TheaterFire

Login with Reddit

Currently browsing tags:

  • LocalLLaMA

  • PSA: Guide for Installing Flash Attention 2 on Windows

    Posted by RokHere@reddit | LocalLLaMA | View on Reddit | 2 comments

  • DISTILLATION is so underrated. I spent an hour and got a neat improvement in accuracy while keeping the costs low

    Posted by Ambitious_Anybody855@reddit | LocalLLaMA | View on Reddit | 19 comments

  • University of Hong Kong releases Dream 7B (Diffusion reasoning model). Highest performing open-source diffusion model to date. You can adjust the number of diffusion timesteps for speed vs accuracy

    Posted by jd_3d@reddit | LocalLLaMA | View on Reddit | 114 comments

  • Anyone with experience combining Nvidia system & mac over llama-rpc?

    Posted by segmond@reddit | LocalLLaMA | View on Reddit | 6 comments

  • LMSYS (LMarena.ai) is highly susceptible to manipulation

    Posted by Economy_Apple_4617@reddit | LocalLLaMA | View on Reddit | 0 comments

  • What are the best value, energy-efficient options with 48GB+ VRAM for AI inference?

    Posted by PangurBanTheCat@reddit | LocalLLaMA | View on Reddit | 55 comments

  • The Candle Test - most LLMs fail to generalise at this simple task

    Posted by Everlier@reddit | LocalLLaMA | View on Reddit | 174 comments

  • Just upgraded my RTX 3060 with 192GB of VRAM

    Posted by Wrong_User_Logged@reddit | LocalLLaMA | View on Reddit | 81 comments

  • SGLang. Some problems, but significantly better performance compared to vLLM

    Posted by Sadeghi85@reddit | LocalLLaMA | View on Reddit | 2 comments

  • Real-Time Speech-to-Speech Chatbot: Whisper, Llama 3.1, Kokoro, and Silero VAD 🚀

    Posted by martian7r@reddit | LocalLLaMA | View on Reddit | 31 comments

  • Are there official (from Google) quantized versions of Gemma 3?

    Posted by lostmsu@reddit | LocalLLaMA | View on Reddit | 8 comments

  • MacBook M3, 24GB ram. What's best for LLM engine?

    Posted by Familyinalicante@reddit | LocalLLaMA | View on Reddit | 42 comments

  • MN-GRAND-Gutenburg-Lyra4-Lyra-23.5B - Long Form Output / NON "AI" prose.

    Posted by Dangerous_Fix_5526@reddit | LocalLLaMA | View on Reddit | 28 comments

  • Mac Studio M3 Ultra 512GB DeepSeek V3-0324 IQ2_XXS (2.0625 bpw) llamacpp performance

    Posted by WhereIsYourMind@reddit | LocalLLaMA | View on Reddit | 12 comments

  • Best bang for the buck GPU

    Posted by Ok-Cucumber-7217@reddit | LocalLLaMA | View on Reddit | 85 comments

  • MacBook M4 Max isn't great for LLMs

    Posted by val_in_tech@reddit | LocalLLaMA | View on Reddit | 247 comments

  • Sharing HallOumi-8B, an open-source hallucination detector usable with any LLM!

    Posted by jeremy_oumi@reddit | LocalLLaMA | View on Reddit | 6 comments

  • Promox or Native Ubuntu

    Posted by pipaman@reddit | LocalLLaMA | View on Reddit | 14 comments

  • ClaudePlaysPokemon Open Sourced - Benchmark AI by letting it play Pokémon

    Posted by MaruluVR@reddit | LocalLLaMA | View on Reddit | 8 comments

  • Dual RTX 3060 Setup for Deep Learning – PCIe x8/x4 Concerns?

    Posted by Ahmedsaed26@reddit | LocalLLaMA | View on Reddit | 1 comments

  • PayPal launches remote and local MCP servers

    Posted by init0@reddit | LocalLLaMA | View on Reddit | 5 comments

  • Instructional Writeup: How to Make LLMs Reason Deep and Build Entire Projects

    Posted by No-Mulberry6961@reddit | LocalLLaMA | View on Reddit | 22 comments

  • Is there any work towards an interactive manga translation tool?

    Posted by Tmmrn@reddit | LocalLLaMA | View on Reddit | 2 comments

  • What is the best model for generating images?

    Posted by rez45gt@reddit | LocalLLaMA | View on Reddit | 11 comments

  • LLM amateur with a multi-GPU question. How to optimize for speed?

