metadata
license: afl-3.0
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: Viet-Mistral/Vistral-7B-Chat
datasets:
- generator
model-index:
- name: vietnamese-news-summarization-vistral-7b
results: []
vietnamese-news-summarization-vistral-7b
This model is a fine-tuned version of Viet-Mistral/Vistral-7B-Chat on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 1.2452
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: 4
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_steps: 0.03
- training_steps: 100
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.3699 | 0.0060 | 20 | 1.3159 |
| 1.4501 | 0.0119 | 40 | 1.2761 |
| 1.2554 | 0.0179 | 60 | 1.2583 |
| 1.1901 | 0.0239 | 80 | 1.2474 |
| 1.4126 | 0.0298 | 100 | 1.2452 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.0
- Pytorch 2.1.2
- Datasets 2.16.0
- Tokenizers 0.19.1