Skip to content

Commit db6a85a

Browse files
nullnonenilNULLnullnonenilNULL
authored
Merge pull request baidu-baige#144 from nullnonenilNULL/master
[docs] chore: refresh model examples listing
2 parents d4ec37a + 746c26f commit db6a85a

7 files changed

Lines changed: 141 additions & 580 deletions

File tree

‎.gitignore‎

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -93,3 +93,6 @@ swanlog/
9393
*.prof
9494
nsys_*/
9595
nsight_*/
96+
97+
# Launch / marketing kit (local only, not for publishing)
98+
/launch-kit/

‎README.md‎

Lines changed: 69 additions & 39 deletions
Original file line numberDiff line numberDiff line change
@@ -136,45 +136,75 @@ Open-source models trained with LoongForge or its predecessor AIAK-Training-LLM:
136136
<a id="models"></a>
137137
## 🏛️ Supported Models
138138

139-
LoongForge supports a broad range of [state-of-the-art models](https://loongforge.readthedocs.io/en/latest/get_started/support_model.html) across LLM, VLM, diffusion, and embodied.
140-
141-
| **Category** | **Architectures** | **Models** |
142-
|---------------|------------------|------------|
143-
| **LLM** | DeepSeek-V2 | deepseek-v2-lite, deepseek-v2 |
144-
| | DeepSeek-V3 | deepseek-v3, deepseek-v32 |
145-
| | DeepSeek-V4 | deepseek-v4-flash, deepseek-v4-pro |
146-
| | LLaMA2 | llama2-7b, llama2-13b, llama2-70b |
147-
| | LLaMA3 | llama3-8b, llama3-70b |
148-
| | LLaMA3.1 | llama3.1-8b, llama3.1-70b, llama3.1-405b |
149-
| | Qwen | qwen-1.8b → qwen-72b |
150-
| | Qwen1.5 | qwen1.5-0.5b → qwen1.5-72b |
151-
| | Qwen2 | qwen2-0.5b → qwen2-72b |
152-
| | Qwen2.5 | qwen2.5-0.5b → qwen2.5-72b |
153-
| | Qwen3 | qwen3-0.6b → qwen3-480b-a35b, qwen3-coder-30b-a3b |
154-
| | Qwen3-Next | qwen3-next-80b-a3b |
155-
| | MiniMax | minimax-m2.1, minimax-m2.5, minimax-m2.7 |
156-
| | MIMO | mimo-7b |
157-
| | GLM | glm5 |
158-
| **VLM** | Qwen2.5-VL | qwen2.5-vl-3b → qwen2.5-vl-72b |
159-
| | Qwen3-VL | qwen3-vl-30b-a3b, qwen3-vl-235b-a22b |
160-
| | Qwen3.5 | qwen3.5-0.8b → qwen3.5-397b-a17b |
161-
| | Qwen3.6 | qwen3.6-27b, qwen3.6-35b-a3b |
162-
| | Kimi-K2.5 | kimi-k2.5, kimi-k2.6 |
163-
| | ERNIE4.5-VL | ernie4.5vl-28b-a3b |
164-
| | LLaVA-OneVision-1.5 | llava-onevision-1.5-4b |
165-
| | InternVL2.5 | internvl2.5-8b → internvl2.5-78b |
166-
| | InternVL3.5 | internvl3.5-8b → internvl3.5-241b-a28b |
167-
| | CustomCombinedModel | Flexible ViT + LLM backbone configuration ([example](https://github.com/baidu-baige/LoongForge/blob/master/configs/models/custom/qwen_vit_llama3_8b.yaml)) |
168-
| **Diffusion** | Wan2.1 | wan2-1-i2v |
169-
| | Wan2.2 | wan2-2-i2v |
170-
| | Qwen-Image | qwen-image-edit-2511 |
171-
| **Embodied** | Pi | pi0.5 |
172-
| | GR00T | groot_n1_6, groot_n1_7 |
173-
| | XVLA | xvla |
174-
| | FastWAM | fastwam |
175-
| | LingBot-VA | lingbot_va |
176-
| | Cosmos3 | cosmos3_nano |
177-
| | DreamZero | dreamzero_lora_wan22_5b, dreamzero_full_wan22_5b, dreamzero_lora_wan21_14b, dreamzero_full_wan21_14b, dreamzero_libero_wan22_5b, dreamzero_agibot_wan21_14b, dreamzero_yam_wan21_14b |
139+
