RSS Reader

Mold 2.0 is out today as a major update to this high performance linker developed by Rui Ueyama. Mold has consistently shown to outperform GNU's Gold and LLVM's LLD linkers while today is making another shift with it now turning to MIT licensing.

Mold previously resorted to AGPL licensing as the linker project struggled with commercial funding. With Mold 2.0, the AGPL code has been relicensed now to being MIT licensed. Rui Ueyama wrote in this morning's v2.0 announcement:

"With

Open Source LLM

Alibaba's cloud unit brings Meta's AI model Llama to its clients - Reuters

Reuters | July 26, 2023, 6:55 a.m. | Original

Alibaba's cloud unit brings Meta's AI model Llama to its clients Reuters

Open Source LLM

LLVMpipe Now Exposes Shared Virtual Memory Support

Phoronix | July 22, 2023, 11:45 a.m. | Original

Mesa's LLVMpipe software driver is now exposing system Shared Virtual Memory (SVM) support with the necessary API bits being in place for the modern Rusticl OpenCL driver as well as the older Clover code. Plus with being a CPU-based driver there isn't any added work or complications around shared virtual memory.

The change to advertise system SVM support was merged this week to Mesa 23.3-devel. SVM is one of the requirements of OpenCL 2.0. Karol Herbst of Red Hat notes that

Open Source LLM

(PR) Qualcomm & Meta Collaborating on Next-gen Snapdragon AI Experiences

TechPowerUp News | July 19, 2023, 12:28 p.m. | Original

Qualcomm Technologies, Inc. and Meta are working to optimize the execution of Meta's Llama 2 large language models directly on-device - without relying on the sole use of cloud services. The ability to run generative AI models like Llama 2 on devices such as smartphones, PCs, VR/AR headsets, and vehicles allows developers to save on cloud costs, and to provide users with private, more reliable, and personalized experiences.

As a result, Qualcomm Technologies plans to make available on-device Llama 2-

Open Source LLM

Meta和微軟聯手發表可免費商用的大型語言模型Llama 2

ITHome | July 18, 2023, 9:07 p.m. | Original

Meta宣佈開源可免費商用大型語言模型Llama 2，並找來微軟當作首批發表合作夥伴，展示雙方於人工智慧領域的擴大合作。用户現在能夠方便地在Azure和Windows上部署Llama 2模型，而這有望使企業開發人工智慧應用的成本與障礙大幅降低。Meta

Open Source LLM

WormGPT Might Become Hackers' New Best Imaginary Friend

Tom's Hardware | July 17, 2023, 3:23 p.m. | Original

(Image credit: InformationAge)

A new, custom-trained version of an LLM (Large Language Model) is making the rounds, but for the worst possible reasons. WormGPT, as it's been christened by its creator, is a new conversational tool -- based on the 2021-released GPT-J language model -- that's been trained and developed with the sole purpose to write and deploy black-hat coding and tools . The promise is that it'll allow its users

Open Source LLM

uutils 0.0.20 Improves GNU Coreutils Compatibility For This Rust-Written Replacement

Phoronix | July 14, 2023, 4:47 p.m. | Original

The uutils project continues advancing as a modern, drop-in replacement to the GNU Coreutils utilities that is written in the Rust programming language.

With today's release of uutils 0.0.20, there continues to be a focus on increasing the GNU test suite compatibility for ensuring uutils can effectively serve as a replacement to GNU Coreutils without any user impact. With uutils 0.0.20 another eight tests are now passing, bringing the total test suite pass rate to 373 while still having

Open Source LLM

Rusticl Capable Of Running Tinygrad For LLaMA Model

Phoronix | July 14, 2023, 6:05 a.m. | Original

Mesa's Rusticl OpenGL implementation written in Rust it turns out can already run the Tinygrad open-source software with its OpenCL back-end for running the LLaMA model.

Tinygrad is the deep learning framework out of Tiny Corp that is run by George Hotz. He's been working on Tinygrad with an emphasis on using the AMD ROCm compute stack but it turns out Mesa's Rusticl OpenCL state tracker for Gallium3D is good enough for Tinygrad too.

Red Hat's David Airlie has begun testing and profiling

Open Source LLM