I tried running the demo_graph.py and got the following error.

python demo_graph.py 
Traceback (most recent call last):
  File "demo_graph.py", line 56, in <module>
    w.add_graph(model, (dummy_input, ))
  File "/home/dana/Desktop/tensorboard-pytorch/tensorboardX/writer.py", line 400, in add_graph
    self.file_writer.add_graph(graph(model, input_to_model, verbose))
  File "/home/dana/Desktop/tensorboard-pytorch/tensorboardX/graph.py", line 52, in graph
    trace, _ = torch.jit.trace(model, args)
TypeError: 'function' object is not iterable

The return of torch.jit.trace is this wrapper, from this commit

I'm using the Python 3.6 on Ubuntu 16.04

• torch==0.4.0 built from source
• tensorboardX==1.1
• tensorboard==1.6.0

Edit: Updated my pytorch version

tensorboardX1.2 may not support upsample module in pytorch0.4

File "/home/wanghongzhen/anaconda2/lib/python2.7/site-packages/tensorboardX/writer.py", line 422, in add_graph self.file_writer.add_graph(graph(model, input_to_model, verbose)) File "/home/wanghongzhen/anaconda2/lib/python2.7/site-packages/tensorboardX/graph.py", line 94, in graph torch.onnx._optimize_trace(trace, False) File "/home/wanghongzhen/anaconda2/lib/python2.7/site-packages/torch/onnx/init.py", line 30, in _optimize_trace trace.set_graph(utils._optimize_graph(trace.graph(), aten)) File "/home/wanghongzhen/anaconda2/lib/python2.7/site-packages/torch/onnx/utils.py", line 95, in _optimize_graph graph = torch._C._jit_pass_onnx(graph, aten) File "/home/wanghongzhen/anaconda2/lib/python2.7/site-packages/torch/onnx/init.py", line 40, in _run_symbolic_function return utils._run_symbolic_function(*args, **kwargs) File "/home/wanghongzhen/anaconda2/lib/python2.7/site-packages/torch/onnx/utils.py", line 368, in _run_symbolic_function return fn(g, *inputs, **attrs) TypeError: upsample_bilinear2d() got an unexpected keyword argument 'align_corners' (occurred when translating upsample_bilinear2d)

I am pretty sure that the inputs are lists. I am not sure whether I am using tensorboard rightly.

for epoch in range(epochs):    
    batch_loss_list = []
    
    batch_list = seqHelper.gen_batch_list_of_lists(train_list,batch_size,(random_seed+epoch))

    #run through training minibatches
    for counter, train_batch in enumerate(batch_list):
        x_atom,x_bonds,x_atom_index,x_bond_index,x_mask,y_val = seqHelper.get_info_with_smiles_list(train_batch,\
                smiles_to_measurement,smiles_to_atom_info,smiles_to_bond_info,\
                smiles_to_atom_neighbors,smiles_to_bond_neighbors,smiles_to_atom_mask)

        atoms_prediction, mol_prediction = model(x_atom,x_bonds,x_atom_index,x_bond_index,x_mask)
        with SummaryWriter(comment='Fingerprint') as w:
            w.add_graph(model, (x_atom,x_bonds,x_atom_index,x_bond_index,x_mask))

It returns:

RuntimeError                              Traceback (most recent call last)
<ipython-input-6-fd2c7c65a5f0> in <module>()
      1 with SummaryWriter(comment='Fingerprint') as w:
----> 2     w.add_graph(model, (x_atom,x_bonds,x_atom_index,x_bond_index,x_mask))
      3 

~/anaconda3/envs/deepchem/lib/python3.5/site-packages/tensorboardX/writer.py in add_graph(self, model, input_to_model, verbose)
    398                 print('add_graph() only supports PyTorch v0.2.')
    399                 return
--> 400         self.file_writer.add_graph(graph(model, input_to_model, verbose))
    401 
    402     @staticmethod

