word-level timestamps in transcribe() (#869)

* word-level timestamps in `transcribe()`

* moving to `timing.py`

* numba implementation for dtw, replacing dtw-python

* triton implementation for dtw

* add test for dtw implementations

* triton implementation of median_filter

* a simple word-level timestamps test

* add scipy as dev dependency

* installs an older version of Triton if CUDA < 11.4

* fix broken merge

* loosen nvcc version match regex

* find_alignment() function

* miscellaneous improvements

* skip median filtering when the input is too small

* Expose punctuation options in cli and transcribe() (#973)

* fix merge error

* fix merge error 2

* annotating that word_timestamps is experimental

---------

Co-authored-by: ryanheise <ryan@ryanheise.com>

tests/test_transcribe.py

@@ -13,10 +13,22 @@ def test_transcribe(model_name: str):
     audio_path = os.path.join(os.path.dirname(__file__), "jfk.flac")
 
     language = "en" if model_name.endswith(".en") else None
-    result = model.transcribe(audio_path, language=language, temperature=0.0)
+    result = model.transcribe(audio_path, language=language, temperature=0.0, word_timestamps=True)
     assert result["language"] == "en"
 
     transcription = result["text"].lower()
     assert "my fellow americans" in transcription
     assert "your country" in transcription
     assert "do for you" in transcription
+
+    timing_checked = False
+    for segment in result["segments"]:
+        for timing in segment["words"]:
+            assert timing["start"] < timing["end"]
+            if timing["word"].strip(" ,") == "Americans":
+                assert timing["start"] <= 1.8
+                assert timing["end"] >= 1.8
+                print(timing)
+                timing_checked = True
+
+    assert timing_checked