been chipping away at https://t.co/npFzBJd3zR - the little free thing i made because i was tired of guessing whether a model would fit my GPU.
since then i mostly just listened and fixed: added EXL2/EXL3, a fine-tune (QLoRA) VRAM mode, a "what can i run on my card" view, and a bunch more GPUs after people pointed out gaps.
still rough, feedback very welcome: https://t.co/BWpq1DqBdA
"Will it even run on my GPU?" β the question every new LLM release brings.
So I built https://t.co/TZnMR4s8RY: a free VRAM fit-calculator + real tokens/sec, with benchmarks & cost for local and API models. No signup.
Check your rig π
https://t.co/mO9NG4LPbZ
#LocalLLM#AI#GPU
Dynamic workflows in Claude Code are now generally available.
For complex tasks like codebase-wide bug hunts, Claude writes its own orchestration and runs subagents in parallel, verifying the work before it reaches you.
Read more: https://t.co/nbNpvkfRBZ
Founders in India, this is for you:
Applications are open for the @GoogleStartups Immersion - powered by @AntlerIndia - a two-phased technical sprint designed for founders with a product in the wild and the ambition to lead the AI-native era.
Four sessions. Two weeks. Hands-on with Google's full AI stack.
Applications close May 22. Find out more and apply now: https://t.co/T3DBqMEpRG
New model drops β "will it even run on my GPU?"
https://t.co/TZnMR4s8RY answers it: a free VRAM fit-calculator + real tokens/sec, benchmarks & cost for local and API LLMs. No signup.
See what your rig can run π
https://t.co/mO9NG4LPbZ
#LocalLLM#AI#GPU
Introducing Claude Fable 5: a Mythos-class model that weβve made safe for general use.
Its capabilities exceed those of any model weβve ever made generally available.
Most businesses lose time managing their online presence in pieces. Site with one provider, automations with another, follow-up somewhere else.
I run all of it from one system. Every client site, every automation, every record. One overview, always online.
One system. No stress.