CoolFace
Modelpublic

lgris/wav2vec2-xls-r-pt-cv7-from-bp400h

sourceHugging Faceapache-2.0updated 5y agoView on Hugging Face
0likes52downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

wav2vec2-xls-r-pt-cv7-from-bp400h

This model is a fine-tuned version of lgris/bp_400h_xlsr2_300M on the common_voice dataset. It achieves the following results on the evaluation set:

  • —Loss: 0.1535
  • —Wer: 0.1254

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: 0.0001
  • —trainbatchsize: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —gradientaccumulationsteps: 2
  • —totaltrainbatch_size: 16
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 100
  • —training_steps: 5000

Training results

Training LossEpochStepValidation LossWer
0.49910.131000.17740.1464
0.46550.262000.18840.1568
0.46890.393000.22820.1672
0.46620.524000.19970.1584
0.45920.655000.19890.1663
0.45330.786000.20040.1698
0.43910.917000.18880.1642
0.46551.048000.19210.1624
0.41381.179000.19500.1602
0.3741.310000.20770.1658
0.40641.4311000.19450.1596
0.39221.5612000.20690.1665
0.42261.6913000.19620.1573
0.39741.8214000.19190.1553
0.36311.9515000.18540.1573
0.37972.0816000.19020.1550
0.32872.2117000.19260.1598
0.35682.3418000.18880.1534
0.34152.4719000.18340.1502
0.35452.620000.19060.1560
0.33442.7321000.18040.1524
0.33082.8622000.17410.1485
0.3442.9923000.17870.1455
0.3093.1224000.17730.1448
0.3123.2525000.17380.1440
0.30663.3826000.17270.1417
0.29993.5127000.16920.1436
0.29853.6428000.17320.1430
0.30583.7729000.17540.1402
0.29433.930000.16910.1379
0.28134.0331000.17540.1376
0.27334.1632000.16390.1363
0.25924.2933000.16750.1349
0.26974.4234000.16180.1360
0.25384.5535000.16580.1348
0.27464.6736000.16740.1325
0.26554.837000.16550.1319
0.27454.9338000.16650.1316
0.26175.0639000.16000.1311
0.26745.1940000.16230.1311
0.2375.3241000.15910.1315
0.26695.4542000.15840.1295
0.24765.5843000.15720.1285
0.24455.7144000.15800.1271
0.22075.8445000.15670.1269
0.22895.9746000.15360.1260
0.24386.147000.15300.1260
0.2276.2348000.15440.1249
0.22566.3649000.15430.1254
0.21846.4950000.15350.1254

Framework versions

  • —Transformers 4.15.0
  • —Pytorch 1.10.0+cu111
  • —Datasets 1.18.0
  • —Tokenizers 0.10.3