No_Afternoon_4260

• HF flagged safetensors as unsafe? wtf?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 5 comments

• Llm wiki by karpathy

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 11 comments

• PSA: DDR5 RDIMM price passed the point were 3090 are less expensive per gb..

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 220 comments

• Roleplay in 2026

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 8 comments

• Quickie: my first week with some sparks

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 2 comments

• On apple is mlx-lm still sota

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 2 comments

• PSA dgx spark ~= rdimm $\go, rtx pro only 4x the price

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 2 comments

• Has anyone tried to saturate a threadripper pro/epyc with pcie 5.0 nvme and see what happens? Theoretically it should have storage bandwidth just under epyc's ram bandwidth

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 6 comments

• Dev on mac, seems promising but what will I miss?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 26 comments

• Best ocr

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 47 comments

• A100 vs rtx pro 6000?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 30 comments

• Anyone on arm?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 2 comments

• Have you tried that new devstral?! Myyy! The next 8x7b?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 48 comments

• Xeon 6 6900, 12mrdimm 8800, amx.. worth it?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 20 comments

• 4 5090 or rtx pro 6000?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 28 comments

• Somebody running kimi locally?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 16 comments

• Level1tech runs deepseek on am5 and it's not that bad!

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 24 comments

• Grok open source

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 19 comments

• Is rocm better supported on arch through a AUR package?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 5 comments

• Help me find this meme of a company that want to implement ia features and become a ia company

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 4 comments

• Amazed by llamacon

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 4 comments

• Who tf is qasar alpha?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 12 comments

• Big moe models => cpu/mac inference?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 2 comments

• Nous Deephermes 24b and 3b are out !

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 55 comments

• Computer vision, vllm and conventional programming

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 2 comments

• Fairydreaming's very instructive post on threadripper 7000 ram bandwidth comparaison

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 6 comments

• Can someone explain what is mixtral 8x7B?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 3 comments

• Thinking is challenging (how to run deepseek and qwq)

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 5 comments

• It's not that mistral 24b is dry, it's parsable and it rocks!

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 33 comments

• Nous DeepHermes-3 8B

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 32 comments

• The light based computer that supports pytorch

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 23 comments

• Voice as fingerprint?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 6 comments

• What all about this deepseek craze?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 6 comments

• Grokking at the Edge of Numerical Stability

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 3 comments

• How come llm space don't have a comfy UI?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 18 comments

• Rig building, immersion cooling

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 11 comments

• 🚶🏃⛹️Smart little people, why small language model

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 14 comments

• Agency swarm

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 0 comments

• Running multimodal quantized

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 8 comments

• Anyone benchmarked these webgpu implementation vs proper backend with nvidia driver?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 0 comments

• Snapdragon X performance?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 8 comments

• Slack's au with confidential data?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 0 comments

• Whisper on SBC?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 4 comments

• What's happening at apple?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 15 comments

• Threadripper 1920x worth it?

  Posted by No_Afternoon_4260@reddit | LocalLLaMA | View on Reddit | 5 comments

• FAI, fully automatic installation with open ssh port and key

  Posted by No_Afternoon_4260@reddit | linux | View on Reddit | 4 comments