pakawadeep/mt5-small-finetuned-ctfl-backtranslation_7k

This model is a fine-tuned version of google/mt5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 4.4260
• Validation Loss: 4.1643
• Train Bleu: 0.0
• Train Gen Len: 127.0
• Epoch: 29

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', 'learning_rate': 2e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}
• training_precision: float32

Training results

Train Loss	Validation Loss	Train Bleu	Train Gen Len	Epoch
18.4580	9.0813	0.0013	3.0	0
10.6168	8.0760	0.0008	50.0	1
8.9972	7.4303	0.0003	127.0	2
8.2615	7.1119	0.0010	3.0	3
7.8727	6.9450	0.0	2.0	4
7.6314	6.8080	0.0	2.0	5
7.4404	6.6029	0.0	2.0	6
7.2126	6.2358	0.0003	127.0	7
6.9077	5.6784	0.0003	127.0	8
6.5300	5.3165	0.0003	127.0	9
6.2332	5.0961	0.0003	127.0	10
5.9571	4.9619	0.0003	127.0	11
5.7344	4.8588	0.0	7.0	12
5.5347	4.7700	0.0	127.0	13
5.4059	4.7055	0.0005	127.0	14
5.2839	4.6485	0.0004	127.0	15
5.1769	4.5958	0.0	127.0	16
5.0864	4.5455	0.0	127.0	17
5.0025	4.5005	0.0	127.0	18
4.9161	4.4589	0.6412	127.0	19
4.8523	4.4206	0.8846	127.0	20
4.7860	4.3832	0.0	10.0	21
4.7340	4.3479	0.0	10.0	22
4.6707	4.3200	0.0	10.0	23
4.6181	4.2951	0.0	10.0	24
4.5780	4.2621	0.0	14.0	25
4.5309	4.2380	0.0	127.0	26
4.4944	4.2097	5.1292	29.0	27
4.4613	4.1865	0.0	127.0	28
4.4260	4.1643	0.0	127.0	29

Framework versions

• Transformers 4.44.2
• TensorFlow 2.17.0
• Datasets 3.0.0
• Tokenizers 0.19.1