Hugging Face
Models
Datasets
Spaces
Community
Docs
Enterprise
Pricing
Log In
Sign Up
ocisd4
/
llama-2-tokenizer-dataprep
like
0
Follow
ocisd4
30
Model card
Files
Files and versions
xet
Community
main
llama-2-tokenizer-dataprep
/
README.md
AlexHung29629
Update README.md
96e482f
over 2 years ago
preview
code
|
raw
Copy download link
history
blame
contribute
delete
Safe
179 Bytes
-
關閉自動添加
`<s>`
,方便產生megatron-deepspeed訓練用檔案
-
指定pad token為
`<unk>`
,訓練時token數才會正確,以及finetune
_t0.py才能正確pack_
sample