e98b626ba5b1b0dc9b084c7c20f72584

This model is a fine-tuned version of google-bert/bert-large-uncased-whole-word-masking on the nyu-mll/glue [qqp] dataset. It achieves the following results on the evaluation set:

• Loss: 0.6660
• Data Size: 0.25
• Epoch Runtime: 294.1163
• Accuracy: 0.6320
• F1 Macro: 0.3872
• Rouge1: 0.6318
• Rouge2: 0.0
• Rougel: 0.6319
• Rougelsum: 0.6317

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	0.6727	0	31.3530	0.6320	0.3876	0.6318	0.0	0.6320	0.6318
0.638	1	11370	0.4364	0.0078	40.2418	0.7936	0.7827	0.7936	0.0	0.7936	0.7936
0.4353	2	22740	0.4083	0.0156	48.5550	0.8179	0.8006	0.8179	0.0	0.8179	0.8179
0.4062	3	34110	0.4983	0.0312	65.1620	0.8030	0.7662	0.8029	0.0	0.8029	0.8031
0.3759	4	45480	0.4365	0.0625	98.1457	0.8402	0.8214	0.8401	0.0	0.8402	0.8403
0.6697	5	56850	0.6598	0.125	161.4430	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.6608	6	68220	0.6660	0.25	294.1163	0.6320	0.3872	0.6318	0.0	0.6319	0.6317

Framework versions

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