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_17
results: []
library_name: peft
codet5p-770m-py-sanitized-codebleu-1-True-5e-05-0.1-lora-layer_17
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: 0.7150
- Codebleu: 0.0985
- Ngram Match Score: 0.0231
- Weighted Ngram Match Score: 0.0493
- Syntax Match Score: 0.1217
- Dataflow Match Score: 0.1064
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 |
---|---|---|---|---|---|---|---|---|
0.9723 | 1.0 | 15 | 0.9246 | 0.0072 | 0.0000 | 0.0000 | 0.0079 | 0.0100 |
0.9649 | 2.0 | 30 | 0.9232 | 0.0072 | 0.0000 | 0.0000 | 0.0079 | 0.0100 |
0.9719 | 3.0 | 45 | 0.9202 | 0.0072 | 0.0000 | 0.0000 | 0.0079 | 0.0100 |
0.9435 | 4.0 | 60 | 0.9142 | 0.0080 | 0.0000 | 0.0000 | 0.0079 | 0.0120 |
0.9529 | 5.0 | 75 | 0.9023 | 0.0096 | 0.0000 | 0.0002 | 0.0079 | 0.0161 |
0.9443 | 6.0 | 90 | 0.8827 | 0.0380 | 0.0003 | 0.0176 | 0.0423 | 0.0482 |
0.9339 | 7.0 | 105 | 0.8558 | 0.0731 | 0.0080 | 0.0338 | 0.0899 | 0.0823 |
0.8783 | 8.0 | 120 | 0.8325 | 0.0960 | 0.0224 | 0.0519 | 0.1230 | 0.0984 |
0.8592 | 9.0 | 135 | 0.8139 | 0.0969 | 0.0188 | 0.0423 | 0.1164 | 0.1104 |
0.8398 | 10.0 | 150 | 0.7967 | 0.0972 | 0.0160 | 0.0379 | 0.1111 | 0.1185 |
0.8187 | 11.0 | 165 | 0.7857 | 0.1118 | 0.0199 | 0.0470 | 0.1362 | 0.1265 |
0.8064 | 12.0 | 180 | 0.7785 | 0.0996 | 0.0176 | 0.0449 | 0.1270 | 0.1064 |
0.7903 | 13.0 | 195 | 0.7729 | 0.1082 | 0.0179 | 0.0450 | 0.1402 | 0.1145 |
0.7737 | 14.0 | 210 | 0.7677 | 0.1076 | 0.0169 | 0.0457 | 0.1349 | 0.1185 |
0.7851 | 15.0 | 225 | 0.7628 | 0.1028 | 0.0163 | 0.0435 | 0.1296 | 0.1124 |
0.7865 | 16.0 | 240 | 0.7590 | 0.1031 | 0.0170 | 0.0435 | 0.1323 | 0.1104 |
0.7691 | 17.0 | 255 | 0.7552 | 0.0979 | 0.0175 | 0.0435 | 0.1230 | 0.1064 |
0.749 | 18.0 | 270 | 0.7522 | 0.0979 | 0.0176 | 0.0435 | 0.1230 | 0.1064 |
0.7587 | 19.0 | 285 | 0.7489 | 0.0950 | 0.0169 | 0.0419 | 0.1164 | 0.1064 |
0.7513 | 20.0 | 300 | 0.7465 | 0.0983 | 0.0174 | 0.0453 | 0.1217 | 0.1084 |
0.7368 | 21.0 | 315 | 0.7439 | 0.1015 | 0.0179 | 0.0452 | 0.1257 | 0.1124 |
0.7351 | 22.0 | 330 | 0.7411 | 0.1048 | 0.0195 | 0.0469 | 0.1310 | 0.1145 |
0.7347 | 23.0 | 345 | 0.7395 | 0.1011 | 0.0200 | 0.0471 | 0.1257 | 0.1104 |
0.7342 | 24.0 | 360 | 0.7384 | 0.1011 | 0.0200 | 0.0471 | 0.1257 | 0.1104 |
0.7561 | 25.0 | 375 | 0.7361 | 0.1075 | 0.0219 | 0.0503 | 0.1323 | 0.1185 |
0.7272 | 26.0 | 390 | 0.7345 | 0.1041 | 0.0211 | 0.0486 | 0.1323 | 0.1104 |
0.7172 | 27.0 | 405 | 0.7330 | 0.1007 | 0.0220 | 0.0485 | 0.1257 | 0.1084 |
0.724 | 28.0 | 420 | 0.7320 | 0.1007 | 0.0220 | 0.0485 | 0.1257 | 0.1084 |
0.7098 | 29.0 | 435 | 0.7316 | 0.1024 | 0.0231 | 0.0486 | 0.1257 | 0.1124 |
0.7111 | 30.0 | 450 | 0.7312 | 0.1033 | 0.0235 | 0.0492 | 0.1257 | 0.1145 |
