Downloads · 30 days
8
4% of all-time downloads
akshatladdha16/Llama-3.2-3B-Chemistry-Tutor-LoRA
Llama-3.2-3B-Chemistry-Tutor-LoRA is a text generation model from akshatladdha16. Use it when you need the model to write or continue text. It is set up for peft.
This is a PEFT (Parameter Efficient Fine-Tuning) adapter trained on chemistry educational content using QLoRA (Quantized Low-Rank Adaptation) technique. The adapter is designed to enhance Llama-3.2-3B's capabilities i…
Downloads · 30 days
8
4% of all-time downloads
All-time downloads
205
Public
Repo size
115 MB
Likes
5
Public
Click a slice to open those files.
.safetensors97.3 MB · 84%
From the Hugging Face model README
This is a PEFT (Parameter Efficient Fine-Tuning) adapter trained on chemistry educational content using QLoRA (Quantized Low-Rank Adaptation) technique. The adapter is designed to enhance Llama-3.2-3B's capabilities in answering chemistry-related questions.
This model is a QLoRA fine-tuned version of Meta-Llama-3.2-3B specifically optimized for chemistry question-answering tasks. The adapter layers were trained on a diverse chemistry dataset containing 4.4k+ educational Q&A pairs covering fundamental to advanced chemistry concepts.
Created a dedicated Google Collab notebook for anyone to infer the fine tuned adapter layer with base model. Use your personal access token from huggingface account Google colab Notebook: [https://colab.research.google.com/drive/16N_lnLKieJjMunvIXb59LtGavifx96nx#scrollTo=nd3kQhZbm2z9]
The model was fine-tuned on a curated dataset of chemistry educational content, focusing on:
[https://huggingface.co/datasets/KadamParth/NCERT_Chemistry_12th]
### Conversation:
User: {chemistry_question}
Assistant: {response}
Training Configuration:
LoRA Settings:
NVIDIA RTX 3050 with 4GB VRAM
If you use this model, please cite:
@misc{llama-chemistry-adapter,
author = {Akshat Rai Laddha},
title = {Chemistry QLoRA Adapter for Llama-3.2-3B},
year = {2025},
publisher = {Hugging Face},
journal = {Hugging Face Model Hub},
}