dyula-french-translation

This model is a fine-tuned version of google-t5/t5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2685
• Bleu: 1.0635

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss
1.2724	1.0	505	0.3235
0.3418	2.0	1010	0.3107
0.3266	3.0	1515	0.3046
0.3213	4.0	2020	0.3003
0.3139	5.0	2525	0.2969
0.3114	6.0	3030	0.2941
0.3068	7.0	3535	0.2917
0.3034	8.0	4040	0.2896
0.3009	9.0	4545	0.2878
0.2962	10.0	5050	0.2863
0.2964	11.0	5555	0.2847
0.2923	12.0	6060	0.2835
0.2899	13.0	6565	0.2823
0.29	14.0	7070	0.2813
0.2874	15.0	7575	0.2803
0.2831	16.0	8080	0.2792
0.2843	17.0	8585	0.2785
0.2824	18.0	9090	0.2778
0.2805	19.0	9595	0.2769
0.2786	20.0	10100	0.2760
0.2782	21.0	10605	0.2754
0.2761	22.0	11110	0.2749
0.2749	23.0	11615	0.2744
0.2744	24.0	12120	0.2737
0.2743	25.0	12625	0.2735
0.271	26.0	13130	0.2729
0.2713	27.0	13635	0.2723
0.2702	28.0	14140	0.2719
0.2697	29.0	14645	0.2718
0.2679	30.0	15150	0.2713
0.267	31.0	15655	0.2709
0.2688	32.0	16160	0.2706
0.2658	33.0	16665	0.2703
0.2647	34.0	17170	0.2702
0.264	35.0	17675	0.2698
0.265	36.0	18180	0.2696
0.2659	37.0	18685	0.2694
0.2638	38.0	19190	0.2694
0.2637	39.0	19695	0.2691
0.2629	40.0	20200	0.2692
0.2611	41.0	20705	0.2689
0.2625	42.0	21210	0.2689
0.2601	43.0	21715	0.2688
0.2633	44.0	22220	0.2687
0.261	45.0	22725	0.2687
0.2603	46.0	23230	0.2686
0.2605	47.0	23735	0.2686
0.2628	48.0	24240	0.2685
0.26	49.0	24745	0.2685
0.2602	50.0	25250	0.2685

Framework versions

• Transformers 4.42.3
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1