BatchalignHK 0.7.19.post12__tar.gz → 0.7.19.post14__tar.gz

{batchalignhk-0.7.19.post12 → batchalignhk-0.7.19.post14}/BatchalignHK.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: BatchalignHK
-Version: 0.7.19.post12
+Version: 0.7.19.post14
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalignhk-0.7.19.post12 → batchalignhk-0.7.19.post14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: BatchalignHK
-Version: 0.7.19.post12
+Version: 0.7.19.post14
 Summary: Python Speech Language Sample Analysis
 Author: Brian MacWhinney, Houjun Liu
 Author-email: macw@cmu.edu, houjun@cmu.edu

{batchalignhk-0.7.19.post12 → batchalignhk-0.7.19.post14}/batchalign/cli/cli.py

@@ -111,7 +111,8 @@ batchalign.add_command(train, "models")
 @click.option("--wav2vec/--whisper_fa",
               default=True, help="Use Whisper instead of Wav2Vec for English (defaults for Whisper for non-English)")
 @click.option("--pauses", type=bool, default=False, help="Should we try to bullet each word or should we try to add pauses in between words by grouping them? Default: no pauses.", is_flag=True)
-
+@click.option("--wor/--nowor",
+              default=True, help="Should we write word level alignment line? Default to yes.")
 @click.pass_context
 def align(ctx, in_dir, out_dir, whisper, wav2vec, **kwargs):
     """Align transcripts against corresponding media files."""
@@ -122,7 +123,7 @@ def align(ctx, in_dir, out_dir, whisper, wav2vec, **kwargs):
         )
 
     def writer(doc, output):
-        CHATFile(doc=doc).write(output)
+        CHATFile(doc=doc).write(output, write_wor=kwargs.get("wor", True))
 
     if not wav2vec:
         _dispatch("align", "eng", 1,
@@ -344,7 +345,9 @@ def benchmark(ctx, in_dir, out_dir, lang, num_speakers, whisper, tencent, **kwar
         cha = p.with_suffix(".cha")
         # if there are not cha file found, we complain
         if not cha.exists():
-            raise FileNotFoundError(f"No gold .cha transcript found, we cannot do benchmarking. audio: {p.name}, desired cha: {cha.name}, looked in: {str(cha)}.")
+            cha = (Path(in_dir)/p.name).with_suffix(".cha")
+            if not cha.exists():
+                raise FileNotFoundError(f"No gold .cha transcript found, we cannot do benchmarking. audio: {p.name}, desired cha: {cha.name}, looked in: {str(cha)}.")
         # otherwise, load the goald along with the input file
         return file, {"gold": CHATFile(path=str(cha), special_mor_=True).doc}
 

batchalignhk-0.7.19.post14/batchalign/version

@@ -0,0 +1,3 @@
+0.7.19-post.14
+May 27th, 2025
+--wor flag

batchalignhk-0.7.19.post12/batchalign/version

@@ -1,3 +0,0 @@
-0.7.19-post.12
-May 27th, 2025
-tencent: tokenize roman words together