Codestin Search App

naymaraq · 2025-09-26T09:38:41Z

Important

The Update branch button must only be pressed in very rare occassions.
An outdated branch is never blocking the merge of a PR.
Please reach out to the automation team before pressing that button.

What does this PR do ?

This PR contains inference support for streaming ASR, including Buffered CTC/RNN-T/TDT and Cache-Aware CTC/RNN-T

Collection: [ASR]

Changelog

Add specific line by line info of high level changes in this PR.

Support for Buffered CTC / RNN-T / TDT + Cache-Aware CTC / RNN-T
Word and Segment level metadata - timestamps, conf scores
EoU support - Blank based and VAD based
For buffered RNN-T/TDT - stateful and stateless modes
For Cache-Aware - chunk size configurability
Streaming ITN and Streaming PnC
Only greedy inference for now
All of them are portable in RIVA

Usage

You can potentially add a usage example below

# Add a code snippet demonstrating how to use this

GitHub Actions CI

The Jenkins CI system has been replaced by GitHub Actions self-hosted runners.

The GitHub Actions CI will run automatically when the "Run CICD" label is added to the PR.
To re-run CI remove and add the label again.
To run CI on an untrusted fork, a NeMo user with write access must first click "Approve and run".

Before your PR is "Ready for review"

Pre checks:

Make sure you read and followed Contributor guidelines
Did you write any new necessary tests?
Did you add or update any necessary documentation?
Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- Reviewer: Does the PR have correct import guards for all optional libraries?

PR Type:

New Feature
Bugfix
Documentation

If you haven't finished some of the above items you can still open "Draft" PR.

Who can review?

Anyone in the community is free to review the PR once the checks have passed.
Contributor guidelines contains specific people who can review PRs to various areas.

Additional Information

Related to # (issue)

Signed-off-by: naymaraq <[email protected]>

chtruong814 · 2025-11-01T20:23:20Z

Last test failure was unrelated to this change. Moving forward with merging it.

