iva_mt_wslot-m2m100_418M-en-pl-plaintext_10e

This model is a fine-tuned version of facebook/m2m100_418M on the iva_mt_wslot dataset. It achieves the following results on the evaluation set:

• Loss: 0.0169
• Bleu: 41.3124
• Gen Len: 15.5197

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.0169	1.0	5091	0.0162	36.663	15.6444
0.0124	2.0	10182	0.0151	38.36	15.6314
0.0086	3.0	15273	0.0150	39.3808	15.5507
0.0069	4.0	20364	0.0152	39.6307	15.5235
0.0049	5.0	25455	0.0156	40.4441	15.5911
0.0038	6.0	30546	0.0159	40.3781	15.47
0.0027	7.0	35637	0.0163	40.1339	15.4722
0.0021	8.0	40728	0.0166	41.4429	15.4906
0.0016	9.0	45819	0.0168	41.1024	15.5249
0.0012	10.0	50910	0.0169	41.3124	15.5197

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3