Change whisper-subtitles-generation.ipynb to genai pipeline

openvinotoolkit · Oct 3, 2024 · dc2c6bc · dc2c6bc
1 parent af144be
commit dc2c6bc
Show file tree

Hide file tree

Showing 2 changed files with 221 additions and 591 deletions.
diff --git a/notebooks/whisper-asr-genai/whisper-asr-genai.ipynb b/notebooks/whisper-asr-genai/whisper-asr-genai.ipynb
@@ -942,8 +942,7 @@
     "+model = OVModelForSpeechSeq2Seq.from_pretrained(model_id, export=True)\n",
     "```\n",
     "\n",
-    "Like the original PyTorch model, the OpenVINO model is also compatible with HuggingFace [pipeline](https://huggingface.co/docs/transformers/main_classes/pipelines#transformers.AutomaticSpeechRecognitionPipeline) interface for `automatic-speech-recognition`. \n",
-    "Pipeline can be used for long audio transcription. Distil-Whisper uses a chunked algorithm to transcribe long-form audio files. In practice, this chunked long-form algorithm is 9x faster than the sequential algorithm proposed by OpenAI in the Whisper paper. To enable chunking, pass the chunk_length_s parameter to the pipeline. For Distil-Whisper, a chunk length of 15 seconds is optimal. To activate batching, pass the argument batch_size."
+    "Like the original PyTorch model, the OpenVINO model is also compatible with HuggingFace [pipeline](https://huggingface.co/docs/transformers/main_classes/pipelines#transformers.AutomaticSpeechRecognitionPipeline) interface for `automatic-speech-recognition`. "
    ]
   },
   {
@@ -1049,14 +1048,6 @@
     "from datasets import load_dataset\n",
     "from tqdm.notebook import tqdm\n",
     "\n",
-    "def extract_input_features(sample):\n",
-    "    input_features = processor(\n",
-    "        sample[\"audio\"][\"array\"],\n",
-    "        sampling_rate=sample[\"audio\"][\"sampling_rate\"],\n",
-    "        return_tensors=\"pt\",\n",
-    "    ).input_features\n",
-    "    return input_features\n",
-    "\n",
     "\n",
     "\n",
     "CALIBRATION_DATASET_SIZE = 30\n",