848717d1bb6dcb417941b32247e995ac

This model is a fine-tuned version of albert/albert-xxlarge-v1 on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 0.3848
• Data Size: 1.0
• Epoch Runtime: 307.6984
• Accuracy: 0.9088
• F1 Macro: 0.9087
• Rouge1: 0.9088
• Rouge2: 0.0
• Rougel: 0.9088
• Rougelsum: 0.9088

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	0.7221	0	4.0115	0.5048	0.3416	0.5046	0.0	0.5046	0.5044
No log	1	3273	0.4753	0.0078	6.2258	0.7871	0.7834	0.7869	0.0	0.7868	0.7875
0.0091	2	6546	0.3686	0.0156	8.3708	0.8386	0.8353	0.8386	0.0	0.8386	0.8386
0.3873	3	9819	0.2283	0.0312	12.7594	0.9204	0.9204	0.9204	0.0	0.9202	0.9204
0.2796	4	13092	0.2260	0.0625	21.0479	0.9175	0.9175	0.9175	0.0	0.9176	0.9176
0.2356	5	16365	0.1960	0.125	38.0719	0.9267	0.9266	0.9268	0.0	0.9267	0.9267
0.2192	6	19638	0.2178	0.25	70.9393	0.9226	0.9226	0.9230	0.0	0.9224	0.9228
0.19	7	22911	0.1886	0.5	153.8451	0.9278	0.9277	0.9278	0.0	0.9279	0.9278
0.1771	8.0	26184	0.1892	1.0	308.0908	0.9263	0.9263	0.9263	0.0	0.9263	0.9265
0.0974	9.0	29457	0.2689	1.0	307.7917	0.9187	0.9187	0.9187	0.0	0.9186	0.9189
0.0794	10.0	32730	0.2984	1.0	303.5691	0.9197	0.9196	0.9199	0.0	0.9195	0.9197
0.1164	11.0	36003	0.3848	1.0	307.6984	0.9088	0.9087	0.9088	0.0	0.9088	0.9088

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1