exp4_10partition_modelo6000

This model is a fine-tuned version of Helsinki-NLP/opus-mt-es-es on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8540
• Model Preparation Time: 0.0034
• Bleu Msl: 0
• Bleu 1 Msl: 0.7743
• Bleu 2 Msl: 0.6603
• Bleu 3 Msl: 0.5062
• Bleu 4 Msl: 0.3243
• Ter Msl: 31.0669
• Bleu Asl: 0
• Bleu 1 Asl: 0.9697
• Bleu 2 Asl: 0.9518
• Bleu 3 Asl: 0.9303
• Bleu 4 Asl: 0.9050
• Ter Asl: 3.3960

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: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Bleu Msl	Bleu 1 Msl	Bleu 2 Msl	Bleu 3 Msl	Bleu 4 Msl	Ter Msl	Bleu Asl	Bleu 1 Asl	Bleu 2 Asl	Bleu 3 Asl	Bleu 4 Asl	Ter Asl
No log	1.0	150	0.9614	0.0034	0	0.0575	0.0404	0.0273	0.0150	413.5300	0	0.5393	0.4921	0.4484	0.3962	36.1891
No log	2.0	300	0.7098	0.0034	0	0.1656	0.1236	0.0875	0.0535	155.2391	0	0.9490	0.9150	0.8798	0.8420	6.1300
No log	3.0	450	0.7208	0.0034	0	0.7392	0.6461	0.5257	0.3665	33.0621	0	0.9504	0.9186	0.8840	0.8450	5.9823
0.4901	4.0	600	0.7311	0.0034	0	0.7283	0.6380	0.5091	0.3382	33.2655	0	0.9563	0.9269	0.8954	0.8598	5.3176
0.4901	5.0	750	0.6964	0.0034	0	0.7530	0.6468	0.5295	0.3664	33.1638	0	0.9543	0.9221	0.8876	0.8509	5.3914
0.4901	6.0	900	0.6565	0.0034	0	0.7484	0.6664	0.5516	0.3933	31.1292	0	0.9595	0.9318	0.9018	0.8675	4.8006
0.0617	7.0	1050	0.7010	0.0034	0	0.7308	0.6316	0.5135	0.3597	36.5209	0	0.9587	0.9296	0.8986	0.8636	5.0960
0.0617	8.0	1200	0.6944	0.0034	0	0.7498	0.6633	0.5413	0.3752	30.6205	0	0.9608	0.9335	0.9028	0.8668	4.6529
0.0617	9.0	1350	0.6936	0.0034	0	0.7514	0.6582	0.5402	0.3830	33.3672	0	0.9582	0.9311	0.9012	0.8686	5.0222
0.0301	10.0	1500	0.7215	0.0034	0	0.7439	0.6563	0.5385	0.3807	32.4517	0	0.9626	0.9359	0.9069	0.8753	4.5052
0.0301	11.0	1650	0.7079	0.0034	0	0.7598	0.6823	0.5774	0.4262	30.0102	0	0.9594	0.9316	0.9004	0.8657	4.9483
0.0301	12.0	1800	0.7150	0.0034	0	0.7616	0.6783	0.5604	0.4030	30.6205	0	0.9557	0.9268	0.8959	0.8614	5.2437
0.0301	13.0	1950	0.6878	0.0034	0	0.7558	0.6760	0.5588	0.3985	31.0275	0	0.9562	0.9263	0.8931	0.8577	5.3914
0.0185	14.0	2100	0.6908	0.0034	0	0.7564	0.6823	0.5656	0.4076	30.3154	0	0.9620	0.9349	0.9039	0.8694	4.5790
0.0185	15.0	2250	0.7289	0.0034	0	0.7666	0.6871	0.5777	0.4227	30.2136	0	0.9627	0.9361	0.9064	0.8734	4.5052
0.0185	16.0	2400	0.7149	0.0034	0	0.7602	0.6877	0.5749	0.4198	29.3998	0	0.9607	0.9334	0.9039	0.8707	4.6529
0.0118	17.0	2550	0.7480	0.0034	0	0.7504	0.6748	0.5651	0.4074	30.0102	0	0.9602	0.9325	0.9039	0.8714	4.6529
0.0118	18.0	2700	0.7143	0.0034	0	0.7706	0.6901	0.5786	0.4223	29.8067	0	0.9607	0.9335	0.9048	0.8722	4.6529
0.0118	19.0	2850	0.7106	0.0034	0	0.7609	0.6820	0.5660	0.4065	30.4171	0	0.9621	0.9371	0.9108	0.8816	4.4313
0.0091	20.0	3000	0.7359	0.0034	0	0.7469	0.6675	0.5524	0.3941	31.3327	0	0.9608	0.9344	0.9058	0.8742	4.5790
0.0091	21.0	3150	0.7470	0.0034	0	0.7387	0.6599	0.5459	0.3924	31.0275	0	0.9613	0.9345	0.9049	0.8717	4.6529
0.0091	22.0	3300	0.7382	0.0034	0	0.7455	0.6702	0.5544	0.3947	30.1119	0	0.9607	0.9333	0.9042	0.8723	4.7267
0.0091	23.0	3450	0.7153	0.0034	0	0.7442	0.6742	0.5593	0.4001	30.0102	0	0.9626	0.9368	0.9089	0.8783	4.4313
0.0067	24.0	3600	0.7121	0.0034	0	0.7486	0.6714	0.5575	0.4013	30.7223	0	0.9639	0.9382	0.9104	0.8801	4.3575
0.0067	25.0	3750	0.7364	0.0034	0	0.7464	0.6708	0.5578	0.4019	30.0102	0	0.9645	0.9393	0.9118	0.8819	4.2836
0.0067	26.0	3900	0.7256	0.0034	0	0.7516	0.6781	0.5660	0.4088	29.6033	0	0.9639	0.9387	0.9112	0.8812	4.3575
0.0054	27.0	4050	0.7313	0.0034	0	0.7474	0.6720	0.5599	0.4011	30.0102	0	0.9639	0.9383	0.9106	0.8804	4.3575
0.0054	28.0	4200	0.7272	0.0034	0	0.7529	0.6780	0.5654	0.4063	29.7050	0	0.9639	0.9387	0.9112	0.8812	4.3575
0.0054	29.0	4350	0.7279	0.0034	0	0.7528	0.6789	0.5663	0.4066	29.5015	0	0.9627	0.9364	0.9083	0.8777	4.5052
0.0045	30.0	4500	0.7309	0.0034	0	0.7530	0.6746	0.5601	0.4003	29.9084	0	0.9627	0.9364	0.9083	0.8777	4.5052

Framework versions

• Transformers 4.50.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1