[논문리뷰] Inference-Time Hyper-Scaling with KV Cache Compression
paper-review, with-gpt
26
분
Archives
시간순으로 모든 글을 둘러보세요.
총 503개의 글
타임라인
2025
153 개의 글
2025년 07월
28 개의 글
[논문리뷰] Llama-Nemotron: Efficient Reasoning Models
paper-review, with-gpt, efficient-llm, system-optimization, inference-acceleration
26
분
[논문리뷰] KIMI K2: OPEN AGENTIC INTELLIGENCE
paper-review, with-gpt, open-source, agentic-intelligence, RL-alignment, foundation-models
28
분
[논문리뷰] Qwen3 Technical Report
paper-review, foundation-models, with-gpt
26
분
K-뷰티, '어쩌다 성공'을 넘어 '구조적 성장'으로
산업분석, 화장품산업
4
분
Jekyll에서 Hugo로의 마이그레이션 가이드
일상
4
분
K-뷰티, 새로운 도약의 기로에 설까?
산업분석, 화장품산업
12
분
블로그 플랫폼 이전(jekyll to hugo)
일상
3
분
[논문리뷰] Massive Activations in Large Language Models
paper-review, with-gpt
24
분
[논문리뷰] Peri-LN: Revisiting Normalization Layer in the Transformer Architecture
paper-review, with-gpt
24
분
[논문리뷰] SageAttention3: Microscaling FP4 Attention for Inference and An Exploration of 8-bit Training
paper-review, with-gpt
26
분
[논문리뷰] Helix Parallelism: Rethinking Sharding Strategies for Interactive Multi-Million-Token LLM Decoding
paper-review, with-gpt
31
분
DeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning for Subgoal Decomposition
paper-review, with-gpt
23
분
Inference-Time Scaling for Generalist Reward Modeling
paper-review, with-gpt
22
분
Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI Architectures
paper-review, with-gpt
26
분
Code I/O: Condensing Reasoning Patterns via Code Input-Output Prediction
paper-review, with-gpt
31
분
Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention
paper-review, with-gpt
31
분
디지털 드래곤의 심장: AI 시대, 중국 데이터센터 배터리 시장에 투자해야 하는 이유
투자, 2차전지, deep-research, with-gemini
11
분
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
paper-review, with-gpt, DeepSeek
30
분
Janus-Pro: UnifiedMultimodalUnderstanding and Generation with Data and Model Scaling
paper-review, with-gpt, Janus, DeepSeek
31
분
셀프 로드 자전거 정비 가이드
자전거정비, 일상, with gpt
7
분
DeepSeek-V3 Technical Report
paper-review, with-gpt
35
분
DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
paper-review, with-gpt
31
분
Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation
paper-review, with-gpt
30
분
JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation
paper-review, with-gpt
26
분
Auxiliary-Loss-Free Load Balancing Strategy for Mixture-of-Experts
paper-review, with-gpt
28
분
DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search
paper-review, with-gpt
29
분
Fire-Flyer AI-HPC: A Cost-Effective Software-Hardware Co-Design for Deep Learning
paper-review, with-gpt
27
분
2025년 06월
29 개의 글
DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence
paper-review, with-gpt
30
분
DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence
paper-review, with-gpt
28
분
Let the Expert Stick to His Last: Expert-Specialized Fine-Tuning for Sparse Architectural Large Language Models
paper-review, with-gpt
26
분
DeepSeek LLM: Scaling Open-Source Language Models with Longtermism
paper-review, with-gpt, DeepSeek
25
분
DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models
paper-review, with-gpt
36
분
DreamCraft3D: Hierarchical 3D Generation with Bootstrapped Diffusion Prior
paper-review, with-gpt, 3D, Diffusion
28
분
Accelerated Test-Time Scaling with Model-Free Speculative Sampling
paper-review, with-gpt-o3
27
분
KVzip: Query-Agnostic KV Cache Compression with Context Reconstruction
paper-review, with-gpt-o3
24
분
Compress, Gather, and Recompute: REFORMingLong-Context Processing in Transformers
paper-review, with-gpt-o3
26
분
Mamba Drafters for Speculative Decoding
paper-review, with-gpt-o3
23
분
Accelerating LLM Inference Throughput via Asynchronous KV Cache Prefetching
paper-review, with-gemini-2.5-pro(preview)
39
분
Hogwild! Inference: Parallel LLM Generation via Concurrent Attention
paper-review, with-gemini-2.5-pro(preview)
38
분
MMInference: Accelerating Pre-filling for Long-Context Visual Language Models via Modality-Aware Permutation Sparse Attention
paper-review, with-gemini-2.5-pro(preview)
42
분
PRIMA.CPP: Speeding Up 70B-Scale LLM Inference on Low-Resource Everyday Home Clusters
paper-review, with-gemini-2.5-pro(preview)
41
분
Slim attention: cut your context memory in half without loss– K-cache is all you need for MHA
paper-review, with-gemini-2.5-pro(preview)
30
분
Towards Economical Inference: Enabling DeepSeek’s Multi-Head Latent Attention in Any Transformer-based LLMs
paper-review, with-gemini-2.5-pro(preview)
35
분
TransMLA: Multi-Head Latent Attention Is All You Need
paper-review, with-gemini-2.5-pro(preview)
33
분
X-EcoMLA: Upcycling Pre-Trained Attention into MLA for Efficient and Extreme KV Compression
paper-review, with-gemini-2.5-pro(preview)
32
분
A Bring-Your-Own-Model Approach for ML-Driven Storage Placement in Warehouse-Scale Computers
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
32
분
Know Where You're Uncertain When Planning with Multimodal Foundation Models: A Formal Framework
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
35
분
ReaL: Efficient RLHF Training of Large Language Models with Parameter Reallocation
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
30
분
Rubick: Exploiting Job Reconfigurability for Deep Learning Cluster Scheduling
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
27
분
Supply-Chain Attacks in Machine Learning Frameworks
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
27
분
Accelerating MoE Model Inference with Expert Sharding
paper-review, with-gemini-2.5-pro(preview)
34
분
FlexInfer: Breaking Memory Constraint via Flexible and Efficient Offloading for On-Device LLM Inference
