File size: 1,957 Bytes
0470b7c |
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 |
---
base_model: klue/roberta-large
tags:
- generated_from_trainer
datasets:
- klue
model-index:
- name: sts_roberta-large_lr1e-05_wd1e-03_ep7_ckpt
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. -->
# sts_roberta-large_lr1e-05_wd1e-03_ep7_ckpt
This model is a fine-tuned version of [klue/roberta-large](https://huggingface.co/klue/roberta-large) on the klue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3621
- Mse: 0.3621
- Mae: 0.4438
- R2: 0.8342
## 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: 1e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 7
### Training results
| Training Loss | Epoch | Step | Validation Loss | Mse | Mae | R2 |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|
| 1.8712 | 1.0 | 183 | 0.5118 | 0.5118 | 0.5409 | 0.7656 |
| 0.1606 | 2.0 | 366 | 0.4621 | 0.4621 | 0.5142 | 0.7884 |
| 0.1111 | 3.0 | 549 | 0.4687 | 0.4687 | 0.5088 | 0.7854 |
| 0.0837 | 4.0 | 732 | 0.4317 | 0.4317 | 0.4906 | 0.8023 |
| 0.0681 | 5.0 | 915 | 0.4662 | 0.4662 | 0.5091 | 0.7865 |
| 0.0559 | 6.0 | 1098 | 0.3742 | 0.3742 | 0.4524 | 0.8286 |
| 0.0485 | 7.0 | 1281 | 0.3621 | 0.3621 | 0.4438 | 0.8342 |
### Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.0
- Tokenizers 0.13.3
|