llama_3_3_70b

What it is

• A module that defines a preconfigured ChatModel wrapper for Meta’s Llama-3.3-70B-Instruct using langchain_ollama.ChatOllama.

Public API

• Constants

  • ID: Model identifier ("meta-llama/Llama-3.3-70B-Instruct").
  • NAME: Ollama model name ("llama-3.3-70b-instruct").
  • DESCRIPTION: Human-readable model description.
  • IMAGE: Image URL for the model.
  • CONTEXT_WINDOW: Context window size (131072).
  • PROVIDER: Provider label ("meta").
  • TEMPERATURE: Default temperature (0).
  • MAX_TOKENS: Defined but not used in configuration (4096).
  • MAX_RETRIES: Defined but not used in configuration (2).

• Objects

  • model: ChatModel: A naas_abi_core.models.Model.ChatModel instance configured with:
    • model_id, name, description, image, provider
    • model=ChatOllama(model=NAME, temperature=TEMPERATURE)
    • context_window=CONTEXT_WINDOW

Configuration/Dependencies

• Dependencies
  • langchain_ollama.ChatOllama
  • naas_abi_core.models.Model.ChatModel
• Runtime expectation
  • The configured ChatOllama backend must be available and have the model named llama-3.3-70b-instruct accessible.

Usage

from naas_abi_marketplace.ai.llama.models.llama_3_3_70b import model

# `model` is a ChatModel wrapper around a ChatOllama instance
print(model.name)
print(model.model_id)

Caveats

• MAX_TOKENS and MAX_RETRIES are defined in the module but are not applied to the ChatOllama or ChatModel configuration here.
• The exact invocation interface (e.g., .invoke(...)) depends on ChatModel and ChatOllama implementations, which are external to this module.