refiners/tests/training_utils/mock_config_2_models.toml

[mock_model1]
requires_grad = true
learning_rate = 1e-5

[mock_model2]
requires_grad = true

[clock]
verbose = false

[training]
duration = "100:epoch"
seed = 0
batch_size = 4
gradient_accumulation = "4:step"
evaluation_interval = "5:epoch"
evaluation_seed = 1
gradient_clipping_max_norm = 1.0

[optimizer]
optimizer = "SGD"
learning_rate = 1

[lr_scheduler]
type = "ConstantLR"
update_interval = "1:iteration"
refactor register_model decorator 2024-02-12 13:17:51 +00:00			`[mock_model1]`
Allow optional train ModelConfig + forbid extra input for configs 2024-02-10 14:53:18 +00:00			`requires_grad = true`
lr, betas, eps, weight_decay at model level Co-authored-by: Cédric Deltheil <355031+deltheil@users.noreply.github.com> 2024-02-01 08:26:59 +00:00			`learning_rate = 1e-5`

refactor register_model decorator 2024-02-12 13:17:51 +00:00			`[mock_model2]`
Allow optional train ModelConfig + forbid extra input for configs 2024-02-10 14:53:18 +00:00			`requires_grad = true`
lr, betas, eps, weight_decay at model level Co-authored-by: Cédric Deltheil <355031+deltheil@users.noreply.github.com> 2024-02-01 08:26:59 +00:00
add @register_model and @register_callback decorators Refactor ClockTrainer to include Callback 2024-02-12 08:28:41 +00:00			`[clock]`
			`verbose = false`

lr, betas, eps, weight_decay at model level Co-authored-by: Cédric Deltheil <355031+deltheil@users.noreply.github.com> 2024-02-01 08:26:59 +00:00			`[training]`
			`duration = "100:epoch"`
			`seed = 0`
			`batch_size = 4`
			`gradient_accumulation = "4:step"`
			`evaluation_interval = "5:epoch"`
			`evaluation_seed = 1`
Switch gradient clipping to native torch torch.nn.utils.clip_grad_norm_ 2024-03-19 16:34:34 +00:00			`gradient_clipping_max_norm = 1.0`
lr, betas, eps, weight_decay at model level Co-authored-by: Cédric Deltheil <355031+deltheil@users.noreply.github.com> 2024-02-01 08:26:59 +00:00
			`[optimizer]`
			`optimizer = "SGD"`
			`learning_rate = 1`

rename Scheduler -> LRScheduler 2024-02-15 09:48:12 +00:00			`[lr_scheduler]`
			`type = "ConstantLR"`
Refactor TimeValue 2024-04-18 14:41:57 +00:00			`update_interval = "1:iteration"`