ashercn97/giraffe-7b

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Jul 26, 2023Architecture:Transformer Cold

The ashercn97/giraffe-7b is a 7 billion parameter language model fine-tuned on an Orca dataset and a multilingual code dataset. This model is designed to leverage the instruction-following capabilities of Orca-style training combined with enhanced code generation and understanding. It is particularly suited for tasks requiring both general instruction adherence and multi-language coding proficiency.

Loading preview...

Model Overview

The ashercn97/giraffe-7b is a 7 billion parameter language model developed by ashercn97. This model distinguishes itself through its fine-tuning process, which involved a combination of an Orca dataset and a dedicated multilingual code dataset. This dual-dataset approach aims to imbue the model with strong instruction-following capabilities, characteristic of Orca-trained models, alongside robust performance in coding tasks across multiple programming languages.

Key Capabilities

  • Instruction Following: Benefits from Orca-style fine-tuning, enhancing its ability to understand and execute complex instructions.
  • Multilingual Code Proficiency: Trained on a specific multilingual code dataset, making it suitable for various coding-related applications.
  • Accessible Training: Developed using Axolotl, demonstrating efficient training on consumer-grade hardware (2 RTX 4090 GPUs).

Good For

  • Developers seeking a 7B model with a balance of general instruction adherence and coding capabilities.
  • Applications requiring code generation, code completion, or code explanation in multiple languages.
  • Experimentation with models fine-tuned on diverse datasets for specific task performance.