Downloads · 30 days
49
8% of all-time downloads
turboderp/Devstral-2-123B-Instruct-2512-exl3
Devstral-2-123B-Instruct-2512-exl3 is a machine learning model from turboderp. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as other.
EXL3 quants of Devstral-2-123B-Instruct-2512
Downloads · 30 days
49
8% of all-time downloads
All-time downloads
649
Public
Repo size
370 GB
Likes
11
Public
Click a slice to open those files.
.svg15.8 KB · 77%
From the Hugging Face model README
EXL3 quants of Devstral-2-123B-Instruct-2512
2.00 bits per weight
2.25 bits per weight
2.50 bits per weight
3.00 bits per weight
3.50 bits per weight
4.00 bits per weight
5.00 bits per weight
| . | KL-div | HumanEval @1 | MMLU |
|---|---|---|---|
| 2.00 bpw | 0.824 | 86.59% | 83.00% ±2.64%¹ |
| 2.25 bpw | 0.523 | 90.24% | 84.50% ±2.54%¹ |
| 2.50 bpw | 0.346 | 90.85% | 84.50% ±2.54%¹ |
| 3.00 bpw | 0.149 | 91.46% | 87.50% ±2.32%¹ |
| 3.50 bpw | 0.060 | 89.02% | 88.00% ±2.28%¹ |
| 4.00 bpw | 0.028 | 92.68% | 88.50% ±2.24%¹ |
| 5.00 bpw | 0.008 | 90.85% | 88.00% ±2.28%¹ |
¹ 200 random samples, 95% CI