Launch HN: Moonglow (YC S24) – Serverless Jupyter Notebooks
Moonglow is a platform that allows local Jupyter notebooks to run on remote cloud GPUs, integrating with VSCode and currently supporting Runpod and AWS, with free API keys available for testing.
Moonglow is a platform developed by Leila and Trevor that allows users to run local Jupyter notebooks on remote cloud GPUs, streamlining the process of scaling experiments for data scientists and machine learning researchers. The service integrates with VSCode, enabling users to start and stop pre-configured remote machines that function like standard Jupyter kernels. This innovation addresses common challenges faced by researchers, such as the need to frequently switch between cloud providers for GPU availability and the cumbersome setup of remote machines and Jupyter servers. Moonglow simplifies this by orchestrating the provisioning of cloud machines and establishing a connection to the Jupyter kernel. The platform currently supports Runpod and AWS, with plans to add GCP and Azure in the future. Users can access a demo video showcasing Moonglow's capabilities, including training a ResNet model on the CIFAR-10 dataset. For a limited time, Moonglow is offering free API keys for users to test the service without needing their own compute resources. The founders are seeking feedback to improve the platform and plan to implement a pricing model for individual users and teams in the future.
- Moonglow enables local Jupyter notebooks to run on remote cloud GPUs.
- The platform integrates with VSCode for easy management of cloud resources.
- It addresses challenges in scaling experiments for data scientists and ML researchers.
- Currently supports Runpod and AWS, with plans for GCP and Azure.
- Free API keys are available for users to test the service during the launch.
Related
Show HN: We made glhf.chat – run almost any open-source LLM, including 405B
The platform allows running various large language models via Hugging Face repo links using vLLM and GPU scheduler. Offers free beta access with plans for competitive pricing post-beta using multi-tenant model running.
GitHub Models: A new generation of AI engineers building on GitHub
GitHub has launched GitHub Models, providing developers access to advanced language models for AI experimentation, enhancing coding practices while ensuring privacy and security in development processes.
Launch HN: Release (YC W20) – Orchestrate AI Infrastructure and Applications
Release.ai, founded by Erik, Tommy, and David, offers a platform for orchestrating AI applications, providing free GPU cycles, prioritizing data security, and featuring a workflow engine with deployment templates.
Show HN: Attaching to a Virtual GPU over TCP
Thunder Compute provides a flexible, cost-efficient cloud-based GPU service with instant scaling, pay-per-use billing, high utilization rates, and strong security, benefiting enterprises by minimizing idle GPU time.
Launch HN: Outerport (YC S24) – Instant hot-swapping for AI model weights
Towaki and Allen are developing Outerport, a distribution network that optimizes AI model deployment, enabling quick model swapping and potentially reducing GPU costs by 40% through efficient management and orchestration.
- Users appreciate the UI and express curiosity about its design and usability compared to other platforms like Modal and Google Colab.
- Several comments focus on technical aspects, such as handling Python environments, filesystem access, and SSH key management.
- There are questions regarding the product's differentiation from similar services like Syncthing, Databricks, and brev.dev.
- Some users express excitement about the potential of the platform while questioning its long-term market viability.
- Overall, there is a mix of congratulations on the launch and constructive feedback on features and comparisons to existing tools.
What do you think of this compared with running a Jupyter server on Modal? (I think Modal is slightly harder, ie, you run a terminal command, but curious!) https://modal.com/docs/guide/notebooks
I suspect that’s a matter of time right?
1. How is this different from Syncthing and similar solutions? Syncthing is free, open-source, cloud agnostic and easy to use to accomplish what seems to be the same task as Moonglow. 2. What is serverless about this? It is not clear from the pitch above.
This is brilliant and "obvious" in a good way along those lines, congrats on the launch!
Curious, are the SSH keys stored on Moonglow's internal servers?
I really like their Jupyter repl format because it separates the cells with python comments so it’s much easier to deploy your code when you are done versus a notebook.
Related
Show HN: We made glhf.chat – run almost any open-source LLM, including 405B
The platform allows running various large language models via Hugging Face repo links using vLLM and GPU scheduler. Offers free beta access with plans for competitive pricing post-beta using multi-tenant model running.
GitHub Models: A new generation of AI engineers building on GitHub
GitHub has launched GitHub Models, providing developers access to advanced language models for AI experimentation, enhancing coding practices while ensuring privacy and security in development processes.
Launch HN: Release (YC W20) – Orchestrate AI Infrastructure and Applications
Release.ai, founded by Erik, Tommy, and David, offers a platform for orchestrating AI applications, providing free GPU cycles, prioritizing data security, and featuring a workflow engine with deployment templates.
Show HN: Attaching to a Virtual GPU over TCP
Thunder Compute provides a flexible, cost-efficient cloud-based GPU service with instant scaling, pay-per-use billing, high utilization rates, and strong security, benefiting enterprises by minimizing idle GPU time.
Launch HN: Outerport (YC S24) – Instant hot-swapping for AI model weights
Towaki and Allen are developing Outerport, a distribution network that optimizes AI model deployment, enabling quick model swapping and potentially reducing GPU costs by 40% through efficient management and orchestration.