0.7076 | 31.0 | 465 | 0.7299 | 0.1017 | 0.0229 | 0.0492 | 0.1257 | 0.1104 |
0.7104 | 32.0 | 480 | 0.7282 | 0.1029 | 0.0236 | 0.0502 | 0.1283 | 0.1104 |
0.7059 | 33.0 | 495 | 0.7272 | 0.1046 | 0.0242 | 0.0503 | 0.1283 | 0.1145 |
0.7082 | 34.0 | 510 | 0.7254 | 0.1029 | 0.0236 | 0.0502 | 0.1283 | 0.1104 |
0.6904 | 35.0 | 525 | 0.7245 | 0.1090 | 0.0244 | 0.0517 | 0.1349 | 0.1185 |
0.7197 | 36.0 | 540 | 0.7237 | 0.1029 | 0.0236 | 0.0502 | 0.1283 | 0.1104 |
0.7051 | 37.0 | 555 | 0.7231 | 0.1029 | 0.0236 | 0.0502 | 0.1283 | 0.1104 |
0.7083 | 38.0 | 570 | 0.7227 | 0.1029 | 0.0236 | 0.0502 | 0.1283 | 0.1104 |
0.7018 | 39.0 | 585 | 0.7216 | 0.0994 | 0.0239 | 0.0501 | 0.1217 | 0.1084 |
0.7033 | 40.0 | 600 | 0.7209 | 0.1000 | 0.0238 | 0.0503 | 0.1230 | 0.1084 |
0.6924 | 41.0 | 615 | 0.7208 | 0.0994 | 0.0239 | 0.0501 | 0.1217 | 0.1084 |
0.689 | 42.0 | 630 | 0.7209 | 0.1003 | 0.0239 | 0.0505 | 0.1217 | 0.1104 |
0.6935 | 43.0 | 645 | 0.7202 | 0.1034 | 0.0236 | 0.0505 | 0.1296 | 0.1104 |
0.6927 | 44.0 | 660 | 0.7197 | 0.1000 | 0.0238 | 0.0503 | 0.1230 | 0.1084 |
0.6877 | 45.0 | 675 | 0.7193 | 0.1034 | 0.0236 | 0.0505 | 0.1296 | 0.1104 |
0.6835 | 46.0 | 690 | 0.7188 | 0.1034 | 0.0236 | 0.0505 | 0.1296 | 0.1104 |
0.6889 | 47.0 | 705 | 0.7182 | 0.1034 | 0.0236 | 0.0505 | 0.1296 | 0.1104 |
0.6913 | 48.0 | 720 | 0.7182 | 0.1034 | 0.0236 | 0.0505 | 0.1296 | 0.1104 |
0.6874 | 49.0 | 735 | 0.7178 | 0.1034 | 0.0234 | 0.0505 | 0.1296 | 0.1104 |
0.6801 | 50.0 | 750 | 0.7175 | 0.1034 | 0.0234 | 0.0505 | 0.1296 | 0.1104 |
0.704 | 51.0 | 765 | 0.7172 | 0.1034 | 0.0234 | 0.0505 | 0.1296 | 0.1104 |
0.6981 | 52.0 | 780 | 0.7169 | 0.1034 | 0.0234 | 0.0505 | 0.1296 | 0.1104 |
0.6927 | 53.0 | 795 | 0.7164 | 0.1030 | 0.0226 | 0.0494 | 0.1310 | 0.1084 |
0.6638 | 54.0 | 810 | 0.7161 | 0.1030 | 0.0226 | 0.0494 | 0.1310 | 0.1084 |
0.6859 | 55.0 | 825 | 0.7159 | 0.1030 | 0.0226 | 0.0494 | 0.1310 | 0.1084 |
0.688 | 56.0 | 840 | 0.7157 | 0.1030 | 0.0226 | 0.0494 | 0.1310 | 0.1084 |
0.6903 | 57.0 | 855 | 0.7155 | 0.0995 | 0.0229 | 0.0493 | 0.1243 | 0.1064 |
0.6784 | 58.0 | 870 | 0.7153 | 0.0995 | 0.0229 | 0.0493 | 0.1243 | 0.1064 |
0.7028 | 59.0 | 885 | 0.7152 | 0.0995 | 0.0229 | 0.0493 | 0.1243 | 0.1064 |
0.6744 | 60.0 | 900 | 0.7152 | 0.1019 | 0.0228 | 0.0494 | 0.1283 | 0.1084 |
0.693 | 61.0 | 915 | 0.7151 | 0.0985 | 0.0231 | 0.0493 | 0.1217 | 0.1064 |
0.6801 | 62.0 | 930 | 0.7151 | 0.0985 | 0.0231 | 0.0493 | 0.1217 | 0.1064 |
0.6842 | 63.0 | 945 | 0.7150 | 0.0985 | 0.0231 | 0.0493 | 0.1217 | 0.1064 |
0.674 | 64.0 | 960 | 0.7150 | 0.0985 | 0.0231 | 0.0493 | 0.1217 | 0.1064 |
Framework versions
- PEFT 0.4.0
- Transformers 4.31.0
- Pytorch 2.0.1
- Datasets 2.14.4
- Tokenizers 0.13.3