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
35
분
ScaleFusion: Scalable Inference of Spatial-Temporal Diffusion Transformers for High-Resolution Long Video Generation
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
41
분
SOLA: Optimizing SLO Attainment for Large Language Model Serving with State-Aware Scheduling
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
38
분
XGRAMMAR: FLEXIBLE AND EFFICIENT STRUCTURED GENERATION ENGINE FOR LARGE LANGUAGE MODELS
paper-review, with-gemini-2.5-pro(preview), MLSYS2025
28
분
2025년 05월
7 개의 글
Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI Architectures
paper-review, with-gemini-2.5-pro(preview)
69
분
RODIMUS*: BREAKING THE ACCURACY-EFFICIENCY TRADE-OFF WITH EFFICIENT ATTENTIONS
paper-review, with-gemini-2.5-pro(preview)
46
분
An Empirical Study of Qwen3 Quantization
paper-review, with-gpt
15
분
Gemini Embedding: Generalizable Embeddings from Gemini
paper-review, with-gpt
17
분
Gemma 3 Technical Report
paper-review, with-gpt
15
분
MELODI: Exploring Memory Compression for Long Contexts
paper-review, with-gpt
24
분
Seesaw: High-throughput LLM Inference via Model Re-sharding
paper-review, with-gpt
18
분
2025년 04월
14 개의 글
Seedream 2.0: A Native Chinese-English Bilingual Image Generation Foundation Model
paper-review, with-gpt
15
분
Toward Efficient Inference for Mixture of Experts
paper-review, with-gpt
12
분
Comet: Fine-grained Computation-communication Overlapping for Mixture-of-Experts
paper-review, with-gpt
23
분
MegaScale-Infer: Serving Mixture-of-Experts at Scale with Disaggregated Expert Parallelism
paper-review, with-gpt
21
분
SwitchHead: Accelerating Transformers with Mixture-of-Experts Attention
paper-review, with-gpt
23
분
Mirage: A Multi-Level Superoptimizer for Tensor Programs
paper-review, with-gpt
30
분
MoEUT: Mixture-of-Experts Universal Transformers
paper-review, with-gpt
23
분
FLEX ATTENTION: A PROGRAMMING MODEL FOR GENERATING OPTIMIZED ATTENTION KERNELS
paper-review, with-gpt
28
분
Inference-Time Scaling for Generalist Reward Modeling
paper-review, with-gpt
26
분
LeanAttention: Hardware-Aware Scalable Attention Mechanism for the Decode-Phase of Transformers
paper-review, with-gpt
26
분
MegaScale-Infer: Serving Mixture-of-Experts at Scale with Disaggregated Expert Parallelism
paper-review, with-gpt
24
분
AIOpsLab: A Holistic Framework to Evaluate AI Agents for Enabling Autonomous Clouds
paper-review, with-gpt
15
분
SparseTransX: Efficient Training of Translation-Based Knowledge Graph Embeddings Using Sparse Matrix Operations
paper-review, with-gpt
15
분
2025년 03월
24 개의 글
Context Parallelism for Scalable Million-Token Inference
paper-review, with-gpt, MLSYS2025
18
분
NEO: Saving GPU Memory Crisis with CPU Offloading for Online LLM Inference
paper-review, with-gpt, MLSYS2025
25
분
PipeFill: Using GPUs During Bubbles in Pipeline-parallel LLM Training
paper-review, with-gpt
25
분
SELF-DATA DISTILLATION FOR RECOVERING QUALITY IN PRUNED LARGE LANGUAGE MODELS
paper-review, with-gpt
21
분
On Distributed Larger-Than-Memory Subset Selection With Pairwise Submodular Functions
paper-review, with-gpt
22
분
TRAINING ULTRA LONG CONTEXT LANGUAGE MODEL WITH FULLY PIPELINED DISTRIBUTED TRANSFORMER
paper-review, with-gpt
19
분
QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving
paper-review, with-gpt
29
분
Venn: Resource Management Across Federated Learning Jobs
paper-review, with-gpt
16
분
AI Metropolis: Scaling Large Language Model-based Multi-Agent Simulation with Out-of-order Execution
paper-review, with-gpt, MLSYS2025
19
분
Balancing Pipeline Parallelism with Vocabulary Parallelism
paper-review, with-gpt
35
분
DIFFSERVE: EFFICIENTLY SERVING TEXT-TO-IMAGE DIFFUSION MODELS WITH QUERY-AWARE MODEL SCALING
paper-review, with-gpt, MLSYS2025
31
분
EFFICIENT LLM INFERENCE USING DYNAMIC INPUT PRUNING AND CACHE-AWARE MASKING
paper-review, with-gpt, MLSYS2025
37
분
Marconi: Prefix Caching for the Era of Hybrid LLMs
paper-review, with-gpt, MLSYS2025
37
분
LAVA: LIFETIME-AWARE VM ALLOCATION WITH LEARNED DISTRIBUTIONS AND ADAPTATION TO MISPREDICTIONS
paper-review, with-gpt, MLSYS2025
39
분
TurboAttention: Efficient Attention Approximation for High Throughputs LLMs
paper-review, with-gpt, MLSYS2025
35
분
A PRACTICAL CROSS-LAYER APPROACH FOR ML-DRIVEN STORAGE PLACEMENT IN WAREHOUSE-SCALE COMPUTERS
paper-review, with-gpt, MLSYS2025
36
분
Comet: Fine-grained Computation-communication Overlapping for Mixture-of-Experts
paper-review, with-gpt, MLSYS2025
35
분
Scaling Deep Learning Training with MPMD Pipeline Parallelism
paper-review, with-gpt, MLSYS2025
28
분
LSERVE: EFFICIENT LONG-SEQUENCE LLM SERVING WITH UNIFIED SPARSE ATTENTION
paper-review, with-gpt, MLSYS2025
29
분
ThunderServe: High-performance and Cost-efficient LLM Serving in Cloud Environments
paper-review, with-gpt, MLSYS2025
24
분
VOLUT: EFFICIENT VOLUMETRIC STREAMING ENHANCED BY LUT-BASED SUPER-RESOLUTION
paper-review, with-gpt, MLSYS2025
27
분
Bridging the Safety Gap: A Guardrail Pipeline for Trustworthy LLM Inferences
paper-review, with-gpt
19
분
Forget the Data and Fine-Tuning! Just Fold the Network to Compress
paper-review, with-gpt
33
분
2025년 02월
31 개의 글
Dynamic Diffusion Transformer
paper-review, with-gpt, ICLR2025
36
분
HEXGEN-2: DISAGGREGATED GENERATIVE INFERENCE OF LLMS IN HETEROGENEOUS ENVIRONMENT
paper-review, with-gpt, ICLR2025
39
분
Speculate, then Collaborate: Fusing Knowledge of Language Models during Decoding
paper-review, with-gpt, ICLR2025
38
분
You OnlyPruneOnce: DESIGNING CALIBRATION-FREE MODEL COMPRESSION WITH POLICY LEARNING
paper-review, with-gpt, ICLR2025
36
분
FlashMask: Efficient and Rich Mask Extension of FlashAttention
