemma-7b), [Gemma 7B it](https://huggingface.co/google/gemma-7b-it), [Gemma 2B](https://huggingface.co/google/gemma-2b), [Gemma 2B it](https://huggingface.co/google/gemma-2b-it) | [Technical report](https://storage.googleapis.com/deepmind-media/gemma/gemma-report.pdf) | 2-7 | [8192](https://storage.googleapis.com/deepmind-media/gemma/gemma-report.pdf) | [Gemma Terms of Use](https://ai.google.dev/gemma/terms) 免费，但有使用限制；基于 Gemma 输出训练的模型将成为 Gemma 衍生模型，并受本许可证约束。 | |
| Grok-1 | 2024/03<!--17--> | [Grok-1](https://huggingface.co/xai-org/grok-1) | [Open Release of Grok-1](https://x.ai/blog/grok-os) | 314 | [8192](https://github.com/xai-org/grok-1/blob/main/run.py) | Apache 2.0 | |
| Qwen1.5 MoE | 2024/03<!--28--> | [Qwen1.5-MoE-A2.7B](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B), [Qwen1.5-MoE-A2.7B-Chat](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B-Chat) | [Qwen1.5-MoE: Matching 7B Model Performance with 1/3 Activated Parameters](https://qwenlm.github.io/blog/qwen-moe/) | 14.3 | [8192](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B/blob/main/config.json) | [Custom](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B/blob/main/LICENSE) 若用户数低于 1 亿则可免费使用，且不得使用 Qwen 输出训练除 Qwen 及其衍生模型以外的其他 LLM | |
| Jamba 0.1 | 2024/03<!--28--> | [Jamba-v0.1](https://huggingface.co/ai21labs/Jamba-v0.1) | [Introducing Jamba: AI21's Groundbreaking SSM-Transformer Model](https://www.ai21.com/blog/announcing-jamba) | 52 | [256k](https://huggingface.co/ai21labs/Jamba-v0.1/blob/main/config.json) | Apache 2.0 | |
| Qwen1.5 32B | 2024/04<!--02--> | [Qwen1.5-32B](https://huggingface.co/Qwen/Qwen1.5-32B), [Qwen1.5-32B-Chat](https://huggingface.co/Qwen/Qwen1.5-32B-Chat) | [Qwen1.5-32B: Fitting the Capstone of the Qwen1.5 Language Model Series](https://qwenlm.github.io/blog/qwen1.5-32b/) | 32 | [32k](https://huggingface.co/Qwen/Qwen1.5-32B-Chat/blob/main/config.json) | [Custom](https://huggingface.co/Qwen/Qwen1.5-32B-Chat/blob/main/LICENSE) 若用户数低于 1 亿则可免费使用，且不得使用 Qwen 输出训练除 Qwen 及其衍生模型以外的其他 LLM | |
| Mamba-7B | 2024/04<!--15--> | [mamba-7b-rw](https://huggingface.co/TRI-ML/mamba-7b-rw) | [Toyota Research Institute](https://huggingface.co/TRI-ML/mamba-7b-rw) | 7 | [unlimited(RNN), trained on 2048](https://huggingface.co/TRI-ML/mamba-7b-rw) | Apache 2.0 | |
| Mixtral8x22B v0.1 | 2024/04<!--17--> | [Mixtral-8x22B-v0.1](https://huggingface.co/mistralai/Mixtral-8x22B-v0.1), [Mixtral-8x22B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x22B-Instruct-v0.1) | [Cheaper, Better, Faster, Stronger](https://mistral.ai/news/mixtral-8x22b/) | 141 | [64k](https://mistral.ai/news/mixtral-8x22b/) | Apache 2.0 | |
| Llama 3 | 2024/04<!--18--> | [Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B), [Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct), [Llama-3-70B](https://huggingface.co/meta-llama/Meta-Llama-3-70B), [Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct), [Llama-Guard-2-8B](https://huggingface.co/meta-llama/Meta-Llama-Guard-2-8B) | [Introducing Meta Llama 3](https://ai.meta.com/blog/meta-llama-3/), [Meta Llama 3](https://llama.meta.com/llama3/) | 8, 70 | [8192](https://github.com/meta-llama/llama3) | [Meta Llama 3 Community License Agreement](https://github.com/meta-llama/llama3/blob/main/LICENSE) 若用户数低于 7 亿则可免费使用，且不得使用 LLaMA 3 输出训练除 LLaMA 3 及其衍生模型以外的其他 LLM | |
