Quasar, operating as Subnet 24 on the Bittensor network, has developed a decentralized AI training marketplace that claims to reduce pre-training costs by 99.5% compared to traditional centralized methods. The project focuses on training long-context AI models with context lengths around 2 million tokens, using a novel Quasar Attention method and hybrid architecture. Quasar plans a massive 10-trillion-token training run, far surpassing GPT-3's 300 billion tokens, leveraging a competitive miner reward system to improve model performance. Investors should watch for benchmark results and progress on the ambitious training run as a test case for decentralized AI training viability.
Audiera (BEAT), a decentralized music streaming token, surged 21.33% to $3.48, reaching a market cap of $1.07 billion. The rise follows new partnerships with major artists, boosting investor interest. Other altcoins like Lighter (LIT), Bittensor (TAO...
Bittensor's token TAO is approaching a critical breakout point after weeks of declining selling pressure, forming a falling wedge pattern near $182.5 support. Technical indicators like RSI and MACD suggest weakening bearish momentum, but volume remai...
Bittensor's TAO token has recently fallen more than 2% despite being among the top trending tokens with increased transaction volume. Liquidity analysis shows significant resistance between $220-$255 and support near $185-$190, but the token remains ...
