metadata
license: apache-2.0
base_model: mistralai/Mistral-7B-Instruct-v0.2
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: mistralit2_550_STEPS_5e8_SFT_SFT
results: []
mistralit2_550_STEPS_5e8_SFT_SFT
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5702
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: 5e-08
- train_batch_size: 4
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 550
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.4484 | 0.1 | 50 | 1.4370 |
| 1.3055 | 0.2 | 100 | 1.2880 |
| 1.1109 | 0.29 | 150 | 1.0891 |
| 0.9328 | 0.39 | 200 | 0.9223 |
| 0.7797 | 0.49 | 250 | 0.7676 |
| 0.6719 | 0.59 | 300 | 0.6567 |
| 0.5895 | 0.68 | 350 | 0.5927 |
| 0.5801 | 0.78 | 400 | 0.5714 |
| 0.5676 | 0.88 | 450 | 0.5703 |
| 0.5723 | 0.98 | 500 | 0.5702 |
| 0.5668 | 1.07 | 550 | 0.5702 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.0.0+cu117
- Datasets 2.18.0
- Tokenizers 0.15.2