CeLLaTe3.0_Base_no_vague_adapted_pubmed_gaz

This model is a fine-tuned version of Mardiyyah/cellate2.0-tapt_base-LR_5e-05 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1429
• Precision: 0.8007
• Recall: 0.8511
• F1: 0.8251
• Accuracy: 0.9656

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: 32
• 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
1.8192	0.2625	100	1.0343	0.0	0.0	0.0	0.8644
0.6494	0.5249	200	0.4040	0.2492	0.1613	0.1958	0.8851
0.3634	0.7874	300	0.2682	0.4259	0.3941	0.4094	0.9181
0.2352	1.0499	400	0.1609	0.7297	0.7455	0.7375	0.9588
0.1488	1.3123	500	0.1463	0.7468	0.8138	0.7789	0.9599
0.1157	1.5748	600	0.1597	0.8004	0.7518	0.7754	0.9610
0.0962	1.8373	700	0.1630	0.7967	0.7250	0.7592	0.9596
0.0742	2.0997	800	0.1551	0.8050	0.7888	0.7968	0.9636
0.0564	2.3622	900	0.1445	0.8005	0.8511	0.8250	0.9656
0.0518	2.6247	1000	0.1700	0.7614	0.8448	0.8009	0.9605
0.0448	2.8871	1100	0.1857	0.7895	0.8102	0.7997	0.9620
0.0317	3.1496	1200	0.1977	0.8013	0.8240	0.8125	0.9619

Framework versions

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