Interesting. 📱
#damus #yakihonne #nos #nostur #wherostr #primal #nostr #ios
A melhor segunda lĂngua.
https://video.nostr.build/f55880ba8f2eab6dd988b80c82e981f84504551ba67718906c635d8eadc0f23d.mp4
From a reliable source.
That’s because the KJV follows the Textus Receptus (compiled by Erasmus of Rotterdam). Erasmus likely forged a greek manuscript to support a portion of 1 John 5:7-8. The missing verses come from following critical editions of the greek NT. These verses are lacking in the oldest sources.
Don’t write unit tests.
https://video.nostr.build/b94b841c79abf9c4d13c7cd87d7bcf0eeed6ad472dbf5911f50656a7a6ecd7b3.mp4
Oh, I see. By dataset I was thinking of the [WoT filtered] raw data after cleaning/curation and post-processing.
Just a quick update. I’ve been doing some testing and I noticed that I’m getting ActivityPub content from both wss://relay.nostr.band and wss://ditto.pub/relay too.
Yes, you can. These notebooks use smaller models only to take advantage of the Tesla T4 (free tier). You can mod the notebook and use it locally. You can use their bigger models or any other that you want when you feel more comfortable with the different model templates. https://docs.unsloth.ai/get-started/all-our-models
If you don’t mind me giving you a suggestion. An easy way to get started is by using Unsloth’s Google Colab notebooks. Just by inspecting the code of some of their many notebooks you can get a solid starting point about the fine-tunneling steps, including the dataset formats. https://unsloth.ai
Wanna see what Nostr does to AI?
Check out the pics. The red ones are coming from default Llama 3.1. The green ones are after training with Nostr notes.
If you want to download the current version of the LLM:
https://huggingface.co/some1nostr/Nostr-Llama-3.1-8B
The trainings are continuing and current version is far from complete. After more trainings there will be more questions where it flipped its opinion..
These are not my curation, it is coming from a big portion of Nostr (I only did the web of trust filtration).
Isn’t it better to use an uncensored base model for the training? Will you opensource the dataset?
Wanna see what Nostr does to AI?
Check out the pics. The red ones are coming from default Llama 3.1. The green ones are after training with Nostr notes.
If you want to download the current version of the LLM:
https://huggingface.co/some1nostr/Nostr-Llama-3.1-8B
The trainings are continuing and current version is far from complete. After more trainings there will be more questions where it flipped its opinion..
These are not my curation, it is coming from a big portion of Nostr (I only did the web of trust filtration).
Did you use a system prompt in these examples?
đź‘€
nostr:note1c4tkz3qz0scdrzhmq6u89chkpdu0gflfl3e47r4qnd7yux4l2yes33mwja
Oh, that’s right. I saw both on https://mostr.pub but only wss://relay.mostr.pub makes the bridge. 
wss://relay.mostr.pub and wss://ditto.pub/relay
#asknostr is there any open source mobile client that gives more control over the metadata you share with relays? Like showing the raw data, creating new templates, encrypted data as custom event kind, encrypted social graphs, relay policy management…
nostr:npub1xtscya34g58tk0z605fvr788k263gsu6cy9x0mhnm87echrgufzsevkk5s is this expected to happen?
I keep getting *failed to decrypt content* in my dm/requests with dates like 3 weeks to 3 months ago. I got one from nostr:npub1gcxzte5zlkncx26j68ez60fzkvtkm9e0vrwdcvsjakxf9mu9qewqlfnj5z and another one from nostr:npub1r0rs5q2gk0e3dk3nlc7gnu378ec6cnlenqp8a3cjhyzu6f8k5sgs4sq9ac. The dms are gone now. #damus
This is what I’m talking about.
Note ID: nostr:note1mz933ec7kvvcch8vplrulvuxpqc0zj5ltkpgvfvtmcn933854flqds3m7t
#damus 
rStar-Math: Small LLMs Can Master Math Reasoning
with Self-Evolved Deep Thinking
https://arxiv.org/pdf/2501.04519
#llm #slm #ai #math #microsoft #research 
I don't have a purple account and I'm using it right now. Just clone the repository and compile. https://github.com/damus-io/notedeck