Skip to content

GitLab

Commit a99300bd authored by zhuwenwen's avatar zhuwenwen
Browse files

Merge tag 'v0.10.2rc1' into v0.10.2rc1-dev

parents cc3e01c7 5438967f
Hide whitespace changes
Inline Side-by-side
Showing with 411 additions and 317 deletions
tests/lora/test_lora_manager.py
View file @ a99300bd
...@@ -21,6 +21,8 @@ from vllm.lora.worker_manager import (LRUCacheWorkerLoRAManager, ...@@ -21,6 +21,8 @@ from vllm.lora.worker_manager import (LRUCacheWorkerLoRAManager,
WorkerLoRAManager) WorkerLoRAManager)
from vllm.platforms import current_platform from vllm.platforms import current_platform
from .utils import create_peft_lora
EMBEDDING_MODULES = { EMBEDDING_MODULES = {
"embed_tokens": "input_embeddings", "embed_tokens": "input_embeddings",
"lm_head": "output_embeddings", "lm_head": "output_embeddings",
...@@ -35,17 +37,6 @@ DEVICES = ([ ...@@ -35,17 +37,6 @@ DEVICES = ([
DEFAULT_DTYPE = torch.get_default_dtype() DEFAULT_DTYPE = torch.get_default_dtype()
@pytest.fixture(scope="function", autouse=True)
def use_v0_only(monkeypatch: pytest.MonkeyPatch):
"""
Some tests depend on V0 internals. Since both V0 and V1 use the same
LoRAModelManager it is okay to just test V0.
"""
with monkeypatch.context() as m:
m.setenv('VLLM_USE_V1', '0')
yield
@pytest.mark.parametrize("device", DEVICES) @pytest.mark.parametrize("device", DEVICES)
def test_from_lora_tensors(sql_lora_files, device): def test_from_lora_tensors(sql_lora_files, device):
tensors = load_file( tensors = load_file(
...@@ -326,7 +317,6 @@ def test_lru_lora_model_manager(dist_init, dummy_model, device): ...@@ -326,7 +317,6 @@ def test_lru_lora_model_manager(dist_init, dummy_model, device):
max_loras=2, max_loras=2,
lora_dtype=DEFAULT_DTYPE), lora_dtype=DEFAULT_DTYPE),
device=device) device=device)
assert all(x is None for x in manager.lora_index_to_id) assert all(x is None for x in manager.lora_index_to_id)
# Add up to capacity # Add up to capacity
...@@ -430,32 +420,40 @@ def test_lru_lora_model_manager(dist_init, dummy_model, device): ...@@ -430,32 +420,40 @@ def test_lru_lora_model_manager(dist_init, dummy_model, device):
@pytest.mark.parametrize("device", DEVICES) @pytest.mark.parametrize("device", DEVICES)
def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings, def test_lru_cache_worker_adapter_manager(dist_init, dummy_model, device,
sql_lora_files, device): tmp_path):
lora_config = LoRAConfig(max_lora_rank=8, lora_config = LoRAConfig(max_lora_rank=8,
max_cpu_loras=4, max_cpu_loras=4,
max_loras=4, max_loras=4,
lora_dtype=DEFAULT_DTYPE) lora_dtype=DEFAULT_DTYPE)
dummy_lora_files = f"{tmp_path}/lora_adapter"
os.makedirs(dummy_lora_files, exist_ok=True)
create_peft_lora(
dummy_model,
save_dir=dummy_lora_files,
target_modules=["layer1.dense1", "dense2"],
lora_dtype=DEFAULT_DTYPE,
)
worker_adapter_manager = LRUCacheWorkerLoRAManager( worker_adapter_manager = LRUCacheWorkerLoRAManager(
4, 2, llama_2_7b_model_extra_embeddings.unpadded_vocab_size - 4, 2,
lora_config.lora_extra_vocab_size, lora_config, device, dummy_model.unpadded_vocab_size - lora_config.lora_extra_vocab_size,
EMBEDDING_MODULES, EMBEDDING_PADDING_MODULES) lora_config, device, EMBEDDING_MODULES, EMBEDDING_PADDING_MODULES)
worker_adapter_manager.create_lora_manager( worker_adapter_manager.create_lora_manager(dummy_model)
llama_2_7b_model_extra_embeddings)
mapping = LoRAMapping([], []) mapping = LoRAMapping([], [])
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("2", 2, sql_lora_files) LoRARequest("2", 2, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 2} assert worker_adapter_manager.list_adapters() == {1, 2}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
assert worker_adapter_manager._adapter_manager.lora_index_to_id[1] == 2 assert worker_adapter_manager._adapter_manager.lora_index_to_id[1] == 2
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("3", 3, sql_lora_files), LoRARequest("3", 3, dummy_lora_files),
LoRARequest("4", 4, sql_lora_files) LoRARequest("4", 4, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 2, 3, 4} assert worker_adapter_manager.list_adapters() == {1, 2, 3, 4}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -464,9 +462,9 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -464,9 +462,9 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
assert worker_adapter_manager._adapter_manager.lora_index_to_id[3] == 4 assert worker_adapter_manager._adapter_manager.lora_index_to_id[3] == 4
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("2", 2, sql_lora_files), LoRARequest("2", 2, dummy_lora_files),
LoRARequest("5", 5, sql_lora_files) LoRARequest("5", 5, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 2, 4, 5} assert worker_adapter_manager.list_adapters() == {1, 2, 4, 5}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -475,9 +473,9 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -475,9 +473,9 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
assert worker_adapter_manager._adapter_manager.lora_index_to_id[3] == 4 assert worker_adapter_manager._adapter_manager.lora_index_to_id[3] == 4
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("1", 1, sql_lora_files) LoRARequest("1", 1, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 2, 4, 5} assert worker_adapter_manager.list_adapters() == {1, 2, 4, 5}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -486,9 +484,9 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -486,9 +484,9 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
assert worker_adapter_manager._adapter_manager.lora_index_to_id[3] == 4 assert worker_adapter_manager._adapter_manager.lora_index_to_id[3] == 4
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("6", 6, sql_lora_files), LoRARequest("6", 6, dummy_lora_files),
LoRARequest("7", 7, sql_lora_files), LoRARequest("7", 7, dummy_lora_files),
LoRARequest("8", 8, sql_lora_files) LoRARequest("8", 8, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 6, 7, 8} assert worker_adapter_manager.list_adapters() == {1, 6, 7, 8}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -499,11 +497,11 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -499,11 +497,11 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
# Over capacity # Over capacity
with pytest.raises(RuntimeError): with pytest.raises(RuntimeError):
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("10", 10, sql_lora_files), LoRARequest("10", 10, dummy_lora_files),
LoRARequest("11", 11, sql_lora_files), LoRARequest("11", 11, dummy_lora_files),
LoRARequest("12", 12, sql_lora_files), LoRARequest("12", 12, dummy_lora_files),
LoRARequest("13", 13, sql_lora_files), LoRARequest("13", 13, dummy_lora_files),
LoRARequest("14", 14, sql_lora_files) LoRARequest("14", 14, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.device == device assert worker_adapter_manager.device == device
...@@ -512,33 +510,41 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -512,33 +510,41 @@ def test_lru_cache_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
@pytest.mark.parametrize("device", DEVICES) @pytest.mark.parametrize("device", DEVICES)
def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings, def test_worker_adapter_manager(dist_init, dummy_model_gate_up, device,
sql_lora_files, device): tmp_path):
