Qwen2.5-3B-Instruct-mlx-16bit
16-bit MLX conversion of Qwen/Qwen2.5-3B-Instruct for Apple Silicon.
Converted by: SirSahOl Source model: Qwen/Qwen2.5-3B-Instruct Framework: MLX by Apple Quantization: 16-bit Format: safetensors License: other
Quick Start
Installation
pip install mlx-lm
CLI Usage
# Chat interactively
mlx_lm.chat --model SirSahOl/Qwen2.5-3B-Instruct-chat-mlx-16bit
# Generate text
mlx_lm.generate --model SirSahOl/Qwen2.5-3B-Instruct-chat-mlx-16bit --prompt "Your prompt here"
Python Usage
from mlx_lm import load, generate
model, tokenizer = load("SirSahOl/Qwen2.5-3B-Instruct-chat-mlx-16bit")
response = generate(model, tokenizer, prompt="Your prompt here", max_tokens=256)
print(response)
Performance Benchmarks
Benchmarks coming soon.
Who Should Use This?
| Your Hardware | Recommended Quantization |
|---|---|
| M1/M2 (8GB) | 4-bit — Best balance of quality and memory usage |
| M1/M2 Pro/Max (16-32GB) | 8-bit — Higher quality with reasonable memory |
| M2/M3/M4 Ultra (64GB+) | 16-bit — Full precision, no quality loss |
General guidance:
- Use 4-bit if you want to run this model alongside other applications
- Use 8-bit if you have the memory and want better quality
- Use 16-bit for research, evaluation, or if memory isn't a concern
Other Quantization Variants
| Variant | Link |
|---|---|
| 16-bit | SirSahOl/Qwen2.5-3B-Instruct-chat-mlx-16bit |
Conversion Details
| Property | Value |
|---|---|
| Source Model | Qwen/Qwen2.5-3B-Instruct |
| Quantization | 16-bit |
| mlx-lm Version | 0.31.3 |
| Conversion Time | 42.73s |
| Output Size | 5.8 GB |
| Date | 2026-09-11T19:05:57.898413+00:00 |
Reproduction
To reproduce this conversion:
pip install mlx-lm==0.31.3
python3 -m mlx_lm.convert --hf-path Qwen/Qwen2.5-3B-Instruct --mlx-path output/Qwen2.5-3B-Instruct-mlx-16bit
Limitations & Known Issues
- Performance may degrade with very long contexts (>8K tokens) at lower quantization levels.
- This is a weight-only conversion; the model architecture and behavior are inherited from the source model.
- Quantization introduces a small quality loss compared to the original model. Lower bit counts = more loss.
- This model requires Apple Silicon (M1 or later) to run with MLX.
License
This model conversion inherits the license of the source model: other.
See the original model card for full license details.
Changelog
| Version | Date | Changes |
|---|---|---|
| v1.0 | 2026-09-11 | Initial conversion |
Converted with MLX Foundry — a professional pipeline for converting models to Apple MLX format.