removed # pylint: disable=useless-parent-delegation in FinetuneDialog…

…Agent

examples/conversation_with_agent_with_finetuned_model/FinetuneDialogAgent.py

@@ -43,14 +43,14 @@ def __init__(
  Note:
  Refer to `class DialogAgent(AgentBase)` for more information.
  """
- # pylint: disable=useless-parent-delegation
  super().__init__(
  name,
  sys_prompt,
  model_config_name,
  use_memory,
  memory_config,
  )
+ self.finetune = True
 
  def load_model(
  self,

examples/conversation_with_agent_with_finetuned_model/conversation_with_agent_with_finetuned_model.py

@@ -9,7 +9,10 @@
 Features include model and tokenizer loading,
 and fine-tuning on the lima dataset with adjustable parameters.
 """
-
+# This import is necessary for AgentScope to properly use
+# HuggingFaceWrapper even though it's not explicitly used in this file.
+# To remove the pylint disable without causing issues
+# HuggingFaceWrapper needs to be put under src/agentscope/agents.
 # pylint: disable=unused-import
 from huggingface_model import HuggingFaceWrapper
 from FinetuneDialogAgent import FinetuneDialogAgent
@@ -52,6 +55,7 @@ def main() -> None:
  # loading a model saved as lora model
  "fine_tune_config": {
  "continue_lora_finetuning": False,
+ "max_seq_length": 4096,
  "lora_config": {
  "r": 16,
  "lora_alpha": 32,

examples/conversation_with_agent_with_finetuned_model/huggingface_model.py

@@ -80,10 +80,13 @@ def __init__(
  fine-tuning the model.
  **kwargs: Additional keyword arguments.
  """
- super().__init__(config_name=config_name)
+ super().__init__(
+ config_name=config_name,
+ model_name=pretrained_model_name_or_path,
+ )
  self.model = None
  self.tokenizer = None
- self.max_length = max_length # Set max_length as an attribute
+ self.max_length = max_length
  self.pretrained_model_name_or_path = pretrained_model_name_or_path
  self.local_model_path = local_model_path
  self.lora_config = None
@@ -358,6 +361,8 @@ def load_tokenizer(
  f" from '{local_model_path}'",
  )
  self.tokenizer.padding_side = "right"
+ if self.tokenizer.pad_token is None:
+ self.tokenizer.pad_token = self.tokenizer.eos_token
 
  except Exception as e:
  # Handle exceptions during model loading,
@@ -666,10 +671,13 @@ def fine_tune_training(
  "optim": "paged_adamw_8bit",
  "logging_steps": 1,
  }
+ max_seq_length_default = 4096
 
  lora_config_default = {}
 
  if fine_tune_config is not None:
+ if fine_tune_config.get("max_seq_length") is not None:
+ max_seq_length_default = fine_tune_config["max_seq_length"]
  if fine_tune_config.get("training_args") is not None:
  training_defaults.update(fine_tune_config["training_args"])
  if fine_tune_config.get("lora_config") is not None:
@@ -709,7 +717,7 @@ def fine_tune_training(
  else {}
  ),
  args=trainer_args,
- max_seq_length=4096,
+ max_seq_length=max_seq_length_default,
  )
 
  logger.info(
@@ -718,14 +726,7 @@ def fine_tune_training(
  timestamp=datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
  )
 
- try:
- trainer.train()
- except Exception as e:
- import traceback
-
- logger.error(f"Error during training: {e}")
- traceback.print_exc()
- raise
+ trainer.train()
 
  now = datetime.now()
  time_string = now.strftime("%Y-%m-%d_%H-%M-%S")