Hugging Face's logo

NexaAIDev
/
DeepSeek-R1-Distill-Qwen-1.5B-NexaQuant

Transformers
GGUF
deepseek
qwen
qwen2
GGUF
Inference Endpoints
conversational
Model card Files Community
3
yli-nexa4ai commited on
Commit
6889276
·
verified ·
1 Parent(s): 0dd2af4

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -7
README.md CHANGED
@@ -54,16 +54,10 @@ The benchmarks show that NexaQuant’s 4-bit model preserves the reasoning capac
54
  |----------------------------|------------|-------------------|-------------------|
55
  | **HellaSwag** | 35.81 | 34.31 | 34.60 |
56
  | **MMLU** | 37.31 | 35.49 | 37.41 |
57
- | **Humanities** | 31.86 | 34.87 | 30.97 |
58
- | **Social Sciences** | 41.50 | 38.17 | 42.09 |
59
- | **STEM** | 38.60 | 35.74 | 39.26 |
60
  | **ARC Easy** | 67.55 | 54.20 | 65.53 |
61
  | **MathQA** | 41.04 | 28.51 | 39.87 |
62
  | **PIQA** | 65.56 | 61.70 | 65.07 |
63
- | **IFEval - Inst - Loose** | 25.06 | 24.77 | 28.54 |
64
- | **IFEval - Inst - Strict** | 23.62 | 22.94 | 27.94 |
65
- | **IFEval - Prom - Loose** | 13.86 | 10.29 | 15.71 |
66
- | **IFEval - Prom - Strict** | 12.57 | 8.09 | 15.16 |
67
 
68
  ## Run locally
69
 
 
54
  |----------------------------|------------|-------------------|-------------------|
55
  | **HellaSwag** | 35.81 | 34.31 | 34.60 |
56
  | **MMLU** | 37.31 | 35.49 | 37.41 |
 
 
 
57
  | **ARC Easy** | 67.55 | 54.20 | 65.53 |
58
  | **MathQA** | 41.04 | 28.51 | 39.87 |
59
  | **PIQA** | 65.56 | 61.70 | 65.07 |
60
+ | **IFEval** | 18.78 | 16.52 | 21.84 |
 
 
 
61
 
62
  ## Run locally
63