| Phi-3 Mini | 2024/04<!--23--> | [Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct), [Phi-3-mini-128k-instruct](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct) | [Introducing Phi-3](https://azure.microsoft.com/en-us/blog/introducing-phi-3-redefining-whats-possible-with-slms/), [Technical Report](https://arxiv.org/abs/2404.14219) | 3.8 | [4096, 128k](https://arxiv.org/abs/2404.14219) | MIT | |
| OpenELM | 2024/04<!--24--> | [OpenELM-270M](https://huggingface.co/apple/OpenELM-270M), [OpenELM-270M-Instruct](https://huggingface.co/apple/OpenELM-270M-Instruct), [OpenELM-450M](https://huggingface.co/apple/OpenELM-450M), [OpenELM-450M-Instruct](https://huggingface.co/apple/OpenELM-450M-Instruct), [OpenELM-1_1B](https://huggingface.co/apple/OpenELM-1_1B), [OpenELM-1_1B-Instruct](https://huggingface.co/apple/OpenELM-1_1B-Instruct), [OpenELM-3B](https://huggingface.co/apple/OpenELM-3B), [OpenELM-3B-Instruct](https://huggingface.co/apple/OpenELM-3B-Instruct) | [OpenELM: An Efficient Language Model Family with Open Training and Inference Framework](https://machinelearning.apple.com/research/openelm) | 0.27, 0.45, 1.1, 3 | [2048](https://arxiv.org/html/2404.14619v2) | [Custom open license](https://huggingface.co/apple/OpenELM-270M/blob/main/LICENSE) 无使用或训练限制
| Snowflake Arctic | 2024/04<!--24--> | [snowflake-arctic-base](https://huggingface.co/Snowflake/snowflake-arctic-base), [snowflake-arctic-instruct](https://huggingface.co/Snowflake/snowflake-arctic-instruct) | [Snowflake Arctic: The Best LLM for Enterprise AI — Efficiently Intelligent, Truly Open](https://www.snowflake.com/blog/arctic-open-efficient-foundation-language-models-snowflake/) | 480 | [4096](https://huggingface.co/Snowflake/snowflake-arctic-base/blob/main/config.json) | Apache 2.0 | |
| Qwen1.5 110B | 2024/04<!--25--> | [Qwen1.5-110B](https://huggingface.co/Qwen/Qwen1.5-110B), [Qwen1.5-110B-Chat](https://huggingface.co/Qwen/Qwen1.5-110B-Chat) | [Qwen1.5-110B: The First 100B+ Model of the Qwen1.5 Series](https://qwenlm.github.io/blog/qwen1.5-110b/) | 110 | [32k](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/config.json) | [Custom](https://huggingface.co/Qwen/Qwen1.5-110B-Chat/blob/main/LICENSE) 若用户数低于 1 亿则可免费使用，且不得使用 Qwen 输出训练除 Qwen 及其衍生模型以外的其他 LLM | |
| RWKV 6 v2.1 | 2024/05<!--06--> | [rwkv-6-world-1.6b-2.1, rwkv-6-world-3b-2.1, rwkv-6-world-7b-2.1](https://huggingface.co/BlinkDL/rwkv-6-world) | [RWKV 6](https://www.rwkv.com/) | 1.6, 3, 7 | [unlimited(RNN), trained on 4096](https://huggingface.co/BlinkDL/rwkv-6-world/tree/main) | Apache 2.0 | |
