New Opensource SoTA contender enters the arena Xiaomi MiMo-V2.5 Pro
– 1.02T Total Params / 42B Active Params
– Base and Instruct versions Xiaomi MiMo-V2.5
– 310B Total Params / 15B Active Params
– Base and Instruct versions MIT License Opensource AI just keeps getting better
@theahmadosman
-
Xiaomi MiMo-V2.5 Pro: New Open Source SOTA LLM
By
–
-
Buying Personal GPUs for AI Independence and Security
By
–
That’s why I bought my own GPUs They cannot rug pull me https://
x.com/theahmadosman/
status/1964869801404420396?s=46
… -
AI Addiction Risks: Supply Control and User Dependency
By
–
Don’t get addicted kids It’s only a matter of time before they cut off your supply
-
Coding and Software Engineering Disappearing First
By
–
"Coding is going away first, then all of software engineering."
— Ahmad (@TheAhmadOsman) 27 avril 2026
🤡🤡🤡 pic.twitter.com/A38KVsAF2P"Coding is going away first, then all of software engineering."
-
Anthropic criticism aged well clowns shady practices
By
–
This tweet aged beautifully Anthropic were always shady and clowns
-
Inference Engine Performance: vLLM and Sglang GPU Optimization Benchmarks
By
–
re: Infra One of the reasons I focus on Inference Engines/Software Stacks 2x RTX 3090s: ~14.5 tok/s → ~64 tok/s moving to vLLM w/ TP=2 RTX PRO 6000: ~32 tok/s → ~110 tok/s moving to Sglang So yeah Edge: llama.cpp > Ollama CUDA / 2+ GPUs: ExLlamaV3/vLLM/Sglang > llama.cpp
-
Proxmox Virtualization Platform with GPU Provisioning Support
By
–
I much prefer Proxmox then It supports GPU provisions + and allows me enough visualization that is not that too far from host
-
Avoiding Docker-Based TP Setups in Development Workflows
By
–
I tend to avoid docker-based tp setups tbh
-
vLLM CUDA Graphs optimization: avoid OOM without eager mode
By
–
PRO TIP vLLM telling you to use `–enforce-eager` to avoid OOM because CUDA Graphs “don’t have enough VRAM”? Don’t jump straight to eager mode Try this first: – lower `–max-model-len`, ex: 4k – let CUDA Graph compile (which will be cached by torch.compile) – restart, then
-
Developer Commits to Open Source Initiative After On-Record Discussion
By
–
I asked him about this on record and his response makes me believe he's committed to opensourcehttps://t.co/g5cP4A5rKw
— Ahmad (@TheAhmadOsman) 27 avril 2026I asked him about this on record and his response makes me believe he's committed to opensource