Add greedy search for streaming zipformer CTC.

.github/scripts/multi-zh-hans.sh

@@ -50,6 +50,26 @@ log "----------------------------------------"
 
 ls -lh $repo/exp/
 
+log "------------------------------------------------------------"
+log "Test exported streaming ONNX CTC models (greedy search)     "
+log "------------------------------------------------------------"
+
+test_wavs=(
+DEV_T0000000000.wav
+DEV_T0000000001.wav
+DEV_T0000000002.wav
+TEST_MEETING_T0000000113.wav
+TEST_MEETING_T0000000219.wav
+TEST_MEETING_T0000000351.wav
+)
+
+for w in ${test_wavs[@]}; do
+  ./zipformer/onnx_pretrained-streaming-ctc.py \
+    --model-filename $repo/exp/ctc-epoch-20-avg-1-chunk-16-left-128.int8.onnx \
+    --tokens $repo/data/lang_bpe_2000/tokens.txt \
+    $repo/test_wavs/$w
+done
+
 log "Upload onnx CTC models to huggingface"
 url=https://huggingface.co/k2-fsa/sherpa-onnx-streaming-zipformer-ctc-multi-zh-hans-2023-12-13
 GIT_LFS_SKIP_SMUDGE=1 git clone $url

.github/workflows/multi-zh-hans.yml

@@ -4,7 +4,6 @@ on:
   push:
     branches:
       - master
-      - streaming-ctc-decoding
 
   workflow_dispatch:
 

egs/librispeech/ASR/zipformer/onnx_pretrained-streaming-ctc.py

@@ -49,7 +49,7 @@ You can use either the ``int8.onnx`` model or just the ``.onnx`` model.
 
 ./zipformer/onnx_pretrained-streaming-ctc.py \
   --model-filename $repo/exp/ctc-epoch-20-avg-1-chunk-16-left-128.onnx \
-  --tokens $repo/data/lang_bpe_500/tokens.txt \
+  --tokens $repo/data/lang_bpe_2000/tokens.txt \
   $repo/test_wavs/DEV_T0000000001.wav
 
 Note: Even though this script only supports decoding a single file,
@@ -58,9 +58,8 @@ the exported ONNX models do support batch processing.
 
 import argparse
 import logging
-from typing import Dict, List, Optional, Tuple
+from typing import Dict, List, Tuple
 
-import k2
 import numpy as np
 import onnxruntime as ort
 import torch
@@ -111,7 +110,7 @@ class OnnxModel:
 
         self.init_model(model_filename)
 
-    def init_model(self, encoder_model_filename: str):
+    def init_model(self, model_filename: str):
         self.model = ort.InferenceSession(
             model_filename,
             sess_options=self.session_opts,
@@ -207,7 +206,7 @@ class OnnxModel:
         x: torch.Tensor,
     ) -> Tuple[Dict[str, np.ndarray], List[str]]:
         model_input = {"x": x.numpy()}
-        model_output = ["model_out"]
+        model_output = ["log_probs"]
 
         def build_inputs_outputs(tensors, i):
             assert len(tensors) == 6, len(tensors)
@@ -262,18 +261,18 @@ class OnnxModel:
     def _update_states(self, states: List[np.ndarray]):
         self.states = states
 
-    def run_encoder(self, x: torch.Tensor) -> torch.Tensor:
+    def __call__(self, x: torch.Tensor) -> torch.Tensor:
         """
         Args:
           x:
             A 3-D tensor of shape (N, T, C)
         Returns:
-          Return a 3-D tensor of shape (N, T', joiner_dim) where
+          Return a 3-D tensor containing log_probs. Its shape is (N, T, vocab_size)
-          T' is usually equal to ((T-7)//2+1)//2
+          where T' is usually equal to ((T-7)//2 - 3)//2
         """
         model_input, model_output_names = self._build_model_input_output(x)
 
-        out = self.encoder.run(model_output_names, model_input)
+        out = self.model.run(model_output_names, model_input)
 
         self._update_states(out[1:])
 
