Hogwild! Inference: Parallel LLM Generation via Concurrent Attention

• URL: http://arxiv.org/abs/2504.06261v2
• Date: Wed, 09 Apr 2025 17:56:08 GMT
• Title: Hogwild! Inference: Parallel LLM Generation via Concurrent Attention
• Authors: Gleb Rodionov, Roman Garipov, Alina Shutova, George Yakushev, Vage Egiazarian, Anton Sinitsin, Denis Kuznedelev, Dan Alistarh,
• Abstract summary: Large Language Models (LLMs) tackle increasingly complex tasks through advanced reasoning, long-form content generation, and tool use.<n>Inference: a parallel LLM inference engine where multiple instances of the same LLM run in parallel with the same attention cache, with "instant" access to each other's generated tokens.<n>Hogwild! inference takes advantage of Rotary Position Embeddings (RoPE) to avoid recomputation while improving parallel hardware utilization.
• Score: 28.52740266774425
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Large Language Models (LLMs) have demonstrated the ability to tackle increasingly complex tasks through advanced reasoning, long-form content generation, and tool use. Solving these tasks often involves long inference-time computations. In human problem solving, a common strategy to expedite work is collaboration: by dividing the problem into sub-tasks, exploring different strategies concurrently, etc. Recent research has shown that LLMs can also operate in parallel by implementing explicit cooperation frameworks, such as voting mechanisms or the explicit creation of independent sub-tasks that can be executed in parallel. However, each of these frameworks may not be suitable for all types of tasks, which can hinder their applicability. In this work, we propose a different design approach: we run LLM "workers" in parallel , allowing them to synchronize via a concurrently-updated attention cache and prompt these workers to decide how best to collaborate. Our approach allows the instances to come up with their own collaboration strategy for the problem at hand, all the while "seeing" each other's partial progress in the concurrent cache. We implement this approach via Hogwild! Inference: a parallel LLM inference engine where multiple instances of the same LLM run in parallel with the same attention cache, with "instant" access to each other's generated tokens. Hogwild! inference takes advantage of Rotary Position Embeddings (RoPE) to avoid recomputation while improving parallel hardware utilization. We find that modern reasoning-capable LLMs can perform inference with shared Key-Value cache out of the box, without additional fine-tuning.

Related papers

• Multi-Bin Batching for Increasing LLM Inference Throughput [19.652542432683234]
  Large language models (LL) grow in popularity improving the efficiency of their systems.<n> requests is a critical step in scheduling jobs on servers.<n> requests often have varying generation lengths, causing resource underutilization.<n>We formalize this problem from a queueing-theoretic perspective, and aim to design a throughput control policy.
  arXiv  Detail & Related papers  (2024-12-03T03:16:12Z)
• Benchmarking Agentic Workflow Generation [80.74757493266057]
  We introduce WorfBench, a unified workflow generation benchmark with multi-faceted scenarios and intricate graph workflow structures. We also present WorfEval, a systemic evaluation protocol utilizing subsequence and subgraph matching algorithms. We observe that the generated can enhance downstream tasks, enabling them to achieve superior performance with less time during inference.
  arXiv  Detail & Related papers  (2024-10-10T12:41:19Z)
• Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition [31.741274626546666]
  Large Language Models (LLMs) have demonstrated remarkable in-context learning capabilities. LLMs can perform multiple, computationally distinct ICL tasks simultaneously, during a single inference call, a capability we term "task superposition"
  arXiv  Detail & Related papers  (2024-10-08T01:28:57Z)
• Beyond Prompts: Dynamic Conversational Benchmarking of Large Language Models [0.0]
  We introduce a dynamic benchmarking system for conversational agents that evaluates their performance through a single, simulated, and lengthy user interaction. We context switch regularly to interleave the tasks, which constructs a realistic testing scenario in which we assess the Long-Term Memory, Continual Learning, and Information Integration capabilities of the agents.
  arXiv  Detail & Related papers  (2024-09-30T12:01:29Z)
• Can Long-Context Language Models Subsume Retrieval, RAG, SQL, and More? [54.667202878390526]
  Long-context language models (LCLMs) have the potential to revolutionize our approach to tasks traditionally reliant on external tools like retrieval systems or databases. We introduce LOFT, a benchmark of real-world tasks requiring context up to millions of tokens designed to evaluate LCLMs' performance on in-context retrieval and reasoning. Our findings reveal LCLMs' surprising ability to rival state-of-the-art retrieval and RAG systems, despite never having been explicitly trained for these tasks.
  arXiv  Detail & Related papers  (2024-06-19T00:28:58Z)
• An LLM-Tool Compiler for Fused Parallel Function Calling [1.990293258268139]
  State-of-the-art sequential reasoning in Large Language Models (LLMs) has expanded the capabilities of Copilots beyond conversational tasks to complex function calling. We propose LLM-Tool Compiler, which fuses similar types of tool operations under a single function at runtime, presenting them as a unified task to the LLM. Benchmarked on a large-scale Copilot platform, LLM-Tool Compiler achieves up to four times more parallel calls than existing methods, reducing token costs and latency by up to 40% and 12%, respectively.
  arXiv  Detail & Related papers  (2024-05-07T18:55:50Z)
• Small LLMs Are Weak Tool Learners: A Multi-LLM Agent [73.54562551341454]
  Large Language Model (LLM) agents significantly extend the capabilities of standalone LLMs. We propose a novel approach that decomposes the aforementioned capabilities into a planner, caller, and summarizer. This modular framework facilitates individual updates and the potential use of smaller LLMs for building each capability.
  arXiv  Detail & Related papers  (2024-01-14T16:17:07Z)
• Distributed Inference and Fine-tuning of Large Language Models Over The Internet [91.00270820533272]
  Large language models (LLMs) are useful in many NLP tasks and become more capable with size. These models require high-end hardware, making them inaccessible to most researchers. We develop fault-tolerant inference algorithms and load-balancing protocols that automatically assign devices to maximize the total system throughput.
  arXiv  Detail & Related papers  (2023-12-13T18:52:49Z)
• An LLM Compiler for Parallel Function Calling [68.04566807806071]
  We introduce LLMCompiler, which executes functions in parallel to efficiently orchestrate multiple function calls. We observe consistent latency speedup of up to 3.7x, cost savings of up to 6.7x, and accuracy improvement of up to 9% compared to ReAct.
  arXiv  Detail & Related papers  (2023-12-07T18:32:04Z)
• Fast Chain-of-Thought: A Glance of Future from Parallel Decoding Leads to Answers Faster [61.83949316226113]
  FastCoT is a model-agnostic framework based on parallel decoding. We show that FastCoT saves inference time by nearly 20% with only a negligible performance drop compared to the regular approach.
  arXiv  Detail & Related papers  (2023-11-14T15:56:18Z)
• Answer-Set Programming for Lexicographical Makespan Optimisation in Parallel Machine Scheduling [18.286430978487388]
  We deal with a challenging scheduling problem on parallel machines with sequence-dependent setup times and release dates. We put the individual machine spans in non-ascending order and lexicographically minimise the resulting robustnesss. Our experimental results show that ASP is indeed a promising KRR paradigm for this problem and is competitive with state-of-the-art CP and MIP solvers.
  arXiv  Detail & Related papers  (2022-12-18T12:43:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.