| DeepSeek-V2 | 2024/05<!--06--> | [DeepSeek-V2](https://huggingface.co/deepseek-ai/DeepSeek-V2), [DeepSeek-V2-Chat](https://huggingface.co/deepseek-ai/DeepSeek-V2-Chat) | [DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model](https://github.com/deepseek-ai/DeepSeek-V2) | 236 | [128k](https://github.com/deepseek-ai/DeepSeek-V2) | [Custom](https://github.com/deepseek-ai/DeepSeek-V2/blob/main/LICENSE-MODEL) 免费，但有使用限制；基于 DeepSeek 输出训练的模型将成为 DeepSeek 衍生模型，并受本许可证约束。 | |
| Fugaku-LLM | 2024/05<!--13--> | [Fugaku-LLM-13B](https://huggingface.co/Fugaku-LLM/Fugaku-LLM-13B), [Fugaku-LLM-13B-instruct](https://huggingface.co/Fugaku-LLM/Fugaku-LLM-13B-instruct) | [Release of "Fugaku-LLM" – a large language model trained on the supercomputer "Fugaku"](https://www.titech.ac.jp/english/news/2024/069223) | 13 | [2048](https://huggingface.co/Fugaku-LLM/Fugaku-LLM-13B/blob/main/config.json) | [Custom](https://huggingface.co/Fugaku-LLM/Fugaku-LLM-13B-instruct/blob/main/LICENSE) 免费，但有使用限制 | |
| Falcon 2 | 2024/05<!--13--> | [falcon2-11B](https://huggingface.co/tiiuae/falcon-11B) | [Meet Falcon 2: TII Releases New AI Model Series, Outperforming Meta’s New Llama 3](https://falconllm.tii.ae/falcon-2.html) | 11 | [8192](https://huggingface.co/tiiuae/falcon-11B#training-data) | [Custom Apache 2.0](https://falconllm-staging.tii.ae/falcon-2-terms-and-conditions.html) 附带轻度可接受使用政策 | |
| Yi-1.5 | 2024/05<!--15--> | [Yi-1.5-6B](https://huggingface.co/01-ai/Yi-1.5-6B), [Yi-1.5-6B-Chat](https://huggingface.co/01-ai/Yi-1.5-6B-Chat), [Yi-1.5-9B](https://huggingface.co/01-ai/Yi-1.5-9B), [Yi-1.5-9B-Chat](https://huggingface.co/01-ai/Yi-1.5-9B-Chat), [Yi-1.5-34B](https://huggingface.co/01-ai/Yi-1.5-34B), [Yi-1.5-34B-Chat](https://huggingface.co/01-ai/Yi-1.5-34B-Chat) | [Yi-1.5](https://github.com/01-ai/Yi-1.5) | 6, 9, 34 | [4096](https://huggingface.co/01-ai/Yi-1.5-6B/blob/main/config.json) | Apache 2.0 | |
| DeepSeek-V2-Lite | 2024/05<!--16--> | [DeepSeek-V2-Lite](https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite), [DeepSeek-V2-Lite-Chat](https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite-Chat) | [DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model](https://github.com/deepseek-ai/DeepSeek-V2) | 16 | [32k](https://github.com/deepseek-ai/DeepSeek-V2) | [Custom](https://github.com/deepseek-ai/DeepSeek-V2/blob/main/LICENSE-MODEL) 免费，但有使用限制；基于 DeepSeek 输出训练的模型将成为 DeepSeek 衍生模型，并受本许可证约束。 | |
| Phi-3 small/medium | 2024/05<!--21--> | [Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct), [Phi-3-mini-128k-instruct](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct), [Phi-3-medium-4k-instruct](https://huggingface.co/microsoft/Phi-3-medium-4k-instruct), [Phi-3-medium-128k-instruct](https://huggingface.co/microsoft/Phi-3-medium-128k-instruct) | [New models added to the Phi-3 family, available on Microsoft Azure](https://azure.microsoft.com/en-us/blog/new-models-added-to-the-phi-3-family-available-on-microsoft-azure/), [Technical Report](https://arxiv.org/abs/2404.14219) | 7, 14 | [4096, 128k](https://arxiv.org/abs/2404.14219) | MIT | |
