my_fyp_test_run

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

• Loss: 2.5756
• Accuracy: 0.2925

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 512
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	7	3.8445	0.0537
3.5752	2.0	14	3.6710	0.0675
3.3893	3.0	21	3.5345	0.0675
3.3893	4.0	28	3.4432	0.0675
2.9682	5.0	35	3.3577	0.0675
3.1131	6.0	42	3.2758	0.0675
3.1131	7.0	49	3.2017	0.0675
2.7654	8.0	56	3.1359	0.0762
2.909	9.0	63	3.0826	0.0762
2.6036	10.0	70	3.0339	0.08
2.6036	11.0	77	2.9885	0.1237
2.7703	12.0	84	2.9562	0.1363
2.7149	13.0	91	2.9275	0.1462
2.7149	14.0	98	2.8886	0.1625
2.4505	15.0	105	2.8333	0.185
2.6106	16.0	112	2.7990	0.2037
2.6106	17.0	119	2.7640	0.21
2.3455	18.0	126	2.7388	0.23
2.4896	19.0	133	2.7288	0.2313
2.2448	20.0	140	2.6880	0.24
2.2448	21.0	147	2.6840	0.2462
2.3857	22.0	154	2.6485	0.2637
2.3275	23.0	161	2.6562	0.2525
2.3275	24.0	168	2.6276	0.275
2.0947	25.0	175	2.6293	0.2787
2.2405	26.0	182	2.5991	0.285
2.2405	27.0	189	2.5928	0.2975
2.0203	28.0	196	2.5915	0.2787
2.1706	29.0	203	2.5839	0.3
1.9635	30.0	210	2.5835	0.2875
1.9635	31.0	217	2.5870	0.2838
2.1132	32.0	224	2.6006	0.2787
2.0928	33.0	231	2.5880	0.2888
2.0928	34.0	238	2.5841	0.2888
1.8937	35.0	245	2.5847	0.28
2.0696	36.0	252	2.5785	0.2863
2.0696	37.0	259	2.5786	0.2913
1.8714	38.0	266	2.5762	0.2938
2.0447	39.0	273	2.5764	0.2913
1.873	40.0	280	2.5769	0.29
1.873	41.0	287	2.5762	0.2913
2.0296	42.0	294	2.5756	0.2925
2.0317	42.96	300	2.5756	0.2925

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1