metadata
license: mit
base_model: microsoft/Phi-3-mini-128k-instruct
tags:
- generated_from_trainer
model-index:
- name: CodePhi-3-miniFull0.1k
results: []
CodePhi-3-miniFull0.1k
This model is a fine-tuned version of microsoft/Phi-3-mini-128k-instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6586
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-06
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- training_steps: 100
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.7277 | 0.1 | 10 | 0.7000 |
| 0.6698 | 0.2 | 20 | 0.6784 |
| 0.6505 | 0.3 | 30 | 0.6673 |
| 0.6621 | 0.4 | 40 | 0.6620 |
| 0.5436 | 0.5 | 50 | 0.6590 |
| 0.3544 | 0.6 | 60 | 0.6591 |
| 0.4525 | 0.7 | 70 | 0.6588 |
| 0.4208 | 0.8 | 80 | 0.6586 |
| 0.5922 | 0.9 | 90 | 0.6586 |
| 0.6187 | 1.0 | 100 | 0.6586 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1