in optimum/tpu/modeling_llama.py [0:0]
def __init__(self, config, rank=0, world_size=1):
super().__init__()
self.config = config
self.hidden_size = config.hidden_size
self.intermediate_size = config.intermediate_size
self.rank = rank
self.world_size = world_size
self.gate_proj = ColumnParallelLinear.create(
self.hidden_size,
self.intermediate_size,
bias=False,
rank=rank,
world_size=world_size,
)
self.up_proj = ColumnParallelLinear.create(
self.hidden_size,
self.intermediate_size,
bias=False,
rank=rank,
world_size=world_size,
)
self.down_proj = RowParallelLinear.create(
self.intermediate_size,
self.hidden_size,
bias=False,
rank=rank,
world_size=world_size,
)
self.act_fn = ACT2FN[config.hidden_act]