metadata
license: bsd-3-clause
base_model: Salesforce/codet5p-770m-py
tags:
- generated_from_trainer
datasets:
- mbpp
model-index:
- name: codet5p-770m-py-sanitized-codebleu-1-True-5e-05-0.1-lora-layer_18
results: []
library_name: peft
codet5p-770m-py-sanitized-codebleu-1-True-5e-05-0.1-lora-layer_18
This model is a fine-tuned version of Salesforce/codet5p-770m-py on the mbpp dataset. It achieves the following results on the evaluation set:
- Loss: 0.7028
- Codebleu: 0.1061
- Ngram Match Score: 0.0250
- Weighted Ngram Match Score: 0.0538
- Syntax Match Score: 0.1310
- Dataflow Match Score: 0.1145
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 64
Training results
Training Loss | Epoch | Step | Validation Loss | Codebleu | Ngram Match Score | Weighted Ngram Match Score | Syntax Match Score | Dataflow Match Score |
---|---|---|---|---|---|---|---|---|
0.9722 | 1.0 | 15 | 0.9246 | 0.0072 | 0.0000 | 0.0000 | 0.0079 | 0.0100 |
0.9623 | 2.0 | 30 | 0.9232 | 0.0072 | 0.0000 | 0.0000 | 0.0079 | 0.0100 |
0.9702 | 3.0 | 45 | 0.9197 | 0.0072 | 0.0000 | 0.0000 | 0.0079 | 0.0100 |
0.944 | 4.0 | 60 | 0.9118 | 0.0080 | 0.0000 | 0.0000 | 0.0079 | 0.0120 |
0.953 | 5.0 | 75 | 0.8956 | 0.0413 | 0.0007 | 0.0230 | 0.0370 | 0.0602 |
0.9402 | 6.0 | 90 | 0.8719 | 0.0750 | 0.0134 | 0.0426 | 0.0913 | 0.0823 |
0.9279 | 7.0 | 105 | 0.8508 | 0.0990 | 0.0215 | 0.0480 | 0.1217 | 0.1084 |
0.8723 | 8.0 | 120 | 0.8362 | 0.0989 | 0.0221 | 0.0519 | 0.1243 | 0.1044 |
0.8598 | 9.0 | 135 | 0.8227 | 0.0998 | 0.0231 | 0.0521 | 0.1243 | 0.1064 |
0.8459 | 10.0 | 150 | 0.8105 | 0.1034 | 0.0148 | 0.0396 | 0.1283 | 0.1165 |
0.8238 | 11.0 | 165 | 0.7952 | 0.0991 | 0.0162 | 0.0411 | 0.1190 | 0.1145 |
0.8176 | 12.0 | 180 | 0.7802 | 0.1020 | 0.0185 | 0.0459 | 0.1243 | 0.1145 |
0.7965 | 13.0 | 195 | 0.7719 | 0.1017 | 0.0173 | 0.0441 | 0.1243 | 0.1145 |
0.7755 | 14.0 | 210 | 0.7616 | 0.1017 | 0.0181 | 0.0441 | 0.1243 | 0.1145 |
0.7876 | 15.0 | 225 | 0.7538 | 0.1018 | 0.0186 | 0.0440 | 0.1243 | 0.1145 |
0.7887 | 16.0 | 240 | 0.7509 | 0.0992 | 0.0174 | 0.0438 | 0.1204 | 0.1124 |
0.7723 | 17.0 | 255 | 0.7467 | 0.1015 | 0.0182 | 0.0438 | 0.1217 | 0.1165 |
0.753 | 18.0 | 270 | 0.7440 | 0.1015 | 0.0182 | 0.0438 | 0.1217 | 0.1165 |
0.762 | 19.0 | 285 | 0.7405 | 0.1045 | 0.0200 | 0.0458 | 0.1243 | 0.1205 |
0.7537 | 20.0 | 300 | 0.7377 | 0.1029 | 0.0182 | 0.0447 | 0.1230 | 0.1185 |
0.7381 | 21.0 | 315 | 0.7344 | 0.1056 | 0.0188 | 0.0449 | 0.1296 | 0.1185 |
0.7389 | 22.0 | 330 | 0.7314 | 0.1048 | 0.0189 | 0.0449 | 0.1296 | 0.1165 |
0.7352 | 23.0 | 345 | 0.7301 | 0.1030 | 0.0189 | 0.0449 | 0.1270 | 0.1145 |
0.7421 | 24.0 | 360 | 0.7288 | 0.1029 | 0.0188 | 0.0449 | 0.1270 | 0.1145 |
0.7569 | 25.0 | 375 | 0.7271 | 0.1053 | 0.0190 | 0.0441 | 0.1270 | 0.1205 |
0.7339 | 26.0 | 390 | 0.7256 | 0.1071 | 0.0233 | 0.0529 | 0.1323 | 0.1165 |
0.7169 | 27.0 | 405 | 0.7238 | 0.1072 | 0.0236 | 0.0530 | 0.1323 | 0.1165 |
