balassarprofile
This model is a fine-tuned version of TheBloke/Mistral-7B-Instruct-v0.2-GPTQ on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.8042
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: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.8802 | 1.0 | 1 | 3.6270 |
| 3.861 | 2.0 | 2 | 3.5746 |
| 3.7758 | 3.0 | 3 | 3.4416 |
| 3.5819 | 4.0 | 4 | 3.3048 |
| 3.3879 | 5.0 | 5 | 3.1740 |
| 3.2106 | 6.0 | 6 | 3.0575 |
| 3.0652 | 7.0 | 7 | 2.9588 |
| 2.94 | 8.0 | 8 | 2.8822 |
| 2.8566 | 9.0 | 9 | 2.8301 |
| 2.7926 | 10.0 | 10 | 2.8042 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 2
Model tree for Balassar/balassarprofile
Base model
mistralai/Mistral-7B-Instruct-v0.2
Quantized
TheBloke/Mistral-7B-Instruct-v0.2-GPTQ