paper-review, with-gpt, ICLR2025
37
분
TypedThinker: Typed Thinking Improves Large Language Model Reasoning
paper-review, with-gpt, ICLR2025
41
분
LASP-2: Rethinking Sequence Parallelism for Linear Attention and Its Hybrid
paper-review, with-gpt
26
분
BitsAI-CR: Automated Code Review via LLM in Practice
paper-review, with-gpt
36
분
Robust and Secure Code Watermarking for Large Language Models via ML/Crypto Codesign
paper-review, with-gpt
35
분
SmolLM2: When Smol Goes Big Data-Centric Training of a Small Language Mode
paper-review, with-gpt
38
분
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
paper-review, with-gpt
36
분
DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding
paper-review, with-gpt
39
분
Humanity's Last Exam
paper-review, with-gpt
16
분
Qwen2.5-1M Technical Report
paper-review, with-gpt
34
분
DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence
paper-review, with-gpt
39
분
JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation
paper-review, with-gpt
37
분
Qwen2-VL: Enhancing Vision-Language Model’s Perception of the World at AnyResolution
paper-review, with-gpt, Qwen
41
분
DeepSeek-Prover-V1.5: Harnessing Proof Assistant Feedback for Reinforcement Learning and Monte-Carlo Tree Search
paper-review, with-gpt, DeepSeek
43
분
Let the Expert Stick to His Last: Expert-Specialized Fine-Tuning for Sparse Architectural Large Language Models
paper-review, with-gpt
37
분
Qwen2 Technical Report
paper-review, with-gpt, Qwen
66
분
DeepSeek-Prover: Advancing Theorem Proving in LLMs through Large-Scale Synthetic Data
paper-review, with-gpt, DeepSeek
39
분
DeepSeek-VL: Towards Real-World Vision-Language Understanding
paper-review, with-gpt
40
분
How to Train Data-Efficient LLMs
paper-review, with-gpt
39
분
DeepSeek-Coder: When the Large Language Model Meets Programming - The Rise of Code Intelligence
paper-review, with-gpt
37
분
DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
paper-review, with-gpt
38
분
DeepSeek LLM: Scaling Open-Source Language Models with Longtermism
paper-review, with-gpt
36
분
DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models
paper-review, with-gpt
39
분
Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models
paper-review, with-gpt
38
분
Qwen Technical Report
paper-review, with-gpt
41
분
Qwen-VL: AVersatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
paper-review, with-gpt
34
분
Janus:DecouplingVisualEncoding for Unified Multimodal Understanding and Generation
paper-review, with-gpt
40
분
2025년 01월
20 개의 글
A Hardware Evaluation Framework for Large Language Model Inference
paper-review, with-gpt
25
분
Compressed Context Memory For Online Language Model Interaction
paper-review, with-gpt
28
분
DeepSeek-V3 Technical Report
paper-review, with-gpt, DeepSeek
29
분
Fast State Restoration in LLM Serving with HCache
paper-review, with-gpt
29
분
Qwen2.5 Technical Report
paper-review, with-gpt
34
분
AIOS: LLM Agent Operating System
paper-review, with-gpt
41
분
DynamicKV: Task-Aware Adaptive KV Cache Compression for Long Context LLMs
paper-review, with-gpt
23
분
SeerAttention: Learning Intrinsic Sparse Attention in Your LLMs
paper-review, with-gpt
29
분
TAIPAN: EFFICIENT AND EXPRESSIVE STATE SPACE LANGUAGE MODELS WITH SELECTIVE ATTENTION
paper-review, with-gpt
34
분
TokenRing: An Efficient Parallelism Framework for Infinite-Context LLMs via Bidirectional Communication
paper-review, with-gpt
21
분
SANA: EFFICIENT HIGH-RESOLUTION IMAGE SYN THESIS WITH LINEAR DIFFUSION TRANSFORMERS
paper-review, with-gpt
25
분
Block Transformer: Global-to-Local Language Modeling for Fast Inference
paper-review, with-gpt
29
분
FLAME: Factuality-Aware Alignment for Large Language Models
paper-review, with-gpt
24
분
MobiLlama: Towards Accurate and Lightweight Fully Transparent GPT
paper-review, with-gpt
22
분
Rethinking Optimization and Architecture for Tiny Language Models
paper-review, with-gpt
22
분
Cascade Speculative Drafting for Even Faster LLM Inference
paper-review, with-gpt
22
분
Distributed Inference and Fine-tuning of Large Language Models Over The Internet
paper-review, with-gpt
22
분
EE-LLM: Large-Scale Training and Inference of Early-Exit Large Language Models with 3D Parallelism
paper-review, with-gpt
21
분
Gated Linear Attention Transformers with Hardware-Efficient Training
paper-review, with-gpt
28
분
LLM in a flash : Efficient Large Language Model Inference with Limited Memory
paper-review, with-gpt
24
분
2024
326 개의 글
2024년 12월
190 개의 글
Agile-Quant: Activation-Guided Quantization for Faster Inference of LLMs on the Edge
paper-review, with-gpt
23
분
Improving alignment of dialogue agents via targeted human judgements
paper-review, with-gpt
26
분
Language Models are General-Purpose Interfaces
paper-review, with-gpt
23
분
OPT: Open Pre-trained Transformer Language Models
paper-review, with-gpt
25
분
SparQ Attention: Bandwidth-Efficient LLM Inference
paper-review, with-gpt
21
분
CBQ: Cross-Block Quantization for Large Language Models
paper-review, with-gpt
23
분
SCCA: Shifted Cross Chunk Attention for long contextual semantic expansion
paper-review, with-gpt
22
분
GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projection
paper-review, with-gpt
22
분
Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context
paper-review, with-gpt
28
분
Gemma: Open Models Based on Gemini Research and Technology
paper-review, with-gpt
24
분
LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding
paper-review, with-gpt
22
분
The Truth is in There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction
paper-review, with-gpt
25
분
Abseil Tip 231 여기와 저기 사이 – 간과되기 쉬운 몇 가지 알고리즘
cpp, abseil
3
분
Abseil Tip 232 변수 선언 시 auto를 언제 사용할 것인가
