simonycl's picture
update model card README.md
e75f028
|
raw
history blame
10.6 kB
metadata
license: mit
base_model: roberta-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: best_model-yelp_polarity-64-21
    results: []

best_model-yelp_polarity-64-21

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7237
  • Accuracy: 0.9375

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: 1e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 150

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 4 0.4953 0.9531
No log 2.0 8 0.4901 0.9609
0.4161 3.0 12 0.4944 0.9609
0.4161 4.0 16 0.5364 0.9531
0.3982 5.0 20 0.5743 0.9531
0.3982 6.0 24 0.5888 0.9531
0.3982 7.0 28 0.5943 0.9531
0.271 8.0 32 0.5953 0.9531
0.271 9.0 36 0.5948 0.9531
0.3643 10.0 40 0.5942 0.9531
0.3643 11.0 44 0.5936 0.9531
0.3643 12.0 48 0.5918 0.9531
0.2103 13.0 52 0.5912 0.9531
0.2103 14.0 56 0.5900 0.9531
0.1932 15.0 60 0.5847 0.9531
0.1932 16.0 64 0.5810 0.9531
0.1932 17.0 68 0.5774 0.9531
0.1372 18.0 72 0.5731 0.9531
0.1372 19.0 76 0.5691 0.9531
0.0774 20.0 80 0.5697 0.9531
0.0774 21.0 84 0.5627 0.9531
0.0774 22.0 88 0.5599 0.9531
0.0831 23.0 92 0.5587 0.9531
0.0831 24.0 96 0.5821 0.9453
0.0236 25.0 100 0.5533 0.9531
0.0236 26.0 104 0.5497 0.9531
0.0236 27.0 108 0.5459 0.9531
0.0245 28.0 112 0.5447 0.9531
0.0245 29.0 116 0.5385 0.9531
0.0123 30.0 120 0.5433 0.9453
0.0123 31.0 124 0.5401 0.9453
0.0123 32.0 128 0.5369 0.9453
0.0 33.0 132 0.5347 0.9453
0.0 34.0 136 0.5363 0.9453
0.0001 35.0 140 0.5268 0.9531
0.0001 36.0 144 0.5327 0.9531
0.0001 37.0 148 0.5355 0.9531
0.0 38.0 152 0.5369 0.9531
0.0 39.0 156 0.5374 0.9531
0.0 40.0 160 0.5374 0.9531
0.0 41.0 164 0.5372 0.9531
0.0 42.0 168 0.5366 0.9531
0.0 43.0 172 0.5345 0.9531
0.0 44.0 176 0.5323 0.9531
0.0 45.0 180 0.5295 0.9453
0.0 46.0 184 0.5441 0.9453
0.0 47.0 188 0.5519 0.9453
0.0 48.0 192 0.5562 0.9453
0.0 49.0 196 0.5588 0.9453
0.0 50.0 200 0.5607 0.9453
0.0 51.0 204 0.5622 0.9453
0.0 52.0 208 0.5632 0.9453
0.0 53.0 212 0.5640 0.9453
0.0 54.0 216 0.5660 0.9453
0.0001 55.0 220 0.5577 0.9531
0.0001 56.0 224 0.6090 0.9453
0.0001 57.0 228 0.5699 0.9453
0.0 58.0 232 0.5844 0.9453
0.0 59.0 236 0.6061 0.9375
0.0318 60.0 240 0.5903 0.9453
0.0318 61.0 244 0.5835 0.9453
0.0318 62.0 248 0.5701 0.9453
0.0 63.0 252 0.5625 0.9531
0.0 64.0 256 0.5609 0.9531
0.0 65.0 260 0.5609 0.9531
0.0 66.0 264 0.5975 0.9375
0.0 67.0 268 0.6321 0.9297
0.0194 68.0 272 0.6293 0.9375