~/anaconda3/envs/deepchem/lib/python3.5/site-packages/tensorboardX/graph.py in graph(model, args, verbose)
     50     if LooseVersion(torch.__version__) >= LooseVersion("0.4"):
     51         with torch.onnx.set_training(model, False):
---> 52             trace, _ = torch.jit.get_trace_graph(model, args)
     53         torch.onnx._optimize_trace(trace, False)
     54     else:

~/anaconda3/envs/deepchem/lib/python3.5/site-packages/torch/jit/__init__.py in get_trace_graph(f, args, kwargs, nderivs)
    251     if not isinstance(args, tuple):
    252         args = (args,)
--> 253     return LegacyTracedModule(f, nderivs=nderivs)(*args, **kwargs)
    254 
    255 

~/anaconda3/envs/deepchem/lib/python3.5/site-packages/torch/nn/modules/module.py in __call__(self, *input, **kwargs)
    369             result = self._slow_forward(*input, **kwargs)
    370         else:
--> 371             result = self.forward(*input, **kwargs)
    372         for hook in self._forward_hooks.values():
    373             hook_result = hook(self, input, result)

~/anaconda3/envs/deepchem/lib/python3.5/site-packages/torch/jit/__init__.py in forward(self, *args)
    277     def forward(self, *args):
    278         global _tracing
--> 279         in_vars, in_desc = _flatten(args)
    280         # NOTE: use full state, because we need it for BatchNorm export
    281         # This differs from the compiler path, which doesn't support it at the moment.

RuntimeError: Only tuples, lists and Variables supported as JIT inputs, but got numpy.ndarray\

Describe the bug add_graph() with torch 1.2.0 will show error 'torch._C.Value' object has no attribute 'uniqueName' but work well with torch 1.1.0

Minimal runnable code to reproduce the behavior

from tensorboardX import SummaryWriter
class LeNet(nn.Module):
    def __init__(self):
        super(LeNet, self).__init__()
        self.conv1 = nn.Sequential(     #input_size=(1*28*28)
            nn.Conv2d(1, 6, 5, 1, 2),
            nn.ReLU(),      #(6*28*28)
            nn.MaxPool2d(kernel_size=2, stride=2),  #output_size=(6*14*14)
        )
        self.conv2 = nn.Sequential(
            nn.Conv2d(6, 16, 5),
            nn.ReLU(),      #(16*10*10)
            nn.MaxPool2d(2, 2)  #output_size=(16*5*5)
        )
        self.fc1 = nn.Sequential(
            nn.Linear(16 * 5 * 5, 120),
            nn.ReLU()
        )
        self.fc2 = nn.Sequential(
            nn.Linear(120, 84),
            nn.ReLU()
        )
        self.fc3 = nn.Linear(84, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = self.conv2(x)
        x = x.view(x.size()[0], -1)
        x = self.fc1(x)
        x = self.fc2(x)
        x = self.fc3(x)
        return x

dummy_input = torch.rand(13, 1, 28, 28)
model = LeNet()
with SummaryWriter(comment='Net', log_dir='/output') as w:
    w.add_graph(model, (dummy_input, ))

Expected behavior

Screenshots

Environment

Python environment Python 3.6

Additional context Add any other context about the problem here.

The code is add_histogram(name,data, niter) as examples. But it can not work. The error msg is File "/opt/pytorch/lib/python3.5/site-packages/tensorboardX/writer.py", line 305, in add_histogram self.file_writer.add_summary(histogram(tag, values, bins), global_step) File "/opt/pytorch/lib/python3.5/site-packages/tensorboardX/summary.py", line 113, in histogram hist = make_histogram(values.astype(float), bins) File "/opt/pytorch/lib/python3.5/site-packages/tensorboardX/summary.py", line 131, in make_histogram bucket=counts) TypeError: 0 has type numpy.int64, but expected one of: int, long, float

After I installed tensorflow 1.3 ,tensorboard can be used in terminal But after I installed tensorboard-pytorch, it can not work:

Traceback (most recent call last):
  File "/usr/local/bin/tensorboard", line 7, in <module>
    from tensorboard.main import main
ImportError: No module named 'tensorboard.main'

If it's possible, supporting the newly added hyper params feature would be amazing :)

https://www.tensorflow.org/tensorboard/r2/hyperparameter_tuning_with_hparams

Hey, thanks for the great work!