* init inference folders Signed-off-by: naymaraq <[email protected]> * added base asr inference Signed-off-by: naymaraq <[email protected]> * add ctc and rnnt inference classes Signed-off-by: naymaraq <[email protected]> * small changes for ctc/rnnt inference Signed-off-by: naymaraq <[email protected]> * add cache aware ctc/rnnt inference classes Signed-off-by: naymaraq <[email protected]> * finilize asr inference part Signed-off-by: naymaraq <[email protected]> * add word class Signed-off-by: naymaraq <[email protected]> * add enums file Signed-off-by: naymaraq <[email protected]> * add alignment preserving itn Signed-off-by: naymaraq <[email protected]> * add punctuation/capitalization model Signed-off-by: naymaraq <[email protected]> * add audio_io and progressbar files Signed-off-by: naymaraq <[email protected]> * add framing and buffering files Signed-off-by: naymaraq <[email protected]> * mv common/inference/utils into asr/inference/utils Signed-off-by: naymaraq <[email protected]> * add StreamingState objects Signed-off-by: naymaraq <[email protected]> * temporary rm enhancement stuff Signed-off-by: naymaraq <[email protected]> * rm common/inference Signed-off-by: naymaraq <[email protected]> * add greedy decoders for CTC/RNNT Signed-off-by: naymaraq <[email protected]> * add endpointing files Signed-off-by: naymaraq <[email protected]> * add text processing Signed-off-by: naymaraq <[email protected]> * mv itn_utils into utils Signed-off-by: naymaraq <[email protected]> * add bpe_decoder, context_manager for cache aware, recognizer_utils Signed-off-by: naymaraq <[email protected]> * add base_recognizer and recognizer interface files Signed-off-by: naymaraq <[email protected]> * add recognizers Signed-off-by: naymaraq <[email protected]> * add factory Signed-off-by: naymaraq <[email protected]> * add inference example and asr_client.py Signed-off-by: naymaraq <[email protected]> * minor fix Signed-off-by: naymaraq <[email protected]> * minor fixes Signed-off-by: naymaraq <[email protected]> * add example usage Signed-off-by: naymaraq <[email protected]> * add jsonl support Signed-off-by: naymaraq <[email protected]> * rm niva prefix Signed-off-by: naymaraq <[email protected]> * fix docstrings Signed-off-by: naymaraq <[email protected]> * mv RequestType into enums.py Signed-off-by: naymaraq <[email protected]> * rm redundant setters Signed-off-by: naymaraq <[email protected]> * add a log_level to config.yaml Signed-off-by: naymaraq <[email protected]> * setup log_level in RecognizerBuilder Signed-off-by: naymaraq <[email protected]> * add comments in multi stream and fix docstrings in buffering Signed-off-by: naymaraq <[email protected]> * conditional import for diskcache Signed-off-by: naymaraq <[email protected]> * set log level to INFO Signed-off-by: naymaraq <[email protected]> * add MPS device support Signed-off-by: naymaraq <[email protected]> * add tests Signed-off-by: naymaraq <[email protected]> * move inference into examples/asr/asr_chunked_inference/ctc Signed-off-by: naymaraq <[email protected]> * rm duplicated create_partial_transcript method Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> Signed-off-by: naymaraq <[email protected]> * resolve flake8 errors Signed-off-by: naymaraq <[email protected]> * resolve return type Signed-off-by: naymaraq <[email protected]> * fix imports in tests Signed-off-by: naymaraq <[email protected]> * optimize bpe_decoder Signed-off-by: naymaraq <[email protected]> * optimize log prob normalization Signed-off-by: naymaraq <[email protected]> * optimize split_text function Signed-off-by: naymaraq <[email protected]> * fix parital batching, improved GPU utilization Signed-off-by: naymaraq <[email protected]> * simplify ctc greedy decoder Signed-off-by: naymaraq <[email protected]> * add a method to perform ITN on a list of texts Signed-off-by: naymaraq <[email protected]> * remove duplicated code in enums Signed-off-by: naymaraq <[email protected]> * remove unnecessary pad_to logging Signed-off-by: naymaraq <[email protected]> * modified update_punctuation_and_language_tokens_timestamps function to ensure correct global timestamps for eou calculation Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] conditional import for pynini and nemo_text_processing Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] fix configs, added asr_output_granularity Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] write segment/word level output into json instead of ctm Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add output granuality to request options Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add segment related fields to state Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add remove repeated punctuation function Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add TextSegment class Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] update bpe decoder