PanGu-Σ: Towards Trillion Parameter Language Model with Sparse Heterogeneous Computing
The scaling of large language models has greatly improved natural language understanding, generation, and reasoning. In this work, we develop a system that trained a trillion-parameter language model on a cluster of Ascend 910 AI processors and MindSpore framework, and present the language model with 1.085T parameters named PanGu-Σ. With parameter inherent from PanGu-α, we extend the dense Transfo
実務ã«ãŠã‘ã‚‹RAG 〜å¦ã³ã¨ç¾å ´ã®ãƒŽã‚¦ãƒã‚¦ã€œ | ドクセル
ã‚¹ãƒ©ã‚¤ãƒ‰æ¦‚è¦ å®Ÿå‹™ã«ãŠã„ã¦RAG（Retrieval-Augumented Generation）をãŸã£ã·ã‚ŠçµŒé¨“ã—ãŸhoxo-mã®å¦ã³ã¨ç¾å ´ã®ãƒŽã‚¦ãƒã‚¦ã‚’ã¾ã¨ã‚ãŸ
LLM Inference Series: 4. KV caching, a deeper look
In the previous post, we introduced KV caching, a common optimization of the inference process of LLMs that make compute requirements of the (self-)attention mechanism to scale linearly rather than quadratically in the total sequence length (prompt + generated completions). More concretely, KV caching consists to spare the recomputation of key and value tensors of past tokens at each generation st
1
ランã‚ング
ランã‚ング
メンテナンス
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
Provable unlearning in topic modeling and downstream tasks
UnSTAR: Unlearning with Self-Taught Anti-Sample Reasoning for LLMs
Low Compute Unlearning via Sparse Representations
Controllable Unlearning for Image-to-Image Generative Models via...
ICLR 2025 A Probabilistic Perspective on Unlearning and Alignment for Large Language Models Oral
PanGu-$Ï€$: Enhancing Language Model Architectures via Nonlinearity Compensation
Evaluating World Models with LLM for Decision Making
https://dl.acm.org/doi/abs/10.1145/3664647.3681488
Is Your LLM Secretly a World Model of the Internet? Model-Based Planning for Web Agents
WorldCoder, a Model-Based LLM Agent: Building World Models by Writing Code and Interacting with the Environment
What if we drop the causal mask in auto-regressive Transformer?
Pretraining LLMs - DeepLearning.AI
GitHub - janakiramm/Intel-Xeon-LLM-RAG-Inference-Setup: This repository provides a comprehensive guide to setting up and running a LLM inference server optimized for Intel Xeon machines, with a focus on Retrieval Augmented Generation (RAG). The repository
{{#tags}}- {{label}}
{{/tags}}