423A35C7/LLaMA-Factory
1
0
Fork 0
mirror of https://github.com/hiyouga/LLaMA-Factory.git synced 2025-08-05 21:22:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #741 from hiyouga/feature-addDatasetCheck

Browse Source 
Feature add dataset check

Former-commit-id: 701a9d60cb030a18238b5426477752d41657f85d
This commit is contained in:
codingma 2023-08-31 20:57:36 +08:00 committed by GitHub
commit 5e1a8b1c74
2 changed files with 13 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/llmtuner/extras/template.py
View File

@ -285,7 +285,7 @@ register_template(
"Always answer as helpfully as possible, while being safe. " "Always answer as helpfully as possible, while being safe. "
"Your answers should not include any harmful, unethical, " "Your answers should not include any harmful, unethical, "
"racist, sexist, toxic, dangerous, or illegal content. " "racist, sexist, toxic, dangerous, or illegal content. "
"Please ensure that your responses are socially unbiased and positive in nature.\n" "Please ensure that your responses are socially unbiased and positive in nature.\n\n"
"If a question does not make any sense, or is not factually coherent, " "If a question does not make any sense, or is not factually coherent, "
"explain why instead of answering something not correct. " "explain why instead of answering something not correct. "
"If you don't know the answer to a question, please don't share false information." "If you don't know the answer to a question, please don't share false information."

12
src/llmtuner/tuner/core/parser.py
View File

@ -186,6 +186,18 @@ def get_train_args(
else: else:
model_args.compute_dtype = torch.float16 model_args.compute_dtype = torch.float16
# transfer training stage to dataset stage
dataset_stage = general_args.stage
if general_args.stage == "ppo":
dataset_stage = "sft"
elif general_args.stage == "dpo":
dataset_stage = "rm"
for dataset_attr in data_args.dataset_list:
if dataset_attr.stage and dataset_attr.stage != dataset_stage:
raise ValueError("Dataset {} is not supported for the stage {}"
.format(dataset_attr.dataset_name, general_args.stage))
model_args.model_max_length = data_args.max_source_length + data_args.max_target_length model_args.model_max_length = data_args.max_source_length + data_args.max_target_length
# Log on each process the small summary: # Log on each process the small summary: