[Port HIL-SERL] Add resnet-10 as default encoder for HIL-SERL (#696)

Co-authored-by: Khalil Meftah <kmeftah.khalil@gmail.com>
Co-authored-by: Adil Zouitine <adilzouitinegm@gmail.com>
Co-authored-by: Michel Aractingi <michel.aractingi@huggingface.co>
Co-authored-by: Ke Wang <superwk1017@gmail.com>

lerobot/common/policies/hilserl/classifier/configuration_classifier.py

@@ -10,7 +10,7 @@ class ClassifierConfig:
     num_classes: int = 2
     hidden_dim: int = 256
     dropout_rate: float = 0.1
-    model_name: str = "microsoft/resnet-50"
+    model_name: str = "helper2424/resnet10"
     device: str = "cpu"
     model_type: str = "cnn"  # "transformer" or "cnn"
     num_cameras: int = 2

lerobot/common/policies/sac/configuration_sac.py

@@ -61,7 +61,7 @@ class SACConfig:
     )
     camera_number: int = 1
     # Add type annotations for these fields:
-    vision_encoder_name: str | None = field(default="microsoft/resnet-18")
+    vision_encoder_name: str | None = field(default="helper2424/resnet10")
     freeze_vision_encoder: bool = True
     image_encoder_hidden_dim: int = 32
     shared_encoder: bool = True

lerobot/configs/policy/hilserl_classifier.yaml

@@ -36,7 +36,7 @@ eval:
 
 policy:
   name: "hilserl/classifier/push_green_cube_hf_cropped_resized" #"hilserl/classifier/pick_place_lego_cube_1"
-  model_name: "facebook/convnext-base-224"
+  model_name: "helper2424/resnet10"
   model_type: "cnn"
   num_cameras: 2 # Has to be len(training.image_keys)