Hugging Face Trending Papers
Resuming Vision-Language Inference across Onboard Interruptions
Read the original on Hugging Face Trending Papers →The Flow has not summarised this story yet — read it at Hugging Face Trending Papers.
Related stories
arXiv AI
Oct 23
Resuming Vision-Language Inference across Onboard Interruptions
OrbitWeaver is a framework that enables vision‑language models to resume inference after onboard interruptions on low‑Earth orbit satellites. It uses phase‑aware checkpointing and dependency‑aware dynamic recovery to preserve progress across power, thermal, and fault‑induced shutdowns, reducing restart latency and SSD write volume. Experiments on satellite payload hardware show up to a 4.07‑fold speedup and a 3.56‑fold reduction in write traffic, and a 72‑hour deployment confirmed its ability to maintain inference progress across repeated bus‑enforced shutdowns.
arXiv Machine Learning
Jun 19
Execution-State Capsules: Graph-Bound Execution-State Checkpoint and Restore for Low-Latency, Small-Batch, On-Device Physical-AI Serving
arXiv:2606. 20537v1 Announce Type: new Abstract: Mainstream LLM serving systems reuse prefix work mainly through paged or radix key-value (KV) caches.
Hugging Face Trending Papers
Jul 5
Sangam: Efficiently Serving Diffusion LLMs with the AR Stack
Diffusion language models (dLLMs) generate text by iteratively denoising a masked response and can commit multiple output positions per model invocation. Their bidirectional attention prevents exact autoregressive-style KV caching, since committing one position shifts the KV activations of all others.
Hugging Face Trending Papers
Jun 22
Concordia: JIT-Compiled Persistent-Kernel Checkpointing for Fault-Tolerant LLM Inference
Long-running LLM agents keep valuable state resident on GPUs: KV caches, request schedulers, communication state, and sometimes online adapters. Losing this state after a GPU or communicator failure can discard minutes to hours of work, yet existing recovery mechanisms either restart the whole serving stack or require application-specific checkpoint logic inside every attention and runtime component.
arXiv AI
Oct 5
Zephon: Elastic Determinism for Online, Stateful Foundation Model Data Loading Pipelines
Zephon is a data loader designed for foundation model training that ensures deterministic ordering of training data batches even when GPU resources change, checkpoints are resumed, or different processing backends are used. It handles online, stateful pipelines—where tokenization, packing, and mixing of samples create complex n‑to‑m transformations—by partitioning the data stream into topology‑independent lanes, serializing ordering decisions, and checkpointing only bounded in‑flight state. Experiments on text and vision‑language tasks show Zephon delivers competitive throughput while offering guarantees that existing loaders lack for such pipelines.
arXiv Machine Learning
Jul 8
PHOENIX: Resilient LLM Training with Hot-Swapping via Zero-Overhead Checkpoint
arXiv:2607. 01646v2 Announce Type: replace Abstract: State-of-the-art large language model (LLM) training takes tens of thousands of graphics processing units (GPUs) for months and encounters failures across the software and hardware stack.