emre/llama-2-13b-mini
The emre/llama-2-13b-mini is a 13 billion parameter Llama-2-13b-chat-hf model fine-tuned using QLoRA (4-bit precision) with a context length of 4096 tokens. Developed by emre, this model is primarily designed for educational purposes and internal use within BBVA Group, GarantiBBVA, and its subsidiaries. It serves as an example of fine-tuning a Llama-2 base model for specific organizational or learning objectives.
Loading preview...
Model Overview
The emre/llama-2-13b-mini is a 13 billion parameter language model based on the Llama-2-13b-chat-hf architecture. It has been fine-tuned using the QLoRA method with 4-bit precision, making it a more efficient adaptation of the larger Llama 2 model.
Key Characteristics
- Base Model: Llama-2-13b-chat-hf
- Fine-tuning Method: QLoRA (4-bit precision)
- Training Environment: Colab Pro+
- Context Length: The base model supports a context length of 4096 tokens.
- Training Parameters: Utilizes specific QLoRA configurations including
lora_r=8,lora_alpha=16, andlora_dropout=0.05.
Intended Use
This model is primarily developed for educational purposes. While it can be used for inference, its application is specifically restricted to BBVA Group, GarantiBBVA, and their subsidiaries. It demonstrates the process of adapting a large language model for specialized, internal organizational use cases rather than broad public deployment.