From b1ef434983e874130d5ff8be3ec4f37740b2ba18 Mon Sep 17 00:00:00 2001
From: xiongxinlei <xiongxinlei@baidu.com>
Date: Tue, 17 May 2022 17:38:56 +0800
Subject: [PATCH] update the max len compute method, test=doc

---
 paddlespeech/cli/asr/infer.py       | 21 +++++++++++++++------
 paddlespeech/s2t/modules/encoder.py |  2 +-
 2 files changed, 16 insertions(+), 7 deletions(-)

diff --git a/paddlespeech/cli/asr/infer.py b/paddlespeech/cli/asr/infer.py
index 0569653a..863a933f 100644
--- a/paddlespeech/cli/asr/infer.py
+++ b/paddlespeech/cli/asr/infer.py
@@ -187,13 +187,7 @@ class ASRExecutor(BaseExecutor):
                     vocab=self.config.vocab_filepath,
                     spm_model_prefix=self.config.spm_model_prefix)
                 self.config.decode.decoding_method = decode_method
-                self.max_len = 5000
-                if self.config.encoder_conf.get("max_len", None):
-                    self.max_len = self.config.encoder_conf.max_len
 
-                logger.info(f"max len: {self.max_len}")
-                # we assumen that the subsample rate is 4 and every frame step is 40ms
-                self.max_len = 40 * self.max_len / 1000
             else:
                 raise Exception("wrong type")
         model_name = model_type[:model_type.rindex(
@@ -208,6 +202,21 @@ class ASRExecutor(BaseExecutor):
         model_dict = paddle.load(self.ckpt_path)
         self.model.set_state_dict(model_dict)
 
+        # compute the max len limit
+        if "conformer" in model_type or "transformer" in model_type or "wenetspeech" in model_type:
+            # in transformer like model, we may use the subsample rate cnn network
+            subsample_rate = self.model.subsampling_rate()
+            frame_shift_ms = self.config.preprocess_config.process[0][
+                'n_shift'] / self.config.preprocess_config.process[0]['fs']
+            max_len = self.model.encoder.embed.pos_enc.max_len
+
+            if self.config.encoder_conf.get("max_len", None):
+                max_len = self.config.encoder_conf.max_len
+
+            self.max_len = frame_shift_ms * max_len * subsample_rate
+            logger.info(
+                f"The asr server limit max duration len: {self.max_len}")
+
     def preprocess(self, model_type: str, input: Union[str, os.PathLike]):
         """
         Input preprocess and return paddle.Tensor stored in self.input.
diff --git a/paddlespeech/s2t/modules/encoder.py b/paddlespeech/s2t/modules/encoder.py
index 8266a2bc..669a12d6 100644
--- a/paddlespeech/s2t/modules/encoder.py
+++ b/paddlespeech/s2t/modules/encoder.py
@@ -332,7 +332,7 @@ class BaseEncoder(nn.Layer):
         # fake mask, just for jit script and compatibility with `forward` api
         masks = paddle.ones([1, ys.shape[1]], dtype=paddle.bool)
         masks = masks.unsqueeze(1)
-        return ys, masks, offset
+        return ys, masks
 
 
 class TransformerEncoder(BaseEncoder):
-- 
GitLab