Image-to-Video
Cosmos
world-model
autonomous-driving
video-prediction
thermal
lidar
adverse-weather
lora
Daishi974 commited on
Commit
3dd1aff
·
verified ·
1 Parent(s): 75e742c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -49,7 +49,7 @@ Given **5 observed frames** (0.5 s at 10 Hz), it forecasts the **next 20 frames
49
  |---|---|
50
  | `tricosmos_ema_bf16.pt` | Model weights (EMA, bf16, 3.9 GB): the base model with the trained LoRA weights, training step 500. This is the checkpoint behind all reported results. |
51
  | `tricosmos_infer.py` | Inference script: one forecast from a 2×2 grid video; writes the forecast grid and one video per sensor. Also runs the zero-shot base model for comparison. |
52
- | `notebooks/TriCosmos_inference.ipynb` | Google Colab notebook: install, download, forecast, and a side-by-side comparison video. |
53
  | `examples/light_snow_example.mp4` | Test window from the DSERT-RoLL validation set (Light Snow, night): 5 observed + 20 real future frames. |
54
  | `examples/light_snow_example.json` | The exact caption used for that window. |
55
  | `config/experiment.json` | Model options used in training. Needed to rebuild the model. |
 
49
  |---|---|
50
  | `tricosmos_ema_bf16.pt` | Model weights (EMA, bf16, 3.9 GB): the base model with the trained LoRA weights, training step 500. This is the checkpoint behind all reported results. |
51
  | `tricosmos_infer.py` | Inference script: one forecast from a 2×2 grid video; writes the forecast grid and one video per sensor. Also runs the zero-shot base model for comparison. |
52
+ | `TriCosmos_inference.ipynb` | Google Colab notebook: install, download, forecast, and a side-by-side comparison video. |
53
  | `examples/light_snow_example.mp4` | Test window from the DSERT-RoLL validation set (Light Snow, night): 5 observed + 20 real future frames. |
54
  | `examples/light_snow_example.json` | The exact caption used for that window. |
55
  | `config/experiment.json` | Model options used in training. Needed to rebuild the model. |