0.7278 | 28.0 | 420 | 0.7224 | 0.1062 | 0.0242 | 0.0535 | 0.1296 | 0.1165 |
0.7185 | 29.0 | 435 | 0.7215 | 0.1063 | 0.0250 | 0.0536 | 0.1296 | 0.1165 |
0.713 | 30.0 | 450 | 0.7204 | 0.1098 | 0.0254 | 0.0533 | 0.1283 | 0.1265 |
0.7145 | 31.0 | 465 | 0.7195 | 0.1101 | 0.0266 | 0.0551 | 0.1283 | 0.1265 |
0.7128 | 32.0 | 480 | 0.7180 | 0.1149 | 0.0282 | 0.0588 | 0.1389 | 0.1265 |
0.7131 | 33.0 | 495 | 0.7172 | 0.1140 | 0.0276 | 0.0588 | 0.1389 | 0.1245 |
0.7116 | 34.0 | 510 | 0.7154 | 0.1140 | 0.0276 | 0.0588 | 0.1389 | 0.1245 |
0.6959 | 35.0 | 525 | 0.7140 | 0.1148 | 0.0276 | 0.0588 | 0.1389 | 0.1265 |
0.7259 | 36.0 | 540 | 0.7131 | 0.1148 | 0.0276 | 0.0588 | 0.1389 | 0.1265 |
0.7094 | 37.0 | 555 | 0.7127 | 0.1148 | 0.0276 | 0.0588 | 0.1389 | 0.1265 |
0.7151 | 38.0 | 570 | 0.7119 | 0.1148 | 0.0276 | 0.0588 | 0.1389 | 0.1265 |
0.7082 | 39.0 | 585 | 0.7108 | 0.1140 | 0.0280 | 0.0586 | 0.1389 | 0.1245 |
0.7101 | 40.0 | 600 | 0.7102 | 0.1136 | 0.0271 | 0.0576 | 0.1402 | 0.1225 |
0.696 | 41.0 | 615 | 0.7098 | 0.1136 | 0.0271 | 0.0576 | 0.1402 | 0.1225 |
0.6955 | 42.0 | 630 | 0.7090 | 0.1136 | 0.0271 | 0.0576 | 0.1402 | 0.1225 |
0.6967 | 43.0 | 645 | 0.7083 | 0.1136 | 0.0271 | 0.0576 | 0.1402 | 0.1225 |
0.7006 | 44.0 | 660 | 0.7077 | 0.1136 | 0.0271 | 0.0576 | 0.1402 | 0.1225 |
0.6921 | 45.0 | 675 | 0.7072 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6866 | 46.0 | 690 | 0.7069 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6971 | 47.0 | 705 | 0.7067 | 0.1068 | 0.0247 | 0.0540 | 0.1310 | 0.1165 |
0.6943 | 48.0 | 720 | 0.7065 | 0.1060 | 0.0248 | 0.0538 | 0.1310 | 0.1145 |
0.6988 | 49.0 | 735 | 0.7060 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6865 | 50.0 | 750 | 0.7056 | 0.1068 | 0.0245 | 0.0539 | 0.1310 | 0.1165 |
0.7076 | 51.0 | 765 | 0.7053 | 0.1060 | 0.0248 | 0.0538 | 0.1310 | 0.1145 |
0.7049 | 52.0 | 780 | 0.7050 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.7008 | 53.0 | 795 | 0.7045 | 0.1119 | 0.0266 | 0.0576 | 0.1402 | 0.1185 |
0.6728 | 54.0 | 810 | 0.7043 | 0.1119 | 0.0266 | 0.0576 | 0.1402 | 0.1185 |
0.6918 | 55.0 | 825 | 0.7040 | 0.1119 | 0.0266 | 0.0576 | 0.1402 | 0.1185 |
0.6952 | 56.0 | 840 | 0.7037 | 0.1119 | 0.0266 | 0.0576 | 0.1402 | 0.1185 |
0.6986 | 57.0 | 855 | 0.7034 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.687 | 58.0 | 870 | 0.7032 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.7026 | 59.0 | 885 | 0.7030 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6817 | 60.0 | 900 | 0.7030 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6994 | 61.0 | 915 | 0.7029 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6843 | 62.0 | 930 | 0.7029 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6876 | 63.0 | 945 | 0.7028 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
0.6783 | 64.0 | 960 | 0.7028 | 0.1061 | 0.0250 | 0.0538 | 0.1310 | 0.1145 |
Framework versions
- PEFT 0.4.0
- Transformers 4.31.0
- Pytorch 2.0.1
- Datasets 2.14.4
- Tokenizers 0.13.3