apac_5sents_XLS-R_2

slplab/apac_5sents_XLS-R_2

apac_5sents_XLS-R_2 is a automatic speech recognition model from slplab. 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

1% of all-time downloads

All-time downloads

298

Public

Repo size

101 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin1.3 GB · 100%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: apache-2.0
Model type: wav2vec2
Access: Public
Created: Nov 8, 2023
Updated: Nov 9, 2023
SHA: e530de1c

Model card

From the Hugging Face model README

apac_5sents_XLS-R_2

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the None dataset. It achieves the following results on the evaluation set:

Loss: 274.3813
Wer: 1.0

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: 1e-06
train_batch_size: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 128
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
training_steps: 5000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
281.5958	5.54	100	273.7941	1.0
281.366	11.11	200	269.9106	1.0
273.1232	16.65	300	263.4907	1.0
261.9693	22.22	400	249.4823	1.0
219.9941	27.76	500	176.5986	1.0
147.673	33.32	600	122.9458	1.0
114.1416	38.86	700	101.6788	1.0
99.6196	44.43	800	91.0054	1.0
90.5221	49.97	900	84.6282	1.0
85.6017	55.54	1000	80.2764	1.0
81.5146	61.11	1100	77.0657	1.0
77.7573	66.65	1200	74.5696	1.0
76.1933	72.22	1300	72.5473	1.0
73.6424	77.76	1400	70.8357	1.0
72.4248	83.32	1500	69.3329	1.0
70.3529	88.86	1600	67.9791	1.0
69.4685	94.43	1700	66.7548	1.0
68.0717	99.97	1800	65.6130	1.0
67.2376	105.54	1900	64.5488	1.0
66.0277	111.11	2000	63.5492	1.0
64.7994	116.65	2100	62.6000	1.0
64.0574	122.22	2200	61.7027	1.0
62.9425	127.76	2300	60.8545	1.0
62.254	133.32	2400	60.0502	1.0
61.3336	138.86	2500	59.2789	1.0
60.7559	144.43	2600	58.5486	1.0
59.7298	149.97	2700	57.8531	1.0
59.4008	155.54	2800	57.1906	1.0
58.6672	161.11	2900	56.5612	1.0
57.6835	166.65	3000	55.9709	1.0
57.4515	172.22	3100	55.4099	1.0
56.7266	177.76	3200	54.8743	1.0
56.3382	183.32	3300	54.3800	1.0
55.4538	188.86	3400	53.9105	1.0
55.4566	194.43	3500	53.4741	1.0
54.6733	199.97	3600	53.0646	1.0
54.5231	205.54	3700	52.6835	1.0
54.1944	211.11	3800	52.3376	1.0
53.5359	216.65	3900	52.0120	1.0
53.4527	222.22	4000	51.7225	1.0
53.0497	227.76	4100	51.4579	1.0
52.9911	233.32	4200	51.2190	1.0
52.3869	238.86	4300	51.0154	1.0
52.4158	244.43	4400	50.8332	1.0
52.1746	249.97	4500	50.6797	1.0
52.2056	255.54	4600	50.5553	1.0
52.1142	261.11	4700	50.4548	1.0
51.7802	266.65	4800	50.3846	1.0
51.9224	272.22	4900	50.3413	1.0
51.7253	277.76	5000	50.3258	1.0

Framework versions

Transformers 4.26.1
Pytorch 2.0.1+cu118
Datasets 2.11.0
Tokenizers 0.13.3

Task: Automatic Speech Recognition
Library: transformers
Type: wav2vec2
License: apache-2.0
Created: Nov 8, 2023
Updated: Nov 9, 2023