CeLLaTe3.0_no_vague_adapted_tok

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
1.0902	1.0	201	0.3240	0.7557	0.7769	0.7662	0.9090
0.2236	2.0	402	0.1587	0.8654	0.9202	0.8919	0.9594
0.1178	3.0	603	0.1405	0.9020	0.9375	0.9194	0.9639
0.0779	4.0	804	0.1647	0.8984	0.9267	0.9124	0.9608
0.0543	5.0	1005	0.1751	0.9070	0.9261	0.9165	0.9633
0.04	6.0	1206	0.2010	0.8966	0.9239	0.9101	0.9600
0.0306	7.0	1407	0.2047	0.9005	0.9200	0.9101	0.9595
0.0242	8.0	1608	0.2117	0.9028	0.9172	0.9099	0.9621
0.019	9.0	1809	0.2070	0.9126	0.9316	0.9220	0.9652
0.0145	10.0	2010	0.2219	0.9177	0.9296	0.9236	0.9649
0.0113	11.0	2211	0.2304	0.9045	0.9299	0.9170	0.9627
0.0089	12.0	2412	0.2482	0.9026	0.9216	0.9120	0.9618
0.0079	13.0	2613	0.2594	0.9057	0.9214	0.9135	0.9613
0.0068	14.0	2814	0.3033	0.8992	0.9096	0.9044	0.9596
0.0053	15.0	3015	0.2722	0.9062	0.9200	0.9130	0.9624
0.0051	16.0	3216	0.2651	0.9080	0.9238	0.9158	0.9637
0.0041	17.0	3417	0.2697	0.9121	0.9211	0.9166	0.9635
0.0038	18.0	3618	0.2755	0.9078	0.9246	0.9161	0.9634
0.0032	19.0	3819	0.2719	0.9078	0.9261	0.9168	0.9633
0.0032	20.0	4020	0.2706	0.9103	0.9252	0.9177	0.9636

Safetensors

Model size

0.1B params

Tensor type

F32

Base model

Finetuned

Finetuned

(4)

this model