cf5fd067a432029aba2a9df19727e81a

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

• Loss: 0.6330
• Data Size: 1.0
• Epoch Runtime: 400.7248
• Accuracy: 0.7962
• F1 Macro: 0.7958
• Rouge1: 0.7962
• Rouge2: 0.0
• Rougel: 0.7961
• Rougelsum: 0.7962

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	1.1106	0	3.6559	0.3292	0.2436	0.3294	0.0	0.3288	0.3296
1.0382	1	12271	0.8957	0.0078	8.1470	0.6014	0.6014	0.6015	0.0	0.6015	0.6017
0.8627	2	24542	0.8030	0.0156	10.2993	0.6544	0.6464	0.6545	0.0	0.6547	0.6545
0.7559	3	36813	0.7186	0.0312	16.0539	0.6962	0.6936	0.6960	0.0	0.6962	0.6963
0.6997	4	49084	0.6661	0.0625	28.5904	0.7256	0.7249	0.7254	0.0	0.7260	0.7256
0.6106	5	61355	0.6235	0.125	52.8673	0.7426	0.7408	0.7426	0.0	0.7426	0.7427
0.632	6	73626	0.6339	0.25	102.2006	0.7452	0.7461	0.7450	0.0	0.7452	0.7450
0.5155	7	85897	0.5847	0.5	199.6007	0.7639	0.7637	0.7638	0.0	0.7637	0.7641
0.5208	8.0	98168	0.5472	1.0	395.4411	0.7825	0.7828	0.7824	0.0	0.7825	0.7825
0.4672	9.0	110439	0.5268	1.0	402.2054	0.7889	0.7883	0.7886	0.0	0.7889	0.7889
0.4283	10.0	122710	0.5447	1.0	403.6526	0.7913	0.7901	0.7912	0.0	0.7911	0.7913
0.3874	11.0	134981	0.5530	1.0	399.0921	0.7972	0.7977	0.7972	0.0	0.7974	0.7972
0.3645	12.0	147252	0.5262	1.0	399.8399	0.8011	0.8005	0.8012	0.0	0.8011	0.8012
0.3413	13.0	159523	0.5466	1.0	401.2812	0.8007	0.8004	0.8006	0.0	0.8009	0.8007
0.3217	14.0	171794	0.6170	1.0	402.6584	0.7897	0.7891	0.7896	0.0	0.7896	0.7899
0.2613	15.0	184065	0.6549	1.0	399.1038	0.7905	0.7911	0.7905	0.0	0.7907	0.7908
0.2669	16.0	196336	0.6330	1.0	400.7248	0.7962	0.7958	0.7962	0.0	0.7961	0.7962

Framework versions

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