vllm 0.10.0 transformer 4.54.0 throw ValueError after update model file with this repository
[1;36m(VllmWorker rank=5 pid=6878)[0;0m INFO 07-26 19:05:47 [default_loader.py:262] Loading weights took 115.32 seconds
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] WorkerProc failed to start.
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] Traceback (most recent call last):
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] File "/usr/local/lib64/python3.9/site-packages/vllm/v1/executor/multiproc_executor.py", line 485, in worker_main
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] worker = WorkerProc(*args, **kwargs)
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] File "/usr/local/lib64/python3.9/site-packages/vllm/v1/executor/multiproc_executor.py", line 382, in init
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] self.worker.load_model()
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] File "/usr/local/lib64/python3.9/site-packages/vllm/v1/worker/gpu_worker.py", line 201, in load_model
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] self.model_runner.load_model(eep_scale_up=eep_scale_up)
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] File "/usr/local/lib64/python3.9/site-packages/vllm/v1/worker/gpu_model_runner.py", line 1876, in load_model
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] self.model = model_loader.load_model(
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] File "/usr/local/lib64/python3.9/site-packages/vllm/model_executor/model_loader/base_loader.py", line 49, in load_model
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] self.load_weights(model, model_config)
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] File "/usr/local/lib64/python3.9/site-packages/vllm/model_executor/model_loader/default_loader.py", line 271, in load_weights
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] raise ValueError("Following weights were not initialized from "
[1;36m(VllmWorker rank=5 pid=6878)[0;0m ERROR 07-26 19:05:47 [multiproc_executor.py:511] ValueError: Following weights were not initialized from checkpoint: {'model.layers.47.mlp.shared_experts.gate_up_proj.weight', 'model.layers.3.mlp.shared_experts.gate_up_proj.weight', 'model.layers.45.mlp.shared_experts.down_proj.weight', 'model.layers.28.mlp.shared_experts.down_proj.weight', 'model.layers.42.mlp.shared_experts.down_proj.weight', 'model.layers.30.mlp.shared_experts.down_proj.weight', 'model.layers.41.mlp.shared_experts.down_proj.weight', 'model.layers.33.mlp.shared_experts.down_proj.weight', 'model.layers.24.mlp.shared_experts.gate_up_proj.weight', 'model.layers.19.mlp.shared_experts.gate_up_proj.weight', 'model.layers.38.mlp.shared_experts.gate_up_proj.weight', 'model.layers.27.mlp.shared_experts.gate_up_proj.weight', 'model.layers.5.mlp.shared_experts.down_proj.weight', 'model.layers.26.mlp.shared_experts.down_proj.weight', 'model.layers.44.mlp.shared_experts.gate_up_proj.weight', 'model.layers.39.mlp.shared_experts.down_proj.weight', 'model.layers.15.mlp.shared_experts.down_proj.weight', 'model.layers.50.mlp.shared_experts.gate_up_proj.weight', 'model.layers.11.mlp.shared_experts.gate_up_proj.weight', 'model.layers.35.mlp.shared_experts.down_proj.weight', 'model.layers.21.mlp.shared_experts.gate_up_proj.weight', 'model.layers.17.mlp.shared_experts.gate_up_proj.weight', 'model.layers.40.mlp.shared_experts.down_proj.weight', 'model.layers.15.mlp.shared_experts.gate_up_proj.weight', 'model.layers.23.mlp.shared_experts.gate_up_proj.weight', 'model.layers.20.mlp.shared_experts.gate_up_proj.weight', 'model.layers.48.mlp.shared_experts.gate_up_proj.weight', 'model.layers.40.mlp.shared_experts.gate_up_proj.weight', 'model.layers.51.mlp.shared_experts.gate_up_proj.weight', 'model.layers.37.mlp.shared_experts.gate_up_proj.weight', 'model.layers.12.mlp.shared_experts.gate_up_proj.weight', 'model.layers.46.mlp.shared_experts.down_proj.weight', 'model.layers.22.mlp.shared_experts.gate_up_proj.weight', 'model.layers.29.mlp.shared_experts.gate_up_proj.weight', 'model.layers.34.mlp.shared_experts.gate_up_proj.weight', 'model.layers.22.mlp.shared_experts.down_proj.weight', 'model.layers.26.mlp.shared_experts.gate_up_proj.weight', 'model.layers.36.mlp.shared_experts.down_proj.weight', 'model.layers.50.mlp.shared_experts.down_proj.weight', 'model.layers.13.mlp.shared_experts.down_proj.weight', 'model.layers.9.mlp.shared_experts.gate_up_proj.weight', 'model.layers.47.mlp.shared_experts.down_proj.weight', 'model.layers.52.mlp.shared_experts.gate_up_proj.weight', 'model.layers.43.mlp.shared_experts.gate_up_proj.weight', 'model.layers.53.mlp.shared_experts.gate_up_proj.weight', 'model.layers.52.mlp.shared_experts.down_proj.weight', 'model.layers.45.mlp.shared_experts.gate_up_proj.weight', 'model.layers.33.mlp.shared_experts.gate_up_proj.weight', 'model.layers.8.mlp.shared_experts.gate_up_proj.weight', 'model.layers.12.mlp.shared_experts.down_proj.weight', 'model.layers.7.mlp.shared_experts.down_proj.weight', 