penfever/nl2bash-verified-GLM-4_6-traces-32ep-32k-dft

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Nov 22, 2025License:apache-2.0Architecture:Transformer Open Weights Cold

This model, developed by penfever, is an 8 billion parameter language model fine-tuned from Qwen/Qwen3-8B. It is specifically trained on the nl2bash-verified-GLM-4.6-traces-32ep-32k dataset, indicating a specialization in natural language to bash command translation. With a context length of 32768 tokens, it is optimized for generating verified bash commands from natural language inputs.

Loading preview...

Model Overview

This model, nl2bash-verified-GLM-4_6-traces-32ep-32k-dft, is a specialized 8 billion parameter language model. It is a fine-tuned variant of the Qwen/Qwen3-8B architecture, developed by penfever. The model's primary differentiation lies in its training on the penfever/nl2bash-verified-GLM-4.6-traces-32ep-32k dataset, which suggests a strong focus on natural language to bash command translation with an emphasis on verified outputs.

Key Training Details

The model underwent training with the following notable hyperparameters:

  • Base Model: Qwen/Qwen3-8B
  • Learning Rate: 4e-05
  • Optimizer: ADAMW_TORCH_FUSED
  • Epochs: 7.0
  • Distributed Training: Utilized 16 GPUs with a total train batch size of 16.

Intended Use Cases

Given its specific fine-tuning, this model is particularly suited for applications requiring the conversion of natural language instructions into executable bash commands. Its training on a 'verified' dataset implies a focus on accuracy and reliability in the generated commands. Developers looking to automate command-line tasks or build interfaces that translate user queries into shell scripts could find this model highly relevant.