EdukaAI

Fine-Tune with Hugging Face

Use the "GitHub of AI" for cloud training, sharing models, and accessing a huge ecosystem of tools.

📋 Prerequisites

📦

1. Dataset Ready

Export your dataset from the EdukaAI application Export page. Hugging Face accepts JSON files directly.

Open the EdukaAI app and go to Export.

👤

2. Hugging Face Account

Create a free account at huggingface.co

1 Method A: AutoTrain (Easiest)

AutoTrain is Hugging Face's no-code solution. Just upload your data and click train!

  1. 1.Go to AutoTrain
  2. 2.Select "LLM Fine-tuning" project type
  3. 3.Upload your exported JSON file
  4. 4.Choose a base model (TinyLlama-1.1B recommended for testing)
  5. 5.Click "Start Training"

✅ Done! AutoTrain handles everything: preprocessing, training, and saving your model.

2 Method B: Training Notebooks

For more control, use Hugging Face's training notebooks in Google Colab or locally.

Quick Steps:

  1. 1. Search "LLM fine-tuning Hugging Face" in their docs
  2. 2. Open a notebook in Google Colab
  3. 3. Upload your EdukaAI dataset
  4. 4. Run cells sequentially
  5. 5. Download your fine-tuned model

📚 Popular Notebooks

  • • TRL (Transformer Reinforcement Learning): SFTTrainer for supervised fine-tuning
  • • PEFT: LoRA and QLoRA examples
  • • Unsloth: 2x faster training notebooks

3 Method C: Local Training with HF Tools

Use Hugging Face libraries on your own computer:

# Install libraries
pip install transformers datasets peft accelerate

# Load your EdukaAI dataset
from datasets import load_dataset
dataset = load_dataset('json', data_files='your_exported_file_alpaca.json')

# Fine-tune with PEFT/LoRA
from peft import LoraConfig
from transformers import AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained('TinyLlama/TinyLlama-1.1B-Chat-v1.0')
# ... training code ...

4 Share Your Model

One of Hugging Face's best features: easily share your fine-tuned model with the world!

from huggingface_hub import HfApi api = HfApi() api.upload_folder(folder_path="./model-output", repo_id="your-username/my-edukaai-model")

🌟 Benefits of sharing: Others can download your model, you get feedback, and you build a portfolio!

💡 Hugging Face Tips

🤝 Community

Join discussions and get help from the community on model pages.

🏆 Leaderboards

Compare models on Open LLM Leaderboard to see how yours ranks.

📊 Inference API

Test models instantly on the website without downloading.

🔒 Private Models

Keep models private while testing, then make public when ready.

← All Methods