pacsum_led_model / README.md
alwaysaditi's picture
End of training
f31464a verified
metadata
license: apache-2.0
base_model: allenai/led-base-16384
tags:
  - generated_from_trainer
model-index:
  - name: DATASET_PACSUM
    results: []

DATASET_PACSUM

This model is a fine-tuned version of allenai/led-base-16384 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 2.5461

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: 2
  • eval_batch_size: 2
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 8
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 5
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
2.8648 0.1 10 2.8816
2.9889 0.2 20 2.7866
3.0516 0.3 30 2.7394
2.6605 0.4 40 2.7132
2.8093 0.5 50 2.6759
2.9206 0.6 60 2.6607
2.8094 0.7 70 2.6576
2.5233 0.8 80 2.6327
2.6508 0.9 90 2.6117
2.8456 1.0 100 2.5861
2.4622 1.1 110 2.5942
2.2871 1.2 120 2.5751
2.4482 1.3 130 2.5776
2.4079 1.4 140 2.5777
2.2842 1.5 150 2.5621
2.6267 1.6 160 2.5463
2.3895 1.7 170 2.5503
2.2786 1.8 180 2.5470
2.3628 1.9 190 2.5420
2.2809 2.0 200 2.5367
2.2726 2.1 210 2.5405
2.1934 2.2 220 2.5676
2.2447 2.3 230 2.5399
2.4508 2.4 240 2.5435
2.2969 2.5 250 2.5490
2.4206 2.6 260 2.5317
2.0131 2.7 270 2.5378
2.0025 2.8 280 2.5492
2.2179 2.9 290 2.5280
2.2082 3.0 300 2.5190
1.9491 3.1 310 2.5608
2.291 3.2 320 2.5448
2.0431 3.3 330 2.5319
2.0671 3.4 340 2.5529
2.1939 3.5 350 2.5388
2.0606 3.6 360 2.5306
2.0088 3.7 370 2.5557
2.1919 3.8 380 2.5317
2.2516 3.9 390 2.5290
1.9401 4.0 400 2.5404
2.1101 4.1 410 2.5354
1.8906 4.2 420 2.5520
1.9808 4.3 430 2.5488
1.8195 4.4 440 2.5496
1.8512 4.5 450 2.5535
2.0464 4.6 460 2.5519
2.0176 4.7 470 2.5450
2.0686 4.8 480 2.5460
2.0267 4.9 490 2.5463
1.8617 5.0 500 2.5461

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.3.0+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1