# Should remove every LoRA not specified in the request. # Should remove every LoRA not specified in the request.
lora_config = LoRAConfig(max_lora_rank=8, lora_config = LoRAConfig(max_lora_rank=8,
max_cpu_loras=4, max_cpu_loras=4,
max_loras=4, max_loras=4,
lora_dtype=DEFAULT_DTYPE) lora_dtype=DEFAULT_DTYPE)
worker_adapter_manager = WorkerLoRAManager( worker_adapter_manager = WorkerLoRAManager(
4, 2, llama_2_7b_model_extra_embeddings.unpadded_vocab_size - 4, 2, dummy_model_gate_up.unpadded_vocab_size -
lora_config.lora_extra_vocab_size, lora_config, device, lora_config.lora_extra_vocab_size, lora_config, device,
EMBEDDING_MODULES, EMBEDDING_PADDING_MODULES) EMBEDDING_MODULES, EMBEDDING_PADDING_MODULES)
worker_adapter_manager.create_lora_manager( worker_adapter_manager.create_lora_manager(dummy_model_gate_up)
llama_2_7b_model_extra_embeddings)
dummy_lora_files = f"{tmp_path}/lora_adapter"
os.makedirs(dummy_lora_files, exist_ok=True)
create_peft_lora(
dummy_model_gate_up,
save_dir=dummy_lora_files,
target_modules=["layer1.dense1", "dense2"],
lora_dtype=DEFAULT_DTYPE,
)
mapping = LoRAMapping([], []) mapping = LoRAMapping([], [])
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("2", 2, sql_lora_files) LoRARequest("2", 2, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 2} assert worker_adapter_manager.list_adapters() == {1, 2}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
assert worker_adapter_manager._adapter_manager.lora_index_to_id[1] == 2 assert worker_adapter_manager._adapter_manager.lora_index_to_id[1] == 2
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("3", 3, sql_lora_files), LoRARequest("3", 3, dummy_lora_files),
LoRARequest("4", 4, sql_lora_files) LoRARequest("4", 4, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 3, 4} assert worker_adapter_manager.list_adapters() == {1, 3, 4}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -546,9 +552,9 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -546,9 +552,9 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
assert worker_adapter_manager._adapter_manager.lora_index_to_id[2] == 4 assert worker_adapter_manager._adapter_manager.lora_index_to_id[2] == 4
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("2", 2, sql_lora_files), LoRARequest("2", 2, dummy_lora_files),
LoRARequest("5", 5, sql_lora_files) LoRARequest("5", 5, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1, 2, 5} assert worker_adapter_manager.list_adapters() == {1, 2, 5}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -556,9 +562,9 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -556,9 +562,9 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
assert worker_adapter_manager._adapter_manager.lora_index_to_id[2] == 5 assert worker_adapter_manager._adapter_manager.lora_index_to_id[2] == 5
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("1", 1, sql_lora_files), LoRARequest("1", 1, dummy_lora_files),
LoRARequest("1", 1, sql_lora_files) LoRARequest("1", 1, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {1} assert worker_adapter_manager.list_adapters() == {1}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 1
...@@ -566,9 +572,9 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -566,9 +572,9 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
assert worker_adapter_manager._adapter_manager.lora_index_to_id[2] is None assert worker_adapter_manager._adapter_manager.lora_index_to_id[2] is None
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("6", 6, sql_lora_files), LoRARequest("6", 6, dummy_lora_files),
LoRARequest("7", 7, sql_lora_files), LoRARequest("7", 7, dummy_lora_files),
LoRARequest("8", 8, sql_lora_files) LoRARequest("8", 8, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.list_adapters() == {6, 7, 8} assert worker_adapter_manager.list_adapters() == {6, 7, 8}
assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 8 assert worker_adapter_manager._adapter_manager.lora_index_to_id[0] == 8
...@@ -578,11 +584,11 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings, ...@@ -578,11 +584,11 @@ def test_worker_adapter_manager(llama_2_7b_model_extra_embeddings,
# Over capacity # Over capacity
with pytest.raises(RuntimeError): with pytest.raises(RuntimeError):
worker_adapter_manager.set_active_adapters([ worker_adapter_manager.set_active_adapters([
LoRARequest("10", 10, sql_lora_files), LoRARequest("10", 10, dummy_lora_files),
LoRARequest("11", 11, sql_lora_files), LoRARequest("11", 11, dummy_lora_files),
LoRARequest("12", 12, sql_lora_files), LoRARequest("12", 12, dummy_lora_files),
LoRARequest("13", 13, sql_lora_files), LoRARequest("13", 13, dummy_lora_files),
LoRARequest("14", 14, sql_lora_files) LoRARequest("14", 14, dummy_lora_files)
], mapping) ], mapping)
assert worker_adapter_manager.device == device assert worker_adapter_manager.device == device
......
tests/lora/test_mixtral.py
View file @ a99300bd
...@@ -52,7 +52,6 @@ def test_mixtral_lora(mixtral_lora_files, tp_size): ...@@ -52,7 +52,6 @@ def test_mixtral_lora(mixtral_lora_files, tp_size):
max_loras=4, max_loras=4,
distributed_executor_backend="ray", distributed_executor_backend="ray",
tensor_parallel_size=tp_size, tensor_parallel_size=tp_size,
enable_chunked_prefill=True,
) )
expected_lora_output = [ expected_lora_output = [
......
tests/lora/test_worker.py
View file @ a99300bd
...@@ -4,17 +4,15 @@ ...@@ -4,17 +4,15 @@
import os import os
import random import random
import tempfile import tempfile
from typing import Union
from unittest.mock import patch from unittest.mock import patch
import vllm.envs as envs
from vllm.config import (CacheConfig, DeviceConfig, LoadConfig, LoRAConfig, from vllm.config import (CacheConfig, DeviceConfig, LoadConfig, LoRAConfig,
ModelConfig, ParallelConfig, SchedulerConfig, ModelConfig, ParallelConfig, SchedulerConfig,
VllmConfig) VllmConfig)
from vllm.lora.models import LoRAMapping from vllm.lora.models import LoRAMapping
from vllm.lora.request import LoRARequest from vllm.lora.request import LoRARequest
from vllm.v1.worker.gpu_worker import Worker as V1Worker
from vllm.worker.worker import Worker from vllm.v1.worker.gpu_worker import Worker
from ..utils import models_path_prefix from ..utils import models_path_prefix
NUM_LORAS = 16 NUM_LORAS = 16
...@@ -23,18 +21,11 @@ NUM_LORAS = 16 ...@@ -23,18 +21,11 @@ NUM_LORAS = 16
@patch.dict(os.environ, {"RANK": "0"}) @patch.dict(os.environ, {"RANK": "0"})
def test_worker_apply_lora(sql_lora_files): def test_worker_apply_lora(sql_lora_files):
def set_active_loras(worker: Union[Worker, V1Worker], def set_active_loras(worker: Worker, lora_requests: list[LoRARequest]):
lora_requests: list[LoRARequest]):
lora_mapping = LoRAMapping([], []) lora_mapping = LoRAMapping([], [])
if isinstance(worker, Worker):
# v0 case
worker.model_runner.set_active_loras(lora_requests, lora_mapping)
else:
# v1 case
worker.model_runner.lora_manager.set_active_adapters(
lora_requests, lora_mapping)
worker_cls = V1Worker if envs.VLLM_USE_V1 else Worker worker.model_runner.lora_manager.set_active_adapters(
lora_requests, lora_mapping)
vllm_config = VllmConfig( vllm_config = VllmConfig(
model_config=ModelConfig( model_config=ModelConfig(
...@@ -63,7 +54,7 @@ def test_worker_apply_lora(sql_lora_files): ...@@ -63,7 +54,7 @@ def test_worker_apply_lora(sql_lora_files):
max_cpu_loras=NUM_LORAS, max_cpu_loras=NUM_LORAS,
max_loras=NUM_LORAS), max_loras=NUM_LORAS),
) )
worker = worker_cls( worker = Worker(
vllm_config=vllm_config, vllm_config=vllm_config,
local_rank=0, local_rank=0,
rank=0, rank=0,
......
tests/lora/utils.py
View file @ a99300bd
# SPDX-License-Identifier: Apache-2.0 # SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
import json
import os
from dataclasses import dataclass from dataclasses import dataclass
from typing import Optional, Union from typing import Optional, Union
import torch import torch
from safetensors.torch import save_file
from vllm.lora.lora import LoRALayerWeights, PackedLoRALayerWeights from vllm.lora.lora import LoRALayerWeights, PackedLoRALayerWeights
...@@ -340,3 +343,76 @@ def generate_data_for_nslices( ...@@ -340,3 +343,76 @@ def generate_data_for_nslices(
seq_len_tensor, seq_len_tensor,
indices, indices,
) )
def create_peft_lora(
model: torch.nn.Module,
save_dir: str,
target_modules: list[str],
rank: int = 8,
alpha: int = 16,
dropout: float = 0.1,
lora_dtype: torch.dtype = torch.float16,
) -> dict[str, torch.Tensor]:
lora_weights = {}
adapter_config = {
"peft_type": "LORA",
"auto_mapping": None,
"base_model_name_or_path": "dummy_model",
"revision": None,
"task_type": "CAUSAL_LM",
"inference_mode": False,
"r": rank,
"lora_alpha": alpha,
"lora_dropout": dropout,
"fan_in_fan_out": False,
"bias": "none",
"modules_to_save": None,
"init_lora_weights": True,
"layers_to_transform": None,
"layers_pattern": None,
"target_modules": target_modules,
"exclude_modules": None,
"use_rslora": False,
"use_dora": False,
"loftq_config": None,
}
for module_name in target_modules:
module = model
for attr in module_name.split("."):
module = getattr(module, attr)
if hasattr(module, "input_size") and hasattr(module, "output_size"):
in_features = module.input_size
out_features = module.output_size
elif hasattr(module, "embedding_dim") and hasattr(
module, "num_embeddings"):
# ParallelLMHead
in_features = module.embedding_dim
out_features = module.num_embeddings
else:
raise ValueError(
f"Unable to determine dimensions for module {module_name}")
lora_A = torch.randn(rank, in_features, dtype=lora_dtype)
torch.nn.init.kaiming_uniform_(lora_A, a=5**0.5)
lora_B = torch.zeros(out_features, rank, dtype=lora_dtype)
# PEFT style
lora_weights[f"base_model.model.{module_name}.lora_A.weight"] = lora_A
lora_weights[f"base_model.model.{module_name}.lora_B.weight"] = lora_B
config_path = os.path.join(save_dir, "adapter_config.json")
with open(config_path, "w", encoding="utf-8") as f:
json.dump(adapter_config, f, indent=2, ensure_ascii=False)
weights_path = os.path.join(save_dir, "adapter_model.safetensors")
save_file(lora_weights, weights_path)
return lora_weights
tests/models/language/generation/test_common.py
View file @ a99300bd
...@@ -94,7 +94,8 @@ AITER_MODEL_LIST = [ ...@@ -94,7 +94,8 @@ AITER_MODEL_LIST = [
pytest.param( pytest.param(
os.path.join(models_path_prefix, "allenai/OLMoE-1B-7B-0924-Instruct"), os.path.join(models_path_prefix, "allenai/OLMoE-1B-7B-0924-Instruct"),
marks=[pytest.mark.cpu_model], marks=[pytest.mark.cpu_model],
) ),
pytest.param("swiss-ai/Apertus-8B"), # apertus
]) ])
@pytest.mark.parametrize("max_tokens", [32]) @pytest.mark.parametrize("max_tokens", [32])
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
......
