Kwen-1.5B
Model Description
Kwen-1.5B is a fine-tuned version of the Qwen2.5-1.5B base model. This model has been specifically trained to embody the identity of Kwen, an AI assistant developed by The Kwen Foundation.
Through fine-tuning, the model has integrated its new identity while maintaining the high reasoning and linguistic capabilities of the original Qwen architecture.
Key Features
- Identity-Aware: Recognizes itself as Kwen and references the Kwen Foundation.
- Efficient Size: Optimized for consumer GPUs (like the RTX 4060).
- Architecture: Based on Qwen2.5, utilizing Causal Language Modeling with advanced attention mechanisms.
Model Lineage
- Base Model: Qwen/Qwen2.5-1.5B
- Developer: The Kwen Foundation
- Fine-tuning Method: LoRA/QLoRA (via Unsloth)
Quants (GGUF)
The weights provided in this repository include quantized versions created using llama.cpp.
| File Name | Method | Size | Description |
|---|---|---|---|
kwen_q4_k_m.gguf |
Q4_K_M | ~935 MB | High compression, balanced quality (Recommended) |
kwen_f16.gguf |
F16 | ~3.0 GB | Uncompressed weights for maximum precision |
How to use with Ollama
- Run ollama run hf.co/TheKwenFoundation/Kwen-1.5B:Q4_K_M
- Downloads last month
- 5
Hardware compatibility
Log In to add your hardware
4-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 1 Ask for provider support
Model tree for TheKwenFoundation/Kwen-1.5B
Base model
Qwen/Qwen2.5-1.5B