跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-06AI 評分46

NeMo-DCR: Bit-Exact Delta-Compressed Refit for Scalable Agentic RL at Trillion-Parameter Scale

NeMo-DCR: Bit-Exact Delta-Compressed Refit for Scalable Agentic RL at Trillion-Parameter Scale

AI 導讀

NeMo-DCR 以增量壓縮方式同步模型權重,接收端取得的參數和緩衝區位元與完整權重同步一致。對 30B-1T 模型,即使權重變動率為 3% 或 5%,速度仍比完整檢查點傳輸快 12-40 倍;1T 模型在 3% 變動率下需時 150 s,完整傳輸則需 87.5 min。

正文

Published on Oct 6

Authors:

,

,

,

,

,

,

Abstract

Agentic reinforcement learning (RL) disaggregates training from rollout, so each policy update must reach the rollout clusters before the next batch. Transferring a full 1T checkpoint for such weight synchronization (refit) takes 87.5 min between two AWS regions. Measurements of BF16 training show that about 1% of weights change their stored values per step. Recent systems exploit this sparsity but fall short on placement, exactness, or efficiency: they reimplement placement rules, assemble full tensors, rebuild values arithmetically, or use a cross-cluster collective, and none fully recovers from mid-refit failures. We present NeMo-DCR (Delta-Compressed Refit), which sends only changes yet is bit-exact: receivers obtain the same parameter and buffer bits as a dense refit. For placement, fixed affine mappings project changes from training shards into the checkpoint's canonical coordinates, residual conversion covers the other changes, and the serving runtime's native loader places all changes in receiver storage. For exactness, compressible XOR masks carry affine changes whose projection and loader preserve stored bits, and overwrites carry the others. Receivers apply both in place, retries overwrite partial writes, and a joint commit binds the policy to the baseline for the next delta. For efficiency, object storage or a relay tree streams payloads during delta construction, without a cross-cluster collective. Even at 3% and 5% change rates, NeMo-DCR refits of 30B-1T models are 12-40times faster than a transport-only full-checkpoint reference. A 1T relay-tree refit at 3% takes 150 s instead of 87.5 min, making refits practical for cross-cluster agentic RL at trillion-parameter scale.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2610.08430

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.08430 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.08430 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.08430 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co