Add Task.init auto_connect_streams controlling stdout/stderr/logging capture. Issue #181

trains/backend_interface/logger.py

@@ -1,10 +1,10 @@
 import logging
 import sys
 import threading
+from time import time
 
-from ..backend_interface.task.development.worker import DevWorker
-from ..backend_interface.task.log import TaskHandler
-from ..config import running_remotely
+from ..binding.frameworks import _patched_call  # noqa
+from ..config import running_remotely, config
 
 
 class StdStreamPatch(object):
@@ -14,51 +14,62 @@ class StdStreamPatch(object):
     _stderr_original_write = None
 
     @staticmethod
-    def patch_std_streams(logger):
-        if DevWorker.report_stdout and not PrintPatchLogger.patched and not running_remotely():
-            StdStreamPatch._stdout_proxy = PrintPatchLogger(sys.stdout, logger, level=logging.INFO)
-            StdStreamPatch._stderr_proxy = PrintPatchLogger(sys.stderr, logger, level=logging.ERROR)
-            logger._task_handler = TaskHandler(task=logger._task, capacity=100)
+    def patch_std_streams(a_logger, connect_stdout=True, connect_stderr=True):
+        if (connect_stdout or connect_stderr) and not PrintPatchLogger.patched and not running_remotely():
+            StdStreamPatch._stdout_proxy = PrintPatchLogger(sys.stdout, a_logger, level=logging.INFO) \
+                if connect_stdout else None
+            StdStreamPatch._stderr_proxy = PrintPatchLogger(sys.stderr, a_logger, level=logging.ERROR) \
+                if connect_stderr else None
+
+            if StdStreamPatch._stdout_proxy:
                 # noinspection PyBroadException
                 try:
                     if StdStreamPatch._stdout_original_write is None:
                         StdStreamPatch._stdout_original_write = sys.stdout.write
-                if StdStreamPatch._stderr_original_write is None:
-                    StdStreamPatch._stderr_original_write = sys.stderr.write
+
                     # this will only work in python 3, guard it with try/catch
                     if not hasattr(sys.stdout, '_original_write'):
                         sys.stdout._original_write = sys.stdout.write
                     sys.stdout.write = StdStreamPatch._stdout__patched__write__
+                except Exception:
+                    pass
+                sys.stdout = StdStreamPatch._stdout_proxy
+                # noinspection PyBroadException
+                try:
+                    sys.__stdout__ = sys.stdout
+                except Exception:
+                    pass
+
+            if StdStreamPatch._stderr_proxy:
+                # noinspection PyBroadException
+                try:
+                    if StdStreamPatch._stderr_original_write is None:
+                        StdStreamPatch._stderr_original_write = sys.stderr.write
                     if not hasattr(sys.stderr, '_original_write'):
                         sys.stderr._original_write = sys.stderr.write
                     sys.stderr.write = StdStreamPatch._stderr__patched__write__
                 except Exception:
                     pass
-            sys.stdout = StdStreamPatch._stdout_proxy
                 sys.stderr = StdStreamPatch._stderr_proxy
+
                 # patch the base streams of sys (this way colorama will keep its ANSI colors)
                 # noinspection PyBroadException
                 try:
                     sys.__stderr__ = sys.stderr
                 except Exception:
                     pass
-            # noinspection PyBroadException
-            try:
-                sys.__stdout__ = sys.stdout
-            except Exception:
-                pass
 
             # now check if we have loguru and make it re-register the handlers
             # because it stores internally the stream.write function, which we cant patch
             # noinspection PyBroadException
             try:
-                from loguru import logger
+                from loguru import logger  # noqa
                 register_stderr = None
                 register_stdout = None
-                for k, v in logger._handlers.items():
-                    if v._name == '<stderr>':
+                for k, v in logger._handlers.items():  # noqa
+                    if connect_stderr and v._name == '<stderr>':  # noqa
                         register_stderr = k
-                    elif v._name == '<stdout>':
+                    elif connect_stdout and v._name == '<stdout>':  # noqa
                         register_stderr = k
                 if register_stderr is not None:
                     logger.remove(register_stderr)
@@ -69,12 +80,30 @@ class StdStreamPatch(object):
             except Exception:
                 pass
 
