Nov 19
How Abundant designed efficient rollout architecture with Modal Sandboxes
Abundant builds RL environments that run on Modal Sandboxes, powering training for frontier labs.
Modal · Company blog · 8.1 a month · 158 posts since Jun 15, 2023
Nov 19
Abundant builds RL environments that run on Modal Sandboxes, powering training for frontier labs.
Nov 12
Why you're paying for capacity you aren't using, and how to pay only for what you use.
Nov 11
Modal runs upwards of 40 million gVisor containers every day. Here's why we bet on it, what we've built into it, and where we want to take it next.
Nov 6
Modal powers inference for the open-weight models that Snorkel is evaluating as well as GPU and CPU sandboxes for millions of simulations per month.
Oct 22
Spindle is an open-source API for training and fine-tuning open models, deploy a server with a single command and run any Tinker-compatible training code against it.
Oct 21
Reinforcement learning for LLMs is hard, so we built Modal Dojo, a library that gives you the building blocks, tuned defaults, and observability to get started with production-grade LLM training.
Oct 20
Voice agents, multiplayer games, and other workloads can now build on Sticky Sessions. Achieve strong affinity without losing autoscaling. Here’s how we built this on Modal Servers.
Oct 19
To test a legal agent that works for hours across thousands of documents, Legora boots its entire application, pinned to a specific commit, inside a separate Modal Sandbox for every eval run.
Oct 19
We benchmarked NVIDIA's new Arm-based Vera CPU on agentic Sandbox workloads. Under heavy load, it completes far more tasks than the x86 and Graviton alternatives, without sacrificing latency.
Oct 16
How we made VM Sandbox filesystems fast by speaking in blocks instead of files.
Oct 15
How to run a LiveKit video avatar with open-weights models on Modal GPUs, using Sticky Sessions for container affinity.
Oct 15
Muse Spark Open, a coding-optimized multimodal model from Meta, is now available on Modal Auto Endpoints.
Oct 14
Modal has crossed $1B in annualized run rate and raised a $750M Series D led by Accel to build new infrastructure for AI applications. We can't think of a better time to be building software.
Oct 13
We trained a model on agentic coding tasks using fully asynchronous RL and mixed-precision, mixed-accelerator samplers. With off-policy correction, training stayed stable through convergence, on par with homogeneous samplers, at a lower...
Oct 12
Test and promote inference endpoints to evaluate quality, latency, throughput, and cost.
Oct 5
TypeSafe AI launched Jev, a new class of AI model built for automation, and scaled to trillions of tokens over a weekend on Modal.
Oct 1
Sidecars are isolated containers that run alongside your main Sandbox on the same host.
Oct 1
Everything announced live at Modal's inaugural conference, Runtime.
Oct 1
VM Sandboxes are built for those who need to give their agents the power of a full computer.
Oct 1
Multi-node GPU clusters with RDMA, gang scheduled from Modal's shared capacity pool and billed by the second, behind a single decorator.
Sep 24
Maximizing perf on AI-SQL queries with the KV-optimal left-deep join
Sep 23
Learn how we optimized performance and efficiency serving the workload that is changing software engineering forever, and you can too.
Modal's own blog on modal.com, found by the daily check.
AI and agents (14), Infrastructure (11), Product and company (10) in the last six months.