|
|
--- |
|
|
base_model: llama-3.2-3b-instruct-bnb-4bit |
|
|
tags: |
|
|
- text-generation-inference |
|
|
- transformers |
|
|
- unsloth |
|
|
- llama |
|
|
- gguf |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
--- |
|
|
|
|
|
<div align="center"> |
|
|
<img src="https://cdn-uploads.huggingface.co/production/uploads/669777597cb32718c20d97e9/4emWK_PB-RrifIbrCUjE8.png" |
|
|
alt="Title card" |
|
|
style="width: 500px; |
|
|
height: auto; |
|
|
object-position: center top;"> |
|
|
</div> |
|
|
|
|
|
**Website - https://www.alphaai.biz** |
|
|
|
|
|
# Uploaded model |
|
|
|
|
|
- **Developed by:** alphaaico |
|
|
- **License:** apache-2.0 |
|
|
- **Finetuned from model :** llama-3.2-3b-instruct-bnb-4bit |
|
|
|
|
|
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library. |
|
|
|
|
|
**AlphaAI-Chatty-INT1** |
|
|
|
|
|
Overview |
|
|
AlphaAI-Chatty-INT1 is a fine-tuned LLaMA 3B Small model optimized for chatty and engaging conversations. This model has been trained on a proprietary conversational dataset, making it well-suited for local deployments that require a natural, interactive dialogue experience. |
|
|
|
|
|
The model is available in GGUF format and has been quantized to different levels to support various hardware configurations. |
|
|
|
|
|
**Model Details** |
|
|
- Base Model: LLaMA 3B Small |
|
|
- Fine-tuned By: Alpha AI |
|
|
- Training Framework: Unsloth |
|
|
|
|
|
Quantization Levels Available: |
|
|
- q4_k_m |
|
|
- q5_k_m |
|
|
- q8_0 |
|
|
- 16-bit (full precision) https://huggingface.co/alphaaico/AlphaAI-Chatty-INT1-16bit |
|
|
|
|
|
Format: GGUF (Optimized for local deployments) |
|
|
|
|
|
Use Cases: |
|
|
- Conversational AI – Ideal for chatbots, virtual assistants, and customer support. |
|
|
- Local AI Deployments – Runs efficiently on local machines without requiring cloud-based inference. |
|
|
- Research & Experimentation – Suitable for studying conversational AI and fine-tuning on domain-specific datasets. |
|
|
|
|
|
Model Performance |
|
|
The model has been optimized for chat-style interactions, ensuring: |
|
|
- Engaging and context-aware responses |
|
|
- Efficient performance on consumer hardware |
|
|
- Balanced coherence and creativity in conversations |
|
|
|
|
|
Limitations & Biases |
|
|
This model, like any AI system, may have biases from the training data. It is recommended to use it responsibly and fine-tune further if needed for specific applications. |
|
|
|
|
|
License |
|
|
This model is released under a permissible license. Please check the Hugging Face repository for more details. |
|
|
|
|
|
Acknowledgments |
|
|
Special thanks to the Unsloth team for providing an optimized training pipeline for LLaMA models. |