-        elif DevWorker.report_stdout and not running_remotely():
-            logger._task_handler = TaskHandler(task=logger._task, capacity=100)
-            if StdStreamPatch._stdout_proxy:
-                StdStreamPatch._stdout_proxy.connect(logger)
-            if StdStreamPatch._stderr_proxy:
-                StdStreamPatch._stderr_proxy.connect(logger)
+        elif (connect_stdout or connect_stderr) and not running_remotely():
+            if StdStreamPatch._stdout_proxy and connect_stdout:
+                StdStreamPatch._stdout_proxy.connect(a_logger)
+            if StdStreamPatch._stderr_proxy and connect_stderr:
+                StdStreamPatch._stderr_proxy.connect(a_logger)
+
+    @staticmethod
+    def patch_logging_formatter(a_logger, logging_handler=None):
+        if not logging_handler:
+            import logging
+            logging_handler = logging.Handler
+        logging_handler.format = _patched_call(logging_handler.format, HandlerFormat(a_logger))
+
+    @staticmethod
+    def remove_patch_logging_formatter(logging_handler=None):
+        if not logging_handler:
+            import logging
+            logging_handler = logging.Handler
+        # remove the function, Hack calling patched logging.Handler.format() returns the original function
+        # noinspection PyBroadException
+        try:
+            logging_handler.format = logging_handler.format()  # noqa
+        except Exception:
+            pass
 
     @staticmethod
     def remove_std_logger(logger=None):
@@ -105,13 +134,33 @@ class StdStreamPatch(object):
     def _stdout__patched__write__(*args, **kwargs):
         if StdStreamPatch._stdout_proxy:
             return StdStreamPatch._stdout_proxy.write(*args, **kwargs)
-        return sys.stdout._original_write(*args, **kwargs)
+        return sys.stdout._original_write(*args, **kwargs)  # noqa
 
     @staticmethod
     def _stderr__patched__write__(*args, **kwargs):
         if StdStreamPatch._stderr_proxy:
             return StdStreamPatch._stderr_proxy.write(*args, **kwargs)
-        return sys.stderr._original_write(*args, **kwargs)
+        return sys.stderr._original_write(*args, **kwargs)  # noqa
+
+
+class HandlerFormat(object):
+    def __init__(self, logger):
+        self._logger = logger
+
+    def __call__(self, original_format_func, *args):
+        # hack get back original function, so we can remove it
+        if all(a is None for a in args):
+            return original_format_func
+        if len(args) == 1:
+            record = args[0]
+            msg = original_format_func(record)
+        else:
+            handler = args[0]
+            record = args[1]
+            msg = original_format_func(handler, record)
+
+        self._logger.report_text(msg=msg, level=record.levelno, print_console=False)
+        return msg
 
 
 class PrintPatchLogger(object):
@@ -122,6 +171,7 @@ class PrintPatchLogger(object):
     patched = False
     lock = threading.Lock()
     recursion_protect_lock = threading.RLock()
+    lf_flush_period = config.get("development.worker.console_lf_flush_period", 0)
 
     def __init__(self, stream, logger=None, level=logging.INFO):
         PrintPatchLogger.patched = True
@@ -129,23 +179,35 @@ class PrintPatchLogger(object):
         self._log = logger
         self._log_level = level
         self._cur_line = ''
+        self._force_lf_flush = False
 
     def write(self, message):
         # make sure that we do not end up in infinite loop (i.e. log.console ends up calling us)
-        if self._log and not PrintPatchLogger.recursion_protect_lock._is_owned():
+        if self._log and not PrintPatchLogger.recursion_protect_lock._is_owned():  # noqa
             try:
                 self.lock.acquire()
                 with PrintPatchLogger.recursion_protect_lock:
                     if hasattr(self._terminal, '_original_write'):
-                        self._terminal._original_write(message)
+                        self._terminal._original_write(message)  # noqa
                     else:
                         self._terminal.write(message)
 
                 do_flush = '\n' in message
                 do_cr = '\r' in message
                 self._cur_line += message
-                if (not do_flush and not do_cr) or not message:
+
+                if not do_flush and do_cr and PrintPatchLogger.lf_flush_period and self._force_lf_flush:
+                    self._cur_line += '\n'
+                    do_flush = True
+
+                self._force_lf_flush = False
+
+                if (not do_flush and (PrintPatchLogger.lf_flush_period or not do_cr)) or not message:
                     return
+
+                if PrintPatchLogger.lf_flush_period and self._cur_line:
+                    self._cur_line = '\n'.join(line.split('\r')[-1] for line in self._cur_line.split('\n'))
+
                 last_lf = self._cur_line.rindex('\n' if do_flush else '\r')
                 next_line = self._cur_line[last_lf + 1:]
                 cur_line = self._cur_line[:last_lf + 1].rstrip()
@@ -158,13 +220,14 @@ class PrintPatchLogger(object):
                     # noinspection PyBroadException
                     try:
                         if self._log:
+                            # noinspection PyProtectedMember
                             self._log._console(cur_line, level=self._log_level, omit_console=True)
                     except Exception:
                         # what can we do, nothing
                         pass
         else:
             if hasattr(self._terminal, '_original_write'):
-                self._terminal._original_write(message)
+                self._terminal._original_write(message)  # noqa
             else:
                 self._terminal.write(message)
 
