From e6eb721f05ba0d4286f12b2dc92f7f2fd64686e6 Mon Sep 17 00:00:00 2001
From: piotrmp <piotr.m.przybyla@gmail.com>
Date: Fri, 4 Nov 2022 14:36:04 +0100
Subject: [PATCH] Command-line options for LAMBO segmentation.
---
combo/main.py | 6 +++---
combo/predict.py | 2 +-
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/combo/main.py b/combo/main.py
index c0365f3..03f2a85 100644
--- a/combo/main.py
+++ b/combo/main.py
@@ -88,8 +88,8 @@ flags.DEFINE_boolean(name="silent", default=True,
flags.DEFINE_enum(name="predictor_name", default="combo-spacy",
enum_values=["combo", "combo-spacy", "combo-lambo"],
help="Use predictor with whitespace, spacy or LAMBO tokenizer.")
-flags.DEFINE_string(name="lambo_model_name", default="en",
- help="LAMBO model name (if LAMBO used for segmentation).")
+#flags.DEFINE_string(name="lambo_model_name", default="en",
+# help="LAMBO model name (if LAMBO used for segmentation).")
def run(_):
"""Run model."""
@@ -176,7 +176,7 @@ def _get_predictor() -> predictors.Predictor:
)
return predictors.Predictor.from_archive(
- archive, FLAGS.predictor_name, "validation", True, {"lambo_model_name" : FLAGS.lambo_model_name}
+ archive, FLAGS.predictor_name#, extra_args={"lambo_model_name" : FLAGS.lambo_model_name}
)
diff --git a/combo/predict.py b/combo/predict.py
index 68bdef3..0423437 100644
--- a/combo/predict.py
+++ b/combo/predict.py
@@ -238,7 +238,7 @@ class COMBO(predictor.Predictor):
@classmethod
def with_lambo_tokenizer(cls, model: models.Model,
- dataset_reader: allen_data.DatasetReader, lambo_model_name : str):
+ dataset_reader: allen_data.DatasetReader, lambo_model_name : str = 'en'):
return cls(model, dataset_reader, lambo.LamboTokenizer(lambo_model_name))
@classmethod
--
GitLab