Mistral 7b Science Mcq GGUF
This repository contains a fine-tuned version of Mistral-7B-Instruct-v0.1 optimized for high-school level science multiple-choice question (MCQ) answering, covering Biology, Chemistry, and Physics.
Fuente del modelo
- Tipo
- Other
- Tamaño del archivo
- 4,07 GB
- Versión
- vRevision 6213e3d
- Actualizado
- 2 oct 2026
- Publicador
- @HangTuahMalayWarrior
- Descargas
- 7
- Parámetros
- 7B
- Fuente
- Hugging Face
Descripción de la fuente
For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/modelcard.md?plain=1
Doc / guide: https://huggingface.co/docs/hub/model-cards
{}
📚 Mistral-7B Science MCQ Fine-Tuned (GGUF)
Model Overview
This repository contains a fine-tuned version of Mistral-7B-Instruct-v0.1 optimized for high-school level science multiple-choice question (MCQ) answering, covering Biology, Chemistry, and Physics.
The model is provided in GGUF format for efficient local inference (e.g., LM Studio, llama.cpp-based runtimes).
🧠 Base Model
The base model used for fine-tuning:
- https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF
- Original architecture: Mistral-7B Instruct v0.1
📊 Training Dataset
The model was fine-tuned using selected subsets of the MMLU (Massive Multitask Language Understanding) benchmark:
-
High School Biology https://huggingface.co/datasets/cais/mmlu/viewer/high_school_biology
-
High School Chemistry https://huggingface.co/datasets/cais/mmlu/viewer/high_school_chemistry
-
High School Physics https://huggingface.co/datasets/cais/mmlu/viewer/high_school_physics
These datasets consist of multiple-choice science questions designed to evaluate reasoning and subject knowledge at a high-school academic level.
⚙️ Model Variants
1. Hugging Face Fine-Tuned Model (Transformers Format)
https://huggingface.co/HangTuahMalayWarrior/mistral-7b-science-mcq-finetuned
2. GGUF Model (LM Studio / llama.cpp Compatible)
Recommended quantization: Q4_K_M
💻 How to Use in LM Studio
- Open LM Studio
- Go to the Models tab
- Search:
mistral-7b-science-mcq-gguf
- Download and load the model
- Start prompting for science MCQs
🧪 Intended Use
This model is optimized for:
- High school science MCQ answering
- Educational Q&A
- Science tutoring assistance
- Exam-style question practice
⚠️ Limitations
- Not guaranteed to be 100% accurate in scientific reasoning
image
- May hallucinate or guess answers when uncertain
- Not a replacement for verified textbooks or academic sources
- Performance depends on prompt quality and context
📦 Why GGUF Version?
The GGUF version is provided for:
- Fast CPU/GPU inference
- Local deployment (offline usage)
- Compatibility with LM Studio and llama.cpp-based tools
- Lower memory requirements via quantization
📌 Notes
We also uploaded this model for open archival and public access, allowing anyone to download and experiment with the fine-tuned science MCQ model.
🙌 Acknowledgements
- Mistral AI for the base model
- TheBloke
- CAIS MMLU dataset contributors
Fuentes
1 fuente
Fuentes
1 fuenteVerificado 2 oct
Artefactos del modelo
1 artefacto
Artefactos del modelo
1 artefactomistral-7b-science-mcq-finetuned.Q4_K_M.gguf
gguf · 4,07 GB · SHA-256 f3766e52d300…a526 · Hugging Face
DescargarExtractos de fuentes
3 extractos
Extractos de fuentes
3 extractos--- # For reference on model card metadata, see the spec: https://github.com/huggingface/hub-docs/blob/main/modelcard.md?plain=1 # Doc / guide: https://huggingface.co/docs/hub/model-cards {} --- --- # 📚 Mistral-7B Science MCQ Fine-Tuned (GGUF) ## Model Overview This repository contains a fine-tuned version of **Mistral-7B-Instruct-v0.1** optimized for **high-school level science multiple-choice question (MCQ) answering**, covering Biology, Chemistry, and Physics. The model is provided in **GGUF format** for efficient local inference (e.g., LM Studio, llama.cpp-based runtimes). --- ## 🧠 Base Model The base model used for fine-tuning: * [https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF](https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF) * Original architecture: Mistral-7B Instruct v0.1 --- ## 📊 Training Dataset The model was fine-tuned using selected subsets of the **MMLU (Massive Multitask Language Understanding)** benchmark: * High School Biology [https://huggingface.co/datasets/cais/mmlu/viewer/high_school_biology](https://huggingface.co/datasets/cais/mmlu/viewer/high_school_biology) * High School Chemistry [https://huggingface.co/datasets/cais/mmlu/viewer/high_school_chemistry](https://huggingface.co/datasets/cais/mmlu/viewer/high_school_chemistry) * High School Physics [https://huggingface.co/datasets/cais/mmlu/viewer/high_school_physics](https://huggingface.co/datasets/cais/mmlu/viewer/high_school_physics) These datasets consist of multiple-choice science questions designed to evaluate reasoning and subject knowledge at a high-school academic level. --- ## ⚙️ Model Variants ### 1. Hugging Face Fine-Tuned Model (Transformers Format) [https://huggingface.co/HangTuahMalayWarrior/mistral-7b-science-mcq-finetuned](https://huggingface.co/HangTuahMalayWarrior/mistral-7b-science-mcq-finetuned) ### 2. GGUF Model (LM Studio / llama.cpp Compatible) [https://huggingface.co/HangTuahMalayWarrior/mistral-7b-science-mcq-gguf/blob/main/mistral-7b-science-mcq-finetuned.Q4_K_M.gguf](https://huggingface.co/HangTuahMalayWarrior/mistral-7b-science-mcq-gguf/blob/main/mistral-7b-science-mcq-finetuned.Q4_K_M.gguf) Recommended quantization: **Q4_K_M** --- ## 💻 How to Use in LM Studio 1. Open **LM Studio** 2. Go to the **Models** tab 3. Search: ``` mistral-7b-science-mcq-gguf ``` 4. Download and load the model 5. Start...
Source context: 7 downloads · 0 likes · Repo HangTuahMalayWarrior/mistral-7b-science-mcq-gguf
Source context: 21 downloads · 0 likes · Repo HangTuahMalayWarrior/mistral-7b-science-mcq-gguf
