generated_from_trainer

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

small-vanilla-target-glue-mnli

This model is a fine-tuned version of google/bert_uncased_L-4_H-512_A-8 on the None dataset. It achieves the following results on the evaluation set:

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:

Training results

Training Loss Epoch Step Validation Loss Accuracy
0.9249 0.04 500 0.8197 0.6419
0.8154 0.08 1000 0.7776 0.6651
0.7747 0.12 1500 0.7455 0.6749
0.7523 0.16 2000 0.7315 0.6853
0.7445 0.2 2500 0.7097 0.6977
0.7337 0.24 3000 0.6981 0.7026
0.7165 0.29 3500 0.6876 0.7117
0.704 0.33 4000 0.6669 0.7192
0.68 0.37 4500 0.6773 0.7249
0.6834 0.41 5000 0.6513 0.7228
0.6808 0.45 5500 0.6467 0.7260
0.671 0.49 6000 0.6416 0.7275
0.6733 0.53 6500 0.6434 0.7372
0.6714 0.57 7000 0.6358 0.7394
0.6552 0.61 7500 0.6243 0.7423
0.6519 0.65 8000 0.6272 0.7380
0.6546 0.69 8500 0.6207 0.7440
0.6457 0.73 9000 0.6293 0.7424
0.6458 0.77 9500 0.6084 0.7492
0.6349 0.81 10000 0.6248 0.7431
0.6376 0.86 10500 0.6035 0.7531
0.6294 0.9 11000 0.5969 0.7561
0.6264 0.94 11500 0.5981 0.7518
0.6297 0.98 12000 0.5953 0.7539
0.5935 1.02 12500 0.6093 0.7602
0.5726 1.06 13000 0.5991 0.7628
0.5571 1.1 13500 0.5921 0.7606
0.5535 1.14 14000 0.6011 0.7559
0.5427 1.18 14500 0.6020 0.7618

Framework versions