本地部署DeepSeek-R1-Distill-Llama-70B,使用Atlas 300V或300I(3010)
收藏回复举报
本地部署DeepSeek-R1-Distill-Llama-70B,使用Atlas 300V或300I(3010)
t('forum.solved') 已解决
新人帖
发表于2025-02-21 10:06:01
0 查看

我参考您文章和官方文档,在atlas 300I上面部署deepseek r1 70b,但是运行mindieservice_daemon的时候报错:

[ERROR] model.py:39 - [Model] >>> Exception:Unsupported soc version: Ascend310。

然后前面有一个未初始化的日志

2025-02-21 08:11:05,610 [INFO] config.py:35 - do not init dmi config

镜像使用的是swr.cn-south-1.myhuaweicloud.com/ascendhub/mindie:1.0.0-300I-Duo-py311-openeuler24.03-lts

运行的详细日志:

...
2025-02-21 10:04:17,615 [INFO] config.py:30 - model_config {'backend_bin_path': '/usr/local/Ascend/mindie/1.0.0/mindie-llm/bin/', 'backend_log_file': '/usr/local/Ascend/mindie/1.0.0/mindie-service/logs/mindservice.log', 'backend_modelInstance_id': '0', 'backend_type': 'atb', 'block_size': '128', 'cpu_mem': '5', 'deploy_type': 'INTER_PROCESS', 'dp': '0', 'ep': '0', 'executor_type': 'LLM_EXECUTOR_PYTHON', 'globalRankIds': '', 'globalWorldSize': '0', 'interNodeKmcKsfMaster': 'tools/pmt/master/ksfa', 'interNodeKmcKsfStandby': 'tools/pmt/standby/ksfb', 'interNodeTLSEnabled': '1', 'interNodeTlsCaFiles': 'ca.pem,', 'interNodeTlsCaPath': 'security/grpc/ca/', 'interNodeTlsCert': 'security/grpc/certs/server.pem', 'interNodeTlsCrlFiles': 'server_crl.pem,', 'interNodeTlsCrlPath': 'security/grpc/certs/', 'interNodeTlsPk': 'security/grpc/keys/server.key.pem', 'interNodeTlsPkPwd': 'security/grpc/pass/mindie_server_key_pwd.txt', 'isMaster': '0', 'localIP': '', 'local_rank': '0', 'log_error': '1', 'log_file_num': '20', 'log_file_size': '20', 'log_info': '1', 'log_verbose': '0', 'log_warning': '1', 'masterIP': '', 'max_input_len': '2048', 'max_iter_times': '512', 'max_prefill_tokens': '8192', 'max_seq_len': '2560', 'model_cut_policy': 'standard', 'model_id': '/data/models/DeepSeek-R1-Distill-Llama-70B', 'model_instance_number': '1', 'model_instance_type': 'Standard', 'model_name': 'llama_65b', 'multiNodesInferEnabled': '0', 'multiNodesInferPort': '1120', 'npu_device_id': '0', 'npu_device_ids': '0,1,2,3', 'npu_mem': '-1', 'pp': '0', 'rank': '0', 'slaveIPs': '', 'speculation_gamma': '0', 'tp': '0', 'trust_remote_code': '0', 'world_size': '4'}
...
2025-02-21 10:04:31,802 [ERROR] model.py:39 - [Model]   >>> Exception:Unsupported soc version: Ascend310
Traceback (most recent call last):
  File "/usr/local/lib/python3.11/site-packages/model_wrapper/model.py", line 37, in initialize
    return self.python_model.initialize(config)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/model_wrapper/standard_model.py", line 146, in initialize
    self.generator = Generator(
                     ^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/generator.py", line 79, in __init__
    self.generator_backend = get_generator_backend(model_config)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/__init__.py", line 26, in get_generator_backend
    return generator_cls(model_config)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/generator_torch.py", line 96, in __init__
    super().__init__(model_config)
  File "/usr/local/lib/python3.11/site-packages/mindie_llm/text_generator/adapter/generator_backend.py", line 107, in __init__
    self.model_wrapper = get_model_wrapper(model_config, backend_type)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/mindie_llm/modeling/model_wrapper/__init__.py", line 15, in get_model_wrapper
    return wrapper_cls(**model_config)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/mindie_llm/modeling/model_wrapper/atb/atb_model_wrapper.py", line 24, in __init__
    self.model_runner = ModelRunner(
                        ^^^^^^^^^^^^
  File "/usr/local/Ascend/atb-models/atb_llm/runner/model_runner.py", line 132, in __init__
    self.process_group, self.device = initialize_distributed(self.rank, self.npu_id, world_size)
                                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/Ascend/atb-models/atb_llm/utils/dist.py", line 76, in initialize_distributed
    torch.npu.set_device(device)
  File "/usr/local/lib64/python3.11/site-packages/torch_npu/npu/utils.py", line 58, in set_device
    torch_npu._C._npu_setDevice(device_id)
RuntimeError: Unsupported soc version: Ascend310
...

我要发帖子