apider/Llama-3.2-1B-Instruct-FineTome-100k
The apider/Llama-3.2-1B-Instruct-FineTome-100k is a 1 billion parameter instruction-tuned Llama model developed by apider. It was fine-tuned from unsloth/Llama-3.2-1B-Instruct using the mlabonne/FineTome-100k dataset for one epoch. This model leverages Unsloth and Huggingface's TRL library for accelerated training, making it suitable for tasks requiring a compact yet capable instruction-following language model.
Loading preview...
Overview
The apider/Llama-3.2-1B-Instruct-FineTome-100k is a 1 billion parameter instruction-tuned language model developed by apider. It is based on the Llama architecture, specifically fine-tuned from the unsloth/Llama-3.2-1B-Instruct model.
Key Characteristics
- Base Model: Fine-tuned from
unsloth/Llama-3.2-1B-Instruct. - Training Data: Utilizes the
mlabonne/FineTome-100kfull dataset for one epoch of fine-tuning. - Training Efficiency: Training was significantly accelerated (2x faster) by employing Unsloth and Huggingface's TRL library.
- License: Distributed under the Apache-2.0 license.
Use Cases
This model is designed for instruction-following tasks, benefiting from its fine-tuning on a dedicated instruction dataset. Its compact size (1 billion parameters) combined with optimized training makes it a candidate for applications where computational resources are a consideration, or for further domain-specific fine-tuning.