It would be even better if there is an option to optionally remove a particular experiment like what we have in Crayon.

from pycrayon import CrayonClient

# Connect to the server
cc = CrayonClient(hostname="server_machine_address")
cc.remove_experiment(xp_name) # remove the 'xp_name' experiment

Thanks!

Hi, I'm a Pytorch beginner (previously working on th and tf) using your tensorboard-pytorch bridge to get some info during neural net training. It works like a charm for everything i've needed since now :). However i'm having some troubles with the graph visualization for some ConvNet and i've a few questions:

• Is it possible to define scope for module? (i.e. if I'have a nested module can i give it a name to obtain a compress visualization of everything inside it?) ;
• using the torch.cat() method i get some strange behaviours related to the order of the elements in the list of the first argument of cat() (see attached images, where the second one is correct for the code reported) .

Btw great work :)

def forward(self,i):
        # i stand as the input
        # conv_j is a module
        x = self.conv_0(i)
        x = self.conv_1(x)
        y = self.conv_r1(i)
        z = torch.cat((y,x),1)
        z = z.view(len(z),-1)
        z = self.fc1(z)
        z = F.relu(z)
        z = self.fc2(z)
        z = F.log_softmax(z)
        return z

This is originally found when our CI server installed protobuf 4.12.0, which was released today, as a dependency of the tensorboardX.

$ pip3 install tensorboardX  # installs the newest protobuf as a dependency
$ python3
Python 3.8.13 (default, Mar 17 2022, 16:53:17) 
[Clang 9.1.0 (clang-902.0.39.2)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> from tensorboardX import SummaryWriter
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/__init__.py", line 5, in <module>
    from .torchvis import TorchVis
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/torchvis.py", line 11, in <module>
    from .writer import SummaryWriter
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/writer.py", line 17, in <module>
    from .comet_utils import CometLogger
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/comet_utils.py", line 7, in <module>
    from .summary import _clean_tag
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/summary.py", line 13, in <module>
    from .proto.summary_pb2 import Summary
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/proto/summary_pb2.py", line 16, in <module>
    from tensorboardX.proto import tensor_pb2 as tensorboardX_dot_proto_dot_tensor__pb2
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/proto/tensor_pb2.py", line 16, in <module>
    from tensorboardX.proto import resource_handle_pb2 as tensorboardX_dot_proto_dot_resource__handle__pb2
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/tensorboardX/proto/resource_handle_pb2.py", line 36, in <module>
    _descriptor.FieldDescriptor(
  File "/Users/dtakahashi/Library/Python/3.8/lib/python/site-packages/google/protobuf/descriptor.py", line 560, in __new__
    _message.Message._CheckCalledFromGeneratedFile()
TypeError: Descriptors cannot not be created directly.
If this call came from a _pb2.py file, your generated code is out of date and must be regenerated with protoc >= 3.19.0.
If you cannot immediately regenerate your protos, some other possible workarounds are:
 1. Downgrade the protobuf package to 3.20.x or lower.
 2. Set PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python (but this will use pure-Python parsing and will be much slower).

More information: https://developers.google.com/protocol-buffers/docs/news/2022-05-06#python-updates
>>> 

Environment

$ pip3 list | grep -E "torch|proto|tensor"
protobuf                      4.21.0
tensorboard                   2.7.0
tensorboard-data-server       0.6.1
tensorboard-plugin-wit        1.8.0
tensorboardX                  2.5
torch                         1.10.1
$ python3 --version
Python 3.8.13

Thank you very much in advance.

I am trying to run the demo example, but when changing to the projector tab, the browser freezes while trying to compute PCA.

I am using Ubuntu 18.04, python 3.6, tensorflow-gpu 1.11, pytorch 4.1, any ideas on what might cause the problem?

