Sidd Karamcheti

By default, EGLInitialize() (in containers like Docker/Singularity or natively) looks to bind to GPU:0 (first absolute device) on a given machine. When running MatterSim under HPC (e.g., via SLURM or...

Shell config

3

If you don't open atom from the command line, and launch the term plugin, any shell customization is gone (i.e zsh) and pane is misaligned. Yet when you open atom...

Add results on all bAbI Tasks

Hi! This is an amazing effort, and I’m excited to try using this platform in my upcoming work; however, I can’t find the `method` directory with the BC agent implementation....

Hey @lucidrains - hope you're doing well! I have some time to hack the next couple weeks, just wanted to get a sense of: - Current state of benchmarking (what...

Get sentence level exact match accuracy at train/validation?

2

Is there an easy way to get exact match of translating from source to target at train time/validation?

Example-based Reweighting

2

Is there a way to reweight individual examples (sentence pairs) by some value? E.g. scale the loss for a given example by 0.5 (arbitrary scalars).

Non-Transporter Baselines

1

Hey Mohit, I’m looking to test some models that do language conditioned imitation without the underlying Transporter networks inductive bias (state, language —> action). Is this feasible from say ~100...

Adding new robots

1

Love the Ravens work & ability to script demonstrations with the provided robot. I'm looking to work on cross-embodiment transfer with language, specifically trying to use the existing Ravens workspace...

This is an incredible project! For reproducibility, and for some of my own work, would you mind sharing/pointing me to code for fine-tuning VQGAN models (e.g., `vqgan_imagenet_f16_16384`) on custom datasets?...