Richard Saavedra
End of training
31f9a7d
---
tags:
- generated_from_trainer
datasets:
- amazon_reviews_multi
metrics:
- accuracy
model-index:
- name: electricidad-small-finetuned-amazon-review-classification
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: amazon_reviews_multi
type: amazon_reviews_multi
args: es
metrics:
- name: Accuracy
type: accuracy
value: 0.581
---
<!-- 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. -->
# electricidad-small-finetuned-amazon-review-classification
This model is a fine-tuned version of [mrm8488/electricidad-small-discriminator](https://huggingface.co/mrm8488/electricidad-small-discriminator) on the amazon_reviews_multi dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9601
- Accuracy: 0.581
## 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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 1.0136 | 1.0 | 25000 | 1.0153 | 0.5414 |
| 0.9416 | 2.0 | 50000 | 0.9942 | 0.5576 |
### Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 2.0.0
- Tokenizers 0.11.6