Hugging Face's logo

NexaAIDev
/
DeepSeek-R1-Distill-Llama-8B-NexaQuant

Transformers
GGUF
deepseek
llama
llama-3
meta
GGUF
Inference Endpoints
conversational
Model card Files Community
1
yli-nexa4ai commited on
Commit
0a76fab
·
verified ·
1 Parent(s): 94e62a8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -7
README.md CHANGED
@@ -55,16 +55,10 @@ The benchmarks show that NexaQuant’s 4-bit model preserves the reasoning capac
55
  |----------------------------|------------|-------------------|-------------------|
56
  | **HellaSwag** | 57.07 | 52.12 | 54.56 |
57
  | **MMLU** | 55.59 | 52.82 | 54.94 |
58
- | **Humanities** | 50.48 | 44.74 | 50.24 |
59
- | **Social Sciences** | 65.32 | 65.62 | 64.74 |
60
- | **STEM** | 47.73 | 52.50 | 46.75 |
61
  | **ARC Easy** | 74.49 | 69.32 | 71.72 |
62
  | **MathQA** | 35.34 | 30.00 | 32.46 |
63
  | **PIQA** | 78.56 | 76.09 | 77.68 |
64
- | **IFEval - Inst - Loose** | 44.24 | 44.95 | 42.45 |
65
- | **IFEval - Inst - Strict** | 42.57 | 44.95 | 40.05 |
66
- | **IFEval - Prom - Loose** | 30.31 | 25.74 | 28.47 |
67
- | **IFEval - Prom - Strict** | 27.91 | 25.74 | 25.51 |
68
 
69
  ## Run locally
70
 
 
55
  |----------------------------|------------|-------------------|-------------------|
56
  | **HellaSwag** | 57.07 | 52.12 | 54.56 |
57
  | **MMLU** | 55.59 | 52.82 | 54.94 |
 
 
 
58
  | **ARC Easy** | 74.49 | 69.32 | 71.72 |
59
  | **MathQA** | 35.34 | 30.00 | 32.46 |
60
  | **PIQA** | 78.56 | 76.09 | 77.68 |
61
+ | **IFEval** | 36.26 | 35.35 | 34.12 |
 
 
 
62
 
63
  ## Run locally
64