0.0194 69.0 276 0.6356 0.9297
0.0134 70.0 280 0.5923 0.9453
0.0134 71.0 284 0.5733 0.9453
0.0134 72.0 288 0.5553 0.9531
0.0 73.0 292 0.5595 0.9453
0.0 74.0 296 0.5778 0.9453
0.0001 75.0 300 0.6930 0.9297
0.0001 76.0 304 0.6281 0.9375
0.0001 77.0 308 0.6218 0.9375
0.0018 78.0 312 0.5614 0.9453
0.0018 79.0 316 0.5087 0.9531
0.0206 80.0 320 0.4872 0.9531
0.0206 81.0 324 0.4978 0.9531
0.0206 82.0 328 0.5067 0.9531
0.0 83.0 332 0.5116 0.9531
0.0 84.0 336 0.5143 0.9531
0.0 85.0 340 0.5159 0.9531
0.0 86.0 344 0.5175 0.9531
0.0 87.0 348 0.5206 0.9531
0.0 88.0 352 0.5255 0.9453
0.0 89.0 356 0.5319 0.9453
0.0 90.0 360 0.5390 0.9375
0.0 91.0 364 0.5455 0.9375
0.0 92.0 368 0.5516 0.9375
0.0 93.0 372 0.5572 0.9375
0.0 94.0 376 0.5623 0.9375
0.0 95.0 380 0.5664 0.9375
0.0 96.0 384 0.5692 0.9375
0.0 97.0 388 0.5712 0.9375
0.0 98.0 392 0.5734 0.9375
0.0 99.0 396 0.5754 0.9375
0.0 100.0 400 0.5765 0.9375
0.0 101.0 404 0.5815 0.9375
0.0 102.0 408 0.5821 0.9375
0.0 103.0 412 0.5819 0.9375
0.0 104.0 416 0.5818 0.9375
0.0 105.0 420 0.5805 0.9375
0.0 106.0 424 0.5984 0.9375
0.0 107.0 428 0.5581 0.9453
0.0231 108.0 432 0.5229 0.9531
0.0231 109.0 436 0.4868 0.9453
0.0 110.0 440 0.5184 0.9531
0.0 111.0 444 0.5554 0.9453
0.0 112.0 448 0.7197 0.9375
0.0001 113.0 452 0.7466 0.9375
0.0001 114.0 456 0.7533 0.9375
0.0 115.0 460 0.7535 0.9375
0.0 116.0 464 0.7472 0.9375
0.0 117.0 468 0.7407 0.9375
0.0 118.0 472 0.7366 0.9375
0.0 119.0 476 0.7347 0.9297
0.0 120.0 480 0.7338 0.9297
0.0 121.0 484 0.7366 0.9297
0.0 122.0 488 0.7408 0.9297
0.0 123.0 492 0.7434 0.9297
0.0 124.0 496 0.7454 0.9297
0.0073 125.0 500 0.6336 0.9453
0.0073 126.0 504 0.5907 0.9453
0.0073 127.0 508 0.6316 0.9453
0.0023 128.0 512 0.6673 0.9453
0.0023 129.0 516 0.6764 0.9453
0.0 130.0 520 0.6814 0.9453
0.0 131.0 524 0.6917 0.9375
0.0 132.0 528 0.7031 0.9375
0.0 133.0 532 0.7111 0.9375
0.0 134.0 536 0.7161 0.9375
0.0 135.0 540 0.7153 0.9375
0.0 136.0 544 0.7137 0.9375
0.0 137.0 548 0.7130 0.9375
0.0 138.0 552 0.7126 0.9375
0.0 139.0 556 0.7126 0.9375
0.0 140.0 560 0.7127 0.9375
0.0 141.0 564 0.7154 0.9375
0.0 142.0 568 0.7190 0.9375
0.0 143.0 572 0.7211 0.9375
0.0 144.0 576 0.7223 0.9375
0.0 145.0 580 0.7230 0.9375
0.0 146.0 584 0.7233 0.9375
0.0 147.0 588 0.7235 0.9375
0.0 148.0 592 0.7236 0.9375
0.0 149.0 596 0.7237 0.9375
0.0 150.0 600 0.7237 0.9375

Framework versions

  • Transformers 4.32.0.dev0
  • Pytorch 2.0.1+cu118
  • Datasets 2.4.0
  • Tokenizers 0.13.3