move AED chunked infer script (#9367)

Signed-off-by: stevehuang52 <heh@nvidia.com>

examples/asr/asr_chunked_inference/README.md

@@ -1,6 +1,6 @@
-# Streaming / Buffered ASR
+# Streaming / Buffered / Chunked ASR
 
-Contained within this directory are scripts to perform streaming or buffered inference of audio files using CTC / Transducer ASR models.
+Contained within this directory are scripts to perform streaming or buffered inference of audio files using CTC / Transducer ASR models, and chunked inference for MultitaskAED models (e.g., "nvidia/canary-1b").
 
 ## Difference between streaming and buffered ASR
 
@@ -9,3 +9,7 @@ While we primarily showcase the defaults of these models in buffering mode, note
 If you reduce your chunk size, the latency for your first prediction is reduced, and the model appears to predict the text with shorter delay. On the other hand, since the amount of information in the chunk is reduced, it causes higher WER.
 
 On the other hand, if you increase your chunk size, then the delay between spoken sentence and the transcription increases (this is buffered ASR). While the latency is increased, you are able to obtain more accurate transcripts since the model has more context to properly transcribe the text.
+
+## Chunked Inference
+
+For MultitaskAED models, we provide a script to perform chunked inference. This script will split the input audio into non-overlapping chunks and perform inference on each chunk. The script will then concatenate the results to provide the final transcript.

examples/asr/speech_multitask/speech_to_text_aed_chunked_infer.py → examples/asr/asr_chunked_inference/aed/speech_to_text_aed_chunked_infer.py

@@ -88,7 +88,9 @@ class TranscriptionConfig:
 
     # Chunked configs
     chunk_len_in_secs: float = 40.0  # Chunk length in seconds
-    model_stride: int = 8  # Model downsampling factor, 8 for Citrinet and FasConformer models and 4 for Conformer models.
+    model_stride: int = (
+        8  # Model downsampling factor, 8 for Citrinet and FasConformer models and 4 for Conformer models.
+    )
 
     # Decoding strategy for MultitaskAED models
     decoding: MultiTaskDecodingConfig = MultiTaskDecodingConfig()
@@ -209,7 +211,12 @@ def autocast(*args, **kwargs):
     with autocast(dtype=amp_dtype):
         with torch.no_grad():
             hyps = get_buffered_pred_feat_multitaskAED(
-                frame_asr, model_cfg.preprocessor, model_stride_in_secs, asr_model.device, manifest, filepaths,
+                frame_asr,
+                model_cfg.preprocessor,
+                model_stride_in_secs,
+                asr_model.device,
+                manifest,
+                filepaths,
             )
 
     output_filename, pred_text_attr_name = write_transcription(