|
|
--- |
|
|
license: mit |
|
|
--- |
|
|
|
|
|
# MirrorAPI-Cache |
|
|
|
|
|
This model is a fine-tuned version of [StableToolBench-MirrorAPI](https://huggingface.co/stabletoolbench/MirrorAPI). |
|
|
|
|
|
### Training and evaluation data |
|
|
|
|
|
[The training data of MirrorAPI-Cache](https://huggingface.co/datasets/stabletoolbench/MirrorAPI/tree/main/train) is `train_cache.json`. |
|
|
|
|
|
[The evaluation data of MirrorAPI-Cache](https://huggingface.co/datasets/stabletoolbench/MirrorAPI/tree/main/test) is `test_cache.json`. |
|
|
|
|
|
### Training hyperparameters |
|
|
|
|
|
The following hyperparameters were used during training: |
|
|
- learning_rate: 1e-05 |
|
|
- train_batch_size: 8 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- distributed_type: multi-GPU |
|
|
- num_devices: 8 |
|
|
- gradient_accumulation_steps: 1 |
|
|
- total_train_batch_size: 64 |
|
|
- total_eval_batch_size: 64 |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: cosine |
|
|
- lr_scheduler_warmup_ratio: 0.1 |
|
|
- lr_scheduler_warmup_steps: 0 |
|
|
- num_epochs: 3.0 |