@@ -177,13 +240,16 @@ class PrintPatchLogger(object):
         if not logger or self._log == logger:
             self.connect(None)
 
+    def force_lf_flush(self):
+        self._force_lf_flush = True
+
     def __getattr__(self, attr):
-        if attr in ['_log', '_terminal', '_log_level', '_cur_line']:
+        if attr in ['_log', '_terminal', '_log_level', '_cur_line', '_cr_overwrite', '_force_lf_flush']:
             return self.__dict__.get(attr)
         return getattr(self._terminal, attr)
 
     def __setattr__(self, key, value):
-        if key in ['_log', '_terminal', '_log_level', '_cur_line']:
+        if key in ['_log', '_terminal', '_log_level', '_cur_line', '_cr_overwrite', '_force_lf_flush']:
             self.__dict__[key] = value
         else:
             return setattr(self._terminal, key, value)
@@ -197,6 +263,11 @@ class LogFlusher(threading.Thread):
         self._period = period
         self._logger = logger
         self._exit_event = threading.Event()
+        self._lf_last_flush = 0
+        try:
+            self._lf_flush_period = float(PrintPatchLogger.lf_flush_period)
+        except (ValueError, TypeError):
+            self._lf_flush_period = 0
 
     @property
     def period(self):
@@ -208,7 +279,15 @@ class LogFlusher(threading.Thread):
         while True:
             period = self._period
             while not self._exit_event.wait(period or 1.0):
+                if self._lf_flush_period and time() - self._lf_last_flush > self._lf_flush_period:
+                    if isinstance(sys.stdout, PrintPatchLogger):
+                        sys.stdout.force_lf_flush()
+                    if isinstance(sys.stderr, PrintPatchLogger):
+                        sys.stderr.force_lf_flush()
+                    self._lf_last_flush = time()
+                # now signal the real flush
                 self._logger.flush()
+
             # check if period is negative or None we should exit
             if self._period is None or self._period < 0:
                 break

trains/backend_interface/metrics/events.py

@@ -1,6 +1,7 @@
 import abc
 import hashlib
 import time
+from logging import getLevelName
 from multiprocessing import Lock
 
 import attr
@@ -132,6 +133,24 @@ class ScalarEvent(MetricsEventAdapter):
             **self._get_base_dict())
 
 
+class ConsoleEvent(MetricsEventAdapter):
+    """ Console log event adapter """
+
+    def __init__(self, message, level, worker, **kwargs):
+        self._value = str(message)
+        self._level = getLevelName(level) if isinstance(level, int) else str(level)
+        self._worker = worker
+        super(ConsoleEvent, self).__init__(metric=None, variant=None, iter=0, **kwargs)
+
+    def get_api_event(self):
+        return events.TaskLogEvent(
+            task=self._task,
+            timestamp=self._timestamp,
+            level=self._level,
+            worker=self._worker,
+            msg=self._value)
+
+
 class VectorEvent(MetricsEventAdapter):
     """ Vector event adapter """
 

trains/backend_interface/metrics/reporter.py

@@ -1,8 +1,9 @@
 import json
+import logging
 import math
 
 try:
-    from collections.abc import Iterable
+    from collections.abc import Iterable  # noqa
 except ImportError:
     from collections import Iterable
 
@@ -17,7 +18,8 @@ from ...utilities.plotly_reporter import create_2d_histogram_plot, create_value_
     create_2d_scatter_series, create_3d_scatter_series, create_line_plot, plotly_scatter3d_layout_dict, \
     create_image_plot, create_plotly_table
 from ...utilities.py3_interop import AbstractContextManager
-from .events import ScalarEvent, VectorEvent, ImageEvent, PlotEvent, ImageEventNoUpload, UploadEvent, MediaEvent
+from .events import ScalarEvent, VectorEvent, ImageEvent, PlotEvent, ImageEventNoUpload, \
+    UploadEvent, MediaEvent, ConsoleEvent
 from ...config import config
 
