a94f1a4aab14c6de0f35752493bfbc3d

This model is a fine-tuned version of albert/albert-large-v1 on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 0.7659
• Data Size: 0.125
• Epoch Runtime: 15.7110
• Accuracy: 0.5093
• F1 Macro: 0.3374
• Rouge1: 0.5093
• Rouge2: 0.0
• Rougel: 0.5093
• Rougelsum: 0.5081

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	Accuracy	F1 Macro	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	0.6850	0	1.0444	0.5718	0.5666	0.5729	0.0	0.5729	0.5729
No log	1	2104	0.5461	0.0078	2.5640	0.7766	0.7765	0.7766	0.0	0.7755	0.7778
No log	2	4208	0.6346	0.0156	2.9685	0.6678	0.6622	0.6678	0.0	0.6690	0.6678
0.0144	3	6312	0.9086	0.0312	4.8693	0.6944	0.6733	0.6956	0.0	0.6956	0.6956
0.7168	4	8416	0.7077	0.0625	8.6248	0.5093	0.3374	0.5093	0.0	0.5093	0.5081
0.6975	5	10520	0.7659	0.125	15.7110	0.5093	0.3374	0.5093	0.0	0.5093	0.5081

Framework versions

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