'model.layers.13.mlp.shared_experts.gate_up_proj.weight', 'model.layers.18.mlp.shared_experts.down_proj.weight', 'model.layers.4.mlp.shared_experts.gate_up_proj.weight', 'model.layers.42.mlp.shared_experts.gate_up_proj.weight', 'model.layers.5.mlp.shared_experts.gate_up_proj.weight', 'model.layers.14.mlp.shared_experts.gate_up_proj.weight', 'model.layers.10.mlp.shared_experts.down_proj.weight', 'model.layers.24.mlp.shared_experts.down_proj.weight', 'model.layers.48.mlp.shared_experts.down_proj.weight', 'model.layers.19.mlp.shared_experts.down_proj.weight', 'model.layers.20.mlp.shared_experts.down_proj.weight', 'model.layers.38.mlp.shared_experts.down_proj.weight', 'model.layers.36.mlp.shared_experts.gate_up_proj.weight', 'model.layers.39.mlp.shared_experts.gate_up_proj.weight', 'model.layers.3.mlp.shared_experts.down_proj.weight', 'model.layers.34.mlp.shared_experts.down_proj.weight', 'model.layers.44.mlp.shared_experts.down_proj.weight', 'model.layers.21.mlp.shared_experts.down_proj.weight', 'model.layers.11.mlp.shared_experts.down_proj.weight', 'model.layers.18.mlp.shared_experts.gate_up_proj.weight', 'model.layers.7.mlp.shared_experts.gate_up_proj.weight', 'model.layers.25.mlp.shared_experts.gate_up_proj.weight', 'model.layers.31.mlp.shared_experts.gate_up_proj.weight', 'model.layers.30.mlp.shared_experts.gate_up_proj.weight', 'model.layers.8.mlp.shared_experts.down_proj.weight', 'model.layers.16.mlp.shared_experts.down_proj.weight', 'model.layers.28.mlp.shared_experts.gate_up_proj.weight', 'model.layers.49.mlp.shared_experts.gate_up_proj.weight', 'model.layers.10.mlp.shared_experts.gate_up_proj.weight', 'model.layers.4.mlp.shared_experts.down_proj.weight', 'model.layers.17.mlp.shared_experts.down_proj.weight', 'model.layers.32.mlp.shared_experts.gate_up_proj.weight', 'model.layers.14.mlp.shared_experts.down_proj.weight', 'model.layers.53.mlp.shared_experts.down_proj.weight', 'model.layers.37.mlp.shared_experts.down_proj.weight', 'model.layers.35.mlp.shared_experts.gate_up_proj.weight', 'model.layers.6.mlp.shared_experts.gate_up_proj.weight', 'model.layers.6.mlp.shared_experts.down_proj.weight', 'model.layers.23.mlp.shared_experts.down_proj.weight', 'model.layers.32.mlp.shared_experts.down_proj.weight', 'model.layers.46.mlp.shared_experts.gate_up_proj.weight', 'model.layers.25.mlp.shared_experts.down_proj.weight', 'model.layers.49.mlp.shared_experts.down_proj.weight', 'model.layers.27.mlp.shared_experts.down_proj.weight', 'model.layers.41.mlp.shared_experts.gate_up_proj.weight', 'model.layers.51.mlp.shared_experts.down_proj.weight', 'model.layers.29.mlp.shared_experts.down_proj.weight', 'model.layers.16.mlp.shared_experts.gate_up_proj.weight', 'model.layers.9.mlp.shared_experts.down_proj.weight', 'model.layers.31.mlp.shared_experts.down_proj.weight', 'model.layers.43.mlp.shared_experts.down_proj.weight'}
The problem seems to be that the newly released vlmm 0.10 is not compatible?
However, vlmm 0.9.2 is currently not compatible with the latest version of transformer either. its throw
Traceback (most recent call last):
File "/usr/local/bin/vllm", line 5, in
from vllm.entrypoints.cli.main import main
File "/usr/local/lib64/python3.9/site-packages/vllm/entrypoints/cli/init.py", line 3, in
from vllm.entrypoints.cli.benchmark.latency import BenchmarkLatencySubcommand
File "/usr/local/lib64/python3.9/site-packages/vllm/entrypoints/cli/benchmark/latency.py", line 5, in
from vllm.benchmarks.latency import add_cli_args, main
File "/usr/local/lib64/python3.9/site-packages/vllm/benchmarks/latency.py", line 16, in
from vllm import LLM, SamplingParams
File "", line 1055, in _handle_fromlist
File "/usr/local/lib64/python3.9/site-packages/vllm/init.py", line 64, in getattr
module = import_module(module_name, package)
File "/usr/lib64/python3.9/importlib/init.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "/usr/local/lib64/python3.9/site-packages/vllm/entrypoints/llm.py", line 20, in
from vllm.config import (CompilationConfig, ModelDType, TokenizerMode,
File "/usr/local/lib64/python3.9/site-packages/vllm/config.py", line 37, in
from vllm.transformers_utils.config import (
File "/usr/local/lib64/python3.9/site-packages/vllm/transformers_utils/config.py", line 33, in
from vllm.transformers_utils.configs import (ChatGLMConfig, Cohere2Config,
File "/usr/local/lib64/python3.9/site-packages/vllm/transformers_utils/configs/init.py", line 26, in
from vllm.transformers_utils.configs.ovis import OvisConfig
File "/usr/local/lib64/python3.9/site-packages/vllm/transformers_utils/configs/ovis.py", line 76, in
AutoConfig.register("aimv2", AIMv2Config)
File "/usr/local/lib/python3.9/site-packages/transformers/models/auto/configuration_auto.py", line 1306, in register
CONFIG_MAPPING.register(model_type, config, exist_ok=exist_ok)
File "/usr/local/lib/python3.9/site-packages/transformers/models/auto/configuration_auto.py", line 993, in register
raise ValueError(f"'{key}' is already used by a Transformers config, pick another name.")
ValueError: 'aimv2' is already used by a Transformers config, pick another name.