分布式训练相关

更新时间：2025-11-26

查询训练任务列表

使用以下代码可以查询训练任务列表。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量（可选，推荐使用）
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息和服务端点
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'  # 百舸AIHC服务地址
AK = os.getenv('AK') or 'your-access-key-id'                # 您的Access Key ID
SK = os.getenv('SK') or 'your-secret-access-key'            # 您的Secret Access Key

# 创建BCE客户端配置
config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)

# 创建 aihc client
aihc_client = AihcClient(config)

# 查询训练任务列表
try:
    __logger.info('--------------------------------DescribeJobs start--------------------------------')
    resource_pool_id = "cce-xxx"  # 替换为实际的资源池ID
    keyword = ""
    response = aihc_client.job.DescribeJobs(resourcePoolId=resource_pool_id, keyword=keyword)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务列表

查询训练任务详情

使用以下代码可以查询训练任务详情。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 查询训练任务详情
try:
    __logger.info('--------------------------------DescribeJob start--------------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    queue_id = "default"         # 替换为实际的队列ID或名称
    job_id = "job-xxx"           # 替换为实际的任务ID
    need_detail = True
    response = aihc_client.job.DescribeJob(resourcePoolId=resource_pool_id, queueID=queue_id, jobId=job_id,
                                            needDetail=need_detail)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务详情

删除训练任务

使用以下代码可以删除训练任务。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 删除训练任务
try:
    __logger.info('----------------------------------DeleteJob start-----------------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    response = aihc_client.job.DeleteJob(resourcePoolId=resource_pool_id, jobId=job_id)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为删除训练任务

更新训练任务

使用以下代码可以更新训练任务（如修改优先级）。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 更新训练任务
try:
    __logger.info('---------------------------------ModifyJob start----------------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    priority = "high"            # 新的优先级
    response = aihc_client.job.ModifyJob(resourcePoolId=resource_pool_id, jobId=job_id, priority=priority)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为更新训练任务

查询训练任务事件

使用以下代码可以查询训练任务事件。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 查询训练任务事件
try:
    __logger.info('--------------------------DescribeJobEvents start-----------------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    start_time = "1758532230"    # 起始时间戳
    end_time = "1758618650"      # 结束时间戳
    response = aihc_client.job.DescribeJobEvents(resourcePoolId=resource_pool_id, jobId=job_id,
                                                    startTime=start_time, endTime=end_time)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务事件

查询训练任务日志

使用以下代码可以查询训练任务日志。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 查询训练任务日志
try:
    __logger.info('--------------------------DescribeJobLogs start---------------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    pod_name = "xxx-test-copy2-master-0" # 替换为实际的pod名称
    keywords = "xxx"             # 搜索关键字
    response = aihc_client.job.DescribeJobLogs(resourcePoolId=resource_pool_id,
                                                jobId=job_id,
                                                keywords=keywords,
                                                podName=pod_name)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务日志

查询训练任务Pod事件

使用以下代码可以查询训练任务Pod事件。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 查询训练任务Pod事件
try:
    __logger.info('---------------DescribeJobPodEvents start---------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    pod_name = "job-xxx-master-0" # 替换为实际的pod名称
    job_id = "job-xxx"           # 替换为实际的任务ID
    response = aihc_client.job.DescribeJobPodEvents(resourcePoolId=resource_pool_id,
                                                    jobId=job_id,
                                                    podName=pod_name)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务Pod事件

停止训练任务

使用以下代码可以停止训练任务。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 停止训练任务
try:
    __logger.info('stop job')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    response = aihc_client.job.StopJob(resourcePoolId=resource_pool_id, jobId=job_id)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为停止训练任务

查询训练任务所在节点列表

使用以下代码可以查询训练任务所在节点列表。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 查询训练任务所在节点列表
try:
    __logger.info('-------------------DescribeJobNodes start--------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    response = aihc_client.job.DescribeJobNodes(resourcePoolId=resource_pool_id, jobId=job_id)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务所在节点列表

获取训练任务WebTerminal地址

使用以下代码可以获取训练任务WebTerminal地址。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 获取训练任务WebTerminal地址
try:
    __logger.info('-------------------DescribeJobWebterminal start--------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    podName = "xxx-test-bb1-rerun1-master-0" # 替换为实际的pod名称
    handshake_timeout_second = "30"
    ping_timeout_second = "900"
    response = aihc_client.job.DescribeJobWebterminal(
        resourcePoolId=resource_pool_id,
        jobId=job_id,
        podName=podName,
        pingTimeoutSecond=ping_timeout_second,
        handshakeTimeoutSecond=handshake_timeout_second,
    )
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为获取训练任务WebTerminal地址

