---
library_name: transformers
license: apache-2.0
base_model: deepvk/RuModernBERT-base
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: RuHalluBERT-base-v4
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# RuHalluBERT-base-v4

This model is a fine-tuned version of [deepvk/RuModernBERT-base](https://huggingface.co/deepvk/RuModernBERT-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9101
- F1 Macro: 0.7177
- F1 Class1: 0.6720
- F1 Class0: 0.7634
- Accuracy: 0.7251
- Precision Macro: 0.7252
- Recall Macro: 0.7160

## 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: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 0.06
- num_epochs: 10

### Training results

| Training Loss | Epoch | Step | Validation Loss | F1 Macro | F1 Class1 | F1 Class0 | Accuracy | Precision Macro | Recall Macro |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:---------:|:--------:|:---------------:|:------------:|
| 11.2782       | 1.0   | 113  | 0.6490          | 0.6090   | 0.5076    | 0.7105    | 0.6353   | 0.6385          | 0.6155       |
| 10.4259       | 2.0   | 226  | 0.6731          | 0.5534   | 0.6783    | 0.4286    | 0.5884   | 0.7154          | 0.6228       |
| 8.9919        | 3.0   | 339  | 0.5547          | 0.7192   | 0.6884    | 0.75      | 0.7226   | 0.7197          | 0.7188       |
| 7.0874        | 4.0   | 452  | 0.5604          | 0.7349   | 0.6963    | 0.7734    | 0.7405   | 0.7396          | 0.7333       |
| 5.1057        | 5.0   | 565  | 0.6218          | 0.7091   | 0.7059    | 0.7124    | 0.7092   | 0.7144          | 0.7153       |
| 4.0620        | 6.0   | 678  | 0.6958          | 0.7181   | 0.7175    | 0.7188    | 0.7181   | 0.7250          | 0.7252       |
| 2.7632        | 7.0   | 791  | 0.7622          | 0.7370   | 0.6982    | 0.7758    | 0.7427   | 0.7421          | 0.7353       |
| 2.2915        | 8.0   | 904  | 0.8325          | 0.7164   | 0.6942    | 0.7386    | 0.7181   | 0.7160          | 0.7175       |
| 1.9114        | 9.0   | 1017 | 0.8926          | 0.7336   | 0.7079    | 0.7592    | 0.7360   | 0.7334          | 0.7338       |
| 1.2978        | 10.0  | 1130 | 0.9135          | 0.7269   | 0.7012    | 0.7526    | 0.7293   | 0.7266          | 0.7272       |


### Framework versions

- Transformers 5.8.1
- Pytorch 2.11.0+cu130
- Datasets 4.8.5
- Tokenizers 0.22.2