DeepSeek V4.1 Flash 1017 now available for post-training on Together AI with custom RL/SFT loops
togethercompute · x · 2026-11-18
Together AI added DeepSeek V4.1 Flash 1017 to its Post-Training platform. Users can run SFT or RL (GRPO/PPO or custom losses) with their own data, rewards, verifiers and environments, writing the training loop in their own Python. Together handles GPU capacity, training-rollout consistency, parallelism and tuned rollout engines; supports Multi-LoRA or full-weight training, parallel experiments, W&B integration and Tinker compatibility.
More from Infra
- Microsoft launches Azure AI Gateway Labs: 27 hands-on labs for governing models and agents — adnan_hashmi · 2026-11-18
- vLLM ships Day-0 support for Tencent Hy4 Flash with 8x H200 deployment recipe — vllm_project · 2026-11-18
- SpaceX CFO says power is the real compute bottleneck, eyes top AI infra spot — XFreeze · 2026-11-18
- AI2 releases MoE training tech report: rebuilt Olmo-core scales to 1.2T params on 512 B300 GPUs — AkariAsai · 2026-11-18
- SGLang hits v1.0 as NVIDIA, AMD, Intel, Google and Meta form its consortium — nikola_mr64990 · 2026-11-18
- Report: Nvidia sales staff introduced Alice Huang to Alibaba pre-ban, pre-ChatGPT — HanchungLee · 2026-11-18