LoongForge supports a broad range of model families across LLM, VLM, diffusion, and embodied. Select a model below to open its training examples. For complete usage instructions, see the [User Guide](https://loongforge.readthedocs.io/en/latest/index.html) and the full [model support matrix](https://loongforge.readthedocs.io/en/latest/get_started/support_model.html).
140+
141+
<table width="100%" style="table-layout: fixed; border-collapse: collapse;">
142+
<colgroup>
143+
<col width="25%">
144+
<col width="25%">
145+
<col width="25%">
146+
<col width="25%">
147+
</colgroup>
148+
<thead align="center" valign="bottom">
149+
<tr><th width="25%">LLM</th><th width="25%">VLM</th><th width="25%">Diffusion</th><th width="25%">Embodied</th></tr>
150+
</thead>
151+
<tbody valign="top">
152+
<tr>
153+
<td valign="top">
154+
<ul>
155+
<li><a href="examples/deepseek_v2/">DeepSeek-V2</a> ✅</li>
156+
<li><a href="examples/deepseek_v3/">DeepSeek-V3/V3.2</a> ✅</li>
157+
<li><a href="examples/deepseek_v4/">DeepSeek-V4</a> ✅</li>
158+
<li><a href="examples/llama2/">LLaMA2</a> ✅</li>
159+
<li><a href="examples/llama3/">LLaMA3</a> ✅</li>
160+
<li><a href="examples/llama3.1/">LLaMA3.1</a> ✅</li>
161+
<li><a href="examples/qwen/">Qwen</a> ✅</li>
162+
<li><a href="examples/qwen1.5/">Qwen1.5</a> ✅</li>
163+
<li><a href="examples/qwen2/">Qwen2</a> ✅</li>
164+
<li><a href="examples/qwen2.5/">Qwen2.5</a> ✅</li>
165+
<li><a href="examples/qwen3/">Qwen3</a> ✅</li>
166+
<li><a href="examples/qwen3_next/">Qwen3-Next</a> ✅</li>
167+
<li><a href="examples/minimax/">MiniMax-M2.1/2.5/2.7</a> ✅</li>
168+
<li><a href="examples/mimo/">MIMO</a> ✅</li>
169+
<li><a href="examples/glm5/">GLM-5</a> ✅</li>
170+
</ul>
171+
</td>
172+
<td valign="top">
173+
<ul>
174+
<li><a href="examples/qwen2.5_vl/">Qwen2.5-VL</a> ✅</li>
175+
<li><a href="examples/qwen3_vl/">Qwen3-VL</a> ✅</li>
176+
<li><a href="examples/qwen3.5/">Qwen3.5</a> ✅</li>
177+
<li><a href="examples/qwen3.6/">Qwen3.6</a> ✅</li>
178+
<li><a href="examples/kimi_k2.x/kimi_k2.5/">Kimi-K2.5/2.6</a> ✅</li>
179+
<li><a href="examples/ernie4.5/">ERNIE4.5-VL</a> ✅</li>
180+
<li><a href="examples/llava_onevision_1.5/">LLaVA-OneVision-1.5</a> ✅</li>
181+
<li><a href="examples/internvl2.5/">InternVL2.5</a> ✅</li>
182+
<li><a href="examples/internvl3.5/">InternVL3.5</a> ✅</li>
183+
<li><a href="examples/custom/">CustomCombinedModel Example</a> ✅</li>
184+
</ul>
185+
</td>
186+
<td valign="top">
187+
<ul>
188+
<li><a href="examples/wan/">Wan2.1</a> ✅</li>
189+
<li><a href="examples/wan/">Wan2.2</a> ✅</li>
190+
<li><a href="examples/qwen_image/">Qwen-Image</a> ✅</li>
191+
</ul>
192+
</td>
193+
<td valign="top">
194+
<ul>
195+
<li><a href="examples/embodied/pi05/">Pi0.5</a> ✅</li>
196+
<li><a href="examples/embodied/groot_n1_6/">GR00T-N1.6</a> ✅</li>
197+
<li><a href="examples/embodied/groot_n1_7/">GR00T-N1.7</a> ✅</li>
198+
<li><a href="examples/embodied/xvla/">xVLA</a> ✅</li>
199+
<li><a href="examples/embodied/fastwam/">FastWAM</a> ✅</li>
200+
<li><a href="examples/embodied/lingbot_va/">LingBot-VA</a> ✅</li>
201+
<li><a href="examples/embodied/cosmos3/">Cosmos3</a> ✅</li>
202+
<li><a href="examples/embodied/dreamzero/">DreamZero</a> ✅</li>
203+
</ul>
204+
</td>
205+
</tr>
206+
</tbody>
207+
</table>
178208

