Megatron-LM
Megatron-LM copied to clipboard
Why is gather_output not supported in ColumnParallelLinear when using sequence parallelism?
https://github.com/NVIDIA/Megatron-LM/blob/6bf8448ba065a0a37b2b874f49fd65ca9547b5c0/megatron/core/tensor_parallel/layers.py#L907