to support text segment Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] update recognizers Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] update text processing to support segment-level output Signed-off-by: naymaraq <[email protected]> * rm unused and duplicated code Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * code cleanup Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * rm unused code and code cleanup Signed-off-by: naymaraq <[email protected]> * Set num_slots to 1024 and add a num_slots parameter to the config files Signed-off-by: naymaraq <[email protected]> * removed hyp.alignment processing codes Signed-off-by: naymaraq <[email protected]> * disable amp Signed-off-by: naymaraq <[email protected]> * mv diskcache req into requirements_asr.txt Signed-off-by: naymaraq <[email protected]> * set use_amp to true and make typing consistent Signed-off-by: naymaraq <[email protected]> * use match/case for readability Signed-off-by: naymaraq <[email protected]> * rm lambdas from punctuation_capitalization_config.py Signed-off-by: naymaraq <[email protected]> * rm detect_eou method from RNNTGreedyEndpointing Signed-off-by: naymaraq <[email protected]> * reuse read_manifest from manifest_utils Signed-off-by: naymaraq <[email protected]> * use librosa instead of soundfile Signed-off-by: naymaraq <[email protected]> * unfreeze ASRRequestOptions dataclass Signed-off-by: naymaraq <[email protected]> * set use_amp to false for buffered CTC/RNNT recognizers, improved throughput Signed-off-by: naymaraq <[email protected]> * change matmul precision to high for cache aware models Signed-off-by: naymaraq <[email protected]> * optimized audio buffer shifting Signed-off-by: naymaraq <[email protected]> * Move running scripts and YAML files out of the ctc folder Signed-off-by: naymaraq <[email protected]> * reorganize file structure Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * Minor code simplifications Signed-off-by: naymaraq <[email protected]> * rm duplicated initializations from recognizers Signed-off-by: naymaraq <[email protected]> * remove package version for diskcache Signed-off-by: naymaraq <[email protected]> * move tqdm import to the top Signed-off-by: naymaraq <[email protected]> * simplify millisecond_to_frames function Signed-off-by: naymaraq <[email protected]> * raise a ValueError in case of stream_id > n_audio_files Signed-off-by: naymaraq <[email protected]> * fix return types Signed-off-by: naymaraq <[email protected]> * use list/dict/... instead of List/Dict/... Signed-off-by: naymaraq <[email protected]> * use keyword argument passing to create CacheFeatureBufferer Signed-off-by: naymaraq <[email protected]> * clean up state resetting logic Signed-off-by: naymaraq <[email protected]> * reuse normalize_batch Signed-off-by: naymaraq <[email protected]> * rename verbatim_transcripts and automatic_punctuation Signed-off-by: naymaraq <[email protected]> * rename recognizers to pipelines Signed-off-by: naymaraq <[email protected]> * rename asr/*_inference -> model_wrappers/*_inference_wrapper Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * reorgonize pnc, itn, text_processing params Signed-off-by: naymaraq <[email protected]> * improved code readability in pipeline initializations Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * add CI script for testing Signed-off-by: naymaraq <[email protected]> * add output_dir in CI test Signed-off-by: naymaraq <[email protected]> * move python running script into new folder Signed-off-by: naymaraq <[email protected]> * renamed asr_streaming_infer -> asr_streaming_inference Signed-off-by: naymaraq <[email protected]> * correct path in CI test Signed-off-by: naymaraq <[email protected]> * fix: variable may be used before it is initialized Signed-off-by: naymaraq <[email protected]> * fix docstring in itn/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in model_wrappers/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in utils/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in pipelines/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in streaming/ folder Signed-off-by: naymaraq <[email protected]> * remove PnC codes since nlp models are no longer supported Signed-off-by: naymaraq <[email protected]> * minor changes Signed-off-by: naymaraq <[email protected]> * return step output from transcribe_step method Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * fix functional_test Signed-off-by: naymaraq <[email protected]> * increase timeout for L0_Unit_Tests_CPU_ASR Signed-off-by: naymaraq <[email protected]> * rm cache aware inference from functional test Signed-off-by: naymaraq <[email protected]> --------- Signed-off-by: naymaraq <[email protected]> Signed-off-by: naymaraq <[email protected]> Co-authored-by: naymaraq <[email protected]> Co-authored-by: naymaraq <[email protected]> Signed-off-by: quanpham <[email protected]>

