MCPcopy Create free account
hub / github.com/Xiaobin-Rong/gtcrn / forward

Method forward

stream/modules/convolution.py:85–93  ·  view source on GitHub ↗

x: [bs, C, 1, F] cache: [bs, C, T_size-1, F]

(self, x, cache)

Source from the content-addressed store, hash-verified

83 bias = bias)
84
85 def forward(self, x, cache):
86 """
87 x: [bs, C, 1, F]
88 cache: [bs, C, T_size-1, F]
89 """
90 inp = torch.cat([cache, x], dim=2)
91 outp = self.Conv2d(inp)
92 out_cache = inp[:,:, 1:]
93 return outp, out_cache
94
95## Version 1
96## The inference of this implementation is slow, according to https://github.com/Xiaobin-Rong/gtcrn/issues/37

Callers

nothing calls this directly

Calls

no outgoing calls

Tested by

no test coverage detected