179209
## 📂 Repository Layout
180210

‎README_zh.md‎

Lines changed: 69 additions & 39 deletions
Original file line numberDiff line numberDiff line change
@@ -135,45 +135,75 @@
135135
<a id="models"></a>
136136
## 🏛️ 支持的模型
137137

138-
LoongForge 已支持 LLM、VLM、Diffusion 与 Embodied 等多模态的[广泛的 SOTA 模型](https://loongforge.readthedocs.io/zh-cn/latest/get_started/support_model.html)。
139-
140-
| **类别** | **架构** | **模型** |
141-
|---------------|------------------|------------|
142-
| **LLM** | DeepSeek-V2 | deepseek-v2-lite, deepseek-v2 |
143-
| | DeepSeek-V3 | deepseek-v3, deepseek-v32 |
144-
| | DeepSeek-V4 | deepseek-v4-flash, deepseek-v4-pro |
145-
| | LLaMA2 | llama2-7b, llama2-13b, llama2-70b |
146-
| | LLaMA3 | llama3-8b, llama3-70b |
147-
| | LLaMA3.1 | llama3.1-8b, llama3.1-70b, llama3.1-405b |
148-
| | Qwen | qwen-1.8b → qwen-72b |
149-
| | Qwen1.5 | qwen1.5-0.5b → qwen1.5-72b |
150-
| | Qwen2 | qwen2-0.5b → qwen2-72b |
151-
| | Qwen2.5 | qwen2.5-0.5b → qwen2.5-72b |
152-
| | Qwen3 | qwen3-0.6b → qwen3-480b-a35b, qwen3-coder-30b-a3b |
153-
| | Qwen3-Next | qwen3-next-80b-a3b |
154-
| | MiniMax | minimax-m2.1, minimax-m2.5, minimax-m2.7 |
155-
| | MIMO | mimo-7b |
156-
| | GLM | glm5 |
157-
| **VLM** | Qwen2.5-VL | qwen2.5-vl-3b → qwen2.5-vl-72b |
158-
| | Qwen3-VL | qwen3-vl-30b-a3b, qwen3-vl-235b-a22b |
159-
| | Qwen3.5 | qwen3.5-0.8b → qwen3.5-397b-a17b |
160-
| | Qwen3.6 | qwen3.6-27b, qwen3.6-35b-a3b |
161-
| | Kimi-K2.5 | kimi-k2.5, kimi-k2.6 |
162-
| | ERNIE4.5-VL | ernie4.5vl-28b-a3b |
163-
| | LLaVA-OneVision-1.5 | llava-onevision-1.5-4b |
164-
| | InternVL2.5 | internvl2.5-8b → internvl2.5-78b |
165-
| | InternVL3.5 | internvl3.5-8b → internvl3.5-241b-a28b |
166-
| | CustomCombinedModel | ViT + LLM backbone 灵活组合（[示例](https://github.com/baidu-baige/LoongForge/blob/master/configs/models/custom/qwen_vit_llama3_8b.yaml)） |
167-
| **Diffusion** | Wan2.1 | wan2-1-i2v |
168-
| | Wan2.2 | wan2-2-i2v |
169-
| | Qwen-Image | qwen-image-edit-2511 |
170-
| **Embodied** | Pi | pi0.5 |
171-
| | GR00T | groot_n1_6, groot_n1_7 |
172-
| | XVLA | xvla |
173-
| | FastWAM | fastwam |
174-
| | LingBot-VA | lingbot_va |
175-
| | Cosmos3 | cosmos3_nano |
176-
| | DreamZero | dreamzero_lora_wan22_5b, dreamzero_full_wan22_5b, dreamzero_lora_wan21_14b, dreamzero_full_wan21_14b, dreamzero_libero_wan22_5b, dreamzero_agibot_wan21_14b, dreamzero_yam_wan21_14b |
138+
LoongForge 已支持 LLM、VLM、Diffusion 与 Embodied 等类别的广泛模型家族。点击模型名称可查看对应的训练示例；完整的使用说明请参阅[用户手册](https://loongforge.readthedocs.io/zh-cn/latest/index.html)，模型变体请参阅[模型支持矩阵](https://loongforge.readthedocs.io/zh-cn/latest/get_started/support_model.html)。
139+
140+
<table width="100%" style="table-layout: fixed; border-collapse: collapse;">
141+
<colgroup>
142+
<col width="25%">
143+
<col width="25%">
144+
<col width="25%">
145+
<col width="25%">
146+
</colgroup>
147+