    Posted by William-Riker@reddit | LocalLLaMA | View on Reddit | 6 comments

  • Smallest model capable of detecting profane/nsfw language?

    Posted by ohcrap___fk@reddit | LocalLLaMA | View on Reddit | 60 comments

  • koboldcpp-1.87.1: Merged Qwen2.5VL support! :)

    Posted by Snail_Inference@reddit | LocalLLaMA | View on Reddit | 4 comments

  • Why isn't the whole industry focusing on online-learning?

    Posted by unraveleverything@reddit | LocalLLaMA | View on Reddit | 15 comments

  • LiveBench team just dropped a leaderboard for coding agent tools

    Posted by ihexx@reddit | LocalLLaMA | View on Reddit | 55 comments

  • Qwen3 will be released in the second week of April

    Posted by AaronFeng47@reddit | LocalLLaMA | View on Reddit | 77 comments

  • Anyone try 5090 yet

    Posted by 4hometnumberonefan@reddit | LocalLLaMA | View on Reddit | 12 comments

  • PAI: your personal AI 100% local inspired by Google's Project Astra

    Posted by Such_Advantage_6949@reddit | LocalLLaMA | View on Reddit | 8 comments

  • Whats the fastest inference API provider (Not OpenAI)

    Posted by Hungry-Connection645@reddit | LocalLLaMA | View on Reddit | 12 comments

  • An idea: an LLM trapped in the past

    Posted by Vehnum@reddit | LocalLLaMA | View on Reddit | 49 comments

  • Nemotron-49B uses 70% less KV cache compare to source Llama-70B

    Posted by Ok_Warning2146@reddit | LocalLLaMA | View on Reddit | 43 comments

  • How do I tell llama-stack where to install models?

    Posted by ImpossibleBritches@reddit | LocalLLaMA | View on Reddit | 1 comments

  • You can now check if your Laptop/ Rig can run a GGUF directly from Hugging Face! 🤗

    Posted by vaibhavs10@reddit | LocalLLaMA | View on Reddit | 64 comments

  • canvas for code and local model

    Posted by sunole123@reddit | LocalLLaMA | View on Reddit | 3 comments

  • Another coding model, Achieves strong performance on software engineering tasks, including 37.2% resolve rate on SWE-Bench Verified.

    Posted by Ornery_Local_6814@reddit | LocalLLaMA | View on Reddit | 18 comments

  • Does the new Jetson Orin Nano Super make sense for a home setup?

    Posted by Initial-Image-1015@reddit | LocalLLaMA | View on Reddit | 84 comments

  • Recitation over Reasoning: How Cutting-Edge Language Models Can Fail on Elementary School-Level Reasoning Problems?

    Posted by ninjasaid13@reddit | LocalLLaMA | View on Reddit | 3 comments

  • Best way to do Multi GPU

    Posted by SalmonSoup15@reddit | LocalLLaMA | View on Reddit | 7 comments

  • How to do citations in local Web Search?

    Posted by tilmx@reddit | LocalLLaMA | View on Reddit | 1 comments

  • How do I tell the 'llama' command where to install models?

    Posted by ImpossibleBritches@reddit | LocalLLaMA | View on Reddit | 1 comments

  • DeepMind will delay sharing research to remain competitive

    Posted by mayalihamur@reddit | LocalLLaMA | View on Reddit | 118 comments

  • I made a Grammarly alternative without clunky UI. Completely free with Gemini Nano (in-browser AI). Helps you with writing emails, articles, social media posts, etc.

    Posted by WordyBug@reddit | LocalLLaMA | View on Reddit | 75 comments

  • Kyutai Labs finally release finetuning code for Moshi - We can now give it any voice we wish!

    Posted by JawGBoi@reddit | LocalLLaMA | View on Reddit | 7 comments

  • SOTA 3d?

    Posted by Charuru@reddit | LocalLLaMA | View on Reddit | 11 comments

  • Thinking about running dual 4060TIs 16gb. But is there a way to limit power on linux? Am I going to sweat myself to death in the summer?

    Posted by LanceThunder@reddit | LocalLLaMA | View on Reddit | 9 comments

  • Top reasoning LLMs failed horribly on USA Math Olympiad (maximum 5% score)

    Posted by Kooky-Somewhere-2883@reddit | LocalLLaMA | View on Reddit | 211 comments

Next