| Phi-4 | 2024/12 | [Phi-4](https://huggingface.co/microsoft/phi-4)| [Introducing Phi-4: Microsoft’s Newest Small Language Model Specializing in Complex Reasoning](https://techcommunity.microsoft.com/blog/aiplatformblog/introducing-phi-4-microsoft%E2%80%99s-newest-small-language-model-specializing-in-comple/4357090), [Technical Report](https://arxiv.org/pdf/2412.08905) | 14 | [4096](https://arxiv.org/pdf/2412.08905) | MIT | |
| YuLan-Mini | 2024/12 | [YuLan-Mini](https://huggingface.co/yulan-team/YuLan-Mini) | [YuLan-Mini: An Open Data-efficient Language Model](https://arxiv.org/abs/2412.17743), [GitHub](https://github.com/RUC-GSAI/YuLan-Mini) | 14 | [28672](https://github.com/RUC-GSAI/YuLan-Mini) | MIT | [YuLan-Mini](https://huggingface.co/yulan-team/YuLan-Mini) |
| Selene Mini | 2025/01 | [Selene Mini](https://huggingface.co/AtlaAI/Selene-1-Mini-Llama-3.1-8B) | [Atla Selene Mini: A General Purpose Evaluation Model](https://arxiv.org/abs/2501.17195v1), [GitHub](https://github.com/atla-ai/selene-mini) | 8 | [128K](https://github.com/atla-ai/selene-mini) | Apache 2.0 | [Hugging Face Space](https://huggingface.co/spaces/AtlaAI/selene) |
## 用于代码的开源大语言模型（Open LLMs）
| Language Model | Release Date | Checkpoints | Paper/Blog | Params (B) | Context Length | Licence | Try it |
| --- | --- | --- | --- | --- |----------------------------------------------------------------------------------------| --- |-------------------------------------------------------------------------------------------|
| SantaCoder | 2023/01 | [santacoder](https://huggingface.co/bigcode/santacoder) |[SantaCoder: don't reach for the stars!](https://arxiv.org/abs/2301.03988) | 1.1 | [2048](https://huggingface.co/bigcode/santacoder/blob/main/README.md#model-summary) | [OpenRAIL-M v1](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | [SantaCoder](https://github.com/slai-labs/get-beam/tree/main/examples/santacoder) |
| CodeGen2 | 2023/04 | [codegen2 1B-16B](https://github.com/salesforce/CodeGen2) | [CodeGen2: Lessons for Training LLMs on Programming and Natural Languages](https://arxiv.org/abs/2305.02309) | 1 - 16 | [2048](https://arxiv.org/abs/2305.02309) | [Apache 2.0](https://github.com/salesforce/CodeGen2/blob/main/LICENSE)| |
| StarCoder | 2023/05 | [starcoder](https://huggingface.co/bigcode/starcoder) | [StarCoder: A State-of-the-Art LLM for Code](https://huggingface.co/blog/starcoder), [StarCoder: May the source be with you!](https://drive.google.com/file/d/1cN-b9GnWtHzQRoE7M7gAEyivY0kl4BYs/view) | 1.1-15 | [8192](https://huggingface.co/bigcode/starcoder#model-summary) | [OpenRAIL-M v1](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | |
| StarChat Alpha | 2023/05 | [starchat-alpha](https://huggingface.co/HuggingFaceH4/starchat-alpha) | [Creating a Coding Assistant with StarCoder](https://huggingface.co/blog/starchat-alpha) | 16 | [8192](https://huggingface.co/bigcode/starcoder#model-summary) | [OpenRAIL-M v1](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | |
