README.md

metadata

license: mit
base_model: microsoft/mdeberta-v3-base
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: mdeberta-v3-base_binary_2_seed7_NL-IT
    results: []

mdeberta-v3-base_binary_2_seed7_NL-IT

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

• Loss: 0.5531
• Accuracy: 0.7234
• F1: 0.7266
• Precision: 0.7324
• Recall: 0.7234

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-06
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.7125	0.2105	100	0.6482	0.6667	0.5333	0.4444	0.6667
0.6448	0.4211	200	0.6268	0.6667	0.5333	0.4444	0.6667
0.6334	0.6316	300	0.5938	0.6667	0.5333	0.4444	0.6667
0.6066	0.8421	400	0.5962	0.6667	0.5333	0.4444	0.6667
0.5887	1.0526	500	0.5849	0.7106	0.6807	0.6942	0.7106
0.5683	1.2632	600	0.5597	0.7034	0.6406	0.7043	0.7034
0.578	1.4737	700	0.5500	0.7177	0.7172	0.7167	0.7177
0.5565	1.6842	800	0.5487	0.6916	0.6992	0.7202	0.6916
0.5505	1.8947	900	0.5365	0.7117	0.7062	0.7035	0.7117
0.5137	2.1053	1000	0.5331	0.7236	0.7269	0.7322	0.7236
0.5162	2.3158	1100	0.5339	0.7307	0.7304	0.7300	0.7307
0.5022	2.5263	1200	0.5303	0.7307	0.7336	0.7379	0.7307
0.5103	2.7368	1300	0.5346	0.7426	0.7353	0.7340	0.7426
0.4983	2.9474	1400	0.5239	0.7331	0.7350	0.7374	0.7331
0.4902	3.1579	1500	0.5232	0.7367	0.7397	0.7447	0.7367
0.4496	3.3684	1600	0.5384	0.7497	0.7460	0.7443	0.7497
0.4522	3.5789	1700	0.5386	0.7497	0.7496	0.7495	0.7497
0.4597	3.7895	1800	0.5583	0.7426	0.7373	0.7354	0.7426

Framework versions

• Transformers 4.40.2
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.19.1