LevonHakobyan/bs_32_adapter_head_l23_const_lr_1e-3

<!-- 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. -->

bs32adapterheadl23constlr_1e-3

This model is a fine-tuned version of facebook/w2v-bert-2.0 on the commonvoice17_0 dataset. It achieves the following results on the evaluation set:

• —eval_loss: 1.8753
• —eval_wer: 1.0722
• —eval_cer: 0.2745
• —eval_runtime: 222.9507
• —evalsamplesper_second: 19.202
• —evalstepsper_second: 2.404
• —epoch: 30.7692
• —step: 5000

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
• —trainbatchsize: 32
• —evalbatchsize: 8
• —seed: 42
• —gradientaccumulationsteps: 2
• —totaltrainbatch_size: 64
• —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• —lrschedulertype: constant
• —num_epochs: 1000
• —mixedprecisiontraining: Native AMP

Framework versions

• —Transformers 4.41.2
• —Pytorch 2.3.0+cu121
• —Datasets 2.20.0
• —Tokenizers 0.19.1