Fix for setting adam_beta3 and adam_epsilon2 for CAME Optimizer (#2654) [skip ci]
* make setting `adam_beta3` and `adam_epsilon2` work correctly * update config docs so users know args are specific to CAME optim --------- Co-authored-by: Wing Lian <wing@axolotl.ai>
This commit is contained in:
@@ -633,7 +633,9 @@ weight_decay:
|
||||
# adamw hyperparams
|
||||
adam_beta1:
|
||||
adam_beta2:
|
||||
adam_beta3: # only used for CAME Optimizer
|
||||
adam_epsilon:
|
||||
adam_epsilon2: # only used for CAME Optimizer
|
||||
# Gradient clipping max norm
|
||||
max_grad_norm:
|
||||
|
||||
|
||||
Reference in New Issue
Block a user