2023-08-31 11:29:05 +08:00
|
|
|
from opencompass.models import HuggingFaceCausalLM
|
|
|
|
|
|
|
|
|
|
# Please note that we have specified the revision here. Recently (on 20230827),
|
|
|
|
|
# during our evaluations, we found that the newer revision models have a drop
|
|
|
|
|
# of more than 5 points on datasets like GaokaoBench / mbpp.
|
|
|
|
|
# We are not yet sure whether this drop is due to incorrect logic in OpenCompass
|
|
|
|
|
# calling qwen or some other reasons. We would like to highlight this.
|
|
|
|
|
|
|
|
|
|
models = [
|
|
|
|
|
dict(
|
|
|
|
|
type=HuggingFaceCausalLM,
|
|
|
|
|
abbr='qwen-7b-hf',
|
|
|
|
|
path="Qwen/Qwen-7B",
|
|
|
|
|
tokenizer_path='Qwen/Qwen-7B',
|
|
|
|
|
tokenizer_kwargs=dict(
|
|
|
|
|
padding_side='left',
|
|
|
|
|
truncation_side='left',
|
|
|
|
|
trust_remote_code=True,
|
|
|
|
|
use_fast=False,
|
|
|
|
|
revision='39fc5fdcb95c8c367bbdb3bfc0db71d96266de09'
|
|
|
|
|
),
|
2023-09-05 10:15:19 +08:00
|
|
|
pad_token_id=151643,
|
2023-08-31 11:29:05 +08:00
|
|
|
max_out_len=100,
|
|
|
|
|
max_seq_len=2048,
|
|
|
|
|
batch_size=8,
|
|
|
|
|
model_kwargs=dict(
|
|
|
|
|
device_map='auto',
|
|
|
|
|
trust_remote_code=True,
|
|
|
|
|
revision='39fc5fdcb95c8c367bbdb3bfc0db71d96266de09'
|
|
|
|
|
),
|
|
|
|
|
run_cfg=dict(num_gpus=1, num_procs=1),
|
|
|
|
|
)
|
|
|
|
|
]
|