* init inference folders Signed-off-by: naymaraq <[email protected]> * added base asr inference Signed-off-by: naymaraq <[email protected]> * add ctc and rnnt inference classes Signed-off-by: naymaraq <[email protected]> * small changes for ctc/rnnt inference Signed-off-by: naymaraq <[email protected]> * add cache aware ctc/rnnt inference classes Signed-off-by: naymaraq <[email protected]> * finilize asr inference part Signed-off-by: naymaraq <[email protected]> * add word class Signed-off-by: naymaraq <[email protected]> * add enums file Signed-off-by: naymaraq <[email protected]> * add alignment preserving itn Signed-off-by: naymaraq <[email protected]> * add punctuation/capitalization model Signed-off-by: naymaraq <[email protected]> * add audio_io and progressbar files Signed-off-by: naymaraq <[email protected]> * add framing and buffering files Signed-off-by: naymaraq <[email protected]> * mv common/inference/utils into asr/inference/utils Signed-off-by: naymaraq <[email protected]> * add StreamingState objects Signed-off-by: naymaraq <[email protected]> * temporary rm enhancement stuff Signed-off-by: naymaraq <[email protected]> * rm common/inference Signed-off-by: naymaraq <[email protected]> * add greedy decoders for CTC/RNNT Signed-off-by: naymaraq <[email protected]> * add endpointing files Signed-off-by: naymaraq <[email protected]> * add text processing Signed-off-by: naymaraq <[email protected]> * mv itn_utils into utils Signed-off-by: naymaraq <[email protected]> * add bpe_decoder, context_manager for cache aware, recognizer_utils Signed-off-by: naymaraq <[email protected]> * add base_recognizer and recognizer interface files Signed-off-by: naymaraq <[email protected]> * add recognizers Signed-off-by: naymaraq <[email protected]> * add factory Signed-off-by: naymaraq <[email protected]> * add inference example and asr_client.py Signed-off-by: naymaraq <[email protected]> * minor fix Signed-off-by: naymaraq <[email protected]> * minor fixes Signed-off-by: naymaraq <[email protected]> * add example usage Signed-off-by: naymaraq <[email protected]> * add jsonl support Signed-off-by: naymaraq <[email protected]> * rm niva prefix Signed-off-by: naymaraq <[email protected]> * fix docstrings Signed-off-by: naymaraq <[email protected]> * mv RequestType into enums.py Signed-off-by: naymaraq <[email protected]> * rm redundant setters Signed-off-by: naymaraq <[email protected]> * add a log_level to config.yaml Signed-off-by: naymaraq <[email protected]> * setup log_level in RecognizerBuilder Signed-off-by: naymaraq <[email protected]> * add comments in multi stream and fix docstrings in buffering Signed-off-by: naymaraq <[email protected]> * conditional import for diskcache Signed-off-by: naymaraq <[email protected]> * set log level to INFO Signed-off-by: naymaraq <[email protected]> * add MPS device support Signed-off-by: naymaraq <[email protected]> * add tests Signed-off-by: naymaraq <[email protected]> * move inference into examples/asr/asr_chunked_inference/ctc Signed-off-by: naymaraq <[email protected]> * rm duplicated create_partial_transcript method Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> Signed-off-by: naymaraq <[email protected]> * resolve flake8 errors Signed-off-by: naymaraq <[email protected]> * resolve return type Signed-off-by: naymaraq <[email protected]> * fix imports in tests Signed-off-by: naymaraq <[email protected]> * optimize bpe_decoder Signed-off-by: naymaraq <[email protected]> * optimize log prob normalization Signed-off-by: naymaraq <[email protected]> * optimize split_text function Signed-off-by: naymaraq <[email protected]> * fix parital batching, improved GPU utilization Signed-off-by: naymaraq <[email protected]> * simplify ctc greedy decoder Signed-off-by: naymaraq <[email protected]> * add a method to perform ITN on a list of texts Signed-off-by: naymaraq <[email protected]> * remove duplicated code in enums Signed-off-by: naymaraq <[email protected]> * remove unnecessary pad_to logging Signed-off-by: naymaraq <[email protected]> * modified update_punctuation_and_language_tokens_timestamps function to ensure correct global timestamps for eou calculation Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] conditional import for pynini and nemo_text_processing Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] fix configs, added asr_output_granularity Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] write segment/word level output into json instead of ctm Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add output granuality to request options Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add segment related fields to state Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add remove repeated punctuation function Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] add TextSegment class Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] update bpe decoder to support text