File size: 1,928 Bytes
b7f7157 |
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 70 71 72 |
---
base_model: scb10x/typhoon-7b
library_name: peft
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: local_model_directory
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. -->
# local_model_directory
This model is a fine-tuned version of [scb10x/typhoon-7b](https://huggingface.co/scb10x/typhoon-7b) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4505
## 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: 0.0002
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2
- num_epochs: 100
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 5.0407 | 0.9524 | 5 | 4.8168 |
| 3.5268 | 1.9048 | 10 | 2.1234 |
| 1.9397 | 2.8571 | 15 | 1.2800 |
| 1.0223 | 4.0 | 21 | 0.7717 |
| 0.8135 | 4.9524 | 26 | 0.5757 |
| 0.6574 | 5.9048 | 31 | 0.5178 |
| 0.5833 | 6.8571 | 36 | 0.4854 |
| 0.4549 | 8.0 | 42 | 0.4494 |
| 0.5012 | 8.9524 | 47 | 0.4294 |
| 0.514 | 9.9048 | 52 | 0.4505 |
### Framework versions
- PEFT 0.11.1
- Transformers 4.42.3
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1 |