a45732f67a9eaf8b4b98af424d12ac13

This model is a fine-tuned version of distilbert/distilbert-base-cased-distilled-squad on the nyu-mll/glue [sst2] dataset. It achieves the following results on the evaluation set:

• Loss: 0.5315
• Data Size: 1.0
• Epoch Runtime: 55.7050
• Accuracy: 0.8912
• F1 Macro: 0.8909
• Rouge1: 0.8900
• Rouge2: 0.0
• Rougel: 0.8912
• Rougelsum: 0.8912

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	0.6996	0	0.8819	0.4907	0.3292	0.4907	0.0	0.4907	0.4919
No log	1	2104	0.4867	0.0078	1.9768	0.7720	0.7720	0.7720	0.0	0.7720	0.7720
No log	2	4208	0.4103	0.0156	1.8108	0.8137	0.8113	0.8137	0.0	0.8137	0.8137
0.0095	3	6312	0.3374	0.0312	2.8309	0.8542	0.8542	0.8542	0.0	0.8547	0.8542
0.341	4	8416	0.3014	0.0625	4.5492	0.8738	0.8737	0.8738	0.0	0.8738	0.8738
0.2747	5	10520	0.3625	0.125	7.8829	0.8588	0.8587	0.8588	0.0	0.8588	0.8576
0.2077	6	12624	0.3917	0.25	14.7922	0.8623	0.8613	0.8634	0.0	0.8623	0.8623
0.2061	7	14728	0.2977	0.5	28.0708	0.8877	0.8876	0.8877	0.0	0.8877	0.8883
0.1573	8.0	16832	0.2965	1.0	54.7423	0.9005	0.9005	0.9005	0.0	0.9005	0.9005
0.1121	9.0	18936	0.3981	1.0	54.7366	0.8889	0.8889	0.8889	0.0	0.8889	0.8889
0.1021	10.0	21040	0.4421	1.0	54.9207	0.8831	0.8831	0.8831	0.0	0.8831	0.8831
0.0536	11.0	23144	0.4679	1.0	56.0739	0.8808	0.8803	0.8808	0.0	0.8808	0.8808
0.0655	12.0	25248	0.5315	1.0	55.7050	0.8912	0.8909	0.8900	0.0	0.8912	0.8912

Framework versions

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