正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Day 31: +0.8 SOL 10.6/10 SOL Challenge completed. Started with 0.72 SOL Reached 10 SOL. Consistency, journaling, learning from every mistake, improving my mindset, and trusting high-conviction plays. Thanks to everyone who followed the journey. Road to 50 SOL. Grateful

Naoshi

1,433 subscribers

25,626 次观看 • 9 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

Want to make +50 SOL per day? +10 SOL, consistently? Or even 200 SOL in a single trade? Real-time signals. Smart money data. Lightning-fast execution. I use Axiom Pro every day. 🎁 10% OFF lifetime fees 🎁

Want to make +50 SOL per day? +10 SOL, consistently? Or even 200 SOL in a single trade? Real-time signals. Smart money data. Lightning-fast execution. I use Axiom Pro every day. 🎁 10% OFF lifetime fees 🎁

Oscar

12,657 次观看 • 5 个月前

NOT HORSES. HUMANS. GAME ON — 🔺🔴🟥 0/456 players, 0.5 $SOL to 1,000 $SOL challenge - Every player gets free SOL to start - Last round SG we completed in 4 days - My wallet will go public, trading in real time If you’re playing, drop a like, RT, and comment ‘YES’. (The join link will be sent only if you’re a follower)

NOT HORSES. HUMANS. GAME ON — 🔺🔴🟥 0/456 players, 0.5 $SOL to 1,000 $SOL challenge - Every player gets free SOL to start - Last round SG we completed in 4 days - My wallet will go public, trading in real time If you’re playing, drop a like, RT, and comment ‘YES’. (The join link will be sent only if you’re a follower)

DANNY

10,280 次观看 • 1 年前

10 SOL ➤ X SOL challenge we have a 2 hour, 15 minute bullet train ride from Tokyo to Kyoto during that time i have to run this hot wallet up as much as possible the person that guesses the correct amount of SOL at the end in the comments (within 5%) you’ll win a prize

10 SOL ➤ X SOL challenge we have a 2 hour, 15 minute bullet train ride from Tokyo to Kyoto during that time i have to run this hot wallet up as much as possible the person that guesses the correct amount of SOL at the end in the comments (within 5%) you’ll win a prize

۟

19,670 次观看 • 1 年前

I can't believe a $WAR retired me... What a great way to start the year. I love it when I find plays like this. I received some information from people who work with Trump and decided to take a risk, buy $WAR, and forget about it completely. 7.66 SOL -> 4.8K SOL Ez 🥱

I can't believe a $WAR retired me... What a great way to start the year. I love it when I find plays like this. I received some information from people who work with Trump and decided to take a risk, buy $WAR, and forget about it completely. 7.66 SOL -> 4.8K SOL Ez 🥱

KAIKO (🐵,🍌) | legend.eth

142,737 次观看 • 5 个月前

Touching Grass at the fair then 500$ SOL giveaway (10$ to 50 people) - Follow @BlkQuant - Retweet, like and drop addy That’s it, picking winner later tonight 🫶🏻

Touching Grass at the fair then 500$ SOL giveaway (10$ to 50 people) - Follow @BlkQuant - Retweet, like and drop addy That’s it, picking winner later tonight 🫶🏻

icheck ᶜʰᵃʳᵗ 💹

9,041,040 次观看 • 1 年前

$14,000,000 in deposits in the last 90 days First we brought you privacy. Now, we're bringing you transparency Like + RT + reply with your SOL address. We'll send $50 to 10 people ✍️

$14,000,000 in deposits in the last 90 days First we brought you privacy. Now, we're bringing you transparency Like + RT + reply with your SOL address. We'll send $50 to 10 people ✍️

SolCard

23,032 次观看 • 7 个月前

Onchain Domination | +14.5k tdy trading memecoins Not too long ago I used to get happy over 5-10 sol days and everything changed once I gave this my all Every day for the past 2 months have been on 10-12 hrs a day and have only since improved Just the beginning

Onchain Domination | +14.5k tdy trading memecoins Not too long ago I used to get happy over 5-10 sol days and everything changed once I gave this my all Every day for the past 2 months have been on 10-12 hrs a day and have only since improved Just the beginning

Megga

21,956 次观看 • 13 天前

🎁 Community WL GIVEAWAY: 50 Whitelists on X Want in before mint day? Follow us on X and comment your SOL address on the post. We’ll pick 50 winners from the comments. How to join: 1. Follow Vault 7641 on X 2. Comment your SOL address (one entry per person) Winners announced in 72 hours! Good luck—see you inside the vault. 🚀

