pollen-ner2-400

This model is a fine-tuned version of DeepPavlov/bert-base-bg-cs-pl-ru-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6034
• Precision: 0.4117
• Recall: 0.4960
• F1: 0.4499

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• 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
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1
No log	1.0	50	0.8021	0.3428	0.1948	0.2484
No log	2.0	100	0.7462	0.3190	0.2390	0.2732
No log	3.0	150	0.7122	0.3388	0.3293	0.3340
No log	4.0	200	0.6697	0.3721	0.3594	0.3657
No log	5.0	250	0.6542	0.3978	0.4297	0.4131
No log	6.0	300	0.6287	0.4071	0.4357	0.4210
No log	7.0	350	0.6155	0.4011	0.4518	0.4249
No log	8.0	400	0.6096	0.4068	0.4779	0.4395
No log	9.0	450	0.6042	0.4132	0.4920	0.4491
1.1014	10.0	500	0.6034	0.4117	0.4960	0.4499

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.7.0+cu128
• Datasets 3.5.0
• Tokenizers 0.21.1