segment Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] update recognizers Signed-off-by: naymaraq <[email protected]> * [refactor: segment-level output] update text processing to support segment-level output Signed-off-by: naymaraq <[email protected]> * rm unused and duplicated code Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * code cleanup Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * rm unused code and code cleanup Signed-off-by: naymaraq <[email protected]> * Set num_slots to 1024 and add a num_slots parameter to the config files Signed-off-by: naymaraq <[email protected]> * removed hyp.alignment processing codes Signed-off-by: naymaraq <[email protected]> * disable amp Signed-off-by: naymaraq <[email protected]> * mv diskcache req into requirements_asr.txt Signed-off-by: naymaraq <[email protected]> * set use_amp to true and make typing consistent Signed-off-by: naymaraq <[email protected]> * use match/case for readability Signed-off-by: naymaraq <[email protected]> * rm lambdas from punctuation_capitalization_config.py Signed-off-by: naymaraq <[email protected]> * rm detect_eou method from RNNTGreedyEndpointing Signed-off-by: naymaraq <[email protected]> * reuse read_manifest from manifest_utils Signed-off-by: naymaraq <[email protected]> * use librosa instead of soundfile Signed-off-by: naymaraq <[email protected]> * unfreeze ASRRequestOptions dataclass Signed-off-by: naymaraq <[email protected]> * set use_amp to false for buffered CTC/RNNT recognizers, improved throughput Signed-off-by: naymaraq <[email protected]> * change matmul precision to high for cache aware models Signed-off-by: naymaraq <[email protected]> * optimized audio buffer shifting Signed-off-by: naymaraq <[email protected]> * Move running scripts and YAML files out of the ctc folder Signed-off-by: naymaraq <[email protected]> * reorganize file structure Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * Minor code simplifications Signed-off-by: naymaraq <[email protected]> * rm duplicated initializations from recognizers Signed-off-by: naymaraq <[email protected]> * remove package version for diskcache Signed-off-by: naymaraq <[email protected]> * move tqdm import to the top Signed-off-by: naymaraq <[email protected]> * simplify millisecond_to_frames function Signed-off-by: naymaraq <[email protected]> * raise a ValueError in case of stream_id > n_audio_files Signed-off-by: naymaraq <[email protected]> * fix return types Signed-off-by: naymaraq <[email protected]> * use list/dict/... instead of List/Dict/... Signed-off-by: naymaraq <[email protected]> * use keyword argument passing to create CacheFeatureBufferer Signed-off-by: naymaraq <[email protected]> * clean up state resetting logic Signed-off-by: naymaraq <[email protected]> * reuse normalize_batch Signed-off-by: naymaraq <[email protected]> * rename verbatim_transcripts and automatic_punctuation Signed-off-by: naymaraq <[email protected]> * rename recognizers to pipelines Signed-off-by: naymaraq <[email protected]> * rename asr/*_inference -> model_wrappers/*_inference_wrapper Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * reorgonize pnc, itn, text_processing params Signed-off-by: naymaraq <[email protected]> * improved code readability in pipeline initializations Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * add CI script for testing Signed-off-by: naymaraq <[email protected]> * add output_dir in CI test Signed-off-by: naymaraq <[email protected]> * move python running script into new folder Signed-off-by: naymaraq <[email protected]> * renamed asr_streaming_infer -> asr_streaming_inference Signed-off-by: naymaraq <[email protected]> * correct path in CI test Signed-off-by: naymaraq <[email protected]> * fix: variable may be used before it is initialized Signed-off-by: naymaraq <[email protected]> * fix docstring in itn/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in model_wrappers/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in utils/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in pipelines/ folder Signed-off-by: naymaraq <[email protected]> * fix docstring in streaming/ folder Signed-off-by: naymaraq <[email protected]> * remove PnC codes since nlp models are no longer supported Signed-off-by: naymaraq <[email protected]> * minor changes Signed-off-by: naymaraq <[email protected]> * return step output from transcribe_step method Signed-off-by: naymaraq <[email protected]> * Apply isort and black reformatting Signed-off-by: naymaraq <[email protected]> * fix functional_test Signed-off-by: naymaraq <[email protected]> * increase timeout for L0_Unit_Tests_CPU_ASR Signed-off-by: naymaraq <[email protected]> * rm cache aware inference from functional test Signed-off-by: naymaraq <[email protected]> --------- Signed-off-by: naymaraq <[email protected]> Signed-off-by: naymaraq <[email protected]> Co-authored-by: naymaraq <[email protected]> Co-authored-by: naymaraq <[email protected]>

