Update README.md
Browse files
README.md
CHANGED
@@ -13,6 +13,87 @@ base_model:
|
|
13 |
Spaetzle-v31-7b is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
14 |
* [yleo/EmertonMonarch-7B](https://huggingface.co/yleo/EmertonMonarch-7B)
|
15 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
## 🧩 Configuration
|
17 |
|
18 |
```yaml
|
|
|
13 |
Spaetzle-v31-7b is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
14 |
* [yleo/EmertonMonarch-7B](https://huggingface.co/yleo/EmertonMonarch-7B)
|
15 |
|
16 |
+
| Model |AGIEval|GPT4All|TruthfulQA|Bigbench|Average|
|
17 |
+
|--------------------------------------------------------------|------:|------:|---------:|-------:|------:|
|
18 |
+
|[Spaetzle-v31-7b](https://huggingface.co/cstr/Spaetzle-v31-7b)| 46.23| 76.6| 69.58| 46.79| 59.8|
|
19 |
+
|
20 |
+
### AGIEval
|
21 |
+
| Task |Version| Metric |Value| |Stderr|
|
22 |
+
|------------------------------|------:|--------|----:|---|-----:|
|
23 |
+
|agieval_aqua_rat | 0|acc |28.74|± | 2.85|
|
24 |
+
| | |acc_norm|27.56|± | 2.81|
|
25 |
+
|agieval_logiqa_en | 0|acc |39.63|± | 1.92|
|
26 |
+
| | |acc_norm|40.25|± | 1.92|
|
27 |
+
|agieval_lsat_ar | 0|acc |24.35|± | 2.84|
|
28 |
+
| | |acc_norm|24.35|± | 2.84|
|
29 |
+
|agieval_lsat_lr | 0|acc |54.31|± | 2.21|
|
30 |
+
| | |acc_norm|54.12|± | 2.21|
|
31 |
+
|agieval_lsat_rc | 0|acc |65.80|± | 2.90|
|
32 |
+
| | |acc_norm|66.54|± | 2.88|
|
33 |
+
|agieval_sat_en | 0|acc |79.13|± | 2.84|
|
34 |
+
| | |acc_norm|79.61|± | 2.81|
|
35 |
+
|agieval_sat_en_without_passage| 0|acc |46.12|± | 3.48|
|
36 |
+
| | |acc_norm|45.15|± | 3.48|
|
37 |
+
|agieval_sat_math | 0|acc |35.00|± | 3.22|
|
38 |
+
| | |acc_norm|32.27|± | 3.16|
|
39 |
+
|
40 |
+
Average: 46.23%
|
41 |
+
|
42 |
+
### GPT4All
|
43 |
+
| Task |Version| Metric |Value| |Stderr|
|
44 |
+
|-------------|------:|--------|----:|---|-----:|
|
45 |
+
|arc_challenge| 0|acc |64.76|± | 1.40|
|
46 |
+
| | |acc_norm|66.89|± | 1.38|
|
47 |
+
|arc_easy | 0|acc |86.66|± | 0.70|
|
48 |
+
| | |acc_norm|82.83|± | 0.77|
|
49 |
+
|boolq | 1|acc |87.80|± | 0.57|
|
50 |
+
|hellaswag | 0|acc |67.43|± | 0.47|
|
51 |
+
| | |acc_norm|85.85|± | 0.35|
|
52 |
+
|openbookqa | 0|acc |38.00|± | 2.17|
|
53 |
+
| | |acc_norm|48.80|± | 2.24|
|
54 |
+
|piqa | 0|acc |83.57|± | 0.86|
|
55 |
+
| | |acc_norm|84.71|± | 0.84|
|
56 |
+
|winogrande | 0|acc |79.32|± | 1.14|
|
57 |
+
|
58 |
+
Average: 76.6%
|
59 |
+
|
60 |
+
### TruthfulQA
|
61 |
+
| Task |Version|Metric|Value| |Stderr|
|
62 |
+
|-------------|------:|------|----:|---|-----:|
|
63 |
+
|truthfulqa_mc| 1|mc1 |53.37|± | 1.75|
|
64 |
+
| | |mc2 |69.58|± | 1.48|
|
65 |
+
|
66 |
+
Average: 69.58%
|
67 |
+
|
68 |
+
### Bigbench
|
69 |
+
| Task |Version| Metric |Value| |Stderr|
|
70 |
+
|------------------------------------------------|------:|---------------------|----:|---|-----:|
|
71 |
+
|bigbench_causal_judgement | 0|multiple_choice_grade|56.84|± | 3.60|
|
72 |
+
|bigbench_date_understanding | 0|multiple_choice_grade|66.94|± | 2.45|
|
73 |
+
|bigbench_disambiguation_qa | 0|multiple_choice_grade|44.57|± | 3.10|
|
74 |
+
|bigbench_geometric_shapes | 0|multiple_choice_grade|21.17|± | 2.16|
|
75 |
+
| | |exact_str_match | 0.28|± | 0.28|
|
76 |
+
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|31.80|± | 2.08|
|
77 |
+
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|22.57|± | 1.58|
|
78 |
+
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|56.00|± | 2.87|
|
79 |
+
|bigbench_movie_recommendation | 0|multiple_choice_grade|45.40|± | 2.23|
|
80 |
+
|bigbench_navigate | 0|multiple_choice_grade|52.80|± | 1.58|
|
81 |
+
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|70.65|± | 1.02|
|
82 |
+
|bigbench_ruin_names | 0|multiple_choice_grade|50.67|± | 2.36|
|
83 |
+
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|30.66|± | 1.46|
|
84 |
+
|bigbench_snarks | 0|multiple_choice_grade|71.27|± | 3.37|
|
85 |
+
|bigbench_sports_understanding | 0|multiple_choice_grade|74.34|± | 1.39|
|
86 |
+
|bigbench_temporal_sequences | 0|multiple_choice_grade|49.80|± | 1.58|
|
87 |
+
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|22.16|± | 1.18|
|
88 |
+
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|18.57|± | 0.93|
|
89 |
+
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|56.00|± | 2.87|
|
90 |
+
|
91 |
+
Average: 46.79%
|
92 |
+
|
93 |
+
Average score: 59.8%
|
94 |
+
|
95 |
+
Elapsed time: 02:09:50
|
96 |
+
|
97 |
## 🧩 Configuration
|
98 |
|
99 |
```yaml
|