tests/models/language/generation/test_hybrid.py
View file @ a99300bd
...@@ -33,6 +33,7 @@ HYBRID_MODELS = [ ...@@ -33,6 +33,7 @@ HYBRID_MODELS = [
os.path.join(models_path_prefix, "hmellor/tiny-random-BambaForCausalLM"), os.path.join(models_path_prefix, "hmellor/tiny-random-BambaForCausalLM"),
os.path.join(models_path_prefix, "ibm-granite/granite-4.0-tiny-preview"), os.path.join(models_path_prefix, "ibm-granite/granite-4.0-tiny-preview"),
os.path.join(models_path_prefix, "tiiuae/Falcon-H1-0.5B-Base"), os.path.join(models_path_prefix, "tiiuae/Falcon-H1-0.5B-Base"),
os.path.join(models_path_prefix, "LiquidAI/LFM2-1.2B"),
] ]
HF_UNSUPPORTED_MODELS = [ HF_UNSUPPORTED_MODELS = [
...@@ -47,26 +48,28 @@ HF_UNSUPPORTED_MODELS = [ ...@@ -47,26 +48,28 @@ HF_UNSUPPORTED_MODELS = [
] ]
V1_SUPPORTED_MODELS = [ V1_SUPPORTED_MODELS = [
os.path.join(models_path_prefix, "state-spaces/mamba-130m-hf"), os.path.join(models_path_prefix,"state-spaces/mamba-130m-hf"),
os.path.join(models_path_prefix, "ai21labs/Jamba-tiny-dev"), os.path.join(models_path_prefix,"ai21labs/Jamba-tiny-dev"),
os.path.join(models_path_prefix, "yujiepan/mamba2-codestral-v0.1-tiny-random"), os.path.join(models_path_prefix,"yujiepan/mamba2-codestral-v0.1-tiny-random"),
os.path.join(models_path_prefix, "Zyphra/Zamba2-1.2B-instruct"), os.path.join(models_path_prefix,"Zyphra/Zamba2-1.2B-instruct"),
os.path.join(models_path_prefix, "hmellor/tiny-random-BambaForCausalLM"), os.path.join(models_path_prefix,"hmellor/tiny-random-BambaForCausalLM"),
os.path.join(models_path_prefix, "ibm-granite/granite-4.0-tiny-preview"), os.path.join(models_path_prefix,"ibm-granite/granite-4.0-tiny-preview"),
os.path.join(models_path_prefix, "tiiuae/Falcon-H1-0.5B-Base"), os.path.join(models_path_prefix,"tiiuae/Falcon-H1-0.5B-Base"),
os.path.join(models_path_prefix,"LiquidAI/LFM2-1.2B"),
]
FULL_CUDA_GRAPH_MODELS = [
os.path.join(models_path_prefix,"ai21labs/Jamba-tiny-dev"),
os.path.join(models_path_prefix,"Zyphra/Zamba2-1.2B-instruct"),
] ]
V0_UNSUPPORTED_MODELS = [
os.path.join(models_path_prefix,"LiquidAI/LFM2-1.2B"),
]
# Avoid OOM # Avoid OOM
MAX_NUM_SEQS = 4 MAX_NUM_SEQS = 4
# Once we add support for FCG in Mamba1, this list will be removed and tests
# all test cases will use enforce_eager=False
ENFORCE_EAGER_MODELS_V1 = [
os.path.join(models_path_prefix, "state-spaces/mamba-130m-hf"),
os.path.join(models_path_prefix, "ai21labs/Jamba-tiny-dev"),
]
@pytest.mark.parametrize("model", SSM_MODELS + HYBRID_MODELS) @pytest.mark.parametrize("model", SSM_MODELS + HYBRID_MODELS)
@pytest.mark.parametrize("max_tokens", [64]) @pytest.mark.parametrize("max_tokens", [64])
...@@ -99,31 +102,23 @@ def test_models( ...@@ -99,31 +102,23 @@ def test_models(
else: else:
hf_outputs = None hf_outputs = None
with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model: with monkeypatch.context() as m:
vllm_v0_outputs = vllm_model.generate_greedy_logprobs( m.setenv("VLLM_USE_V1", "0")
example_prompts, max_tokens, num_logprobs) if model not in V0_UNSUPPORTED_MODELS:
with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model:
vllm_v0_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
else:
vllm_v0_outputs = None
if model in V1_SUPPORTED_MODELS: if model in V1_SUPPORTED_MODELS:
enforce_eager = False with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model:
with monkeypatch.context() as m: vllm_v1_outputs = vllm_model.generate_greedy_logprobs(
m.setenv("VLLM_USE_V1", "1") example_prompts, max_tokens, num_logprobs)
if model in HYBRID_MODELS:
# required due to reorder_batch behaviour
m.setenv("VLLM_ATTENTION_BACKEND", "FLASHINFER")
if model in ENFORCE_EAGER_MODELS_V1:
enforce_eager = True
with vllm_runner(model,
max_num_seqs=MAX_NUM_SEQS,
enforce_eager=enforce_eager,
enable_prefix_caching=False) as vllm_model:
vllm_v1_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
else: else:
vllm_v1_outputs = None vllm_v1_outputs = None
if hf_outputs is not None: if hf_outputs is not None and vllm_v0_outputs is not None:
check_logprobs_close( check_logprobs_close(
outputs_0_lst=hf_outputs, outputs_0_lst=hf_outputs,
outputs_1_lst=vllm_v0_outputs, outputs_1_lst=vllm_v0_outputs,
...@@ -133,6 +128,7 @@ def test_models( ...@@ -133,6 +128,7 @@ def test_models(
if model in V1_SUPPORTED_MODELS: if model in V1_SUPPORTED_MODELS:
ref_outputs = hf_outputs if hf_outputs is not None else vllm_v0_outputs ref_outputs = hf_outputs if hf_outputs is not None else vllm_v0_outputs
assert ref_outputs is not None
check_logprobs_close( check_logprobs_close(
outputs_0_lst=ref_outputs, outputs_0_lst=ref_outputs,
outputs_1_lst=vllm_v1_outputs, outputs_1_lst=vllm_v1_outputs,
...@@ -141,7 +137,7 @@ def test_models( ...@@ -141,7 +137,7 @@ def test_models(
) )
@pytest.mark.parametrize("model", SSM_MODELS + HYBRID_MODELS) @pytest.mark.parametrize("model", [SSM_MODELS[0], HYBRID_MODELS[0]])
@pytest.mark.parametrize("max_tokens", [64]) @pytest.mark.parametrize("max_tokens", [64])
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
def test_batching( def test_batching(
...@@ -151,7 +147,6 @@ def test_batching( ...@@ -151,7 +147,6 @@ def test_batching(
max_tokens: int, max_tokens: int,
num_logprobs: int, num_logprobs: int,
) -> None: ) -> None:
try: try:
model_info = HF_EXAMPLE_MODELS.find_hf_info(model) model_info = HF_EXAMPLE_MODELS.find_hf_info(model)
model_info.check_available_online(on_fail="skip") model_info.check_available_online(on_fail="skip")
...@@ -189,29 +184,32 @@ def test_chunked_prefill( ...@@ -189,29 +184,32 @@ def test_chunked_prefill(
max_tokens: int, max_tokens: int,
num_logprobs: int, num_logprobs: int,
chunked_prefill_token_size: int, chunked_prefill_token_size: int,
monkeypatch,
) -> None: ) -> None:
max_num_seqs = chunked_prefill_token_size max_num_seqs = chunked_prefill_token_size
max_num_batched_tokens = chunked_prefill_token_size max_num_batched_tokens = chunked_prefill_token_size
with vllm_runner(model, with monkeypatch.context() as m:
enable_chunked_prefill=True, m.setenv("VLLM_USE_V1", "0")
max_num_batched_tokens=max_num_batched_tokens, with vllm_runner(model,
max_num_seqs=max_num_seqs) as vllm_model: enable_chunked_prefill=True,
chunked = vllm_model.generate_greedy_logprobs(example_prompts, max_num_batched_tokens=max_num_batched_tokens,
max_tokens, num_logprobs) max_num_seqs=max_num_seqs) as vllm_model:
chunked = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
with vllm_runner(model, with vllm_runner(model,
enable_chunked_prefill=False, enable_chunked_prefill=False,
max_num_seqs=max_num_seqs) as vllm_model: max_num_seqs=max_num_seqs) as vllm_model:
non_chunked = vllm_model.generate_greedy_logprobs( non_chunked = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs) example_prompts, max_tokens, num_logprobs)
check_logprobs_close( check_logprobs_close(
outputs_0_lst=chunked, outputs_0_lst=chunked,
outputs_1_lst=non_chunked, outputs_1_lst=non_chunked,
name_0="chunked", name_0="chunked",
name_1="non_chunked", name_1="non_chunked",
) )
@pytest.mark.parametrize("model", [SSM_MODELS[0], HYBRID_MODELS[0]]) @pytest.mark.parametrize("model", [SSM_MODELS[0], HYBRID_MODELS[0]])
...@@ -282,25 +280,29 @@ def test_models_preemption_recompute( ...@@ -282,25 +280,29 @@ def test_models_preemption_recompute(
example_prompts, example_prompts,
model: str, model: str,
max_tokens: int, max_tokens: int,
monkeypatch,
) -> None: ) -> None:
""" """
Tests that outputs are identical with and w/o preemptions (recompute). Tests that outputs are identical with and w/o preemptions (recompute).
