metadata

license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
  - generated_from_trainer
datasets:
  - audiofolder
metrics:
  - wer
model-index:
  - name: wav2vec2-large-mms-1b-all-lingala-ojpl
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: audiofolder
          type: audiofolder
          config: default
          split: train
          args: default
        metrics:
          - name: Wer
            type: wer
            value: 0.2697881828316611

wav2vec2-large-mms-1b-all-lingala-ojpl

This model is a fine-tuned version of facebook/mms-1b-all on the audiofolder dataset. It achieves the following results on the evaluation set:

Loss: 0.8394
Wer: 0.2698

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
train_batch_size: 1
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: 100
num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.5442	0.13	100	0.9396	0.3307
0.9882	0.27	200	0.9189	0.3389
0.5845	0.4	300	0.9322	0.3129
0.4162	0.54	400	1.0742	0.2939
0.506	0.67	500	0.9626	0.3077
0.8789	0.81	600	1.0502	0.3055
0.6166	0.94	700	0.9560	0.2984
0.4101	1.08	800	0.9520	0.2995
0.6536	1.21	900	1.1213	0.2988
0.4921	1.34	1000	1.0319	0.3010
0.856	1.48	1100	0.9514	0.3043
0.4479	1.61	1200	0.9079	0.2843
0.7249	1.75	1300	0.9612	0.2895
0.5384	1.88	1400	0.9050	0.2928
0.709	2.02	1500	0.9844	0.2735
0.6575	2.15	1600	0.9377	0.2772
0.6115	2.28	1700	0.9690	0.2876
0.3119	2.42	1800	0.9222	0.2798
0.3591	2.55	1900	0.9358	0.2783
0.3979	2.69	2000	0.9156	0.2702
0.7541	2.82	2100	0.8838	0.2761
0.81	2.96	2200	0.8460	0.2813
0.2224	3.09	2300	0.9377	0.2694
0.2338	3.23	2400	0.8870	0.2746
0.5315	3.36	2500	0.8782	0.2672
0.4045	3.49	2600	0.8811	0.2653
0.4874	3.63	2700	0.9059	0.2620
0.304	3.76	2800	0.8801	0.2690
1.4688	3.9	2900	0.8394	0.2698

Framework versions

Transformers 4.32.0.dev0
Pytorch 1.13.1+cu117
Datasets 2.13.1
Tokenizers 0.13.3