BatchalignHK 0.8.0.post3__tar.gz → 0.8.0.post5__tar.gz

{batchalignhk-0.8.0.post3 → batchalignhk-0.8.0.post5}/BatchalignHK.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: BatchalignHK
-Version: 0.8.0.post3
+Version: 0.8.0.post5
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalignhk-0.8.0.post3 → batchalignhk-0.8.0.post5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: BatchalignHK
-Version: 0.8.0.post3
+Version: 0.8.0.post5
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalignhk-0.8.0.post3 → batchalignhk-0.8.0.post5}/batchalign/cli/cli.py

@@ -158,6 +158,8 @@ def align(ctx, in_dir, out_dir, whisper, wav2vec, iic, wav2vec_yue, tencent, fun
               default=False, help="Use Alibaba instead of Rev.AI (default). Superceeds --whisper.")
 @click.option("--funaudio/--rev",
               default=False, help="Use FunAudio instead of Rev.AI (default). Superceeds --whisper.")
+@click.option("--paraformer/--rev",
+              default=False, help="Use FunAudio instead of Rev.AI (default). Superceeds --whisper.")
 @click.option("--diarize/--nodiarize",
               default=False, help="Perform speaker diarization (this flag is ignored with Rev.AI)")
 @click.option("--wor/--nowor",
@@ -189,6 +191,8 @@ def transcribe(ctx, in_dir, out_dir, lang, num_speakers, **kwargs):
         asr = "aliyun"
     if kwargs["funaudio"]:
         asr = "funaudio"
+    if kwargs["paraformer"]:
+        asr = "paraformer"
 
     def writer(doc, output):
         doc.content.insert(0, CustomLine(id="Comment", type=CustomLineType.INDEPENDENT,

{batchalignhk-0.8.0.post3 → batchalignhk-0.8.0.post5}/batchalign/cli/dispatch.py

@@ -312,6 +312,22 @@ def _dispatch(command, lang, num_speakers,
               **kwargs):
 
     C = console
+    worker_handled = {
+        "align",
+        "transcribe",
+        "transcribe_s",
+        "translate",
+        "morphotag",
+        "utseg",
+        "coref",
+        "benchmark",
+        "opensmile",
+    }
+    if command in worker_handled:
+        # Avoid pickling CLI-local loader/writer functions when the worker
+        # implements the command-specific IO logic.
+        loader = None
+        writer = None
     from batchalign.constants import FORCED_CONVERSION
     from batchalign.document import TaskFriendlyName
 

{batchalignhk-0.8.0.post3 → batchalignhk-0.8.0.post5}/batchalign/pipelines/dispatch.py

@@ -152,7 +152,9 @@ def dispatch_pipeline(pkg_str, lang, num_speakers=None, **arg_overrides):
         elif engine == "aliyun":
             engines.append(AliyunEngine())
         elif engine == "funaudio":
-            engines.append(FunAudioEngine())
+            engines.append(FunAudioEngine(model="FunAudioLLM/SenseVoiceSmall"))
+        elif engine == "paraformer":
+            engines.append(FunAudioEngine(model="paraformer-zh"))
         elif engine ==  "funaudio_utr":
             engines.append(FunAudioUTREngine(lang=lang))
         elif engine == "pyannote":

batchalignhk-0.8.0.post5/batchalign/version

@@ -0,0 +1,3 @@
+0.8.0-post.5
+Jan 20th, 2025
+Funaudio support

batchalignhk-0.8.0.post3/batchalign/version

@@ -1,3 +0,0 @@
-0.8.0-post.3
-Jan 16th, 2025
-Patch regression?