c01f8e34aac09a94e0066c0e05588317

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

• Loss: 0.7051
• Data Size: 1.0
• Epoch Runtime: 20.3889
• Accuracy: 0.8301
• F1 Macro: 0.7801
• Rouge1: 0.8301
• Rouge2: 0.0
• Rougel: 0.8301
• Rougelsum: 0.8301

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	0.7508	0	1.3434	0.3115	0.2375	0.3105	0.0	0.3115	0.3115
No log	1	267	0.6534	0.0078	1.6531	0.6885	0.4078	0.6895	0.0	0.6885	0.6885
No log	2	534	0.7352	0.0156	1.7671	0.6885	0.4078	0.6895	0.0	0.6885	0.6885
No log	3	801	0.6148	0.0312	2.2797	0.6885	0.4078	0.6895	0.0	0.6885	0.6885
No log	4	1068	0.5366	0.0625	2.9324	0.6904	0.4145	0.6904	0.0	0.6904	0.6904
0.0343	5	1335	0.6498	0.125	4.2355	0.7295	0.5479	0.7295	0.0	0.7295	0.7305
0.4681	6	1602	0.4779	0.25	6.6104	0.7959	0.7216	0.7959	0.0	0.7959	0.7959
0.4279	7	1869	0.5325	0.5	11.3156	0.7773	0.6644	0.7773	0.0	0.7773	0.7773
0.3358	8.0	2136	0.4419	1.0	21.0993	0.8213	0.7759	0.8203	0.0	0.8213	0.8213
0.2166	9.0	2403	0.4798	1.0	20.6877	0.8242	0.7731	0.8242	0.0	0.8242	0.8242
0.2188	10.0	2670	0.5300	1.0	20.6852	0.8359	0.7945	0.8359	0.0	0.8359	0.8359
0.164	11.0	2937	0.4815	1.0	20.2574	0.8164	0.7685	0.8164	0.0	0.8154	0.8174
0.1603	12.0	3204	0.7051	1.0	20.3889	0.8301	0.7801	0.8301	0.0	0.8301	0.8301

Framework versions

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