Hugging Face's logo

haryoaw
/
scenario-NON-KD-PR-COPY-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_

Transformers
PyTorch
Safetensors
deberta-v2
Generated from Trainer
Inference Endpoints
Model card Files Community
scenario-NON-KD-PR-COPY-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_ / README.md
haryoaw's picture
haryoaw
Initial Commit
a9d29df verified
preview code
|
raw
history blame
4.86 kB
metadata
license: mit
base_model: microsoft/mdeberta-v3-base
tags:
  - generated_from_trainer
datasets:
  - tweet_sentiment_multilingual
metrics:
  - accuracy
  - f1
model-index:
  - name: >-
      scenario-NON-KD-PR-COPY-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_
    results: []

scenario-NON-KD-PR-COPY-CDF-ALL-D2_data-cardiffnlp_tweet_sentiment_multilingual_

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the tweet_sentiment_multilingual dataset. It achieves the following results on the evaluation set:

  • Loss: 4.7763
  • Accuracy: 0.5525
  • F1: 0.5514

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: 32
  • eval_batch_size: 32
  • seed: 66
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Accuracy F1
1.0802 1.09 500 1.0761 0.4776 0.4641
0.9869 2.17 1000 1.0145 0.5320 0.5250
0.8881 3.26 1500 0.9953 0.5432 0.5391
0.7877 4.35 2000 1.0837 0.5340 0.5311
0.6671 5.43 2500 1.2702 0.5394 0.5411
0.5716 6.52 3000 1.4643 0.5440 0.5409
0.4817 7.61 3500 1.6304 0.5448 0.5336
0.4102 8.7 4000 1.7103 0.5301 0.5267
0.3334 9.78 4500 2.0038 0.5343 0.5334
0.2776 10.87 5000 1.8016 0.5475 0.5472
0.2349 11.96 5500 2.0203 0.5282 0.5280
0.2017 13.04 6000 2.4490 0.5359 0.5334
0.1727 14.13 6500 2.5313 0.5378 0.5382
0.1491 15.22 7000 2.3797 0.5390 0.5388
0.1425 16.3 7500 2.4724 0.5444 0.5446
0.1265 17.39 8000 2.9398 0.5413 0.5389
0.1185 18.48 8500 2.3527 0.5370 0.5370
0.1038 19.57 9000 3.2756 0.5482 0.5442
0.1071 20.65 9500 3.0308 0.5432 0.5441
0.0865 21.74 10000 3.1408 0.5297 0.5296
0.0813 22.83 10500 3.3928 0.5436 0.5434
0.0831 23.91 11000 3.4793 0.5320 0.5339
0.0736 25.0 11500 3.2782 0.5451 0.5452
0.0672 26.09 12000 3.4270 0.5428 0.5396
0.0616 27.17 12500 3.7192 0.5471 0.5425
0.0588 28.26 13000 3.3739 0.5421 0.5424
0.0537 29.35 13500 3.5891 0.5421 0.5393
0.0534 30.43 14000 3.5400 0.5436 0.5391
0.0503 31.52 14500 4.1166 0.5409 0.5378
0.0431 32.61 15000 4.1346 0.5374 0.5339
0.0423 33.7 15500 3.9483 0.5478 0.5456
0.0371 34.78 16000 4.0371 0.5436 0.5429
0.0339 35.87 16500 4.0302 0.5478 0.5480
0.0381 36.96 17000 4.0057 0.5432 0.5425
0.0274 38.04 17500 4.5734 0.5521 0.5520
0.0288 39.13 18000 4.4791 0.5502 0.5472
0.0203 40.22 18500 4.7187 0.5536 0.5538
0.0248 41.3 19000 4.7855 0.5486 0.5490
0.025 42.39 19500 4.4324 0.5502 0.5471
0.0211 43.48 20000 4.7410 0.5475 0.5470
0.0215 44.57 20500 4.6235 0.5478 0.5483
0.0188 45.65 21000 4.6657 0.5517 0.5499
0.0163 46.74 21500 4.7207 0.5509 0.5505
0.0136 47.83 22000 4.7870 0.5525 0.5523
0.0131 48.91 22500 4.8396 0.5505 0.5501
0.0207 50.0 23000 4.7763 0.5525 0.5514

Framework versions

  • Transformers 4.33.3
  • Pytorch 2.1.1+cu121
  • Datasets 2.14.5
  • Tokenizers 0.13.3