speecht5_finetuned_2500

This model is a fine-tuned version of microsoft/speecht5_tts on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7095

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: 0.0001
• train_batch_size: 4
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• training_steps: 3500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.584	88.8889	100	0.6268
0.4839	177.7778	200	0.5803
0.4304	266.6667	300	0.6589
0.3976	355.5556	400	0.6201
0.3594	444.4444	500	0.6107
0.3468	533.3333	600	0.6492
0.3353	622.2222	700	0.6362
0.3194	711.1111	800	0.6373
0.3065	800.0	900	0.6431
0.2985	888.8889	1000	0.6736
0.289	977.7778	1100	0.6697
0.2833	1066.6667	1200	0.7000
0.2798	1155.5556	1300	0.6875
0.2775	1244.4444	1400	0.7149
0.275	1333.3333	1500	0.6630
0.2722	1422.2222	1600	0.7117
0.2622	1511.1111	1700	0.6647
0.2481	1600.0	1800	0.6985
0.2474	1688.8889	1900	0.7020
0.2511	1777.7778	2000	0.6847
0.2359	1866.6667	2100	0.6940
0.2336	1955.5556	2200	0.6996
0.2338	2044.4444	2300	0.7134
0.23	2133.3333	2400	0.7031
0.2305	2222.2222	2500	0.6993
0.2273	2311.1111	2600	0.7099
0.2229	2400.0	2700	0.6907
0.2151	2488.8889	2800	0.6985
0.2232	2577.7778	2900	0.6972
0.2234	2666.6667	3000	0.7091
0.2165	2755.5556	3100	0.7103
0.2143	2844.4444	3200	0.7168
0.22	2933.3333	3300	0.6877
0.2225	3022.2222	3400	0.7050
0.2192	3111.1111	3500	0.7095

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1