Hi, I trained my model for 50 epochs. Now I want to continue training for other 15 epochs. so I start training with model.load_from_checkpoint("path") but I don't know how to show continuation of logs with tensorboard I thought if I write same path of previous file logs, it continue but it didn't show me the previous logs.

I'd be appreciate if you could help me because I need to see the previous training logs along with the new ones. Part of source code is:

checkpoint_callback=ModelCheckpoint(dirpath="model",filename="newmodel_1",save_last=True,verbose=True,monitor="val_ce_loss",mode="min")
logger=TensorBoardLogger("training-logs",name="test_1")
model1 = model(config)
model_with_previous = model1.load_from_checkpoint(base_path_ckpt)
trainer = pl.Trainer(
     max_steps = max_steps,
     max_epochs=N_EPOCHS,
     gpus=(1 if torch.cuda.is_available() else 0),
     logger=logger,
     callbacks=[checkpoint_callback], 
)
%load_ext tensorboard
%tensorboard --logdir ./lightning_logs

Any reason why protobuf version is restricted to 3.20.1? https://github.com/lanpa/tensorboardX/blob/9356270bce9e9d7e4f4ecedc85931b89f19a8a14/setup.py#L46 And can it be changed to >=3.8.0,<4 ?

The latest release of onnx requires protobuf >= 3.20.2, < 4 https://github.com/onnx/onnx/blob/ee7d2cdfa34b8b3c7e0b68b70daf72aaa48c23ac/requirements.txt#L2 Therefore, it is not possible to install both tensorboardX and onnx in the same environment without error messages.

Describe the bug When I was using tensorboardX in pytorch to record hyper parameters, I got the error as the title described.

Minimal runnable code to reproduce the behavior

import tensorboardX
writer = tensorboardX.SummaryWriter()
hparams = {'batch_size': args.batch_size,
           'lr': args.lr,
           'momentum': args.momentum,
           'weight_decay': args.weight_decay}
writer.add_hparams(hparams, {'best_acc1': best_acc1})
...

Expected behavior A clear and concise description of what you expected to happen.

Screenshots

Environment What is the result of

protobuf 3.19.6 tensorboard 2.10.1 tensorboard-data-server 0.6.1 tensorboard-plugin-wit 1.8.1 tensorboardX 2.2 torch 1.11.0 torch-tb-profiler 0.4.0 torchaudio 0.11.0 torchvision 0.12.0

Python environment Which version of python are you using? Did you use Andconda or Virtualenv? Andconda Additional context Add any other context about the problem here.

Describe the bug Any hparams not present in the first run will never be shown. This is frustrating if extra information is added as you realise it might be relevant during the experiment. You may also have different experiments in the same folder.

Minimal runnable code to reproduce the behavior

# Example 1
from tensorboardX import SummaryWriter
with SummaryWriter() as w:
    w.add_hparams({'param1': 0.1}, {"metric1": 1})
    w.add_hparams({'param1': 2, "param2": 4}, {"metric1": 3, "metric2": 2}) 

rm -rf runs/*

# Example 2
from tensorboardX import SummaryWriter
with SummaryWriter() as w:
    w.add_hparams({'param1': 0.1}, {"metric1": 1})
with SummaryWriter() as w:
    w.add_hparams({'param1': 2, "param2": 4}, {"metric1": 3, "metric2": 2})

Expected behavior The first example shows param1, param2, metric1, metric2. The second only shows param1 and metric1. I would expect the tables to contain all parameters and metrics present in all runs. They can be displayed as blanks if not present for a given run, as is the case for the first example.

Screenshots Example 1

Example 2

Environment

protobuf                      3.19.4
pytorch-ignite                0.4.9
tensorboard                   2.10.0
tensorboard-data-server       0.6.1
tensorboard-plugin-wit        1.8.1
tensorboardX                  2.5.1
torch                         1.12.1
torchvision                   0.13.1

Python environment Conda 3.9

Bumps numpy from 1.21.0 to 1.22.0.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Today when there are NaN or Inf values, it draw as 0. In Tensorboard, NaN or Inf are draw as Triangle Link

I wish to help but don't know where. Thanks, Roni