library_name: transformers
license: apache-2.0
base_model: open-thoughts/OpenThinker-7B
tags:
- llama-factory
- full
- generated_from_trainer
- mlx
- mlx-my-repo
datasets:
- open-thoughts/open-thoughts-114k
model-index:
- name: OpenThinker-7B
results: []
About:
A fully open-source family of reasoning models built using a dataset derived by distilling DeepSeek-R1.
This model is a fine-tuned version of Qwen/Qwen2.5-7B-Instruct on the OpenThoughts-114k dataset dataset. This model improves upon the Bespoke-Stratos-7B model, which used 17k examples (Bespoke-Stratos-17k dataset).
Special thanks to the folks at Open Thoughts for fine-tuning this version of Qwen/Qwen2.5-7B-Instruct. More information about it can be found here:
https://huggingface.co/open-thoughts/OpenThinker-7B (Base Model)
https://github.com/open-thoughts/open-thoughts (Open Thoughts Git Repo)
I simply converted it to MLX format (using mlx-lm version 0.20.5.) with a quantization of 8-bit for better performance on Apple Silicon Macs (M1,M2,M3,M4 Chips).
Other Types:
| Link | Type | Size | Notes |
|---|---|---|---|
| [MLX] (https://huggingface.co/AlejandroOlmedo/OpenThinker-7B-8bit-mlx) | 8-bit | 8.10 GB | Best Quality |
| [MLX] (https://huggingface.co/AlejandroOlmedo/OpenThinker-7B-4bit-mlx) | 4-bit | 4.30 GB | Good Quality |
AlejandroOlmedo/OpenThinker-7B-8bit-mlx
The Model AlejandroOlmedo/OpenThinker-7B-8bit-mlx was converted to MLX format from open-thoughts/OpenThinker-7B using mlx-lm version 0.20.5.
Use with mlx
pip install mlx-lm
from mlx_lm import load, generate
model, tokenizer = load("AlejandroOlmedo/OpenThinker-7B-8bit-mlx")
prompt="hello"
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True)