创建训练任务

使用以下代码可以创建训练任务。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.services.aihc.modules.job.job_model import Datasource, Env, JobSpec
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 创建训练任务
try:
    __logger.info('---------------------------------CreateJob start------------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    queue_id = "default"         # 替换为实际的队列ID或名称
    name = "python-sdk-test-xxx" # 任务名称
    command = "sleep 5m"         # 启动命令
    
    # 配置环境变量
    envs = [Env(name="NCCL_DEBUG", value="DEBUG"), Env(name="NCCL_IB_DISABLE", value="0")]
    
    # 配置数据源
    data_source = Datasource(type="pfs", name="pfs-pxE6jz", mountPath="/mnt/cluster")
    
    # 配置任务规格
    job_spec = JobSpec(
        image="registry.baidubce.com/aihc-aiak/aiak-megatron:ubuntu20.04-cu11.8-torch1.14.0-py38_v1.2.7.12_release",
        replicas=1,
        resources=[],
        envs=envs,
        enableRDMA=False
    )
    
    job_type = "PyTorchJob"
    labels = []
    priority = "normal"
    dataSources = [
        data_source,
    ]
    enable_bccl = False
    fault_tolerance = False
    fault_tolerance_args = {}
    tensorboard_config = {}
    retention_period = "5m"
    
    response = aihc_client.job.CreateJob(
        resourcePoolId=resource_pool_id,
        queueID=queue_id,
        name=name,
        command=command,
        jobSpec=job_spec,
        jobType=job_type,
        labels=labels,
        priority=priority,
        dataSources=dataSources,
        enableBccl=enable_bccl,
        faultTolerance=fault_tolerance,
        faultToleranceArgs=fault_tolerance_args,
        tensorboardConfig=tensorboard_config,
        retentionPeriod=retention_period,
    )
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为创建训练任务

查询训练任务监控

使用以下代码可以查询训练任务监控数据。

                Python
                
            

                import logging
import os

import dotenv

from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials
from baidubce.services.aihc.aihc_client import AihcClient
from baidubce.exception import BceHttpClientError, BceServerError

# 加载环境变量
dotenv.load_dotenv()

logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s', force=True)
logging.getLogger().setLevel(logging.INFO)
logging.getLogger("baidubce").setLevel(logging.INFO)
__logger = logging.getLogger(__name__)
__logger.setLevel(logging.INFO)

# 配置认证信息
HOST = os.getenv('HOST') or 'https://aihc.bj.baidubce.com'
AK = os.getenv('AK') or 'your-access-key-id'
SK = os.getenv('SK') or 'your-secret-access-key'

config = BceClientConfiguration(credentials=BceCredentials(AK, SK), endpoint=HOST)
aihc_client = AihcClient(config)

# 查询训练任务监控
try:
    __logger.info('--------------------DescribeJobMetrics start------------------------')
    resource_pool_id = "cce-xxx" # 替换为实际的资源池ID
    job_id = "job-xxx"           # 替换为实际的任务ID
    start_time = "1758359060"    # 起始时间戳
    end_time = "1758445563"      # 结束时间戳
    time_step = "5m"             # 监控数据间隔
    metric_type = "GpuUsage"     # 监控指标类型
    rate_interval = "5m"         # 指标变化周期频率
    response = aihc_client.job.DescribeJobMetrics(resourcePoolId=resource_pool_id,
                                                    jobId=job_id,
                                                    metricType=metric_type,
                                                    startTime=start_time,
                                                    endTime=end_time,
                                                    timeStep=time_step,
                                                    rateInterval=rate_interval)
    print(response)
except BceHttpClientError as e:
    if isinstance(e.last_error, BceServerError):
        __logger.error('send request failed. Response %s, code: %s, msg: %s'
                        % (e.last_error.status_code, e.last_error.code, str(e.last_error)))
    else:
        __logger.error('send request failed. Unknown exception: %s' % e)
            

注意: 根据接口文档去填写具体的访问参数，接口链接为查询训练任务监控

资源池相关

在线服务部署相关

百度智能云

百度百舸 · AI计算平台