Skip to content
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.

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

Add subsets for lcb #587

Open
wants to merge 5 commits into
base: main
Choose a base branch
from
Open
Changes from all commits
Commits
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
40 changes: 24 additions & 16 deletions src/lighteval/tasks/extended/lcb/main.py
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,7 @@

import numpy as np
from aenum import extend_enum
from datasets import get_dataset_config_names

from lighteval.metrics.metrics import MetricCategory, Metrics, MetricUseCase, SampleLevelMetric
from lighteval.tasks.extended.lcb.codegen_metrics import (
Expand Down Expand Up @@ -114,21 +115,28 @@ def codegen_metric(predictions: list[str], formatted_doc: Doc, **kwargs) -> floa

extend_enum(Metrics, "lcb_codegen_metric", lcb_codegen_metric)


task = LightevalTaskConfig(
name="lcb:codegeneration",
suite=["extended"],
prompt_function=lcb_codegeneration_prompt_fn,
hf_repo="livecodebench/code_generation_lite",
hf_subset="v4_v5", # https://github.com/LiveCodeBench/LiveCodeBench/tree/main?tab=readme-ov-file#dataset-versions
hf_avail_splits=["test"],
evaluation_splits=["test"],
generation_size=32768,
metric=[Metrics.lcb_codegen_metric],
stop_sequence=[], # no stop sequence, will use EOS token
trust_dataset=True,
version=0,
)
configs = get_dataset_config_names("livecodebench/code_generation_lite", trust_remote_code=True)

tasks = []

for subset in configs:
# To keep the base subset as the default, the others are named "lcb:codegeneration_v4", "lcb:codegeneration_v5"... etc
name = "lcb:codegeneration" if subset == "v4_v5" else f"lcb:codegeneration_{subset}"
task = LightevalTaskConfig(
name=name,
suite=["extended"],
prompt_function=lcb_codegeneration_prompt_fn,
hf_repo="livecodebench/code_generation_lite",
hf_subset=subset, # https://github.com/LiveCodeBench/LiveCodeBench/tree/main?tab=readme-ov-file#dataset-versions
hf_avail_splits=["test"],
evaluation_splits=["test"],
generation_size=32768,
metric=[Metrics.lcb_codegen_metric],
stop_sequence=[], # no stop sequence, will use EOS token
trust_dataset=True,
version=0,
)
tasks.append(task)


TASKS_TABLE = [task]
TASKS_TABLE = tasks