tango
tango copied to clipboard
allenai
Organize your experiments into discrete steps that can be cached and reused throughout the lifetime of your research project.
### 🐛 Describe the bug `BeakerExecutor` checks for a dirty working directory before launching, because it requires that all (most) changes are available on GitHub. But if you make a...
For example, ref steps can now look like `{type: "ref", ref: some_step, key: 0}` Closes #334
Or, as @dirkgr suggested, we could just not stream the logs. But I think we should merge this either way so that we have this fix *for the record*.
Fixes a bug that prevented users from specifying the output artifact type when using a wandb workspace. The artifact type can be set from the `step.metadata['artifact_kind']`. If that is not...
Bumps [black](https://github.com/psf/black) from 22.6.0 to 22.8.0. Release notes Sourced from black's releases. 22.8.0 Highlights Python 3.11 is now supported, except for blackd as aiohttp does not support 3.11 as of...
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot] avatar"){kind=avatar}
### 🚀 The feature, motivation and pitch If an object has an input argument of type `torch.Tensor` it currently cannot be constructed from jsonnet configs since the array-type from jsonnet...
Config files for machine learning projects can often get quite long (data paths, model config, training hyper-parameters). In past projects, I have used frameworks like [Hydra](https://github.com/facebookresearch/hydra) to compose smaller config...
### 🚀 The feature, motivation and pitch If a Tango step returns multiple values, we currently have to pass all the returned values to another Step. The goal is to...
