README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
model-index:
  - name: ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1
    results: []

ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1

This model is a fine-tuned version of gary109/ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5583
• Wer: 0.2403

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: 4e-06
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 50.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.4806	1.0	1154	0.5674	0.2700
0.4109	2.0	2308	0.5667	0.2740
0.4183	3.0	3462	0.5651	0.2742
0.4618	4.0	4616	0.5653	0.2675
0.4179	5.0	5770	0.5541	0.2697
0.4376	6.0	6924	0.5799	0.2691
0.4341	7.0	8078	0.5593	0.2649
0.4234	8.0	9232	0.5489	0.2652
0.3965	9.0	10386	0.5606	0.2601
0.4383	10.0	11540	0.5630	0.2607
0.4245	11.0	12694	0.5647	0.2561
0.3946	12.0	13848	0.5526	0.2635
0.3722	13.0	15002	0.5786	0.2565
0.3558	14.0	16156	0.5589	0.2638
0.392	15.0	17310	0.5581	0.2593
0.401	16.0	18464	0.5528	0.2566
0.4391	17.0	19618	0.5584	0.2554
0.4082	18.0	20772	0.5610	0.2528
0.4316	19.0	21926	0.5408	0.2507
0.3136	20.0	23080	0.5592	0.2525
0.4249	21.0	24234	0.5684	0.2560
0.4993	22.0	25388	0.5521	0.2529
0.4006	23.0	26542	0.5391	0.2540
0.3415	24.0	27696	0.5622	0.2503
0.3938	25.0	28850	0.5533	0.2541
0.4509	26.0	30004	0.5567	0.2486
0.3583	27.0	31158	0.5641	0.2523
0.4084	28.0	32312	0.5687	0.2521
0.3916	29.0	33466	0.5656	0.2495
0.4298	30.0	34620	0.5718	0.2499
0.3544	31.0	35774	0.5595	0.2445
0.3748	32.0	36928	0.5464	0.2461
0.3839	33.0	38082	0.5652	0.2475
0.3691	34.0	39236	0.5551	0.2428
0.3076	35.0	40390	0.5554	0.2463
0.3886	36.0	41544	0.5588	0.2461
0.41	37.0	42698	0.5601	0.2507
0.3595	38.0	43852	0.5645	0.2425
0.4174	39.0	45006	0.5561	0.2422
0.3801	40.0	46160	0.5554	0.2427
0.3513	41.0	47314	0.5599	0.2420
0.3739	42.0	48468	0.5669	0.2418
0.3702	43.0	49622	0.5646	0.2441
0.3444	44.0	50776	0.5658	0.2413
0.3152	45.0	51930	0.5559	0.2402
0.4	46.0	53084	0.5599	0.2429
0.4154	47.0	54238	0.5568	0.2430
0.3158	48.0	55392	0.5608	0.2428
0.3577	49.0	56546	0.5593	0.2412
0.3456	50.0	57700	0.5583	0.2403

Framework versions

• Transformers 4.21.0.dev0
• Pytorch 1.9.1+cu102
• Datasets 2.3.3.dev0
• Tokenizers 0.12.1