--- language: - en license: mit tags: - generated_from_trainer datasets: - glue metrics: - accuracy model-index: - name: roberta-base-mnli results: - task: name: Text Classification type: text-classification dataset: name: GLUE MNLI type: glue args: mnli metrics: - name: Accuracy type: accuracy value: 0.8657445077298617 --- # roberta-base-mnli This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the GLUE MNLI dataset. It achieves the following results on the evaluation set: - Loss: 0.3617 - Accuracy: 0.8657 ## 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: 2e-05 - train_batch_size: 16 - 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_ratio: 0.06 - num_epochs: 10.0 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:------:|:---------------:|:--------:| | 1.0993 | 0.02 | 500 | 1.0983 | 0.3321 | | 1.099 | 0.04 | 1000 | 1.0932 | 0.4276 | | 1.011 | 0.06 | 1500 | 0.8352 | 0.6732 | | 0.7551 | 0.08 | 2000 | 0.6018 | 0.7615 | | 0.6343 | 0.1 | 2500 | 0.5726 | 0.7813 | | 0.5884 | 0.12 | 3000 | 0.5349 | 0.7926 | | 0.5548 | 0.14 | 3500 | 0.4925 | 0.8078 | | 0.5244 | 0.16 | 4000 | 0.4806 | 0.8161 | | 0.5198 | 0.18 | 4500 | 0.4614 | 0.8257 | | 0.5168 | 0.2 | 5000 | 0.4713 | 0.8177 | | 0.5194 | 0.22 | 5500 | 0.4344 | 0.8323 | | 0.485 | 0.24 | 6000 | 0.4527 | 0.8316 | | 0.4909 | 0.26 | 6500 | 0.4377 | 0.8376 | | 0.49 | 0.29 | 7000 | 0.4649 | 0.8266 | | 0.4897 | 0.31 | 7500 | 0.4162 | 0.8413 | | 0.4672 | 0.33 | 8000 | 0.4163 | 0.8425 | | 0.4699 | 0.35 | 8500 | 0.4060 | 0.8451 | | 0.4729 | 0.37 | 9000 | 0.4412 | 0.8387 | | 0.4733 | 0.39 | 9500 | 0.4353 | 0.8401 | | 0.4699 | 0.41 | 10000 | 0.4060 | 0.8476 | | 0.4759 | 0.43 | 10500 | 0.4226 | 0.8358 | | 0.461 | 0.45 | 11000 | 0.4220 | 0.8423 | | 0.4608 | 0.47 | 11500 | 0.4404 | 0.8319 | | 0.462 | 0.49 | 12000 | 0.4280 | 0.8455 | | 0.4533 | 0.51 | 12500 | 0.4128 | 0.8468 | | 0.4691 | 0.53 | 13000 | 0.4155 | 0.8437 | | 0.4552 | 0.55 | 13500 | 0.4385 | 0.8348 | | 0.4573 | 0.57 | 14000 | 0.4498 | 0.8424 | | 0.4562 | 0.59 | 14500 | 0.4162 | 0.8442 | | 0.4665 | 0.61 | 15000 | 0.4417 | 0.8432 | | 0.4569 | 0.63 | 15500 | 0.4113 | 0.8492 | | 0.4705 | 0.65 | 16000 | 0.4454 | 0.8399 | | 0.4685 | 0.67 | 16500 | 0.4055 | 0.8451 | | 0.4475 | 0.69 | 17000 | 0.4426 | 0.8383 | | 0.4641 | 0.71 | 17500 | 0.4256 | 0.8471 | | 0.4299 | 0.73 | 18000 | 0.4260 | 0.8478 | | 0.4439 | 0.75 | 18500 | 0.4218 | 0.8454 | | 0.4628 | 0.77 | 19000 | 0.4087 | 0.8479 | | 0.4502 | 0.79 | 19500 | 0.4238 | 0.8450 | | 0.4299 | 0.81 | 20000 | 0.4091 | 0.8485 | | 0.4496 | 0.84 | 20500 | 0.4160 | 0.8439 | | 0.4492 | 0.86 | 21000 | 0.4109 | 0.8469 | | 0.432 | 0.88 | 21500 | 0.4499 | 0.8493 | | 0.4343 | 0.9 | 22000 | 0.4136 | 0.8465 | | 0.4445 | 0.92 | 22500 | 0.4095 | 0.8433 | | 0.4378 | 0.94 | 23000 | 0.3999 | 0.8483 | | 0.4367 | 0.96 | 23500 | 0.3962 | 0.8509 | | 0.4428 | 0.98 | 24000 | 0.3958 | 0.8504 | | 0.4356 | 1.0 | 24500 | 0.3998 | 0.8558 | | 0.3715 | 1.02 | 25000 | 0.4016 | 0.8589 | | 0.3649 | 1.04 | 25500 | 0.4368 | 0.8582 | | 0.3565 | 1.06 | 26000 | 0.4084 | 0.8519 | | 0.3626 | 1.08 | 26500 | 0.4302 | 0.8438 | | 0.3535 | 1.1 | 27000 | 0.4206 | 0.8557 | | 0.3684 | 1.12 | 27500 | 0.4117 | 0.8561 | | 0.3649 | 1.14 | 28000 | 0.4300 | 0.8527 | | 0.3791 | 1.16 | 28500 | 0.3916 | 0.8585 | | 0.366 | 1.18 | 29000 | 0.4101 | 0.8592 | | 0.3777 | 1.2 | 29500 | 0.3946 | 0.8561 | | 0.3672 | 1.22 | 30000 | 0.4417 | 0.8530 | | 0.3688 | 1.24 | 30500 | 0.4066 | 0.8523 | | 0.3525 | 1.26 | 31000 | 0.4299 | 0.8581 | | 0.3688 | 1.28 | 31500 | 0.3870 | 0.8553 | | 0.3699 | 1.3 | 32000 | 0.3781 | 0.8627 | | 0.3547 | 1.32 | 32500 | 0.4311 | 0.8526 | | 0.3653 | 1.34 | 33000 | 0.4034 | 0.8603 | | 0.3738 | 1.36 | 33500 | 0.4103 | 0.8554 | | 0.3824 | 1.39 | 34000 | 0.3719 | 0.8618 | | 0.3591 | 1.41 | 34500 | 0.4244 | 0.8615 | | 0.3697 | 1.43 | 35000 | 0.4689 | 0.8451 | | 0.3598 | 1.45 | 35500 | 0.4149 | 0.8532 | | 0.3586 | 1.47 | 36000 | 0.4070 | 0.8591 | | 0.3519 | 1.49 | 36500 | 0.4133 | 0.8545 | | 0.3681 | 1.51 | 37000 | 0.3889 | 0.8601 | | 0.3611 | 1.53 | 37500 | 0.3934 | 0.8591 | | 0.3696 | 1.55 | 38000 | 0.4313 | 0.8552 | | 0.3798 | 1.57 | 38500 | 0.3784 | 0.8602 | | 0.3601 | 1.59 | 39000 | 0.3994 | 0.8600 | | 0.3696 | 1.61 | 39500 | 0.4206 | 0.8577 | | 0.368 | 1.63 | 40000 | 0.3903 | 0.8627 | | 0.3473 | 1.65 | 40500 | 0.3813 | 0.8655 | | 0.3604 | 1.67 | 41000 | 0.3930 | 0.8551 | | 0.3741 | 1.69 | 41500 | 0.3644 | 0.8618 | | 0.3551 | 1.71 | 42000 | 0.3936 | 0.8583 | | 0.378 | 1.73 | 42500 | 0.3826 | 0.8607 | | 0.3609 | 1.75 | 43000 | 0.3815 | 0.8618 | | 0.3678 | 1.77 | 43500 | 0.3961 | 0.8578 | | 0.3633 | 1.79 | 44000 | 0.4011 | 0.8603 | | 0.3792 | 1.81 | 44500 | 0.4061 | 0.8592 | | 0.3675 | 1.83 | 45000 | 0.4155 | 0.8631 | | 0.3576 | 1.85 | 45500 | 0.4061 | 0.8589 | | 0.3546 | 1.87 | 46000 | 0.3862 | 0.8623 | | 0.3564 | 1.89 | 46500 | 0.3937 | 0.8607 | | 0.3602 | 1.91 | 47000 | 0.3851 | 0.8646 | | 0.3494 | 1.94 | 47500 | 0.4015 | 0.8541 | | 0.3499 | 1.96 | 48000 | 0.4266 | 0.8545 | | 0.3672 | 1.98 | 48500 | 0.3761 | 0.8588 | | 0.3661 | 2.0 | 49000 | 0.4121 | 0.8567 | | 0.2759 | 2.02 | 49500 | 0.4653 | 0.8645 | | 0.2927 | 2.04 | 50000 | 0.4652 | 0.8597 | | 0.2736 | 2.06 | 50500 | 0.4547 | 0.8597 | | 0.2749 | 2.08 | 51000 | 0.4896 | 0.8565 | | 0.2757 | 2.1 | 51500 | 0.4814 | 0.8639 | | 0.2833 | 2.12 | 52000 | 0.4110 | 0.8656 | | 0.2797 | 2.14 | 52500 | 0.4316 | 0.8636 | | 0.2643 | 2.16 | 53000 | 0.4317 | 0.8599 | | 0.2791 | 2.18 | 53500 | 0.4557 | 0.8617 | | 0.2737 | 2.2 | 54000 | 0.4102 | 0.8624 | | 0.2748 | 2.22 | 54500 | 0.4187 | 0.8585 | | 0.2619 | 2.24 | 55000 | 0.4412 | 0.8590 | | 0.2718 | 2.26 | 55500 | 0.4707 | 0.8618 | | 0.2662 | 2.28 | 56000 | 