""" """
with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model: with monkeypatch.context() as m:
scheduler = vllm_model.llm.llm_engine.scheduler[0] m.setenv("VLLM_USE_V1", "0")
scheduler.ENABLE_ARTIFICIAL_PREEMPT = True with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model:
preempt_vllm_outputs = vllm_model.generate_greedy( scheduler = vllm_model.llm.llm_engine.scheduler[0]
example_prompts, max_tokens) scheduler.ENABLE_ARTIFICIAL_PREEMPT = True
preempt_vllm_outputs = vllm_model.generate_greedy(
scheduler.ENABLE_ARTIFICIAL_PREEMPT = False example_prompts, max_tokens)
vllm_outputs = vllm_model.generate_greedy(example_prompts, max_tokens)
scheduler.ENABLE_ARTIFICIAL_PREEMPT = False
check_outputs_equal( vllm_outputs = vllm_model.generate_greedy(example_prompts,
outputs_0_lst=preempt_vllm_outputs, max_tokens)
outputs_1_lst=vllm_outputs,
name_0="vllm_preepmtions", check_outputs_equal(
name_1="vllm", outputs_0_lst=preempt_vllm_outputs,
) outputs_1_lst=vllm_outputs,
name_0="vllm_preepmtions",
name_1="vllm",
)
@pytest.mark.parametrize("model", [SSM_MODELS[0], HYBRID_MODELS[0]]) @pytest.mark.parametrize("model", [SSM_MODELS[0], HYBRID_MODELS[0]])
...@@ -376,7 +378,7 @@ def test_distributed_correctness( ...@@ -376,7 +378,7 @@ def test_distributed_correctness(
) )
@pytest.mark.parametrize("model", ["Zyphra/Zamba2-1.2B-instruct"]) @pytest.mark.parametrize("model", FULL_CUDA_GRAPH_MODELS)
@pytest.mark.parametrize("max_tokens", [64]) @pytest.mark.parametrize("max_tokens", [64])
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
def test_full_cuda_graph( def test_full_cuda_graph(
...@@ -403,23 +405,20 @@ def test_full_cuda_graph( ...@@ -403,23 +405,20 @@ def test_full_cuda_graph(
else: else:
hf_outputs = None hf_outputs = None
with monkeypatch.context() as m:
m.setenv("VLLM_USE_V1", "0")
if model not in V0_UNSUPPORTED_MODELS:
with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model:
vllm_v0_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
else:
vllm_v0_outputs = None
with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model: with vllm_runner(model, max_num_seqs=MAX_NUM_SEQS) as vllm_model:
vllm_v0_outputs = vllm_model.generate_greedy_logprobs( vllm_v1_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs) example_prompts, max_tokens, num_logprobs)
with monkeypatch.context() as m: if hf_outputs is not None and vllm_v0_outputs is not None:
m.setenv("VLLM_USE_V1", "1")
if model in HYBRID_MODELS:
# required due to reorder_batch behaviour
m.setenv("VLLM_ATTENTION_BACKEND", "FLASHINFER")
with vllm_runner(model,
max_num_seqs=MAX_NUM_SEQS,
compilation_config={'full_cuda_graph': True},
enable_prefix_caching=False) as vllm_model:
vllm_v1_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
if hf_outputs is not None:
check_logprobs_close( check_logprobs_close(
outputs_0_lst=hf_outputs, outputs_0_lst=hf_outputs,
outputs_1_lst=vllm_v0_outputs, outputs_1_lst=vllm_v0_outputs,
...@@ -428,6 +427,7 @@ def test_full_cuda_graph( ...@@ -428,6 +427,7 @@ def test_full_cuda_graph(
) )
ref_outputs = hf_outputs if hf_outputs is not None else vllm_v0_outputs ref_outputs = hf_outputs if hf_outputs is not None else vllm_v0_outputs
assert ref_outputs is not None
check_logprobs_close( check_logprobs_close(
outputs_0_lst=ref_outputs, outputs_0_lst=ref_outputs,
outputs_1_lst=vllm_v1_outputs, outputs_1_lst=vllm_v1_outputs,
...@@ -463,24 +463,20 @@ def test_fp32_state( ...@@ -463,24 +463,20 @@ def test_fp32_state(
else: else:
hf_outputs = None hf_outputs = None
with vllm_runner(model,
max_num_seqs=MAX_NUM_SEQS,
mamba_ssm_cache_dtype="float32") as vllm_model:
vllm_v0_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
with monkeypatch.context() as m: with monkeypatch.context() as m:
m.setenv("VLLM_USE_V1", "1") m.setenv("VLLM_USE_V1", "0")
if model in HYBRID_MODELS:
# required due to reorder_batch behaviour
m.setenv("VLLM_ATTENTION_BACKEND", "FLASHINFER")
with vllm_runner(model, with vllm_runner(model,
max_num_seqs=MAX_NUM_SEQS, max_num_seqs=MAX_NUM_SEQS,
mamba_ssm_cache_dtype="float32", mamba_ssm_cache_dtype="float32") as vllm_model:
enable_prefix_caching=False) as vllm_model: vllm_v0_outputs = vllm_model.generate_greedy_logprobs(
vllm_v1_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs) example_prompts, max_tokens, num_logprobs)
with vllm_runner(model,
max_num_seqs=MAX_NUM_SEQS,
mamba_ssm_cache_dtype="float32") as vllm_model:
vllm_v1_outputs = vllm_model.generate_greedy_logprobs(
example_prompts, max_tokens, num_logprobs)
if hf_outputs is not None: if hf_outputs is not None:
check_logprobs_close( check_logprobs_close(
outputs_0_lst=hf_outputs, outputs_0_lst=hf_outputs,
......
tests/models/language/pooling/embed_utils.py
View file @ a99300bd
...@@ -51,6 +51,9 @@ def correctness_test_embed_models(hf_runner, ...@@ -51,6 +51,9 @@ def correctness_test_embed_models(hf_runner,
vllm_extra_kwargs = vllm_extra_kwargs or {} vllm_extra_kwargs = vllm_extra_kwargs or {}
vllm_extra_kwargs["dtype"] = model_info.dtype vllm_extra_kwargs["dtype"] = model_info.dtype
if model_info.hf_overrides is not None:
vllm_extra_kwargs["hf_overrides"] = model_info.hf_overrides
with vllm_runner(model_info.name, with vllm_runner(model_info.name,
runner="pooling", runner="pooling",
max_model_len=None, max_model_len=None,
......
