Upload folder using huggingface_hub

README.md

@@ -1,41 +1,3 @@
----
-license: mit
-datasets:
-- clip-rt/modified_libero_hdf5
-language:
-- en
-tags:
-- robotics
-- vla
-- clip
-- contrastive_learning
----
-
-# CLIP-RT Finetuned on LIBERO-Goal
-
-This model was produced by fine-tuning the [CLIP-RT model](https://clip-rt.github.io/) with a 0.3B parameter action decoder added to enable continuous action prediction on the LIBERO-Goal dataset from the [LIBERO simulation benchmark](https://libero-project.github.io/main.html). 
-
-## Hyperparemeters
-
-| Category             | Details                                                             |
-|----------------------|---------------------------------------------------------------------|
-| **Hardware**         | 8 × H100 GPUs with 80GB memory                                      |
-| **Model size**       | 1.3B (CLIP-RT base + 0.3B action decoder)                           |
-| **Action dimension** | 7D per step × 8 steps (chunked)                                     |
-| **Loss**             | L1 regression                                                       |
-| **Batch size**       | 256                                                                 |
-| **Epochs**           | 128                                                                 |             
-
-## Usage Instructions
-To evaluate this model on the LIBERO simulator or in your own imitation learning pipeline, use the action decoder module with precomputed CLIP image and language embeddings. Refer to the original CLIP-RT GitHub repository for code and inference scripts.
-
-## Citation
-
-```bibtex
-@article{kang2024cliprt,
-  title={CLIP-RT: Learning Language-Conditioned Robotic Policies from Natural Language Supervision},
-  author={Kang, Gi-Cheon and Kim, Junghyun and Shim, Kyuhwan and Lee, Jun Ki and Zhang, Byoung-Tak},
-  journal={arXiv preprint arXiv:2411.00508},
-  year = {2024}
-}
-```
+---
+license: mit
+---

cliprt_libero_goal.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0b142300bf8d12afad7a1d6e610de0aa780c72e82e2545ef5f82c587e0cc7bd
+size 16203732598