dhiya96's picture
dhiya96/zephyr_summarisation_finetuned
b081fbf verified
metadata
license: apache-2.0
base_model: google-t5/t5-base
tags:
  - generated_from_trainer
metrics:
  - rouge
model-index:
  - name: t5-base-finetuned-stocknews_1900_100
    results: []

t5-base-finetuned-stocknews_1900_100

This model is a fine-tuned version of google-t5/t5-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 1.2997
  • Rouge1: 16.6203
  • Rouge2: 8.7831
  • Rougel: 13.9116
  • Rougelsum: 15.4831
  • Gen Len: 19.0

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: 2e-05
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 40
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 102 1.5488 14.6381 6.8963 12.1802 13.6527 19.0
No log 2.0 204 1.4139 15.0451 6.9216 12.6068 14.1445 19.0
No log 3.0 306 1.3627 15.3864 7.115 12.6537 14.267 19.0
No log 4.0 408 1.3288 15.6891 7.5106 13.0451 14.6203 19.0
1.8685 5.0 510 1.3087 15.8071 7.6382 13.103 14.7587 19.0
1.8685 6.0 612 1.2938 15.6775 7.6448 13.0823 14.6034 19.0
1.8685 7.0 714 1.2870 15.7672 7.89 13.3325 14.7821 19.0
1.8685 8.0 816 1.2779 16.1616 8.1642 13.4471 15.0305 19.0
1.8685 9.0 918 1.2731 16.3679 8.4804 13.7618 15.3468 19.0
1.1991 10.0 1020 1.2695 16.2821 8.456 13.7692 15.2461 19.0
1.1991 11.0 1122 1.2647 16.4056 8.5019 13.7217 15.3711 19.0
1.1991 12.0 1224 1.2667 16.4259 8.6692 13.8396 15.4122 19.0
1.1991 13.0 1326 1.2654 16.6988 8.9574 14.0239 15.6864 19.0
1.1991 14.0 1428 1.2648 16.7394 9.0588 14.0529 15.6644 19.0
1.0382 15.0 1530 1.2642 16.6864 9.106 13.9046 15.5687 19.0
1.0382 16.0 1632 1.2662 16.6786 8.8288 13.9603 15.5724 19.0
1.0382 17.0 1734 1.2651 16.7446 8.9211 13.9999 15.6617 19.0
1.0382 18.0 1836 1.2702 16.6361 8.8503 14.0324 15.546 19.0
1.0382 19.0 1938 1.2676 16.7046 9.0089 14.073 15.6342 19.0
0.9273 20.0 2040 1.2732 16.4339 8.6714 13.8422 15.44 19.0
0.9273 21.0 2142 1.2743 16.5655 8.7747 13.839 15.4958 19.0
0.9273 22.0 2244 1.2781 16.7749 8.9154 14.1216 15.6395 19.0
0.9273 23.0 2346 1.2814 16.535 8.7436 13.971 15.5056 19.0
0.9273 24.0 2448 1.2795 16.6612 8.7045 14.0096 15.5692 19.0
0.8539 25.0 2550 1.2844 16.6083 8.6106 13.9202 15.5641 19.0
0.8539 26.0 2652 1.2817 16.6449 8.8127 14.0562 15.5792 19.0
0.8539 27.0 2754 1.2856 16.6185 8.7475 14.0134 15.5439 19.0
0.8539 28.0 2856 1.2868 16.4913 8.7293 13.9367 15.4702 19.0
0.8539 29.0 2958 1.2905 16.4887 8.6461 13.8893 15.4342 19.0
0.8006 30.0 3060 1.2893 16.5861 8.695 13.9081 15.4307 19.0
0.8006 31.0 3162 1.2919 16.5972 8.8314 13.9069 15.4967 19.0
0.8006 32.0 3264 1.2940 16.5957 8.789 13.9202 15.4839 19.0
0.8006 33.0 3366 1.2946 16.6313 8.8011 13.9684 15.5256 19.0
0.8006 34.0 3468 1.2945 16.6711 8.8915 14.0228 15.5394 19.0
0.7598 35.0 3570 1.2970 16.67 8.891 13.9749 15.5174 19.0
0.7598 36.0 3672 1.2975 16.6223 8.7522 13.9528 15.4761 19.0
0.7598 37.0 3774 1.2987 16.6444 8.8594 13.9567 15.5117 19.0
0.7598 38.0 3876 1.2993 16.6444 8.8594 13.9567 15.5117 19.0
0.7598 39.0 3978 1.2996 16.6196 8.8108 13.9213 15.4806 19.0
0.7463 40.0 4080 1.2997 16.6203 8.7831 13.9116 15.4831 19.0

Framework versions

  • Transformers 4.38.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.18.0
  • Tokenizers 0.15.2