| # TODO(yair): Currenly not compatible with grad clipping | |
| _target_: lightning.pytorch.strategies.FSDPStrategy | |
| sharding_strategy: SHARD_GRAD_OP | |
| # TODO(yair): Currenly not compatible with grad clipping | |
| _target_: lightning.pytorch.strategies.FSDPStrategy | |
| sharding_strategy: SHARD_GRAD_OP | |