exp1_10partition_modelo_msl6000

This model is a fine-tuned version of vgaraujov/bart-base-spanish on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4599
• Model Preparation Time: 0.0035
• Bleu Msl: 0
• Bleu 1 Msl: 0.6392
• Bleu 2 Msl: 0.5125
• Bleu 3 Msl: 0.4224
• Bleu 4 Msl: 0.3388
• Ter Msl: 49.8480
• Bleu Asl: 0
• Bleu 1 Asl: 0.9246
• Bleu 2 Asl: 0.8881
• Bleu 3 Asl: 0.8511
• Bleu 4 Asl: 0.8099
• Ter Asl: 8.7289

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	1.4201	0.0035	0	0.5855	0.4688	0.3803	0.2617	49.7061	0	0.8418	0.7628	0.6958	0.6270	19.7203
No log	2.0	300	1.0165	0.0035	0	0.7396	0.6342	0.5464	0.4526	37.6994	0	0.8953	0.8417	0.7961	0.7461	13.1469
No log	3.0	450	0.8755	0.0035	0	0.7379	0.6574	0.5803	0.4725	35.7683	0	0.9177	0.8716	0.8293	0.7826	10.7692
1.6824	4.0	600	0.8238	0.0035	0	0.7738	0.6926	0.6179	0.5189	30.9824	0	0.9363	0.9006	0.8660	0.8267	8.2517
1.6824	5.0	750	0.8042	0.0035	0	0.7149	0.6312	0.5438	0.4332	35.1805	0	0.9374	0.9093	0.8824	0.8507	7.9720
1.6824	6.0	900	0.8480	0.0035	0	0.7469	0.6742	0.6091	0.5018	30.3107	0	0.9375	0.9025	0.8698	0.8320	7.8322
0.2441	7.0	1050	0.8531	0.0035	0	0.6312	0.5500	0.4770	0.3797	41.4777	0	0.9406	0.9063	0.8734	0.8372	7.6923
0.2441	8.0	1200	0.8495	0.0035	0	0.7302	0.6401	0.5653	0.4614	36.9437	0	0.9477	0.9165	0.8873	0.8551	6.9231
0.2441	9.0	1350	0.8623	0.0035	0	0.6935	0.6149	0.5451	0.4552	39.4626	0	0.9402	0.9086	0.8778	0.8415	7.6224
0.1022	10.0	1500	0.8533	0.0035	0	0.6700	0.6052	0.5381	0.4471	37.9513	0	0.9453	0.9139	0.8848	0.8516	6.9231
0.1022	11.0	1650	0.9126	0.0035	0	0.6731	0.5920	0.5202	0.4231	38.9589	0	0.9525	0.9238	0.8951	0.8631	6.0839
0.1022	12.0	1800	0.9068	0.0035	0	0.7141	0.6329	0.5539	0.4452	37.1956	0	0.9473	0.9165	0.8869	0.8533	6.7832
0.1022	13.0	1950	0.9214	0.0035	0	0.6425	0.5591	0.4849	0.3882	43.6608	0	0.9471	0.9164	0.8881	0.8574	6.9930
0.0478	14.0	2100	0.9292	0.0035	0	0.6367	0.5527	0.4761	0.3818	42.1495	0	0.9409	0.9071	0.8741	0.8374	7.9021
0.0478	15.0	2250	0.9934	0.0035	0	0.6377	0.5528	0.4728	0.3743	42.6532	0	0.9438	0.9115	0.8803	0.8446	7.1329
0.0478	16.0	2400	0.9738	0.0035	0	0.6457	0.5657	0.4902	0.3998	42.5693	0	0.9443	0.9124	0.8817	0.8465	7.3427
0.0274	17.0	2550	0.8840	0.0035	0	0.6418	0.5598	0.4831	0.3861	41.4777	0	0.9474	0.9185	0.8896	0.8553	6.8531
0.0274	18.0	2700	0.9120	0.0035	0	0.6428	0.5599	0.4832	0.3895	43.2410	0	0.9442	0.9110	0.8812	0.8489	7.3427
0.0274	19.0	2850	0.9104	0.0035	0	0.6790	0.6017	0.5266	0.4279	40.4702	0	0.9480	0.9174	0.8880	0.8555	6.7133
0.017	20.0	3000	0.9158	0.0035	0	0.6488	0.5616	0.4771	0.3772	40.8900	0	0.9425	0.9084	0.8769	0.8420	7.6224
0.017	21.0	3150	0.9166	0.0035	0	0.6541	0.5670	0.4891	0.3918	40.2183	0	0.9431	0.9103	0.8804	0.8472	7.4825
0.017	22.0	3300	0.8951	0.0035	0	0.6413	0.5573	0.4797	0.3779	41.3938	0	0.9427	0.9095	0.8785	0.8441	7.0629
0.017	23.0	3450	0.9018	0.0035	0	0.6557	0.5731	0.4941	0.3926	41.1419	0	0.9444	0.9122	0.8818	0.8481	7.1329
0.012	24.0	3600	0.8967	0.0035	0	0.6830	0.6007	0.5199	0.4124	37.9513	0	0.9479	0.9183	0.8900	0.8582	6.6434
0.012	25.0	3750	0.9093	0.0035	0	0.6570	0.5781	0.5005	0.3941	39.5466	0	0.9504	0.9225	0.8954	0.8646	6.3636
0.012	26.0	3900	0.8968	0.0035	0	0.6863	0.6044	0.5242	0.4228	38.3711	0	0.9474	0.9174	0.8889	0.8570	6.7133
0.0089	27.0	4050	0.8918	0.0035	0	0.6705	0.5902	0.5122	0.4065	39.7145	0	0.9498	0.9196	0.8913	0.8597	6.4336
0.0089	28.0	4200	0.8972	0.0035	0	0.6690	0.5861	0.5066	0.4014	39.7145	0	0.9480	0.9176	0.8887	0.8564	6.7133
0.0089	29.0	4350	0.9027	0.0035	0	0.6605	0.5778	0.4965	0.3926	40.8060	0	0.9474	0.9170	0.8876	0.8550	6.7133
0.0075	30.0	4500	0.8992	0.0035	0	0.6606	0.5775	0.4969	0.3932	40.3862	0	0.9480	0.9180	0.8889	0.8566	6.6434

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1