Edit model card

SadNews_2_20epochs

This model is a fine-tuned version of csebuetnlp/mT5_multilingual_XLSum on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 3.2760

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
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 1000
  • num_epochs: 20

Training results

Training Loss Epoch Step Validation Loss
3.3998 0.76 500 2.8591
2.8316 1.52 1000 2.5640
2.3969 2.29 1500 2.4411
1.8762 3.05 2000 2.4241
1.9914 3.81 2500 2.3928
1.8294 4.57 3000 2.4093
1.7505 5.34 3500 2.4674
1.313 6.1 4000 2.5457
1.5311 6.86 4500 2.4944
1.2645 7.62 5000 2.5876
1.0581 8.38 5500 2.6490
1.2508 9.15 6000 2.7311
1.3671 9.91 6500 2.7163
0.9799 10.67 7000 2.7751
1.03 11.43 7500 2.8911
0.8382 12.2 8000 2.9649
1.0784 12.96 8500 2.9205
0.7969 13.72 9000 3.0321
0.7965 14.48 9500 3.0564
0.7737 15.24 10000 3.1322
0.591 16.01 10500 3.1051
0.8767 16.77 11000 3.1922
0.6458 17.53 11500 3.2355
0.591 18.29 12000 3.2301
0.6428 19.05 12500 3.2603
0.7789 19.82 13000 3.2760

Framework versions

  • Transformers 4.38.1
  • Pytorch 2.1.0+cu121
  • Datasets 2.17.1
  • Tokenizers 0.15.2
Downloads last month
2
Safetensors
Model size
582M 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 tjkmitl/SadNews_2_20epochs

Finetuned
(28)
this model