|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: mnli_IndE |
|
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. --> |
|
|
|
# mnli_IndE |
|
|
|
This model is a fine-tuned version of [WillHeld/roberta-base-mnli](https://huggingface.co/WillHeld/roberta-base-mnli) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5406 |
|
- Acc: 0.8536 |
|
|
|
## 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: 32 |
|
- eval_batch_size: 32 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Acc | |
|
|:-------------:|:-----:|:-----:|:---------------:|:------:| |
|
| 0.3901 | 0.17 | 2000 | 0.4456 | 0.8354 | |
|
| 0.3758 | 0.33 | 4000 | 0.4508 | 0.8356 | |
|
| 0.3668 | 0.5 | 6000 | 0.4372 | 0.8425 | |
|
| 0.3653 | 0.67 | 8000 | 0.4357 | 0.8400 | |
|
| 0.3543 | 0.83 | 10000 | 0.4030 | 0.8517 | |
|
| 0.3559 | 1.0 | 12000 | 0.4242 | 0.8472 | |
|
| 0.2523 | 1.17 | 14000 | 0.4746 | 0.8464 | |
|
| 0.2521 | 1.33 | 16000 | 0.4780 | 0.8470 | |
|
| 0.2525 | 1.5 | 18000 | 0.4664 | 0.8507 | |
|
| 0.2464 | 1.67 | 20000 | 0.4806 | 0.8484 | |
|
| 0.2495 | 1.83 | 22000 | 0.4868 | 0.8464 | |
|
| 0.2451 | 2.0 | 24000 | 0.4794 | 0.8508 | |
|
| 0.1737 | 2.17 | 26000 | 0.5492 | 0.8491 | |
|
| 0.1727 | 2.33 | 28000 | 0.5552 | 0.8531 | |
|
| 0.1736 | 2.5 | 30000 | 0.5418 | 0.8515 | |
|
| 0.1746 | 2.67 | 32000 | 0.5511 | 0.8516 | |
|
| 0.1717 | 2.83 | 34000 | 0.5406 | 0.8536 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.24.0 |
|
- Pytorch 1.13.0+cu117 |
|
- Datasets 2.7.1 |
|
- Tokenizers 0.12.1 |
|
|
