Modelpublic

bigmorning/whisper_input_decoder_equal_labels_no_force__0005

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face

0likes7downloads

Model Card

whisperinputdecoderequallabelsnoforce__0005

This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:

—Train Loss: 0.0018
—Train Accuracy: 0.0362
—Train Wermet: 23.1500
—Validation Loss: 0.0013
—Validation Accuracy: 0.0266
—Validation Wermet: 44.9304
—Epoch: 4

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:

—optimizer: {'name': 'AdamWeightDecay', 'learningrate': 1e-05, 'decay': 0.0, 'beta1': 0.9, 'beta2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weightdecay_rate': 0.01}
—training_precision: float32

Training results

Train Loss	Train Accuracy	Train Wermet	Validation Loss	Validation Accuracy	Validation Wermet	Epoch
0.7688	0.0332	32.0424	0.0164	0.0265	57.8712	0
0.0114	0.0362	31.1067	0.0062	0.0266	54.8371	1
0.0048	0.0362	27.0610	0.0030	0.0266	50.1428	2
0.0027	0.0362	24.9672	0.0018	0.0266	48.0741	3
0.0018	0.0362	23.1500	0.0013	0.0266	44.9304	4

Framework versions

—Transformers 4.33.0.dev0
—TensorFlow 2.13.0
—Tokenizers 0.13.3