[MM][CG] Support ViT full CUDA graph for DeepSeek-VL2 (#46005)

Signed-off-by: littlecircle0730 <littlecircle0730@gmail.com>
Signed-off-by: littlecircle0730 <43994952+littlecircle0730@users.noreply.github.com>
Co-authored-by: Shanshan Shen <467638484@qq.com>
This commit is contained in:
littlecircle0730
2026-10-15 21:38:52 +08:00
committed by GitHub
co-authored by Shanshan Shen
parent 0951fc183c
commit 71536bd5f8
5 changed files with 482 additions and 4 deletions
+1
View File
@@ -12,6 +12,7 @@ For two-tower vision encoders (e.g., DeepSeek-OCR's SAM + CLIP with dynamic tili
| Architecture | Models | CG for Image | CG for Video | Multi-Path Graph |
| ------------ | ------ | ------------ | ------------ | --------------- |
| `DeepseekOCRForCausalLM` | `DeepSeek-OCR` | ✅︎ | ❌︎ | ✅︎ |
| `DeepseekVLV2ForCausalLM` | `DeepSeek-VL2` | ✅︎ | ❌︎ | ❌︎ |
| `DeepseekV41ForCausalLM` | `DeepSeek-V4.1-Flash` | ✅︎ | ❌︎ | ❌︎ |
| `Ernie4_5_VLMoeForConditionalGeneration` | `ERNIE-4.5-VL` | ✅︎ | ❌︎ | ❌︎ |
| `Gemma3ForConditionalGeneration` | `Gemma3` | ✅︎ | ❌︎ | ❌︎ |