init

format_text.py

@@ -1,10 +1,20 @@
+import os
 import csv
 import json
 from glob import glob
 import pandas as pd
 
-df = pd.concat([pd.read_csv(i, quoting=csv.QUOTE_NONE, encoding='utf-8', sep='\t', header=None, on_bad_lines='skip') for i in glob('seamless.dataset.metadata.public.jpn.batch_*.tsv')])
-line_no = [i.split(" ")[3] for i in df[0]]
+
+direction_speech = os.getenv("DIRECTION_SPEECH", "enA")
+direction_text = os.getenv("DIRECTION_TEXT", "jpn")
+direction = os.getenv("DIRECTION", "enA-jpn")
+
+df = pd.concat([
+    pd.read_csv(i, quoting=csv.QUOTE_NONE, encoding='utf-8', sep='\t', header=None, on_bad_lines='skip')
+    for i in glob(f'metadata.{direction_speech}-{direction_text}.batch_*')
+])
+line_no = [int(i.split(" ")[-1]) for i in df[0]]
 text = df[1].values.tolist()
-with open("text.enA-jpn.json", "w") as f:
+os.makedirs("text_corpus", exist_ok=True)
+with open(f"text_corpus/text.{direction_speech}-{direction_text}.json", "w") as f:
     json.dump({l: t for l, t in zip(line_no, text)}, f)

main_s2s.sh

@@ -19,13 +19,7 @@ do
   python fetch_dataset_s2s.py
 done
 # tokenize
-for i in $(seq 1 30);
-do
-  export DATASET_ID=${i}
-  export DIRECTION="enA-jaA"
-  python tokenize_dataset_s2s.py
-done
-for i in $(seq 30 60);
+for i in $(seq 120 144);
 do
   export DATASET_ID=${i}
   export DIRECTION="enA-jaA"
@@ -42,7 +36,7 @@ export LINE_NO_START=0
 export LINE_NO_END=10
 python fetch_dataset_s2s.py
 # main
-for i in $(seq 1 40);
+for i in $(seq 121 149);
 do
   export N_POOL=15
   export DATASET_ID=${i}
@@ -52,7 +46,7 @@ do
   echo ${LINE_NO_START}
   python fetch_dataset_s2s.py
 done
-for i in $(seq 41 80);
+for i in 114 77 78 79 80;
 do
   export N_POOL=15
   export DATASET_ID=${i}
@@ -62,6 +56,14 @@ do
   echo ${LINE_NO_START}
   python fetch_dataset_s2s.py
 done
+# tokenize
+for i in $(seq 50 120);
+do
+  export DATASET_ID=${i}
+  export DIRECTION="enA-viA"
+  python tokenize_dataset_s2s.py
+done
+
 
 ####################
 # enA-koA: 511_358 #
@@ -104,7 +106,7 @@ export LINE_NO_START=0
 export LINE_NO_END=10
 python fetch_dataset_s2s.py
 # main
-for i in $(seq 1 50);
+for i in $(seq 1 91);
 do
   export N_POOL=15
   export DATASET_ID=${i}
@@ -114,15 +116,19 @@ do
   echo ${LINE_NO_START}
   python fetch_dataset_s2s.py
 done
-for i in $(seq 51 91);
+# tokenize
+for i in $(seq 21 50);
 do
-  export N_POOL=15
   export DATASET_ID=${i}
   export DIRECTION="enA-hiA"
-  export LINE_NO_START=$(((DATASET_ID-1) * 2500))
-  export LINE_NO_END=$((DATASET_ID * 2500))
-  echo ${LINE_NO_START}
-  python fetch_dataset_s2s.py
+  python tokenize_dataset_s2s.py
+done
+
+for i in $(seq 71 91);
+do
+  export DATASET_ID=${i}
+  export DIRECTION="enA-hiA"
+  python tokenize_dataset_s2s.py
 done
 
 ######################
@@ -135,11 +141,11 @@ export LINE_NO_START=0
 export LINE_NO_END=10
 python fetch_dataset_s2s.py
 # main
-for i in $(seq 1 100);
+for i in $(seq 1 258);
 do
   export N_POOL=15
   export DATASET_ID=${i}
-  export DIRECTION="enA-viA"
+  export DIRECTION="enA-zhA"
   export LINE_NO_START=$(((DATASET_ID-1) * 2500))
   export LINE_NO_END=$((DATASET_ID * 2500))
   echo ${LINE_NO_START}
@@ -156,11 +162,21 @@ export LINE_NO_START=0
 export LINE_NO_END=10
 python fetch_dataset_s2s.py
 # main
-for i in $(seq 1 100);
+for i in $(seq 1 300);
 do
   export N_POOL=15
   export DATASET_ID=${i}
-  export DIRECTION="enA-viA"
+  export DIRECTION="enA-frA"
+  export LINE_NO_START=$(((DATASET_ID-1) * 2500))
+  export LINE_NO_END=$((DATASET_ID * 2500))
+  echo ${LINE_NO_START}
+  python fetch_dataset_s2s.py
+done
+for i in $(seq 300 600);
+do
+  export N_POOL=15
+  export DATASET_ID=${i}
+  export DIRECTION="enA-frA"
   export LINE_NO_START=$(((DATASET_ID-1) * 2500))
   export LINE_NO_END=$((DATASET_ID * 2500))
   echo ${LINE_NO_START}

main_s2t.sh

@@ -51,9 +51,9 @@ export CHUNK_SIZE=10
 python download_s2t_metadata.py
 for i in $(seq 1 ${CHUNK_SIZE});
 do
-  cat seamless.dataset.metadata.public.${DIRECTION_SPEECH}-${DIRECTION_TEXT}.withduration.reordered.batch_${i}.tsv | egrep ^crawl-data | tr '\t' ' ' | wet_lines | tee metadata.${DIRECTION_SPEECH}-${DIRECTION_TEXT}.batch_1.tsv &
+  cat seamless.dataset.metadata.public.${DIRECTION_SPEECH}-${DIRECTION_TEXT}.withduration.reordered.batch_${i}.tsv | egrep ^crawl-data | tr '\t' ' ' | wet_lines | tee metadata.${DIRECTION_SPEECH}-${DIRECTION_TEXT}.batch_${i}.tsv &
 done
-#python format_text.py
+python format_text.py
 
 ########
 # NLLB #

text.enA-jpn.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:96b5def10895c3cf05bf6d8a3b9626b7832b6e225f834700a87c39d6a2ef99fc
-size 1094842