diff options
Diffstat (limited to 'gi/pyp-topics/src/train-contexts.cc')
| -rw-r--r-- | gi/pyp-topics/src/train-contexts.cc | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/gi/pyp-topics/src/train-contexts.cc b/gi/pyp-topics/src/train-contexts.cc index 0a48d3d9..5e98d02f 100644 --- a/gi/pyp-topics/src/train-contexts.cc +++ b/gi/pyp-topics/src/train-contexts.cc @@ -54,6 +54,7 @@ int main(int argc, char **argv) ("freq-cutoff-end", value<int>()->default_value(0), "final frequency cutoff.") ("freq-cutoff-interval", value<int>()->default_value(0), "number of iterations between frequency decrement.") ("max-threads", value<int>()->default_value(1), "maximum number of simultaneous threads allowed") + ("max-contexts-per-document", value<int>()->default_value(0), "Only sample the n most frequent contexts for a document.") ("num-jobs", value<int>()->default_value(1), "allows finer control over parallelization") ; @@ -110,7 +111,8 @@ int main(int argc, char **argv) model.sample_corpus(contexts_corpus, vm["samples"].as<int>(), vm["freq-cutoff-start"].as<int>(), vm["freq-cutoff-end"].as<int>(), - vm["freq-cutoff-interval"].as<int>()); + vm["freq-cutoff-interval"].as<int>(), + vm["max-contexts-per-document"].as<int>()); if (vm.count("document-topics-out")) { ogzstream documents_out(vm["document-topics-out"].as<string>().c_str()); |