sft_10000
This model is a fine-tuned version of mistralai/Mistral-Nemo-Instruct-2407 on the heat_transfer_10000_fs dataset. It achieves the following results on the evaluation set:
- Loss: 0.0005
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: 0.0001
- train_batch_size: 12
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- total_train_batch_size: 24
- total_eval_batch_size: 2
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- num_epochs: 2
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
0.0104 | 0.0533 | 20 | 0.0039 |
0.0037 | 0.1067 | 40 | 0.0035 |
0.0034 | 0.16 | 60 | 0.0030 |
0.0019 | 0.2133 | 80 | 0.0022 |
0.0025 | 0.2667 | 100 | 0.0022 |
0.0017 | 0.32 | 120 | 0.0012 |
0.0017 | 0.3733 | 140 | 0.0012 |
0.0014 | 0.4267 | 160 | 0.0016 |
0.0011 | 0.48 | 180 | 0.0012 |
0.0011 | 0.5333 | 200 | 0.0010 |
0.0008 | 0.5867 | 220 | 0.0010 |
0.0012 | 0.64 | 240 | 0.0011 |
0.0009 | 0.6933 | 260 | 0.0012 |
0.0012 | 0.7467 | 280 | 0.0010 |
0.001 | 0.8 | 300 | 0.0008 |
0.0008 | 0.8533 | 320 | 0.0007 |
0.0007 | 0.9067 | 340 | 0.0009 |
0.0011 | 0.96 | 360 | 0.0007 |
0.0008 | 1.0133 | 380 | 0.0010 |
0.0008 | 1.0667 | 400 | 0.0008 |
0.0007 | 1.12 | 420 | 0.0008 |
0.0005 | 1.1733 | 440 | 0.0006 |
0.0007 | 1.2267 | 460 | 0.0007 |
0.0009 | 1.28 | 480 | 0.0006 |
0.0006 | 1.3333 | 500 | 0.0006 |
0.0006 | 1.3867 | 520 | 0.0006 |
0.0007 | 1.44 | 540 | 0.0006 |
0.0006 | 1.4933 | 560 | 0.0006 |
0.0004 | 1.5467 | 580 | 0.0006 |
0.0005 | 1.6 | 600 | 0.0005 |
0.0005 | 1.6533 | 620 | 0.0006 |
0.0006 | 1.7067 | 640 | 0.0005 |
0.0005 | 1.76 | 660 | 0.0005 |
0.0006 | 1.8133 | 680 | 0.0005 |
0.0005 | 1.8667 | 700 | 0.0005 |
0.0005 | 1.92 | 720 | 0.0005 |
0.0004 | 1.9733 | 740 | 0.0005 |
Framework versions
- PEFT 0.12.0
- Transformers 4.46.0
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.20.1
- Downloads last month
- 7
Model tree for Howard881010/heat_transfer_sft_10000_fs
Base model
mistralai/Mistral-Nemo-Base-2407
Finetuned
mistralai/Mistral-Nemo-Instruct-2407