Update README.md
Browse files
README.md
CHANGED
|
@@ -3,3 +3,17 @@ license: other
|
|
| 3 |
license_name: waymo-open-dataset-non-commercial-use
|
| 4 |
license_link: https://waymo.com/open/terms/
|
| 5 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
license_name: waymo-open-dataset-non-commercial-use
|
| 4 |
license_link: https://waymo.com/open/terms/
|
| 5 |
---
|
| 6 |
+
|
| 7 |
+
**Folder structure:**
|
| 8 |
+
|
| 9 |
+
```bash
|
| 10 |
+
<InfiniCube_repo>/checkpoints
|
| 11 |
+
βββ gsm_vs02_res512_view1_dual_branch_sky_mlp_modulator.ckpt # Stage 3, dual branch reconstruction model
|
| 12 |
+
βββ vae_epoch7_step6250.ckpt # Stage 1, 3D voxel VAE
|
| 13 |
+
βββ voxel_diffusion.ckpt # Stage 1, HD Map conditioned voxel diffusion model
|
| 14 |
+
βββ wan14b-i2v-buffer-step-900.safetensors # Stage 2, 14B buffer-guided image-to-video wan2.1 model.
|
| 15 |
+
βββ wan14b-t2v-buffer-step-1200.safetensors # Stage 2, 14B buffer-guided text-to-video wan2.1 model.
|
| 16 |
+
βββ wan1pt3b-t2v-buffer-step-3500.safetensors # Stage 2, 1.3B buffer-guided image-to-video wan2.1 model (not performant)
|
| 17 |
+
```
|
| 18 |
+
|
| 19 |
+
In current inference script in stage 2, we only use `wan14b-t2v-buffer-step-1200.safetensors` by default.
|