--- license: apache-2.0 tags: - generated_from_trainer base_model: jonatasgrosman/wav2vec2-large-xlsr-53-english metrics: - accuracy model-index: - name: wav2vec2-large-xlsr-53-english-ser-cosine results: [] --- # wav2vec2-large-xlsr-53-english-ser-cosine This model is a fine-tuned version of [jonatasgrosman/wav2vec2-large-xlsr-53-english](https://huggingface.co/jonatasgrosman/wav2vec2-large-xlsr-53-english) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 0.4215 - Accuracy: 0.8611 ## 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: 0.0001076429938136877 - train_batch_size: 4 - eval_batch_size: 4 - seed: 42 - gradient_accumulation_steps: 2 - total_train_batch_size: 8 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: cosine_with_restarts - lr_scheduler_warmup_steps: 18 - num_epochs: 2.0 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | 1.781 | 0.01 | 10 | 1.8028 | 0.1545 | | 1.7854 | 0.02 | 20 | 1.7883 | 0.1964 | | 1.8096 | 0.02 | 30 | 1.7266 | 0.2555 | | 1.7726 | 0.03 | 40 | 1.7654 | 0.2219 | | 1.7558 | 0.04 | 50 | 1.6892 | 0.3180 | | 1.7778 | 0.05 | 60 | 1.6563 | 0.3336 | | 1.6491 | 0.06 | 70 | 1.6236 | 0.3665 | | 1.5512 | 0.07 | 80 | 1.5289 | 0.3804 | | 1.6337 | 0.07 | 90 | 1.4650 | 0.3977 | | 1.4708 | 0.08 | 100 | 1.3707 | 0.4700 | | 1.4622 | 0.09 | 110 | 1.4187 | 0.4412 | | 1.409 | 0.1 | 120 | 1.2115 | 0.5793 | | 1.3799 | 0.11 | 130 | 1.4589 | 0.3681 | | 1.1948 | 0.12 | 140 | 1.2008 | 0.5563 | | 1.1255 | 0.12 | 150 | 1.3140 | 0.5004 | | 1.3201 | 0.13 | 160 | 1.1924 | 0.5546 | | 1.137 | 0.14 | 170 | 0.9202 | 0.6820 | | 0.9879 | 0.15 | 180 | 0.8952 | 0.6713 | | 1.0591 | 0.16 | 190 | 1.1175 | 0.6261 | | 1.0489 | 0.16 | 200 | 1.0495 | 0.6228 | | 1.145 | 0.17 | 210 | 1.0476 | 0.6048 | | 1.0471 | 0.18 | 220 | 1.0145 | 0.6360 | | 1.071 | 0.19 | 230 | 0.8197 | 0.7206 | | 1.0695 | 0.2 | 240 | 0.8922 | 0.6820 | | 0.9588 | 0.21 | 250 | 0.9974 | 0.6270 | | 0.9946 | 0.21 | 260 | 0.8327 | 0.7083 | | 0.8376 | 0.22 | 270 | 0.7972 | 0.7157 | | 0.9653 | 0.23 | 280 | 1.1024 | 0.6442 | | 0.9783 | 0.24 | 290 | 0.9703 | 0.6746 | | 1.1273 | 0.25 | 300 | 0.8766 | 0.6960 | | 1.0978 | 0.25 | 310 | 0.8021 | 0.7124 | | 0.7481 | 0.26 | 320 | 0.8639 | 0.6878 | | 0.9392 | 0.27 | 330 | 0.7483 | 0.7346 | | 0.8972 | 0.28 | 