From 4a0a5980cf81ea0764911845c016bf314f535848 Mon Sep 17 00:00:00 2001 From: Victor Chahuneau Date: Fri, 10 Aug 2012 19:12:29 -0400 Subject: [python] Parallel decoding --- python/examples/cdec-mt.py | 41 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 41 insertions(+) create mode 100644 python/examples/cdec-mt.py (limited to 'python/examples/cdec-mt.py') diff --git a/python/examples/cdec-mt.py b/python/examples/cdec-mt.py new file mode 100644 index 00000000..9621df80 --- /dev/null +++ b/python/examples/cdec-mt.py @@ -0,0 +1,41 @@ +#!/usr/bin/env python +import sys +import argparse +import logging +import multiprocessing as mp +import cdec + +decoder = None +def make_decoder(config, weights): + global decoder + decoder = cdec.Decoder(config) + decoder.read_weights(weights) + +def translate(sentence): + global decoder + return decoder.translate(sentence).viterbi() + +def main(): + logging.basicConfig(level=logging.INFO, format='%(message)s') + + parser = argparse.ArgumentParser(description='Run multiple decoders concurrentely') + parser.add_argument('-c', '--config', required=True, + help='decoder configuration') + parser.add_argument('-w', '--weights', required=True, + help='feature weights') + parser.add_argument('-j', '--jobs', type=int, default=mp.cpu_count(), + help='number of decoder instances') + parser.add_argument('-s', '--chunksize', type=int, default=10, + help='number of sentences / chunk') + args = parser.parse_args() + + with open(args.config) as config: + config = config.read() + logging.info('Starting %d workers; chunk size: %d', args.jobs, args.chunksize) + pool = mp.Pool(args.jobs, make_decoder, (config, args.weights)) + for output in pool.imap(translate, sys.stdin, args.chunksize): + print(output.encode('utf8')) + logging.info('Shutting down workers...') + +if __name__ == '__main__': + main() -- cgit v1.2.3