add qwen2 models
This commit is contained in:
hiyouga
parent
74f96efef9
commit
8e95648850
|
|
@ -71,14 +71,16 @@ Compared to ChatGLM's [P-Tuning](https://github.com/THUDM/ChatGLM2-6B/tree/main/
|
|||
|
||||
## Changelog
|
||||
|
||||
[24/06/07] We supported fine-tuning the **[Qwen-2](https://qwenlm.github.io/blog/qwen2/)** series models.
|
||||
|
||||
[24/06/05] We supported fine-tuning the **[GLM-4-9B/GLM-4-9B-Chat](https://github.com/THUDM/GLM-4)** models.
|
||||
|
||||
[24/05/26] We supported **[SimPO](https://arxiv.org/abs/2405.14734)** algorithm for preference learning. See [examples](examples/README.md) for usage.
|
||||
|
||||
[24/05/20] We supported fine-tuning the **PaliGemma** series models. Note that the PaliGemma models are pre-trained models, you need to fine-tune them with `gemma` template for chat completion.
|
||||
|
||||
<details><summary>Full Changelog</summary>
|
||||
|
||||
[24/05/20] We supported fine-tuning the **PaliGemma** series models. Note that the PaliGemma models are pre-trained models, you need to fine-tune them with `gemma` template for chat completion.
|
||||
|
||||
[24/05/18] We supported **[KTO](https://arxiv.org/abs/2402.01306)** algorithm for preference learning. See [examples](examples/README.md) for usage.
|
||||
|
||||
[24/05/14] We supported training and inference on the Ascend NPU devices. Check [installation](#installation) section for details.
|
||||
|
|
@ -172,6 +174,7 @@ Compared to ChatGLM's [P-Tuning](https://github.com/THUDM/ChatGLM2-6B/tree/main/
|
|||
| [Phi-3](https://huggingface.co/microsoft) | 4B/7B/14B | phi |
|
||||
| [Qwen](https://huggingface.co/Qwen) | 1.8B/7B/14B/72B | qwen |
|
||||
| [Qwen1.5 (Code/MoE)](https://huggingface.co/Qwen) | 0.5B/1.8B/4B/7B/14B/32B/72B/110B | qwen |
|
||||
| [Qwen2 (MoE)](https://huggingface.co/Qwen) | 0.5B/1.5B/7B/57B/72B | qwen |
|
||||
| [StarCoder2](https://huggingface.co/bigcode) | 3B/7B/15B | - |
|
||||
| [XVERSE](https://huggingface.co/xverse) | 7B/13B/65B | xverse |
|
||||
| [Yi (1/1.5)](https://huggingface.co/01-ai) | 6B/9B/34B | yi |
|
||||
|
|
|
|||
|
|
@ -71,14 +71,16 @@ https://github.com/hiyouga/LLaMA-Factory/assets/16256802/ec36a9dd-37f4-4f72-81bd
|
|||
|
||||
## 更新日志
|
||||
|
||||
[24/06/07] 我们支持了 **[Qwen-2](https://qwenlm.github.io/blog/qwen2/)** 系列模型的微调。
|
||||
|
||||
[24/06/05] 我们支持了 **[GLM-4-9B/GLM-4-9B-Chat](https://github.com/THUDM/GLM-4)** 模型的微调。
|
||||
|
||||
[24/05/26] 我们支持了 **[SimPO](https://arxiv.org/abs/2405.14734)** 偏好对齐算法。详细用法请参照 [examples](examples/README_zh.md)。
|
||||
|
||||
[24/05/20] 我们支持了 **PaliGemma** 系列模型的微调。注意 PaliGemma 是预训练模型，你需要使用 `gemma` 模板进行微调使其获得对话能力。
|
||||
|
||||
<details><summary>展开日志</summary>
|
||||
|
||||
[24/05/20] 我们支持了 **PaliGemma** 系列模型的微调。注意 PaliGemma 是预训练模型，你需要使用 `gemma` 模板进行微调使其获得对话能力。
|
||||
|
||||
[24/05/18] 我们支持了 **[KTO](https://arxiv.org/abs/2402.01306)** 偏好对齐算法。详细用法请参照 [examples](examples/README_zh.md)。
|
||||
|
||||
[24/05/14] 我们支持了昇腾 NPU 设备的训练和推理。详情请查阅[安装](#安装-llama-factory)部分。
|
||||
|
|
@ -172,6 +174,7 @@ https://github.com/hiyouga/LLaMA-Factory/assets/16256802/ec36a9dd-37f4-4f72-81bd
|
|||
| [Phi-3](https://huggingface.co/microsoft) | 4B/7B/14B | phi |
|
||||
| [Qwen](https://huggingface.co/Qwen) | 1.8B/7B/14B/72B | qwen |
|
||||
| [Qwen1.5 (Code/MoE)](https://huggingface.co/Qwen) | 0.5B/1.8B/4B/7B/14B/32B/72B/110B | qwen |
|
||||
| [Qwen2 (MoE)](https://huggingface.co/Qwen) | 0.5B/1.5B/7B/57B/72B | qwen |
|
||||
| [StarCoder2](https://huggingface.co/bigcode) | 3B/7B/15B | - |
|
||||
| [XVERSE](https://huggingface.co/xverse) | 7B/13B/65B | xverse |
|
||||
| [Yi (1/1.5)](https://huggingface.co/01-ai) | 6B/9B/34B | yi |
