Hyponatremia_L3_1000steps_1e5rate_05beta_DPO
This model is a fine-tuned version of tsavage68/Hyponatremia_L3_450steps_1e7rate_SFT on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0000
- Rewards/chosen: 3.1005
- Rewards/rejected: -14.4818
- Rewards/accuracies: 1.0
- Rewards/margins: 17.5823
- Logps/rejected: -68.3904
- Logps/chosen: -16.5158
- Logits/rejected: -1.0107
- Logits/chosen: -0.9178
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: 1e-05
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.0 | 0.2667 | 50 | 0.0000 | 2.6888 | -10.9218 | 1.0 | 13.6106 | -61.2704 | -17.3392 | -1.0060 | -0.9176 |
| 0.0 | 0.5333 | 100 | 0.0000 | 2.8123 | -12.0070 | 1.0 | 14.8193 | -63.4408 | -17.0922 | -1.0062 | -0.9166 |
| 0.0 | 0.8 | 150 | 0.0000 | 2.8718 | -12.6309 | 1.0 | 15.5027 | -64.6887 | -16.9733 | -1.0077 | -0.9178 |
| 0.0 | 1.0667 | 200 | 0.0000 | 2.9222 | -12.9997 | 1.0 | 15.9220 | -65.4263 | -16.8724 | -1.0083 | -0.9180 |
| 0.0 | 1.3333 | 250 | 0.0000 | 2.9587 | -13.2786 | 1.0 | 16.2372 | -65.9839 | -16.7994 | -1.0085 | -0.9179 |
| 0.0 | 1.6 | 300 | 0.0000 | 2.9748 | -13.5117 | 1.0 | 16.4865 | -66.4503 | -16.7673 | -1.0094 | -0.9185 |
| 0.0 | 1.8667 | 350 | 0.0000 | 3.0038 | -13.7311 | 1.0 | 16.7350 | -66.8891 | -16.7092 | -1.0097 | -0.9181 |
| 0.0 | 2.1333 | 400 | 0.0000 | 3.0287 | -13.8707 | 1.0 | 16.8994 | -67.1683 | -16.6595 | -1.0097 | -0.9178 |
| 0.0 | 2.4 | 450 | 0.0000 | 3.0555 | -14.0219 | 1.0 | 17.0774 | -67.4707 | -16.6059 | -1.0096 | -0.9174 |
| 0.0 | 2.6667 | 500 | 0.0000 | 3.0689 | -14.1391 | 1.0 | 17.2081 | -67.7051 | -16.5790 | -1.0110 | -0.9186 |
| 0.0 | 2.9333 | 550 | 0.0000 | 3.0728 | -14.2357 | 1.0 | 17.3085 | -67.8981 | -16.5711 | -1.0101 | -0.9176 |
| 0.0 | 3.2 | 600 | 0.0000 | 3.0755 | -14.3397 | 1.0 | 17.4152 | -68.1062 | -16.5658 | -1.0104 | -0.9180 |
| 0.0 | 3.4667 | 650 | 0.0000 | 3.0977 | -14.3908 | 1.0 | 17.4884 | -68.2083 | -16.5214 | -1.0106 | -0.9180 |
| 0.0 | 3.7333 | 700 | 0.0000 | 3.1035 | -14.4417 | 1.0 | 17.5452 | -68.3102 | -16.5099 | -1.0117 | -0.9189 |
| 0.0 | 4.0 | 750 | 0.0000 | 3.0881 | -14.4574 | 1.0 | 17.5455 | -68.3416 | -16.5406 | -1.0099 | -0.9170 |
| 0.0 | 4.2667 | 800 | 0.0000 | 3.1048 | -14.4756 | 1.0 | 17.5804 | -68.3780 | -16.5072 | -1.0102 | -0.9176 |
| 0.0 | 4.5333 | 850 | 0.0000 | 3.0963 | -14.4856 | 1.0 | 17.5819 | -68.3980 | -16.5242 | -1.0096 | -0.9168 |
| 0.0 | 4.8 | 900 | 0.0000 | 3.1097 | -14.4788 | 1.0 | 17.5885 | -68.3844 | -16.4973 | -1.0104 | -0.9175 |
| 0.0 | 5.0667 | 950 | 0.0000 | 3.1005 | -14.4818 | 1.0 | 17.5823 | -68.3904 | -16.5158 | -1.0107 | -0.9178 |
| 0.0 | 5.3333 | 1000 | 0.0000 | 3.1005 | -14.4818 | 1.0 | 17.5823 | -68.3904 | -16.5158 | -1.0107 | -0.9178 |
Framework versions
- Transformers 4.42.3
- Pytorch 2.0.0+cu117
- Datasets 2.20.0
- Tokenizers 0.19.1
- Downloads last month
- 5
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Model tree for tsavage68/Hyponatremia_L3_1000steps_1e5rate_05beta_DPO
Base model
meta-llama/Meta-Llama-3-8B-Instruct