File size: 2,003 Bytes
7be3ad6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 |
---
license: cc-by-4.0
library_name: peft
tags:
- generated_from_trainer
base_model: EMBEDDIA/crosloengual-bert
metrics:
- accuracy
- f1
model-index:
- name: lora_fine_tuned_boolq_croslo
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. -->
# lora_fine_tuned_boolq_croslo
This model is a fine-tuned version of [EMBEDDIA/crosloengual-bert](https://huggingface.co/EMBEDDIA/crosloengual-bert) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5923
- Accuracy: 0.7778
- F1: 0.6806
## 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
- training_steps: 400
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|:-------------:|:-------:|:----:|:---------------:|:--------:|:------:|
| 0.7055 | 4.1667 | 50 | 0.6182 | 0.7222 | 0.7325 |
| 0.6596 | 8.3333 | 100 | 0.5842 | 0.8333 | 0.8243 |
| 0.6565 | 12.5 | 150 | 0.5833 | 0.8333 | 0.8243 |
| 0.6642 | 16.6667 | 200 | 0.5852 | 0.7778 | 0.6806 |
| 0.6495 | 20.8333 | 250 | 0.5873 | 0.7778 | 0.6806 |
| 0.6477 | 25.0 | 300 | 0.5892 | 0.7778 | 0.6806 |
| 0.652 | 29.1667 | 350 | 0.5918 | 0.7778 | 0.6806 |
| 0.6362 | 33.3333 | 400 | 0.5923 | 0.7778 | 0.6806 |
### Framework versions
- PEFT 0.10.1.dev0
- Transformers 4.40.1
- Pytorch 2.3.0
- Datasets 2.19.0
- Tokenizers 0.19.1 |