SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

add low_cpu_mem_usage option in run_clm.py example which will benefit… (#22288)

Browse Source 
* add low_cpu_mem_usage option in run_clm.py example which will benefit LLM loading

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>

* update all the example and README under language-modeling

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>

---------

Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>
This commit is contained in:
Wang, Yi
2023-03-22 18:42:39 +08:00
committed by GitHub
parent 8472a224fb
commit 4ccaf268fb
6 changed files with 52 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
examples/pytorch/language-modeling/README.md
View File

@@ -178,6 +178,10 @@ sure all your batches have the same length.
To use the streaming dataset mode which can be very useful for large datasets, add `--streaming` to the command line. This is currently supported by `run_mlm.py` and `run_clm.py`.
## Low Cpu Memory Usage
To use low cpu memory mode which can be very useful for LLM, add `--low_cpu_mem_usage` to the command line. This is currently supported by `run_clm.py`,`run_mlm.py`, `run_plm.py`,`run_mlm_no_trainer.py` and `run_clm_no_trainer.py`.
## Creating a model on the fly
When training a model from scratch, configuration values may be overridden with the help of `--config_overrides`: