train.yaml

# @package _global_


# specify here default configuration
# order of defaults determines the order in which configs override each other
defaults:
  - _self_
  - data: mnist.yaml
  - model_task: mnist.yaml
  - callbacks: default.yaml
  - logger: null # set logger here or use command line (e.g. `python train.py logger=tensorboard`)
  - trainer: gpu.yaml # cpu, gpu, mps 
  - paths: default.yaml
  - extras: default.yaml
  - hydra: default.yaml

  # experiment configs allow for version control of specific hyperparameters
  # e.g. best hyperparameters for given model and datamodule
  - experiment: null

  # config for hyperparameter optimization
  - hparams_search: null

  # optional local config for machine/user specific settings
  # it's optional since it doesn't need to exist and is excluded from version control
  - optional local: default.yaml

  # debugging config (enable through command line, e.g. `python train.py debug=default)
  - debug: null

# task name, determines output directory path
task_name: "train"

# tags to help you identify your experiments
# you can overwrite this in experiment configs
# overwrite from command line with `python train.py tags="[first_tag, second_tag]"`
tags: ["dev"]

# set False to skip model training
train: True

# evaluate on test set, using best model weights achieved during training
# lightning chooses best weights based on the metric specified in checkpoint callback
test: True

# compile model for faster training with pytorch 2.0
compile: False

# simply provide checkpoint path to resume training
ckpt_path: null

# seed for random number generators in pytorch, numpy and python.random
seed: null


# torch matmul precision: You are using a CUDA device ('NVIDIA A100 80GB PCIe') that has Tensor Cores. To properly utilize them, you should set `torch.set_float32_matmul_precision('medium' | 'high')` which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision
torch_matmul_precision: medium # medium, high, or None