t5-small-finetuned-en-to-it-b32

This model is a fine-tuned version of t5-small on the ccmatrix dataset. It achieves the following results on the evaluation set:

• Loss: 2.1496
• Bleu: 9.6816
• Gen Len: 56.5347

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 40
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	282	2.9409	2.6764	69.2487
3.3809	2.0	564	2.8277	2.4974	87.428
3.3809	3.0	846	2.7483	2.6851	89.7887
3.1255	4.0	1128	2.6831	3.1801	85.6927
3.1255	5.0	1410	2.6293	3.6949	79.9467
2.9965	6.0	1692	2.5809	4.0149	76.852
2.9965	7.0	1974	2.5403	4.3463	74.6487
2.9002	8.0	2256	2.5033	4.838	72.6053
2.8229	9.0	2538	2.4694	5.2829	67.984
2.8229	10.0	2820	2.4421	5.4964	68.986
2.76	11.0	3102	2.4135	5.8118	66.528
2.76	12.0	3384	2.3897	6.1966	65.052
2.7051	13.0	3666	2.3667	6.452	64.2273
2.7051	14.0	3948	2.3465	6.6428	63.516
2.6568	15.0	4230	2.3265	6.9467	61.8673
2.6183	16.0	4512	2.3101	7.2029	60.7393
2.6183	17.0	4794	2.2954	7.4982	60.0327
2.5757	18.0	5076	2.2799	7.7555	59.968
2.5757	19.0	5358	2.2660	7.8406	60.0307
2.5534	20.0	5640	2.2558	8.0679	59.0793
2.5534	21.0	5922	2.2426	8.3325	58.5367
2.5159	22.0	6204	2.2324	8.3538	58.6893
2.5159	23.0	6486	2.2217	8.5867	57.7627
2.4983	24.0	6768	2.2135	8.8324	56.7367
2.4791	25.0	7050	2.2052	8.8113	57.4373
2.4791	26.0	7332	2.1981	9.0909	57.0173
2.4529	27.0	7614	2.1908	9.0056	57.802
2.4529	28.0	7896	2.1856	9.2696	56.9773
2.4395	29.0	8178	2.1780	9.2824	57.0007
2.4395	30.0	8460	2.1722	9.2106	56.9893
2.4277	31.0	8742	2.1685	9.4668	56.406
2.4181	32.0	9024	2.1646	9.4992	56.2327
2.4181	33.0	9306	2.1616	9.5054	56.3033
2.4071	34.0	9588	2.1578	9.5093	56.548
2.4071	35.0	9870	2.1554	9.5227	56.7807
2.3991	36.0	10152	2.1532	9.5762	56.756
2.3991	37.0	10434	2.1518	9.6659	56.5913
2.3955	38.0	10716	2.1506	9.7199	56.5753
2.3955	39.0	10998	2.1498	9.6715	56.558
2.3913	40.0	11280	2.1496	9.6816	56.5347

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1
• Datasets 2.5.1
• Tokenizers 0.11.0