Add files using upload-large-folder tool

hn_eng_bn/tok_en-bn-hn/tok_files/tokenizer_117.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_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/tokenizer117/
+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_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer117/ \
+  --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_120.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_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/tokenizer120/
+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_splitac_splitae \
+ --output-prefix $FINAL_DIR/tokenizer120/ \
+  --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_130.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_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/tokenizer130/
+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_splitac_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer130/ \
+  --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_148.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_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/tokenizer148/
+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_splitab_splitab \
+ --output-prefix $FINAL_DIR/tokenizer148/ \
+  --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_173.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_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/tokenizer173/
+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_splitad \
+ --output-prefix $FINAL_DIR/tokenizer173/ \
+  --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_191.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_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/tokenizer191/
+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_splitab_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer191/ \
+  --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_220.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_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/tokenizer220/
+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_splitab \
+ --output-prefix $FINAL_DIR/tokenizer220/ \
+  --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_225.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_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/tokenizer225/
+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_splitab \
+ --output-prefix $FINAL_DIR/tokenizer225/ \
+  --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_241.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//finalal_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/tokenizer241/
+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//finalal_splitad_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer241/ \
+  --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_254.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_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/tokenizer254/
+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_splitac_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer254/ \
+  --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_271.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_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/tokenizer271/
+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_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer271/ \
+  --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_279.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_splitac_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/tokenizer279/
+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_splitac_splitad \
+ --output-prefix $FINAL_DIR/tokenizer279/ \
+  --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_284.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_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/tokenizer284/
+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_splitad \
+ --output-prefix $FINAL_DIR/tokenizer284/ \
+  --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_306.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_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/tokenizer306/
+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_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer306/ \
+  --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_307.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_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/tokenizer307/
+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_splitae_splitad \
+ --output-prefix $FINAL_DIR/tokenizer307/ \
+  --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_312.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_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/tokenizer312/
+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_splitac \
+ --output-prefix $FINAL_DIR/tokenizer312/ \
+  --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_313.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_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/tokenizer313/
+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_splitad \
+ --output-prefix $FINAL_DIR/tokenizer313/ \
+  --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_315.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_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/tokenizer315/
+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_splitab_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer315/ \
+  --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_319.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_splitab_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/tokenizer319/
+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_splitab_splitae \
+ --output-prefix $FINAL_DIR/tokenizer319/ \
+  --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_321.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_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/tokenizer321/
+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_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer321/ \
+  --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_341.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_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/tokenizer341/
+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_splitab_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer341/ \
+  --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_352.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_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/tokenizer352/
+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_splitab \
+ --output-prefix $FINAL_DIR/tokenizer352/ \
+  --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_38.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_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/tokenizer38/
+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_splitae \
+ --output-prefix $FINAL_DIR/tokenizer38/ \
+  --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_380.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_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/tokenizer380/
+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_splitaa_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer380/ \
+  --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_382.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_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/tokenizer382/
+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_splitaa_splitac \
+ --output-prefix $FINAL_DIR/tokenizer382/ \
+  --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_387.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_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/tokenizer387/
+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_splitab_splitac \
+ --output-prefix $FINAL_DIR/tokenizer387/ \
+  --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_388.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_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/tokenizer388/
+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_splitab_splitad \
+ --output-prefix $FINAL_DIR/tokenizer388/ \
+  --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_390.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_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/tokenizer390/
+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_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer390/ \
+  --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_394.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_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/tokenizer394/
+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_splitae \
+ --output-prefix $FINAL_DIR/tokenizer394/ \
+  --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_405.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_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/tokenizer405/
+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_splitaf \
+ --output-prefix $FINAL_DIR/tokenizer405/ \
+  --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_412.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//finalat_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/tokenizer412/
+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//finalat_splitad_splitad \
+ --output-prefix $FINAL_DIR/tokenizer412/ \
+  --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_413.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//finalat_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/tokenizer413/
+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//finalat_splitad_splitae \
+ --output-prefix $FINAL_DIR/tokenizer413/ \
+  --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_430.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_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/tokenizer430/
+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_splitae \
+ --output-prefix $FINAL_DIR/tokenizer430/ \
+  --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_432.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_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/tokenizer432/
+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_splitad_splitab \
+ --output-prefix $FINAL_DIR/tokenizer432/ \
+  --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_442.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_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/tokenizer442/
+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_splitaa_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer442/ \
+  --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_453.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_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/tokenizer453/
+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_splitab \
+ --output-prefix $FINAL_DIR/tokenizer453/ \
+  --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_461.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_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/tokenizer461/
+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_splitae \
+ --output-prefix $FINAL_DIR/tokenizer461/ \
+  --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_462.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_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/tokenizer462/
+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_splitae_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer462/ \
+  --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_466.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_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/tokenizer466/
+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_splitae_splitae \
+ --output-prefix $FINAL_DIR/tokenizer466/ \
+  --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_468.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_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/tokenizer468/
+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_splitaa_splitaa \
+ --output-prefix $FINAL_DIR/tokenizer468/ \
+  --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_471.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_splitaa_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/tokenizer471/
+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_splitaa_splitad \
+ --output-prefix $FINAL_DIR/tokenizer471/ \
+  --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_479.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_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/tokenizer479/
+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_splitac_splitab \
+ --output-prefix $FINAL_DIR/tokenizer479/ \
+  --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_489.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_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/tokenizer489/
+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_splitae_splitab \
+ --output-prefix $FINAL_DIR/tokenizer489/ \
+  --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_490.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_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/tokenizer490/
+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_splitae_splitac \
+ --output-prefix $FINAL_DIR/tokenizer490/ \
+  --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_506.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_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/tokenizer506/
+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_splitae_splitae \
+ --output-prefix $FINAL_DIR/tokenizer506/ \
+  --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_513.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_splitab_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/tokenizer513/
+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_splitab_splitae \
+ --output-prefix $FINAL_DIR/tokenizer513/ \
+  --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_531.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_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/tokenizer531/
+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_splitaa_splitab \
+ --output-prefix $FINAL_DIR/tokenizer531/ \
+  --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_534.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_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/tokenizer534/
+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_splitaa_splitae \
+ --output-prefix $FINAL_DIR/tokenizer534/ \
+  --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_539.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_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/tokenizer539/
+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_splitae \
+ --output-prefix $FINAL_DIR/tokenizer539/ \
+  --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_544.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_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/tokenizer544/
+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_splitad_splitad \
+ --output-prefix $FINAL_DIR/tokenizer544/ \
+  --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 \
+