Spaces:
Sleeping
Sleeping
metadata
title: GenSim
emoji: π
colorFrom: purple
colorTo: indigo
sdk: gradio
python_version: 3.9.13
sdk_version: 3.41.2
app_file: app.py
pinned: false
license: apache-2.0
Generative Simulation Interactive Demo
This demo is from the project:
GenSim: Generating Robotic Simulation Tasks via Large Language Models
Preparations
- Obtain an OpenAI API Key
Usage
- Click Run-Example will simulate one example of pre-saved tasks in the task library and render videos.
- Top-Down Model:
- Type in the desired task name in the box. Then GenSim will try to run through the pipeline to generate the task.
- The task name has the form word separated by a dash. Example: 'place-blue-in-yellow' and 'align-rainbow-along-line'.
- Bottom-Up Model: No need to type in desired task. GenSim will try to generate novel tasks that are different from the task library.
- Usage: Always click on "Setup/Reset Simulation" and then click "Run".
Guideline
- The first output is the current stage of the task generation pipeline.
- The second output shows the generated code from Gen-Sim
- If there are errors in the generation stage above, you will see an error log on the top right.
- If the orange borders are still on, then the task is being simulated and rendered.
- The rendered video will come out in a stream, i.e. it will render and re-render in a sequence. Each new update takes 15 seconds.
Known Limitations
- Code generation can fail or generate infeasible tasks. The success rate is around 0.5.
- The low-level pick place primitive does not do collision checking and cannot pick up certain objects.
- Top-down generation is typically more challenging if the task name is too vague or too distant from the primitives.
Note
For GPT-4 model, each inference costs about $\$$0.03. For GPT-3.5 model, each inference costs about $\$$0.005. You can select which LLM model you would like to use.