@@ -323,51 +322,24 @@ def create_streaming_feature_extractor() -> OnlineFeature:
 
 
 def greedy_search(
-    model: OnnxModel,
+    log_probs: torch.Tensor,
-    model_out: torch.Tensor,
-    context_size: int,
-    decoder_out: Optional[torch.Tensor] = None,
-    hyp: Optional[List[int]] = None,
 ) -> List[int]:
-    """Greedy search in batch mode. It hardcodes --max-sym-per-frame=1.
+    """Greedy search for a single utterance.
     Args:
-      model:
+      log_probs:
-        The transducer model.
+        A 3-D tensor of shape (1, T, vocab_size)
-      model_out:
-        A 3-D tensor of shape (1, T, joiner_dim)
-      context_size:
-        The context size of the decoder model.
-      decoder_out:
-        Optional. Decoder output of the previous chunk.
-      hyp:
-        Decoding results for previous chunks.
     Returns:
-      Return the decoded results so far.
+      Return the decoded result.
     """
+    assert log_probs.ndim == 3, log_probs.shape
+    assert log_probs.shape[0] == 1, log_probs.shape
+
+    max_indexes = log_probs[0].argmax(dim=1)
+    unique_indexes = torch.unique_consecutive(max_indexes)
 
     blank_id = 0
-
+    unique_indexes = unique_indexes[unique_indexes != blank_id]
-    if decoder_out is None:
+    return unique_indexes.tolist()
-        assert hyp is None, hyp
-        hyp = [blank_id] * context_size
-        decoder_input = torch.tensor([hyp], dtype=torch.int64)
-        decoder_out = model.run_decoder(decoder_input)
-    else:
-        assert hyp is not None, hyp
-
-    model_out = model_out.squeeze(0)
-    T = model_out.size(0)
-    for t in range(T):
-        cur_encoder_out = model_out[t : t + 1]
-        joiner_out = model.run_joiner(cur_encoder_out, decoder_out).squeeze(0)
-        y = joiner_out.argmax(dim=0).item()
-        if y != blank_id:
-            hyp.append(y)
-            decoder_input = hyp[-context_size:]
-            decoder_input = torch.tensor([decoder_input], dtype=torch.int64)
-            decoder_out = model.run_decoder(decoder_input)
-
-    return hyp, decoder_out
 
 
 @torch.no_grad()
@@ -376,11 +348,7 @@ def main():
     args = parser.parse_args()
     logging.info(vars(args))
 
-    model = OnnxModel(
+    model = OnnxModel(model_filename=args.model_filename)
-        encoder_model_filename=args.encoder_model_filename,
-        decoder_model_filename=args.decoder_model_filename,
-        joiner_model_filename=args.joiner_model_filename,
-    )
 
     sample_rate = 16000
 
@@ -400,9 +368,7 @@ def main():
     segment = model.segment
     offset = model.offset
 
-    context_size = model.context_size
+    hyp = []
-    hyp = None
-    decoder_out = None
 
     chunk = int(1 * sample_rate)  # 1 second
     start = 0
@@ -423,18 +389,28 @@ def main():
             num_processed_frames += offset
             frames = torch.cat(frames, dim=0)
             frames = frames.unsqueeze(0)
-            model_out = model.run_encoder(frames)
+            log_probs = model(frames)
-            hyp = greedy_search(
-                model,
-                model_out,
-                hyp,
-            )
 
-    token_table = k2.SymbolTable.from_file(args.tokens)
+            hyp += greedy_search(log_probs)
 
-    text = ""
+    # To handle byte-level BPE, we convert string tokens to utf-8 encoded bytes
-    for i in hyp[context_size:]:
+    id2token = {}
-        text += token_table[i]
+    with open(args.tokens, encoding="utf-8") as f:
+        for line in f:
+            token, idx = line.split()
+            if token[:3] == "<0x" and token[-1] == ">":
+                token = int(token[1:-1], base=16)
+                assert 0 <= token < 256, token
+                token = token.to_bytes(1, byteorder="little")
+            else:
+                token = token.encode(encoding="utf-8")
+
+            id2token[int(idx)] = token
+
+    text = b""
+    for i in hyp:
+        text += id2token[i]
+    text = text.decode(encoding="utf-8")
     text = text.replace("▁", " ").strip()
 
     logging.info(args.sound_file)

egs/librispeech/ASR/zipformer/onnx_pretrained-streaming.py

@@ -326,7 +326,7 @@ class OnnxModel:
             A 3-D tensor of shape (N, T, C)
         Returns:
           Return a 3-D tensor of shape (N, T', joiner_dim) where
-          T' is usually equal to ((T-7)//2+1)//2
+          T' is usually equal to ((T-7)//2-3)//2
         """
         encoder_input, encoder_output_names = self._build_encoder_input_output(x)