diff --git a/test/train-sets/ref/help.stdout b/test/train-sets/ref/help.stdout index 5d94ecce144..96601833d2e 100644 --- a/test/train-sets/ref/help.stdout +++ b/test/train-sets/ref/help.stdout @@ -121,6 +121,21 @@ Input Options: keep) --flatbuffer Data file will be interpreted as a flatbuffer file (type: bool, experimental) + --csv Data file will be interpreted as a CSV file (type: bool, experimental) + --csv_separator arg CSV Parser: Specify field separator in one character, " | : are + not allowed for reservation. (type: str, default: ,, experimental) + --csv_no_file_header CSV Parser: First line is NOT a header. By default, CSV files + are assumed to have a header with feature and/or namespaces names. + You MUST specify the header with --csv_header if you use this + option. (type: bool, default: false, experimental) + --csv_header arg CSV Parser: Override the CSV header by providing (namespace, + '|' and) feature name separated with ','. By default, CSV files + are assumed to have a header with feature and/or namespaces names + in the CSV first line. You can override it by specifying here. + Combined with --csv_no_file_header, we assume that there is no + header in the CSV file. (type: str, experimental) + --csv_ns_value arg CSV Parser: Scale the namespace values by specifying the float + ratio. e.g. --csv_ns_value=a:0.5,b:0.3,:8 (type: str, experimental) Logging Options: --quiet Don't output diagnostics and progress updates. Supplying this implies --log_level off and --driver_output_off. Supplying this @@ -613,6 +628,14 @@ Weight Options: uint, keep, necessary) --replay_m_count arg How many times (in expectation) should each example be played (default: 1 = permuting) (type: uint, default: 1) +[Reduction] Experimental: Contextual Bandit Exploration with ADF with graph feedback Options: + --cb_explore_adf Online explore-exploit for a contextual bandit problem with multiline + action dependent features (type: bool, keep, necessary) + --gamma_scale arg Sets CB with graph feedback gamma parameter to gamma=[gamma_scale]*[num + examples]^1/2 (type: float, default: 1, keep) + --gamma_exponent arg Exponent on [num examples] in CB with graph feedback parameter + gamma (type: float, default: 0.5, keep) + --graph_feedback Graph feedback pdf (type: bool, keep, necessary, experimental) [Reduction] Experimental: Contextual Bandit Exploration with ADF with large action space filtering Options: --cb_explore_adf Online explore-exploit for a contextual bandit problem with multiline action dependent features (type: bool, keep, necessary)