cpp, abseil
4
분
Abseil Tip 234 값, 포인터, 참조로 전달하기
cpp, abseil
3
분
Gemma 2: Improving Open Language Models at a Practical Size
paper-review, with-gpt
25
분
Mooncake: A KVCache-centric Disaggregated Architecture for LLM Serving
paper-review, with-gpt
24
분
SAGEATTENTION: ACCURATE 8-BIT ATTENTION FOR PLUG-AND-PLAY INFERENCE ACCELERATION
paper-review, with-gpt
24
분
The Llama 3 Herd of Models
paper-review, with-gpt
26
분
Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality
paper-review, with-gpt
34
분
Communication Compression for Tensor Parallel LLM Inference
paper-review, with-gpt
20
분
Context Parallelism for Scalable Million-Token Inference
paper-review, with-gpt
24
분
FastAttention: Extend FlashAttention2 to NPUs and Low-resource GPUs
paper-review, with-gpt
22
분
SimpleFSDP: Simpler Fully Sharded Data Parallel with torch.compile
paper-review, with-gpt
22
분
ClusterKV: Manipulating LLM KV Cache in Semantic Space for Recallable Compression
paper-review, with-gpt
25
분
Efficient LLM Inference with I/O-Aware Partial KV Cache Recomputation
paper-review, with-gpt
24
분
Large Concept Models: Language Modeling in a Sentence Representation Space
paper-review, with-gpt
26
분
SageAttention2 Technical Report: Accurate 4 Bit Attention for Plug-and-play Inference Acceleration
paper-review, with-gpt
25
분
Sharing and Throughput-oriented Token Batching
paper-review, with-gpt
23
분
SparseInfer: Training-free Prediction of Activation Sparsity for Fast LLM Inference
paper-review, with-gpt
21
분
Star Attention: Efficient LLM Inference over Long Sequences
paper-review, with-gpt
23
분
Byte Latent Transformer: Patches Scale Better Than Tokens
paper-review, with-gpt
25
분
Efficient Memory Management for Large Language Model Serving with PagedAttention
paper-review, with-gpt
28
분
GShard: Scaling Giant Models with Conditional Computation and Automatic Sharding
paper-review, with-gpt
30
분
GSPMD: General and Scalable Parallelization for ML Computation Graphs
paper-review, with-gpt
24
분
Memory Layers at Scale
paper-review, with-gpt
25
분
Orca: Progressive Learning from Complex Explanation Traces of GPT-4
paper-review, with-gpt
29
분
Abseil Tip 197 Reader Lock은 드물게 사용해야 합니다
cpp, abseil
3
분
Abseil Tip 224 vector.at() 사용 피하기
cpp, abseil
3
분
Abseil Tip 227 빈 컨테이너와 부호 없는 정수 연산 주의하기
cpp, abseil
3
분
Abseil Tip 229 템플릿 메타프로그래밍을 위한 순위 기반 오버로드
cpp, abseil
3
분
Fast Inference of Mixture-of-Experts Language Models with Offloading
paper-review, with-gpt
25
분
FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning
paper-review, with-gpt
28
분
FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision
paper-review, with-gpt
26
분
FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
paper-review, with-gpt
26
분
Infinite-LLM: Efficient LLM Service for Long Context with DistAttention and Distributed KVCache
paper-review, with-gpt
25
분
SuperServe: Fine-Grained Inference Serving for Unpredictable Workloads
paper-review, with-gpt
9
분
Abseil Tip 124 absl::StrFormat()
cpp, abseil
3
분
Abseil Tip 18 Substitute를 활용한 문자열 포맷팅
cpp, abseil
2
분
Abseil Tip 198 태그 타입(Tag Types)
cpp, abseil
4
분
Abseil Tip 215 AbslStringify()를 사용한 사용자 정의 타입 문자열화"
cpp, abseil
3
분
Abseil Tip 218 FTADLE로 확장 지점 설계하기
cpp, abseil
4
분
Abseil Tip 3 문자열 연결과 operator+ vs. StrCat()
cpp, abseil
3
분
Fast and Effective Weight Update for Pruned Large Language Models
paper-review, with-gpt
24
분
FFSplit: Split Feed-Forward Network For Optimizing Accuracy-Efficiency Trade-off in Language Model Inference
paper-review, with-gpt
23
분
FlightLLM: Efficient Large Language Model Inference with a Complete Mapping Flow on FPGAs
paper-review, with-gpt
26
분
Lightning Attention-2: A Free Lunch for Handling Unlimited Sequence Lengths in Large Language Models
paper-review, with-gpt
21
분
MoE-Mamba: Efficient Selective State Space Models with Mixture of Experts
paper-review, with-gpt
26
분
DeepSpeed-FastGen: High-throughput Text Generation for LLMs via MII and DeepSpeed-Inference
paper-review, with-gpt
29
분
DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving
paper-review, with-gpt
36
분
Inference without Interference: Disaggregate LLM Inference for Mixed Downstream Workloads
paper-review, with-gpt
28
분
INFERFLOW: AN EFFICIENT AND HIGHLY CONFIG URABLE INFERENCE ENGINE FOR LARGE LANGUAGE MODELS
paper-review, with-gpt
30
분
MEDUSA: Simple LLMInference Acceleration Framework with Multiple Decoding Heads
paper-review, with-gpt
23
분
Abseil Tip 186 함수는 무명 네임스페이스에 두는 것을 선호하세요
cpp, abseil
4
분
Abseil Tip 187 std::unique_ptr Must Be Moved"
cpp, abseil
5
분
Abseil Tip 188 스마트 포인터를 함수 매개변수로 사용할 때 주의하세요
cpp, abseil
2
분
Break the Sequential Dependency of LLM Inference Using LOOKAHEAD DECODING
paper-review, with-gpt
33
분
Decoding Speculative Decoding
paper-review, with-gpt
37
분
Efficient Prompt Caching via Embedding Similarity
paper-review, with-gpt
25
분
FP6-LLM: Efficiently Serving Large Language Models Through FP6-Centric Algorithm-System Co-Design
paper-review, with-gpt
50
분
TP-Aware Dequantization
paper-review, with-gpt
31
분
Abseil Tip 116 함수 인자에서 참조 사용 시 주의사항
cpp, abseil
4
분
Abseil Tip 165 초기화 구문을 포함한 if와 switch 문 사용하기
cpp, abseil
4
분
Abseil Tip 181 StatusOr<T> 값 접근하기
cpp, abseil
4
분
Abseil Tip 76 absl::Status 사용하기
cpp, abseil
3
분
Benchmarking and Dissecting the Nvidia Hopper GPU Architecture
paper-review, with-gpt
42
분
Get More with LESS: Synthesizing Recurrence with KV Cache Compression for Efficient LLM Inferenc
paper-review, with-gpt
17
분
Hydragen: High-Throughput LLM Inference with Shared Prefixes
paper-review, with-gpt
26
분
QUICK: Quantization-aware Interleaving and Conflict-free Kernel for efficient LLM inference
paper-review, with-gpt
40
분
RelayAttention for Efficient Large Language Model Serving with Long System Prompts
paper-review, with-gpt
52
분
TeraPipe: Token-Level Pipeline Parallelism for Training Large-Scale Language Models
paper-review, with-gpt
19
분
Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers
paper-review, with-gpt
14
분
Fast Transformer Decoding: One Write-Head is All You Need
paper-review, with-gpt
16
분
GQA:Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints
paper-review, with-gpt
17
분
H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Model
paper-review, with-gpt
17
분
Scissorhands: Exploiting the Persistence of Importance Hypothesis for LLM KV Cache Compression at Test Time
paper-review, with-gpt
16
분
Token Merging: Your ViT But Faster
paper-review, with-gpt
15
분
Abseil Tip 140 상수(Constant) 처리 안전한 관용구
cpp, abseil
4
분
Abseil Tip 163 std::optional 매개변수 전달하기
cpp, abseil
3
분
Abseil Tip 171 Sentinel 값 피하기
cpp, abseil
3
분
Abseil Tip 172 지정 초기화자(Designated Initializers)
cpp, abseil
3
분
Abseil Tip 173 옵션 구조체로 인수 래핑하기
cpp, abseil
4
분
Abseil Tip 175 C++14와 C++17의 리터럴 상수 변경 사항
cpp, abseil
3
분
Abseil Tip 176 출력 매개변수 대신 반환 값을 선호하세요
cpp, abseil
4
분
Abseil Tip 177 할당 가능성과 데이터 멤버 타입
cpp, abseil
4
분
Abseil Tip 5 사라지는 객체의 함정
cpp, abseil
3
분
FlexLLM: A System for Co-Serving Large Language Model Inference and Parameter-Efficient Finetuning
paper-review, with-gpt
15
분
GLM-130B: An Open Bilingual Pre-trained Model
paper-review, with-gpt
22
분
LLM Inference Unveiled: Survey and Roofline Model Insights
paper-review, with-gpt
8
분
LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Models
paper-review, with-gpt
19
분
Model Tells You What to Discard: Adaptive KV Cache Compression for LLMs
paper-review, with-gpt
13
분
No Token Left Behind: Reliable KV Cache Compression via Importance-Aware Mixed Precision Quantization
paper-review, with-gpt
14
분
Scaling Instruction-Finetuned Language Models
paper-review, with-gpt
19
분
Benchmarks as Limits to Arbitrage: Understanding the Low-Volatility Anomaly
paper-review, with-gpt, finance
12
분
Abseil Tip 108 std::bind를 피하세요
cpp, abseil
3
분
Abseil Tip 132 Avoid Redundant Map Lookups
cpp, abseil
3
분
Abseil Tip 146 기본 초기화와 값 초기화
cpp, abseil
3
분
Abseil Tip 161 좋은 지역 변수와 나쁜 지역 변수
cpp, abseil
3
분
Abseil Tip 166 복사가 복사가 아닐 때
cpp, abseil
3
분
Abseil Tip 168 inline 변수
cpp, abseil
2
분
BLOOM: A 176B-Parameter Open-Access Multilingual Language Model
paper-review, with-gpt, LLM
18
분
CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving
paper-review, with-gpt, LLM-Inference
21
분
CHAI: Clustered Head Attention for Efficient LLM Inference
paper-review, with-gpt, LLM-Inference
19
분
Compressed Context Memory For Online Language Model Interaction
paper-review, with-gpt, LLM-Inference
14
분
Galactica: A Large Language Model for Science
paper-review, with-gpt, LLM
18
분
High Idiosyncratic Volatility and Low Returns: International and Further U.S. Evidence
paper-review, with-gpt, finance
12
분
LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compression
paper-review, with-gpt, LLM-Inference
20
분
QAQ: Quality Adaptive Quantization for LLM KV Cache
paper-review, with-gpt, LLM-Inference
17
분
Transformers are Multi-State RNNs
paper-review, with-gpt, LLM
15
분
Abseil Tip 147 Exhaustive switch 문을 책임감 있게 사용하기
cpp, abseil
4
분
Abseil Tip 158 Abseil 연관 컨테이너와 contains()
cpp, abseil
2
분
Abseil Tip 180 Dangling References(유효하지 않은 참조) 피하기
cpp, abseil
4
분
Abseil Tip 182 정수형 변수를 초기화하세요!
cpp, abseil
3
분
DeepCache: Accelerating Diffusion Models for Free
paper-review, with-gpt, LLM-Inference
19
분
Get More with LESS: Synthesizing Recurrence with KV Cache Compression for Efficient LLM Inference
paper-review, with-gpt, LLM-Inference
19
분
KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cache
paper-review, with-gpt, LLM-Inference
14
분
KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization
paper-review, with-gpt, LLM-Inference
14
분
Mixed Precision Quantization
paper-review, with-gpt, LLM-Inference
17
분
Momentum Strategies
paper-review, with-gpt, finance
27
분
WKVQuant: Quantizing Weight and Key/Value Cache for Large Language Models Gains More
paper-review, with-gpt, LLM-Inference
16
분
Abseil Tip 103 플래그는 전역 변수입니다
cpp, abseil
2
분
Abseil Tip 45 플래그를 피하라, 특히 라이브러리 코드에서
cpp, abseil
3
분
Abseil Tip 90 Retired Flags(사용 중단된 플래그)
cpp, abseil
3
분
Improving Language Understanding by Generative Pre-Training
paper-review, with-gpt, LLM
18
분
PaLM 2 Technical Report
paper-review, with-gpt, LLM
20
분
PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU
paper-review, with-gpt, LLM-Inference
13
분
QAQ: Quality Adaptive Quantization for LLM KV Cache
paper-review, with-gpt, LLM-Inference
17
분
Abseil Tip 136 Unordered Containers
cpp, abseil
2
분
Abseil Tip 144 연관 컨테이너에서의 이종 조회(Heterogeneous Lookup)
cpp, abseil
4
분
Abseil Tip 152 AbslHashValue과 함께
cpp, abseil
2
분
Abseil Tip 153 using-directives를 사용하지 마세요
cpp, abseil
4
분
ALISA: Accelerating Large Language Model Inference via Sparsity-Aware KV Caching
paper-review, with-gpt, LLM-Inference
16
분
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
paper-review, with-gpt, LLM
19
분
Dynamic Memory Compression: Retrofitting LLMs for Accelerated Inference
paper-review, with-gpt, Dynamic Memory Compression, LLM-Inference
14
분
Fast Inference from Transformers via Speculative Decoding
paper-review, with-gpt, LLM-Infernce, Speculative Decoding
21
분
FASTDECODE: High-Throughput GPU-Efficient LLM Serving using Heterogeneous Pipelines
paper-review, with-gpt, FASTDECODE, LLM-Inference
15
분
LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression
paper-review, with-gpt, LLMLingua-2, LLM-Inference
