|
|
--- |
|
|
license: apache-2.0 |
|
|
pipeline_tag: text-generation |
|
|
library_name: mlx |
|
|
tags: |
|
|
- vllm |
|
|
- mlx |
|
|
base_model: openai/gpt-oss-120b |
|
|
--- |
|
|
**See gpt-oss-120b 6.5bit MLX in action - [demonstration video](https://youtube.com/xcreate)** |
|
|
|
|
|
*q6.5bit quant typically achieves 1.128 perplexity in our testing which is equivalent to q8 perplexity (1.128).* |
|
|
| Quantization | Perplexity | |
|
|
|:------------:|:----------:| |
|
|
| **q2** | 41.293 | |
|
|
| **q3** | 1.900 | |
|
|
| **q4** | 1.168 | |
|
|
| **q6** | 1.128 | |
|
|
| **q8** | 1.128 | |
|
|
|
|
|
## Usage Notes |
|
|
|
|
|
* Built with a modified version of [MLX](https://github.com/ml-explore/mlx) 0.26 |
|
|
* Memory usage: ~95 GB |
|
|
* Expect ~60 tokens/s |
|
|
* For more details see [demonstration video](https://youtube.com/xcreate) or visit [Open AI gpt-oss-20b](https://huggingface.co/openai/gpt-oss-120b). |
