HuiResearch
按照步骤安装出错
drake@DESKTOP-05BHGRQ:~$ flashtts serve --model_path Spark-TTS-0.5B --backend vllm --llm_device cuda --tokenizer_device cuda --detokenizer_device cuda --wav2vec_attn_implementation sdpa --llm_attn_implementation sdpa --torch_dtype "bfloat16" --max_length 32768 --llm_gpu_memory_utilization 0.6 --fix_voice --host 0.0.0.0 --port 8000
[FlashTTS] 2025-08-16 13:33:12 [INFO] [serve:227] >> Starting FlashTTS service...
[FlashTTS] 2025-08-16 13:33:12 [INFO] [serve:228] >> Serving args: Namespace(model_path='Spark-TTS-0.5B', backend='vllm', lang=None, llm_tensorrt_path=None, snac_path=None, llm_device='cuda', tokenizer_device='cuda', detokenizer_device='cuda', wav2vec_attn_implementation='sdpa', llm_attn_implementation='sdpa', max_length=32768, llm_gpu_memory_utilization=0.6, torch_dtype='bfloat16', cache_implementation=None, seed=0, batch_size=1, llm_batch_size=8, wait_timeout=0.01, model_name=None, role_dir=None, db_path='SPEAKERS.bin', api_key=None, fix_voice=True, host='0.0.0.0', port=8000, ssl_keyfile=None, ssl_certfile=None, func=<function serve_command_factory at 0x7b4cb39c11b0>)
INFO: Started server process [87907]
INFO: Waiting for application startup.
[FlashTTS] 2025-08-16 13:33:12 [INFO] [auto_engine:120] >> Initializing AutoEngine(engine=spark) with config: (model_path='Spark-TTS-0.5B', max_length=32768, llm_device='cuda', llm_tensorrt_path=None, backend='vllm', llm_attn_implementation='sdpa', torch_dtype='bfloat16', llm_gpu_memory_utilization=0.6, cache_implementation=None, batch_size=1, llm_batch_size=8, seed=0, tokenizer_device='cuda', detokenizer_device='cuda', wav2vec_attn_implementation='sdpa', wait_timeout=0.01)
/home/drake/.local/lib/python3.10/site-packages/torch/nn/utils/weight_norm.py:143: FutureWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
WeightNorm.apply(module, name, dim)
INFO 08-16 13:33:14 [init.py:235] Automatically detected platform cuda.
ERROR: Traceback (most recent call last):
File "/home/drake/.local/lib/python3.10/site-packages/starlette/routing.py", line 694, in lifespan
async with self.lifespan_context(app) as maybe_state:
File "/usr/lib/python3.10/contextlib.py", line 199, in aenter
return await anext(self.gen)
File "/home/drake/.local/lib/python3.10/site-packages/fastapi/routing.py", line 134, in merged_lifespan
async with original_context(app) as maybe_original_state:
File "/usr/lib/python3.10/contextlib.py", line 199, in aenter
return await anext(self.gen)
File "/home/drake/.local/lib/python3.10/site-packages/fastapi/routing.py", line 134, in merged_lifespan
async with original_context(app) as maybe_original_state:
File "/usr/lib/python3.10/contextlib.py", line 199, in aenter
return await anext(self.gen)
File "/home/drake/.local/lib/python3.10/site-packages/flashtts/commands/serve.py", line 107, in lifespan
engine = AutoEngine(
File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/auto_engine.py", line 122, in init
self._engine = engine_cls(**engine_kwargs)
File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/spark_engine.py", line 285, in init
super().init(
File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/base_engine.py", line 177, in init
self.generator = initialize_llm(
File "/home/drake/.local/lib/python3.10/site-packages/flashtts/llm/init_llm.py", line 39, in initialize_llm
return VllmGenerator(
File "/home/drake/.local/lib/python3.10/site-packages/flashtts/llm/vllm_generator.py", line 32, in init
async_args = AsyncEngineArgs(**engine_kwargs)
TypeError: AsyncEngineArgs.init() got an unexpected keyword argument 'device'
Exception ignored in: <function BaseEngine.del at 0x7b4b34f84ee0> Traceback (most recent call last): File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/base_engine.py", line 249, in del self.shutdown() File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/base_engine.py", line 246, in shutdown self.generator.shutdown() AttributeError: 'AsyncSparkEngine' object has no attribute 'generator' Exception ignored in: <function AutoEngine.del at 0x7b4b34f856c0> Traceback (most recent call last): File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/auto_engine.py", line 144, in del self.shutdown() File "/home/drake/.local/lib/python3.10/site-packages/flashtts/engine/auto_engine.py", line 141, in shutdown self._engine.shutdown() AttributeError: 'AutoEngine' object has no attribute '_engine' ERROR: Application startup failed. Exiting.
用的wsl，重装了好多次好像都是这个问题。请大佬指点下怎么回事，感谢！
是vllm不接受--llm_device这个参数，去修改flashtts源文件vllm_generator.py的第32行async_args = AsyncEngineArgs(**engine_kwargs)上面加一行代码engine_kwargs.pop("device", None)，就可以解决了
是vllm不接受--llm_device这个参数，去修改flashtts源文件vllm_generator.py的第32行async_args = AsyncEngineArgs(**engine_kwargs)上面加一行代码engine_kwargs.pop("device", None)，就可以解决了
@HuiResearch 作者你好，这个问题是否可以跟进一下，我的vllm版本为0.10.0，torch版本2.7.1