zaddyzaddy/hushGPT

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:8kPublished:Jun 14, 2024License:apache-2.0Architecture:Transformer Open Weights Cold

hushGPT is an 8 billion parameter Llama-3 based language model developed by zaddyzaddy. It was fine-tuned using Unsloth and Huggingface's TRL library, enabling 2x faster training. This model is optimized for efficient performance within its parameter class, leveraging advanced training techniques.

Loading preview...

hushGPT: An Efficient Llama-3 Fine-tune

hushGPT is an 8 billion parameter language model developed by zaddyzaddy. It is fine-tuned from the unsloth/llama-3-8b-bnb-4bit base model, leveraging the Unsloth library in conjunction with Huggingface's TRL library. This combination allowed for a significantly accelerated training process, reportedly achieving 2x faster fine-tuning.

Key Characteristics

  • Base Model: Llama-3 8B
  • Developer: zaddyzaddy
  • Training Efficiency: Utilizes Unsloth for 2x faster fine-tuning.
  • License: Apache-2.0

Potential Use Cases

Given its Llama-3 foundation and efficient training, hushGPT is suitable for applications requiring a capable 8B parameter model where rapid deployment and fine-tuning are beneficial. Its Apache-2.0 license provides flexibility for various commercial and research applications.