tests/models/language/pooling/mteb_utils.py
View file @ a99300bd
...@@ -172,6 +172,9 @@ def mteb_test_embed_models(hf_runner, ...@@ -172,6 +172,9 @@ def mteb_test_embed_models(hf_runner,
vllm_extra_kwargs = vllm_extra_kwargs or {} vllm_extra_kwargs = vllm_extra_kwargs or {}
vllm_extra_kwargs["dtype"] = model_info.dtype vllm_extra_kwargs["dtype"] = model_info.dtype
if model_info.hf_overrides is not None:
vllm_extra_kwargs["hf_overrides"] = model_info.hf_overrides
with vllm_runner(model_info.name, with vllm_runner(model_info.name,
runner="pooling", runner="pooling",
max_model_len=None, max_model_len=None,
...@@ -284,6 +287,9 @@ def mteb_test_rerank_models(hf_runner, ...@@ -284,6 +287,9 @@ def mteb_test_rerank_models(hf_runner,
vllm_extra_kwargs = vllm_extra_kwargs or {} vllm_extra_kwargs = vllm_extra_kwargs or {}
vllm_extra_kwargs["dtype"] = model_info.dtype vllm_extra_kwargs["dtype"] = model_info.dtype
if model_info.hf_overrides is not None:
vllm_extra_kwargs["hf_overrides"] = model_info.hf_overrides
with vllm_runner(model_info.name, with vllm_runner(model_info.name,
runner="pooling", runner="pooling",
max_model_len=None, max_model_len=None,
......
tests/models/language/pooling/test_bge_reranker_v2_gemma.py
View file @ a99300bd
...@@ -13,7 +13,14 @@ from .mteb_utils import VllmMtebEncoder, mteb_test_rerank_models ...@@ -13,7 +13,14 @@ from .mteb_utils import VllmMtebEncoder, mteb_test_rerank_models
RERANK_MODELS = [ RERANK_MODELS = [
LASTPoolingRerankModelInfo("BAAI/bge-reranker-v2-gemma", LASTPoolingRerankModelInfo("BAAI/bge-reranker-v2-gemma",
architecture="GemmaForSequenceClassification"), architecture="GemmaForSequenceClassification",
hf_overrides={
"architectures":
["GemmaForSequenceClassification"],
"classifier_from_token": ["Yes"],
"method":
"no_post_processing",
}),
] ]
PROMPT = "Given a query A and a passage B, determine whether the passage contains an answer to the query by providing a prediction of either 'Yes' or 'No'." # noqa: E501 PROMPT = "Given a query A and a passage B, determine whether the passage contains an answer to the query by providing a prediction of either 'Yes' or 'No'." # noqa: E501
...@@ -119,22 +126,9 @@ class GemmaMtebEncoder(VllmMtebEncoder): ...@@ -119,22 +126,9 @@ class GemmaMtebEncoder(VllmMtebEncoder):
@pytest.mark.parametrize("model_info", RERANK_MODELS) @pytest.mark.parametrize("model_info", RERANK_MODELS)
def test_rerank_models_mteb(vllm_runner, model_info: RerankModelInfo, def test_rerank_models_mteb(vllm_runner, model_info: RerankModelInfo) -> None:
monkeypatch) -> None:
monkeypatch.setenv("VLLM_USE_V1", "0")
assert model_info.architecture == "GemmaForSequenceClassification"
vllm_extra_kwargs: dict[str, Any] = {
"hf_overrides": {
"architectures": ["GemmaForSequenceClassification"],
"classifier_from_token": ["Yes"],
"method": "no_post_processing",
}
}
mteb_test_rerank_models(GemmaRerankerHfRunner, mteb_test_rerank_models(GemmaRerankerHfRunner,
vllm_runner, vllm_runner,
model_info, model_info,
vllm_extra_kwargs,
vllm_mteb_encoder=GemmaMtebEncoder) vllm_mteb_encoder=GemmaMtebEncoder)
tests/models/language/pooling/test_embedding.py
View file @ a99300bd
...@@ -13,14 +13,6 @@ from vllm.platforms import current_platform ...@@ -13,14 +13,6 @@ from vllm.platforms import current_platform
from ...utils import check_embeddings_close, check_transformers_version from ...utils import check_embeddings_close, check_transformers_version
@pytest.fixture(autouse=True)
def v1(run_with_both_engines):
# Simple autouse wrapper to run both engines for each test
# This can be promoted up to conftest.py to run for every
# test in a package
pass
@pytest.mark.parametrize( @pytest.mark.parametrize(
"model", "model",
[ [
...@@ -35,21 +27,15 @@ def v1(run_with_both_engines): ...@@ -35,21 +27,15 @@ def v1(run_with_both_engines):
os.path.join(models_path_prefix, "intfloat/e5-mistral-7b-instruct"), os.path.join(models_path_prefix, "intfloat/e5-mistral-7b-instruct"),
# CPU v1 doesn't support sliding window # CPU v1 doesn't support sliding window
marks=[pytest.mark.core_model]), marks=[pytest.mark.core_model]),
# the qwen models interfere with each other (see PR
# https://github.com/vllm-project/vllm/pull/18720).
# To avoid this problem, for now we skip v0 since it will be
# deprecated anyway.
pytest.param(os.path.join(models_path_prefix, "ssmits/Qwen2-7B-Instruct-embed-base"), pytest.param(os.path.join(models_path_prefix, "ssmits/Qwen2-7B-Instruct-embed-base"),
marks=[pytest.mark.skip_v0, pytest.mark.cpu_model]), marks=[pytest.mark.cpu_model]),
# [Encoder-only] # [Encoder-only]
pytest.param(os.path.join(models_path_prefix, "BAAI/bge-base-en-v1.5"), marks=[pytest.mark.core_model]), pytest.param(os.path.join(models_path_prefix, "BAAI/bge-base-en-v1.5"), marks=[pytest.mark.core_model]),
pytest.param(os.path.join(models_path_prefix, "sentence-transformers/all-MiniLM-L12-v2")), pytest.param(os.path.join(models_path_prefix, "sentence-transformers/all-MiniLM-L12-v2")),
pytest.param(os.path.join(models_path_prefix, "intfloat/multilingual-e5-small")), pytest.param(os.path.join(models_path_prefix, "intfloat/multilingual-e5-small")),
pytest.param(os.path.join(models_path_prefix, "Alibaba-NLP/gte-Qwen2-1.5B-instruct"), pytest.param(os.path.join(models_path_prefix, "Alibaba-NLP/gte-Qwen2-1.5B-instruct")),
marks=[pytest.mark.skip_v1]),
# [Cross-Encoder] # [Cross-Encoder]
pytest.param(os.path.join(models_path_prefix, "sentence-transformers/stsb-roberta-base-v2"), pytest.param(os.path.join(models_path_prefix, "sentence-transformers/stsb-roberta-base-v2")),
marks=[pytest.mark.skip_v1]),
], ],
) )
......
tests/models/language/pooling/test_gritlm.py
View file @ a99300bd
...@@ -18,6 +18,7 @@ from ....utils import models_path_prefix ...@@ -18,6 +18,7 @@ from ....utils import models_path_prefix
MODEL_NAME = os.path.join(models_path_prefix, "parasail-ai/GritLM-7B-vllm") MODEL_NAME = os.path.join(models_path_prefix, "parasail-ai/GritLM-7B-vllm")
MAX_MODEL_LEN = 4000 MAX_MODEL_LEN = 4000
ATOL = 0.002
def _arr(arr): def _arr(arr):
...@@ -101,16 +102,16 @@ def get_test_data(): ...@@ -101,16 +102,16 @@ def get_test_data():
def validate_embed_output(q_rep: list[list[float]], d_rep: list[list[float]]): def validate_embed_output(q_rep: list[list[float]], d_rep: list[list[float]]):
cosine_sim_q0_d0 = 1 - cosine(q_rep[0], d_rep[0]) cosine_sim_q0_d0 = 1 - cosine(q_rep[0], d_rep[0])
assert cosine_sim_q0_d0 == pytest.approx(0.609, abs=0.001) assert cosine_sim_q0_d0 == pytest.approx(0.609, abs=ATOL)
cosine_sim_q0_d1 = 1 - cosine(q_rep[0], d_rep[1]) cosine_sim_q0_d1 = 1 - cosine(q_rep[0], d_rep[1])
assert cosine_sim_q0_d1 == pytest.approx(0.101, abs=0.001) assert cosine_sim_q0_d1 == pytest.approx(0.101, abs=ATOL)
cosine_sim_q1_d0 = 1 - cosine(q_rep[1], d_rep[0]) cosine_sim_q1_d0 = 1 - cosine(q_rep[1], d_rep[0])
assert cosine_sim_q1_d0 == pytest.approx(0.120, abs=0.001) assert cosine_sim_q1_d0 == pytest.approx(0.120, abs=ATOL)
cosine_sim_q1_d1 = 1 - cosine(q_rep[1], d_rep[1]) cosine_sim_q1_d1 = 1 - cosine(q_rep[1], d_rep[1])
assert cosine_sim_q1_d1 == pytest.approx(0.534, abs=0.001) assert cosine_sim_q1_d1 == pytest.approx(0.534, abs=ATOL)
def test_gritlm_offline_embedding(vllm_runner): def test_gritlm_offline_embedding(vllm_runner):
......
