Create README.md

README.md

@@ -0,0 +1,36 @@
+---
+license: apache-2.0
+---
+
+# fineweb-edu_default_Llama2_Tokenizer
+
+The original `fineweb-edu_default_Llama2_Tokenizer.tar.gz` archive (≈1.9T on Ubuntu) was split into smaller 40 GB chunks for easier upload to Hugging Face.
+```bash
+sudo apt install git-lfs
+pip install -U huggingface_hub  # `hf version`==1.1.4
+tar cvf - fineweb-edu_default_Llama2_Tokenizer/ | pigz -p 16 > fineweb-edu_default_Llama2_Tokenizer.tar.gz
+split -b 40G -d -a 3 fineweb-edu_default_Llama2_Tokenizer.tar.gz fineweb-edu_default_Llama2_Tokenizer/fineweb-edu_default_Llama2_Tokenizer_part_
+
+# Upload files sequentially using a loop:
+count=0
+for file in fineweb-edu_default_Llama2_Tokenizer/fineweb-edu_default_Llama2_Tokenizer_part_*; do
+    count=$((count + 1))
+    # if [ $count -lt 3 ]; then
+    #     continue
+    # fi
+    filename=$(basename "$file")
+    echo "Uploading $file (file #$count) to fineweb-edu_default_Llama2_Tokenizer/$filename ..."
+    hf upload jsun/fineweb-edu_default_Llama2_Tokenizer "$file" "fineweb-edu_default_Llama2_Tokenizer/$filename" --repo-type dataset
+done
+```
+
+Once all parts are downloaded, you can reconstruct the original tar.gz file using:
+```bash
+cat fineweb-edu_default_Llama2_Tokenizer_part_* > fineweb-edu_default_Llama2_Tokenizer.tar.gz
+```
+
+Then extract the archive as usual:
+
+```bash
+pigz -dc -p 16 fineweb-edu_default_Llama2_Tokenizer.tar.gz | tar xvf -
+```