metadata

library_name: transformers
license: apache-2.0
base_model: openai/whisper-small
tags:
  - generated_from_trainer
datasets:
  - balbus-classifier
metrics:
  - accuracy
model-index:
  - name: miosipof/whisper-small-ft-balbus-sep28k-v1.5
    results:
      - task:
          name: Audio Classification
          type: audio-classification
        dataset:
          name: Apple dataset
          type: balbus-classifier
          config: default
          split: train
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8121278596051394

miosipof/whisper-small-ft-balbus-sep28k-v1.5

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

Loss: 0.1218
Accuracy: 0.8121

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: 3e-06
train_batch_size: 32
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.5
training_steps: 2000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.173	0.1253	50	0.1725	0.5682
0.1729	0.2506	100	0.1715	0.5655
0.1707	0.3759	150	0.1701	0.5680
0.1692	0.5013	200	0.1684	0.5708
0.1661	0.6266	250	0.1666	0.5762
0.1642	0.7519	300	0.1613	0.6346
0.158	0.8772	350	0.1572	0.6543
0.1522	1.0025	400	0.1475	0.6963
0.1402	1.1278	450	0.1324	0.7455
0.1258	1.2531	500	0.1230	0.7733
0.1222	1.3784	550	0.1157	0.7918
0.1091	1.5038	600	0.1112	0.8008
0.1123	1.6291	650	0.1089	0.8070
0.1107	1.7544	700	0.1177	0.7860
0.1137	1.8797	750	0.1086	0.8062
0.1061	2.0050	800	0.1063	0.8126
0.0981	2.1303	850	0.1071	0.8140
0.0957	2.2556	900	0.1097	0.8099
0.1006	2.3810	950	0.1055	0.8134
0.0974	2.5063	1000	0.1123	0.8092
0.0965	2.6316	1050	0.1078	0.8128
0.1	2.7569	1100	0.1109	0.8030
0.0985	2.8822	1150	0.1075	0.8098
0.1006	3.0075	1200	0.1058	0.8167
0.0832	3.1328	1250	0.1097	0.8151
0.0841	3.2581	1300	0.1097	0.8113
0.08	3.3835	1350	0.1104	0.8112
0.0843	3.5088	1400	0.1097	0.8139
0.0816	3.6341	1450	0.1125	0.8135
0.083	3.7594	1500	0.1097	0.8135
0.0854	3.8847	1550	0.1112	0.8170
0.0848	4.0100	1600	0.1083	0.8118
0.072	4.1353	1650	0.1161	0.8112
0.0734	4.2607	1700	0.1171	0.8126
0.0689	4.3860	1750	0.1208	0.8149
0.0682	4.5113	1800	0.1208	0.8118
0.0686	4.6366	1850	0.1215	0.8115
0.0698	4.7619	1900	0.1208	0.8120
0.0669	4.8872	1950	0.1219	0.8118
0.0698	5.0125	2000	0.1218	0.8121

Framework versions

Transformers 4.45.2
Pytorch 2.2.0
Datasets 3.2.0
Tokenizers 0.20.3