| Replit Code | 2023/05 | [replit-code-v1-3b](https://huggingface.co/replit/replit-code-v1-3b) | [Training a SOTA Code LLM in 1 week and Quantifying the Vibes — with Reza Shabani of Replit](https://www.latent.space/p/reza-shabani#details) | 2.7 | [infinity? (ALiBi)](https://huggingface.co/replit/replit-code-v1-3b#model-description) | CC BY-SA-4.0 | [Replit-Code-v1-3B](https://github.com/slai-labs/get-beam/tree/main/examples/replit-code) |
| CodeT5+ | 2023/05 | [CodeT5+](https://github.com/salesforce/CodeT5/tree/main/CodeT5+) | [CodeT5+: Open Code Large Language Models for Code Understanding and Generation](https://arxiv.org/abs/2305.07922) | 0.22 - 16 | [512](https://arxiv.org/abs/2305.07922) | [BSD-3-Clause](https://github.com/salesforce/CodeT5/blob/main/LICENSE.txt) | [Codet5+-6B](https://github.com/slai-labs/get-beam/tree/main/examples/codeT5%2B) |
| XGen-7B | 2023/06 | [XGen-7B-8K-Base](https://huggingface.co/Salesforce/xgen-7b-8k-base) | [Long Sequence Modeling with XGen: A 7B LLM Trained on 8K Input Sequence Length](https://blog.salesforceairesearch.com/xgen/) | 7 | [8192](https://huggingface.co/Salesforce/xgen-7b-8k-base/blob/main/config.json) | [Apache 2.0](https://github.com/salesforce/xgen/blob/main/LICENSE) |
| CodeGen2.5 | 2023/07 | [CodeGen2.5-7B-multi](https://huggingface.co/Salesforce/codegen25-7b-multi) | [CodeGen2.5: Small, but mighty](https://blog.salesforceairesearch.com/codegen25/) | 7 | [2048](https://huggingface.co/Salesforce/codegen25-7b-multi/blob/main/config.json) | [Apache 2.0](https://huggingface.co/Salesforce/codegen25-7b-multi/blob/main/README.md) |
| DeciCoder-1B | 2023/08 | [DeciCoder-1B](https://huggingface.co/Deci/DeciCoder-1b#how-to-use) | [Introducing DeciCoder: The New Gold Standard in Efficient and Accurate Code Generation](https://deci.ai/blog/decicoder-efficient-and-accurate-code-generation-llm/) | 1.1 | [2048](https://huggingface.co/Deci/DeciCoder-1b#model-architecture) | Apache 2.0 | [DeciCoder Demo](https://huggingface.co/spaces/Deci/DeciCoder-Demo)|
| Code Llama | 2023/08 | [Inference Code for CodeLlama models]([https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://github.com/facebookresearch/codellama)) | [Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) | 7 - 34 | [4096](https://scontent-zrh1-1.xx.fbcdn.net/v/t39.2365-6/369856151_1754812304950972_1159666448927483931_n.pdf?_nc_cat=107&ccb=1-7&_nc_sid=3c67a6&_nc_ohc=wURKmnWKaloAX-ib5XW&_nc_ht=scontent-zrh1-1.xx&oh=00_AfAN1GB2K_XwIz54PqXTr-dhilI3CfCwdQoaLMyaYEEECg&oe=64F0A68F) | [Custom](https://github.com/facebookresearch/llama/blob/main/LICENSE) Free if you have under 700M users and you cannot use LLaMA outputs to train other LLMs besides LLaMA and its derivatives | [HuggingChat](https://huggingface.co/blog/codellama) |
## 用于预训练的开源大语言模型（LLM）数据集
| Name | Release Date | Paper/Blog | Dataset | Tokens (T) | License |
| --- | --- | --- | --- | --- | ---- |
| RedPajama | 2023/04 | [RedPajama, a project to create leading open-source models, starts by reproducing LLaMA training dataset of over 1.2 trillion tokens](https://www.together.xyz/blog/redpajama) | [RedPajama-Data](https://github.com/togethercomputer/RedPajama-Data) | 1.2 | Apache 2.0 |