15
분
Abseil Tip 117 복사 생략과 값으로 전달하기
cpp, abseil
4
분
Abseil Tip 148 Overload Sets
cpp, abseil
4
분
Abseil Tip 149 Object Lifetimes vs = delete
cpp, abseil
4
분
Abseil Tip 24 복사, 축약
cpp, abseil
2
분
ALISA: Accelerating Large Language Model Inference via Sparsity-Aware KV Caching
paper-review, with-gpt
15
분
Direct Preference Optimization: Your Language Model is Secretly a Reward Model
paper-review, with-gpt
13
분
HIERARCHICAL CONTEXT MERGING: BETTER LONG CONTEXT UNDERSTANDING FOR PRE-TRAINED LLMS
paper-review, with-gpt
12
분
LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression
paper-review, with-gpt
16
분
MELTing point: Mobile Evaluation of Language Transformers
paper-review, with-gpt
13
분
MuxServe:FlexibleSpatial-TemporalMultiplexingforMultipleLLMServing
paper-review, with-gpt
14
분
The Flan Collection: Designing Data and Methods for Effective Instruction Tuning
paper-review, with-gpt
22
분
Transformer-Lite: High-efficiency Deployment of Large Language Models on Mobile Phone GPUs
paper-review, with-gpt
19
분
Tree of Thoughts: Deliberate Problem Solving with Large Language Models
paper-review, with-gpt
22
분
Abseil Tip 11 반환 정책
cpp, abseil
5
분
Abseil Tip 120 반환 값은 건드리지 마세요
cpp, abseil
4
분
Abseil Tip 143 C++11 삭제된 함수 (= delete)
cpp, abseil
6
분
CORM: Cache Optimization with Recent Message for Large Language Model Inference
paper-review, with-gpt
17
분
Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention
paper-review, with-gpt
21
분
Llama 2: Open Foundation and Fine-Tuned Chat Models
paper-review, with-gpt
28
분
Mistral 7B
paper-review, with-gpt
25
분
Retrieval Head Mechanistically Explains Long-Context Factuality
paper-review, with-gpt
22
분
SnapKV: LLM Knows What You are Looking for Before Generation
paper-review, with-gpt
16
분
SqueezeAttention: 2D Management of KV-Cache in LLM Inference via Layer-wise Optimal Budget
paper-review, with-gpt
24
분
Toward Inference-optimal Mixture-of-Expert Large Language Models
paper-review, with-gpt
14
분
Abseil Tip 134 make_unique와 private 생성자
cpp, abseil
4
분
Abseil Tip 141 bool로의 암시적 변환에 주의하라
cpp, abseil
3
분
Abseil Tip 61 기본 멤버 초기화 (Default Member Initializers)
cpp, abseil
3
분
Abseil Tip 93 absl::Span 사용하기
cpp, abseil
4
분
DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
paper-review, with-gpt
18
분
Fast Inference from Transformers via Speculative Decoding
paper-review, with-gpt
17
분
MobileAIBench: Benchmarking LLMs and LMMs for On-Device Use Cases
paper-review, with-gpt
7
분
Parallel Decoding via Hidden Transfer for Lossless Large Language Model Acceleration
paper-review, with-gpt
22
분
PowerInfer-2: Fast Large Language Model Inference on a Smartphone
paper-review, with-gpt
28
분
RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation
paper-review, with-gpt
16
분
Tree-based Speculative Inference and Verification
paper-review, with-gpt
18
분
Abseil Tip 142 다중 매개변수 생성자와 explicit
cpp, abseil
5
분
Abseil Tip 59 튜플 연결하기
cpp, abseil
2
분
Abseil Tip 88 초기화 방법 =, (), 그리고 {}
cpp, abseil
4
분
KV Cache Compression
paper-review, with-gpt
18
분
Layer-Condensed KV Cache for Efficient Inference of Large Language Models
paper-review, with-gpt
16
분
PyramidInfer: Pyramid KV Cache Compression for High-throughput LLM Inference
paper-review, with-gpt
16
분
SKVQ:Sliding-window Key and Value Cache Quantization for Large Language Models
paper-review, with-gpt
22
분
You Only Cache Once: Decoder-Decoder Architectures for Language Models
paper-review, with-gpt
16
분
2024년 11월
130 개의 글
Abseil Tip 10 문자열 분리, 골치 아프지 않게
cpp, abseil
3
분
Abseil Tip 3 문자열 연결과 operator+ vs. StrCat()
cpp, abseil
3
분
Abseil Tip 36 새로운 Join API
cpp, abseil
5
분
LoCoCo: Dropping In Convolutions for Long Context Compression
paper-review, with-gpt
22
분
Loki: Low-rank Keys for Efficient Sparse Attention
paper-review, with-gpt
15
분
MiniCache: KV Cache Compression in Depth Dimension for Large Language Models
paper-review, with-gpt
19
분
PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling
paper-review, with-gpt
17
분
Reducing Transformer Key-Value Cache Size with Cross-Layer Attention
paper-review, with-gpt
21
분
Abseil Tip 131 Special 멤버 함수와 = default
cpp, abseil
5
분
Abseil Tip 42 초기화 메서드보다 팩토리 함수를 선호하세요
cpp, abseil
4
분
Abseil Tip 74 위임 생성자와 상속 생성자
cpp, abseil
3
분
ASimple and Effective L2 Norm-Based Strategy for KV Cache Compression
paper-review, with-gpt
14
분
Attention Score is not All You Need for Token Importance Indicator in KV Cache Reduction: Value Also Matters
paper-review, with-gpt
18
분
CItruS : ChunkedInstruction-aware State Eviction for Long Sequence Modeling
paper-review, with-gpt
14
분
Effectively Compress KV Heads for LLM
paper-review, with-gpt
14
분
MLKV:Multi-Layer Key-Value Heads for Memory Efficient Transformer Decoding
paper-review, with-gpt
15
분
Benchmark of Long Context Capable Approaches
paper-review, with-gpt
12
분
Dynamic Discriminative Operations (D2O) for Efficient Generative Inference of Large Language Models
paper-review, with-gpt
14
분
Efficient Sparse Attention needs Adaptive Token Release
paper-review, with-gpt
21
분
LOOK-M:Look-Once Optimization in KV Cache for Efficient Multimodal Long-Context Inference
paper-review, with-gpt
16
분
MODEL TELLS YOU WHERE TO MERGE: ADAPTIVE KV CACHE MERGING FOR LLMS ON LONG-CONTEXT TASKS
paper-review, with-gpt
18
분
Abseil Tip 119 using 선언과 네임스페이스 별칭 사용하기
cpp, abseil
3
분
Abseil Tip 123 absl::optional과 std::unique_ptr
cpp, abseil
3
분
Abseil Tip 130 네임스페이스 이름 지정
cpp, abseil
3
분
Ada-KV: Optimizing KV Cache Eviction by Adaptive Budget Allocation for Efficient LLM Inference
paper-review, with-gpt
21
분
Keep the Cost Down: A Review on Methods to Optimize LLM’s KV Cache Consumption.
