File size: 2,180 Bytes
82f6a75 |
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 72 |
---
license: mit
base_model: FacebookAI/xlm-roberta-base
tags:
- generated_from_trainer
model-index:
- name: xlm-roberta-base_hau_loss_5e-06
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. -->
# xlm-roberta-base_hau_loss_5e-06
This model is a fine-tuned version of [FacebookAI/xlm-roberta-base](https://huggingface.co/FacebookAI/xlm-roberta-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0248
- Spearman Corr: 0.7591
## 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-06
- train_batch_size: 32
- eval_batch_size: 128
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Spearman Corr |
|:-------------:|:-----:|:----:|:---------------:|:-------------:|
| No log | 0.95 | 200 | 0.0249 | 0.7564 |
| No log | 1.91 | 400 | 0.0258 | 0.7583 |
| 0.0014 | 2.86 | 600 | 0.0239 | 0.7618 |
| 0.0014 | 3.82 | 800 | 0.0246 | 0.7601 |
| 0.0013 | 4.77 | 1000 | 0.0242 | 0.7588 |
| 0.0013 | 5.73 | 1200 | 0.0241 | 0.7574 |
| 0.0013 | 6.68 | 1400 | 0.0248 | 0.7605 |
| 0.0013 | 7.64 | 1600 | 0.0255 | 0.7623 |
| 0.0014 | 8.59 | 1800 | 0.0247 | 0.7585 |
| 0.0014 | 9.55 | 2000 | 0.0252 | 0.7582 |
| 0.0014 | 10.5 | 2200 | 0.0248 | 0.7591 |
### Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2
|