Skip to content

Modify MMLU prompt template to improve accuracy on DeepSeek models; E… #235

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
bzgoogle merged 2 commits into from
Apr 4, 2025
Merged

Modify MMLU prompt template to improve accuracy on DeepSeek models; E… #235

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b71cf5c
Modify MMLU prompt template to improve accuracy on DeepSeek models; E…
bzgoogle Apr 4, 2025
1bff78f
change the default value of num_prompts to -1
bzgoogle Apr 4, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 8 additions & 3 deletions benchmarks/benchmark_serving.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -342,7 +342,7 @@ def gen_mmlu_qa(data: Any, mmlu_method: str = "") -> str:
f"(D) {row['D']}\n"
)

output += "\nCorrect answer: "
output += "\nCorrect answer:"

if mmlu_method == "HELM":
output += f"({row['answer']})\n\n"
Expand Down Expand Up @@ -938,7 +938,7 @@ def parse_args() -> argparse.Namespace:
parser.add_argument(
"--num-prompts",
type=int,
default=1000,
default=-1,
help=(
"Number of prompts to process. (number of sample requests we randomly"
" collect from dataset)"
Expand Down Expand Up @@ -1133,11 +1133,16 @@ def main(args: argparse.Namespace):
# A given args.max_output_length value is the max generation step,
# when the args.max_output_length is default to None, the sample's golden
# output length will be used to decide the generation step.
if args.num_prompts == -1:
Comment thread
bzgoogle marked this conversation as resolved.
num_requests = len(dataset)
else:
num_requests = args.num_prompts

input_requests = sample_requests(
dataset=dataset,
tokenizer=tokenizer,
use_chat_template=use_chat_template,
num_requests=args.num_prompts,
num_requests=num_requests,
dataset_type=args.dataset,
max_output_length=args.max_output_length,
min_input_length=args.min_input_length,
Expand Down
Loading