 
@@ -143,7 +145,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :param value: Reported value
         :type value: float
         :param iter: Iteration number
-        :type value: int
+        :type iter: int
         """
         ev = ScalarEvent(metric=self._normalize_name(title), variant=self._normalize_name(series), value=value,
                          iter=iter)
@@ -157,9 +159,9 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :param series: Series (AKA variant)
         :type series: str
         :param values: Reported values
-        :type value: [float]
+        :type values: [float]
         :param iter: Iteration number
-        :type value: int
+        :type iter: int
         """
         if not isinstance(values, Iterable):
             raise ValueError('values: expected an iterable')
@@ -190,7 +192,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :type plot: str or dict
         :param iter: Iteration number
         :param round_digits: number of digits after the dot to leave
-        :type value: int
+        :type round_digits: int
         """
         def floatstr(o):
             if o != o:
@@ -251,7 +253,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
             for presentation of processing. Currently only http(s), file and s3 schemes are supported.
         :type src: str
         :param iter: Iteration number
-        :type value: int
+        :type iter: int
         """
         ev = ImageEventNoUpload(metric=self._normalize_name(title), variant=self._normalize_name(series), iter=iter,
                                 src=src)
@@ -268,7 +270,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
             for presentation of processing. Currently only http(s), file and s3 schemes are supported.
         :type src: str
         :param iter: Iteration number
-        :type value: int
+        :type iter: int
         """
         ev = ImageEventNoUpload(metric=self._normalize_name(title), variant=self._normalize_name(series), iter=iter,
                                 src=src)
@@ -290,6 +292,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :type path: str
         :param image: Image data. Required unless filename is provided.
         :type image: A PIL.Image.Image object or a 3D numpy.ndarray object
+        :param upload_uri: Destination URL
         :param max_image_history: maximum number of image to store per metric/variant combination
         use negative value for unlimited. default is set in global configuration (default=5)
         :param delete_after_upload: if True, one the file was uploaded the local copy will be deleted
@@ -321,6 +324,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :param path: A path to an image file. Required unless matrix is provided.
         :type path: str
         :param stream: File/String stream
+        :param upload_uri: Destination URL
         :param file_extension: file extension to use when stream is passed
         :param max_history: maximum number of files to store per metric/variant combination
         use negative value for unlimited. default is set in global configuration (default=5)
@@ -353,7 +357,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
             A row for each dataset(bar in a bar group). A column for each bucket.
         :type histogram: numpy array
         :param iter: Iteration number
-        :type value: int
+        :type iter: int
         :param labels: The labels for each bar group.
         :type labels: list of strings.
         :param xlabels: The labels of the x axis.
@@ -466,7 +470,7 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :param series: Series (AKA variant)
         :type series: str
         :param data: A scattered data: pairs of x,y as rows in a numpy array
-        :type scatter: ndarray
+        :type data: ndarray
         :param iter: Iteration number
         :type iter: int
         :param mode: (type str) 'lines'/'markers'/'lines+markers'
@@ -520,16 +524,17 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :param xtitle: optional x-axis title
         :param ytitle: optional y-axis title
         :param ztitle: optional z-axis title
+        :param fill: optional
         :param comment: comment underneath the title
         :param layout_config: optional dictionary for layout configuration, passed directly to plotly
         :type layout_config: dict or None
         """
         data_series = data if isinstance(data, list) else [data]
 
-        def get_labels(i):
+        def get_labels(a_i):
             if labels and isinstance(labels, list):
                 try:
-                    item = labels[i]
+                    item = labels[a_i]
                 except IndexError:
                     item = labels[-1]
                 if isinstance(item, list):
@@ -666,11 +671,11 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
         :param series: Series (AKA variant)
         :type series: str
         :param iter: Iteration number
-        :type value: int
+        :type iter: int
         :param path: A path to an image file. Required unless matrix is provided.
         :type path: str
         :param matrix: A 3D numpy.ndarray object containing image data (RGB). Required unless filename is provided.
-        :type matrix: str
+        :type matrix: np.ndarray
         :param upload_uri: upload image destination (str)
         :type upload_uri: str
         :param max_image_history: maximum number of image to store per metric/variant combination
@@ -686,8 +691,8 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
                       file_history_size=max_image_history)
 
         if matrix is not None:
-            width = matrix.shape[1]
-            height = matrix.shape[0]
+            width = matrix.shape[1]  # noqa
+            height = matrix.shape[0]  # noqa
         else:
             # noinspection PyBroadException
             try:
@@ -722,6 +727,21 @@ class Reporter(InterfaceBase, AbstractContextManager, SetupUploadMixin, AsyncMan
             iter=iter,
         )
 