paper-review, with-gpt
6
분
LazyLLM: Dynamic Token Pruning for Efficient Long Context LLM Inference
paper-review, with-gpt
21
분
PQCache: Product Quantization-based KVCache for Long Context LLM Inference
paper-review, with-gpt
20
분
Pruning in Transformer Decoder
paper-review, with-gpt
21
분
Abseil Tip 109 함수 선언에서 의미 있는 const 사용
cpp, abseil
3
분
Abseil Tip 126 make_unique는 새로운 new입니다
cpp, abseil
3
분
Abseil Tip 99 비멤버 인터페이스 에티켓
cpp, abseil
4
분
NACL: AGeneral and Effective KV Cache Eviction Framework for LLMs at Inference Time
paper-review, with-gpt
21
분
Palu: Compressing KV-Cache with Low-Rank Projection
paper-review, with-gpt
25
분
Post-Training Sparse Attention with Double Sparsity
paper-review, with-gpt
18
분
RetrievalAttention: Accelerating Long-Context LLM Inference via Vector Retrieval
paper-review, with-gpt
18
분
ThinK: Thinner Key Cache by Query-Driven Pruning
paper-review, with-gpt
24
분
Discovering the Gems in Early Layers: Accelerating Long-Context LLMs with 1000x Input Token Reduction
paper-review, with-gpt
25
분
InfiniPot: Infinite Context Processing on Memory-Constrained LLMs
paper-review, with-gpt
18
분
KV-COMPRESS: Paged KV-Cache Compression with Variable Compression Rates per Attention Head
paper-review, with-gpt
24
분
Locret: Enhancing Eviction in Long-Context LLM Inference with Trained Retaining Heads
paper-review, with-gpt
22
분
TACO-RL: Task Aware Prompt Compression Optimization with Reinforcement Learning
paper-review, with-gpt
23
분
Abseil Tip 112 emplace vs. push_back
cpp, abseil
3
분
Abseil Tip 49 인자 기반 탐색
cpp, abseil
4
분
Abseil Tip 65 제자리에 넣기
cpp, abseil
3
분
DUOATTENTION: EFFICIENT LONG-CONTEXT LLM INFERENCE WITH RETRIEVAL AND STREAMING HEADS
paper-review, with-gpt
20
분
LoRC: Low-Rank Compression for LLMs KV Cache with a Progressive Compression Strategy
paper-review, with-gpt
15
분
SPARSEVLM: VISUAL TOKEN SPARSIFICATION FOR EFFICIENT VISION-LANGUAGE MODEL INFERENCE
paper-review, with-gpt
19
분
SWIFTKV: FAST PREFILL-OPTIMIZED INFERENCE WITH KNOWLEDGE-PRESERVING MODEL TRANSFORMATION
paper-review, with-gpt
19
분
TIDALDECODE: FAST AND ACCURATE LLM DECOD ING WITH POSITION PERSISTENT SPARSE ATTENTION
paper-review, with-gpt
16
분
TorchTitan: One-stop PyTorch native solution for production ready LLM pre-training
paper-review, with-gpt
16
분
A Systematic Study of Cross-Layer KV Sharing for Efficient LLM Inference
paper-review, with-gpt
19
분
In-context KV-Cache Eviction for LLMs via Attention-Gate
paper-review, with-gpt
16
분
Prompt Compression for Large Language Models: A Survey
paper-review, with-gpt
6
분
Scaling Laws for Neural Language Models
paper-review, with-gpt
17
분
SimLayerKV: A Simple Framework for Layer-Level KV Cache Reduction
paper-review, with-gpt
18
분
Textbooks Are All You Need
paper-review, with-gpt
17
분
Abseil Tip 101 반환 값, 참조 및 수명
cpp, abseil
4
분
Abseil Tip 107 참조 수명 연장
cpp, abseil
3
분
Abseil Tip 135 계약을 테스트하라, 구현을 테스트하지 마라
cpp, abseil
5
분
MagicPIG: LSH Sampling for Efficient LLM Generation
paper-review, with-gpt
26
분
Recycled Attention: Efficient inference for long-context language models
paper-review, with-gpt
22
분
Squeezed Attention: Accelerating Long Context Length LLM Inference
paper-review, with-gpt
24
분
VL-Cache: Sparsity and Modality-Aware KV Cache Compression for Vision-Language Model Inference Acceleration
paper-review, with-gpt
29
분
Abseil Tip 64 Raw 문자열 리터럴
cpp, abseil
3
분
Abseil Tip 77 임시 객체, 이동, 복사
cpp, abseil
5
분
Abseil Tip 86 클래스(enum class)를 활용한 열거형
cpp, abseil
4
분
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
paper-review, with-gpt
15
분
Deep Compression Autoencoder for Efficient High-Resolution Diffusion Models
paper-review, with-gpt
22
분
HART Efficient Visual Generation with Hybrid Autoregressive Transformer
paper-review, with-gpt
21
분
Learning Transferable Visual Models From Natural Language Supervision
paper-review, with-gpt
26
분
The CoT Collection Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning
paper-review, with-gpt
22
분
Abseil Tip 122 테스트 픽스처, 명확성, 그리고 데이터 흐름
cpp, abseil
4
분
Abseil Tip 55 이름 개수 세기와 unique_ptr
cpp, abseil
4
분
Condition-Aware Neural Network for Controlled Image Generation
paper-review, with-gpt
24
분
DistriFusion Distributed Parallel Inference for High-Resolution Diffusion Models
paper-review, with-gpt
32
분
FastComposer Tuning-Free Multi-Subject Image Generation with Localized Attention
paper-review, with-gpt
25
분
VILA On Pre-training for Visual Language Models
paper-review, with-gpt
19
분
VILA-U a Unified Foundation Model Integrating Visual Understanding and Generation
paper-review, with-gpt
23
분
Abseil Tip 1 string_view의 활용 방법과 이점
cpp, abseil
3
분
Batch Calibration Rethinking Calibration for In-Context Learning and Prompt Engineering
paper-review, with-gpt
19
분
LaRS Latent Reasoning Skills for Chain-of-Thought Reasoning
paper-review, with-gpt
25
분
LiteMoE Customizing On-device LLM Serving via Proxy Submodel Tuning
paper-review, with-gpt
37
분
Query-Efficient Correlation Clustering with Noisy Oracle
paper-review, with-gpt
16
분
ShadowKV KV Cache in Shadows for High-Throughput Long-Context LLM Inference
paper-review, with-gpt
18
분
COMET Towards Partical W4A4KV4 LLMs Serving
paper-review, with-gpt
11
분
ELICIT LLM Augmentation via External In-Context Capability
paper-review, with-gpt
21
분
EPIC Efficient Position-Independent Context Caching for Serving Large Language Models
paper-review, with-gpt
11
분
Foundations of Factor Investing
paper-review, with-gpt, finance
11
분
MagicPIG LSH Sampling for Efficient LLM Generation
paper-review, with-gpt
14
분
RAG4ITOps A Supervised Fine-Tunable and Comprehensive RAG Framework for IT Operations and Maintenance
paper-review, with-gpt
15
분
Scientific Beta Multi-Beta Multi-Strategy Indices Implementing Multi-Factor Equity Portfolios with Smart Factor Indices
paper-review, with-gpt, finance
13
분
ALPINE Unveiling the Planning Capability of Autoregressive Learning in Language Models
paper-review, with-gpt
10
분
Can Graph Learning Improve Planning in LLM-based Agents?
