---
library_name: transformers
base_model: Fsoft-AIC/videberta-xsmall
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: videberta-xsmall_v2
  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. -->

# videberta-xsmall_v2

This model is a fine-tuned version of [Fsoft-AIC/videberta-xsmall](https://huggingface.co/Fsoft-AIC/videberta-xsmall) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3825
- Accuracy: 0.9021
- Precision Macro: 0.7867
- Recall Macro: 0.7246
- F1 Macro: 0.7460
- F1 Weighted: 0.8970

## 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: 3e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- optimizer: Use 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: 20
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision Macro | Recall Macro | F1 Macro | F1 Weighted |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------------:|:------------:|:--------:|:-----------:|
| 0.8576        | 1.0   | 90   | 0.5078          | 0.8263   | 0.5549          | 0.5804       | 0.5639   | 0.8071      |
| 0.462         | 2.0   | 180  | 0.4000          | 0.8667   | 0.5773          | 0.6061       | 0.5912   | 0.8464      |
| 0.3895        | 3.0   | 270  | 0.3928          | 0.8711   | 0.5819          | 0.6073       | 0.5939   | 0.8503      |
| 0.334         | 4.0   | 360  | 0.3652          | 0.8793   | 0.5871          | 0.6133       | 0.5996   | 0.8584      |
| 0.2906        | 5.0   | 450  | 0.3542          | 0.8844   | 0.7567          | 0.6261       | 0.6205   | 0.8658      |
| 0.2912        | 6.0   | 540  | 0.3562          | 0.8895   | 0.8269          | 0.6502       | 0.6619   | 0.8747      |
| 0.2564        | 7.0   | 630  | 0.3404          | 0.8932   | 0.7944          | 0.7308       | 0.7537   | 0.8890      |
| 0.2424        | 8.0   | 720  | 0.3381          | 0.8970   | 0.8492          | 0.6758       | 0.6998   | 0.8854      |
| 0.2167        | 9.0   | 810  | 0.3292          | 0.9015   | 0.8294          | 0.7088       | 0.7377   | 0.8938      |
| 0.2052        | 10.0  | 900  | 0.3621          | 0.9021   | 0.8239          | 0.7163       | 0.7459   | 0.8953      |
| 0.1976        | 11.0  | 990  | 0.3453          | 0.9002   | 0.8251          | 0.7113       | 0.7408   | 0.8930      |
| 0.1904        | 12.0  | 1080 | 0.3754          | 0.9015   | 0.8426          | 0.7040       | 0.7345   | 0.8931      |
| 0.176         | 13.0  | 1170 | 0.3586          | 0.9046   | 0.8177          | 0.7101       | 0.7378   | 0.8971      |
| 0.1783        | 14.0  | 1260 | 0.3635          | 0.8958   | 0.7590          | 0.7239       | 0.7379   | 0.8922      |
| 0.1566        | 15.0  | 1350 | 0.4087          | 0.8926   | 0.7601          | 0.7089       | 0.7270   | 0.8874      |
| 0.1509        | 16.0  | 1440 | 0.3878          | 0.9033   | 0.8019          | 0.7172       | 0.7427   | 0.8970      |
| 0.1463        | 17.0  | 1530 | 0.3730          | 0.8989   | 0.7670          | 0.7348       | 0.7481   | 0.8959      |
| 0.1493        | 18.0  | 1620 | 0.3801          | 0.8996   | 0.7687          | 0.7225       | 0.7397   | 0.8952      |
| 0.1465        | 19.0  | 1710 | 0.3995          | 0.8983   | 0.7738          | 0.7175       | 0.7372   | 0.8932      |
| 0.133         | 20.0  | 1800 | 0.3825          | 0.9021   | 0.7867          | 0.7246       | 0.7460   | 0.8970      |


### Framework versions

- Transformers 4.55.0
- Pytorch 2.7.0+cu126
- Datasets 4.0.0
- Tokenizers 0.21.4