340 | 0.8086 | 0.7083 | | 0.812 | 0.29 | 350 | 0.8079 | 0.7206 | | 0.9077 | 0.3 | 360 | 1.0001 | 0.6598 | | 0.7214 | 0.3 | 370 | 0.8035 | 0.7338 | | 0.9227 | 0.31 | 380 | 0.9332 | 0.6910 | | 0.7574 | 0.32 | 390 | 0.7768 | 0.7206 | | 1.0059 | 0.33 | 400 | 0.7643 | 0.7280 | | 0.9047 | 0.34 | 410 | 0.8035 | 0.7141 | | 0.9737 | 0.35 | 420 | 0.7310 | 0.7395 | | 0.732 | 0.35 | 430 | 0.8227 | 0.7165 | | 0.9809 | 0.36 | 440 | 0.7379 | 0.7502 | | 0.9453 | 0.37 | 450 | 0.7537 | 0.7264 | | 0.7107 | 0.38 | 460 | 0.7420 | 0.7272 | | 0.7221 | 0.39 | 470 | 0.8797 | 0.7075 | | 0.7188 | 0.39 | 480 | 0.7679 | 0.7379 | | 0.8938 | 0.4 | 490 | 0.6450 | 0.7617 | | 0.7478 | 0.41 | 500 | 0.7466 | 0.7453 | | 0.685 | 0.42 | 510 | 0.8612 | 0.7058 | | 0.8602 | 0.43 | 520 | 0.6979 | 0.7568 | | 0.6247 | 0.44 | 530 | 0.6357 | 0.7740 | | 0.8188 | 0.44 | 540 | 0.7325 | 0.7379 | | 0.7733 | 0.45 | 550 | 0.6679 | 0.7568 | | 0.6555 | 0.46 | 560 | 0.6318 | 0.7707 | | 0.7855 | 0.47 | 570 | 0.6164 | 0.7675 | | 0.8602 | 0.48 | 580 | 0.7241 | 0.7535 | | 0.7176 | 0.49 | 590 | 0.6710 | 0.7642 | | 0.639 | 0.49 | 600 | 0.6418 | 0.7806 | | 0.6366 | 0.5 | 610 | 0.7135 | 0.7601 | | 0.559 | 0.51 | 620 | 0.7705 | 0.7329 | | 0.8654 | 0.52 | 630 | 0.8205 | 0.7313 | | 0.7747 | 0.53 | 640 | 0.7320 | 0.7592 | | 0.8225 | 0.53 | 650 | 0.6535 | 0.7642 | | 0.7234 | 0.54 | 660 | 0.6321 | 0.7666 | | 0.7549 | 0.55 | 670 | 0.6618 | 0.7592 | | 0.7577 | 0.56 | 680 | 0.7642 | 0.7436 | | 0.8584 | 0.57 | 690 | 0.6483 | 0.7847 | | 0.8096 | 0.58 | 700 | 0.7530 | 0.7576 | | 0.5032 | 0.58 | 710 | 0.8088 | 0.7264 | | 0.9413 | 0.59 | 720 | 0.6103 | 0.7806 | | 0.6731 | 0.6 | 730 | 0.6943 | 0.7609 | | 0.7774 | 0.61 | 740 | 0.5902 | 0.7938 | | 0.556 | 0.62 | 750 | 0.5710 | 0.7929 | | 0.609 | 0.62 | 760 | 0.6431 | 0.7757 | | 0.7012 | 0.63 | 770 | 0.6323 | 0.7798 | | 0.6209 | 0.64 | 780 | 0.6324 | 0.7847 | | 0.6434 | 0.65 | 790 | 0.7455 | 0.7560 | | 0.6942 | 0.66 | 800 | 0.7024 | 0.7650 | | 0.6962 | 0.67 | 810 | 0.5922 | 0.7995 | | 0.5535 | 0.67 | 820 | 0.6350 | 0.7839 | | 0.664 | 0.68 | 830 | 0.6658 | 0.7740 | | 0.959 | 0.69 | 840 | 0.6383 | 0.7716 | | 0.9127 | 0.7 | 850 | 0.5963 | 0.7806 | | 0.5255 | 0.71 | 860 | 0.6133 | 0.7724 | | 0.7458 | 0.72 | 870 | 0.7485 | 0.7338 | | 0.8642 | 0.72 | 