<thead align="center" valign="bottom">
148+
<tr><th width="25%">LLM</th><th width="25%">VLM</th><th width="25%">Diffusion</th><th width="25%">Embodied</th></tr>
149+
</thead>
150+
<tbody valign="top">
151+
<tr>
152+
<td valign="top">
153+
<ul>
154+
<li><a href="examples/deepseek_v2/">DeepSeek-V2</a> ✅</li>
155+
<li><a href="examples/deepseek_v3/">DeepSeek-V3/V3.2</a> ✅</li>
156+
<li><a href="examples/deepseek_v4/">DeepSeek-V4</a> ✅</li>
157+
<li><a href="examples/llama2/">LLaMA2</a> ✅</li>
158+
<li><a href="examples/llama3/">LLaMA3</a> ✅</li>
159+
<li><a href="examples/llama3.1/">LLaMA3.1</a> ✅</li>
160+
<li><a href="examples/qwen/">Qwen</a> ✅</li>
161+
<li><a href="examples/qwen1.5/">Qwen1.5</a> ✅</li>
162+
<li><a href="examples/qwen2/">Qwen2</a> ✅</li>
163+
<li><a href="examples/qwen2.5/">Qwen2.5</a> ✅</li>
164+
<li><a href="examples/qwen3/">Qwen3</a> ✅</li>
165+
<li><a href="examples/qwen3_next/">Qwen3-Next</a> ✅</li>
166+
<li><a href="examples/minimax/">MiniMax-M2.1/2.5/2.7</a> ✅</li>
167+
<li><a href="examples/mimo/">MIMO</a> ✅</li>
168+
<li><a href="examples/glm5/">GLM-5</a> ✅</li>
169+
</ul>
170+
</td>
171+
<td valign="top">
172+
<ul>
173+
<li><a href="examples/qwen2.5_vl/">Qwen2.5-VL</a> ✅</li>
174+
<li><a href="examples/qwen3_vl/">Qwen3-VL</a> ✅</li>
175+
<li><a href="examples/qwen3.5/">Qwen3.5</a> ✅</li>
176+
<li><a href="examples/qwen3.6/">Qwen3.6</a> ✅</li>
177+
<li><a href="examples/kimi_k2.x/kimi_k2.5/">Kimi-K2.5/2.6</a> ✅</li>
178+
<li><a href="examples/ernie4.5/">ERNIE4.5-VL</a> ✅</li>
179+
<li><a href="examples/llava_onevision_1.5/">LLaVA-OneVision-1.5</a> ✅</li>
180+
<li><a href="examples/internvl2.5/">InternVL2.5</a> ✅</li>
181+
<li><a href="examples/internvl3.5/">InternVL3.5</a> ✅</li>
182+
<li><a href="examples/custom/">CustomCombinedModel Example</a> ✅</li>
183+
</ul>
184+
</td>
185+
<td valign="top">
186+
<ul>
187+
<li><a href="examples/wan/">Wan2.1</a> ✅</li>
188+
<li><a href="examples/wan/">Wan2.2</a> ✅</li>
189+
<li><a href="examples/qwen_image/">Qwen-Image</a> ✅</li>
190+
</ul>
191+
</td>
192+
<td valign="top">
193+
<ul>
194+
<li><a href="examples/embodied/pi05/">Pi0.5</a> ✅</li>
195+
<li><a href="examples/embodied/groot_n1_6/">GR00T-N1.6</a> ✅</li>
196+
<li><a href="examples/embodied/groot_n1_7/">GR00T-N1.7</a> ✅</li>
197+
<li><a href="examples/embodied/xvla/">xVLA</a> ✅</li>
198+
<li><a href="examples/embodied/fastwam/">FastWAM</a> ✅</li>
199+
<li><a href="examples/embodied/lingbot_va/">LingBot-VA</a> ✅</li>
200+
<li><a href="examples/embodied/cosmos3/">Cosmos3</a> ✅</li>
201+
<li><a href="examples/embodied/dreamzero/">DreamZero</a> ✅</li>
202+
</ul>
203+
</td>
204+
</tr>
205+
</tbody>
206+
</table>
177207

178208
## 📂 代码结构
179209

‎examples/qwen2.5/pretrain/bridge_debug.sh‎

Lines changed: 0 additions & 122 deletions
This file was deleted.

0 commit comments

Comments
 (0)