fix: disable gradient checkpointing (device mismatch with distributed MoE)
This commit is contained in:
@@ -53,5 +53,5 @@ mixed_precision: bf16
|
|||||||
# Evaluation (disable - no test split in dataset)
|
# Evaluation (disable - no test split in dataset)
|
||||||
eval_strategy: "no"
|
eval_strategy: "no"
|
||||||
|
|
||||||
# Gradient Checkpointing
|
# Gradient Checkpointing (disable - causes device issues with distributed MoE)
|
||||||
gradient_checkpointing: true
|
gradient_checkpointing: false
|
||||||
|
|||||||
Reference in New Issue
Block a user