880 | 0.6233 | 0.7806 | | 0.4943 | 0.73 | 890 | 0.7533 | 0.7403 | | 0.5681 | 0.74 | 900 | 0.6017 | 0.7929 | | 0.5809 | 0.75 | 910 | 0.6061 | 0.7773 | | 0.5191 | 0.76 | 920 | 0.8872 | 0.7264 | | 0.8137 | 0.76 | 930 | 0.6620 | 0.7798 | | 0.9125 | 0.77 | 940 | 0.5708 | 0.7987 | | 0.6507 | 0.78 | 950 | 0.5563 | 0.7954 | | 0.5128 | 0.79 | 960 | 0.6318 | 0.7724 | | 0.7114 | 0.8 | 970 | 0.6168 | 0.7740 | | 0.583 | 0.81 | 980 | 0.7461 | 0.7617 | | 0.7679 | 0.81 | 990 | 0.6579 | 0.7913 | | 0.8284 | 0.82 | 1000 | 0.7556 | 0.7354 | | 0.5583 | 0.83 | 1010 | 0.6527 | 0.7691 | | 0.5624 | 0.84 | 1020 | 0.5929 | 0.8069 | | 0.6102 | 0.85 | 1030 | 0.6791 | 0.7847 | | 0.5968 | 0.86 | 1040 | 0.6253 | 0.7888 | | 0.7403 | 0.86 | 1050 | 0.6318 | 0.7888 | | 0.486 | 0.87 | 1060 | 0.6332 | 0.7847 | | 0.5785 | 0.88 | 1070 | 0.6594 | 0.7707 | | 0.7037 | 0.89 | 1080 | 0.6323 | 0.7740 | | 0.5022 | 0.9 | 1090 | 0.6067 | 0.7896 | | 0.5631 | 0.9 | 1100 | 0.7094 | 0.7486 | | 0.7833 | 0.91 | 1110 | 0.5938 | 0.7921 | | 0.7214 | 0.92 | 1120 | 0.5511 | 0.7962 | | 0.7912 | 0.93 | 1130 | 0.5588 | 0.7938 | | 0.684 | 0.94 | 1140 | 0.5046 | 0.8102 | | 0.7606 | 0.95 | 1150 | 0.5403 | 0.7970 | | 0.4331 | 0.95 | 1160 | 0.5822 | 0.7872 | | 0.4767 | 0.96 | 1170 | 0.5382 | 0.8012 | | 0.4303 | 0.97 | 1180 | 0.4929 | 0.8258 | | 0.6541 | 0.98 | 1190 | 0.5382 | 0.8217 | | 0.6647 | 0.99 | 1200 | 0.5436 | 0.8143 | | 0.6649 | 1.0 | 1210 | 0.5499 | 0.7970 | | 0.4763 | 1.0 | 1220 | 0.5227 | 0.8151 | | 0.4252 | 1.01 | 1230 | 0.5697 | 0.8020 | | 0.4634 | 1.02 | 1240 | 0.5495 | 0.8127 | | 0.4511 | 1.03 | 1250 | 0.5456 | 0.8176 | | 0.3716 | 1.04 | 1260 | 0.5608 | 0.8192 | | 0.5631 | 1.04 | 1270 | 0.5308 | 0.8266 | | 0.6632 | 1.05 | 1280 | 0.5098 | 0.8332 | | 0.3734 | 1.06 | 1290 | 0.5800 | 0.8028 | | 0.4876 | 1.07 | 1300 | 0.5907 | 0.8028 | | 0.4039 | 1.08 | 1310 | 0.5270 | 0.8307 | | 0.4644 | 1.09 | 1320 | 0.5837 | 0.8176 | | 0.5113 | 1.09 | 1330 | 0.5672 | 0.8110 | | 0.3963 | 1.1 | 1340 | 0.5400 | 0.8110 | | 0.2716 | 1.11 | 1350 | 0.4932 | 0.8299 | | 0.4686 | 1.12 | 1360 | 0.5798 | 0.8184 | | 0.5707 | 1.13 | 1370 | 0.5204 | 0.8316 | | 0.3846 | 1.13 | 1380 | 0.5716 | 0.7995 | | 0.4161 | 1.14 | 1390 | 0.5645 | 0.8127 | | 0.496 | 1.15 | 1400 | 0.5593 | 0.8209 | | 0.7214 | 1.16 | 1410 | 0.5093 | 0.8250 | | 0.2983 | 1.17 | 1420 | 0.5079 | 0.8332 | | 0.6799 | 1.18 | 1430 | 0.5147 | 0.8225 | | 0.3985 | 1.18 | 1440 | 0.4981 | 0.8233 | | 0.4708 | 1.19 | 1450 | 0.5347 | 0.8151 | | 0.3934 | 1.2 | 1460 | 0.5083 | 0.8291 | | 0.3551 | 1.21 | 1470 | 0.4628 | 0.8389 | | 0.4603 | 1.22 | 1480 | 0.4809 | 0.8291 | | 0.5154 | 1.23 | 1490 | 0.4974 | 0.8324 | | 0.2846 | 1.23 | 1500 | 0.4795 | 0.8340 | | 0.3706 | 1.24 | 1510 | 0.5729 | 0.8118 | | 0.5806 | 1.25 | 1520 | 0.5514 | 0.8135 | | 0.498 | 1.26 | 1530 | 0.4836 | 0.8258 | | 0.5659 | 1.27 | 1540 | 0.4983 | 0.8324 | | 0.4405 | 1.27 | 1550 | 0.4946 | 0.8299 | | 0.2045 | 1.28 | 1560 | 0.5115 | 0.8274 | | 0.2957 | 1.29 | 1570 | 0.5368 | 0.8324 | | 0.4804 | 1.3 | 1580 | 0.5001 | 0.8414 | | 0.4293 | 1.31 | 1590 | 0.5220 | 0.8283 | | 0.5119 | 1.32 | 1600 | 0.5667 | 0.8176 | | 0.3852 | 1.32 | 1610 | 0.5936 | 0.8135 | | 0.5869 | 1.33 | 1620 | 0.5186 | 0.8266 | | 0.6688 | 1.34 | 1630 | 0.5559 | 0.8168 | | 0.3449 | 1.35 | 1640 | 0.5264 | 0.8307 | | 0.387 | 1.36 | 1650 | 0.4626 | 0.8406 | | 0.2209 | 1.37 | 1660 | 0.4919 | 0.8283 | | 0.4323 | 1.37 | 1670 | 0.4556 | 0.8463 | | 0.4921 | 1.38 | 1680 | 0.4789 | 0.8414 | | 0.3967 | 1.39 | 1690 | 0.4715 | 0.8381 | | 0.3463 | 1.4 | 1700 | 0.5216 | 0.8209 | | 0.3331 | 1.41 | 1710 | 0.5219 | 0.8168 | | 0.3396 | 1.41 | 1720 | 0.5132 | 0.8316 | | 0.1933 | 1.42 | 1730 | 0.4971 | 0.8496 | | 0.664 | 1.43 | 1740 | 0.4971 | 0.8439 | | 0.199 | 1.44 | 1750 | 0.5231 | 0.8291 | | 0.5669 | 1.45 | 1760 | 0.4746 | 0.8414 | | 0.3281 | 1.46 | 1770 | 0.4511 | 0.8447 | | 0.4047 | 1.46 | 1780 | 0.4461 | 0.8505 | | 0.5053 | 1.47 | 1790 | 0.4577 | 0.8463 | | 0.3584 | 1.48 | 1800 | 0.4407 | 0.8505 | | 0.4758 | 1.49 | 1810 | 0.4477 | 0.8488 | | 0.3109 | 1.5 | 1820 | 0.4337 | 0.8422 | | 0.3695 | 1.5 | 1830 | 0.4365 | 0.8406 | | 0.6448 | 1.51 | 1840 | 0.4425 | 0.8373 | | 0.5054 | 1.52 | 1850 | 0.4501 | 0.8291 | | 0.3598 | 1.53 | 1860 | 0.4353 | 0.8365 | | 0.3951 | 1.54 | 1870 | 0.4303 | 0.8455 | | 0.4661 | 1.55 | 1880 | 0.4463 | 0.8431 | | 0.425 | 1.55 | 1890 | 0.4427 | 0.8463 | | 0.3758 | 1.56 | 1900 | 0.4388 | 0.8463 | | 0.2416 | 1.57 | 1910 | 0.4330 | 0.8447 | | 0.5203 | 1.58 | 1920 | 0.4369 | 0.8472 | | 0.3258 | 1.59 | 1930 | 0.4340 | 0.8472 | | 