phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-09 11:20:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

eleuther eval fix

Browse source
This commit is contained in:
Xi Yan 2024-10-11 09:30:10 -07:00
parent ad18dc94ac
commit fb565dfb06
3 changed files with 24 additions and 20 deletions
Download patch file Download diff file Expand all files Collapse all files

32
llama_stack/apis/evals/client.py
View file

@ -54,27 +54,29 @@ class EvaluationClient(Evals):
async def run_main(host: str, port: int):
client = EvaluationClient(f"http://{host}:{port}")
# CustomDataset
# Custom Eval Task
# response = await client.run_evals(
# model="Llama3.1-8B-Instruct",
# dataset="mmlu-simple-eval-en",
# task="mmlu",
# eval_task_config=EvaluateTaskConfig(
# n_samples=2,
# ),
# )
# Eleuther Eval Task
response = await client.run_evals(
model="Llama3.1-8B-Instruct",
dataset="mmlu-simple-eval-en",
task="mmlu",
# task="meta_mmlu_pro_instruct",
task="meta_ifeval",
eval_task_config=EvaluateTaskConfig(
n_samples=2,
),
)
cprint(f"evaluate response={response}", "green")
# Eleuther Eval Task
# response = await client.run_evals(
# model="Llama3.1-8B-Instruct",
# task="meta_mmlu_pro_instruct",
# # task="meta_ifeval",
# eval_task_config=EvaluateTaskConfig(
# n_samples=2,
# )
# )
# cprint(response.metrics["metrics_table"], "red")
if response.formatted_report:
cprint(response.formatted_report, "green")
else:
cprint(f"evaluate response={response}", "green")
def main(host: str, port: int):