Together launches Post-Training platform; Deel runs 5 parallel RL experiments on same GPUs
togethercompute · x · 2026-10-23
Together AI announced Together Post-Training, a managed RL/SFT service for open models: you write the training loop in Python (data, rewards, GRPO/PPO or custom losses, Multi-LoRA or full-weight), Together handles GPU capacity, training-rollout consistency, parallelism, and per-model tuned rollout engines. One RL step maps to sample(), yourreward(), forwardbackward(), optimstep(), weightssync(); Tinker-compatible, with parallel LoRA sessions and bring-your-own tracking like W&B.
Case study: Deel's data science team went from setup to real RL experiments in a couple of hours, then ran up to five concurrent experiments on the same GPU capacity, iterating much faster per GPU. On-demand B200s also available.
More from coding & agent
- MCP Tip: One-Line Instructions Beat Long Tool Descriptions, 18/20 vs 3/10 in Evals — colinmcnamara · 2026-10-23
- DynaLM: Open-Source CPU-First LLM Inference Engine Seeks Feedback — Unlikely_Safety_7456 · 2026-10-23
- Software factories as code: Warp runs agents, runners and model routers from markdown/YAML repos — vikvang1 · 2026-10-23
- Dev uses Codex to remotely root a Samsung TV and strip eMudhra root certificates — charliermarsh · 2026-10-23
- Raindrop launches Change Reports to track how agent changes perform in production — HamelHusain · 2026-10-23
- Stop keeping agent sessions local: shared traces enable self-improving agents — vikvang1 · 2026-10-23