7590973e9961be1d27b48be2d0e5d856

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

• Loss: 0.6651
• Data Size: 1.0
• Epoch Runtime: 1076.4548
• 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	0.6809	0	31.4735	0.5719	0.4615	0.5718	0.0	0.5721	0.5716
0.5837	1	11370	0.4359	0.0078	41.2302	0.7960	0.7805	0.7959	0.0	0.7960	0.7959
0.4535	2	22740	0.4315	0.0156	48.2342	0.7959	0.7645	0.7959	0.0	0.7960	0.7958
0.4067	3	34110	0.4051	0.0312	65.1580	0.8192	0.7937	0.8191	0.0	0.8191	0.8192
0.3699	4	45480	0.3823	0.0625	97.2517	0.8451	0.8336	0.8450	0.0	0.8451	0.8450
0.6776	5	56850	0.6576	0.125	161.9829	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.6695	6	68220	0.6618	0.25	292.9002	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.6627	7	79590	0.6637	0.5	550.5814	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.6596	8.0	90960	0.6651	1.0	1076.4548	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