File size: 2,047 Bytes
b728211 a082a3d b728211 a082a3d b728211 a082a3d b728211 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 |
---
base_model: unsloth/gemma-3-4b-it-unsloth-bnb-4bit
tags:
- text-generation-inference
- transformers
- unsloth
- gemma3
license: apache-2.0
language:
- en
---
# 🚀 Fine-tuned Gemma 3 Model (4B, 4-bit) by Webkul
This repository contains a fine-tuned version of [Unsloth's](https://github.com/unslothai/unsloth) `gemma-3-4b-it` model, optimized for lightweight 4-bit inference and instruction tuning using Hugging Face's [TRL](https://github.com/huggingface/trl) and Unsloth's speed-optimized framework.
---
## 🔧 Model Details
- **Base Model:** [`unsloth/gemma-3-4b-it-unsloth-bnb-4bit`](https://huggingface.co/unsloth/gemma-3-4b-it-unsloth-bnb-4bit)
- **Fine-tuned By:** [Webkul](https://webkul.com)
- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0)
- **Language:** English (`en`)
- **Model Size:** 4B parameters (4-bit quantized)
- **Frameworks Used:** Unsloth, Hugging Face Transformers, TRL
---
## 📚 Fine-tuning Dataset
This model was fine-tuned on unopim dev documentation available at [https://devdocs.unopim.com/](https://devdocs.unopim.com/), focusing on structured software documentation and developer support content.
---
## 💡 Intended Use
- Conversational AI assistants trained on UnoPIM developer docs
- API documentation question answering
- Developer tools and chatbot integrations
- Contextual helpdesk or onboarding bots for UnoPIM products
---
## 🧪 How to Use
You can use this model with the Hugging Face `transformers` library:
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
model_name = "webkul/gemma-3-4b-it-unopim-docs"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
input_text = "How do I integrate the UnoPIM API for product syncing?"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=300)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```
📄 License
This model is licensed under the Apache License 2.0.
---
|