|
--- |
|
base_model: models/smolm-mlm/config.json |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: smolm-mlm-bpe-unmask-seed_444 |
|
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. --> |
|
|
|
# smolm-mlm-bpe-unmask-seed_444 |
|
|
|
This model is a fine-tuned version of [models/smolm-mlm/config.json](https://huggingface.co/models/smolm-mlm/config.json) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 2.7042 |
|
- Accuracy: 0.4475 |
|
|
|
## 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: 0.0005 |
|
- train_batch_size: 64 |
|
- eval_batch_size: 512 |
|
- seed: 444 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 24000 |
|
- num_epochs: 10.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:------:|:---------------:|:--------:| |
|
| 3.5523 | 1.0 | 11938 | 3.4904 | 0.3512 | |
|
| 3.3151 | 2.0 | 23876 | 3.3192 | 0.3678 | |
|
| 3.1526 | 3.0 | 35814 | 3.1545 | 0.3894 | |
|
| 3.0031 | 4.0 | 47752 | 3.0706 | 0.3971 | |
|
| 2.9668 | 5.0 | 59690 | 2.9606 | 0.4121 | |
|
| 2.8548 | 6.0 | 71628 | 2.8717 | 0.4244 | |
|
| 2.7686 | 7.0 | 83566 | 2.8147 | 0.4311 | |
|
| 2.7195 | 8.0 | 95504 | 2.7657 | 0.4389 | |
|
| 2.6659 | 9.0 | 107442 | 2.7103 | 0.4483 | |
|
| 2.6254 | 10.0 | 119380 | 2.6990 | 0.4510 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.32.1 |
|
- Pytorch 1.13.1+cu117 |
|
- Datasets 2.12.0 |
|
- Tokenizers 0.13.3 |
|
|
