Improved_output exception occured when training the crew at the last iteration

CrewAI Community Support
,
1

raise Exception(f"An error occurred while training the crew: {e}") Exception: An error occurred while training the crew: ‘improved_output’ An error occurred while training the crew: Command ‘[‘uv’, ‘run’, ‘train’, ‘3’, ‘trained_agents_data.pkl’]’ returned non-zero exit status 1.

2

Thats good but please replace the urls also with complete urls Human feedback: Thats good but please replace the urls also with complete urls 2025-01-05 18:02:52,906 - 856162480 - llm.py-llm:170 - ERROR: LiteLLM call failed: litellm.BadRequestError: Invalid Message bedrock requires at least one non-system message Error during LLM call to classify human feedback: litellm.BadRequestError: Invalid Message bedrock requires at least one non-system message. Retrying... (1/3) 2025-01-05 18:02:52,917 - 856162480 - llm.py-llm:170 - ERROR: LiteLLM call failed: litellm.BadRequestError: Invalid Message bedrock requires at least one non-system message Error during LLM call to classify human feedback: litellm.BadRequestError: Invalid Message bedrock requires at least one non-system message. Retrying... (2/3) 2025-01-05 18:02:52,930 - 856692480 - llm.py-llm:170 - ERROR: LiteLLM call failed: litellm.BadRequestError: Invalid Message bedrock requires at least one non-system message Error during LLM call to classify human feedback: litellm.BadRequestError: Invalid Message bedrock requires at least one non-system message. Retrying... (3/3) Error processing feedback after multiple attempts.

I was getting above error whenever i was trying to give the feedback. May this is why its failing. Could someone help me out here ?

3

I’m facing the same error.

Seems that trained_agents_data.pkl file data is corrupted, or missing a property from a dictionary. I’ll try to use pickle to open and read the files and check if dictionaries are consistent.

I found the following example at crewAI chatGPT:

    import pickle

# Replace 'trained_agents_data.pkl' with your file path
file_path = 'trained_agents_data.pkl'

try:
    with open(file_path, 'rb') as file:
        data = pickle.load(file)
        # Print the loaded data
        print("File Content:")
        print(data)

        # If the file contains a list of items, iterate through them
        if isinstance(data, list):
            for idx, item in enumerate(data):
                print(f"\nItem {idx + 1}:")
                print(item)
        elif isinstance(data, dict):
            print("\nKeys in the dictionary:")
            for key in data.keys():
                print(f"- {key}: {data[key]}")
except Exception as e:
    print(f"Error reading the file: {e}")