eaeb83f3b277bc2b3512d8b157a97deb

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

• Loss: 0.3939
• Data Size: 1.0
• Epoch Runtime: 339.9759
• Accuracy: 0.8825
• F1 Macro: 0.8737
• Rouge1: 0.8825
• Rouge2: 0.0
• Rougel: 0.8826
• Rougelsum: 0.8825

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	0.6694	0	11.1204	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.5897	1	11370	0.5071	0.0078	14.0044	0.7475	0.7248	0.7474	0.0	0.7475	0.7473
0.4661	2	22740	0.4249	0.0156	16.6747	0.7921	0.7747	0.7921	0.0	0.7922	0.7921
0.4266	3	34110	0.4053	0.0312	22.6602	0.8089	0.7945	0.8089	0.0	0.8088	0.8089
0.3957	4	45480	0.3796	0.0625	32.3547	0.8294	0.8173	0.8295	0.0	0.8293	0.8295
0.3596	5	56850	0.3435	0.125	53.8764	0.8427	0.8328	0.8427	0.0	0.8427	0.8428
0.3351	6	68220	0.3206	0.25	96.2496	0.8551	0.8471	0.8550	0.0	0.8552	0.8551
0.2893	7	79590	0.3031	0.5	184.7497	0.8669	0.8590	0.8669	0.0	0.8670	0.8669
0.2831	8.0	90960	0.2863	1.0	350.6842	0.8813	0.8728	0.8812	0.0	0.8814	0.8813
0.2214	9.0	102330	0.2780	1.0	348.1720	0.8863	0.8792	0.8863	0.0	0.8864	0.8863
0.1789	10.0	113700	0.3044	1.0	342.0439	0.8850	0.8773	0.8849	0.0	0.8851	0.8850
0.1555	11.0	125070	0.3529	1.0	343.5793	0.8871	0.8786	0.8870	0.0	0.8872	0.8870
0.1425	12.0	136440	0.3639	1.0	344.7638	0.8856	0.8775	0.8856	0.0	0.8857	0.8855
0.1427	13.0	147810	0.3939	1.0	339.9759	0.8825	0.8737	0.8825	0.0	0.8826	0.8825

Framework versions

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