+    def report_console(self, message, level=logging.INFO):
+        """
+        Report a scalar value
+        :param message: message (AKA metric)
+        :type message: str
+        :param level: log level (int or string, log level)
+        :type level: int
+        """
+        ev = ConsoleEvent(
+            message=message,
+            level=level,
+            worker=self.session.worker,
+        )
+        self._report(ev)
+
     @classmethod
     def _normalize_name(cls, name):
         return name

trains/backend_interface/task/task.py

@@ -172,57 +172,9 @@ class Task(IdObjectBase, AccessMixin, SetupUploadMixin):
         if running_remotely() or DevWorker.report_stdout:
             log_to_backend = False
         self._log_to_backend = log_to_backend
-        self._setup_log(default_log_to_backend=log_to_backend)
         self._artifacts_manager = Artifacts(self)
         self._hyper_params_manager = HyperParams(self)
 
-    def _setup_log(self, default_log_to_backend=None, replace_existing=False):
-        """
-        Setup logging facilities for this task.
-        :param default_log_to_backend: Should this task log to the backend. If not specified, value for this option
-        will be obtained from the environment, with this value acting as a default in case configuration for this is
-        missing.
-        If the value for this option is false, we won't touch the current logger configuration regarding TaskHandler(s)
-        :param replace_existing: If True and another task is already logging to the backend, replace the handler with
-        a handler for this task.
-        """
-        # Make sure urllib is never in debug/info,
-        disable_urllib3_info = config.get('log.disable_urllib3_info', True)
-        if disable_urllib3_info and logging.getLogger('urllib3').isEnabledFor(logging.INFO):
-            logging.getLogger('urllib3').setLevel(logging.WARNING)
-
-        log_to_backend = get_log_to_backend(default=default_log_to_backend) or self._log_to_backend
-        if not log_to_backend:
-            return
-
-        # Handle the root logger and our own logger. We use set() to make sure we create no duplicates
-        # in case these are the same logger...
-        loggers = {logging.getLogger(), LoggerRoot.get_base_logger()}
-
-        # Find all TaskHandler handlers for these loggers
-        handlers = {logger: h for logger in loggers for h in logger.handlers if isinstance(h, TaskHandler)}
-
-        if handlers and not replace_existing:
-            # Handlers exist and we shouldn't replace them
-            return
-
-        # Remove all handlers, we'll add new ones
-        for logger, handler in handlers.items():
-            logger.removeHandler(handler)
-
-        # Create a handler that will be used in all loggers. Since our handler is a buffering handler, using more
-        # than one instance to report to the same task will result in out-of-order log reports (grouped by whichever
-        # handler instance handled them)
-        backend_handler = TaskHandler(task=self)
-
-        # Add backend handler to both loggers:
-        # 1. to root logger root logger
-        # 2. to our own logger as well, since our logger is not propagated to the root logger
-        #    (if we propagate our logger will be caught be the root handlers as well, and
-        #    we do not want that)
-        for logger in loggers:
-            logger.addHandler(backend_handler)
-
     def _validate(self, check_output_dest_credentials=True):
         raise_errors = self._raise_on_validation_errors
         output_dest = self.get_output_destination(raise_on_error=False, log_on_error=False)

trains/config/default/sdk.conf

@@ -176,6 +176,10 @@
             # Log all stdout & stderr
             log_stdout: true
 
+            # Line feed (\r) support. If zero (0) \r treated as \n and flushed to backend
+            # Line feed flush support in seconds, flush consecutive line feeds (\r) every X (default: 10) seconds
+            console_lf_flush_period: 10
+
             # compatibility feature, report memory usage for the entire machine
             # default (false), report only on the running process and its sub-processes
             report_global_mem_used: false

trains/logger.py

@@ -5,13 +5,15 @@ from typing import Any, Sequence, Union, List, Optional, Tuple, Dict, TYPE_CHECK
 
 import numpy as np
 import six
+from PIL import Image
+from pathlib2 import Path
+
+from .debugging.log import LoggerRoot
 
 try:
     import pandas as pd
 except ImportError:
     pd = None
-from PIL import Image
-from pathlib2 import Path
 
