metadata

library_name: peft
base_model: unsloth/gemma-1.1-7b-it-bnb-4bit
datasets:
  - naklecha/minecraft-question-answer-700k
language:
  - en
pipeline_tag: question-answering

Gemma 1.1 7B Minecraft QA Model

This fine-tuned Gemma 7B Instruct model from Unsloth using data from naklecha/minecraft-question-answer-700k to enhance its understanding and generation of Minecraft-related content. This specialized model is now better equipped to provide detailed information and support for Minecraft players and developers.

Model Info

Fine-tuned with NVIDIA A100 with 100 steps, took 5 minutes to train using TRL, QLoRA and quantized.

Note that since the model is quantized, you have to use it with a GPU.