cli: add --stdin-mode for continuous file processing#3637
Open
tsubome wants to merge 3 commits intoggml-org:masterfrom
Open
cli: add --stdin-mode for continuous file processing#3637tsubome wants to merge 3 commits intoggml-org:masterfrom
tsubome wants to merge 3 commits intoggml-org:masterfrom
Conversation
Add --stdin-mode option that allows whisper-cli to load the model once and process multiple audio files by reading paths from stdin. This is useful for: - Batch processing without model reload overhead - Workaround for whisper-server handle leak issues (ggml-org#3358) - Integration with external programs via stdin/stdout Output format (JSON, one per line): {"status":"ready"} {"file":"path","status":"success","text":"..."} {"file":"path","status":"error","error":"..."} {"status":"exit"} Usage: whisper-cli --stdin-mode --model model.bin --language ja # sends file paths via stdin, receives JSON results via stdout
When whisper-cli is invoked via subprocess with piped stdout, the C runtime defaults to full buffering, causing segments to appear in batches. Adding setvbuf(stdout, NULL, _IONBF, 0) ensures each segment is flushed immediately. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
tsubome
force-pushed
the
feature/stdin-mode
branch
from
22dff45 to
c518eca
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add
--stdin-modeoption to whisper-cli that allows loading the model once and processing multiple audio files by reading paths from stdin.Motivation
Usage
Then send file paths via stdin (one per line). Results are output as JSON:
{"status":"ready"} {"file":"path/to/audio.wav","status":"success","text":"transcription result"} {"file":"path/to/audio.wav","status":"error","error":"error message"} {"status":"exit"}Type
exitorquitto terminate, or send EOF.Changes
stdin_modeparameter towhisper_params--stdin-modecommand line optionescape_json_string()helper for JSON output