"docs/git@developer.sourcefind.cn:renzhc/diffusers_dcu.git" did not exist on "50b6513531da7e258204871a9c675a56875d9e69"
Unverified Commit 82a533a6 authored by Xiaoyu Zhang's avatar Xiaoyu Zhang Committed by GitHub
Browse files

add rwkv-5-3b model (#666)

* support rwkv5-3b learnboard

* update rwkv-5-3b config

* update config

* refine

* fix bug

* update config

* refine

* reduce batch size

* refine

* reduce batch size to avoid oom in special datasets

* Update huggingface.py

* Update huggingface.py
parent 4780b39e
from mmengine.config import read_base
with read_base():
from .models.rwkv.rwkv5_3b import models
from .datasets.collections.base_medium_llama import datasets
from .summarizers.leaderboard import summarizer
from opencompass.models import HuggingFaceCausalLM
models = [
dict(
type=HuggingFaceCausalLM,
abbr='rwkv-5-3b',
path="RWKV/rwkv-5-world-3b",
tokenizer_path='RWKV/rwkv-5-world-3b',
model_kwargs=dict(
device_map='auto',
trust_remote_code=True,
),
tokenizer_kwargs=dict(
padding_side='left',
truncation_side='left',
trust_remote_code=True,
use_fast=False,
),
max_out_len=100,
max_seq_len=2048,
batch_padding=True,
batch_size=16,
run_cfg=dict(num_gpus=1, num_procs=1),
)
]
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment