occitan-gemma-3-4b-it-dora

Occitan Gemma-3-4B-IT (DoRA Merged)

This repository contains a fine-tuned version of Google's Gemma-3-4B-IT specifically optimized for the Occitan language.

This version was trained using DoRA (Weight-Decomposed Low-Rank Adaptation). Compared to standard LoRA, DoRA decomposes the weights into magnitude and direction components, allowing the model to learn linguistic patterns that more closely resemble full fine-tuning.

📁 Repository Structure

• Root Directory: Full merged Safetensors weights (compatible with transformers).
• /gguf Folder: Quantized versions (Q4_K_M, Q5_K_M, Q8_0, etc.) for local inference via LM Studio, Ollama, or llama.cpp.
• /adapter Folder: The raw DoRA adapter files.

🚀 How to Use

Using Transformers (Python)

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "julienp79/occitan-gemma-3-4b-it-dora"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

messages = [
    {"role": "user", "content": "Pòdes m'ajudar a escriure un pichon tèxt en occitan?"},
]

inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
outputs = model.generate(inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))