 from .backend_interface.logger import StdStreamPatch, LogFlusher
 from .backend_interface.task import Task as _Task
@@ -19,7 +21,6 @@ from .backend_interface.task.development.worker import DevWorker
 from .backend_interface.task.log import TaskHandler
 from .backend_interface.util import mutually_exclusive
 from .config import running_remotely, get_cache_dir, config
-from .debugging.log import LoggerRoot
 from .errors import UsageError
 from .storage.helper import StorageHelper
 from .utilities.plotly_reporter import SeriesInfo
@@ -29,9 +30,8 @@ warnings.filterwarnings('always', category=DeprecationWarning, module=__name__)
 
 
 if TYPE_CHECKING:
-    from matplotlib.figure import Figure as MatplotlibFigure
-    from matplotlib import pyplot
-    from plotly.graph_objects import Figure
+    from matplotlib.figure import Figure as MatplotlibFigure  # noqa
+    from matplotlib import pyplot  # noqa
 
 
 class Logger(object):
@@ -60,7 +60,7 @@ class Logger(object):
     _tensorboard_logging_auto_group_scalars = False
     _tensorboard_single_series_per_graph = config.get('metrics.tensorboard_single_series_per_graph', False)
 
-    def __init__(self, private_task):
+    def __init__(self, private_task, connect_stdout=True, connect_stderr=True, connect_logging=False):
         """
         .. warning::
             **Do not construct Logger manually!**
@@ -73,12 +73,26 @@ class Logger(object):
         self._default_upload_destination = None
         self._flusher = None
         self._report_worker = None
-        self._task_handler = None
         self._graph_titles = {}
         self._tensorboard_series_force_prefix = None
+        self._task_handler = TaskHandler(task=self._task, capacity=100)
+        self._connect_std_streams = connect_stdout or connect_stderr
+        self._connect_logging = connect_logging
 
-        if self._task.is_main_task():
-            StdStreamPatch.patch_std_streams(self)
+        # Make sure urllib is never in debug/info,
+        disable_urllib3_info = config.get('log.disable_urllib3_info', True)
+        if disable_urllib3_info and logging.getLogger('urllib3').isEnabledFor(logging.INFO):
+            logging.getLogger('urllib3').setLevel(logging.WARNING)
+
+        StdStreamPatch.patch_std_streams(self, connect_stdout=connect_stdout, connect_stderr=connect_stderr)
+
+        if self._connect_logging:
+            StdStreamPatch.patch_logging_formatter(self)
+        elif not self._connect_std_streams:
+            # make sure that at least the main trains logger is connect
+            base_logger = LoggerRoot.get_base_logger()
+            if base_logger and base_logger.handlers:
+                StdStreamPatch.patch_logging_formatter(self, base_logger.handlers[0])
 
     @classmethod
     def current_logger(cls):
@@ -381,6 +395,7 @@ class Logger(object):
             example: extra_layout={'xaxis': {'type': 'date', 'range': ['2020-01-01', '2020-01-31']}}
         """
 
+        # noinspection PyArgumentList
         series = [self.SeriesInfo(**s) if isinstance(s, dict) else s for s in series]
 
         # if task was not started, we have to start it
@@ -832,7 +847,7 @@ class Logger(object):
                 upload_uri = storage.verify_upload(folder_uri=upload_uri)
 
             if isinstance(image, Image.Image):
-                image = np.array(image)
+                image = np.array(image)  # noqa
             # noinspection PyProtectedMember
             self._task._reporter.report_image_and_upload(
                 title=title,
@@ -1068,7 +1083,7 @@ class Logger(object):
         :param float period: The period to flush the logger in seconds. To set no periodic flush,
             specify ``None`` or ``0``.
         """
-        if self._task.is_main_task() and DevWorker.report_stdout and DevWorker.report_period and \
+        if self._task.is_main_task() and self._task_handler and DevWorker.report_period and \
                 not running_remotely() and period is not None:
             period = min(period or DevWorker.report_period, DevWorker.report_period)
 
@@ -1099,6 +1114,30 @@ class Logger(object):
         self.report_image(title=title, series=series, iteration=iteration, local_path=path, image=matrix,
                           max_image_history=max_image_history, delete_after_upload=delete_after_upload)
 
