Update README.md
Browse files
README.md
CHANGED
|
@@ -94,6 +94,87 @@ resize_token_embeddings_to_32x: true
|
|
| 94 |
|
| 95 |
GGUF: https://huggingface.co/cowWhySo/Phi-3-mini-4k-instruct-Friendly-gguf
|
| 96 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 97 |
## Training Summary
|
| 98 |
|
| 99 |
```json
|
|
|
|
| 94 |
|
| 95 |
GGUF: https://huggingface.co/cowWhySo/Phi-3-mini-4k-instruct-Friendly-gguf
|
| 96 |
|
| 97 |
+
## Benchmarks
|
| 98 |
+
|
| 99 |
+
| Model |AGIEval|GPT4All|TruthfulQA|Bigbench|Average|
|
| 100 |
+
|--------------------------------------------------------------------------------------------------|------:|------:|---------:|-------:|------:|
|
| 101 |
+
|[Phi-3-mini-4k-instruct-Friendly](https://huggingface.co/cowWhySo/Phi-3-mini-4k-instruct-Friendly)| 41| 67.56| 46.36| 39.3| 48.56|
|
| 102 |
+
|
| 103 |
+
### AGIEval
|
| 104 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 105 |
+
|------------------------------|------:|--------|----:|---|-----:|
|
| 106 |
+
|agieval_aqua_rat | 0|acc |22.05|± | 2.61|
|
| 107 |
+
| | |acc_norm|22.05|± | 2.61|
|
| 108 |
+
|agieval_logiqa_en | 0|acc |41.01|± | 1.93|
|
| 109 |
+
| | |acc_norm|41.32|± | 1.93|
|
| 110 |
+
|agieval_lsat_ar | 0|acc |22.17|± | 2.75|
|
| 111 |
+
| | |acc_norm|22.17|± | 2.75|
|
| 112 |
+
|agieval_lsat_lr | 0|acc |45.69|± | 2.21|
|
| 113 |
+
| | |acc_norm|45.88|± | 2.21|
|
| 114 |
+
|agieval_lsat_rc | 0|acc |59.48|± | 3.00|
|
| 115 |
+
| | |acc_norm|56.51|± | 3.03|
|
| 116 |
+
|agieval_sat_en | 0|acc |75.24|± | 3.01|
|
| 117 |
+
| | |acc_norm|70.39|± | 3.19|
|
| 118 |
+
|agieval_sat_en_without_passage| 0|acc |39.81|± | 3.42|
|
| 119 |
+
| | |acc_norm|37.86|± | 3.39|
|
| 120 |
+
|agieval_sat_math | 0|acc |33.64|± | 3.19|
|
| 121 |
+
| | |acc_norm|31.82|± | 3.15|
|
| 122 |
+
|
| 123 |
+
Average: 41.0%
|
| 124 |
+
|
| 125 |
+
### GPT4All
|
| 126 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 127 |
+
|-------------|------:|--------|----:|---|-----:|
|
| 128 |
+
|arc_challenge| 0|acc |49.74|± | 1.46|
|
| 129 |
+
| | |acc_norm|50.43|± | 1.46|
|
| 130 |
+
|arc_easy | 0|acc |76.68|± | 0.87|
|
| 131 |
+
| | |acc_norm|73.23|± | 0.91|
|
| 132 |
+
|boolq | 1|acc |79.27|± | 0.71|
|
| 133 |
+
|hellaswag | 0|acc |57.91|± | 0.49|
|
| 134 |
+
| | |acc_norm|77.13|± | 0.42|
|
| 135 |
+
|openbookqa | 0|acc |35.00|± | 2.14|
|
| 136 |
+
| | |acc_norm|43.80|± | 2.22|
|
| 137 |
+
|piqa | 0|acc |77.86|± | 0.97|
|
| 138 |
+
| | |acc_norm|79.54|± | 0.94|
|
| 139 |
+
|winogrande | 0|acc |69.53|± | 1.29|
|
| 140 |
+
|
| 141 |
+
Average: 67.56%
|
| 142 |
+
|
| 143 |
+
### TruthfulQA
|
| 144 |
+
| Task |Version|Metric|Value| |Stderr|
|
| 145 |
+
|-------------|------:|------|----:|---|-----:|
|
| 146 |
+
|truthfulqa_mc| 1|mc1 |31.21|± | 1.62|
|
| 147 |
+
| | |mc2 |46.36|± | 1.55|
|
| 148 |
+
|
| 149 |
+
Average: 46.36%
|
| 150 |
+
|
| 151 |
+
### Bigbench
|
| 152 |
+
| Task |Version| Metric |Value| |Stderr|
|
| 153 |
+
|------------------------------------------------|------:|---------------------|----:|---|-----:|
|
| 154 |
+
|bigbench_causal_judgement | 0|multiple_choice_grade|54.74|± | 3.62|
|
| 155 |
+
|bigbench_date_understanding | 0|multiple_choice_grade|66.67|± | 2.46|
|
| 156 |
+
|bigbench_disambiguation_qa | 0|multiple_choice_grade|29.46|± | 2.84|
|
| 157 |
+
|bigbench_geometric_shapes | 0|multiple_choice_grade|11.98|± | 1.72|
|
| 158 |
+
| | |exact_str_match | 0.00|± | 0.00|
|
| 159 |
+
|bigbench_logical_deduction_five_objects | 0|multiple_choice_grade|28.00|± | 2.01|
|
| 160 |
+
|bigbench_logical_deduction_seven_objects | 0|multiple_choice_grade|17.14|± | 1.43|
|
| 161 |
+
|bigbench_logical_deduction_three_objects | 0|multiple_choice_grade|45.67|± | 2.88|
|
| 162 |
+
|bigbench_movie_recommendation | 0|multiple_choice_grade|24.40|± | 1.92|
|
| 163 |
+
|bigbench_navigate | 0|multiple_choice_grade|53.70|± | 1.58|
|
| 164 |
+
|bigbench_reasoning_about_colored_objects | 0|multiple_choice_grade|68.10|± | 1.04|
|
| 165 |
+
|bigbench_ruin_names | 0|multiple_choice_grade|31.03|± | 2.19|
|
| 166 |
+
|bigbench_salient_translation_error_detection | 0|multiple_choice_grade|15.93|± | 1.16|
|
| 167 |
+
|bigbench_snarks | 0|multiple_choice_grade|77.35|± | 3.12|
|
| 168 |
+
|bigbench_sports_understanding | 0|multiple_choice_grade|52.64|± | 1.59|
|
| 169 |
+
|bigbench_temporal_sequences | 0|multiple_choice_grade|51.50|± | 1.58|
|
| 170 |
+
|bigbench_tracking_shuffled_objects_five_objects | 0|multiple_choice_grade|19.52|± | 1.12|
|
| 171 |
+
|bigbench_tracking_shuffled_objects_seven_objects| 0|multiple_choice_grade|13.89|± | 0.83|
|
| 172 |
+
|bigbench_tracking_shuffled_objects_three_objects| 0|multiple_choice_grade|45.67|± | 2.88|
|
| 173 |
+
|
| 174 |
+
Average: 39.3%
|
| 175 |
+
|
| 176 |
+
Average score: 48.56%
|
| 177 |
+
|
| 178 |
## Training Summary
|
| 179 |
|
| 180 |
```json
|