Rf-frostzone リップル » mbridge github » Mbridge github megatron (54) 사진

Mbridge github megatron (54) 사진

探索在大模型训练中使用Megatron-Core 训练框架提高显存使用效率- NVIDIA 技术博客

eric (@erictang000) / Posts / X

veRL 训推一致性工作及重要性采样代码演进分析_rollout correction-CSDN博客

Anatomy of RL Frameworks

探索在大模型训练中使用Megatron-Core 训练框架提高显存使用效率_深度学习_NVIDIA AI 技术专区-NVIDIA AI 技术专区

Kính Dán An Toàn - Hồng Ngạn Glass

Dev Patel (@devpatelio) / Posts / X

NovaSky (@NovaSkyAI) / Posts / X

Dev Patel (@devpatelio) / Posts / X

探索在大模型训练中使用Megatron-Core 训练框架提高显存使用效率- NVIDIA 技术博客

Philipp Moritz (@pcmoritz) / Posts / X

Kính Dán An Toàn - Hồng Ngạn Glass

Megatron-Bridge/tutorials/training/reduced_precision_training.ipynb at main · NVIDIA-NeMo/Megatron-Bridge | Karin Sevegnani

探索在大模型训练中使用Megatron-Core 训练框架提高显存使用效率- NVIDIA 技术博客

猫头虎AI分享｜智谱开源了为RL scaling 设计的LLM post‑training 框架用于GLM-4.5强化学习训练：slime_slime is an llm post-training framework for rl sca-CSDN博客

FlowRL 用于奖励分布匹配强化学习方法- YouTube

veRL 训推一致性工作及重要性采样代码演进分析_rollout correction-CSDN博客

Power Up FSDP2 as a Flexible Training Backend for Miles | LMSYS Org

eric (@erictang000) / Posts / X

Philipp Moritz (@pcmoritz) / Posts / X

Kính Dán An Toàn - Hồng Ngạn Glass

Kính Dán An Toàn - Hồng Ngạn Glass

Dev Patel (@devpatelio) / Posts / X

Anatomy of RL Frameworks

Tyler Griggs (@tyler_griggs_) / Posts / X

Anatomy of RL Frameworks

Tyler Griggs (@tyler_griggs_) / Posts / X

Support FSDP2 as A Training Backend for slime - 知乎

Philipp Moritz (@pcmoritz) / Posts / X

Anatomy of RL Frameworks

Tyler Griggs (@tyler_griggs_) / Posts / X

NovaSky (@NovaSkyAI) / Posts / X

Dev Patel (@devpatelio) / Posts / X

智谱开源了为RL scaling 设计的LLM post‑training 框架用于GLM-4.5强化学习训练：slime - 文章- 开发者社区- 火山引擎

Anatomy of RL Frameworks

PDF) Mind2Report: A Cognitive Deep Research Agent for Expert-Level Commercial Report Synthesis

Support FSDP2 as A Training Backend for slime - 知乎

Dev Patel (@devpatelio) / Posts / X

探索在大模型训练中使用Megatron-Core 训练框架提高显存使用效率_深度学习_NVIDIA AI 技术专区-NVIDIA AI 技术专区

智谱开源了为RL scaling 设计的LLM post‑training 框架用于GLM-4.5强化学习训练：slime - 文章- 开发者社区- 火山引擎

1T LoRA RL 基于veRL的1T MoE LoRARL工程实践- YouTube

Megatron Bridge — Megatron Bridge

Kính Dán An Toàn - Hồng Ngạn Glass

Mind Lab 用10%的算力，做了件牛逼的事儿-腾讯新闻

eric (@erictang000) / Posts / X

Anatomy of RL Frameworks

Megatron-Bridge/tutorials/training/reduced_precision_training.ipynb at main · NVIDIA-NeMo/Megatron-Bridge | Karin Sevegnani

Philipp Moritz (@pcmoritz) / Posts / X

Megatron 性能调优＆ rollout FP8优化- YouTube

Dev Patel (@devpatelio) / Posts / X

Philipp Moritz (@pcmoritz) / Posts / X

veRL 训推一致性工作及重要性采样代码演进分析_rollout correction-CSDN博客

Dev Patel (@devpatelio) / Posts and Replies / X

0
1
2
3
4
5

16 февраль 0 0