|
||||
|
|
|
|||
|
|
@ -1078,6 +1078,89 @@ register_model_group(
|
|||
)
|
||||
|
||||
|
||||
register_model_group(
|
||||
models={
|
||||
"Qwen2-0.5B": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-0.5B",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-0.5B",
|
||||
},
|
||||
"Qwen2-1.5B": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-1.5B",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-1.5B",
|
||||
},
|
||||
"Qwen2-7B": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-7B",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-7B",
|
||||
},
|
||||
"Qwen2-72B": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-72B",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-72B",
|
||||
},
|
||||
"Qwen2-MoE-57B": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-57B-A14B",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-57B-A14B",
|
||||
},
|
||||
"Qwen2-0.5B-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-0.5B-Instruct",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-0.5B-Instruct",
|
||||
},
|
||||
"Qwen2-1.5B-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-1.5B-Instruct",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-1.5B-Instruct",
|
||||
},
|
||||
"Qwen2-7B-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-7B-Instruct",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-7B-Instruct",
|
||||
},
|
||||
"Qwen2-72B-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-72B-Instruct",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-72B-Instruct",
|
||||
},
|
||||
"Qwen2-MoE-57B-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-57B-A14B-Instruct",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-57B-A14B-Instruct",
|
||||
},
|
||||
"Qwen2-0.5B-int8-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-0.5B-Instruct-GPTQ-Int8",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-0.5B-Instruct-GPTQ-Int8",
|
||||
},
|
||||
"Qwen2-0.5B-int4-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-0.5B-Instruct-AWQ",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-0.5B-Instruct-AWQ",
|
||||
},
|
||||
"Qwen2-1.5B-int8-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-1.5B-Instruct-GPTQ-Int8",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-1.5B-Instruct-GPTQ-Int8",
|
||||
},
|
||||
"Qwen2-1.5B-int4-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-1.5B-Instruct-AWQ",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-1.5B-Instruct-AWQ",
|
||||
},
|
||||
"Qwen2-7B-int8-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-7B-Instruct-GPTQ-Int8",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-7B-Instruct-GPTQ-Int8",
|
||||
},
|
||||
"Qwen2-7B-int4-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-7B-Instruct-AWQ",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-7B-Instruct-AWQ",
|
||||
},
|
||||
"Qwen2-72B-int8-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-72B-Instruct-GPTQ-Int8",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-72B-Instruct-GPTQ-Int8",
|
||||
},
|
||||
"Qwen2-72B-int4-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-72B-Instruct-AWQ",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-72B-Instruct-AWQ",
|
||||
},
|
||||
"Qwen2-MoE-57B-int4-Chat": {
|
||||
DownloadSource.DEFAULT: "Qwen/Qwen2-57B-A14B-Instruct-GPTQ-Int4",
|
||||
DownloadSource.MODELSCOPE: "qwen/Qwen2-57B-A14B-Instruct-GPTQ-Int4",
|
||||
},
|
||||
},
|
||||
template="qwen",
|
||||
)
|
||||
|
||||
|
||||
register_model_group(
|
||||
models={
|
||||
"SOLAR-10.7B": {
|
||||
|
|
|
|||
Loading…
Reference in New Issue