30 stories tagged llama, newest first.
308 points by albinstman | 85 comments
Llmspeed is a web tool that simulates the token generation speed of local LLMs on chosen hardware, model, and quantization settings before buying hardware.
14 points by kmavm | 4 comments
Cedana's interactive calculator that converts a token volume and model choice into an estimated number of GPUs (e.g., H100s) needed to serve it.
620 points by root-parent | 577 comments
Financial Times article on Mark Zuckerberg criticizing closed AI rivals as Meta returns to releasing open AI models; the capture shows a Cloudflare security check.
373 points by spankibalt | 337 comments
Variety news article on a publishers' lawsuit, including author Scott Turow, alleging Mark Zuckerberg personally authorized Meta's copyright infringement in training AI models.
459 points by tzury | 34 comments
Sebastian Raschka's gallery of architecture diagrams and fact sheets for open-weight LLMs such as Llama, OLMo, Qwen, and DeepSeek.
1375 points by ricardbejarano | 326 comments
CanIRun.ai is a web tool that estimates which AI models your machine can run locally, using browser APIs to detect hardware and grading each model's expected performance.
344 points by lukeplato | 121 comments
arXiv paper proposing that deep neural networks trained on diverse tasks converge to shared low-dimensional weight subspaces, based on spectral analysis of 1100+ models.
1118 points by gameshot911 | 804 comments
Ars Technica article on unsealed Meta emails in an authors' copyright case alleging Meta torrented and seeded 81.7 TB of pirated books to train AI.
471 points by egnehots | 112 comments
Meta AI blog post announcing quantized versions of its Llama models that run faster and use less memory.
861 points by nmwnmw | 298 comments
Meta AI blog post announcing Llama 3.2, including vision-capable models and lightweight models for edge and mobile devices.
452 points by felarof | 87 comments
Felafax blog post describing how they fine-tuned Llama 3.1 405B on 8 AMD MI300X GPUs using JAX, covering parameter sharding and LoRA.
331 points by rushingcreek | 135 comments
Phind blog post announcing its Phind-405B model and faster, higher-quality AI search answers; the capture shows a Cloudflare verification page.
2258 points by atgctg | 850 comments
Meta newsroom post by Mark Zuckerberg (July 23, 2024) arguing that open source AI, like Linux before it, is the path forward for AI development.
554 points by fuglede_ | 142 comments
Project page for llama.ttf, a font file that embeds a Llama language model and inference engine via HarfBuzz's Wasm shaper.
448 points by albert_e | 233 comments
A post from Meta AI on X announcing the release of Code Llama 70B, a larger version of its code-generation language model.
377 points by dataminer | 78 comments
GitHub repository for PowerInfer, an SJTU-IPADS engine for high-speed large language model inference on PCs with consumer-grade GPUs.
322 points by amrrs | 296 comments
Meta AI blog post announcing Purple Llama, an umbrella project of open trust and safety tools and evaluations for generative AI.
336 points by danielhanchen | 107 comments
GitHub repository for Unsloth, an open-source library for faster, lower-memory QLoRA/LoRA finetuning of Llama models using Triton kernels.
406 points by victormustar | 120 comments
Landing page for LM Studio, a desktop app for Mac and Windows to discover, download, and run local LLMs such as LLaMa and Falcon in GGUF format.
602 points by nickthegreek | 236 comments
Meta newsroom post from August 24, 2023 announcing Code Llama, a large language model for generating and discussing code.
911 points by marcopicentini | 74 comments
Meta AI blog post from August 24, 2023 announcing Code Llama, a large language model for code generation built on Llama 2.
648 points by birriel | 210 comments
Blog post by Finbarr Timbers explaining the math of why llama.cpp can run LLaMA models locally on consumer hardware, focusing on memory bandwidth and compute limits.
444 points by bkitano19 | 39 comments
A blog tutorial by Brian Kitano on implementing a scaled-down Llama language model from scratch trained on TinyShakespeare, with tips for implementing research papers.
339 points by convexstrictly | 77 comments
A tweet by Thomas Wolf announcing that the Falcon 40B large language model is now licensed under Apache 2.0.
564 points by rain1 | 233 comments
A GitHub Gist with instructions for running Meta's Llama 13B model on a 6GB GPU using llama.cpp's CUDA/cuBLAS layer offloading.
414 points by sadiq | 144 comments
GitHub repository for OpenLLaMA, an Apache-2.0 licensed open reproduction of Meta's LLaMA language model, with a 200B-token preview release.
805 points by tim_sw | 202 comments
Together blog post announcing RedPajama, a project to build open-source LLMs that starts by reproducing the LLaMA training dataset of over 1.2 trillion tokens.
699 points by gardenfelder | 425 comments
A TheSequence newsletter issue discussing how the leak of Meta's LLaMA model spurred a wave of open source ChatGPT alternatives.
1182 points by msoad | 356 comments
GitHub pull request #613 in llama.cpp by jart, merged, which uses mmap to make loading model weights 10-100x faster and reduce RAM usage.
399 points by sirteno | 162 comments
Simon Willison blog post speculating that a GPT-3-class language model could be trained for about $85,000 using LLaMA and Alpaca techniques and run in a browser.