From 95f3b9043c6093a2637c5a2bcae7dab0ad4494ee Mon Sep 17 00:00:00 2001
From: jakubknczny <jakubknczny@gmail.com>
Date: Mon, 31 Jan 2022 18:33:10 +0100
Subject: [PATCH] update marian-train

---
 README.md               |  5 +++--
 scripts/marian-train.sh | 38 ++++++++++++++++++--------------------
 train.sh                |  9 +++------
 3 files changed, 24 insertions(+), 28 deletions(-)

diff --git a/README.md b/README.md
index 56ef77e..514295c 100644
--- a/README.md
+++ b/README.md
@@ -1,5 +1,6 @@
-# PLEWI
-- clone PLEWI repo into ~/
+# NEEDS TO BE UPDATED
+## PLEWI 
+- clone PLEWI repo into ~
 - run following commands:
   - git clone https://git.wmi.amu.edu.pl/s470607/transfix-train.git
   - cd transfix-train
diff --git a/scripts/marian-train.sh b/scripts/marian-train.sh
index f5d3710..a91b430 100755
--- a/scripts/marian-train.sh
+++ b/scripts/marian-train.sh
@@ -2,26 +2,28 @@
 
 # arguments
 # 1. root of gonito.net challenge-like filestructure
-# 2. number of BPE merge operations, E.g. 32000
+# 2. name of the model and therefore a directory that will contain the model
 # 3. expected number of train epochs
 
 # path to corpus
 corpus_path="$1"
-bpe_merges="$2"
+model_name="$2"
 epochs="$3"
 
-source_file="$corpus_path"/train/in.tsv."$bpe_merges"
-source_vocab="$source_file".vocab."$bpe_merges".yml
+source_file="$corpus_path"/train/in.tsv
+source_file_valid="$corpus_path"/test-A/in.tsv
 
-target_file="$corpus_path"/train/expected.tsv."$bpe_merges"
-target_vocab="$target_file".vocab."$bpe_merges".yml
+target_file="$corpus_path"/train/expected.tsv
+target_file_valid="$corpus_path"/test-A/expected.tsv
 
+mkdir "$model_name"
 
 ~/marian/build/marian \
- --type transformer \
+--type transformer \
+--model "$model_name"/model.npz \
 --overwrite \
 --train-sets "$source_file" "$target_file" \
---max-length 100 \
+--max-length 200 \
 --mini-batch-fit -w 10000 --maxi-batch 1000 \
 --valid-freq 5000 \
 --save-freq 5000 \
@@ -36,15 +38,11 @@ target_vocab="$target_file".vocab."$bpe_merges".yml
 --optimizer-params 0.9 0.98 1e-09 --clip-norm 5 \
 --tied-embeddings \
 --exponential-smoothing \
---log ~/train.log \
---after-epochs="$epochs"
-#--vocabs "$source_vocab" "$target_vocab"
-
-#--early-stopping 10 \
-#--model model/model.npz
-#--valid-log model/valid.log \
-#--valid-metrics cross-entropy perplexity translation \
-#--valid-sets data/valid.bpe.en data/valid.bpe.de \
-#--valid-script-path ./scripts/validate.sh \
-#--valid-translation-output data/valid.bpe.en.output --quiet-translation \
-#--valid-mini-batch 64 \
+--log "$model_name"/train.log \
+--after-epochs="$epochs" \
+--vocabs "$model_name"/vocab.in.spm "$model_name"/vocab.expected.spm \
+--valid-log "$model_name"/valid.log \
+--valid-metrics cross-entropy perplexity bleu \
+--valid-mini-batch 64 \
+--valid-sets "$source_file_valid" "$target_file_valid" \
+--valid-translation-output "$model_name"/valid.output --quiet-translation
diff --git a/train.sh b/train.sh
index 2f8e504..8b0f796 100755
--- a/train.sh
+++ b/train.sh
@@ -2,13 +2,10 @@
 
 # arguments
 # 1. root of gonito.net challenge-like filestructure
-# 2. number of BPE merge operations, E.g. 32000
-# 3. expected number of train epochs
+# 2. expected number of train epochs
 
 
 corpus_path="$1"
-bpe_merges="$2"
-epochs="$3"
+epochs="$2"
 
-./scripts/do-fastBPE.sh "$corpus_path" "$bpe_merges"
-./scripts/marian-train.sh "$corpus_path" "$bpe_merges" "$epochs"
\ No newline at end of file
+./scripts/marian-train.sh "$corpus_path" "$epochs"
\ No newline at end of file