d471ee1639b8a836375033fecfa1609a

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

• Loss: 0.3385
• Data Size: 1.0
• Epoch Runtime: 939.8961
• Accuracy: 0.8725
• F1 Macro: 0.8656
• Rouge1: 0.8726
• Rouge2: 0.0
• Rougel: 0.8725
• Rougelsum: 0.8726

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	0.6633	0	29.2838	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.6523	1	11370	0.5581	0.0078	38.1012	0.7229	0.6767	0.7228	0.0	0.7229	0.7227
0.5185	2	22740	0.4794	0.0156	43.6436	0.7653	0.7528	0.7653	0.0	0.7654	0.7654
0.4541	3	34110	0.4153	0.0312	57.9550	0.8107	0.7899	0.8106	0.0	0.8108	0.8107
0.4154	4	45480	0.3803	0.0625	86.4563	0.8323	0.8173	0.8324	0.0	0.8324	0.8323
0.3653	5	56850	0.3528	0.125	143.3803	0.8377	0.8294	0.8376	0.0	0.8377	0.8378
0.3464	6	68220	0.3257	0.25	258.8942	0.8552	0.8481	0.8553	0.0	0.8553	0.8553
0.3336	7	79590	0.3303	0.5	483.1143	0.8493	0.8444	0.8494	0.0	0.8494	0.8494
0.3412	8.0	90960	0.3026	1.0	941.8888	0.8703	0.8609	0.8703	0.0	0.8703	0.8703
0.3038	9.0	102330	0.3066	1.0	941.1943	0.8706	0.8614	0.8706	0.0	0.8706	0.8707
0.3419	10.0	113700	0.3367	1.0	939.7151	0.8644	0.8549	0.8642	0.0	0.8644	0.8644
0.6408	11.0	125070	0.6049	1.0	937.9909	0.6970	0.5646	0.6968	0.0	0.6971	0.6969
0.2879	12.0	136440	0.3385	1.0	939.8961	0.8725	0.8656	0.8726	0.0	0.8725	0.8726

Framework versions

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