|
--- |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- trl |
|
- sft |
|
- unsloth |
|
- generated_from_trainer |
|
base_model: unsloth/gemma-1.1-7b-it-bnb-4bit |
|
model-index: |
|
- name: gemma-ViMMRC-Answer |
|
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. --> |
|
|
|
# gemma-ViMMRC-Answer |
|
|
|
This model is a fine-tuned version of [unsloth/gemma-1.1-7b-it-bnb-4bit](https://huggingface.co./unsloth/gemma-1.1-7b-it-bnb-4bit) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.1034 |
|
- Accuracy: 0.8493 |
|
|
|
## Model description |
|
|
|
More information needed |
|
|
|
## Intended uses & limitations |
|
|
|
More information needed |
|
|
|
## Training and evaluation data |
|
|
|
- ViMMRC train and test set |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 5e-05 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 8 |
|
- seed: 3407 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 64 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_steps: 5 |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 13.1 | 0.3306 | 10 | 5.9870 | |
|
| 2.4875 | 0.6612 | 20 | 1.1997 | |
|
| 0.5062 | 0.9917 | 30 | 0.2423 | |
|
| 0.1602 | 1.3223 | 40 | 0.1251 | |
|
| 0.1289 | 1.6529 | 50 | 0.1156 | |
|
| 0.1234 | 1.9835 | 60 | 0.1000 | |
|
| 0.0727 | 2.3140 | 70 | 0.1068 | |
|
| 0.0945 | 2.6446 | 80 | 0.1035 | |
|
| 0.0785 | 2.9752 | 90 | 0.1034 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.10.0 |
|
- Transformers 4.40.2 |
|
- Pytorch 2.3.0 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |
