From 0982e09066a3c31cb8b2fc32b5ecbc2bb64952fb Mon Sep 17 00:00:00 2001
From: Gustaf Rydholm <gustaf.rydholm@gmail.com>
Date: Thu, 2 Jun 2022 23:43:37 +0200
Subject: Add conformer block

---
 text_recognizer/networks/conformer/block.py | 34 +++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

(limited to 'text_recognizer/networks/conformer')

diff --git a/text_recognizer/networks/conformer/block.py b/text_recognizer/networks/conformer/block.py
index e69de29..d9782e8 100644
--- a/text_recognizer/networks/conformer/block.py
+++ b/text_recognizer/networks/conformer/block.py
@@ -0,0 +1,34 @@
+"""Conformer block."""
+from copy import deepcopy
+from typing import Optional
+
+from torch import nn, Tensor
+from text_recognizer.networks.conformer.conv import ConformerConv
+
+from text_recognizer.networks.conformer.mlp import MLP
+from text_recognizer.networks.conformer.scale import Scale
+from text_recognizer.networks.transformer.attention import Attention
+from text_recognizer.networks.transformer.norm import PreNorm
+
+
+class ConformerBlock(nn.Module):
+    def __init__(
+        self,
+        dim: int,
+        ff: MLP,
+        attn: Attention,
+        conv: ConformerConv,
+    ) -> None:
+        super().__init__()
+        self.attn = PreNorm(dim, attn)
+        self.ff_1 = Scale(0.5, ff)
+        self.ff_2 = deepcopy(self.ff_1)
+        self.conv = conv
+        self.post_norm = nn.LayerNorm(dim)
+
+    def forward(self, x: Tensor, mask: Optional[Tensor]) -> Tensor:
+        x = self.ff_1(x) + x
+        x = self.attn(x, mask=mask) + x
+        x = self.conv(x) + x
+        x = self.ff_2(x) + x
+        return self.post_norm(x)
-- 
cgit v1.2.3-70-g09d2