Downloads · 30 days
3
6% of all-time downloads
qwen-mix-12
cpral/qwen-mix-12
qwen-mix-12 is a machine learning model from cpral. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
All-time downloads
54
Public
Repo size
181 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors180 GB · 100%
At a glance
- Model type
- qwen3_5_moe
- Access
- Public
- Created
- May 6, 2026
- Updated
- May 6, 2026
- SHA
- feb5f40c
Model card
From the Hugging Face model README
-- A perplexity: 3.05385475
-- B perplexity: 3.07927058
-- A label in top-K:
K = 1: 0.7305
K = 2: 0.8217
K = 3: 0.8606
K = 4: 0.8829
K = 5: 0.8976
-- B label in top-K:
K = 1: 0.7284
K = 2: 0.8206
K = 3: 0.8596
K = 4: 0.8823
K = 5: 0.8969
-- Top-K agreement, A vs B:
K = 1: 0.9648
K = 2: 0.8422
K = 3: 0.6789
K = 4: 0.5189
K = 5: 0.3804
-- KL divergence (A, B): 0.01566811
-- KL divergence (B, A): 0.01502759
- Type
- qwen3_5_moe
- Created
- May 6, 2026
- Updated
- May 6, 2026