diff --git a/README.md b/README.md index 814ec809..04a412e8 100644 --- a/README.md +++ b/README.md @@ -51,7 +51,7 @@ be reused in [ms-swift](https://github.com/modelscope/ms-swift). pip install 'twinkle-kit' ``` -### Installation from Source: +### Install from Source: ```shell git clone https://github.com/modelscope/twinkle.git @@ -75,13 +75,14 @@ pip install -e . ## Changelog -- 🎉2026-02-10 Initial version of Twinkle✨ released, including SFT/PT/RL for text models and serverless training capabilities on [ModelScope](https://modelscope.cn). +- 🎉2026-02-13 Initial version of Twinkle✨ released, including SFT/PT/RL support for text models and serverless training capabilities on [ModelScope](https://modelscope.cn). -# ModelScope Community +## Training as a Service on ModelScope -## ModelScope Official Environment - -The ModelScope community provides an official environment for running Twinkle. The API endpoint is: [base_url](https://www.modelscope.cn/twinkle). Developers can refer to our [documentation](docs/source_en/Usage%20Guide/ModelScope-Official-Resources.md) for usage instructions. +We are rolling out training service built atop Twinkle✨ on ModelScope. It is currently in _Beta_. You may +sign up for free access by joining the [Twinkle-Explorers](https://modelscope.cn/organization/twinkle-explorers) organization, and +train via API endpoint `base_url=https://www.modelscope.cn/twinkle`. For more details, please refer to +our [documentation](docs/source_en/Usage%20Guide/ModelScope-Official-Resources.md). ## Supported Hardware @@ -95,29 +96,33 @@ The ModelScope community provides an official environment for running Twinkle. T ## Supported Models We will be adding support for more models as new models are released. The following table lists current models -supported on Twinkle✨ framework. However, the models supported on our serverless training backend may be a -much smaller subset. Please refer to the [doc](link) section for more information. - -| Model Type | Model ID on[ModelScope](https://modelscope.cn) | Requires | Megatron Support | HF Model ID | -| ------------------- | --------------------------------------------------------------------------------------------------------------------- | -------------------- | ---------------- | ---------------------------------------------------------------------------------------------------------- | -| qwen3 series | [Qwen/Qwen3-0.6B-Base](https://modelscope.cn/models/Qwen/Qwen3-0.6B-Base)~32B | transformers>=4.51 | ✔ | [Qwen/Qwen3-0.6B-Base](https://huggingface.co/Qwen/Qwen3-0.6B-Base) | -| qwen3_moe series | [Qwen/Qwen3-30B-A3B-Base](https://modelscope.cn/models/Qwen/Qwen3-30B-A3B-Base) | transformers>=4.51 | ✔ | [Qwen/Qwen3-30B-A3B-Base](https://huggingface.co/Qwen/Qwen3-30B-A3B-Base) | -| | [Qwen/Qwen3-30B-A3B](https://modelscope.cn/models/Qwen/Qwen3-30B-A3B)~235B | transformers>=4.51 | ✔ | [Qwen/Qwen3-30B-A3B](https://huggingface.co/Qwen/Qwen3-30B-A3B) | -| qwen2 series | [Qwen/Qwen2-0.5B-Instruct](https://modelscope.cn/models/Qwen/Qwen2-0.5B-Instruct) ~72B | transformers>=4.37 | ✔ | [Qwen/Qwen2-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2-0.5B-Instruct) | -| | [Qwen/Qwen2.5-0.5B-Instruct](https://modelscope.cn/models/Qwen/Qwen2.5-0.5B-Instruct)~72B | transformers>=4.37 | ✔ | [Qwen/Qwen2.5-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct) | -| | [Qwen/Qwen2.5-0.5B](https://modelscope.cn/models/Qwen/Qwen2.5-0.5B)~72B | transformers>=4.37 | ✔ | [Qwen/Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) | -| qwen2_moe series | [Qwen/Qwen1.5-MoE-A2.7B-Chat](https://modelscope.cn/models/Qwen/Qwen1.5-MoE-A2.7B-Chat) | transformers>=4.40 | ✔ | [Qwen/Qwen1.5-MoE-A2.7B-Chat](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B-Chat) | +supported on Twinkle✨ framework. + +>[!Note] +> For serverless training service accessed via `base_url=https://www.modelscope.cn/twinkle`, it currently supports +> one training base at a time, and currently it is [Qwen3-30B-A3B-Instruct-2507](https://modelscope.cn/models/Qwen/Qwen3-30B-A3B-Instruct-2507). + + +| Model Type | Model ID on [ModelScope](https://modelscope.cn) | Requires | Megatron Support | HF Model ID | +| ------------------- |--------------------------------------------------------------------------------------------------------------------------| -------------------- | ---------------- | ---------------------------------------------------------------------------------------------------------- | +| qwen3 series | [Qwen/Qwen3-0.6B-Base](https://modelscope.cn/models/Qwen/Qwen3-0.6B-Base)~32B | transformers>=4.51 | ✅ | [Qwen/Qwen3-0.6B-Base](https://huggingface.co/Qwen/Qwen3-0.6B-Base) | +| qwen3_moe series | [Qwen/Qwen3-30B-A3B-Base](https://modelscope.cn/models/Qwen/Qwen3-30B-A3B-Base) | transformers>=4.51 | ✅ | [Qwen/Qwen3-30B-A3B-Base](https://huggingface.co/Qwen/Qwen3-30B-A3B-Base) | +| | [Qwen/Qwen3-30B-A3B](https://modelscope.cn/models/Qwen/Qwen3-30B-A3B)~235B | transformers>=4.51 | ✅ | [Qwen/Qwen3-30B-A3B](https://huggingface.co/Qwen/Qwen3-30B-A3B) | +| qwen2 series | [Qwen/Qwen2-0.5B-Instruct](https://modelscope.cn/models/Qwen/Qwen2-0.5B-Instruct) ~72B | transformers>=4.37 | ✅ | [Qwen/Qwen2-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2-0.5B-Instruct) | +| | [Qwen/Qwen2.5-0.5B-Instruct](https://modelscope.cn/models/Qwen/Qwen2.5-0.5B-Instruct)~72B | transformers>=4.37 | ✅ | [Qwen/Qwen2.5-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct) | +| | [Qwen/Qwen2.5-0.5B](https://modelscope.cn/models/Qwen/Qwen2.5-0.5B)~72B | transformers>=4.37 | ✅ | [Qwen/Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B) | +| qwen2_moe series | [Qwen/Qwen1.5-MoE-A2.7B-Chat](https://modelscope.cn/models/Qwen/Qwen1.5-MoE-A2.7B-Chat) | transformers>=4.40 | ✅ | [Qwen/Qwen1.5-MoE-A2.7B-Chat](https://huggingface.co/Qwen/Qwen1.5-MoE-A2.7B-Chat) | | chatglm4 series | [ZhipuAI/glm-4-9b-chat](https://modelscope.cn/models/ZhipuAI/glm-4-9b-chat) | transformers>=4.42 | ✘ | [zai-org/glm-4-9b-chat](https://huggingface.co/zai-org/glm-4-9b-chat) | | | [ZhipuAI/LongWriter-glm4-9b](https://modelscope.cn/models/ZhipuAI/LongWriter-glm4-9b) | transformers>=4.42 | ✘ | [zai-org/LongWriter-glm4-9b](https://huggingface.co/zai-org/LongWriter-glm4-9b) | | glm_edge series | [ZhipuAI/glm-edge-1.5b-chat](https://modelscope.cn/models/ZhipuAI/glm-edge-1.5b-chat) | transformers>=4.46 | ✘ | [zai-org/glm-edge-1.5b-chat](https://huggingface.co/zai-org/glm-edge-1.5b-chat) | | | [ZhipuAI/glm-edge-4b-chat](https://modelscope.cn/models/ZhipuAI/glm-edge-4b-chat) | transformers>=4.46 | ✘ | [zai-org/glm-edge-4b-chat](https://huggingface.co/zai-org/glm-edge-4b-chat) | | internlm2 series | [Shanghai_AI_Laboratory/internlm2-1_8b](https://modelscope.cn/models/Shanghai_AI_Laboratory/internlm2-1_8b) | transformers>=4.38 | ✘ | [internlm/internlm2-1_8b](https://huggingface.co/internlm/internlm2-1_8b) | | | [Shanghai_AI_Laboratory/internlm2-chat-7b](https://modelscope.cn/models/Shanghai_AI_Laboratory/internlm2-chat-7b) | transformers>=4.38 | ✘ | [internlm/internlm2-chat-7b](https://huggingface.co/internlm/internlm2-chat-7b) | -| deepseek_v1 | [deepseek-ai/deepseek-vl-7b-chat](https://modelscope.cn/models/deepseek-ai/deepseek-vl-7b-chat) | transformers>=4.39.4 | ✔ | —— | -| | [deepseek-ai/DeepSeek-V2-Lite](https://modelscope.cn/models/deepseek-ai/DeepSeek-V2-Lite) | transformers>=4.39.3 | ✔ | [deepseek-ai/DeepSeek-V2-Lite](https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite) | -| | [deepseek-ai/DeepSeek-V2.5](https://modelscope.cn/models/deepseek-ai/DeepSeek-V2.5) | transformers>=4.39.3 | ✔ | [deepseek-ai/DeepSeek-V2.5](https://huggingface.co/deepseek-ai/DeepSeek-V2.5) | -| | [deepseek-ai/DeepSeek-R1](https://modelscope.cn/models/deepseek-ai/DeepSeek-R1) | transformers>=4.39.3 | ✔ | [deepseek-ai/DeepSeek-R1](https://huggingface.co/deepseek-ai/DeepSeek-R1) | -| deepSeek-r1-distill | [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://modelscope.cn/models/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) ~32B | transformers>=4.37 | ✔ | [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) | +| deepseek_v1 | [deepseek-ai/deepseek-vl-7b-chat](https://modelscope.cn/models/deepseek-ai/deepseek-vl-7b-chat) | transformers>=4.39.4 | ✅ | —— | +| | [deepseek-ai/DeepSeek-V2-Lite](https://modelscope.cn/models/deepseek-ai/DeepSeek-V2-Lite) | transformers>=4.39.3 | ✅ | [deepseek-ai/DeepSeek-V2-Lite](https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite) | +| | [deepseek-ai/DeepSeek-V2.5](https://modelscope.cn/models/deepseek-ai/DeepSeek-V2.5) | transformers>=4.39.3 | ✅ | [deepseek-ai/DeepSeek-V2.5](https://huggingface.co/deepseek-ai/DeepSeek-V2.5) | +| | [deepseek-ai/DeepSeek-R1](https://modelscope.cn/models/deepseek-ai/DeepSeek-R1) | transformers>=4.39.3 | ✅ | [deepseek-ai/DeepSeek-R1](https://huggingface.co/deepseek-ai/DeepSeek-R1) | +| deepSeek-r1-distill | [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://modelscope.cn/models/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) ~32B | transformers>=4.37 | ✅ | [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) | For a more detailed model support list 👉 [Quick Start.md](https://github.com/modelscope/twinkle/blob/dev/docs/source/%E4%BD%BF%E7%94%A8%E6%8C%87%E5%BC%95/%E5%BF%AB%E9%80%9F%E5%BC%80%E5%A7%8B.md) @@ -141,10 +146,12 @@ twinkle.initialize(mode='ray', groups=device_group, global_device_mesh=device_me def train(): + # to load model from Hugging Face, use 'hf://...' + base_model = 'ms://Qwen/Qwen2.5-7B-Instruct' # 1000 samples dataset = Dataset(dataset_meta=DatasetMeta('ms://swift/self-cognition', data_slice=range(1000))) # Set template to prepare encoding - dataset.set_template('Template', model_id='ms://Qwen/Qwen2.5-7B-Instruct') + dataset.set_template('Template', model_id=base_model) # Preprocess the dataset to standard format dataset.map(SelfCognitionProcessor('twinkle LLM', 'ModelScope Community')) # Encode dataset @@ -152,7 +159,7 @@ def train(): # Global batch size = 8, for GPUs, so 1 sample per GPU dataloader = DataLoader(dataset=dataset, batch_size=8, min_batch_size=8) # Use a TransformersModel - model = TransformersModel(model_id='ms://Qwen/Qwen2.5-7B-Instruct', remote_group='default') + model = TransformersModel(model_id=base_model, remote_group='default') lora_config = LoraConfig( r=8, @@ -184,7 +191,7 @@ if __name__ == '__main__': train() ``` -### Tinker-Like Remote API +### Using Tinker-Like API ```python import os @@ -196,17 +203,19 @@ from twinkle.dataset import Dataset, DatasetMeta from twinkle.preprocessor import SelfCognitionProcessor from twinkle.server.tinker.common import input_feature_to_datum -base_model = "Qwen/Qwen2.5-0.5B-Instruct" +base_model = 'ms://Qwen/Qwen2.5-0.5B-Instruct' +base_url='http://www.modelscope.cn/twinkle' +api_key=os.environ.get('MODELSCOPE_TOKEN') # Use twinkle dataset to load the data dataset = Dataset(dataset_meta=DatasetMeta('ms://swift/self-cognition', data_slice=range(500))) -dataset.set_template('Template', model_id=f'ms://{base_model}', max_length=256) +dataset.set_template('Template', model_id=base_model, max_length=256) dataset.map(SelfCognitionProcessor('twinkle Model', 'twinkle Team'), load_from_cache_file=False) dataset.encode(batched=True, load_from_cache_file=False) dataloader = DataLoader(dataset=dataset, batch_size=8) # Initialize tinker client -service_client = init_tinker_compat_client(base_url='http://www.modelscope.cn/twinkle', api_key=os.environ.get('MODELSCOPE_SDK_TOKEN')) +service_client = init_tinker_compat_client(base_model, api_key) training_client = service_client.create_lora_training_client(base_model=base_model, rank=16) # Training loop: use input_feature_to_datum to transfer the input format @@ -223,12 +232,6 @@ for epoch in range(3): training_client.save_state(f"twinkle-lora-{epoch}").result() ``` -Launch training: - -```shell -python3 train.py -``` - ## Architecture Design diff --git a/README_ZH.md b/README_ZH.md index 134c75a4..1b4cfa40 100644 --- a/README_ZH.md +++ b/README_ZH.md @@ -69,7 +69,7 @@ pip install -e . ## 魔搭社区官方环境 -魔搭社区提供了Twinkle运行的官方环境，调用端点为:[base_url](https://www.modelscope.cn/twinkle)，开发者可以参考我们的[文档](docs/source_zh/使用指引/魔搭官方环境.md)来进行使用。 +魔搭社区提供了Twinkle运行的官方环境，调用端点为:[base_url](https://www.modelscope.cn/twinkle)，开发者可以参考我们的[文档](docs/source_zh/使用指引/训练服务.md)来进行使用。 ## 支持的硬件 diff --git "a/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\345\277\253\351\200\237\345\274\200\345\247\213.md" "b/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\345\277\253\351\200\237\345\274\200\345\247\213.md" index 0c2e5d0a..56abef89 100644 --- "a/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\345\277\253\351\200\237\345\274\200\345\247\213.md" +++ "b/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\345\277\253\351\200\237\345\274\200\345\247\213.md" @@ -1,17 +1,12 @@ -
## ✨ Twinkle 是什么？ 大模型训练组件库。基于 PyTorch，更简洁、更灵活、生产就绪。 -

🧩 松耦合架构 · 标准化接口
🚀 多运行模式 · torchrun / Ray / HTTP
🔌 多框架兼容 · Transformers / Megatron
👥 多租户支持 · 单基座模型部署 -

- -
## Twinkle 适配性 @@ -23,7 +18,6 @@ Twinkle 和 [ms-swift](https://github.com/modelscope/ms-swift) 都是模型训 - 如果你是大模型研究者，希望定制模型或训练方法 - 如果你善于编写 training loop，希望定制训练过程 - 如果你希望提供企业级或商业化训练平台 -- 如果你缺少训练硬件，希望使用社区资源 ### 何时选择ms-swift diff --git "a/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\351\255\224\346\220\255\345\256\230\346\226\271\347\216\257\345\242\203.md" "b/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\350\256\255\347\273\203\346\234\215\345\212\241.md" similarity index 54% rename from "docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\351\255\224\346\220\255\345\256\230\346\226\271\347\216\257\345\242\203.md" rename to "docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\350\256\255\347\273\203\346\234\215\345\212\241.md" index ee8369b8..43f2d84c 100644 --- "a/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\351\255\224\346\220\255\345\256\230\346\226\271\347\216\257\345\242\203.md" +++ "b/docs/source_zh/\344\275\277\347\224\250\346\214\207\345\274\225/\350\256\255\347\273\203\346\234\215\345\212\241.md" @@ -1,24 +1,23 @@ -# 魔搭官方环境 +# ModelScope上的Twinkle训练服务 -在 Twinkle 框架开源的同时，我们在[魔搭社区官网](https://www.modelscope.cn)上提供了样例训练资源。开发者可以通过ModelScope Token进行训练。 +在 Twinkle 框架开源的同时，我们依托ModelScope的后台服务，也提供了托管的模型训练服务(Training as a Service），开发者可以通过这一服务， +免费体验Twinkle的训练API。 目前在集群中运行的模型是[Qwen/Qwen3-30B-A3B-Instruct-2507](https://www.modelscope.cn/models/Qwen/Qwen3-30B-A3B-Instruct-2507)。下面介绍具体的使用方法： -## Step 1. 注册魔搭用户 +## Step 1. 注册ModelScope用户并申请加入 twinkle-explorers 组织 -开发者首先需要注册成为魔搭用户，使用魔搭社区的 token 进行调用。 +开发者首先需要注册成为ModelScope用户，并申请加入 [Twinkle-Explorers](https://modelscope.cn/organization/twinkle-explorers) 组织， +来获取访问权限。当前免费的Serverless训练体验，还在灰度测试中，暂时只向组织内的用户开放。您也可以通过本地部署服务，来使用Twinkle✨。 注册地址：https://www.modelscope.cn/ -调用端点：https://www.modelscope.cn/twinkle +在注册并获批加入[Twinkle-Explorers](https://modelscope.cn/organization/twinkle-explorers) 组织后，在此页面获取 +访问的API-Key（即ModelScope平台的访问Token）：https://www.modelscope.cn/my/access/token 。 -token 在这里获取：https://www.modelscope.cn/my/access/token 拷贝访问令牌。 +调用端点：`base_url="https://www.modelscope.cn/twinkle"` -## Step 2. 加入 twinkle-explorers 组织 - -目前 twinkle-kit 的远程训练能力在灰度测试中，开发者需要加入 [twinkle-explorers](https://www.modelscope.cn/models/twinkle-explorers) 组织，组织内的用户可以进行使用和测试。 - -## Step 3. 查看 Cookbook 并二次定制开发 +## Step 2. 查看 Cookbook 并二次定制开发 我们强烈推荐开发者查看我们的 [cookbook](https://github.com/modelscope/twinkle/tree/main/cookbook/client/)，并根据其中的训练代码进行二次开发。