From ecc1b393f0db116b5fd69ffe4927cb3a13abb3f6 Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Fri, 15 Apr 2022 16:44:46 +0200
Subject: [PATCH] Added new machines templates
---
UD_any/evaluate.sh | 30 ++++++++++---
UD_any/launchBatches.py | 6 +--
UD_any/templates/tagparser_incr/machine.rm | 20 ++++-----
.../tagparser_incr_nopretrained/machine.rm | 33 ++++++++++++++
UD_any/templates/tagparser_seq/machine.rm | 21 ++++-----
.../tagparser_seq_nopretrained/machine.rm | 36 ++++++++++++++++
UD_any/templates/tokeparser_incr/machine.rm | 2 +-
.../tokeparser_incr_nopretrained/machine.rm | 39 +++++++++++++++++
UD_any/templates/tokeparser_seq/machine.rm | 2 +-
.../tokeparser_seq_nopretrained/machine.rm | 43 +++++++++++++++++++
10 files changed, 200 insertions(+), 32 deletions(-)
create mode 100644 UD_any/templates/tagparser_incr_nopretrained/machine.rm
create mode 100644 UD_any/templates/tagparser_seq_nopretrained/machine.rm
create mode 100644 UD_any/templates/tokeparser_incr_nopretrained/machine.rm
create mode 100644 UD_any/templates/tokeparser_seq_nopretrained/machine.rm
diff --git a/UD_any/evaluate.sh b/UD_any/evaluate.sh
index 8b8bdda..7f39096 100755
--- a/UD_any/evaluate.sh
+++ b/UD_any/evaluate.sh
@@ -51,18 +51,34 @@ then
print_usage_and_exit
fi
+MCD="ID,FORM,LEMMA,UPOS,XPOS,FEATS,HEAD,DEPREL"
+NO=""
+for arg in "$@"
+do
+ if [ "$NO" = "1" ]
+ then
+ MCD="$arg"
+ NO=""
+ fi
+ if [ "$arg" = "--mcd" ]
+ then
+ NO="1"
+ fi
+done
+
EVALCONLL="../scripts/conll18_ud_eval.py"
OUTPUT=$EXPPATH"/predicted_eval.tsv"
-if [ "$MODE" = "tsv" ]; then
-macaon decode --model $EXPPATH --inputTSV $REF $@ > $OUTPUT && $EVALCONLL $REF $OUTPUT || exit 1
-exit 0
+INPUT="$REF"
+INPUTARG="--inputTSV"
+if [ "$MODE" = "txt" ]; then
+ INPUT="$REFRAW"
+ INPUTARG="--inputTXT"
fi
-if [ "$MODE" = "txt" ]; then
-macaon decode --model $EXPPATH --inputTXT $REFRAW $@ > $OUTPUT && $EVALCONLL $REF $OUTPUT || exit 1
-exit 0
+if [ ! -f "$OUTPUT" ]; then
+ macaon decode --model $EXPPATH $INPUTARG $INPUT $@ > $OUTPUT || exit 1
fi
-print_usage_and_exit
+$EVALCONLL --mcd $MCD $REF $OUTPUT || exit 1
diff --git a/UD_any/launchBatches.py b/UD_any/launchBatches.py
index 3dd7ebb..157d5d7 100755
--- a/UD_any/launchBatches.py
+++ b/UD_any/launchBatches.py
@@ -26,10 +26,10 @@ def addNamesAndCommandsTrain(names, commands, mode, expName, arguments, seed, pr
###############################################################################
###############################################################################
-def addNamesAndCommandsDecode(names, commands, mode, expName, arguments, pretrained) :
+def addNamesAndCommandsDecode(names, commands, mode, expName, arguments) :
names.append(expName)
- commands.append("./evaluate.sh {} bin/{} pretrained={} --silent {}".format(mode, expName, pretrained,arguments))
+ commands.append("./evaluate.sh {} bin/{} --silent {}".format(mode, expName, arguments))
###############################################################################
###############################################################################
@@ -64,7 +64,7 @@ if __name__ == "__main__" :
prepareExperiment(xp['lang'],xp['template'],xp['expName'])
addNamesAndCommandsTrain(names, commands, xp['mode'],xp['expName'],xp['arguments'],seed=100+i, pretrained=pretrained)
else :
- addNamesAndCommandsDecode(names, commands, xp['mode'],xp['expName'],xp['evalArguments'], pretrained=pretrained)
+ addNamesAndCommandsDecode(names, commands, xp['mode'],xp['expName'],xp['evalArguments'])
launchSlurmArray(names, commands, name, device, nbHours, limit, nbCPU)
###############################################################################
diff --git a/UD_any/templates/tagparser_incr/machine.rm b/UD_any/templates/tagparser_incr/machine.rm
index a4e542a..14406ad 100644
--- a/UD_any/templates/tagparser_incr/machine.rm
