fix: set data processing timeout to 1h

Author: Jensen246

rdagent/app/finetune/llm/conf.py

@@ -39,8 +39,10 @@ class LLMFinetunePropSetting(ExtendedBaseSettings):
     # Timeouts (longer for LLM training, all for Docker container timeout)
     full_timeout: int = 360000
     """Full training timeout in seconds (default 100 hours, env: FT_FULL_TIMEOUT). Used in running stage for complete model training."""
-    data_processing_timeout: int = 14400
-    """Data processing script timeout in seconds (default 4 hours, env: FT_DATA_PROCESSING_TIMEOUT)."""
+    data_processing_timeout: int = 3600
+    """Data processing script timeout in seconds (default 1 hour, env: FT_DATA_PROCESSING_TIMEOUT). Used for full data processing in running stage."""
+    debug_data_processing_timeout: int = 1200
+    """Debug data processing timeout in seconds (default 20 minutes, env: FT_DEBUG_DATA_PROCESSING_TIMEOUT). Used for --debug mode in coding stage."""
     micro_batch_timeout: int = 1800
     """Micro-batch test timeout in seconds (default 30 minutes, env: FT_MICRO_BATCH_TIMEOUT)."""
 

rdagent/components/coder/finetune/conf.py

@@ -197,6 +197,7 @@ def get_ft_env(
     # Select timeout based on operation type
     timeout_map = {
         "data_processing": FT_RD_SETTING.data_processing_timeout,
+        "debug_data_processing": FT_RD_SETTING.debug_data_processing_timeout,
         "micro_batch": FT_RD_SETTING.micro_batch_timeout,
         "full_training": FT_RD_SETTING.full_timeout,
     }
@@ -229,6 +230,7 @@ def get_ft_env(
 
 def get_data_processing_env(
     enable_cache: bool | None = None,
+    is_debug: bool = False,
 ) -> tuple[Env, dict]:
     """Get environment for data processing scripts with LLM API access.
 
@@ -239,13 +241,14 @@ def get_data_processing_env(
 
     Args:
         enable_cache: Whether to enable Docker caching
+        is_debug: Whether running in debug mode (shorter timeout, default 20 min vs 1 hour)
 
     Returns:
         Tuple of (env, env_vars) where env_vars contains LLM API keys
         to be passed to env.run() as the env parameter
     """
     env = get_ft_env(
-        operation="data_processing",
+        operation="debug_data_processing" if is_debug else "data_processing",
         enable_cache=enable_cache,
     )
 

rdagent/components/coder/finetune/eval.py

@@ -77,7 +77,7 @@ def evaluate(
         # So we clear the workspace every time.
 
         # Step 3: Execute script in DEBUG mode (generates ~10 samples for fast validation)
-        env, env_vars = get_data_processing_env()
+        env, env_vars = get_data_processing_env(is_debug=True)
 
         # Clear workspace (except logs and file_dict items) before data processing
         clear_workspace(implementation, env=env)

rdagent/scenarios/finetune/scen/prompts.yaml

@@ -69,6 +69,7 @@ scenario_description: |-
 
   ## Timeout Constraints
   - Full Training Timeout: {{ full_timeout }}
+  - Data Processing Timeout: {{ data_processing_timeout }}
   {% endif %}
 
   ## (Very important!)Sample Size Control (Code-Based, No LLM)

rdagent/scenarios/finetune/scen/scenario.py

@@ -298,6 +298,7 @@ def get_scenario_all_desc(self, enable_dataset_description: bool = False) -> str
             dataset_config=self.dataset_config,
             model_info=self.model_info,
             full_timeout=f"{self.real_full_timeout() / 60 / 60:.2f} hours",
+            data_processing_timeout=f"{FT_RD_SETTING.data_processing_timeout / 60:.0f} minutes",
             enable_dataset_description=enable_dataset_description,
             upper_data_size_limit=FT_RD_SETTING.upper_data_size_limit,
         )