github-actions Bot added the ASR label Sep 26, 2025

github-advanced-security AI found potential problems Sep 26, 2025

View reviewed changes

naymaraq added 28 commits September 26, 2025 15:05

init inference folders

3c8391d

Signed-off-by: naymaraq <[email protected]>

added base asr inference

1b1cb59

Signed-off-by: naymaraq <[email protected]>

add ctc and rnnt inference classes

25c57a8

Signed-off-by: naymaraq <[email protected]>

small changes for ctc/rnnt inference

cb2c7f3

Signed-off-by: naymaraq <[email protected]>

add cache aware ctc/rnnt inference classes

ab47f96

Signed-off-by: naymaraq <[email protected]>

finilize asr inference part

c59bc2b

Signed-off-by: naymaraq <[email protected]>

add word class

7c30d09

Signed-off-by: naymaraq <[email protected]>

add enums file

76ff0c4

Signed-off-by: naymaraq <[email protected]>

add alignment preserving itn

b3aeb99

Signed-off-by: naymaraq <[email protected]>

add punctuation/capitalization model

cd4eb39

Signed-off-by: naymaraq <[email protected]>

add audio_io and progressbar files

3a5bf54

Signed-off-by: naymaraq <[email protected]>

add framing and buffering files

397a950

Signed-off-by: naymaraq <[email protected]>

mv common/inference/utils into asr/inference/utils

08f5203

Signed-off-by: naymaraq <[email protected]>

add StreamingState objects

2f0717d

Signed-off-by: naymaraq <[email protected]>

temporary rm enhancement stuff

f00102c

Signed-off-by: naymaraq <[email protected]>

rm common/inference

77d5ffb

Signed-off-by: naymaraq <[email protected]>

add greedy decoders for CTC/RNNT

6e83d6b

Signed-off-by: naymaraq <[email protected]>

add endpointing files

2cbdf2d

Signed-off-by: naymaraq <[email protected]>

add text processing

c19bf72

Signed-off-by: naymaraq <[email protected]>

mv itn_utils into utils

4800a7c

Signed-off-by: naymaraq <[email protected]>

add bpe_decoder, context_manager for cache aware, recognizer_utils

194a507

Signed-off-by: naymaraq <[email protected]>

add base_recognizer and recognizer interface files

5c6e97f

Signed-off-by: naymaraq <[email protected]>

add recognizers

9ee2364

Signed-off-by: naymaraq <[email protected]>

add factory

8f65ebf

Signed-off-by: naymaraq <[email protected]>

add inference example and asr_client.py

af6e1ef

Signed-off-by: naymaraq <[email protected]>

minor fix

e401e6f

Signed-off-by: naymaraq <[email protected]>

minor fixes

18a3e3b

Signed-off-by: naymaraq <[email protected]>

add example usage

2da1769

Signed-off-by: naymaraq <[email protected]>

naymaraq dismissed stale reviews from artbataev, stevehuang52, nithinraok, and chtruong814 via 5299973 October 31, 2025 06:26

chtruong814 added Run CICD and removed Run CICD labels Oct 31, 2025

chtruong814 temporarily deployed to test October 31, 2025 06:28 — with GitHub Actions Inactive

artbataev previously approved these changes Oct 31, 2025

View reviewed changes

github-actions Bot removed the Run CICD label Oct 31, 2025

naymaraq and others added 2 commits October 31, 2025 19:06

Merge branch 'main' into dkaramyan/inference

d55ad43

increase timeout for L0_Unit_Tests_CPU_ASR

fdc168c

Signed-off-by: naymaraq <[email protected]>

naymaraq dismissed artbataev’s stale review via fdc168c October 31, 2025 15:23

naymaraq added the Run CICD label Oct 31, 2025

naymaraq temporarily deployed to test October 31, 2025 15:26 — with GitHub Actions Inactive

Merge branch 'main' into dkaramyan/inference

0328dbf

chtruong814 added Run CICD and removed Run CICD labels Oct 31, 2025

chtruong814 had a problem deploying to test October 31, 2025 21:08 — with GitHub Actions Error

rm cache aware inference from functional test

5cb1a65

Signed-off-by: naymaraq <[email protected]>

chtruong814 added Run CICD and removed Run CICD labels Oct 31, 2025

chtruong814 temporarily deployed to test October 31, 2025 21:11 — with GitHub Actions Inactive

github-actions Bot removed the Run CICD label Nov 1, 2025

chtruong814 disabled auto-merge November 1, 2025 20:22

chtruong814 approved these changes Nov 1, 2025

View reviewed changes

chtruong814 merged commit 88d2ceb into main Nov 1, 2025
257 of 259 checks passed

chtruong814 deleted the dkaramyan/inference branch November 1, 2025 20:23

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Unified inference of streaming ASR#14817

Unified inference of streaming ASR#14817
chtruong814 merged 129 commits into
mainfrom
dkaramyan/inference

naymaraq commented Sep 26, 2025 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

chtruong814 commented Nov 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Conversation

naymaraq commented Sep 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What does this PR do ?

Changelog

Usage

GitHub Actions CI

Before your PR is "Ready for review"

Who can review?

Additional Information

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

chtruong814 commented Nov 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

naymaraq commented Sep 26, 2025 •

edited

Loading