tests/models/language/pooling/test_gte.py
View file @ a99300bd
# SPDX-License-Identifier: Apache-2.0 # SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
from typing import Any
import pytest import pytest
...@@ -33,12 +32,15 @@ MODELS = [ ...@@ -33,12 +32,15 @@ MODELS = [
########### NewModel ########### NewModel
CLSPoolingEmbedModelInfo("Alibaba-NLP/gte-multilingual-base", CLSPoolingEmbedModelInfo("Alibaba-NLP/gte-multilingual-base",
architecture="GteNewModel", architecture="GteNewModel",
hf_overrides={"architectures": ["GteNewModel"]},
enable_test=True), enable_test=True),
CLSPoolingEmbedModelInfo("Alibaba-NLP/gte-base-en-v1.5", CLSPoolingEmbedModelInfo("Alibaba-NLP/gte-base-en-v1.5",
architecture="GteNewModel", architecture="GteNewModel",
hf_overrides={"architectures": ["GteNewModel"]},
enable_test=True), enable_test=True),
CLSPoolingEmbedModelInfo("Alibaba-NLP/gte-large-en-v1.5", CLSPoolingEmbedModelInfo("Alibaba-NLP/gte-large-en-v1.5",
architecture="GteNewModel", architecture="GteNewModel",
hf_overrides={"architectures": ["GteNewModel"]},
enable_test=True), enable_test=True),
########### Qwen2ForCausalLM ########### Qwen2ForCausalLM
LASTPoolingEmbedModelInfo("Alibaba-NLP/gte-Qwen2-1.5B-instruct", LASTPoolingEmbedModelInfo("Alibaba-NLP/gte-Qwen2-1.5B-instruct",
...@@ -60,11 +62,16 @@ MODELS = [ ...@@ -60,11 +62,16 @@ MODELS = [
] ]
RERANK_MODELS = [ RERANK_MODELS = [
# classifier_pooling: mean
CLSPoolingRerankModelInfo( CLSPoolingRerankModelInfo(
# classifier_pooling: mean
"Alibaba-NLP/gte-reranker-modernbert-base", "Alibaba-NLP/gte-reranker-modernbert-base",
architecture="ModernBertForSequenceClassification", architecture="ModernBertForSequenceClassification",
enable_test=True), enable_test=True),
CLSPoolingRerankModelInfo(
"Alibaba-NLP/gte-multilingual-reranker-base",
architecture="GteNewForSequenceClassification",
hf_overrides={"architectures": ["GteNewForSequenceClassification"]},
enable_test=True),
] ]
...@@ -75,12 +82,7 @@ def test_embed_models_mteb(hf_runner, vllm_runner, ...@@ -75,12 +82,7 @@ def test_embed_models_mteb(hf_runner, vllm_runner,
check_transformers_version(model_info.name, check_transformers_version(model_info.name,
max_transformers_version="4.53.2") max_transformers_version="4.53.2")
vllm_extra_kwargs: dict[str, Any] = {} mteb_test_embed_models(hf_runner, vllm_runner, model_info)
if model_info.architecture == "GteNewModel":
vllm_extra_kwargs["hf_overrides"] = {"architectures": ["GteNewModel"]}
mteb_test_embed_models(hf_runner, vllm_runner, model_info,
vllm_extra_kwargs)
@pytest.mark.parametrize("model_info", MODELS) @pytest.mark.parametrize("model_info", MODELS)
...@@ -91,12 +93,8 @@ def test_embed_models_correctness(hf_runner, vllm_runner, ...@@ -91,12 +93,8 @@ def test_embed_models_correctness(hf_runner, vllm_runner,
check_transformers_version(model_info.name, check_transformers_version(model_info.name,
max_transformers_version="4.53.2") max_transformers_version="4.53.2")
vllm_extra_kwargs: dict[str, Any] = {}
if model_info.architecture == "GteNewModel":
vllm_extra_kwargs["hf_overrides"] = {"architectures": ["GteNewModel"]}
correctness_test_embed_models(hf_runner, vllm_runner, model_info, correctness_test_embed_models(hf_runner, vllm_runner, model_info,
example_prompts, vllm_extra_kwargs) example_prompts)
@pytest.mark.parametrize("model_info", RERANK_MODELS) @pytest.mark.parametrize("model_info", RERANK_MODELS)
......
tests/models/language/pooling/test_multilabel_classification_support.py 0 → 100644
View file @ a99300bd
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
import pytest
import torch
from transformers import AutoModelForSequenceClassification
@pytest.mark.parametrize(
"model",
["Rami/multi-label-class-classification-on-github-issues"],
)
@pytest.mark.parametrize("dtype", ["half"])
def test_classify_models(
hf_runner,
vllm_runner,
example_prompts,
model: str,
dtype: str,
) -> None:
with vllm_runner(model, max_model_len=512, dtype=dtype) as vllm_model:
vllm_outputs = vllm_model.classify(example_prompts)
with hf_runner(model,
dtype=dtype,
auto_cls=AutoModelForSequenceClassification) as hf_model:
hf_outputs = hf_model.classify(example_prompts)
for hf_output, vllm_output in zip(hf_outputs, vllm_outputs):
hf_output = torch.tensor(hf_output)
vllm_output = torch.tensor(vllm_output)
assert torch.allclose(hf_output, vllm_output,
1e-3 if dtype == "float" else 1e-2)
tests/models/language/pooling/test_mxbai_rerank.py
View file @ a99300bd
...@@ -10,12 +10,20 @@ from tests.conftest import HfRunner ...@@ -10,12 +10,20 @@ from tests.conftest import HfRunner
from ...utils import LASTPoolingRerankModelInfo, RerankModelInfo from ...utils import LASTPoolingRerankModelInfo, RerankModelInfo
from .mteb_utils import mteb_test_rerank_models from .mteb_utils import mteb_test_rerank_models
mxbai_rerank_hf_overrides = {
"architectures": ["Qwen2ForSequenceClassification"],
"classifier_from_token": ["0", "1"],
"method": "from_2_way_softmax",
}
RERANK_MODELS = [ RERANK_MODELS = [
LASTPoolingRerankModelInfo("mixedbread-ai/mxbai-rerank-base-v2", LASTPoolingRerankModelInfo("mixedbread-ai/mxbai-rerank-base-v2",
architecture="Qwen2ForSequenceClassification", architecture="Qwen2ForSequenceClassification",
hf_overrides=mxbai_rerank_hf_overrides,
enable_test=True), enable_test=True),
LASTPoolingRerankModelInfo("mixedbread-ai/mxbai-rerank-large-v2", LASTPoolingRerankModelInfo("mixedbread-ai/mxbai-rerank-large-v2",
architecture="Qwen2ForSequenceClassification", architecture="Qwen2ForSequenceClassification",
hf_overrides=mxbai_rerank_hf_overrides,
enable_test=False) enable_test=False)
] ]
...@@ -71,13 +79,4 @@ class MxbaiRerankerHfRunner(HfRunner): ...@@ -71,13 +79,4 @@ class MxbaiRerankerHfRunner(HfRunner):
@pytest.mark.parametrize("model_info", RERANK_MODELS) @pytest.mark.parametrize("model_info", RERANK_MODELS)
def test_rerank_models_mteb(vllm_runner, model_info: RerankModelInfo) -> None: def test_rerank_models_mteb(vllm_runner, model_info: RerankModelInfo) -> None:
vllm_extra_kwargs: dict[str, Any] = {} mteb_test_rerank_models(MxbaiRerankerHfRunner, vllm_runner, model_info)
if model_info.architecture == "Qwen2ForSequenceClassification":
vllm_extra_kwargs["hf_overrides"] = {
"architectures": ["Qwen2ForSequenceClassification"],
"classifier_from_token": ["0", "1"],
"method": "from_2_way_softmax",
}
mteb_test_rerank_models(MxbaiRerankerHfRunner, vllm_runner, model_info,
vllm_extra_kwargs)
tests/models/language/pooling/test_qwen3_reranker.py
View file @ a99300bd
...@@ -11,12 +11,20 @@ from tests.utils import multi_gpu_test ...@@ -11,12 +11,20 @@ from tests.utils import multi_gpu_test
from ...utils import LASTPoolingRerankModelInfo, RerankModelInfo from ...utils import LASTPoolingRerankModelInfo, RerankModelInfo
from .mteb_utils import mteb_test_rerank_models from .mteb_utils import mteb_test_rerank_models
qwen3_reranker_hf_overrides = {
"architectures": ["Qwen3ForSequenceClassification"],
"classifier_from_token": ["no", "yes"],
"is_original_qwen3_reranker": True,
