Optimizer that implements the Muon algorithm.

Note that this optimizer should not be used in the following layers:

Embedding layer
Final output fully connected layer
Any 0- or 1-D variables

These should all be optimized using AdamW.

The Muon optimizer can use both the Muon update step or the AdamW update step based on the following:

For any variable that isn't 2D, 3D or 4D, the AdamW step will be used. This is not configurable.
If the argument exclude_embeddings (defaults to TRUE) is set to TRUE, the AdamW step will be used.
For any variable with a name that matches an expression listed in the argument exclude_layers (a list), the AdamW step will be used.
Any other variable uses the Muon step.

Typically, you only need to pass the name of your densely-connected output layer to exclude_layers, e.g. exclude_layers = "output_dense".

Usage

optimizer_muon(
  learning_rate = 0.001,
  adam_beta_1 = 0.9,
  adam_beta_2 = 0.999,
  epsilon = 1e-07,
  weight_decay = 0.1,
  clipnorm = NULL,
  clipvalue = NULL,
  global_clipnorm = NULL,
  use_ema = FALSE,
  ema_momentum = 0.99,
  ema_overwrite_frequency = NULL,
  loss_scale_factor = NULL,
  gradient_accumulation_steps = NULL,
  name = "muon",
  exclude_layers = NULL,
  exclude_embeddings = TRUE,
  muon_a = 3.4445,
  muon_b = -4.775,
  muon_c = 2.0315,
  adam_lr_ratio = 0.1,
  momentum = 0.95,
  ns_steps = 6L,
  nesterov = TRUE,
  ...
)

Arguments

learning_rate: A float, LearningRateSchedule() instance, or a callable that takes no arguments and returns the actual value to use. The learning rate. Defaults to 0.001.
adam_beta_1: A float value or a constant float tensor, or a callable that takes no arguments and returns the actual value to use. The exponential decay rate for the 1st moment estimates. Defaults to 0.9.
adam_beta_2: A float value or a constant float tensor, ora callable that takes no arguments and returns the actual value to use. The exponential decay rate for the 2nd moment estimates. Defaults to 0.999.
epsilon: A small constant for numerical stability. This is "epsilon hat" in the Kingma and Ba paper (in the formula just before Section 2.1), not the epsilon in Algorithm 1 of the paper. It is used as in AdamW. Defaults to 1e-7.
weight_decay: Float. If set, weight decay is applied.
clipnorm: Float. If set, the gradient of each weight is individually clipped so that its norm is no higher than this value.
clipvalue: Float. If set, the gradient of each weight is clipped to be no higher than this value.
global_clipnorm: Float. If set, the gradient of all weights is clipped so that their global norm is no higher than this value.
use_ema: Boolean, defaults to FALSE. If TRUE, exponential moving average (EMA) is applied. EMA consists of computing an exponential moving average of the weights of the model (as the weight values change after each training batch), and periodically overwriting the weights with their moving average.
ema_momentum: Float, defaults to 0.99. Only used if use_ema = TRUE. This is the momentum to use when computing the EMA of the model's weights: new_average = ema_momentum * old_average + (1 - ema_momentum) * current_variable_value.
ema_overwrite_frequency: Int or NULL, defaults to NULL. Only used if use_ema = TRUE. Every ema_overwrite_frequency steps of iterations, we overwrite the model variable by its moving average. If NULL, the optimizer does not overwrite model variables in the middle of training, and you need to explicitly overwrite the variables at the end of training by calling optimizer$finalize_variable_values() (which updates the model variables in-place). When using the built-in fit() training loop, this happens automatically after the last epoch, and you don't need to do anything.
loss_scale_factor: Float or NULL. If a float, the scale factor will be multiplied the loss before computing gradients, and the inverse of the scale factor will be multiplied by the gradients before updating variables. Useful for preventing underflow during mixed precision training. Alternately, optimizer_loss_scale() will automatically set a loss scale factor.
gradient_accumulation_steps: Int or NULL. If an int, model and optimizer variables will not be updated at every step; instead they will be updated every gradient_accumulation_steps steps, using the average value of the gradients since the last update. This is known as "gradient accumulation". This can be useful when your batch size is very small, in order to reduce gradient noise at each update step. EMA frequency will look at "accumulated" iterations value (optimizer steps // gradient_accumulation_steps). Learning rate schedules will look at "real" iterations value (optimizer steps).
name: String, name for the object
exclude_layers: List of strings, keywords of layer names to exclude. All layers with keywords in their path will use AdamW.
exclude_embeddings: Boolean value. If TRUE, embedding layers will use AdamW.
muon_a: Float, parameter a of the muon algorithm. It is recommended to use the default value.
muon_b: Float, parameter b of the muon algorithm. It is recommended to use the default value.
muon_c: Float, parameter c of the muon algorithm. It is recommended to use the default value.
adam_lr_ratio: Float, the ratio of the learning rate when using Adam to the main learning rate. it is recommended to set it to 0.1.
momentum: Float, momentum used by internal SGD.
ns_steps: Integer, number of Newton-Schulz iterations to run.
nesterov: Boolean, whether to use Nesterov-style momentum.
...: For forward/backward compatibility.

Value

an Optimizer instance

Optimizer that implements the Muon algorithm.

Usage

Arguments

Value

References

See also