README.md

metadata

library_name: transformers
license: mit
base_model: Mardiyyah/cellate1.0-tapt_freeze_llrd_ww_mask-LR_2e-05
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: TAPT_CeLLaTe_llrd_only
    results: []

TAPT_CeLLaTe_llrd_only

This model is a fine-tuned version of Mardiyyah/cellate1.0-tapt_freeze_llrd_ww_mask-LR_2e-05 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1153
• Precision: 0.8168
• Recall: 0.8404
• F1: 0.8285
• Accuracy: 0.9743

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: 16
• eval_batch_size: 16
• seed: 3407
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
1.5314	1.0	55	0.6843	0.0	0.0	0.0	0.8947
0.4343	2.0	110	0.2517	0.3447	0.3526	0.3486	0.9195
0.2147	3.0	165	0.1484	0.6493	0.7204	0.6830	0.9563
0.1396	4.0	220	0.1172	0.7452	0.7599	0.7524	0.9681
0.1112	5.0	275	0.1102	0.7370	0.8176	0.7752	0.9660
0.0892	6.0	330	0.0984	0.7994	0.7994	0.7994	0.9713
0.0747	7.0	385	0.1059	0.8238	0.8100	0.8169	0.9735
0.0643	8.0	440	0.1112	0.7768	0.8252	0.8003	0.9703
0.0533	9.0	495	0.1079	0.8361	0.8298	0.8330	0.9748
0.0473	10.0	550	0.1082	0.8121	0.8343	0.8231	0.9736
0.0445	11.0	605	0.1094	0.8468	0.8146	0.8304	0.9750
0.0375	12.0	660	0.1047	0.8477	0.8374	0.8425	0.9762
0.0312	13.0	715	0.1052	0.8149	0.8298	0.8223	0.9741
0.0299	14.0	770	0.1095	0.8070	0.8389	0.8227	0.9727
0.0269	15.0	825	0.1195	0.7874	0.8389	0.8124	0.9718
0.0238	16.0	880	0.1096	0.8301	0.8389	0.8345	0.9749
0.0218	17.0	935	0.1134	0.8070	0.8450	0.8255	0.9741
0.022	18.0	990	0.1174	0.8038	0.8404	0.8217	0.9736
0.02	19.0	1045	0.1189	0.8151	0.8374	0.8261	0.9741
0.02	20.0	1100	0.1153	0.8168	0.8404	0.8285	0.9743

Framework versions

• Transformers 4.48.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.2
• Tokenizers 0.21.0