RuntimeError: Failed to generate valid JSON output

[micnikolaev]

I got "RuntimeError: Failed to generate valid JSON output" when tried to do python -m graphrag.query --root ./ragtest --method global "What is the release date of the book?"

My config:
settings.txt

Logs:
logs.json

[tingwei161803]

I got this error in the create_final_community_reports process

[brickfrog]

I also got this. Was curious if there was any hardware, etc. that wasn't supported.

[micnikolaev]

on which OS do you run?

[tingwei161803]

I run with Ollama and succeeded with Gemma2-9b, but failed with Llama3-8b. I made the following modifications in setting.yaml:

encoding_model: cl100k_base
skip_workflows: []
llm:
  api_key: ollama
  type: openai_chat # or azure_openai_chat
  model: llama3
  model_supports_json: true # recommended if this is available for your model.
  request_timeout: 100000
  api_base: http://localhost:11434/v1

embeddings:
  async_mode: threaded # or asyncio
  llm:
    api_key: ollama
    type: openai_embedding # or azure_openai_embedding
    model: nomic_embed_text
    api_base: http://localhost:11434/api

[kunjanshah0811]

So with these configuration in LLM and Embeddings it worked, right?

I run with Ollama and succeeded with Gemma2-9b, but failed with Llama3-8b. I made the following modifications in setting.yaml:

encoding_model: cl100k_base
skip_workflows: []
llm:
  api_key: ollama
  type: openai_chat # or azure_openai_chat
  model: llama3
  model_supports_json: true # recommended if this is available for your model.
  request_timeout: 100000
  api_base: http://localhost:11434/v1

embeddings:
  async_mode: threaded # or asyncio
  llm:
    api_key: ollama
    type: openai_embedding # or azure_openai_embedding
    model: nomic_embed_text
    api_base: http://localhost:11434/api