Add files using upload-large-folder tool

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_106.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalae_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer106/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalae_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer106/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_111.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaf_splitaa_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer111/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaf_splitaa_splitab \
+ --output-prefix $FINAL_DIR/tokenizer111/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_121.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaf_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer121/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaf_splitad \
+ --output-prefix $FINAL_DIR/tokenizer121/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_139.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalag_splitad_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer139/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalag_splitad_splitae \
+ --output-prefix $FINAL_DIR/tokenizer139/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_142.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalag_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer142/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalag_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer142/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_160.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalah_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer160/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalah_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer160/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_161.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalah_splitae_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer161/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalah_splitae_splitad \
+ --output-prefix $FINAL_DIR/tokenizer161/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_170.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalai_splitac_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer170/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalai_splitac_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer170/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_175.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalai_splitad_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer175/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalai_splitad_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer175/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_200.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaj_splitac_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer200/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaj_splitac_splitae \
+ --output-prefix $FINAL_DIR/tokenizer200/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_201.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaj_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer201/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaj_splitad \
+ --output-prefix $FINAL_DIR/tokenizer201/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_204.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaj_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer204/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaj_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer204/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_208.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitaa_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer208/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitaa_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer208/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_215.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitac_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer215/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer215/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_221.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitad_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer221/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitad_splitac \
+ --output-prefix $FINAL_DIR/tokenizer221/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_222.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitad_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer222/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitad_splitad \
+ --output-prefix $FINAL_DIR/tokenizer222/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_226.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer226/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer226/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_228.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitae_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer228/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalak_splitae_splitae \
+ --output-prefix $FINAL_DIR/tokenizer228/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_260.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalam_splitad_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer260/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalam_splitad_splitab \
+ --output-prefix $FINAL_DIR/tokenizer260/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_274.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalan_splitab_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer274/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalan_splitab_splitad \
+ --output-prefix $FINAL_DIR/tokenizer274/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_282.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalan_splitad_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer282/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalan_splitad_splitab \
+ --output-prefix $FINAL_DIR/tokenizer282/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_293.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalao_splitab_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer293/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalao_splitab_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer293/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_295.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalao_splitab_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer295/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalao_splitab_splitac \
+ --output-prefix $FINAL_DIR/tokenizer295/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_314.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalap_splitaa_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer314/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalap_splitaa_splitae \
+ --output-prefix $FINAL_DIR/tokenizer314/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_32.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalab_splitac_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer32/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalab_splitac_splitae \
+ --output-prefix $FINAL_DIR/tokenizer32/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_328.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalap_splitad_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer328/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalap_splitad_splitad \
+ --output-prefix $FINAL_DIR/tokenizer328/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_351.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaq_splitad_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer351/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaq_splitad_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer351/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_356.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaq_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer356/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaq_splitae \
+ --output-prefix $FINAL_DIR/tokenizer356/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_358.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitaa_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer358/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitaa_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer358/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_36.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalab_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer36/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalab_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer36/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_364.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitab_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer364/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitab_splitab \
+ --output-prefix $FINAL_DIR/tokenizer364/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_365.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitab_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer365/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitab_splitac \
+ --output-prefix $FINAL_DIR/tokenizer365/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_379.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitaf
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer379/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalar_splitaf \
+ --output-prefix $FINAL_DIR/tokenizer379/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_391.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalas_splitac_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer391/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalas_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer391/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_399.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalas_splitad_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer399/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalas_splitad_splitae \
+ --output-prefix $FINAL_DIR/tokenizer399/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_428.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalau_splitac_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer428/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalau_splitac_splitac \
+ --output-prefix $FINAL_DIR/tokenizer428/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_452.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalav_splitac_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer452/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalav_splitac_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer452/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_460.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalav_splitad_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer460/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalav_splitad_splitad \
+ --output-prefix $FINAL_DIR/tokenizer460/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_484.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaw_splitad_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer484/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaw_splitad_splitab \
+ --output-prefix $FINAL_DIR/tokenizer484/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_493.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaw_splitaf
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer493/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaw_splitaf \
+ --output-prefix $FINAL_DIR/tokenizer493/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_497.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalax_splitac_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer497/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalax_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer497/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_498.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalax_splitac_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer498/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalax_splitac_splitac \
+ --output-prefix $FINAL_DIR/tokenizer498/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_51.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalac_splitac_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer51/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalac_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer51/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_524.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitaa
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer524/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer524/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_525.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer525/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitab \
+ --output-prefix $FINAL_DIR/tokenizer525/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_526.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer526/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer526/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_528.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitae
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer528/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalay_splitae_splitae \
+ --output-prefix $FINAL_DIR/tokenizer528/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_538.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaz_splitab_splitad
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer538/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaz_splitab_splitad \
+ --output-prefix $FINAL_DIR/tokenizer538/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_547.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaz_splitae_splitab
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer547/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalaz_splitae_splitab \
+ --output-prefix $FINAL_DIR/tokenizer547/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_554.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+set -m
+sed -i -e "s/raw_content/text/g" /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalba_splitaa_splitac
+echo "above deepspeed$"
+FINAL_DIR=/mnt/weka/peacock/idc/hineng/tok_en-bn-hn/
+#TOKENIZER=facebook/nllb-200-distilled-600M
+#TOKENIZER_TYPE=HuggingFaceTokenizer
+TOKENIZER_PATH=/mnt/weka/peacock/tokenization/trained-tokenizer/enhiben_50k
+TOKENIZER=$TOKENIZER_PATH/all.model
+VOCAB_FILE=$TOKENIZER_PATH/all.vocab
+TOKENIZER_TYPE=SentencePieceTokenizer
+mkdir -p $FINAL_DIR/tokenizer554/
+cd /mnt/weka/peacock/experiments/llama/Megatron-DeepSpeed
+echo "inside deepspeed"
+pwd
+python3 tools/preprocess_data.py \
+ --input /mnt/weka/peacock/idc/datasets/hn_eng_bn/tok_en-bn-hn/split//finalba_splitaa_splitac \
+ --output-prefix $FINAL_DIR/tokenizer554/ \
+  --tokenizer-model $TOKENIZER \
+  --vocab-file $VOCAB_FILE  \
+    --dataset-impl mmap --tokenizer-type $TOKENIZER_TYPE\
+    --append-eod --workers 8 # --partitions 16 #--chunk-size 50   
+
+  # --merge-file $MERGES_FILE \
+