+    def capture_logging(self):
+        # type: () -> "_LoggingContext"
+        """
+        Return context capturing all the logs (via logging) reported under the context
+
+        :return: a ContextManager
+        """
+        class _LoggingContext(object):
+            def __init__(self, a_logger):
+                self.logger = a_logger
+
+            def __enter__(self, *_, **__):
+                if not self.logger:
+                    return
+                StdStreamPatch.patch_logging_formatter(self.logger)
+
+            def __exit__(self, *_, **__):
+                if not self.logger:
+                    return
+                StdStreamPatch.remove_patch_logging_formatter()
+
+        # Do nothing if we already have full logging support
+        return _LoggingContext(None if self._connect_logging else self)
+
     @classmethod
     def tensorboard_auto_group_scalars(cls, group_scalars=False):
         # type: (bool) -> None
@@ -1137,7 +1176,7 @@ class Logger(object):
     def _remove_std_logger(cls):
         StdStreamPatch.remove_std_logger()
 
-    def _console(self, msg, level=logging.INFO, omit_console=False, *args, **kwargs):
+    def _console(self, msg, level=logging.INFO, omit_console=False, *args, **_):
         # type: (str, int, bool, Any, Any) -> None
         """
         print text to log (same as print to console, and also prints to console)
@@ -1158,29 +1197,32 @@ class Logger(object):
                                msg='Logger failed casting log level "%s" to integer' % str(level))
             level = logging.INFO
 
-        if not running_remotely():
+        # noinspection PyProtectedMember
+        if not running_remotely() or not self._task._is_remote_main_task():
+            if self._task_handler:
                 # noinspection PyBroadException
                 try:
                     record = self._task.log.makeRecord(
                         "console", level=level, fn='', lno=0, func='', msg=msg, args=args, exc_info=None
                     )
                     # find the task handler that matches our task
-                if not self._task_handler:
-                    self._task_handler = [h for h in LoggerRoot.get_base_logger().handlers
-                                          if isinstance(h, TaskHandler) and h.task_id == self._task.id][0]
                     self._task_handler.emit(record)
                 except Exception:
                     # avoid infinite loop, output directly to stderr
+                    # noinspection PyBroadException
                     try:
                         # make sure we are writing to the original stdout
                         StdStreamPatch.stderr_original_write(
                             'trains.Logger failed sending log [level {}]: "{}"\n'.format(level, msg))
                     except Exception:
                         pass
+            else:
+                # noinspection PyProtectedMember
+                self._task._reporter.report_console(message=msg, level=level)
 
         if not omit_console:
             # if we are here and we grabbed the stdout, we need to print the real thing
-            if DevWorker.report_stdout and not running_remotely():
+            if self._connect_std_streams and not running_remotely():
                 # noinspection PyBroadException
                 try:
                     # make sure we are writing to the original stdout
@@ -1218,7 +1260,7 @@ class Logger(object):
         :param path: A path to an image file. Required unless matrix is provided.
         :type path: str
         :param matrix: A 3D numpy.ndarray object containing image data (RGB). Required unless filename is provided.
-        :type matrix: str
+        :type matrix: np.array
         :param max_image_history: maximum number of image to store per metric/variant combination \
         use negative value for unlimited. default is set in global configuration (default=5)
         :type max_image_history: int
@@ -1306,14 +1348,15 @@ class Logger(object):
         pass
 
     def _flush_stdout_handler(self):
-        if self._task_handler and DevWorker.report_stdout:
+        if self._task_handler:
             self._task_handler.flush()
 
     def _close_stdout_handler(self, wait=True):
         # detach the sys stdout/stderr
+        if self._connect_std_streams:
             StdStreamPatch.remove_std_logger(self)
 
-        if self._task_handler and DevWorker.report_stdout:
+        if self._task_handler:
             t = self._task_handler
             self._task_handler = None
             t.close(wait)

trains/task.py

@@ -190,6 +190,7 @@ class Task(_Task):
         auto_connect_arg_parser=True,  # type: Union[bool, Mapping[str, bool]]
         auto_connect_frameworks=True,  # type: Union[bool, Mapping[str, bool]]
         auto_resource_monitoring=True,  # type: bool
+        auto_connect_streams=True,  # type: Union[bool, Mapping[str, bool]]
     ):
         # type: (...) -> Task
         """
@@ -350,6 +351,24 @@ class Task(_Task):
             - ``False`` - Do not automatically create.
             - Class Type - Create ResourceMonitor object of the specified class type.
 