| starcoderdata | 2023/05 | [StarCoder: A State-of-the-Art LLM for Code](https://huggingface.co/blog/starcoder) | [starcoderdata](https://huggingface.co/datasets/bigcode/starcoderdata) | 0.25 | Apache 2.0 |
## 用于指令微调（instruction-tuning）的开源大语言模型（LLM）数据集
| Name | Release Date | Paper/Blog | Dataset | Samples (K) | License |
| --- | --- | --- | --- | --- | ---- |
| OIG (Open Instruction Generalist) | 2023/03 | [THE OIG DATASET](https://laion.ai/blog/oig-dataset/) | [OIG](https://huggingface.co/datasets/laion/OIG) | 44,000 | Apache 2.0 |
| databricks-dolly-15k | 2023/04 | [Free Dolly: Introducing the World's First Truly Open Instruction-Tuned LLM](https://www.databricks.com/blog/2023/04/12/dolly-first-open-commercially-viable-instruction-tuned-llm) | [databricks-dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k) | 15 | CC BY-SA-3.0 |
| MPT-7B-Instruct | 2023/05 | [Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs](https://www.mosaicml.com/blog/mpt-7b) | [dolly_hhrlhf](https://huggingface.co/datasets/mosaicml/dolly_hhrlhf) | 59 | CC BY-SA-3.0 |
## 用于对齐微调（alignment-tuning）的开源大语言模型（LLM）数据集
| Name | Release Date | Paper/Blog | Dataset | Samples (K) | License |
| --- | --- | --- | --- | --- | ---- |
| OpenAssistant Conversations Dataset | 2023/04 | [OpenAssistant Conversations - Democratizing Large Language Model Alignment](https://drive.google.com/file/d/10iR5hKwFqAKhL3umx8muOWSRm7hs5FqX/view) | [oasst1](https://huggingface.co/datasets/OpenAssistant/oasst1) | 161 | Apache 2.0 |
## 开源大语言模型（LLM）评测
- [Leaderboard by lmsys.org](https://chat.lmsys.org/?leaderboard)
- [Evals by MosaicML](https://twitter.com/jefrankle/status/1654631746506301441)
- [Holistic Evaluation of Language Models (HELM)](https://crfm.stanford.edu/helm/latest/?groups=1)
- [LLM-Leaderboard](https://github.com/LudwigStumpp/llm-leaderboard)
- [TextSynth Server Benchmarks](https://bellard.org/ts_server/)
- [Open LLM Leaderboard by Hugging Face](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
---
### 这些许可证分别是什么意思？
- [Apache 2.0](https://en.wikipedia.org/wiki/Apache_License): 允许用户出于任何目的使用软件、分发软件、修改软件，并在许可条款下分发修改后的版本，无需担心版税。
- [MIT](https://en.wikipedia.org/wiki/MIT_License): 与 Apache 2.0 类似，但更短、更简单。此外，与 Apache 2.0 不同的是，它不要求声明对原始代码所做的任何重大更改。
- [CC BY-SA-4.0](https://creativecommons.org/licenses/by-sa/4.0/): 允许 (i) 复制和再分发材料，以及 (ii) 出于任何目的（包括商业目的）对材料进行混编、转换和二次创作。但如果你进行后者，**你必须在与原始材料相同的许可下分发你的贡献。**（因此，可能不适用于内部团队。）
- [OpenRAIL-M v1](https://www.bigcode-project.org/docs/pages/model-license/): 允许免版税访问，并可灵活地对模型及其修改进行下游使用和共享，同时附带一组使用限制（见 [Attachment A](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement))
- [BSD-3-Clause](https://en.wikipedia.org/wiki/BSD_licenses): 该版本允许出于任何目的进行无限制再分发，只要保留其版权声明和许可中的免责声明即可。
**免责声明：** 本仓库提供的信息不构成、且无意构成法律建议。本仓库维护者不对使用相关模型的第三方行为负责。若出于商业目的使用模型，请先咨询律师。
---
### 改进
- [x] 补全上下文长度（context length）相关条目，并使用 `?` 检查条目
- [ ] ~~添加训练 token 数量？~~（参见 [注意事项](https://github.com/eugeneyan/open-llms/issues/7))
- [ ] 添加（指向）训练代码的链接？
- [ ] 添加（指向）评估基准（eval benchmarks）的链接？
Reference in New Issue View Git Blame Copy Permalink