GodComplecs

• Question: Llama cpp, whats good right now for: MTP, KV cache quant, Long context.

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

• What is next for local LLM and AI?

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

• MTP Speed with 3090 Qwen 27B Q4

  Posted by GodComplecs@reddit | LocalLLaMA | View on Reddit | 13 comments

• What it feels like to have to have Qwen 3.6 or Gemma 4 running locally

  Posted by GodComplecs@reddit | LocalLLaMA | View on Reddit | 109 comments

• which is faster and better for coding? Luce-Org/Dflash or noonghunna/qwen36-27b-single-3090

  Posted by GodComplecs@reddit | LocalLLaMA | View on Reddit | 25 comments

• Speculative decoding question, 665% speed increase

  Posted by GodComplecs@reddit | LocalLLaMA | View on Reddit | 40 comments

• Stanford: Self improving Meta-Harness

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

• Gemma 4 31B vs Qwen 3.5 27B vs Qwen Coder Next

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

• Stanford: Meta-Harness, self improving harness

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

• BorisCode, Cherny's CC setup for OpenCode

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

• Best multipurpose local model and specific quant

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

• LLM harness for local inference?

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

• Lets talk about models and their problems

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

• LLMs as an tool for for intelligence mimicking systems?

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

• Leading AGI theories?

  Posted by GodComplecs@reddit | LocalLLaMA | View on Reddit | 39 comments

• Local flair?

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

• 600tk/s+ speed on local hardware with Self speculative decoding (rtx 3090)

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

• Gemini 3.1 Pro HIDDEN thought process exposed

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

• Problem with rtx 3090 and MoE models?

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

• V-JEPA, unsupervised video learning

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

• Open source the strongest: "Frontier AI systems have surpassed the self-replicating red line"

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

• AGI confirmed in paper: The Surprising Effectiveness of Test-Time Training for Abstract Reasoning

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

• Thinking in Code is all you need

  Posted by GodComplecs@reddit | LocalLLaMA | View on Reddit | 56 comments

• Dictionary Lookup Process for better LLM answers.

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

• AGI achieved finally! Singularity near!

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

• Solved Strawberry

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