}
RERANK_MODELS = [ RERANK_MODELS = [
LASTPoolingRerankModelInfo("Qwen/Qwen3-Reranker-0.6B", LASTPoolingRerankModelInfo("Qwen/Qwen3-Reranker-0.6B",
architecture="Qwen3ForSequenceClassification", architecture="Qwen3ForSequenceClassification",
hf_overrides=qwen3_reranker_hf_overrides,
enable_test=True), enable_test=True),
LASTPoolingRerankModelInfo("Qwen/Qwen3-Reranker-4B", LASTPoolingRerankModelInfo("Qwen/Qwen3-Reranker-4B",
architecture="Qwen3ForSequenceClassification", architecture="Qwen3ForSequenceClassification",
hf_overrides=qwen3_reranker_hf_overrides,
enable_test=False) enable_test=False)
] ]
...@@ -74,18 +82,7 @@ class Qwen3RerankerHfRunner(HfRunner): ...@@ -74,18 +82,7 @@ class Qwen3RerankerHfRunner(HfRunner):
@pytest.mark.parametrize("model_info", RERANK_MODELS) @pytest.mark.parametrize("model_info", RERANK_MODELS)
def test_rerank_models_mteb(vllm_runner, model_info: RerankModelInfo) -> None: def test_rerank_models_mteb(vllm_runner, model_info: RerankModelInfo) -> None:
assert model_info.architecture == "Qwen3ForSequenceClassification" mteb_test_rerank_models(Qwen3RerankerHfRunner, vllm_runner, model_info)
vllm_extra_kwargs: dict[str, Any] = {
"hf_overrides": {
"architectures": ["Qwen3ForSequenceClassification"],
"classifier_from_token": ["no", "yes"],
"is_original_qwen3_reranker": True,
}
}
mteb_test_rerank_models(Qwen3RerankerHfRunner, vllm_runner, model_info,
vllm_extra_kwargs)
@pytest.mark.parametrize("model_info", RERANK_MODELS) @pytest.mark.parametrize("model_info", RERANK_MODELS)
...@@ -96,16 +93,8 @@ def test_rerank_models_mteb_tp(vllm_runner, ...@@ -96,16 +93,8 @@ def test_rerank_models_mteb_tp(vllm_runner,
assert model_info.architecture == "Qwen3ForSequenceClassification" assert model_info.architecture == "Qwen3ForSequenceClassification"
vllm_extra_kwargs: dict[str, Any] = { vllm_extra_kwargs: dict[str, Any] = {
"hf_overrides": {
"architectures": ["Qwen3ForSequenceClassification"],
"classifier_from_token": ["no", "yes"],
"is_original_qwen3_reranker": True,
},
"tensor_parallel_size": 2, "tensor_parallel_size": 2,
} }
mteb_test_rerank_models(Qwen3RerankerHfRunner, mteb_test_rerank_models(Qwen3RerankerHfRunner, vllm_runner, model_info,
vllm_runner, vllm_extra_kwargs)
model_info,
vllm_extra_kwargs,
atol=1.2e-2)
tests/models/language/pooling/test_reward.py
View file @ a99300bd
...@@ -13,14 +13,6 @@ from ....conftest import HfRunner ...@@ -13,14 +13,6 @@ from ....conftest import HfRunner
from ...utils import check_transformers_version from ...utils import check_transformers_version
@pytest.fixture(autouse=True)
def v1(run_with_both_engines):
# Simple autouse wrapper to run both engines for each test
# This can be promoted up to conftest.py to run for every
# test in a package
pass
@pytest.fixture @pytest.fixture
def math_step_prompts(): def math_step_prompts():
# ruff: noqa: E501 # ruff: noqa: E501
......
tests/models/language/pooling/test_scoring.py
View file @ a99300bd
...@@ -27,15 +27,6 @@ TEXTS_2 = [ ...@@ -27,15 +27,6 @@ TEXTS_2 = [
"The capital of Germany is Berlin.", "The capital of Germany is Berlin.",
] ]
@pytest.fixture(autouse=True)
def v1(run_with_both_engines):
# Simple autouse wrapper to run both engines for each test
# This can be promoted up to conftest.py to run for every
# test in a package
pass
DTYPE = "half" DTYPE = "half"
......
tests/models/language/pooling/test_st_projector.py 0 → 100644
View file @ a99300bd
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
import pytest
from ...utils import CLSPoolingEmbedModelInfo, EmbedModelInfo
from .mteb_utils import mteb_test_embed_models
# ST models with projector (Dense) layers
ST_PROJECTOR_MODELS = [
CLSPoolingEmbedModelInfo(
"TencentBAC/Conan-embedding-v1",
architecture="BertModel",
enable_test=True,
),
]
@pytest.mark.parametrize("model_info", ST_PROJECTOR_MODELS)
def test_embed_models_mteb(hf_runner, vllm_runner,
model_info: EmbedModelInfo) -> None:
mteb_test_embed_models(hf_runner, vllm_runner, model_info)
tests/models/multimodal/generation/test_common.py
View file @ a99300bd
...@@ -192,23 +192,21 @@ VLM_TEST_SETTINGS = { ...@@ -192,23 +192,21 @@ VLM_TEST_SETTINGS = {
}, },
marks=[pytest.mark.core_model], marks=[pytest.mark.core_model],
), ),
# FIXME(Isotr0py): Enable this test after "idefics3-transformers": VLMTestInfo(
# https://github.com/huggingface/transformers/pull/39470 released models=["HuggingFaceTB/SmolVLM-256M-Instruct"],
# "idefics3-transformers": VLMTestInfo( test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE),
# models=["HuggingFaceTB/SmolVLM-256M-Instruct"], prompt_formatter=lambda img_prompt:f"<|begin_of_text|>User:{img_prompt}<end_of_utterance>\nAssistant:", # noqa: E501
# test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE), img_idx_to_prompt=lambda idx: "<image>",
# prompt_formatter=lambda img_prompt:f"<|begin_of_text|>User:{img_prompt}<end_of_utterance>\nAssistant:", # noqa: E501 max_model_len=8192,
# img_idx_to_prompt=lambda idx: "<image>", max_num_seqs=2,
# max_model_len=8192, auto_cls=AutoModelForImageTextToText,
# max_num_seqs=2, hf_output_post_proc=model_utils.idefics3_trunc_hf_output,
# auto_cls=AutoModelForImageTextToText, image_size_factors=[(0.25, 0.5, 1.0)],
# hf_output_post_proc=model_utils.idefics3_trunc_hf_output, vllm_runner_kwargs={
# image_size_factors=[(0.25, 0.5, 1.0)], "model_impl": "transformers",
# vllm_runner_kwargs={ },
# "model_impl": "transformers", marks=[pytest.mark.core_model],
# }, ),
# marks=[pytest.mark.core_model],
# ),
# Pixel values from processor are not 4D or 5D arrays # Pixel values from processor are not 4D or 5D arrays
"qwen2_5_vl-transformers": VLMTestInfo( "qwen2_5_vl-transformers": VLMTestInfo(
models=["Qwen/Qwen2.5-VL-3B-Instruct"], models=["Qwen/Qwen2.5-VL-3B-Instruct"],
...@@ -225,21 +223,6 @@ VLM_TEST_SETTINGS = { ...@@ -225,21 +223,6 @@ VLM_TEST_SETTINGS = {
}, },
marks=[large_gpu_mark(min_gb=32)], marks=[large_gpu_mark(min_gb=32)],
), ),
# Check "auto" with fallback to transformers
"internvl-transformers": VLMTestInfo(
models=["OpenGVLab/InternVL3-1B-hf"],
test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE),
prompt_formatter=lambda img_prompt: f"<|im_start|>User\n{img_prompt}<|im_end|>\n<|im_start|>Assistant\n", # noqa: E501
img_idx_to_prompt=lambda idx: "<IMG_CONTEXT>",
max_model_len=4096,
use_tokenizer_eos=True,
image_size_factors=[(0.25, 0.5, 1.0)],
vllm_runner_kwargs={
"model_impl": "auto",
},
auto_cls=AutoModelForImageTextToText,
marks=[pytest.mark.core_model],
),
#### Extended model tests #### Extended model tests
"aria": VLMTestInfo( "aria": VLMTestInfo(
models=[os.path.join(models_path_prefix, "rhymes-ai/Aria")], models=[os.path.join(models_path_prefix, "rhymes-ai/Aria")],
...@@ -340,10 +323,6 @@ VLM_TEST_SETTINGS = { ...@@ -340,10 +323,6 @@ VLM_TEST_SETTINGS = {
vllm_output_post_proc=model_utils.fuyu_vllm_to_hf_output, vllm_output_post_proc=model_utils.fuyu_vllm_to_hf_output,
num_logprobs=10, num_logprobs=10,
