Edit model card

flan-t5-2048-input-30-epochs

This model is a fine-tuned version of tanatapanun/flan-t5-2048-input-1-epochs on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7695
  • Rouge1: 0.2731
  • Rouge2: 0.1008
  • Rougel: 0.2328
  • Rougelsum: 0.2319
  • Gen Len: 13.82

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: 30

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Gen Len
No log 1.0 301 1.5244 0.2018 0.0697 0.1768 0.1758 11.98
2.3092 2.0 602 0.8884 0.2155 0.0817 0.1857 0.1837 14.36
2.3092 3.0 903 0.8211 0.2112 0.0739 0.1784 0.1772 15.2
0.9906 4.0 1204 0.8055 0.2205 0.082 0.1854 0.1844 14.71
0.8522 5.0 1505 0.7995 0.2381 0.084 0.2025 0.2026 14.59
0.8522 6.0 1806 0.7937 0.2449 0.0775 0.2046 0.2032 15.15
0.8235 7.0 2107 0.7917 0.2535 0.0882 0.2116 0.2104 14.57
0.8235 8.0 2408 0.7856 0.263 0.0848 0.2198 0.2195 14.36
0.7869 9.0 2709 0.7832 0.2759 0.097 0.2371 0.2347 13.82
0.7702 10.0 3010 0.7810 0.2622 0.0884 0.225 0.2253 14.16
0.7702 11.0 3311 0.7791 0.2807 0.0995 0.2384 0.2376 14.11
0.7721 12.0 3612 0.7786 0.2758 0.0896 0.237 0.2362 14.17
0.7721 13.0 3913 0.7772 0.2851 0.094 0.2466 0.2456 14.05
0.7534 14.0 4214 0.7747 0.2642 0.0929 0.2245 0.2235 13.95
0.7285 15.0 4515 0.7739 0.2681 0.0993 0.2281 0.2289 13.81
0.7285 16.0 4816 0.7746 0.2714 0.0978 0.2309 0.2302 14.02
0.7334 17.0 5117 0.7727 0.2734 0.0957 0.2317 0.2311 14.0
0.7334 18.0 5418 0.7720 0.2729 0.0996 0.2321 0.2321 13.92
0.7193 19.0 5719 0.7726 0.2693 0.0982 0.2298 0.2291 13.88
0.7147 20.0 6020 0.7727 0.2799 0.103 0.2378 0.2368 13.91
0.7147 21.0 6321 0.7715 0.2856 0.1047 0.2465 0.2451 13.83
0.7204 22.0 6622 0.7704 0.2724 0.1016 0.2339 0.233 13.91
0.7204 23.0 6923 0.7705 0.2751 0.1017 0.2344 0.2335 13.96
0.6853 24.0 7224 0.7697 0.2746 0.1011 0.2343 0.2333 13.73
0.7116 25.0 7525 0.7703 0.2757 0.1024 0.2357 0.2348 13.72
0.7116 26.0 7826 0.7698 0.2745 0.1008 0.2341 0.2333 13.72
0.6935 27.0 8127 0.7699 0.2731 0.1008 0.2328 0.2319 13.82
0.6935 28.0 8428 0.7698 0.2731 0.1008 0.2328 0.2319 13.82
0.7051 29.0 8729 0.7695 0.2731 0.1008 0.2328 0.2319 13.82
0.6923 30.0 9030 0.7695 0.2731 0.1008 0.2328 0.2319 13.82

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu118
  • Datasets 2.15.0
  • Tokenizers 0.15.0
Downloads last month
0
Safetensors
Model size
77M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for tanatapanun/flan-t5-2048-input-30-epochs

Finetuned
(1)
this model