From 3e1ff307af70687a53feddcf494ff02d64dc376c Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Sat, 18 Apr 2020 13:47:04 +0200
Subject: [PATCH] added script to save machine descriptions
---
UD_any/tokemorpho_incr/machine.rm | 23 ++++++++++++++++-------
scripts/extractMachines.py | 19 +++++++++++++++++++
2 files changed, 35 insertions(+), 7 deletions(-)
create mode 100755 scripts/extractMachines.py
diff --git a/UD_any/tokemorpho_incr/machine.rm b/UD_any/tokemorpho_incr/machine.rm
index 01d4377..04abba1 100644
--- a/UD_any/tokemorpho_incr/machine.rm
+++ b/UD_any/tokemorpho_incr/machine.rm
@@ -3,7 +3,7 @@ Classifier : tokemorpho
{
Transitions : {data/tokenizer.ts data/tagger.ts data/morpho_parts.ts}
Network type : LSTM
- Unknown value threshold : -1
+ Unknown value threshold : 1
Buffer context : {-3 -2 -1}
Stack context : {}
Columns : {FORM UPOS}
@@ -13,15 +13,24 @@ Classifier : tokemorpho
Max nb elements : {1 10 10}
Raw input left window : 5
Raw input right window : 5
- Embeddings size : 256
- MLP : {8192 0.3}
+ Embeddings size : 64
+ MLP : {2048 0.3}
Context LSTM size : 1024
Focused LSTM size : 256
- Rawinput LSTM size : 32
- Split trans LSTM size : 256
- Num layers : 2
+ Rawinput LSTM size : 16
+ Split trans LSTM size : 128
+ Num layers : 3
BiLSTM : true
- LSTM dropout : 0.3
+ LSTM dropout : 0.1
+ Total input dropout : 0.3
+ Embeddings dropout : 0.3
+ Dropout 2d : false
+ Tree embedding columns : {}
+ Tree embedding buffer : {}
+ Tree embedding stack : {}
+ Tree embedding nb : {}
+ Tree embedding size : 0
+ Optimizer : Adam {0.0005 0.9 0.999 0.00000001 0.00001 true}
}
Splitwords : data/splitwords.ts
Predictions : ID FORM UPOS FEATS EOS
diff --git a/scripts/extractMachines.py b/scripts/extractMachines.py
new file mode 100755
index 0000000..59bff73
--- /dev/null
+++ b/scripts/extractMachines.py
@@ -0,0 +1,19 @@
+#! /usr/bin/python3
+
+import sys
+import os
+import glob
+import shutil
+
+def printUsageAndExit() :
+ print("USAGE : %s bin dest"%sys.argv[0], file=sys.stderr)
+ exit(1)
+
+if __name__ == "__main__" :
+ if len(sys.argv) != 3 :
+ printUsageAndExit()
+
+ for filename in glob.iglob(sys.argv[1]+"/**", recursive=True) :
+ if os.path.splitext(filename)[1] == ".rm" :
+ shutil.copy(filename, sys.argv[2]+"/"+filename.split('/')[-2]+".rm")
+
--
GitLab