0.2536 | 1.6 | 1940 | 0.4303 | 0.8513 | | 0.4079 | 1.6 | 1950 | 0.4336 | 0.8488 | | 0.5333 | 1.61 | 1960 | 0.4404 | 0.8496 | | 0.3799 | 1.62 | 1970 | 0.4395 | 0.8496 | | 0.3737 | 1.63 | 1980 | 0.4426 | 0.8472 | | 0.2297 | 1.64 | 1990 | 0.4382 | 0.8546 | | 0.534 | 1.64 | 2000 | 0.4312 | 0.8554 | | 0.4614 | 1.65 | 2010 | 0.4293 | 0.8521 | | 0.3417 | 1.66 | 2020 | 0.4249 | 0.8513 | | 0.3067 | 1.67 | 2030 | 0.4265 | 0.8537 | | 0.2893 | 1.68 | 2040 | 0.4259 | 0.8570 | | 0.4271 | 1.69 | 2050 | 0.4257 | 0.8611 | | 0.4641 | 1.69 | 2060 | 0.4192 | 0.8587 | | 0.3295 | 1.7 | 2070 | 0.4180 | 0.8578 | | 0.4281 | 1.71 | 2080 | 0.4157 | 0.8595 | | 0.3737 | 1.72 | 2090 | 0.4149 | 0.8554 | | 0.3418 | 1.73 | 2100 | 0.4207 | 0.8537 | | 0.4394 | 1.74 | 2110 | 0.4257 | 0.8554 | | 0.2029 | 1.74 | 2120 | 0.4287 | 0.8554 | | 0.3565 | 1.75 | 2130 | 0.4320 | 0.8537 | | 0.1855 | 1.76 | 2140 | 0.4277 | 0.8570 | | 0.3926 | 1.77 | 2150 | 0.4260 | 0.8546 | | 0.3638 | 1.78 | 2160 | 0.4241 | 0.8587 | | 0.2738 | 1.78 | 2170 | 0.4229 | 0.8587 | | 0.3652 | 1.79 | 2180 | 0.4212 | 0.8587 | | 0.3871 | 1.8 | 2190 | 0.4215 | 0.8595 | | 0.2194 | 1.81 | 2200 | 0.4229 | 0.8587 | | 0.2666 | 1.82 | 2210 | 0.4246 | 0.8570 | | 0.2705 | 1.83 | 2220 | 0.4251 | 0.8595 | | 0.3726 | 1.83 | 2230 | 0.4257 | 0.8603 | | 0.4248 | 1.84 | 2240 | 0.4253 | 0.8611 | | 0.4365 | 1.85 | 2250 | 0.4239 | 0.8595 | | 0.2785 | 1.86 | 2260 | 0.4230 | 0.8603 | | 0.3871 | 1.87 | 2270 | 0.4226 | 0.8611 | | 0.351 | 1.88 | 2280 | 0.4228 | 0.8603 | | 0.4279 | 1.88 | 2290 | 0.4226 | 0.8603 | | 0.4634 | 1.89 | 2300 | 0.4219 | 0.8595 | | 0.3642 | 1.9 | 2310 | 0.4212 | 0.8603 | | 0.4088 | 1.91 | 2320 | 0.4211 | 0.8595 | | 0.5017 | 1.92 | 2330 | 0.4210 | 0.8603 | | 0.3176 | 1.92 | 2340 | 0.4212 | 0.8603 | | 0.2854 | 1.93 | 2350 | 0.4214 | 0.8603 | | 0.3411 | 1.94 | 2360 | 0.4215 | 0.8603 | | 0.418 | 1.95 | 2370 | 0.4216 | 0.8603 | | 0.2308 | 1.96 | 2380 | 0.4216 | 0.8611 | | 0.4071 | 1.97 | 2390 | 0.4216 | 0.8611 | | 0.3283 | 1.97 | 2400 | 0.4215 | 0.8611 | | 0.3909 | 1.98 | 2410 | 0.4215 | 0.8611 | | 0.3095 | 1.99 | 2420 | 0.4215 | 0.8611 | | 0.3992 | 2.0 | 2430 | 0.4215 | 0.8611 | ### Framework versions - Transformers 4.40.0.dev0 - Pytorch 2.2.1+cu121 - Datasets 2.18.1.dev0 - Tokenizers 0.15.2