dvijay/llama-2-7b-miniguanaco

TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kArchitecture:Transformer Cold

dvijay/llama-2-7b-miniguanaco is a 7 billion parameter language model fine-tuned from Llama-2-7b-chat-hf. This model was fine-tuned using the Guanaco 1k dataset, primarily for learning purposes. It is suitable for conversational AI tasks and serves as a demonstration of fine-tuning large language models on consumer-grade hardware.

Loading preview...

dvijay/llama-2-7b-miniguanaco Overview

dvijay/llama-2-7b-miniguanaco is a 7 billion parameter language model based on the Llama-2-7b-chat-hf architecture. This model was fine-tuned using the Guanaco 1k dataset, a common dataset for instruction-tuning LLMs. The fine-tuning process was conducted locally on an RTX 4080 GPU, taking approximately 10 hours.

Key Characteristics

  • Base Model: Llama-2-7b-chat-hf
  • Fine-tuning Dataset: Guanaco 1k
  • Parameter Count: 7 billion parameters
  • Training Environment: Local fine-tuning on an RTX 4080 GPU
  • Training Duration: Approximately 10 hours

Primary Use Case

This model is primarily intended for:

  • Learning and Experimentation: Demonstrating the process and feasibility of fine-tuning large language models on accessible hardware.
  • Conversational AI: Suitable for basic chat applications and generating human-like text responses, given its instruction-tuned base.

It serves as a practical example for developers and researchers interested in understanding the fine-tuning workflow for Llama-2 models.