423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-03 04:02:49 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix example params

Browse Source 
Former-commit-id: 8a45213440ffc960947dd69ecf3b092aa724bef3
This commit is contained in:
hiyouga 2024-03-08 20:41:43 +08:00
parent b268215a0e
commit ccec17f773
4 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
examples/extras/galore/adamw.sh
View File

@ -15,7 +15,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../../src/train_bash.py \
--preprocessing_num_workers 16 \
--per_device_train_batch_size 1 \
--per_device_eval_batch_size 1 \
--gradient_accumulation_steps 8 \
--gradient_accumulation_steps 2 \
--lr_scheduler_type cosine \
--logging_steps 10 \
--warmup_steps 20 \

2
examples/extras/galore/adamw_8bit_bf16.sh
View File

@ -16,7 +16,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../../src/train_bash.py \
--preprocessing_num_workers 16 \
--per_device_train_batch_size 1 \
--per_device_eval_batch_size 1 \
--gradient_accumulation_steps 8 \
--gradient_accumulation_steps 2 \
--lr_scheduler_type cosine \
--logging_steps 10 \
--warmup_steps 20 \

2
examples/extras/galore/galore_adamw.sh
View File

@ -18,7 +18,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../../src/train_bash.py \
--preprocessing_num_workers 16 \
--per_device_train_batch_size 1 \
--per_device_eval_batch_size 1 \
--gradient_accumulation_steps 8 \
--gradient_accumulation_steps 2 \
--lr_scheduler_type cosine \
--logging_steps 10 \
--warmup_steps 20 \

4
examples/extras/galore/galore_adamw_8bit_bf16.sh
View File

@ -10,7 +10,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../../src/train_bash.py \
--finetuning_type full \
--use_galore \
--galore_target mlp,self_attn \
--galore_rank 32 \
--galore_rank 16 \
--optim adamw_8bit \
--output_dir ../../../saves/LLaMA2-7B/galore/sft \
--overwrite_cache \
@ -19,7 +19,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../../src/train_bash.py \
--preprocessing_num_workers 16 \
--per_device_train_batch_size 1 \
--per_device_eval_batch_size 1 \
--gradient_accumulation_steps 8 \
--gradient_accumulation_steps 2 \
--lr_scheduler_type cosine \
--logging_steps 10 \
--warmup_steps 20 \