In collaboration with Bud Ecosystem, we just open-sourced the first version of our Audio generation model, Tansen. The current model is trained on 886 hours of speech data, and it supports high-quality speech generation and in-context voice.
#BUD#generativeal
We're thrilled to announce that our latest open-sourced LLM, GenZ 70B, has clinched the #1 spot on the #HuggingFace leaderboard for Instruction-tuned LLMs and holds the #6 position for open #LLMs in all categories. π π #Bud#GenZ
Commercial license and ease of adoption are crucial for businesses experimenting with #LLMs to build innovative solutions. We've created this leaderboard of Business-Friendly LLMs to help entrepreneurs to find the right models for their requirements.
https://t.co/lc4fYHpQZY
Dear connections, we've listed our latest product, Shan AI, on #ProductHunt, and we'd love to hear your feedback, suggestions on the product and to answer any questions you may have. Your feedback can greatly help us make the product better:https://t.co/z2VRThlmFq
We're delighted to see the responses and feedback we got so far for the beta launch of Shan AI. Thank you all for your support; in the coming days, we'll be rolling out updates to the product incorporating the suggestions we've received.
https://t.co/glGV7gmqoj
Meet Shan, Our AI Virtual Business Analyst. Shan can automate business analysis, turning weeks-long analysis works into just an hour's work, saving you time, effort, and money without compromising work quality. Try it out for free.
More details: https://t.co/glGV7gmqoj
Exciting news! π We're open-sourcing 3 GenZ LLM models (free for commercial use) with 87% ChatGPT accuracy. Our models rival the 70B LLaMa-2 Chat model while being 5x smaller. GenZ-V2-Q models cut GPU needs by 40x, perfect for running on your laptop.
https://t.co/eQ2NV3T9qV
We just released a new instruction fine-tuned model with a 4K input length, #GenZ13B. The model is fine-tuned on top of pretrained #LLaMa2 and has achieved a performance of 86% of #GPT 3.5, with a much smaller model of 13 billion parameters.
HuggingFace: https://t.co/QWuvq1CmIz
The model supports an 8K input token length, is trained on the latest information, and does better coding than other LLMs in the same category.
HuggingFace: https://t.co/cat2wk6tTZ
GitHub: https://t.co/DHiHFVtWuW
We are happy to announce the release of our #SOTA (State of the Art) Instruct fine-tuned commercially usable #LLM of its category - Genz. The model achieved SOTA results in its category for HumanEval and GSM8k benchmarks.
#BUD
We just added Hungry Hungry Hippos (H3), to our #LLM leader quadrant. With 24 points, the model's rank is #11
Surprisingly, #H3 surpasses Transformers in terms of perplexity by an impressive margin of 1.0 on the OpenWebText dataset.
Learn more about H3: https://t.co/jbzDzLCdTk
We just added a new model, #Pythia, to our LLM leaderboard. Surprisingly, our analysis shows that the model outperforms many #LLM leaders. With 55 points, the model ranks 5th on our leaderboard.
Explore Pythia: https://t.co/7xPid1RiGW
Leaderboard: https://t.co/d3yXmUQFaJ
We're excited to announce our partnership with @HBAR_foundation, the ecosystem accelerator of the high-speed, secure, eco-friendlyΒ @hedera Β public network, to enable businesses across the globe to launch NFTs using the Hedera network.
We're diving into the world of #GenerativeAI, and have just taken our first step by automating proposal creation!
Check out the Demo & Results: https://t.co/rRQAQihEwC
We're fine-tuning our models to hit the target report in four weeks. Stay tuned!
We're excited to be recognized as The Best Blockchain Enterprise, 2023, at W3 Asia Summit. This award is a testament to our commitment to delivering our customers the best possible products and service
#accubits#w3summit
After steadily declining for a decade, world hunger is on the rise, affecting nearly 10% of people globally.
Can emerging technologies find a solution for this?