Building Projects Building Game in RN Orbit Mobile App: https://teapawt.net/t.co/bFg6ImURUe Frontend Developer React Native Developer
Check out ๐
Joined February 2018
- Tweets4.8K
- Following4.3K
- Followers1.6K
- Likes18.5K
Mr.Preet - ai/acc retweeted
Aloha! ๐บIntroducing Ornith-1.5, a family of open-source LLMs spanning 9B Dense, 35B MoE, and 397B MoE, trained with self-improving strategies.
It achieves state-of-the-art performance among open-source models of comparable size and delivers performance comparable to Claude Opus 4.8 across reasoning, agentic, and coding tasks:
โ
Terminal-Bench 2.1 (86.1)
โ
SWE-Bench (86 on verified, 65.1 on pro, 79.6 on Multilingual)
โ
DeepSWE (56)
โ
HLE (44.6)
โ
ClawEval (81.4)
โ
Tool Decathlon (71.2)
Ornith-1.5 takes a major step toward training foundation models through end-to-end self-improvement, extending the self-scaffolding strategies introduced in Ornith-1.0 into a more complete self-improvement loop: the model proposes new tasks, generates task-specific scaffolds, and produces solution rollouts for reinforcement learning, continuously creating new learning experiences from which it can improve.
All models, along with their quantized versions (FP8, GGUF, MLX, and NVFP4), have been released under the MIT License, enabling unrestricted commercial and research use.
๐Tech Blog: ornith.ai/ornith_1_5.html
๐คHuggingface: huggingface.co/collections/oโฆ
Mr.Preet - ai/acc retweeted
๐ฅ ๐ฑ๐ผ๐๐๐ฏ-๐ป๐ผ๐๐ฒ ๐ฃ๐ฟ๐ฒ๐๐ถ๐ฒ๐ ๐ถ๐ ๐ณ๐ถ๐ป๐ฎ๐น๐น๐ ๐ผ๐ฝ๐ฒ๐ป-๐๐ผ๐๐ฟ๐ฐ๐ฒ๐ฑ!
I was fortunate to work on its pre-training, so this release feels deeply personal.
dots3-note Preview is the lightest member of the dots3 family, with 280B total parameters and only 16B activated. ๐ฌ๐ฒ๐ ๐ถ๐ ๐ผ๐๐๐ฝ๐ฒ๐ฟ๐ณ๐ผ๐ฟ๐บ๐ ๐๐ฌ๐ฏ, ๐บ๐ฎ๐๐ฐ๐ต๐ฒ๐ ๐ผ๐ฟ ๐ผ๐๐๐ฝ๐ฒ๐ฟ๐ณ๐ผ๐ฟ๐บ๐ ๐๐๐ -๐ฑ.๐ฎ, ๐ฎ๐ป๐ฑ ๐ฟ๐ฒ๐ฎ๐ฐ๐ต๐ฒ๐ ๐๐ต๐ฒ ๐ผ๐๐ฒ๐ฟ๐ฎ๐น๐น ๐ฝ๐ฒ๐ฟ๐ณ๐ผ๐ฟ๐บ๐ฎ๐ป๐ฐ๐ฒ ๐น๐ฒ๐๐ฒ๐น ๐ผ๐ณ ๐๐ต๐ฒ ๐ผ๐ณ๐ณ๐ถ๐ฐ๐ถ๐ฎ๐น ๐๐ฒ๐ฒ๐ฝ๐ฆ๐ฒ๐ฒ๐ธ-๐ฉ๐ฐ-๐๐น๐ฎ๐๐ต ๐ฟ๐ฒ๐น๐ฒ๐ฎ๐๐ฒ.
It also supports a 512K context window and understands text, images, video, and audio.
A model at this scale is shaped by many decisions that never appear in the final model card. Seeing those decisions finally become something that anyone can download, test, and build on is incredibly rewarding.
This is still a preview, and there is plenty to improve. Feel free to try it. Proud of everyone who made this possible. โค๏ธ
๐ Technical blog: studio.dots.ai/dots/dots3-enโฆ
๐ค Model: huggingface.co/dots-studio/dโฆ
๐ป GitHub: github.com/studio-dots-ai/doโฆ
Mr.Preet - ai/acc retweeted
Today, weโre making Voice Agents available to everyone.
With Voice Agents, you can build and scale intelligent, human-like agents that understand context, remember past conversations, and get better at achieving your business outcomes with every interaction.
Our Voice Agents have already powered more than 350 million conversations across enterprise deployments.
Operating at this scale has helped us refine the platform, deepen its capabilities, and build out the complete stack needed to handle the full complexity of voice interactions. The platform brings together the tools and systems to personalise, orchestrate, govern, and launch Voice Agents at scale.
Start building your Voice Agent today: links.sarvam.io/Voiceagents
Mr.Preet - ai/acc retweeted
Introducing VEEDโs Subtitles API.
Worldโs most powerful subtitles now available via API.
โข 100+ languages
โข Highest fidelity
โข 28 style presets
One integration. Endless engagement.
Free @Fal API credits for the first 300 users who retweet and comment โSubsโ.
Mr.Preet - ai/acc retweeted
What we're building at @finz_fun will truly change how creators launch tweet based memecoins
A major portion of total tokens created & volume generated on @pumpdotfun or @bonkfun come from tweets that are simply tokenized
However, these tweetcoin launches are highly competitive & fast paced and generally the first og token created from the tweet gets the most attention from traders
So, speed is the most critical factor for creators/devs when they are tokenizing a tweet
Another problem that we have rn with the current launchpads is devs have very little incentive to launch tokens, launchpads give very small % of rev share to them and so they rely on just profiting via selling or by sniping & bundling to rug the token to make money.
These are the 2 things we are solving with @finz_fun :
- we are building a dedicated launchpad for tokenizing tweet - THE FASTEST - so devs can launch the #1 token for a particular tweet
- they earn a much higher & significant portion of the fees (3x - 4x compared to what the current launchpads offer) so they dont have to just rely on dumping on top of the traders
Creating tokens from tweet is about to get a lot FASTER & more incentivized for devs with @finz_fun - team is cooking so hard.
Launching soon. fade at your own risk
Mr.Preet - ai/acc retweeted
Turn your tweets to tokens with @ TweetToToken - just tag it in your post to tokenize & launch your tweet on @finz_fun
V1 is now live on mainnet
@TweetToToken
Mr.Preet - ai/acc retweeted
Avalanche seems to be playing the right game.
C-chain is an orchestrator of nodes.
Anyone who want their own simple sovereign L1 can deploy and add as many nodes they need for their use-case to validate transactions.
Skipping the centralized sequencers, direct DA on own network and flexible scaling, add validators as the L1 becomes more valuable.
Very overlooked and good learnings here.
The data snowpeer.io/l1s
Mr.Preet - ai/acc retweeted
Join us for Woof Weekly #168 at 1pm ET where we will bork out with @CyberShakti of @Posterdotfun!
teapawt.net/i/spaces/1YpKkByOPZYxj
Mr.Preet - ai/acc retweeted
JUST IN: ๐ฎ๐ณ India to review its stance on cryptocurrency due to global policy changes.