image_size_factors=[(), (0.25,), (0.25, 0.25, 0.25), (0.25, 0.2, 0.15)], image_size_factors=[(), (0.25,), (0.25, 0.25, 0.25), (0.25, 0.2, 0.15)],
# FIXME(Isotr0py): This model is broken in Transformers v4.54.1, we
# should enable this again after the fix is released:
# https://github.com/huggingface/transformers/pull/39915
marks=[pytest.mark.skip("HF model is broken")],
), ),
"gemma3": VLMTestInfo( "gemma3": VLMTestInfo(
models=[os.path.join(models_path_prefix, "google/gemma-3-4b-it")], models=[os.path.join(models_path_prefix, "google/gemma-3-4b-it")],
...@@ -464,6 +443,20 @@ VLM_TEST_SETTINGS = { ...@@ -464,6 +443,20 @@ VLM_TEST_SETTINGS = {
use_tokenizer_eos=True, use_tokenizer_eos=True,
patch_hf_runner=model_utils.internvl_patch_hf_runner, patch_hf_runner=model_utils.internvl_patch_hf_runner,
), ),
"intern_vl-hf": VLMTestInfo(
models=["OpenGVLab/InternVL3-1B-hf"],
test_type=(
VLMTestType.IMAGE,
VLMTestType.MULTI_IMAGE,
VLMTestType.VIDEO,
),
prompt_formatter=lambda img_prompt: f"<|im_start|>User\n{img_prompt}<|im_end|>\n<|im_start|>Assistant\n", # noqa: E501
img_idx_to_prompt=lambda idx: "<IMG_CONTEXT>",
video_idx_to_prompt=lambda idx: "<video>",
max_model_len=8192,
use_tokenizer_eos=True,
auto_cls=AutoModelForImageTextToText,
),
"kimi_vl": VLMTestInfo( "kimi_vl": VLMTestInfo(
models=["moonshotai/Kimi-VL-A3B-Instruct"], models=["moonshotai/Kimi-VL-A3B-Instruct"],
test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE), test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE),
...@@ -624,6 +617,23 @@ VLM_TEST_SETTINGS = { ...@@ -624,6 +617,23 @@ VLM_TEST_SETTINGS = {
hf_model_kwargs={"llm_attn_implementation": "sdpa"}, hf_model_kwargs={"llm_attn_implementation": "sdpa"},
patch_hf_runner=model_utils.ovis_patch_hf_runner, patch_hf_runner=model_utils.ovis_patch_hf_runner,
), ),
"ovis2_5": VLMTestInfo(
models=["AIDC-AI/Ovis2.5-2B"],
test_type=(
VLMTestType.IMAGE,
VLMTestType.MULTI_IMAGE,
VLMTestType.VIDEO
),
prompt_formatter=lambda img_prompt: f"<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n<|im_start|>user\n{img_prompt}<|im_end|>\n<|im_start|>assistant\n", # noqa: E501
img_idx_to_prompt=lambda idx: "<image>\n", # noqa: E501
video_idx_to_prompt=lambda idx: "<video>\n",
max_model_len=4096,
max_num_seqs=2,
dtype="half",
num_logprobs=10,
patch_hf_runner=model_utils.ovis2_5_patch_hf_runner,
hf_model_kwargs={"revision": "refs/pr/5"},
),
"phi3v": VLMTestInfo( "phi3v": VLMTestInfo(
models=["microsoft/Phi-3.5-vision-instruct"], models=["microsoft/Phi-3.5-vision-instruct"],
test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE), test_type=(VLMTestType.IMAGE, VLMTestType.MULTI_IMAGE),
......
tests/models/multimodal/generation/test_mllama.py
View file @ a99300bd
...@@ -6,6 +6,7 @@ from typing import Optional, overload ...@@ -6,6 +6,7 @@ from typing import Optional, overload
import os import os
import pytest import pytest
import torch import torch
from packaging.version import Version
from transformers import AutoConfig, AutoModelForImageTextToText, AutoTokenizer from transformers import AutoConfig, AutoModelForImageTextToText, AutoTokenizer
from transformers import __version__ as TRANSFORMERS_VERSION from transformers import __version__ as TRANSFORMERS_VERSION
...@@ -289,8 +290,8 @@ def clear_cache(): ...@@ -289,8 +290,8 @@ def clear_cache():
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
@pytest.mark.parametrize("attn_backend", LIST_ENC_DEC_SUPPORTED_BACKENDS) @pytest.mark.parametrize("attn_backend", LIST_ENC_DEC_SUPPORTED_BACKENDS)
@pytest.mark.skipif( @pytest.mark.skipif(
TRANSFORMERS_VERSION == "4.55.0", Version(TRANSFORMERS_VERSION) <= Version("4.55.2"),
reason="Transformers v4.55.0 has a regression issue on mllama, " reason="Transformers v4.55 has a regression issue on mllama, "
"see: https://github.com/huggingface/transformers/pull/40083") "see: https://github.com/huggingface/transformers/pull/40083")
def test_models_single_leading_image(hf_runner, vllm_runner, image_assets, def test_models_single_leading_image(hf_runner, vllm_runner, image_assets,
model, sizes, dtype, max_tokens, model, sizes, dtype, max_tokens,
...@@ -321,8 +322,8 @@ def test_models_single_leading_image(hf_runner, vllm_runner, image_assets, ...@@ -321,8 +322,8 @@ def test_models_single_leading_image(hf_runner, vllm_runner, image_assets,
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
@pytest.mark.parametrize("attn_backend", LIST_ENC_DEC_SUPPORTED_BACKENDS) @pytest.mark.parametrize("attn_backend", LIST_ENC_DEC_SUPPORTED_BACKENDS)
@pytest.mark.skipif( @pytest.mark.skipif(
TRANSFORMERS_VERSION == "4.55.0", Version(TRANSFORMERS_VERSION) <= Version("4.55.2"),
reason="Transformers v4.55.0 has a regression issue on mllama, " reason="Transformers v4.55 has a regression issue on mllama, "
"see: https://github.com/huggingface/transformers/pull/40083") "see: https://github.com/huggingface/transformers/pull/40083")
def test_models_multi_leading_images(hf_runner, vllm_runner, image_assets, def test_models_multi_leading_images(hf_runner, vllm_runner, image_assets,
model, dtype, max_tokens, num_logprobs, model, dtype, max_tokens, num_logprobs,
...@@ -374,8 +375,8 @@ def test_models_multi_leading_images(hf_runner, vllm_runner, image_assets, ...@@ -374,8 +375,8 @@ def test_models_multi_leading_images(hf_runner, vllm_runner, image_assets,
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
@pytest.mark.parametrize("attn_backend", LIST_ENC_DEC_SUPPORTED_BACKENDS) @pytest.mark.parametrize("attn_backend", LIST_ENC_DEC_SUPPORTED_BACKENDS)
@pytest.mark.skipif( @pytest.mark.skipif(
TRANSFORMERS_VERSION == "4.55.0", Version(TRANSFORMERS_VERSION) <= Version("4.55.2"),
reason="Transformers v4.55.0 has a regression issue on mllama, " reason="Transformers v4.55 has a regression issue on mllama, "
"see: https://github.com/huggingface/transformers/pull/40083") "see: https://github.com/huggingface/transformers/pull/40083")
def test_models_interleaved_images(hf_runner, vllm_runner, image_assets, model, def test_models_interleaved_images(hf_runner, vllm_runner, image_assets, model,
dtype, max_tokens, num_logprobs, dtype, max_tokens, num_logprobs,
...@@ -418,8 +419,8 @@ def test_models_interleaved_images(hf_runner, vllm_runner, image_assets, model, ...@@ -418,8 +419,8 @@ def test_models_interleaved_images(hf_runner, vllm_runner, image_assets, model,
@pytest.mark.parametrize("max_tokens", [64]) @pytest.mark.parametrize("max_tokens", [64])
@pytest.mark.parametrize("num_logprobs", [5]) @pytest.mark.parametrize("num_logprobs", [5])
@pytest.mark.skipif( @pytest.mark.skipif(
TRANSFORMERS_VERSION == "4.55.0", Version(TRANSFORMERS_VERSION) <= Version("4.55.2"),
reason="Transformers v4.55.0 has a regression issue on mllama, " reason="Transformers v4.55 has a regression issue on mllama, "
"see: https://github.com/huggingface/transformers/pull/40083") "see: https://github.com/huggingface/transformers/pull/40083")
def test_models_distributed( def test_models_distributed(
hf_runner, hf_runner,
......
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