Hugging Face's logo

cportoca
/
MMS_Quechua_finetuned

Automatic Speech Recognition
Transformers
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
MMS_Quechua_finetuned / README.md
cportoca's picture
cportoca
End of training
f81d7b5 verified
preview code
|
raw
history blame contribute delete
10.7 kB
metadata
library_name: transformers
license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: MMS_Quechua_finetuned
    results: []

MMS_Quechua_finetuned

This model is a fine-tuned version of facebook/mms-1b-all on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2531
  • Wer: 0.3172

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: 0.001
  • train_batch_size: 16
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 20
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
10.4014 0.1355 100 4.9727 0.9997
2.7516 0.2710 200 0.6059 0.5024
0.6736 0.4065 300 0.5783 0.4747
0.5917 0.5420 400 0.5185 0.4412
0.5776 0.6775 500 0.4926 0.4249
1.0538 0.8130 600 0.5035 0.4255
0.5216 0.9485 700 0.4767 0.4167
0.6692 1.0840 800 0.4524 0.4274
0.5159 1.2195 900 0.4474 0.4035
0.631 1.3550 1000 0.4456 0.4117
0.6377 1.4905 1100 0.4422 0.4104
0.7765 1.6260 1200 0.4571 0.4082
0.4849 1.7615 1300 0.4563 0.4026
0.4695 1.8970 1400 0.4385 0.4004
0.6693 2.0325 1500 0.4209 0.3928
0.6357 2.1680 1600 0.4203 0.3966
0.4671 2.3035 1700 0.4201 0.3994
0.4672 2.4390 1800 0.4208 0.4038
0.7265 2.5745 1900 0.4195 0.4098
0.4802 2.7100 2000 0.3781 0.3828
0.6319 2.8455 2100 0.3727 0.3844
0.5765 2.9810 2200 0.3976 0.3853
0.5579 3.1165 2300 0.3601 0.3850
0.4431 3.2520 2400 0.3513 0.3881
0.6378 3.3875 2500 0.3406 0.3693
0.6201 3.5230 2600 0.3366 0.3725
0.5285 3.6585 2700 0.3390 0.3731
0.573 3.7940 2800 0.3563 0.3750
0.3998 3.9295 2900 0.4177 0.3731
0.6257 4.0650 3000 0.3899 0.3800
0.5346 4.2005 3100 0.3567 0.3803
0.5731 4.3360 3200 0.3671 0.3866
0.5217 4.4715 3300 0.3429 0.3768
0.4091 4.6070 3400 0.3363 0.3822
0.6284 4.7425 3500 0.3727 0.3794
0.3642 4.8780 3600 0.3273 0.3834
0.4199 5.0136 3700 0.3299 0.3765
0.361 5.1491 3800 0.3164 0.3539
0.5072 5.2846 3900 0.3255 0.3640
0.584 5.4201 4000 0.3168 0.3681
0.7192 5.5556 4100 0.3266 0.3586
0.4023 5.6911 4200 0.3279 0.3765
0.3849 5.8266 4300 0.3274 0.3533
0.5499 5.9621 4400 0.3182 0.3546
0.505 6.0976 4500 0.3199 0.3590
0.3689 6.2331 4600 0.3168 0.3411
0.4963 6.3686 4700 0.3228 0.3455
0.4904 6.5041 4800 0.3248 0.3634
0.3871 6.6396 4900 0.3128 0.3555
0.5636 6.7751 5000 0.3129 0.3552
0.525 6.9106 5100 0.3089 0.3608
0.5762 7.0461 5200 0.3170 0.3527
0.3613 7.1816 5300 0.3156 0.3602
0.4433 7.3171 5400 0.3015 0.3612
0.3692 7.4526 5500 0.3228 0.3608
0.6615 7.5881 5600 0.3052 0.3561
0.4931 7.7236 5700 0.3039 0.3458
0.3608 7.8591 5800 0.3075 0.3464
0.4666 7.9946 5900 0.3047 0.3583
0.3236 8.1301 6000 0.3117 0.3574
0.6959 8.2656 6100 0.3431 0.3499
0.3459 8.4011 6200 0.3075 0.3517
0.4103 8.5366 6300 0.2924 0.3408
0.424 8.6721 6400 0.3148 0.3511
0.3373 8.8076 6500 0.3104 0.3473