+++ b/UD_any/templates/tagparser_incr/machine.rm
@@ -1,31 +1,31 @@
Name : Tagger, Morpho, Lemmatizer, Parser and Segmenter Machine
Classifier : tagparser
{
- Transitions : {tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts lemmatizer_case,data/lemmatizer_case.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
+ Transitions : {tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
LossMultiplier : {}
Network type : Modular
- Contextual : Window{-10 10} Columns{FORM} LSTM{1 1 0 1} In{64} Out{128} w2v{FORM,data/FORM.w2v} Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1}
- Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID} LSTM{1 1 0 1} In{64} Out{64} w2v{}
- Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{UPOS FEATS DEPREL} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{300} Out{128} w2v{FORM,data/W2V/fasttext.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
- History : NbElem{10} LSTM{1 1 0 1} In{32} Out{32}
- InputDropout : 0.3
- MLP : {1600 0.3 1600 0.3}
+ History : NbElem{10} LSTM{1 1 0 1} In{128} Out{64}
+ HistoryMine : NbElem{4} LSTM{1 1 0 1} In{128} Out{64}
+ StateName : Out{64}
+ Distance : FromBuffer{} FromStack{0 1 2} ToBuffer{0} ToStack{} Threshold{15} LSTM{1 1 0.0 1} In{128} Out{64}
+ InputDropout : 0.5
+ MLP : {3200 0.4 1600 0.4}
End
Optimizer : Adagrad {0.01 0.000001 0 0.0000000001}
Type : classification
Loss : crossentropy
}
-Splitwords : data/splitwords.ts
Predictions : UPOS FEATS LEMMA HEAD DEPREL EOS
Strategy
{
Block : End{cannotMove}
tagger morpho * 0
morpho lemmatizer_rules * 0
- lemmatizer_rules lemmatizer_case * 0
- lemmatizer_case parser * 0
+ lemmatizer_rules parser * 0
parser segmenter eager_SHIFT 0
parser segmenter eager_RIGHT_rel 0
parser parser * 0
diff --git a/UD_any/templates/tagparser_incr_nopretrained/machine.rm b/UD_any/templates/tagparser_incr_nopretrained/machine.rm
new file mode 100644
index 0000000..3b513b4
--- /dev/null
+++ b/UD_any/templates/tagparser_incr_nopretrained/machine.rm
@@ -0,0 +1,33 @@
+Name : Tagger, Morpho, Lemmatizer, Parser and Segmenter Machine
+Classifier : tagparser
+{
+ Transitions : {tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
+ LossMultiplier : {}
+ Network type : Modular
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{128} Out{128} w2v{FORM,data/FORM.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
+ Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ History : NbElem{10} LSTM{1 1 0 1} In{128} Out{64}
+ HistoryMine : NbElem{4} LSTM{1 1 0 1} In{128} Out{64}
+ StateName : Out{64}
+ Distance : FromBuffer{} FromStack{0 1 2} ToBuffer{0} ToStack{} Threshold{15} LSTM{1 1 0.0 1} In{128} Out{64}
+ InputDropout : 0.5
+ MLP : {3200 0.4 1600 0.4}
+ End
+ Optimizer : Adagrad {0.01 0.000001 0 0.0000000001}
+ Type : classification
+ Loss : crossentropy
+}
+Predictions : UPOS FEATS LEMMA HEAD DEPREL EOS
+Strategy
+{
+ Block : End{cannotMove}
+ tagger morpho * 0
+ morpho lemmatizer_rules * 0
+ lemmatizer_rules parser * 0
+ parser segmenter eager_SHIFT 0
+ parser segmenter eager_RIGHT_rel 0
+ parser parser * 0
+ segmenter tagger * 1
+}
diff --git a/UD_any/templates/tagparser_seq/machine.rm b/UD_any/templates/tagparser_seq/machine.rm
index 80d2729..a605f22 100644
--- a/UD_any/templates/tagparser_seq/machine.rm
+++ b/UD_any/templates/tagparser_seq/machine.rm
@@ -1,22 +1,24 @@
Name : Tagger, Morpho, Lemmatizer, Parser and Segmenter Machine
Classifier : tagparser
{
- Transitions : {tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts lemmatizer_case,data/lemmatizer_case.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
- LossMultiplier : {}
+ Transitions : {tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
+ LossMultiplier : {}
Network type : Modular
- Contextual : Window{-10 10} Columns{FORM} LSTM{1 1 0 1} In{64} Out{128} w2v{FORM,data/FORM.w2v} Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1}
- Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{300} Out{128} w2v{FORM,data/W2V/fasttext.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
- History : NbElem{10} LSTM{1 1 0 1} In{32} Out{32}
- InputDropout : 0.3
- MLP : {1600 0.3 1600 0.3}
+ History : NbElem{10} LSTM{1 1 0 1} In{128} Out{64}
+ HistoryMine : NbElem{4} LSTM{1 1 0 1} In{128} Out{64}
+ StateName : Out{64}
+ Distance : FromBuffer{} FromStack{0 1 2} ToBuffer{0} ToStack{} Threshold{15} LSTM{1 1 0.0 1} In{128} Out{64}
+ InputDropout : 0.5
+ MLP : {3200 0.4 1600 0.4}
End
Optimizer : Adagrad {0.01 0.000001 0 0.0000000001}
Type : classification
Loss : crossentropy
}
-Splitwords : data/splitwords.ts
Predictions : UPOS FEATS LEMMA HEAD DEPREL EOS
Strategy
{
@@ -25,8 +27,7 @@ Strategy
Block : End{cannotMove}
morpho morpho * 1
Block : End{cannotMove}
- lemmatizer_rules lemmatizer_case * 0
- lemmatizer_case lemmatizer_rules * 1
+ lemmatizer_rules lemmatizer_rules * 1
Block : End{cannotMove}
parser segmenter eager_SHIFT 0
parser segmenter eager_RIGHT_rel 0
diff --git a/UD_any/templates/tagparser_seq_nopretrained/machine.rm b/UD_any/templates/tagparser_seq_nopretrained/machine.rm
new file mode 100644
index 0000000..0691aa7
--- /dev/null
+++ b/UD_any/templates/tagparser_seq_nopretrained/machine.rm
@@ -0,0 +1,36 @@
+Name : Tagger, Morpho, Lemmatizer, Parser and Segmenter Machine
+Classifier : tagparser
+{
+ Transitions : {tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
+ LossMultiplier : {}
+ Network type : Modular
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{128} Out{128} w2v{FORM,data/FORM.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
+ Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ History : NbElem{10} LSTM{1 1 0 1} In{128} Out{64}
+ HistoryMine : NbElem{4} LSTM{1 1 0 1} In{128} Out{64}
+ StateName : Out{64}
+ Distance : FromBuffer{} FromStack{0 1 2} ToBuffer{0} ToStack{} Threshold{15} LSTM{1 1 0.0 1} In{128} Out{64}
+ InputDropout : 0.5
+ MLP : {3200 0.4 1600 0.4}
+ End
+ Optimizer : Adagrad {0.01 0.000001 0 0.0000000001}
+ Type : classification
+ Loss : crossentropy
+}
+Predictions : UPOS FEATS LEMMA HEAD DEPREL EOS
+Strategy
+{
+ Block : End{cannotMove}
+ tagger tagger * 1
+ Block : End{cannotMove}
+ morpho morpho * 1
+ Block : End{cannotMove}
+ lemmatizer_rules lemmatizer_rules * 1
+ Block : End{cannotMove}
+ parser segmenter eager_SHIFT 0
+ parser segmenter eager_RIGHT_rel 0
+ parser parser * 0
+ segmenter parser * 1
+}
diff --git a/UD_any/templates/tokeparser_incr/machine.rm b/UD_any/templates/tokeparser_incr/machine.rm
index ad72b25..f7ea26a 100644
--- a/UD_any/templates/tokeparser_incr/machine.rm
+++ b/UD_any/templates/tokeparser_incr/machine.rm
@@ -4,7 +4,7 @@ Classifier : tokeparser
Transitions : {tokenizer,data/tokenizer.ts tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
LossMultiplier : {}
Network type : Modular
- Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} Concat{1 1 0.0 1} In{300} Out{64} w2v{FORM,data/W2V/fasttextcleanfiltered.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{300} Out{128} w2v{FORM,data/W2V/fasttext.w2v}
Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
diff --git a/UD_any/templates/tokeparser_incr_nopretrained/machine.rm b/UD_any/templates/tokeparser_incr_nopretrained/machine.rm
new file mode 100644
index 0000000..a4a37fc
--- /dev/null
+++ b/UD_any/templates/tokeparser_incr_nopretrained/machine.rm
@@ -0,0 +1,39 @@
+Name : Tokenizer, Tagger, Morpho, Lemmatizer, Parser and Segmenter Machine
+Classifier : tokeparser
+{
+ Transitions : {tokenizer,data/tokenizer.ts tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
+ LossMultiplier : {}
+ Network type : Modular
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{128} Out{128} w2v{FORM,data/FORM.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
+ Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ RawInput : Left{5} Right{10} LSTM{1 1 0.0 1} In{32} Out{32}
+ History : NbElem{10} LSTM{1 1 0 1} In{128} Out{64}
+ HistoryMine : NbElem{4} LSTM{1 1 0 1} In{128} Out{64}
+ StateName : Out{64}
+ Distance : FromBuffer{} FromStack{0 1 2} ToBuffer{0} ToStack{} Threshold{15} LSTM{1 1 0.0 1} In{128} Out{64}
+ SplitTrans : LSTM{1 1 0.0 1} In{128} Out{64}
+ InputDropout : 0.5
+ MLP : {3200 0.4 1600 0.4}
+ End
+ Optimizer : Adagrad {0.01 0.000001 0 0.0000000001}
+ Type : classification
+ Loss : crossentropy
+}
+Splitwords : data/splitwords.ts
+Predictions : ID FORM UPOS FEATS LEMMA HEAD DEPREL EOS
+Strategy
+{
+ Block : End{cannotMove}
+ tokenizer tagger ENDWORD 0
+ tokenizer tagger SPLIT 0
+ tokenizer tokenizer * 0
+ tagger morpho * 0
+ morpho lemmatizer_rules * 0
+ lemmatizer_rules parser * 0
+ parser segmenter eager_SHIFT 0
+ parser segmenter eager_RIGHT_rel 0
+ parser parser * 0
+ segmenter tokenizer * 1
+}
diff --git a/UD_any/templates/tokeparser_seq/machine.rm b/UD_any/templates/tokeparser_seq/machine.rm
index f63c06a..d3e8010 100644
--- a/UD_any/templates/tokeparser_seq/machine.rm
+++ b/UD_any/templates/tokeparser_seq/machine.rm
@@ -4,7 +4,7 @@ Classifier : tokeparser
Transitions : {tokenizer,data/tokenizer.ts tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
LossMultiplier : {}
Network type : Modular
- Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} Concat{1 1 0.0 1} In{300} Out{64} w2v{FORM,data/W2V/fasttextcleanfiltered.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{300} Out{128} w2v{FORM,data/W2V/fasttext.w2v}
Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
diff --git a/UD_any/templates/tokeparser_seq_nopretrained/machine.rm b/UD_any/templates/tokeparser_seq_nopretrained/machine.rm
new file mode 100644
index 0000000..eee655b
--- /dev/null
+++ b/UD_any/templates/tokeparser_seq_nopretrained/machine.rm
@@ -0,0 +1,43 @@
+Name : Tokenizer, Tagger, Morpho, Lemmatizer, Parser and Segmenter Machine
+Classifier : tokeparser
+{
+ Transitions : {tokenizer,data/tokenizer.ts tagger,data/tagger.ts morpho,data/morpho_whole.ts lemmatizer_rules,data/lemmatizer_rules.ts parser,data/parser_eager_rel_strict.ts segmenter,data/segmenter.ts}
+ LossMultiplier : {}
+ Network type : Modular
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{FORM} LSTM{1 1 0.0 1} In{128} Out{128} w2v{FORM,data/FORM.w2v}
+ Context : Targets{b.-3 b.-2 b.-1 b.0 b.1 b.2 s.0 s.1 s.2 b.0.0 s.0.0 s.0.-1 s.1.0 s.1.-1 s.2.0 s.2.-1} Columns{EOS ID UPOS FEATS DEPREL} LSTM{1 1 0 1} In{128} Out{64} w2v{}
+ Focused : Column{prefix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ Focused : Column{suffix3:FORM} NbElem{3} Buffer{0} Stack{} LSTM{1 1 0 1} In{64} Out{64} w2v{}
+ RawInput : Left{5} Right{10} LSTM{1 1 0.0 1} In{32} Out{32}
+ History : NbElem{10} LSTM{1 1 0 1} In{128} Out{64}
+ HistoryMine : NbElem{4} LSTM{1 1 0 1} In{128} Out{64}
+ StateName : Out{64}
+ Distance : FromBuffer{} FromStack{0 1 2} ToBuffer{0} ToStack{} Threshold{15} LSTM{1 1 0.0 1} In{128} Out{64}
+ SplitTrans : LSTM{1 1 0.0 1} In{128} Out{64}
+ InputDropout : 0.5
+ MLP : {3200 0.4 1600 0.4}
+ End
+ Optimizer : Adagrad {0.01 0.000001 0 0.0000000001}
+ Type : classification
+ Loss : crossentropy
+}
+Splitwords : data/splitwords.ts
+Predictions : ID FORM UPOS FEATS LEMMA HEAD DEPREL EOS
+Strategy
+{
+ Block : End{cannotMove}
+ tokenizer tokenizer ENDWORD 1
+ tokenizer tokenizer SPLIT 1
+ tokenizer tokenizer * 0
+ Block : End{cannotMove}
+ tagger tagger * 1
+ Block : End{cannotMove}
+ morpho morpho * 1
+ Block : End{cannotMove}
+ lemmatizer_rules lemmatizer_rules * 1
+ Block : End{cannotMove}
+ parser segmenter eager_SHIFT 0
+ parser segmenter eager_RIGHT_rel 0
+ parser parser * 0
+ segmenter parser * 1
+}
--
GitLab