|
|
--- |
|
|
license: mit |
|
|
base_model: microsoft/deberta-v3-large |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: checkpoints_2 |
|
|
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. --> |
|
|
|
|
|
# checkpoints_2 |
|
|
|
|
|
This model is a fine-tuned version of [microsoft/deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) on the None dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.8543 |
|
|
- Map@3: 0.7167 |
|
|
|
|
|
## 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: 1 |
|
|
- eval_batch_size: 2 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 8 |
|
|
- total_train_batch_size: 8 |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: cosine |
|
|
- lr_scheduler_warmup_ratio: 0.1 |
|
|
- num_epochs: 5 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Map@3 | |
|
|
|:-------------:|:-----:|:----:|:---------------:|:------:| |
|
|
| 1.395 | 0.19 | 25 | 1.3859 | 0.5889 | |
|
|
| 1.3803 | 0.37 | 50 | 1.3840 | 0.6958 | |
|
|
| 1.3842 | 0.56 | 75 | 1.3314 | 0.7194 | |
|
|
| 1.2795 | 0.74 | 100 | 1.0021 | 0.7222 | |
|
|
| 0.9662 | 0.93 | 125 | 0.9006 | 0.6597 | |
|
|
| 0.9574 | 1.11 | 150 | 0.8355 | 0.6903 | |
|
|
| 0.8909 | 1.3 | 175 | 0.8506 | 0.6750 | |
|
|
| 0.8077 | 1.48 | 200 | 0.8180 | 0.7125 | |
|
|
| 0.955 | 1.67 | 225 | 0.8069 | 0.7097 | |
|
|
| 0.8664 | 1.85 | 250 | 0.8186 | 0.7028 | |
|
|
| 0.9396 | 2.04 | 275 | 0.8091 | 0.6986 | |
|
|
| 0.8141 | 2.22 | 300 | 0.8212 | 0.7083 | |
|
|
| 0.7898 | 2.41 | 325 | 0.8531 | 0.7167 | |
|
|
| 0.9143 | 2.59 | 350 | 0.8482 | 0.7125 | |
|
|
| 0.8861 | 2.78 | 375 | 0.8229 | 0.7083 | |
|
|
| 0.8569 | 2.96 | 400 | 0.8372 | 0.7181 | |
|
|
| 0.8381 | 3.15 | 425 | 0.8516 | 0.7153 | |
|
|
| 0.7671 | 3.33 | 450 | 0.8458 | 0.7167 | |
|
|
| 0.8704 | 3.52 | 475 | 0.8651 | 0.7222 | |
|
|
| 0.8733 | 3.7 | 500 | 0.8356 | 0.7153 | |
|
|
| 0.7309 | 3.89 | 525 | 0.8476 | 0.7181 | |
|
|
| 0.7793 | 4.07 | 550 | 0.8566 | 0.7167 | |
|
|
| 0.7849 | 4.26 | 575 | 0.8644 | 0.7167 | |
|
|
| 0.7776 | 4.44 | 600 | 0.8584 | 0.7167 | |
|
|
| 0.7573 | 4.63 | 625 | 0.8546 | 0.7167 | |
|
|
| 0.8115 | 4.81 | 650 | 0.8543 | 0.7167 | |
|
|
| 0.869 | 5.0 | 675 | 0.8543 | 0.7167 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.35.0 |
|
|
- Pytorch 2.0.0 |
|
|
- Datasets 2.1.0 |
|
|
- Tokenizers 0.14.1 |
|
|
|
