metadata
base_model: mistralai/Mistral-7B-Instruct-v0.3
library_name: peft
license: apache-2.0
tags:
- generated_from_trainer
model-index:
- name: mistral-mistralinstruct_climate_finetune
results: []
mistral-mistralinstruct_climate_finetune
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4689
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: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.5446 | 0.4444 | 50 | 1.4019 |
| 1.1129 | 0.8889 | 100 | 0.9769 |
| 0.6082 | 1.3333 | 150 | 0.7359 |
| 0.5066 | 1.7778 | 200 | 0.5834 |
| 0.2138 | 2.2222 | 250 | 0.5273 |
| 0.234 | 2.6667 | 300 | 0.4689 |
Framework versions
- PEFT 0.11.1
- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1