+        :param auto_connect_streams: Control the automatic logging of stdout and stderr
+
+            The values are:
+
+            - ``True`` - Automatically connect (default)
+            -  ``False`` - Do not automatically connect
+            - A dictionary - In addition to a boolean, you can use a dictionary for fined grained control of stdout and
+              stderr. The dictionary keys are 'stdout' , 'stderr' and 'logging', the values are booleans.
+              Keys missing from the dictionary default to ``False``, and an empty dictionary defaults to ``False``.
+              Notice, the default behaviour is logging stdout/stderr the
+              `logging` module is logged as a by product of the stderr logging
+
+            For example:
+
+            .. code-block:: py
+
+               auto_connect_streams={'stdout': True, 'stderr': True`, 'logging': False}
+
         :return: The main execution Task (Task context).
         """
 
@@ -390,7 +409,7 @@ class Task(_Task):
                 # create a new logger (to catch stdout/err)
                 cls.__main_task._logger = None
                 cls.__main_task.__reporter = None
-                cls.__main_task.get_logger()
+                cls.__main_task._get_logger(auto_connect_streams=auto_connect_streams)
                 cls.__main_task._artifacts_manager = Artifacts(cls.__main_task)
                 # unregister signal hooks, they cause subprocess to hang
                 # noinspection PyProtectedMember
@@ -453,7 +472,8 @@ class Task(_Task):
                         continue_last_task=continue_last_task,
                         detect_repo=False if (
                                 isinstance(auto_connect_frameworks, dict) and
-                                not auto_connect_frameworks.get('detect_repository', True)) else True
+                                not auto_connect_frameworks.get('detect_repository', True)) else True,
+                        auto_connect_streams=auto_connect_streams,
                     )
                     # set defaults
                     if cls._offline_mode:
@@ -545,7 +565,7 @@ class Task(_Task):
         # Make sure we start the logger, it will patch the main logging object and pipe all output
         # if we are running locally and using development mode worker, we will pipe all stdout to logger.
         # The logger will automatically take care of all patching (we just need to make sure to initialize it)
-        logger = task.get_logger()
+        logger = task._get_logger(auto_connect_streams=auto_connect_streams)
         # show the debug metrics page in the log, it is very convenient
         if not is_sub_process_task_id:
             if cls._offline_mode:
@@ -2110,7 +2130,7 @@ class Task(_Task):
     @classmethod
     def _create_dev_task(
         cls, default_project_name, default_task_name, default_task_type,
-            reuse_last_task_id, continue_last_task=False, detect_repo=True,
+            reuse_last_task_id, continue_last_task=False, detect_repo=True,  auto_connect_streams=True
     ):
         if not default_project_name or not default_task_name:
             # get project name and task name from repository name and entry_point
@@ -2230,8 +2250,7 @@ class Task(_Task):
         task.reload()
 
         # force update of base logger to this current task (this is the main logger task)
-        task._setup_log(replace_existing=True)
-        logger = task.get_logger()
+        logger = task._get_logger(auto_connect_streams=auto_connect_streams)
         if closed_old_task:
             logger.report_text('TRAINS Task: Closing old development task id={}'.format(default_task.get('id')))
         # print warning, reusing/creating a task
@@ -2271,8 +2290,8 @@ class Task(_Task):
 
         return task
 
-    def _get_logger(self, flush_period=NotSet):
-        # type: (Optional[float]) -> Logger
+    def _get_logger(self, flush_period=NotSet, auto_connect_streams=False):
+        # type: (Optional[float], Union[bool, dict]) -> Logger
         """
         get a logger object for reporting based on the task
 
@@ -2288,10 +2307,15 @@ class Task(_Task):
             # do not recreate logger after task was closed/quit
             if self._at_exit_called:
                 raise ValueError("Cannot use Task Logger after task was closed")
-            # force update of base logger to this current task (this is the main logger task)
-            self._setup_log(replace_existing=self.is_main_task())
             # Get a logger object
-            self._logger = Logger(private_task=self)
+            self._logger = Logger(
+                private_task=self,
+                connect_stdout=(auto_connect_streams is True) or
+                               (isinstance(auto_connect_streams, dict) and auto_connect_streams.get('stdout', False)),
+                connect_stderr=(auto_connect_streams is True) or
+                               (isinstance(auto_connect_streams, dict) and auto_connect_streams.get('stderr', False)),
+                connect_logging=isinstance(auto_connect_streams, dict) and auto_connect_streams.get('logging', False),
+            )
             # make sure we set our reported to async mode
             # we make sure we flush it in self._at_exit
             self._reporter.async_enable = True