|
--- |
|
license: mit |
|
base_model: mikhail-panzo/zlm_b64_le4_s12000 |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: zlm-fil_b64_le5_s8000 |
|
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. --> |
|
|
|
# zlm-fil_b64_le5_s8000 |
|
|
|
This model is a fine-tuned version of [mikhail-panzo/zlm_b64_le4_s12000](https://huggingface.co/mikhail-panzo/zlm_b64_le4_s12000) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4118 |
|
|
|
## 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: 32 |
|
- eval_batch_size: 8 |
|
- 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 |
|
- lr_scheduler_warmup_steps: 2000 |
|
- training_steps: 8000 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:--------:|:----:|:---------------:| |
|
| 0.5529 | 22.2222 | 500 | 0.5000 | |
|
| 0.4974 | 44.4444 | 1000 | 0.4557 | |
|
| 0.4716 | 66.6667 | 1500 | 0.4359 | |
|
| 0.453 | 88.8889 | 2000 | 0.4246 | |
|
| 0.4428 | 111.1111 | 2500 | 0.4196 | |
|
| 0.4332 | 133.3333 | 3000 | 0.4171 | |
|
| 0.4246 | 155.5556 | 3500 | 0.4154 | |
|
| 0.4202 | 177.7778 | 4000 | 0.4133 | |
|
| 0.4223 | 200.0 | 4500 | 0.4145 | |
|
| 0.4127 | 222.2222 | 5000 | 0.4118 | |
|
| 0.418 | 244.4444 | 5500 | 0.4130 | |
|
| 0.4137 | 266.6667 | 6000 | 0.4130 | |
|
| 0.4105 | 288.8889 | 6500 | 0.4127 | |
|
| 0.4164 | 311.1111 | 7000 | 0.4127 | |
|
| 0.4088 | 333.3333 | 7500 | 0.4120 | |
|
| 0.4028 | 355.5556 | 8000 | 0.4118 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.41.0.dev0 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.19.0 |
|
- Tokenizers 0.19.1 |
|
|