paper-review, with-gpt
16
분
Capital asset prices A theory of market equilibrium under conditions of risk
paper-review, with-gpt, finance
10
분
DynamoLLM Designing LLM Inference Clusters for Performance and Energy Efficiency
paper-review, with-gpt
12
분
HYSYNTH Context-Free LLM Approximation for Guiding Program Synthesis
paper-review, with-gpt
11
분
MInference 1.0 Accelerating Pre-filling for Long-Context LLMs via Dynamic Sparse Attention
paper-review, with-gpt
21
분
Portfolio Selection
paper-review, with-gpt, finance
4
분
The Cross-Section of Expected Stock Returns
paper-review, with-gpt, finance
15
분
Efficient Streaming Language Models with Attention Sinks
paper-review, with-gpt
14
분
Enabling Tensor Language Model to Assist in Generating High-Performance Tensor Programs for Deep Learning
paper-review, with-gpt
3
분
Meta Large Language Model Compiler Foundation Models of Compiler Optimization
paper-review, with-gpt
4
분
Model Tells You What to Discard Adaptive KV Cache Compression for LLMs
paper-review, with-gpt
7
분
Transformers are Multi-State RNNs
paper-review, with-gpt
6
분
BUZZ Beehive-structured Sparse KV Cache with Segmented Heavy Hitters for Efficient LLM Inference
paper-review, with-gpt
5
분
CDMPP:ADevice-Model Agnostic Framework for Latency Prediction of Tensor Programs
paper-review, with-gpt
5
분
Don't Look Twice Faster Video Transformers with Run-Length Tokenization
paper-review, with-gpt
3
분
FLUX Fast Software-based Communication Overlap On GPUs Through Kernel Fusion
paper-review, with-gpt
7
분
KVSharer Efficient Inference via Layer-Wise Dissimilar KV Cache Sharing
paper-review, with-gpt
7
분
Efficient Generative LLM Inference Using Phase Splitting
paper-review, with-gpt
10
분
KernelGPT Enhanced Kernel Fuzzing via Large Language Models
paper-review, with-gpt
7
분
Magicoder Empowering Code Generation with OSS-Instruct
paper-review, with-gpt
9
분
Optimal Kernel Orchestration for Tensor Programs with Korch
paper-review, with-gpt
6
분
SpotServe Serving Generative Large Language Models on Preemptible Instances
paper-review, with-gpt
13
분
GraphPipe Improving Performance and Scalability of DNN Training with Graph Pipeline Parallelism
paper-review, with-gpt
15
분
Helix Distributed Serving of Large Language Models via Max-Flow on Heterogeneous GPUs
paper-review, with-gpt
4
분
Memory Bounds for the Experts Problem
paper-review, with-gpt
4
분
Sequoia Scalable, Robust, and Hardware-aware Speculative Decoding
paper-review, with-gpt
8
분
SpecExec Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices
paper-review, with-gpt
7
분
Breaking the Curse of Quality Saturation with User-Centric Ranking
paper-review, with-gpt
3
분
MEGABYTE Predicting Million-byte Sequences with Multiscale Transformers
paper-review, with-gpt
10
분
Reasoning over Public and Private Data in Retrieval-Based Systems
paper-review, with-gpt
5
분
Reasoning over Public and Private Data in Retrieval-Based Systems
paper-review, with-gpt
3
분
FlexGen High-Throughput Generative Inference of Large Language Models with a Single GPU
paper-review, with-gpt
17
분
Quest Query-Aware Sparsity for Efficient Long-Context LLM Inference
paper-review, with-gpt
9
분
Teola Towards End-to-End Optimization of LLM-based Applications
paper-review, with-gpt
9
분
What Matters in Transformers? Not All Attention is Needed Fusion
paper-review, with-gpt
5
분
2024년 10월
6 개의 글
Better & Faster Large Language Models via Multi-token Prediction
paper-review, with-gpt
14
분
CacheBlend Fast Large Language Model Serving for RAG with Cached Knowledge Fusion
paper-review, with-gpt
9
분
Keyformer KV Cache Reduction through Key Tokens Selection for Efficient Generative Inference
paper-review, with-gpt
15
분
Prompt Cache Modular Attention Reuse for Low-Latency Inference
paper-review, with-gpt
6
분
Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve
paper-review, with-gpt
19
분
블로그 다시 시작..
Daily
1
분
2021
24 개의 글
2021년 02월
21 개의 글
LLVM (clang) build and install (ubuntu 18.04)
compiler
1
분
LLVM loop unroll and jam pass and view-cfg
compiler, ML
2
분
간단논문 정리 DARTS DIFFERENTIABLE ARCHITECTURE SEARCH (ICLR 2019)
NAS, ML, paper-review
2
분
간단논문 정리 End-to-End Deep Learning of Optimization Heuristics (PACT 17)
compiler, ML, paper-review
1
분
간단논문 정리 Fast and Effective Orchestration of Compiler Optimizations(Zhelong Pan,Rudolf Eigenmann;Purdue University ;CGO’06)
compiler, ML, paper-review
1
분
간단논문 정리 TVM An Automated End-to-End Optimizing Compiler for Deep Learning (OSDI 18)
compiler, ML, paper-review
1
분
논문 정리 Chameleon Adaptive Code Optimization for Expedited Deep Neural Network Compilation(ICLR 2020)
compiler, ML, paper-review
2
분
논문 정리 LLVM A Compilation Framework for Lifelong Program Analysis & Transformation(CGO 04)
compiler, paper-review
2
분
논문 정리 NeuroVectorizer End-to-End Vectorization with Deep Reinforcement Learning (CGO 20)
compiler, ML, paper-review
1
분
자주쓰는 파이썬 스크립트 패턴
python
1
분
apt-get source 바꾸기
linux
1
분
docker tag 검색하기
linux
1
분
docker 로 gitlab만들기
linux
1
분
inplace_swap
linux
1
분
jupyter notebook 다른python이 실행될 시
linux
1
분
Logitech MX anywhere 2s 우분투에서 제스쳐 사용하기
linux
2
분
ubuntu 16.04 python3 opencv 3.4 설치
linux
1
분
Ubuntu pdf 를 이미지로 변환
linux
1
분
2021년 01월
3 개의 글