Pinned
Knut Jägersberg
147.4K posts
Knut Jägersberg
@JagersbergKnut
- Insanely Fast Whisper Transcribe 300 minutes (5 hours) of audio in less than 98 seconds
- you can download the internet. huggingface.co/datasets/infor…
- Self-evolving Agents with reflective and memory-augmented abilities
- LLMs Do Not Think Step-by-step In Implicit Reasoning arxiv.org/abs/2411.15862
- declare-lab/flan-alpaca-xl Base model: flan-t5, thus no license probs
- Replying to @tsarnickweirdly, I think it is existing celebrities. Sure you can make AI influencers, and they can and will compete, but even with interesting personality, the brands of at least some existing people remain valuable. People are interested in people, even with superinteresting AI around.
- WizardLM: An Instruction-following LLM Using Evol-Instruct uuh "WizardLM-7B outperforms ChatGPT in the high-complexity instructions... Evol-Instruct is a novel method using LLMs instead of humans to automatically mass-produce open-domain instructions"
- TheBloke/galpaca-30B-GPTQ-4bit-128g Tom Jobbins had the kindness to quantize galpaca 30b, it fits in 18gb of vram.
- GeorgiaTechResearchInstitute/galpaca-30b Please somebody 4int this!
- llmware/dragon-mistral-7b-v0 A RAG model
- TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
- Mechanics of Next Token Prediction with Self-Attention This was a key paper. There are more of those. arxiv.org/abs/2403.08081
