blob: c7e99294008b52501ecbc5fd69364240dd781331 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
|
#!/bin/bash -x
#####################################################################
# SUMMARY: Train a model on TSV data with guided alignment
# TAGS: sentencepiece tsv train align
#####################################################################
# Exit on error
set -e
# Remove old artifacts and create working directory
rm -rf train_align0 train_align0.{log,out,diff}
mkdir -p train_align0
# Run marian command
$MRT_MARIAN/marian \
--cost-type ce-mean --no-shuffle --clip-norm 0 --seed 5555 --maxi-batch 1 --maxi-batch-sort none --optimizer adam --learn-rate 0.001 \
--dim-emb 32 --transformer-dim-ffn 64 --type transformer --enc-depth 3 --dec-depth 3 \
-m train_align0/model.npz --tsv -t train2.aln-de-en.tsv -v $MRT_MODELS/rnn-spm/vocab.deen.{spm,spm} \
--after-batches 100 --disp-freq 4 \
--guided-alignment 0 --guided-alignment-weight 1.0 --guided-alignment-cost ce \
--log train_align0.log
# Check if files exist
test -e train_align0/model.npz
test -e train_align0.log
grep -qi "word alignments from" train_align0.log
# Compare the current output with the expected output
cat train_align0.log | $MRT_TOOLS/extract-costs.sh > train_align0.out
$MRT_TOOLS/diff-nums.py train_align0.out train_align.expected -p 0.01 -o train_align0.diff
# Exit with success code
exit 0
|
