[diffusion] model: support LTX2.3 two stage (#22182)

This commit is contained in:
Mick
2026-04-12 22:15:57 +08:00
committed by GitHub
parent 31453bb76a
commit 495ef8ec64
21 changed files with 1541 additions and 859 deletions
+4 -4
View File
@@ -33,16 +33,16 @@ default parameters when initializing and generating videos.
| TurboWan2.1 T2V 14B | `IPostYellow/TurboWan2.1-T2V-14B-Diffusers` | 480p | ✅ | ❌ | ❌ | ❌ | ✅ | ✅ | ⭕ |
| TurboWan2.1 T2V 14B 720P | `IPostYellow/TurboWan2.1-T2V-14B-720P-Diffusers` | 720p | ✅ | ❌ | ❌ | ❌ | ✅ | ✅ | ⭕ |
| TurboWan2.2 I2V A14B | `IPostYellow/TurboWan2.2-I2V-A14B-Diffusers` | 720p | ✅ | ❌ | ❌ | ❌ | ✅ | ✅ | ⭕ |
| LTX-2 | `Lightricks/LTX-2` | 768×512<br>1536×1024 | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
| LTX-2.3 | `Lightricks/LTX-2.3` | 768×512<br>1536×1024 | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
| LTX-2 (one and two stages) | `Lightricks/LTX-2` | 768×512<br>1536×1024 | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
| LTX-2.3 (one and two stages) | `Lightricks/LTX-2.3` | 768×512<br>1536×1024 | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
**Note**:
1. Wan2.2 TI2V 5B has some quality issues when performing I2V generation. We are working on fixing this issue.
2. SageSLA is based on SpargeAttn. Install it first with `pip install git+https://github.com/thu-ml/SpargeAttn.git --no-build-isolation`
3. LTX-2 two-stage generation uses `--pipeline-class-name LTX2TwoStagePipeline`. The spatial upsampler and distilled LoRA are auto-resolved from the model snapshot by default, and can still be overridden with `--spatial-upsampler-path` and `--distilled-lora-path`.
4. `Lightricks/LTX-2.3` is supported through the bundled native overlay materialization path. One-stage generation uses the default `LTX2Pipeline`; two-stage generation uses `--pipeline-class-name LTX2TwoStagePipeline`.
5. For LTX models, the `Resolutions` column uses output video `width×height` semantics, matching `sglang generate --width ... --height ...`. One-stage generation is validated at `768×512`; two-stage generation is validated at `1536×1024`.
3. LTX-2 and LTX-2.3 two-stage generation uses `--pipeline-class-name LTX2TwoStagePipeline`. The spatial upsampler and distilled LoRA are auto-resolved from the model snapshot by default, and can still be overridden with `--spatial-upsampler-path` and `--distilled-lora-path`.
- For LTX models, the `Resolutions` column uses output video `width×height` semantics, matching `sglang generate --width ... --height ...`.
### Image Generation Models