CeLLaTe3.0_no_vague_pubmed

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:

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
0.8103	1.0	402	0.1761	0.5832	0.7740	0.6652	0.9529
0.1291	2.0	804	0.1448	0.6999	0.7842	0.7396	0.9618
0.0705	3.0	1206	0.1829	0.6816	0.8217	0.7451	0.9572
0.0393	4.0	1608	0.1870	0.6990	0.7938	0.7434	0.9620
0.0253	5.0	2010	0.2277	0.7128	0.7826	0.7461	0.9612
0.0155	6.0	2412	0.2306	0.7044	0.7877	0.7437	0.9641
0.0115	7.0	2814	0.2300	0.7706	0.8273	0.7979	0.9672
0.0082	8.0	3216	0.2642	0.7331	0.8121	0.7706	0.9644
0.006	9.0	3618	0.2808	0.7183	0.7887	0.7519	0.9634
0.0046	10.0	4020	0.2718	0.7441	0.8151	0.7780	0.9648
0.0034	11.0	4422	0.2977	0.7215	0.8172	0.7664	0.9635
0.0024	12.0	4824	0.3096	0.7462	0.8151	0.7791	0.9637
0.0022	13.0	5226	0.3120	0.7335	0.8217	0.7751	0.9643
0.0017	14.0	5628	0.3049	0.7420	0.8238	0.7807	0.9652
0.0014	15.0	6030	0.3178	0.7311	0.8273	0.7763	0.9649
0.0014	16.0	6432	0.3172	0.7349	0.8167	0.7736	0.9647
0.0009	17.0	6834	0.3216	0.7516	0.8192	0.7840	0.9644
0.0007	18.0	7236	0.3236	0.7441	0.8228	0.7815	0.9655
0.0007	19.0	7638	0.3225	0.7543	0.8217	0.7866	0.9657
0.0008	20.0	8040	0.3233	0.7513	0.8207	0.7845	0.9655

Safetensors

Model size

0.1B params

Tensor type

F32

Base model

Finetuned

Finetuned

(4)

this model