Zyj

• Unsloth MiniMax M2.7 quants just finished uploading to HF

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 103 comments

• [MIT] RLCR: Teaching AI models to say "I'm not sure"

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 17 comments

• Qwen 3.6 + vLLM + Docker + 2x RTX 3090 setup, working great!

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

• Minimax M2.5 GGUF perform poorly overall

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 36 comments

• Deal on Ryzen 395 w/ 128GB, now 1581€ in Europe

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 78 comments

• Who is doing useful things with local AI and email?

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 10 comments

• MiniMaxAI MiniMax-M2.5 has 230b parameters and 10b active parameters

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 92 comments

• Dual Strix Halo: No Frankenstein setup, no huge power bill, big LLMs

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 61 comments

• SK hynix wins over two-thirds of Nvidia's HBM orders for AI platform this year

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

• A DIY option for the latest beefy LLMs

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

• AMD Medusa Halo, coming after Strix Halo in H1/2026 with (optional) 384bit memory bus

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

• Splitting PCIe 4.0 x16 into 2x PCIe 4.0 x8?

  Posted by Zyj@reddit | buildapc | View on Reddit | 19 comments

• Analysis of (vacation) picture collections w/ local LLM

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

• Improving airflow of Jonsbo UMX3 case

  Posted by Zyj@reddit | buildapc | View on Reddit | 3 comments

• Offering text, images, video, songs, streaming voice chat, speech recognition, all on one AI server?

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

• 4K CPU benchmarks?

  Posted by Zyj@reddit | buildapc | View on Reddit | 4 comments

• The ASRock Radeon RX 7900 XTX Creator

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 45 comments

• Slides from NVIDIA GTC 2025 DGX Spark presentation

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

• FT: Llama 4 w/ voice expected in coming weeks

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 37 comments

• ‘chain of draft’ could cut AI costs by 90%

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

• Contemplating the Radeon 9070 32GB

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 44 comments

• Is anyone working on games?

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

• NVidia APUs for notebooks also just around the corner (May 2025 release!)

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

• Which workstation for 3x 3-slot GPUs?

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

• Open source language learning project with local LLM?

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

• Incorporating a fileserver?

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

• Turn PCIe 5.0 x16 into 2x PCI 4.0 x16?

  Posted by Zyj@reddit | buildapc | View on Reddit | 2 comments

• Watercooling RTX3090s

  Posted by Zyj@reddit | LocalLLaMA | View on Reddit | 27 comments