Open Source LLM

AMD Releases ROCm 6.2.1 With FBGEMM Support, Installation Improvements

Following the AMD ROCm 6.2 release from early August, ROCm 6.2.1 was released on Friday evening as the first point release to that series for this AMD GPU compute stack for Linux systems.

ROCm 6.2.1 rolls out rocAL 2.0 with many new features, support for Facebook General Matrix Multiplication "FBGEMM" and related FBGEMM code, ROCm offline installer improvements, documentation improvements, and various fixes.

For the most part ROCm 6.2.1 is a relatively small release unless
Continue reading...

Open Source LLM|Intel

OpenVINO 2024.4 Prepares For Core Ultra Series 2, New Gen AI Models

Intel engineers today released OpenVINO 2024.4 as the newest version of their open-source AI toolkit. OpenVINO 2024.4 prepares for upcoming Intel Core Ultra Series 2 "Lunar Lake" processors, supports newer Gen AI models, now supports Python 3.12, and finally adds official support for Red Hat Enterprise Linux 9.

OpenVINO 2024.4 adds support for the GLM-4-9B Chat, MiniCPM-1B, Llama 3 and 3.1
Continue reading...

Open Source LLM

Alibaba accelerates AI push by releasing new open-source models, text-to-video - Reuters

Alibaba accelerates AI push by releasing new open-source models, text-to-video Reuters

Open Source LLM

LLVM 19.1 Released With C++17 Support "Complete", More C & C++ Features

LLVM 19.1 is out today as the first stable version of the LLVM 19 compiler stack including the Clang 19 C/C++ compiler.

There's a lot on the table with this six-month update to LLVM. Some of the LLVM 19 release highlights include:

- C++17 support is now considered complete with remaining feature support addressed.

- Various additional C++20 and C++23 features have been implemented and even a few C+

Open Source LLM

Google公佈解決幻覺的AI模型DataGemma

AI模型以產生幻覺及誤導性的答案為人詬病,Google上週 公開 最新模型DataGemma,號稱大幅降低幻覺問題。

Google表示,DataGemma是第一個將大型語言模型(LLM)和來自Google開源知識圖譜(knowledge graph)Data Commons的各種現實世界資料連結的案例,目的在藉由以現實


Open Source LLM

OpenAI debuts "o1" AI models, promising PhD-level reasoning in science and math

Serving tech enthusiasts for over 25 years.TechSpot means tech analysis and advice you can trust .

In a nutshell: OpenAI has unveiled a new series of AI language models named the "o1," specifically engineered to enhance reasoning capabilities, particularly for complex issues in science, coding, and mathematics. The company is so confident in these advancements that it has reset the model version counter to 1, starting anew after GPT-4o, and has notably moved away from the GPT branding

Continue reading...

Open Source LLM

OpenAI o1 具推理能力 AI 模型推出 收費版馬上有得試 + 將來免費版都有得用

OpenAI 日前正式發布旗下首個具有「推理」能力的新模型並將其命名為 o1,它經過訓練後能比人類更快地回答複雜的問題。Open AI 除了推出 o1-preview 外,還推出了一個性能較低且較便宜的 o1-mini 並計劃免費向所有用戶開放使用,但

Continue reading...

Open Source LLM

OpenAI launches new series of AI models with 'reasoning' abilities - Reuters

OpenAI launches new series of AI models with 'reasoning' abilities Reuters

Open Source LLM

Google NotebookLM加入Audio Overview,將用户筆記變成Podcast

Google週四(9/11) 宣佈 AI筆記應用程式NotebookLM加入語音概要(Audio Overview)功能,能將用户的筆記變成語音版小型討論會。

NotebookLM是2023年在Project Tailwind專案下上線,可幫助用户理解複雜資訊。用户上傳筆記後,它能根據用户素材、引用的資訊


Open Source LLM

LLVM Makes Progress On Using ClangIR To Compile GPU Kernels

ClangIR is a new IR for LLVM's Clang compiler built atop MLIR. Thanks to this year's Google Summer of Code, there has been progress on being able to compile GPU kernels using ClangIR as another improvement for heterogeneous programming with this open-source compiler stack.

Student developer Zhi Ma took to the effort of enabling GPU kernel support with ClangIR over the summer. The project was centered around enabling OpenCL C language support in ClangIR that can then be targeted to SPIR-V for executing atop OpenCL/Vulkan drivers