You can deploy this model and start earning money today!

Moodoxyy/
mistral-7b-dsa-finetuned

Name: mistral-7b-dsa-finetuned
Availability: InStock

peft

safetensors

question-answering

Moodoxyy/dsaquizquestionandanswer

mistralai/Mistral-7B-v0.1

adapter

apache-2.0

Views

No views yet

Model Card for QuizBuddy QA Generator (Mistral-7B)

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 using the PEFT (LoRA) technique. It is designed to generate quiz-style questions and answers in various formats (MCQ, fill-in-the-blanks, true/false, etc.) from academic text inputs, primarily targeting BTech CSE core subjects such as DSA, OS, CN, and DBMS.

Model Details

Model Description

Developed by: Nikhil Sharma
License: Apache-2.0
Model type: Causal Language Model (Decoder-only)
Languages: English (en)
Fine-tuned from: mistralai/Mistral-7B-v0.1
Fine-tuning Framework: PEFT with LoRA
Purpose: Educational quiz question generation

Model Sources

Repository: [Private/Internal] (can be linked later)
Dataset: Moodoxyy/dsaquizquestionandanswer

Uses

Direct Use

This model can be used to:

Generate quiz-style questions (MCQs, FIBs, True/False, etc.)
Assist educators in automated test paper generation
Aid students in practice question creation for revision

Downstream Use

Can be plugged into:

QuizBuddy app or API for quiz generation from documents
Educational platforms for automated evaluation tools

Out-of-Scope Use

Generating questions unrelated to computer science domains
Chat-based tutoring or general-purpose reasoning without additional fine-tuning
Use in high-stakes testing environments without human oversight

Bias, Risks, and Limitations

May hallucinate facts outside of dataset context.
Model's understanding is confined to the data distribution of the fine-tuning dataset.
Evaluation metrics show no exact match; best use is assistive, not authoritative.

Recommendations

Always validate generated content before deployment.
Prefer using it in combination with human curation.
Refrain from using it to grade students automatically.

How to Get Started with the Model

python

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

base_model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1", device_map="auto", trust_remote_code=True)
peft_model = PeftModel.from_pretrained(base_model, "Nikhil/quizbuddy-mistral-lora")
tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")

input_text = "Explain stack and its operations"
inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
output = peft_model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(output[0], skip_special_tokens=True))

Moodoxyy/mistral-7b-dsa-finetuned