README.md

metadata

license: mit
base_model: microsoft/deberta-v3-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: deberta-v3-base-p-tuning-isarcasm
    results: []

deberta-v3-base-p-tuning-isarcasm

This model is a fine-tuned version of microsoft/deberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6620
• Accuracy: 0.4476
• F1: 0.3603
• Precision: 0.2266
• Recall: 0.8790

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: 0.001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
No log	1.0	108	0.6889	0.3143	0.2941	0.1786	0.8333
No log	2.0	216	0.7020	0.8286	0.0	0.0	0.0
No log	3.0	324	0.6725	0.6286	0.2353	0.1818	0.3333
No log	4.0	432	0.7169	0.1714	0.2927	0.1714	1.0
0.7087	5.0	540	0.6925	0.5429	0.3846	0.25	0.8333
0.7087	6.0	648	0.6991	0.1714	0.2927	0.1714	1.0
0.7087	7.0	756	0.6780	0.8286	0.0	0.0	0.0
0.7087	8.0	864	0.6851	0.8286	0.0	0.0	0.0
0.7087	9.0	972	0.6712	0.8286	0.0	0.0	0.0
0.7055	10.0	1080	0.6767	0.3143	0.3333	0.2	1.0
0.7055	11.0	1188	0.6720	0.5714	0.4000	0.2632	0.8333
0.7055	12.0	1296	0.6710	0.3714	0.3529	0.2143	1.0
0.7055	13.0	1404	0.6676	0.4857	0.3077	0.2	0.6667
0.6916	14.0	1512	0.6735	0.3714	0.3125	0.1923	0.8333
0.6916	15.0	1620	0.6762	0.3714	0.3529	0.2143	1.0
0.6916	16.0	1728	0.6642	0.6286	0.3158	0.2308	0.5
0.6916	17.0	1836	0.6609	0.5143	0.32	0.2105	0.6667
0.6916	18.0	1944	0.6632	0.4571	0.2963	0.1905	0.6667
0.6798	19.0	2052	0.6640	0.4	0.2759	0.1739	0.6667
0.6798	20.0	2160	0.6644	0.4	0.2759	0.1739	0.6667

Framework versions

• Transformers 4.32.0
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3