Downloads · 30 days
12
10% of all-time downloads
cognita-tinyllama
SyedaArisha/cognita-tinyllama
cognita-tinyllama is a text generation model from SyedaArisha. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as mit.
A fine-tuned TinyLlama 1.1B model trained on CS teaching conversations, built for Pakistani students learning computer science concepts.
All-time downloads
119
Public
Parameters
1.1B
807 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors807 MB · 100%
Parameter types
How the weights are stored.
U8969M · 87%
Try a prompt
Base models
Spaces
Model card
From the Hugging Face model README
Cognita — TinyLlama Fine-Tuned for CS Education
A fine-tuned TinyLlama 1.1B model trained on CS teaching conversations, built for Pakistani students learning computer science concepts.
Model Details
- Base model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
- Fine-tuning method: QLoRA (4-bit quantization)
- Training data: 5,030 CS teaching conversations
- Training loss: 2.34 → 0.55
- Hardware: Google Colab T4 GPU
Use Case
Answers computer science questions in a teaching style. Used as the backend for the Cognita Flutter mobile app.
Live Demo
Try it here: https://huggingface.co/spaces/SyedaArisha/cognita
Training Framework
PEFT, BitsAndBytesConfig, Transformers, HuggingFace Trainer
Developer
Syeda Arisha Hassan — COMSATS University Islamabad
- Task
- Text Generation
- Library
- peft
- Type
- llama
- License
- mit
- Languages
- en, ur
- Created
- Mar 27, 2026
- Updated
- May 30, 2026