Update README.md
Browse files
README.md
CHANGED
|
@@ -6,4 +6,68 @@ tags:
|
|
| 6 |
- open4bits
|
| 7 |
base_model: Qwen/Qwen3-14B-Base
|
| 8 |
pipeline_tag: text-generation
|
| 9 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
- open4bits
|
| 7 |
base_model: Qwen/Qwen3-14B-Base
|
| 8 |
pipeline_tag: text-generation
|
| 9 |
+
---
|
| 10 |
+
Here’s a professional **GitHub-ready `README.md`** for **Open4bits/Qwen3-14B-Base-MLX-FP16**:
|
| 11 |
+
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# Open4bits / Qwen3-14B-Base-MLX-FP16
|
| 15 |
+
|
| 16 |
+
This repository provides the **Qwen3-14B Base model converted to MLX format with FP16 precision**, published by Open4bits to enable efficient high-performance inference with reduced memory usage and broad hardware compatibility.
|
| 17 |
+
|
| 18 |
+
The underlying Qwen3-14B model and architecture are **developed and owned by the original creators**. This repository contains an FP16 precision MLX conversion of the original model weights.
|
| 19 |
+
|
| 20 |
+
Open4bits has started supporting **MLX models** to broaden compatibility with emerging quantization formats and efficient runtimes, allowing improved performance on a range of platforms.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
## Model Overview
|
| 25 |
+
|
| 26 |
+
**Qwen3-14B Base** is a 14-billion parameter transformer-based language model designed for strong general understanding, reasoning, and instruction following.
|
| 27 |
+
This release uses **FP16 precision** in **MLX format**, enabling efficient inference with balanced speed and quality.
|
| 28 |
+
|
| 29 |
+
---
|
| 30 |
+
|
| 31 |
+
## Model Details
|
| 32 |
+
|
| 33 |
+
* **Base Model:** Qwen3-14B
|
| 34 |
+
* **Precision:** FP16 (float16)
|
| 35 |
+
* **Format:** MLX
|
| 36 |
+
* **Task:** Text generation, instruction following
|
| 37 |
+
* **Weight tying:** Preserved
|
| 38 |
+
* **Compatibility:** MLX-enabled inference engines and runtimes
|
| 39 |
+
|
| 40 |
+
The FP16 format provides improved performance and reduced memory consumption compared to full FP32 precision while retaining high generation quality.
|
| 41 |
+
|
| 42 |
+
---
|
| 43 |
+
|
| 44 |
+
## Intended Use
|
| 45 |
+
|
| 46 |
+
This model is intended for:
|
| 47 |
+
|
| 48 |
+
* High-performance text generation and conversational applications
|
| 49 |
+
* CPU-based or accelerator-supported deployments
|
| 50 |
+
* Research, experimentation, and prototyping
|
| 51 |
+
* Offline or self-hosted AI systems
|
| 52 |
+
|
| 53 |
+
---
|
| 54 |
+
|
| 55 |
+
## Limitations
|
| 56 |
+
|
| 57 |
+
* Lower precision compared to non-quantized models
|
| 58 |
+
* Output quality depends on prompt design and inference parameters
|
| 59 |
+
* Not optimized for highly specialized domain-specific tasks without further fine-tuning
|
| 60 |
+
|
| 61 |
+
---
|
| 62 |
+
|
| 63 |
+
## License
|
| 64 |
+
|
| 65 |
+
This model follows the **Apache 2.0** of the base Qwen3-14B model.
|
| 66 |
+
Users must comply with the licensing conditions defined by the original model creators.
|
| 67 |
+
|
| 68 |
+
---
|
| 69 |
+
|
| 70 |
+
## Support
|
| 71 |
+
|
| 72 |
+
If you find this model useful, please consider supporting the project.
|
| 73 |
+
Your support helps Open4bits continue releasing and maintaining high-quality efficient models for the community.
|