b1bc5c61b4f9caa1b61363f03394f260

This model is a fine-tuned version of albert/albert-base-v2 on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 0.5374
• Data Size: 1.0
• Epoch Runtime: 7.6692
• Mse: 0.5376
• Mae: 0.5436
• R2: 0.7595

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	7.1984	0	1.1631	7.1996	2.2495	-2.2207
No log	1	179	3.0938	0.0078	1.4544	3.0947	1.4763	-0.3844
No log	2	358	2.4564	0.0156	1.3761	2.4571	1.3294	-0.0992
No log	3	537	2.2086	0.0312	1.5215	2.2094	1.2730	0.0116
No log	4	716	2.2357	0.0625	1.7500	2.2364	1.2390	-0.0004
No log	5	895	0.9630	0.125	2.0595	0.9634	0.8163	0.5690
0.1082	6	1074	0.7248	0.25	2.9344	0.7250	0.6653	0.6757
1.7546	7	1253	0.6635	0.5	4.4977	0.6637	0.6366	0.7031
0.5454	8.0	1432	0.6060	1.0	7.7242	0.6063	0.6149	0.7288
0.4078	9.0	1611	0.5375	1.0	7.6844	0.5378	0.5597	0.7594
0.3	10.0	1790	0.5301	1.0	7.6488	0.5305	0.5630	0.7627
0.2412	11.0	1969	0.6130	1.0	7.6626	0.6131	0.5868	0.7257
0.223	12.0	2148	0.6040	1.0	7.5864	0.6045	0.5918	0.7296
0.1701	13.0	2327	0.5380	1.0	7.6107	0.5381	0.5461	0.7593
0.1444	14.0	2506	0.5374	1.0	7.6692	0.5376	0.5436	0.7595

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1