0.4517 8.9431 6600 0.3218 0.3546
0.4533 9.0786 6700 0.3196 0.3514
0.4015 9.2141 6800 0.3088 0.3583
0.336 9.3496 6900 0.2927 0.3370
0.5446 9.4851 7000 0.2840 0.3430
0.4258 9.6206 7100 0.3002 0.3430
0.3432 9.7561 7200 0.2911 0.3486
0.3131 9.8916 7300 0.2907 0.3323
0.5729 10.0271 7400 0.2942 0.3326
0.3266 10.1626 7500 0.2914 0.3401
0.3512 10.2981 7600 0.2956 0.3414
0.6843 10.4336 7700 0.2840 0.3392
0.3667 10.5691 7800 0.2857 0.3348
0.3088 10.7046 7900 0.2888 0.3351
0.3679 10.8401 8000 0.2896 0.3361
0.319 10.9756 8100 0.2768 0.3320
0.3045 11.1111 8200 0.2810 0.3348
0.3169 11.2466 8300 0.2813 0.3307
0.3837 11.3821 8400 0.2831 0.3251
0.3687 11.5176 8500 0.2864 0.3351
0.322 11.6531 8600 0.2831 0.3216
0.565 11.7886 8700 0.2776 0.3348
0.363 11.9241 8800 0.2738 0.3270
0.3281 12.0596 8900 0.2785 0.3244
0.3626 12.1951 9000 0.2773 0.3414
0.3201 12.3306 9100 0.2748 0.3222
0.2993 12.4661 9200 0.2833 0.3251
0.5219 12.6016 9300 0.2936 0.3323
0.3078 12.7371 9400 0.2801 0.3329
0.3282 12.8726 9500 0.2890 0.3298
0.3013 13.0081 9600 0.2807 0.3285
0.2689 13.1436 9700 0.3006 0.3389
0.3119 13.2791 9800 0.2885 0.3310
0.3178 13.4146 9900 0.2816 0.3279
0.5885 13.5501 10000 0.2699 0.3188
0.3134 13.6856 10100 0.2857 0.3213
0.3355 13.8211 10200 0.2729 0.3175
0.296 13.9566 10300 0.2732 0.3229
0.3573 14.0921 10400 0.2699 0.3345
0.478 14.2276 10500 0.2692 0.3188
0.3013 14.3631 10600 0.2636 0.3179
0.2978 14.4986 10700 0.2641 0.3175
0.2753 14.6341 10800 0.2697 0.3169
0.3017 14.7696 10900 0.2688 0.3179
0.2897 14.9051 11000 0.2662 0.3135
0.2861 15.0407 11100 0.2650 0.3201
0.2752 15.1762 11200 0.2582 0.3153
0.2908 15.3117 11300 0.2645 0.3219
0.286 15.4472 11400 0.2647 0.3147
0.2828 15.5827 11500 0.2633 0.3169
0.4632 15.7182 11600 0.2628 0.3207
0.2994 15.8537 11700 0.2595 0.3160
0.3075 15.9892 11800 0.2616 0.3201
0.267 16.1247 11900 0.2628 0.3207
0.2825 16.2602 12000 0.2593 0.3191
0.2684 16.3957 12100 0.2554 0.3175
0.4811 16.5312 12200 0.2554 0.3298
0.2904 16.6667 12300 0.2574 0.3160
0.2781 16.8022 12400 0.2612 0.3166
0.2667 16.9377 12500 0.2597 0.3191
0.2945 17.0732 12600 0.2584 0.3150
0.2697 17.2087 12700 0.2546 0.3125
0.2726 17.3442 12800 0.2548 0.3141
0.2679 17.4797 12900 0.2586 0.3119
0.2762 17.6152 13000 0.2588 0.3131
0.2713 17.7507 13100 0.2563 0.3125
0.4666 17.8862 13200 0.2540 0.3125
0.2568 18.0217 13300 0.2613 0.3131
0.4632 18.1572 13400 0.2566 0.3182
0.2926 18.2927 13500 0.2553 0.3166
0.2743 18.4282 13600 0.2535 0.3166
0.2677 18.5637 13700 0.2566 0.3128
0.2763 18.6992 13800 0.2537 0.3125
0.2581 18.8347 13900 0.2550 0.3144
0.2476 18.9702 14000 0.2543 0.3131
0.254 19.1057 14100 0.2548 0.3131
0.2591 19.2412 14200 0.2558 0.3144
0.2728 19.3767 14300 0.2534 0.3147
0.2856 19.5122 14400 0.2523 0.3144
0.2596 19.6477 14500 0.2510 0.3119
0.4273 19.7832 14600 0.2521 0.3153
0.2559 19.9187 14700 0.2531 0.3172

Framework versions

  • Transformers 4.46.3
  • Pytorch 2.4.1+cu121
  • Datasets 3.1.0
  • Tokenizers 0.20.3