7466476ce9dae0a5d5fbfc31c44848fc

This model is a fine-tuned version of distilbert/distilroberta-base on the nyu-mll/glue [stsb] dataset. It achieves the following results on the evaluation set:

• Loss: 0.5177
• Data Size: 1.0
• Epoch Runtime: 8.7167
• Mse: 0.5179
• Mae: 0.5445
• R2: 0.7683

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	6.8550	0	1.2124	6.8562	2.1889	-2.0670
No log	1	179	5.3811	0.0078	1.4966	5.3822	1.9187	-1.4077
No log	2	358	2.5382	0.0156	1.4979	2.5391	1.3658	-0.1358
No log	3	537	2.1043	0.0312	1.8202	2.1051	1.2390	0.0583
No log	4	716	1.4874	0.0625	2.0097	1.4878	1.0198	0.3344
No log	5	895	0.8171	0.125	2.6523	0.8173	0.7220	0.6344
0.1202	6	1074	0.8663	0.25	3.2443	0.8661	0.7130	0.6126
0.6487	7	1253	0.6056	0.5	5.1087	0.6059	0.5884	0.7290
0.4742	8.0	1432	0.5509	1.0	8.7484	0.5511	0.5760	0.7535
0.3265	9.0	1611	0.5833	1.0	8.6051	0.5834	0.5881	0.7390
0.2532	10.0	1790	0.5146	1.0	8.6088	0.5149	0.5462	0.7697
0.2134	11.0	1969	0.5347	1.0	8.6670	0.5350	0.5491	0.7607
0.1698	12.0	2148	0.4767	1.0	8.5253	0.4769	0.5197	0.7866
0.1485	13.0	2327	0.5100	1.0	8.7182	0.5102	0.5439	0.7718
0.1258	14.0	2506	0.5349	1.0	8.7539	0.5350	0.5559	0.7607
0.1026	15.0	2685	0.5021	1.0	8.7144	0.5024	0.5429	0.7753
0.0997	16.0	2864	0.5177	1.0	8.7167	0.5179	0.5445	0.7683

Framework versions

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