0.4754 | 0.8594 | | 0.282 | 2.3 | 56500 | 0.4376 | 0.8617 | | 0.284 | 2.32 | 57000 | 0.4393 | 0.8599 | | 0.2733 | 2.34 | 57500 | 0.4531 | 0.8581 | | 0.2878 | 2.36 | 58000 | 0.4727 | 0.8549 | | 0.2812 | 2.38 | 58500 | 0.4221 | 0.8625 | | 0.2657 | 2.4 | 59000 | 0.4456 | 0.8583 | | 0.2716 | 2.42 | 59500 | 0.4455 | 0.8668 | | 0.2766 | 2.44 | 60000 | 0.4940 | 0.8580 | | 0.2871 | 2.46 | 60500 | 0.4460 | 0.8501 | | 0.2731 | 2.49 | 61000 | 0.4600 | 0.8631 | | 0.2885 | 2.51 | 61500 | 0.4229 | 0.8645 | | 0.2764 | 2.53 | 62000 | 0.4107 | 0.8638 | | 0.2866 | 2.55 | 62500 | 0.4250 | 0.8638 | | 0.2754 | 2.57 | 63000 | 0.4846 | 0.8580 | | 0.3028 | 2.59 | 63500 | 0.4339 | 0.8627 | | 0.2828 | 2.61 | 64000 | 0.4697 | 0.8613 | | 0.2875 | 2.63 | 64500 | 0.4167 | 0.8638 | | 0.2836 | 2.65 | 65000 | 0.5050 | 0.8600 | | 0.2978 | 2.67 | 65500 | 0.4139 | 0.8628 | | 0.2946 | 2.69 | 66000 | 0.4449 | 0.8644 | | 0.2822 | 2.71 | 66500 | 0.4302 | 0.8612 | | 0.3006 | 2.73 | 67000 | 0.4256 | 0.8631 | | 0.2896 | 2.75 | 67500 | 0.4993 | 0.8603 | | 0.2787 | 2.77 | 68000 | 0.4467 | 0.8636 | | 0.3 | 2.79 | 68500 | 0.4196 | 0.8592 | | 0.2939 | 2.81 | 69000 | 0.4234 | 0.8614 | | 0.2841 | 2.83 | 69500 | 0.4173 | 0.8660 | | 0.2935 | 2.85 | 70000 | 0.4054 | 0.8658 | | 0.2977 | 2.87 | 70500 | 0.4400 | 0.8623 | | 0.2853 | 2.89 | 71000 | 0.4322 | 0.8668 | | 0.2779 | 2.91 | 71500 | 0.4460 | 0.8595 | | 0.2923 | 2.93 | 72000 | 0.4279 | 0.8619 | | 0.2915 | 2.95 | 72500 | 0.4324 | 0.8625 | | 0.2927 | 2.97 | 73000 | 0.4108 | 0.8672 | | 0.29 | 2.99 | 73500 | 0.4299 | 0.8579 | | 0.2255 | 3.01 | 74000 | 0.5337 | 0.8637 | | 0.2113 | 3.04 | 74500 | 0.5046 | 0.8624 | | 0.207 | 3.06 | 75000 | 0.6011 | 0.8551 | | 0.2226 | 3.08 | 75500 | 0.5426 | 0.8579 | | 0.2129 | 3.1 | 76000 | 0.5036 | 0.8640 | | 0.2201 | 3.12 | 76500 | 0.5629 | 0.8604 | | 0.2185 | 3.14 | 77000 | 0.5416 | 0.8607 | | 0.21 | 3.16 | 77500 | 0.5457 | 0.8605 | | 0.2372 | 3.18 | 78000 | 0.5337 | 0.8594 | | 0.2237 | 3.2 | 78500 | 0.5060 | 0.8679 | | 0.2277 | 3.22 | 79000 | 0.5647 | 0.8651 | | 0.2301 | 3.24 | 79500 | 0.4906 | 0.8602 | | 0.2238 | 3.26 | 80000 | 0.5231 | 0.8647 | | 0.2365 | 3.28 | 80500 | 0.5628 | 0.8621 | | 0.2189 | 3.3 | 81000 | 0.5496 | 0.8630 | | 0.2233 | 3.32 | 81500 | 0.5418 | 0.8639 | | 0.2216 | 3.34 | 82000 | 0.5032 | 0.8689 | | 0.2314 | 3.36 | 82500 | 0.5437 | 0.8634 | | 0.2351 | 3.38 | 83000 | 0.4863 | 0.8653 | | 0.2378 | 3.4 | 83500 | 0.5158 | 0.8635 | | 0.2357 | 3.42 | 84000 | 0.5142 | 0.8629 | | 0.2484 | 3.44 | 84500 | 0.4536 | 0.8657 | | 0.2261 | 3.46 | 85000 | 0.5619 | 0.8649 | | 0.2323 | 3.48 | 85500 | 0.5371 | 0.8587 | | 0.2336 | 3.5 | 86000 | 0.5562 | 0.8621 | | 0.2259 | 3.52 | 86500 | 0.5339 | 0.8589 | | 0.2371 | 3.54 | 87000 | 0.4711 | 0.8665 | | 0.227 | 3.57 | 87500 | 0.5350 | 0.8644 | | 0.2417 | 3.59 | 88000 | 0.4692 | 0.8665 | | 0.2176 | 3.61 | 88500 | 0.5195 | 0.8655 | | 0.2393 | 3.63 | 89000 | 0.5468 | 0.8588 | | 0.2219 | 3.65 | 89500 | 0.5498 | 0.8646 | | 0.23 | 3.67 | 90000 | 0.5367 | 0.8703 | | 0.2317 | 3.69 | 90500 | 0.4761 | 0.8639 | | 0.2241 | 3.71 | 91000 | 0.4992 | 0.8654 | | 0.2327 | 3.73 | 91500 | 0.5040 | 0.8678 | | 0.2312 | 3.75 | 92000 | 0.4943 | 0.8639 | | 0.2369 | 3.77 | 92500 | 0.4824 | 0.8721 | | 0.2235 | 3.79 | 93000 | 0.5090 | 0.8661 | | 0.2256 | 3.81 | 93500 | 0.5258 | 0.8644 | | 0.236 | 3.83 | 94000 | 0.5490 | 0.8542 | | 0.2313 | 3.85 | 94500 | 0.4672 | 0.8677 | | 0.228 | 3.87 | 95000 | 0.5037 | 0.8623 | | 0.2297 | 3.89 | 95500 | 0.5207 | 0.8545 | | 0.2332 | 3.91 | 96000 | 0.5139 | 0.8698 | | 0.2331 | 3.93 | 96500 | 0.5182 | 0.8615 | | 0.2354 | 3.95 | 97000 | 0.5090 | 0.8657 | | 0.2273 | 3.97 | 97500 | 0.5523 | 0.8637 | | 0.2433 | 3.99 | 98000 | 0.5148 | 0.8691 | | 0.191 | 4.01 | 98500 | 0.6007 | 0.8654 | | 0.1683 | 4.03 | 99000 | 0.6770 | 0.8636 | | 0.1778 | 4.05 | 99500 | 0.6595 | 0.8635 | | 0.1832 | 4.07 | 100000 | 0.6129 | 0.8608 | | 0.1842 | 4.09 | 100500 | 0.6612 | 0.8611 | | 0.1865 | 4.12 | 101000 | 0.6551 | 0.8658 | | 0.1833 | 4.14 | 101500 | 0.6294 | 0.8643 | | 0.1869 | 4.16 | 102000 | 0.6234 | 0.8614 | | 0.1806 | 4.18 | 102500 | 0.6417 | 0.8655 | | 0.1911 | 4.2 | 103000 | 0.6426 | 0.8607 | | 0.1981 | 4.22 | 103500 | 0.6247 | 0.8589 | | 0.1731 | 4.24 | 104000 | 0.6613 | 0.8626 | | 0.1977 | 4.26 | 104500 | 0.5441 | 0.8661 | | 0.1771 | 4.28 | 105000 | 0.6608 | 0.8644 | | 0.1903 | 4.3 | 105500 | 0.6174 | 0.8603 | | 0.1797 | 4.32 | 106000 | 0.6609 | 0.8607 | | 0.188 | 4.34 | 106500 | 0.6059 | 0.8643 | | 0.1863 | 4.36 | 107000 | 0.5723 | 0.8663 | | 0.19 | 4.38 | 107500 | 0.5959 | 0.8652 | | 0.1869 | 4.4 | 108000 | 0.5898 | 0.8698 | | 0.1909 | 4.42 | 108500 | 0.6052 | 0.8659 | | 0.1908 | 4.44 | 109000 | 0.5854 | 0.8690 | | 0.203 | 4.46 | 109500 | 0.5727 | 0.8694 | | 0.1993 | 4.48 | 110000 | 0.5877 | 0.8653 | | 0.1796 | 4.5 | 110500 | 0.6231 | 0.8679 | | 0.1837 | 4.52 | 111000 | 0.5749 | 0.8694 | | 0.1885 | 4.54 | 111500 | 0.6174 | 0.8618 | | 0.1902 | 4.56 | 112000 | 0.5625 | 0.8682 | | 0.2031 | 4.58 | 112500 | 0.6252 | 0.8577 | | 0.1986 | 4.6 | 113000 | 0.6147 | 0.8548 | | 0.1769 | 4.62 | 113500 | 0.6351 | 0.8648 | | 0.1974 | 4.64 | 114000 | 0.6396 | 0.8630 | | 0.1952 | 4.67 | 114500 | 0.6174 | 0.8661 | | 0.1904 | 4.69 | 115000 | 0.6188 | 0.8663 | | 0.191 | 4.71 | 115500 | 0.5860 | 0.8646 | | 0.1869 | 4.73 | 116000 | 0.5978 | 0.8586 | | 0.2056 | 4.75 | 116500 | 0.5985 | 0.8648 | | 0.1837 | 4.77 | 117000 | 0.5742 | 0.8636 | | 0.2038 | 4.79 | 117500 | 0.5726 | 0.8662 | | 0.1939 | 4.81 | 118000 | 0.6097 | 0.8623 | | 0.1869 | 4.83 | 118500 | 0.5820 | 0.8651 | | 0.1897 | 4.85 | 119000 | 0.5766 | 0.8666 | | 0.1792 | 4.87 | 119500 | 0.6093 | 0.8683 | | 0.2056 | 4.89 | 120000 | 0.5890 | 