Upload folder using huggingface_hub

README.md

@@ -7,9 +7,9 @@ tags:
 - pytorch
 ---
 
-# Octo-Base PyTorch Model
+# Octo-Small PyTorch Model
 
-This is the octo-base model converted to PyTorch format.
+This is the octo-small model converted to PyTorch format.
 
 ## Model Description
 
@@ -19,7 +19,7 @@ Octo is a generalist robot policy trained on diverse robot manipulation tasks.
 - **Original JAX Implementation**: [octo-models/octo](https://github.com/octo-models/octo)
 - **Original Pytorch Implementation**: [emb-ai/octo-pytorch](https://github.com/emb-ai/octo-pytorch)
 - **lil'km Implementation**: [s1lent4gnt/octo-pytorch](https://github.com/s1lent4gnt/octo-pytorch)
-- **Model Size**: octo-base
+- **Model Size**: octo-small
 
 ## Usage
 
@@ -60,7 +60,7 @@ model = OctoModel.from_pretrained('lilkm/octo-small-test')
 
 ### Model Architecture
 
-- **Transformer**: 12 layers, 768 dim, 12 heads
+- **Transformer**: 12 layers, 384 dim, 6 heads
 - **Vision Encoder**: Custom CNN (SmallStem16)
 - **Language Encoder**: T5-Base
 - **Action Head**: Diffusion policy with 4 action steps

config.json

@@ -1,10 +1,10 @@
 {
   "model_type": "octo",
-  "model_name": "octo-base",
-  "token_embedding_size": 768,
+  "model_name": "octo-small",
+  "token_embedding_size": 384,
   "num_layers": 12,
-  "num_heads": 12,
-  "mlp_dim": 3072,
+  "num_heads": 6,
+  "mlp_dim": 1536,
   "max_horizon": 10,
   "repeat_task_tokens": true,
   "action_horizon": 4,

dataset_statistics.npy

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc9b6afd6157706a2e8cc489c0c555fac777b1c578567545520e7d72b9314d9d
+size 17516

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f8c8e2f7c7b97b889d413006801ea351cb2213060a69b830ed7b2f83abbceaeb
-size 371352224
+oid sha256:9104dff53dff8c63750e7d373ce35c4efbf10d263af2c74215f2ad543371701b
+size 108199904