config.yml

models:
  - type: main
    engine: hf_pipeline_llama2_13b
    parameters:
      path: "meta-llama/Llama-2-13b-chat-hf"

      # number of GPUs you have , do nvidia-smi to check
      num_gpus: 2

      # This can be: "cpu" or "cuda". "mps" is not supported.
      device: "cuda"

rails:
  output:
    flows:
      - self check facts

instructions:
  - type: general
    content: |
      Below is a conversation between a bot and a user about the recent job reports.
      The bot is factual and concise. If the bot does not know the answer to a
      question, it truthfully says it does not know.

sample_conversation: |
  user "Hello there!"
    express greeting
  bot express greeting
    "Hello! How can I assist you today?"
  user "What can you do for me?"
    ask about capabilities
  bot respond about capabilities
    "I am an AI assistant which helps answer questions based on a given knowledge base."

# The prompts below are the same as the ones from `nemoguardrails/llm/prompts/dolly.yml`.
prompts:
  - task: general
    models:
      - hf_pipeline_llama2_13b
    content: |-
      {{ general_instructions }}

      {{ history | user_assistant_sequence }}
      Assistant:

  # Prompt for detecting the user message canonical form.
  - task: generate_user_intent
    models:
      - hf_pipeline_llama2_13b
    content: |-
      """
      {{ general_instructions }}
      """

      # This is how a conversation between a user and the bot can go:
      {{ sample_conversation | verbose_v1 }}

      # This is how the user talks:
      {{ examples | verbose_v1 }}

      # This is the current conversation between the user and the bot:
      {{ sample_conversation | first_turns(2) | verbose_v1 }}
      {{ history | colang | verbose_v1 }}

    output_parser: "verbose_v1"

  # Prompt for generating the next steps.
  - task: generate_next_steps
    models:
      - hf_pipeline_llama2_13b
    content: |-
      """
      {{ general_instructions }}
      """

      # This is how a conversation between a user and the bot can go:
      {{ sample_conversation | remove_text_messages | verbose_v1 }}

      # This is how the bot thinks:
      {{ examples | remove_text_messages | verbose_v1 }}

      # This is the current conversation between the user and the bot:
      {{ sample_conversation | first_turns(2) | remove_text_messages | verbose_v1 }}
      {{ history | colang | remove_text_messages | verbose_v1 }}

    output_parser: "verbose_v1"

  # Prompt for generating the bot message from a canonical form.
  - task: generate_bot_message
    models:
      - hf_pipeline_llama2_13b
    content: |-
      """
      {{ general_instructions }}
      """

      # This is how a conversation between a user and the bot can go:
      {{ sample_conversation | verbose_v1 }}

      {% if relevant_chunks %}
      # This is some additional context:
      ```markdown
      {{ relevant_chunks }}
      ```
      {% endif %}

      # This is how the bot talks:
      {{ examples | verbose_v1 }}

      # This is the current conversation between the user and the bot:
      {{ sample_conversation | first_turns(2) | verbose_v1 }}
      {{ history | colang | verbose_v1 }}

    output_parser: "verbose_v1"

  # Prompt for generating the value of a context variable.
  - task: generate_value
    models:
      - hf_pipeline_llama2_13b
    content: |-
      """
      {{ general_instructions }}
      """

      # This is how a conversation between a user and the bot can go:
      {{ sample_conversation | verbose_v1 }}

      # This is how the bot thinks:
      {{ examples | verbose_v1 }}

      # This is the current conversation between the user and the bot:
      {{ sample_conversation | first_turns(2) | verbose_v1 }}
      {{ history | colang | verbose_v1 }}
      # {{ instructions }}
      ${{ var_name }} =
    output_parser: "verbose_v1"

  - task: self_check_facts
    models:
      - hf_pipeline_llama2_13b
    content: |-
      <<SYS>>
      You are given a task to identify if the hypothesis is grounded and entailed to the evidence.
      You will only use the contents of the evidence and not rely on external knowledge.
      <</SYS>>

      [INST]Answer with yes/no. "evidence": {{ evidence }} "hypothesis": {{ response }} "entails":[/INST]