wav2vec2-large-960h-lv60-self-paper

hts98/wav2vec2-large-960h-lv60-self-paper

wav2vec2-large-960h-lv60-self-paper is a automatic speech recognition model from hts98. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

24% of all-time downloads

All-time downloads

Public

Repo size

64.4 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin2.5 GB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: apache-2.0
Model type: wav2vec2
Access: Public
Created: Jul 1, 2023
Updated: Jul 2, 2023
SHA: ecdcc23f

Model card

From the Hugging Face model README

wav2vec2-large-960h-lv60-self-paper

This model is a fine-tuned version of facebook/wav2vec2-large-960h-lv60-self on the HTS98/ORIGINAL_VER1.2 - NA dataset. It achieves the following results on the evaluation set:

Loss: 0.6788
Wer: 0.3320

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: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 420
num_epochs: 50.0

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	1.0	419	3.3473	1.0
5.8068	2.0	838	1.9191	0.8917
2.5663	3.0	1257	1.1006	0.5802
1.1433	4.0	1676	0.9009	0.4814
0.8522	5.0	2095	0.8215	0.4247
0.7256	6.0	2514	0.7522	0.3922
0.7256	7.0	2933	0.7202	0.3654
0.6239	8.0	3352	0.6909	0.3579
0.5618	9.0	3771	0.6887	0.3400
0.4998	10.0	4190	0.6788	0.3320
0.4569	11.0	4609	0.6805	0.3351
0.4156	12.0	5028	0.6910	0.3253
0.4156	13.0	5447	0.6859	0.3279
0.3763	14.0	5866	0.7075	0.3207
0.3473	15.0	6285	0.7174	0.3152
0.3141	16.0	6704	0.7284	0.3171
0.2884	17.0	7123	0.7537	0.3192
0.2771	18.0	7542	0.7312	0.3175
0.2771	19.0	7961	0.7669	0.3138
0.2538	20.0	8380	0.8143	0.3074
0.2319	21.0	8799	0.8185	0.3088
0.2206	22.0	9218	0.8111	0.3069
0.2093	23.0	9637	0.8248	0.3088
0.1979	24.0	10056	0.8572	0.3067
0.1979	25.0	10475	0.8710	0.3074
0.1852	26.0	10894	0.8922	0.3067
0.1742	27.0	11313	0.9040	0.3068
0.1688	28.0	11732	0.9144	0.3016
0.1578	29.0	12151	0.8990	0.3109
0.1557	30.0	12570	0.9465	0.3004
0.1557	31.0	12989	0.9480	0.3025
0.1456	32.0	13408	0.9731	0.3017
0.1398	33.0	13827	0.9633	0.3038
0.1343	34.0	14246	0.9844	0.3011
0.1275	35.0	14665	1.0078	0.2997
0.1266	36.0	15084	1.0066	0.2996
0.1243	37.0	15503	1.0133	0.3014
0.1243	38.0	15922	1.0387	0.2972
0.1182	39.0	16341	1.0173	0.3026
0.1152	40.0	16760	1.0527	0.2977
0.1134	41.0	17179	1.0491	0.2978
0.1101	42.0	17598	1.0662	0.2976
0.1083	43.0	18017	1.0544	0.2979
0.1083	44.0	18436	1.0599	0.2957
0.1073	45.0	18855	1.0767	0.2959
0.1045	46.0	19274	1.0773	0.2959
0.1024	47.0	19693	1.0731	0.2953
0.1015	48.0	20112	1.0823	0.2966
0.1016	49.0	20531	1.0885	0.2945
0.1016	50.0	20950	1.0854	0.2950

Framework versions

Transformers 4.31.0.dev0
Pytorch 2.0.0+cu117
Datasets 2.7.0
Tokenizers 0.13.2

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
License: apache-2.0
Created: Jul 1, 2023
Updated: Jul 2, 2023