Questions tagged [wandb]

134 questions
0
votes
1 answer

something wrong with wandb.login() ?

Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving. wandb: Currently logged in as: xxx (use wandb login --relogin to force relogin)
jr q
  • 1
  • 1
  • 1
0
votes
1 answer

How to fix Wandb API giving error 400 when deleting artifacts?

I accidentally logged way too much to wandb and would like to delete some artifacts. I've tried the following script, but I get an error 400 whenever I run it: import wandb wandb.login(key=KEY) api = wandb.Api(overrides={"project": PROJECT,…
dlwh
  • 2,257
  • 11
  • 23
0
votes
1 answer

Weights and Biases - cumulative max (highwater mark) for distributed training and sweep

I have an algorithm that I run 10 times, and return the best run by a cumulative maximum - So for each run, I return the highest validation score of the entire run. For example, this graph: turns to this graph: I ran 7 of these, and grouped them…
0
votes
1 answer

How to show wandb training progress from run folder

After training neural networks with wandb as the logger, I received a link to show the training results and a folder named "run-...", I assume that is the logging of the training process. Now I don't have that link, how to show the wandb training…
Teewip
  • 41
  • 1
  • 8
0
votes
1 answer

How to get wandb to pass arguments by position?

I am trying to explore the results of different parameter settings on my python script "train.py". For that, I use a wandb sweep. Each wandb agent executes the file "train.py" and passes some parameters to it. As per the wandb documentation…
R.Brä
  • 5
  • 1
  • 3
0
votes
0 answers

Pip installations on Colab from local

I'd like to use wandb on Colab, and I've installed it through pip on the command line. However, the import isn't recognized on Colab, so I have to run !pip install wandb each time. How can I install wandb locally so that I don't have to install it…
mpnm
  • 481
  • 1
  • 7
  • 13
0
votes
1 answer

python ray tune unable to stop trial or experiment

I am trying to make ray tune with wandb stop the experiment under certain conditions. stop all experiment if any trial raises an Exception (so i can fix the code and resume) stop if my score gets -999 stop if the variable varcannotbezero gets…
user670186
  • 2,588
  • 6
  • 37
  • 55
0
votes
1 answer

Log two model runs with Keras wandb

I am training multiple models in the same Colab notebook to compare some results. I've written a function to avoid repeating code, and I've added WandbCallback() in the list of callbacks for model_name.fit(). def generic_FE_trainer(model_name,…
mpnm
  • 481
  • 1
  • 7
  • 13
0
votes
1 answer

how to login wandb with another acount using colab?

For example, I have A, B acounts. First, I log in Google Colab with A account. and I want to log in wandb with B acounts. ( using !wandb login ) is it possible??
0
votes
1 answer

wandb loops work differently between Jupyter Notebook and Microsoft Visual Studio Code (Pyhton3)

The following code should loop 3 times. It works three times in Jupyter notbook but only once in VSC. Weird. import wandb N_EPOCHS=10 batch_size=16 from tqdm import tqdm for eachtime in tqdm(range(3)): …
gggg
  • 15
  • 3
0
votes
0 answers

Can I use Wandb in Tracking my YOLOv4 Training?

I want to track my training in YOLOv4 using wandb but I can't see any tutorial on how to do it. I saw a youtube video but it was a training in YOLOv5. My wandb account is now logged in in the YOLOv4 training but I cant see no chart in the wandb…
zeningning
  • 11
  • 2
0
votes
1 answer

wandb.plot.line does not work and it just shows a table

I used this example that was provided by WandB. However, the web interface just shows a table instead of a figure. data = [[i, random.random() + math.sin(i / 10)] for i in range(100)] table = wandb.Table(data=data, columns=["step",…
CVDE
  • 393
  • 3
  • 19
0
votes
1 answer

wandb: artifact.add_reference() option to add specific (not current) versionId or ETag to stop the need for re-upload to s3?

I feel like this should be possible, but I looked through the wandb SDK code and I can't find an easy/logical way to do it. It might be possible to hack it by modifying the manifest entries at some point later (but maybe before the artifact is…
Alex L
  • 4,168
  • 1
  • 9
  • 24
0
votes
0 answers

how to handle wandb Invalid filename characters exception on uploading image

I am using windows 10 & venv & python 3.9.7 this is my code to upload image to wandb wandb_log["Image/train_image"] = wandb.Image('tmp.jpg') wandb.log(wandb_log, step) the full directory of image is…
이준혁
  • 267
  • 4
  • 14
0
votes
1 answer

Model stopped training once I introduced << report_to = 'wandb' >> in TrainingArguments

I am downloading the model https://huggingface.co/microsoft/Multilingual-MiniLM-L12-H384/tree/main microsoft/Multilingual-MiniLM-L12-H384 and then using it. Transformer Version: '4.11.3' I have written the below code: import wandb wandb.login() %env…
MAC
  • 1,345
  • 2
  • 30
  • 60
1 2 3
8
9