cleanup

src/lerobot/policies/rlearn/modeling_rlearn.py

@@ -184,13 +184,22 @@ class RLearNPolicy(PreTrainedPolicy):
         # Layer normalization before reward head to stabilize MLP outputs
         self.pre_reward_norm = nn.LayerNorm(config.dim_model)
         
-        # Regression head (logit mode only)
+        # Temporal-aware regression head (logit mode only)
-        self.reward_head = nn.Linear(config.dim_model, 1)
+        # Concatenates frame embedding with normalized temporal position
+        self.reward_head = nn.Sequential(
+            nn.Linear(config.dim_model + 1, config.dim_model),  # +1 for temporal position
+            nn.ReLU(),
+            nn.Linear(config.dim_model, 1)
+        )
         
-        # Initialize head for logit regression
+        # Initialize temporal-aware head for logit regression
         with torch.no_grad():
-            self.reward_head.weight.normal_(0.0, config.head_weight_init_std)
+            # First layer: embedding + position -> embedding
-            self.reward_head.bias.fill_(0.0)
+            nn.init.normal_(self.reward_head[0].weight, 0.0, config.head_weight_init_std)
+            nn.init.zeros_(self.reward_head[0].bias)
+            # Output layer: embedding -> logit
+            nn.init.normal_(self.reward_head[2].weight, 0.0, config.head_weight_init_std) 
+            nn.init.zeros_(self.reward_head[2].bias)
         
         # Simple frame dropout probability
         self.frame_dropout_p = config.frame_dropout_p
@@ -309,9 +318,19 @@ class RLearNPolicy(PreTrainedPolicy):
         # MLP predictor
         video_frame_embeds = self.mlp_predictor(frame_tokens)
         
-        # Get rewards via logit regression head
+        # Get rewards via temporal-aware logit regression head
         normalized_embeds = self.pre_reward_norm(video_frame_embeds)
-        raw_logits = self.reward_head(normalized_embeds).squeeze(-1)  # (B, T)
+        
+        # Add temporal position information
+        B, T_pred = normalized_embeds.shape[:2]
+        temporal_pos = torch.linspace(0, 1, T_pred, device=normalized_embeds.device)
+        temporal_pos = temporal_pos.unsqueeze(0).unsqueeze(-1).expand(B, T_pred, 1)  # (B, T, 1)
+        
+        # Concatenate embeddings with temporal position
+        temporal_input = torch.cat([normalized_embeds, temporal_pos], dim=-1)  # (B, T, D+1)
+        
+        # Forward through temporal-aware head
+        raw_logits = self.reward_head(temporal_input).squeeze(-1)  # (B, T)
         return torch.sigmoid(raw_logits)  # Apply sigmoid for final predictions
 
     def normalize_inputs(self, batch: dict[str, Tensor]) -> dict[str, Tensor]:
@@ -480,17 +499,35 @@ class RLearNPolicy(PreTrainedPolicy):
 
         # During inference, we might not want to compute loss
         if not self.training and target is None:
-            # Return predictions without loss
+            # Return predictions without loss using temporal-aware head
             normalized_embeds = self.pre_reward_norm(video_frame_embeds)
-            rewards = self.sigmoid(self.reward_head(normalized_embeds)).squeeze(-1)
+            
+            # Add temporal position information
+            B_inf, T_inf = normalized_embeds.shape[:2]
+            temporal_pos = torch.linspace(0, 1, T_inf, device=normalized_embeds.device)
+            temporal_pos = temporal_pos.unsqueeze(0).unsqueeze(-1).expand(B_inf, T_inf, 1)
+            
+            # Concatenate and forward through temporal-aware head
+            temporal_input = torch.cat([normalized_embeds, temporal_pos], dim=-1)
+            raw_logits = self.reward_head(temporal_input).squeeze(-1)
+            rewards = torch.sigmoid(raw_logits)
             return rewards.mean() * 0.0, {"rewards_mean": rewards.mean().item()}
 
         # Calculate loss using logit regression
         loss_start = time.perf_counter()
         
-        # Get model outputs
+        # Get model outputs with temporal-aware head
         normalized_embeds = self.pre_reward_norm(video_frame_embeds)
-        raw_logits = self.reward_head(normalized_embeds).squeeze(-1)  # (B, T_eff)
+        
+        # Add temporal position information
+        temporal_pos = torch.linspace(0, 1, T_eff, device=normalized_embeds.device)
+        temporal_pos = temporal_pos.unsqueeze(0).unsqueeze(-1).expand(B, T_eff, 1)  # (B, T_eff, 1)
+        
+        # Concatenate embeddings with temporal position
+        temporal_input = torch.cat([normalized_embeds, temporal_pos], dim=-1)  # (B, T_eff, D+1)
+        
+        # Forward through temporal-aware head
+        raw_logits = self.reward_head(temporal_input).squeeze(-1)  # (B, T_eff)
         
         # Logit regression: transform targets to logit space and compute MSE on logits
         eps = self.config.logit_eps
@@ -537,9 +574,19 @@ class RLearNPolicy(PreTrainedPolicy):
                 mismatch_tokens = self.frame_mlp(attended_video_mm)  # (B, T, D)
                 mismatch_embeds = self.mlp_predictor(mismatch_tokens)
                 
-                # Predict near-zero progress for mismatched pairs
+                # Predict near-zero progress for mismatched pairs with temporal awareness
                 normalized_mismatch_embeds = self.pre_reward_norm(mismatch_embeds)
-                mismatch_raw_logits = self.reward_head(normalized_mismatch_embeds).squeeze(-1)
+                
+                # Add temporal position information for mismatch computation
+                T_mismatch = normalized_mismatch_embeds.shape[1]
+                temporal_pos_mm = torch.linspace(0, 1, T_mismatch, device=normalized_mismatch_embeds.device)
+                temporal_pos_mm = temporal_pos_mm.unsqueeze(0).unsqueeze(-1).expand(B, T_mismatch, 1)
+                
+                # Concatenate mismatch embeddings with temporal position
+                temporal_input_mm = torch.cat([normalized_mismatch_embeds, temporal_pos_mm], dim=-1)
+                
+                # Forward through temporal-aware head
+                mismatch_raw_logits = self.reward_head(temporal_input_mm).squeeze(-1)
                 
                 # Create mask tensor for loss calculation
                 mismatch_tensor = torch.tensor(mismatch_mask, device=device, dtype=torch.bool)