biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Simplify TokenizerGroup (#16790)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-04-24 12:43:56 +01:00
committed by GitHub
parent 14288d1332
commit 0a05ed57e6
24 changed files with 80 additions and 752 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
tests/lora/test_tokenizer_group.py
View File

@@ -5,17 +5,14 @@ from transformers import AutoTokenizer, PreTrainedTokenizerBase
from vllm.lora.request import LoRARequest
from vllm.transformers_utils.tokenizer import get_lora_tokenizer
from vllm.transformers_utils.tokenizer_group import get_tokenizer_group
from ..conftest import get_tokenizer_pool_config
from vllm.transformers_utils.tokenizer_group import TokenizerGroup
@pytest.mark.asyncio
@pytest.mark.parametrize("tokenizer_group_type", [None, "ray"])
async def test_tokenizer_group_lora(sql_lora_files, tokenizer_group_type):
reference_tokenizer = AutoTokenizer.from_pretrained(sql_lora_files)
tokenizer_group = get_tokenizer_group(
get_tokenizer_pool_config(tokenizer_group_type),
tokenizer_group = TokenizerGroup(
tokenizer_id="gpt2",
enable_lora=True,
max_num_seqs=1,
@@ -60,8 +57,7 @@ def test_get_lora_tokenizer(sql_lora_files, tmp_path):
@pytest.mark.parametrize("max_num_seqs", [1, 2])
@pytest.mark.parametrize("max_loras", [1, 2])
def test_lora_tokenizers(enable_lora, max_num_seqs, max_loras):
tokenizer_group = get_tokenizer_group(
get_tokenizer_pool_config(None),
tokenizer_group = TokenizerGroup(
tokenizer_id="gpt2",
enable_lora=enable_lora,
max_num_seqs=max_num_seqs,