metadata
license: apache-2.0
library_name: peft
tags:
- trl
- sft
- unsloth
- translation
- generated_from_trainer
base_model: unsloth/mistral-7b-bnb-4bit
model-index:
- name: MistralAI_iwslt15_10000
results: []
MistralAI_iwslt15_10000
This model is a fine-tuned version of unsloth/mistral-7b-bnb-4bit on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.5245
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: 4269
- 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: 1
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.1687 | 0.32 | 100 | 1.0937 |
| 1.0905 | 0.64 | 200 | 1.0724 |
| 1.0711 | 0.96 | 300 | 1.0552 |
| 0.9258 | 1.28 | 400 | 1.0648 |
| 0.8979 | 1.6 | 500 | 1.0613 |
| 0.8893 | 1.92 | 600 | 1.0512 |
| 0.7253 | 2.24 | 700 | 1.1353 |
| 0.6713 | 2.56 | 800 | 1.1260 |
| 0.6701 | 2.88 | 900 | 1.1252 |
| 0.5284 | 3.2 | 1000 | 1.2891 |
| 0.446 | 3.52 | 1100 | 1.2803 |
| 0.4454 | 3.84 | 1200 | 1.3040 |
| 0.3663 | 4.16 | 1300 | 1.5203 |
| 0.282 | 4.48 | 1400 | 1.5198 |
| 0.2798 | 4.8 | 1500 | 1.5245 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.3
- Pytorch 2.2.2+cu121
- Datasets 2.16.0
- Tokenizers 0.15.2