0.8633 | | 0.1989 | 4.91 | 120500 | 0.5825 | 0.8674 | | 0.1916 | 4.93 | 121000 | 0.6250 | 0.8641 | | 0.197 | 4.95 | 121500 | 0.5848 | 0.8645 | | 0.1923 | 4.97 | 122000 | 0.5666 | 0.8667 | | 0.1916 | 4.99 | 122500 | 0.6189 | 0.8638 | | 0.1642 | 5.01 | 123000 | 0.7094 | 0.8610 | | 0.1357 | 5.03 | 123500 | 0.6972 | 0.8658 | | 0.1476 | 5.05 | 124000 | 0.6965 | 0.8664 | | 0.1476 | 5.07 | 124500 | 0.7177 | 0.8638 | | 0.1486 | 5.09 | 125000 | 0.6945 | 0.8620 | | 0.1309 | 5.11 | 125500 | 0.7326 | 0.8626 | | 0.1575 | 5.13 | 126000 | 0.6473 | 0.8632 | | 0.1411 | 5.15 | 126500 | 0.6955 | 0.8651 | | 0.1473 | 5.17 | 127000 | 0.6926 | 0.8648 | | 0.153 | 5.19 | 127500 | 0.7010 | 0.8638 | | 0.1488 | 5.22 | 128000 | 0.6643 | 0.8689 | | 0.144 | 5.24 | 128500 | 0.6868 | 0.8668 | | 0.156 | 5.26 | 129000 | 0.6682 | 0.8645 | | 0.1537 | 5.28 | 129500 | 0.6740 | 0.8610 | | 0.1424 | 5.3 | 130000 | 0.7509 | 0.8603 | | 0.1531 | 5.32 | 130500 | 0.6966 | 0.8670 | | 0.1457 | 5.34 | 131000 | 0.7227 | 0.8632 | | 0.1494 | 5.36 | 131500 | 0.6911 | 0.8626 | | 0.1476 | 5.38 | 132000 | 0.6903 | 0.8630 | | 0.1531 | 5.4 | 132500 | 0.6839 | 0.8675 | | 0.1613 | 5.42 | 133000 | 0.6559 | 0.8601 | | 0.1456 | 5.44 | 133500 | 0.7161 | 0.8619 | | 0.1539 | 5.46 | 134000 | 0.7108 | 0.8638 | | 0.1685 | 5.48 | 134500 | 0.6703 | 0.8628 | | 0.1482 | 5.5 | 135000 | 0.6692 | 0.8651 | | 0.1587 | 5.52 | 135500 | 0.6936 | 0.8658 | | 0.152 | 5.54 | 136000 | 0.6844 | 0.8661 | | 0.1619 | 5.56 | 136500 | 0.6632 | 0.8641 | | 0.154 | 5.58 | 137000 | 0.6451 | 0.8666 | | 0.1525 | 5.6 | 137500 | 0.6529 | 0.8686 | | 0.1545 | 5.62 | 138000 | 0.6860 | 0.8603 | | 0.1487 | 5.64 | 138500 | 0.6842 | 0.8668 | | 0.1546 | 5.66 | 139000 | 0.6692 | 0.8655 | | 0.168 | 5.68 | 139500 | 0.6701 | 0.8649 | | 0.1513 | 5.7 | 140000 | 0.6613 | 0.8680 | | 0.1704 | 5.72 | 140500 | 0.6804 | 0.8643 | | 0.1517 | 5.74 | 141000 | 0.6871 | 0.8684 | | 0.1572 | 5.77 | 141500 | 0.6676 | 0.8670 | | 0.1551 | 5.79 | 142000 | 0.6919 | 0.8638 | | 0.1483 | 5.81 | 142500 | 0.6801 | 0.8667 | | 0.1562 | 5.83 | 143000 | 0.6791 | 0.8628 | | 0.1594 | 5.85 | 143500 | 0.6422 | 0.8671 | | 0.1627 | 5.87 | 144000 | 0.6526 | 0.8679 | | 0.1514 | 5.89 | 144500 | 0.6734 | 0.8698 | | 0.1546 | 5.91 | 145000 | 0.6377 | 0.8711 | | 0.146 | 5.93 | 145500 | 0.7214 | 0.8657 | | 0.1608 | 5.95 | 146000 | 0.6756 | 0.8674 | | 0.1648 | 5.97 | 146500 | 0.6387 | 0.8687 | | 0.1547 | 5.99 | 147000 | 0.6871 | 0.8646 | | 0.1304 | 6.01 | 147500 | 0.7543 | 0.8633 | | 0.1059 | 6.03 | 148000 | 0.7576 | 0.8638 | | 0.1089 | 6.05 | 148500 | 0.7530 | 0.8642 | | 0.112 | 6.07 | 149000 | 0.7951 | 0.8640 | | 0.1198 | 6.09 | 149500 | 0.7381 | 0.8636 | | 0.1222 | 6.11 | 150000 | 0.7560 | 0.8623 | | 0.1024 | 6.13 | 150500 | 0.7965 | 0.8669 | | 0.125 | 6.15 | 151000 | 0.7613 | 0.8620 | | 0.1005 | 6.17 | 151500 | 0.7851 | 0.8651 | | 0.1196 | 6.19 | 152000 | 0.7637 | 0.8652 | | 0.1133 | 6.21 | 152500 | 0.7810 | 0.8660 | | 0.1271 | 6.23 | 153000 | 0.7510 | 0.8672 | | 0.1167 | 6.25 | 153500 | 0.7670 | 0.8638 | | 0.1198 | 6.27 | 154000 | 0.7770 | 0.8632 | | 0.1194 | 6.29 | 154500 | 0.7720 | 0.8607 | | 0.1215 | 6.32 | 155000 | 0.7880 | 0.8609 | | 0.1134 | 6.34 | 155500 | 0.8026 | 0.8617 | | 0.1113 | 6.36 | 156000 | 0.7632 | 0.8652 | | 0.1207 | 6.38 | 156500 | 0.7369 | 0.8686 | | 0.1188 | 6.4 | 157000 | 0.7466 | 0.8657 | | 0.1283 | 6.42 | 157500 | 0.7531 | 0.8645 | | 0.1186 | 6.44 | 158000 | 0.7529 | 0.8673 | | 0.135 | 6.46 | 158500 | 0.7706 | 0.8589 | | 0.1116 | 6.48 | 159000 | 0.7754 | 0.8646 | | 0.1295 | 6.5 | 159500 | 0.7026 | 0.8693 | | 0.1309 | 6.52 | 160000 | 0.7342 | 0.8656 | | 0.1172 | 6.54 | 160500 | 0.7828 | 0.8644 | | 0.125 | 6.56 | 161000 | 0.7456 | 0.8671 | | 0.1199 | 6.58 | 161500 | 0.7464 | 0.8701 | | 0.1197 | 6.6 | 162000 | 0.7626 | 0.8639 | | 0.1126 | 6.62 | 162500 | 0.8115 | 0.8609 | | 0.1365 | 6.64 | 163000 | 0.7407 | 0.8681 | | 0.122 | 6.66 | 163500 | 0.7648 | 0.8641 | | 0.1157 | 6.68 | 164000 | 0.7636 | 0.8669 | | 0.118 | 6.7 | 164500 | 0.7688 | 0.8686 | | 0.1173 | 6.72 | 165000 | 0.8051 | 0.8687 | | 0.1137 | 6.74 | 165500 | 0.8101 | 0.8635 | | 0.1412 | 6.76 | 166000 | 0.7004 | 0.8689 | | 0.1131 | 6.78 | 166500 | 0.7589 | 0.8664 | | 0.1232 | 6.8 | 167000 | 0.7657 | 0.8654 | | 0.1343 | 6.82 | 167500 | 0.7547 | 0.8652 | | 0.1208 | 6.84 | 168000 | 0.7407 | 0.8699 | | 0.1284 | 6.87 | 168500 | 0.7182 | 0.8677 | | 0.1182 | 6.89 | 169000 | 0.7248 | 0.8681 | | 0.1166 | 6.91 | 169500 | 0.7385 | 0.8678 | | 0.1289 | 6.93 | 170000 | 0.7293 | 0.8672 | | 0.1243 | 6.95 | 170500 | 0.7178 | 0.8696 | | 0.1256 | 6.97 | 171000 | 0.7291 | 0.8633 | | 0.1162 | 6.99 | 171500 | 0.7515 | 0.8648 | | 0.1013 | 7.01 | 172000 | 0.7824 | 0.8655 | | 0.0811 | 7.03 | 172500 | 0.8297 | 0.8647 | | 0.0831 | 7.05 | 173000 | 0.8144 | 0.8678 | | 0.0872 | 7.07 | 173500 | 0.8176 | 0.8679 | | 0.0868 | 7.09 | 174000 | 0.8405 | 0.8642 | | 0.0756 | 7.11 | 174500 | 0.8867 | 0.8642 | | 0.0882 | 7.13 | 175000 | 0.8185 | 0.8659 | | 0.0879 | 7.15 | 175500 | 0.8653 | 0.8625 | | 0.0831 | 7.17 | 176000 | 0.8323 | 0.8655 | | 0.0847 | 7.19 | 176500 | 0.8358 | 0.8650 | | 0.0938 | 7.21 | 177000 | 0.7967 | 0.8665 | | 0.0908 | 7.23 | 177500 | 0.8147 | 0.8640 | | 0.0809 | 7.25 | 178000 | 0.8325 | 0.8679 | | 0.0993 | 7.27 | 178500 | 0.8131 | 0.8655 | | 0.087 | 7.29 | 179000 | 0.8249 | 0.8628 | | 0.0873 | 7.31 | 179500 | 0.8326 | 0.8661 | | 0.0889 | 7.33 | 180000 | 0.8171 | 0.8685 | | 0.0739 | 7.35 | 180500 | 0.8686 | 0.8642 | | 0.0821 | 7.37 | 181000 | 0.8739 | 0.8669 | | 0.0981 | 7.39 | 181500 | 0.8558 | 0.8639 | | 0.0858 | 7.42 | 182000 | 0.8276 | 0.8673 | | 0.083 | 7.44 | 182500 | 0.8148 | 0.8675 | | 0.0969 | 7.46 | 183000 | 0.8520 | 0.8630 | | 0.0851 | 7.48 | 183500 | 0.8604 | 0.8671 | | 0.0881 | 7.5 | 184000 | 0.8665 | 0.8634 | | 0.1036 | 7.52 | 184500 | 0.8233 | 0.8642 | | 0.0874 | 7.54 | 185000 | 0.8293 | 0.8660 | | 0.0935 | 7.56 | 185500 | 0.8006 | 0.8671 | | 0.0887 | 7.58 | 186000 | 0.8352 | 0.8637 | | 0.0897 | 7.6 | 186500 | 0.8309 | 0.8655 | | 0.0788 | 7.62 | 187000 | 0.8505 | 0.8653 | | 0.0887 | 7.64 | 187500 | 0.8465 | 0.8657 | | 0.0909 | 7.66 | 188000 | 0.8582 | 0.8637 | | 0.0895 | 7.68 | 188500 | 0.8487 | 0.8659 | | 0.0729 | 7.7 | 189000 | 0.8770 | 0.8636 | | 0.0758 | 7.72 | 189500 | 0.8717 | 0.8653 | | 0.0901 | 7.74 | 190000 | 0.8513 | 0.8639 | | 0.0848 | 7.76 | 190500 | 0.8554 | 0.8661 | | 0.0985 | 7.78 | 191000 | 0.8259 | 0.8640 | | 0.091 | 7.8 | 191500 | 0.8483 | 0.8644 | | 0.0868 | 7.82 | 192000 | 0.8776 | 0.8602 | | 0.0898 | 7.84 | 192500 | 0.8470 | 0.8634 | | 0.0959 | 7.86 | 193000 | 0.8344 | 0.8645 | | 0.0939 | 7.88 | 193500 | 0.8419 | 0.8641 | | 0.0769 | 7.9 | 194000 | 0.8355 | 0.8673 | | 0.0808 | 7.92 | 194500 | 0.8642 | 0.8646 | | 0.0797 | 7.94 | 195000 | 0.8401 | 0.8663 | | 0.0875 | 7.97 | 195500 | 0.8598 | 0.8638 | | 0.0896 | 7.99 | 196000 | 0.8624 | 0.8648 | | 0.0762 | 8.01 | 196500 | 0.8645 | 0.8656 | | 0.0552 | 8.03 | 197000 | 0.8844 | 0.8661 | | 0.0598 | 8.05 | 197500 | 0.8870 | 0.8663 | | 0.0528 | 8.07 | 198000 | 0.8866 | 0.8679 | | 0.0679 | 8.09 | 198500 | 0.8835 | 0.8657 | | 0.0628 | 8.11 | 199000 | 0.9017 | 0.8635 | | 0.0644 | 8.13 | 199500 | 0.8979 | 0.8647 | | 0.0446 | 8.15 | 200000 | 0.9144 | 0.8656 | | 0.0524 | 8.17 | 200500 | 0.9116 | 0.8651 | | 0.0561 | 8.19 | 201000 | 0.9281 | 0.8639 | | 0.0525 | 8.21 | 201500 | 0.9115 | 0.8672 | | 0.0646 | 8.23 | 202000 | 0.8933 | 0.8663 | | 0.0691 | 8.25 | 202500 | 0.8591 | 0.8662 | | 0.0708 | 8.27 | 203000 | 0.8525 | 0.8683 | | 0.0598 | 8.29 | 203500 | 0.8663 | 0.8689 | | 0.0513 | 8.31 | 204000 | 0.8671 | 0.8704 | | 0.0564 | 8.33 | 204500 | 0.8597 | 0.8694 | | 0.0619 | 8.35 | 205000 | 0.8645 | 0.8683 | | 0.0563 | 8.37 | 205500 | 0.8848 | 0.8658 | | 0.0615 | 8.39 | 206000 | 0.8728 | 0.8663 | | 0.0668 | 8.41 | 206500 | 0.8925 | 0.8657 | | 0.0592 | 8.43 | 207000 | 0.8644 | 0.8673 | | 0.0668 | 8.45 | 207500 | 0.8601 | 0.8700 | | 0.071 | 8.47 | 208000 | 0.8735 | 0.8682 | | 0.061 | 8.49 | 208500 | 0.8797 | 0.8662 | | 0.0627 | 8.52 | 209000 | 0.8742 | 0.8663 | | 0.0505 | 8.54 | 209500 | 0.9063 | 0.8649 | | 0.0607 | 8.56 | 210000 | 0.8940 | 0.8677 | | 0.0569 | 8.58 | 210500 | 0.8953 | 0.8673 | | 0.0671 | 8.6 | 211000 | 0.8784 | 0.8667 | | 0.0509 | 8.62 | 211500 | 0.8942 | 0.8678 | | 0.0526 | 8.64 | 212000 | 0.8968 | 0.8686 | | 0.0541 | 8.66 | 212500 | 0.8950 | 0.8694 | | 0.0677 | 8.68 | 213000 | 0.8808 | 0.8665 | | 0.0552 | 8.7 | 213500 | 0.8923 | 0.8662 | | 0.053 | 8.72 | 214000 | 0.9118 | 0.8673 | | 0.0608 | 8.74 | 214500 | 0.9023 | 0.8700 | | 