mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-04-21 23:12:07 +00:00
24 lines
544 B
Python
24 lines
544 B
Python
import pytest
|
|
import os
|
|
from text_generation_server.pb import generate_pb2
|
|
|
|
os.environ["USE_PREFIX_CACHING"] = "0"
|
|
os.environ["ATTENTION"] = "flashdecoding"
|
|
|
|
|
|
@pytest.fixture
|
|
def default_pb_parameters():
|
|
return generate_pb2.NextTokenChooserParameters(
|
|
temperature=1.0,
|
|
repetition_penalty=1.0,
|
|
top_k=0,
|
|
top_p=1.0,
|
|
typical_p=1.0,
|
|
do_sample=False,
|
|
)
|
|
|
|
|
|
@pytest.fixture
|
|
def default_pb_stop_parameters():
|
|
return generate_pb2.StoppingCriteriaParameters(stop_sequences=[], max_new_tokens=10)
|