|
--- |
|
license: mit |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
base_model: microsoft/phi-1_5 |
|
model-index: |
|
- name: phi-1_5-finetuned-qlora-cluster-gsm8k |
|
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. --> |
|
|
|
# phi-1_5-finetuned-qlora-cluster-gsm8k |
|
|
|
This model is a fine-tuned version of [microsoft/phi-1_5](https://huggingface.co/microsoft/phi-1_5) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.1536 |
|
|
|
## 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: cosine |
|
- num_epochs: 10 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 1.0717 | 1.0 | 233 | 1.1424 | |
|
| 0.9785 | 2.0 | 467 | 1.1173 | |
|
| 0.9489 | 3.0 | 701 | 1.1147 | |
|
| 0.9226 | 4.0 | 935 | 1.1216 | |
|
| 0.8962 | 5.0 | 1168 | 1.1204 | |
|
| 0.8611 | 6.0 | 1402 | 1.1294 | |
|
| 0.8424 | 7.0 | 1636 | 1.1372 | |
|
| 0.8275 | 8.0 | 1870 | 1.1493 | |
|
| 0.8187 | 9.0 | 2103 | 1.1526 | |
|
| 0.8183 | 9.97 | 2330 | 1.1536 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.11.1 |
|
- Transformers 4.37.2 |
|
- Pytorch 2.3.0 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.15.1 |
