It addresses the unique challenges of agentic RL—long trajectories, tool calls, and weight synchronization—by providing built-in partial rollout, handle-based data movement, and mismatch analysis for rollout-trainer consistency, all supported by production-grade serving and training engines.
English · 简体中文
XYZ AI Lab ·
Technical Blog
AxisRL is an agentic RL post-training framework built on SGLang rollout, Megatron training, and real-world agent workflows.
AxisRL connects high-throughput rollout, large-scale training, weight synchronization, data movement, resource scheduling, and reproducible debugging inside one coherent framework. SGLang and Megatron remain the core serving and training engines; AxisRL handles the system layer around agentic post-training.
LLM post-training workloads are moving beyond single-turn question answering. In agentic RL, a model may interact with a long-lived environment, call tools, observe tool results, update context, and receive a reward only after several turns.
That changes the job of a post-training framework. It has to coordinate multi-turn rollout, environment state, tool calls, verifiers, reward collection, training sample construction, and weight synchronization. It also has to make training behavior observable, because small differences in tokenization, chat templates, logprobs, routing, packing, or weight sync can appear later as loss spikes, reward instability, or rollout-trainer mismatch.
AxisRL is designed for this setting: real agent workflows, SGLang rollout, Megatron training, and the system contracts between them.
At a high level, an AxisRL run follows this loop:
AxisRL is an agentic RL post-training framework built on SGLang rollout, Megatron training, and real-world agent workflows.
Similar projects matched by category, topics, and programming language.
Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.
A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.
mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.