9f67f2622605b6750ccf82aafdd27955

This model is a fine-tuned version of FacebookAI/xlm-roberta-large-finetuned-conll02-spanish on the nyu-mll/glue [stsb] dataset. It achieves the following results on the evaluation set:

• Loss: 2.3875
• Data Size: 1.0
• Epoch Runtime: 33.1412
• Mse: 2.3882
• Mae: 1.2996
• R2: -0.0683

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Mse	Mae	R2
No log	0	0	6.1194	0	2.6683	6.1206	2.0568	-1.7379
No log	1	179	2.4027	0.0078	3.2865	2.4035	1.3175	-0.0752
No log	2	358	10.6271	0.0156	3.8779	10.6274	2.8973	-3.7540
No log	3	537	2.4305	0.0312	5.0272	2.4312	1.2965	-0.0876
No log	4	716	4.3795	0.0625	6.7523	4.3806	1.7340	-0.9596
No log	5	895	2.3035	0.125	9.1773	2.3043	1.2887	-0.0308
0.1567	6	1074	2.3875	0.25	13.0983	2.3882	1.2996	-0.0683
2.2	7	1253	2.2899	0.5	20.6249	2.2907	1.2867	-0.0247
2.2016	8.0	1432	2.3880	1.0	35.8589	2.3888	1.2996	-0.0686
2.2284	9.0	1611	2.5214	1.0	33.8148	2.5221	1.3040	-0.1282
2.2482	10.0	1790	2.2826	1.0	33.5937	2.2834	1.2866	-0.0215
2.1318	11.0	1969	2.3875	1.0	33.2996	2.3882	1.2996	-0.0683
2.0982	12.0	2148	2.6257	1.0	34.1862	2.6264	1.3354	-0.1749
2.1861	13.0	2327	2.2349	1.0	33.8635	2.2357	1.2888	-0.0001
2.2375	14.0	2506	2.2545	1.0	33.4278	2.2553	1.2865	-0.0089
2.2157	15.0	2685	2.7077	1.0	33.6873	2.7084	1.3509	-0.2116
2.2243	16.0	2864	2.3325	1.0	33.8763	2.3332	1.2928	-0.0437
2.1185	17.0	3043	2.3875	1.0	33.1412	2.3882	1.2996	-0.0683

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1