tensorboard has 25.5M monthly downloads and 7K GitHub stars. Version 2.21.0 supports Python 3.9+, first published in 2017.
It is actively maintained under a permissive license, last released in Jun 2026.
It is built on top of absl-py, grpcio and markdown. It is used by 1,079 tracked packages, including nemo-toolkit, torchtnt and tbparse.
Package Insights
((week_daily_avg - month_daily_avg) / month_daily_avg) * 100Weekly Downloads
GitHub Stars
Downloads by OS
Python Versions
Top Countries
Dependencies
- absl-py >=0.4
- grpcio <2.0,>=1.74.0
- markdown >=2.6.8
- numpy >=1.12.0
- packaging
- pillow
- protobuf <8.0.0,>=6.31.1
- setuptools >=41.0.0
- tensorboard-data-server <0.8.0,>=0.7.0
- werkzeug >=1.0.1
Used By
- nemo-toolkit
- torchtnt
- tbparse
- rsl-rl-lib
- autogluon-timeseries
- yolov5
- mjlab
- tensorflow-intel
- tensorflow-cpu-aws
- autogluon-multimodal
- resemble-perth
- tensorflow-aarch64
- pyiqa
- tianshou
- skrl
- coqui-tts-trainer
- model-compression-toolkit
- pypots
- ms-swift
- torchtitan
- neuralprophet
- trainer
- verl
- netcal
- scvi-tools
- simpletransformers
- terratorch
- kauldron
- autotrain-advanced
- aimet-torch
- bigvgan
- nerfstudio
- super-gradients
- mct-nightly
- isaaclab
- fairseq2
- schnetpack
- alphafold3-pytorch
- axolotl
- torchutil
- uni2ts
- deepforest
- sbi
- mindstudio-probe
- opennmt-py
- tensorflow-cloud
- olpai-volunteer-agent
- clarinpl-embeddings
- nemotron-page-elements-v3
- data-gradients