kolerk/tcod_7b_f2b

TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kPublished:Apr 28, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The kolerk/tcod_7b_f2b is a 7.6 billion parameter language model specifically designed and fine-tuned for tasks within the ALFWorld environment. This model leverages its substantial parameter count and a 32768-token context length to process complex interactive scenarios. Its primary differentiation lies in its specialized optimization for ALFWorld, making it highly effective for embodied AI and interactive decision-making tasks within that framework. Developers should consider this model for applications requiring robust performance in ALFWorld-specific challenges.

Loading preview...

kolerk/tcod_7b_f2b: Specialized for ALFWorld

The kolerk/tcod_7b_f2b is a 7.6 billion parameter language model with a 32768-token context length, uniquely developed and optimized for the ALFWorld environment. ALFWorld is a simulated interactive environment designed for embodied AI research, where agents must perform tasks by interacting with objects and navigating virtual spaces.

Key Capabilities

  • ALFWorld Task Performance: This model is specifically fine-tuned to excel in the complex, interactive tasks presented within the ALFWorld framework.
  • Embodied AI Focus: Its design targets scenarios requiring an understanding of physical environments and sequential decision-making.
  • Large Context Window: The 32768-token context length allows for processing extensive interaction histories and detailed environmental descriptions, crucial for navigating and completing tasks in ALFWorld.

Good For

  • Researchers and developers working on embodied AI agents within the ALFWorld simulation.
  • Applications requiring a language model with strong performance in interactive decision-making and planning in simulated environments.
  • Exploring language-guided task completion in virtual worlds.