Hugging Face's logo

Dracones
/
miqu-1-70b-sf_exl2_2.25bpw

Text Generation
Transformers
Safetensors
English
llama
exl2
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
1
Dracones commited on
Commit
ee1f05a
•
1 Parent(s): 8222ae6

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +10 -43
README.md CHANGED
@@ -39,49 +39,16 @@ Below are the perplexity scores for the EXL2 models. A lower score is better.
39
 
40
  Here are the EQ Bench scores for the EXL2 quants using Alpaca, ChatML, Mistral, Vicuna-v1.1 and Vicuna-v0 prompt templates. A higher score is better.
41
 
42
- | Quant Size | Instruct Template | Score |
43
- |------------|-------------------|-------|
44
- | 5.0 | ChatML | 79.91 |
45
- | 5.0 | Alpaca | 81.45 |
46
- | 5.0 | Mistral | 81.11 |
47
- | 5.0 | Vicuna-v1.1 | 78.37 |
48
- | 5.0 | Vicuna-v0 | 76.64 |
49
- | 4.5 | ChatML | 80.64 |
50
- | 4.5 | Alpaca | 80.9 |
51
- | 4.5 | Mistral | 81.65 |
52
- | 4.5 | Vicuna-v1.1 | 77.04 |
53
- | 4.5 | Vicuna-v0 | 74.6 |
54
- | 4.0 | ChatML | 80.78 |
55
- | 4.0 | Alpaca | 79.53 |
56
- | 4.0 | Mistral | 82.78 |
57
- | 4.0 | Vicuna-v1.1 | 79.17 |
58
- | 4.0 | Vicuna-v0 | 76.41 |
59
- | 3.5 | ChatML | 81.11 |
60
- | 3.5 | Alpaca | 82.42 |
61
- | 3.5 | Mistral | 82.34 |
62
- | 3.5 | Vicuna-v1.1 | 81.04 |
63
- | 3.5 | Vicuna-v0 | 78.09 |
64
- | 3.0 | ChatML | 79.13 |
65
- | 3.0 | Alpaca | 77.74 |
66
- | 3.0 | Mistral | 80.11 |
67
- | 3.0 | Vicuna-v1.1 | 79.38 |
68
- | 3.0 | Vicuna-v0 | 77.25 |
69
- | 2.75 | ChatML | 79.6 |
70
- | 2.75 | Alpaca | 77.85 |
71
- | 2.75 | Mistral | 79.71 |
72
- | 2.75 | Vicuna-v1.1 | 76.93 |
73
- | 2.75 | Vicuna-v0 | 75.91 |
74
- | 2.5 | ChatML | 77.45 |
75
- | 2.5 | Alpaca | 77.0 |
76
- | 2.5 | Mistral | 78.4 |
77
- | 2.5 | Vicuna-v1.1 | 75.86 |
78
- | 2.5 | Vicuna-v0 | 75.25 |
79
- | 2.25 | ChatML | 77.18 |
80
- | 2.25 | Alpaca | 74.06 |
81
- | 2.25 | Mistral | 76.75 |
82
- | 2.25 | Vicuna-v1.1 | 75.56 |
83
- | 2.25 | Vicuna-v0 | 74.28 |
84
-
85
 
86
 
87
  ### Perplexity Script
 
39
 
40
  Here are the EQ Bench scores for the EXL2 quants using Alpaca, ChatML, Mistral, Vicuna-v1.1 and Vicuna-v0 prompt templates. A higher score is better.
41
 
42
+ | Quant Size | ChatML | Alpaca | Mistral | Vicuna-v1.1 | Vicuna-v0 |
43
+ |------------|--------|--------|---------|-------------|-----------|
44
+ | 5.0 | 79.91 | 81.45 | 81.11 | 78.37 | 76.64 |
45
+ | 4.5 | 80.64 | 80.9 | 81.65 | 77.04 | 74.6 |
46
+ | 4.0 | 80.78 | 79.53 | 82.78 | 79.17 | 76.41 |
47
+ | 3.5 | 81.11 | 82.42 | 82.34 | 81.04 | 78.09 |
48
+ | 3.0 | 79.13 | 77.74 | 80.11 | 79.38 | 77.25 |
49
+ | 2.75 | 79.6 | 77.85 | 79.71 | 76.93 | 75.91 |
50
+ | 2.5 | 77.45 | 77.0 | 78.4 | 75.86 | 75.25 |
51
+ | 2.25 | 77.18 | 74.06 | 76.75 | 75.56 | 74.28 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52
 
53
 
54
  ### Perplexity Script