Merge pull request #59 from overlabs/master

minor fix

doc/docs/guide/logging/index.md

@@ -23,6 +23,37 @@ output:
 2023-03-18-13:56:03 > test.py > hello world 
 ```
 
+## Using a decorator
+
+`@logger.mention` will mention the decorated function on each call.
+
+```python
+@logger.mention
+def function_a():
+    print('message from a')
+
+def function_b():
+    function_a()
+    print('message from b')
+
+def function_c():
+    function_a()
+    function_b()
+    print('message from c')
+
+function_c()
+```
+output:
+```html
+2023-11-18-06:26:04 > main/function_c >  Currently running: function_a
+message from a
+2023-11-18-06:26:04 > main/function_b >  Currently running: function_a
+message from a
+message from b
+message from c
+```
+
+
 ## Auto tracing
 
 The default `logger` automatically trace back the caller of logger instance.

neetbox/daemon/__init__.py

@@ -29,6 +29,7 @@ def __attach_daemon(daemon_config):
                 "ipython, try to set 'allowIpython' to True."
             )
             return False  # ignore if debugging in ipython
+    pkg.is_installed("flask", try_install_if_not=True)
     _online_status = connect_daemon(daemon_config)  # try to connect daemon
     logger.log("daemon connection status: " + str(_online_status))
     if not _online_status:  # if no daemon online

neetbox/daemon/_local_http_client.py

@@ -1,5 +1,9 @@
 import httpx
+import logging
 
+httpx_logger = logging.getLogger("httpx")
+httpx_logger.setLevel(logging.ERROR)
+
 __no_proxy = {
     "http://": None,
     "https://": None,

neetbox/logging/formatting.py

@@ -4,13 +4,25 @@
 # URL:    https://gong.host
 # Date:   20230318
 
+from dataclasses import dataclass
 import os
 import warnings
 from random import random
 from typing import Optional
 
-
+@dataclass
 class LogStyle:
+    color: Optional[str] = None
+    prefix: str = ""
+    text_style: Optional[str] = None
+    datetime_format: str = "%Y-%m-%d-%H:%M:%S"
+    with_identifier: bool = True
+    trace_level = 3
+    with_datetime: bool = True
+    split_char_cmd = " > "
+    split_char_identity = "/"
+    split_char_txt = " | "
+
     @classmethod
     def get_supported_colors(cls):
         return ["red", "green", "blue", "cyan", "yellow", "magenta"]
@@ -19,18 +31,6 @@ def get_supported_colors(cls):
     def get_supported_text_style(cls):
         return ["bold", "italic", "blink"]
 
-    def __init__(self) -> None:
-        self.color: Optional[str] = None
-        self.prefix: str = ""
-        self.text_style: Optional[str] = None
-        self.datetime_format: str = "%Y-%m-%d-%H:%M:%S"
-        self.with_identifier: bool = True
-        self.trace_level = 3
-        self.with_datetime: bool = True
-        self.split_char_cmd = " > "
-        self.split_char_identity = "/"
-        self.split_char_txt = " | "
-
     def parse(self, pattern: str):
         # todo
         pass

neetbox/torch/nn/functional.py

@@ -0,0 +1,38 @@
+import torch
+from typing import Union
+
+def one_hot(
+    tensor: torch.Tensor,
+    num_classes: int,
+    ignored_label: Union[str, int] = "negative",
+):
+    """An advanced version of F.one_hot with ignore label support. Convert the mask to a one-hot encoded representation by @visualDust
+
+    Args:
+        tensor (torch.Tensor): indexed label image. Should types int
+        num_classes (int): number of classes
+        ignored_label (Union[str|int], optional): specify labels to ignore, or ignore by pattern. Defaults to "negative".
+
+    Returns:
+        torch.Tensor: one hot encoded tensor
+    """
+    original_shape = tensor.shape
+    for _ in range(4 - len(tensor.shape)):
+        tensor = tensor.unsqueeze(0)  # H W -> C H W -> B C H W, if applicable
+    # start to handle ignored label
+    # convert ignored label into positive index bigger than num_classes
+    if type(ignored_label) is int:
+        tensor[tensor == ignored_label] = num_classes
+    elif ignored_label == "negative":
+        tensor[tensor < 0] = num_classes
+
+    # check if mask image is valid
+    if torch.max(tensor) > num_classes:
+        raise RuntimeError("class values must be smaller than num_classes.")
+    B, _, H, W = tensor.shape
+    one_hot = torch.zeros(B, num_classes + 1, H, W)
+    one_hot.scatter_(1, tensor, 1)  # mark 1 on channel(dim=1) with index of mask
+    one_hot = one_hot[:, :num_classes]  # remove ignored label(s)
+    for _ in range(len(one_hot.shape) - len(original_shape)):
+        one_hot.squeeze_(0)  # B C H W -> H W ->  C H W, if applicable
+    return one_hot