CeLLaTe3.0_with_vague_pubmed

This model is a fine-tuned version of microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext on an unknown dataset. It achieves the following results on the evaluation set:

Model description

More information needed

More information needed

More information needed

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 Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.7676	1.0	403	0.1896	0.4931	0.6667	0.5669	0.9484
0.1364	2.0	806	0.1445	0.6858	0.7643	0.7229	0.9613
0.0731	3.0	1209	0.1694	0.6925	0.7972	0.7412	0.9608
0.0395	4.0	1612	0.2014	0.7222	0.7714	0.7460	0.9629
0.0256	5.0	2015	0.2258	0.6952	0.7484	0.7208	0.9598
0.0146	6.0	2418	0.2429	0.7083	0.8047	0.7534	0.9619
0.0104	7.0	2821	0.2540	0.7144	0.7939	0.7521	0.9631
0.007	8.0	3224	0.2652	0.7238	0.7714	0.7468	0.9633
0.0054	9.0	3627	0.2546	0.7247	0.7958	0.7586	0.9638
0.0041	10.0	4030	0.2641	0.7398	0.7930	0.7655	0.9648
0.0028	11.0	4433	0.3089	0.7047	0.7732	0.7374	0.9612
0.0023	12.0	4836	0.3026	0.7031	0.7681	0.7341	0.9615
0.0023	13.0	5239	0.2905	0.7410	0.7967	0.7679	0.9643
0.0018	14.0	5642	0.3225	0.7278	0.7718	0.7491	0.9622
0.0014	15.0	6045	0.3014	0.7595	0.7915	0.7752	0.9648
0.0013	16.0	6448	0.3055	0.7409	0.7962	0.7676	0.9638
0.0009	17.0	6851	0.3146	0.7488	0.7977	0.7724	0.9639
0.0008	18.0	7254	0.3058	0.7519	0.8066	0.7783	0.9647
0.0007	19.0	7657	0.3073	0.7514	0.8061	0.7778	0.9647
0.0006	20.0	8060	0.3135	0.7557	0.8103	0.7821	0.9648

Safetensors

Model size

0.1B params

Tensor type

F32

Base model

Finetuned

(145)

this model