cheesewafer/Llama3-8B-Instruct-sft-alfworld

TEXT GENERATIONPricing:Input $0.2 / Cached $0.028 / Output $0.32Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jun 3, 2025Architecture:Transformer Featherless Exclusive Cold

cheesewafer/Llama3-8B-Instruct-sft-alfworld is an 8 billion parameter language model fine-tuned from Meta-Llama-3.1-8B-Instruct. This model is specifically optimized for tasks within the ALFWorld environment, a text-based interactive simulation for embodied agents. It leverages a 32768 token context length to process complex environmental interactions and instructions. Its primary differentiation lies in its specialized training for agentic behavior and problem-solving within the ALFWorld framework.

Loading preview...

Model Overview

This model, cheesewafer/Llama3-8B-Instruct-sft-alfworld, is a specialized fine-tuned version of the meta-llama/Meta-Llama-3.1-8B-Instruct base model. It features 8 billion parameters and is designed for interaction within the ALFWorld environment, a benchmark for embodied AI agents.

Key Characteristics

  • Base Model: Built upon Meta's Llama 3.1 8B Instruct, inheriting its strong foundational language understanding capabilities.
  • Specialized Fine-tuning: Specifically trained for the ALFWorld environment, indicating an optimization for understanding and executing commands in interactive text-based worlds.
  • Training Parameters: The fine-tuning process utilized a learning rate of 2e-05, a total batch size of 64 across 8 GPUs, and ran for 3 epochs with a cosine learning rate scheduler.

Intended Use Cases

This model is particularly suited for research and development in:

  • Embodied AI: Developing and testing agents that can navigate and interact with virtual environments through natural language.
  • Text-based Game AI: Creating intelligent agents for complex text adventures or interactive fiction.
  • Instruction Following: Tasks requiring precise interpretation and execution of multi-step instructions within a constrained environment.

Due to its specialized training, its performance on general-purpose language tasks may not be its primary strength compared to its base model, but it is expected to excel in ALFWorld-specific challenges.