0.0573 | 8.76 | 215000 | 0.9096 | 0.8681 | | 0.0621 | 8.78 | 215500 | 0.8872 | 0.8684 | | 0.0559 | 8.8 | 216000 | 0.8837 | 0.8672 | | 0.0593 | 8.82 | 216500 | 0.8937 | 0.8675 | | 0.0633 | 8.84 | 217000 | 0.8746 | 0.8685 | | 0.0548 | 8.86 | 217500 | 0.9049 | 0.8662 | | 0.0427 | 8.88 | 218000 | 0.9195 | 0.8685 | | 0.0623 | 8.9 | 218500 | 0.9146 | 0.8669 | | 0.0594 | 8.92 | 219000 | 0.9096 | 0.8672 | | 0.0683 | 8.94 | 219500 | 0.8778 | 0.8679 | | 0.0659 | 8.96 | 220000 | 0.8552 | 0.8699 | | 0.0603 | 8.98 | 220500 | 0.8901 | 0.8679 | | 0.0566 | 9.0 | 221000 | 0.8997 | 0.8677 | | 0.0443 | 9.02 | 221500 | 0.9009 | 0.8683 | | 0.0358 | 9.04 | 222000 | 0.9193 | 0.8680 | | 0.0317 | 9.07 | 222500 | 0.9319 | 0.8687 | | 0.0384 | 9.09 | 223000 | 0.9155 | 0.8699 | | 0.0432 | 9.11 | 223500 | 0.9243 | 0.8685 | | 0.0408 | 9.13 | 224000 | 0.9251 | 0.8693 | | 0.0443 | 9.15 | 224500 | 0.9322 | 0.8677 | | 0.0438 | 9.17 | 225000 | 0.9371 | 0.8666 | | 0.0379 | 9.19 | 225500 | 0.9283 | 0.8693 | | 0.0411 | 9.21 | 226000 | 0.9147 | 0.8703 | | 0.036 | 9.23 | 226500 | 0.9167 | 0.8703 | | 0.0394 | 9.25 | 227000 | 0.9254 | 0.8688 | | 0.0363 | 9.27 | 227500 | 0.9288 | 0.8704 | | 0.0492 | 9.29 | 228000 | 0.9242 | 0.8693 | | 0.0411 | 9.31 | 228500 | 0.9325 | 0.8677 | | 0.0408 | 9.33 | 229000 | 0.9370 | 0.8690 | | 0.0326 | 9.35 | 229500 | 0.9417 | 0.8705 | | 0.038 | 9.37 | 230000 | 0.9480 | 0.8700 | | 0.0412 | 9.39 | 230500 | 0.9398 | 0.8693 | | 0.0588 | 9.41 | 231000 | 0.9174 | 0.8707 | | 0.0417 | 9.43 | 231500 | 0.9204 | 0.8715 | | 0.0362 | 9.45 | 232000 | 0.9319 | 0.8701 | | 0.0283 | 9.47 | 232500 | 0.9562 | 0.8696 | | 0.0353 | 9.49 | 233000 | 0.9525 | 0.8690 | | 0.0384 | 9.51 | 233500 | 0.9561 | 0.8687 | | 0.0406 | 9.53 | 234000 | 0.9375 | 0.8715 | | 0.0356 | 9.55 | 234500 | 0.9575 | 0.8690 | | 0.044 | 9.57 | 235000 | 0.9429 | 0.8708 | | 0.0444 | 9.6 | 235500 | 0.9413 | 0.8690 | | 0.0421 | 9.62 | 236000 | 0.9412 | 0.8689 | | 0.038 | 9.64 | 236500 | 0.9352 | 0.8695 | | 0.0355 | 9.66 | 237000 | 0.9362 | 0.8689 | | 0.04 | 9.68 | 237500 | 0.9403 | 0.8691 | | 0.0356 | 9.7 | 238000 | 0.9402 | 0.8706 | | 0.0383 | 9.72 | 238500 | 0.9466 | 0.8692 | | 0.0534 | 9.74 | 239000 | 0.9378 | 0.8700 | | 0.0383 | 9.76 | 239500 | 0.9390 | 0.8697 | | 0.0418 | 9.78 | 240000 | 0.9404 | 0.8694 | | 0.0335 | 9.8 | 240500 | 0.9390 | 0.8705 | | 0.0398 | 9.82 | 241000 | 0.9430 | 0.8696 | | 0.0336 | 9.84 | 241500 | 0.9438 | 0.8698 | | 0.045 | 9.86 | 242000 | 0.9414 | 0.8703 | | 0.0401 | 9.88 | 242500 | 0.9425 | 0.8696 | | 0.0454 | 9.9 | 243000 | 0.9405 | 0.8696 | | 0.0361 | 9.92 | 243500 | 0.9394 | 0.8696 | | 0.0458 | 9.94 | 244000 | 0.9400 | 0.8690 | | 0.0329 | 9.96 | 244500 | 0.9402 | 0.8693 | | 0.0469 | 9.98 | 245000 | 0.9401 | 0.8691 | ### Framework versions - Transformers 4.21.3 - Pytorch 1.7.1 - Datasets 1.18.3 - Tokenizers 0.11.6