roberta-Planetterp-Reviews

This model is a fine-tuned version of FacebookAI/roberta-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3381

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
1.5161	0.1333	10	1.5380
1.3822	0.2667	20	1.4684
1.5115	0.4	30	1.4511
1.31	0.5333	40	1.4885
1.5454	0.6667	50	1.4811
1.4163	0.8	60	1.5608
1.2721	0.9333	70	1.5238
1.4281	1.0667	80	1.3979
1.2657	1.2	90	1.4822
1.4347	1.3333	100	1.4189
1.5106	1.4667	110	1.3646
1.1319	1.6	120	1.1741
0.8325	1.7333	130	1.0078
0.883	1.8667	140	1.0296
0.926	2.0	150	0.9328
0.9232	2.1333	160	0.8841
0.8484	2.2667	170	1.2787
0.8104	2.4	180	0.9983
0.8624	2.5333	190	1.0868
0.8709	2.6667	200	1.0485
0.8681	2.8	210	0.9021
1.027	2.9333	220	1.0136
0.8836	3.0667	230	0.9574
0.8106	3.2	240	1.2070
1.0653	3.3333	250	1.0175
0.8685	3.4667	260	1.5487
1.1295	3.6	270	0.8569
0.8258	3.7333	280	0.7658
0.7233	3.8667	290	1.1466
0.774	4.0	300	1.3155
0.7213	4.1333	310	1.3360
1.1467	4.2667	320	0.9843
1.0224	4.4	330	0.8038
0.7136	4.5333	340	1.8204
1.1808	4.6667	350	1.1255
0.7814	4.8	360	1.2493
0.5796	4.9333	370	1.1621

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1