github
/
ChatGLM-6B
mirror of https://github.com/THUDM/ChatGLM-6B
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Update batch size

pull/236/merge
duzx16 2023-03-31 11:28:13 +08:00
parent 971a6fbb20
commit d2645d8816
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
ptuning/train.sh
View File

@ -8,14 +8,14 @@ CUDA_VISIBLE_DEVICES=0 python3 main.py \
--prompt_column content \ --prompt_column content \
--response_column summary \ --response_column summary \
--overwrite_cache \ --overwrite_cache \
--model_name_or_path THUDM/chatglm-6b \ --model_name_or_path /mnt/vepfs/workspace/zxdu/chatglm_6b \
--output_dir output/adgen-chatglm-6b-pt-$PRE_SEQ_LEN-$LR-dev \ --output_dir output/adgen-chatglm-6b-pt-$PRE_SEQ_LEN-$LR-dev \
--overwrite_output_dir \ --overwrite_output_dir \
--max_source_length 64 \ --max_source_length 64 \
--max_target_length 64 \ --max_target_length 64 \
--per_device_train_batch_size 8 \ --per_device_train_batch_size 1 \
--per_device_eval_batch_size 1 \ --per_device_eval_batch_size 1 \
--gradient_accumulation_steps 2 \ --gradient_accumulation_steps 16 \
--predict_with_generate \ --predict_with_generate \
--max_steps 3000 \ --max_steps 3000 \
--logging_steps 10 \ --logging_steps 10 \