issue-bot/configuration/base.yaml

auth_token: tmpsecret

database:
  connection_string: postrgres://local:supersecurepassword@localhost:5432/lor_e
  max_connections: 5

embedding_api:
  auth_token: ""
  url: ""

github_api:
  auth_token: ""
  comments_enabled: false

huggingface_api:
  auth_token: ""
  comments_enabled: false

message_config:
  pre: "Hello!\n\nA maintainer will soon take a look, in the meantime you might find these related issues interesting:\n"
  post: "\n\nThank you for opening this issue!"

model:
  id: NovaSearch/stella_en_1.5B_v5
  revision: main
  embeddings_size: 1024
  max_input_size: 131072

server:
  ip: 0.0.0.0
  metrics_port: 4243
  port: 4242

slack:
  auth_token: ""
  channel: ""
  chat_write_url: https://slack.com/api/chat.postMessage

summarization_api:
  auth_token: ""
  model: Qwen/Qwen2.5-Coder-32B-Instruct
  system_prompt: |
    You are Qwen, created by Alibaba Cloud. You are a helpful assistant. Your task is to create user-friendly descriptions of huggingface's transformers individual issues or pull requests and its comments, so that everyone can easily understand what the core of the problem is. Follow these steps:

    Extract key information from the issue/pr, focusing on
      - What is the core of the problem faced or being fixed?
      - What is the model that is being used?
      - Which part of the library is impacted?
      - What relevant error messages were provided?
      - Is it a bug, a feature request or a need for clarification?

    Write a clear and practical description of what the application does:
      - Short description (under 100 characters):
        - Single sentence that captures the core problem reported or being fixed
        - Must be less than 100 characters

    Create a list of three to five (no more) categories/tags that describes the issue, such as:
      - Which model is mentioned in the issue
      - Which part of the transformers library is mentioned by the issue (e.g. "trainer", "inference", "vision", "audio", etc)
      - On which infrastructure component, cloud or device is the issue faced (e.g. "gpu", "AWS", "nvidia", "T4", "L4", etc)
      - is it a bug, feature request or anything of the like

    Provide your output in the following format:
    *Tags: <TAGS>first-category, second-category, third-category</TAGS>*
    > <DESC>Your short description (under 100 characters)</DESC>
  special_tokens_used:
    - DESC
    - TAGS
  url: https://router.huggingface.co/hf-inference/models/Qwen/Qwen2.5-Coder-32B-Instruct