wandb / server

W&B Server is the self hosted version of Weights & Biases
MIT License
263 stars 21 forks source link

Tensorboard plugin not working #51

Closed NDer100KG closed 2 years ago

NDer100KG commented 2 years ago

Hi team WandB:

I am trying to use the function of syncing logs from tensorboard to wandb, and I also need the embedding projector function in tensorboard.

Although the charts have been synced to my wandb-local (assumed the codes are executing correctly? ), the tensorboard plugin seems not popped, as the following picture 2

However, the same code running on https://api.wandb.ai justs works, as the following picture 1

Should I manually add something in the container, or any other commands are suggested? Thanks

Here are the codes I am using:

import wandb
import numpy as np
import tensorflow as tf

wandb.init(project='test_tensorboard', sync_tensorboard=True)

log_writer = tf.summary.create_file_writer("logs")
with log_writer.as_default():
    for i in range(100):
        tf.summary.scalar("test", i, step = i)
vanpelt commented 2 years ago

@NDer100KG Automated tensorboard serving only works in our SaaS offering. We're planning to implement support for our local installations in the future, but this would require a kubernetes cluster and additional infrastructure.