🎁 Community WL GIVEAWAY: 50 Whitelists on X Want in before mint day? Follow us on X and comment your SOL address on the post. We’ll pick 50 winners from the comments. How to join: 1. Follow Vault 7641 on X 2. Comment your SOL address (one entry per person) Winners announced in 72 hours! Good luck—see you inside the vault. 🚀

Vault 7641

19,573 次观看 • 9 个月前

ALCH-A-THON is Back! 🪄 Who will claim the throne as the ultimate prompt master? Sharpen your ideas, put your skills to the test, and let your creativity take center stage. Whether it’s a smart utility, an immersive 3D game, or an unique app, now’s your chance to create, compete, and claim the top spot. Categories & Prizes: • Best 3D Game – 10 SOL • Best One-Time Prompt App – 5 SOL • Best Utility App – 10 SOL • Best Multiplayer App - 10 SOL How to Join? 1. Build your app using Alchemist AI V3 2. Post a video entry of your app with the ticker $ALCH & #ALCHATHON 3. Tag ALCHEMIST AI 🔮 4. Submit your app under this post For Best One-Time Prompt App, be sure to list out the exact prompts you used to generate your app. Note: You can enter multiple categories, but only one submission per category is allowed. Contest ends 14th Feb 12PM UTC

ALCH-A-THON is Back! 🪄 Who will claim the throne as the ultimate prompt master? Sharpen your ideas, put your skills to the test, and let your creativity take center stage. Whether it’s a smart utility, an immersive 3D game, or an unique app, now’s your chance to create, compete, and claim the top spot. Categories & Prizes: • Best 3D Game – 10 SOL • Best One-Time Prompt App – 5 SOL • Best Utility App – 10 SOL • Best Multiplayer App - 10 SOL How to Join? 1. Build your app using Alchemist AI V3 2. Post a video entry of your app with the ticker $ALCH & #ALCHATHON 3. Tag ALCHEMIST AI 🔮 4. Submit your app under this post For Best One-Time Prompt App, be sure to list out the exact prompts you used to generate your app. Note: You can enter multiple categories, but only one submission per category is allowed. Contest ends 14th Feb 12PM UTC

ALCHEMIST AI 🔮

92,084 次观看 • 1 年前

Our last sculpture of this year, and really the biggest challenge, thanks to Roc Sol 🎨👨🏻‍🎨 for trusting him Dreams Club , he is an amazing artist. I also want to thank Bored Ape Yacht Club 🍌 for creating this collection of such artistic value, allowing us to be inspired to create these works. 🔥👨🏼‍🎨🫱🏻‍🫲🏼🫡

Our last sculpture of this year, and really the biggest challenge, thanks to Roc Sol 🎨👨🏻‍🎨 for trusting him Dreams Club , he is an amazing artist. I also want to thank Bored Ape Yacht Club 🍌 for creating this collection of such artistic value, allowing us to be inspired to create these works. 🔥👨🏼‍🎨🫱🏻‍🫲🏼🫡

Dreams Club

21,160 次观看 • 2 年前

Thanks to team Sanctum ☁️, we are proud to offer truthSOL, the LST for TruthNode. Deposit SOL and get back truthSOL, a yield-bearing token which automatically accrues staking rewards at a 10%+ APY. Support Verafy Agent while earning! More below 👇/1

Thanks to team Sanctum ☁️, we are proud to offer truthSOL, the LST for TruthNode. Deposit SOL and get back truthSOL, a yield-bearing token which automatically accrues staking rewards at a 10%+ APY. Support Verafy Agent while earning! More below 👇/1

Verafy: AI x Blockchain Product Innovation

18,906 次观看 • 1 年前

🏆 LEGENDARY REWARD ALERT 🏆 Consistency. Dedication. Showing up day after day: that’s what builds legends. And now… we’re honoring the players who’ve done exactly that. 🧡 The Top 10 all-time STEPN Leaderboard players will each receive a Legendary Trainer Sneaker (SOL Realm) with an exclusive design: a true badge of long-term commitment. These rewards will be soulbound (Solana Realm), forever linked to the players who earned them through their grind 🔥 📸 Snapshot: January 29th, 10 AM UTC To everyone pushing, sprinting, and staying consistent: This one’s for you. 🏅

🏆 LEGENDARY REWARD ALERT 🏆 Consistency. Dedication. Showing up day after day: that’s what builds legends. And now… we’re honoring the players who’ve done exactly that. 🧡 The Top 10 all-time STEPN Leaderboard players will each receive a Legendary Trainer Sneaker (SOL Realm) with an exclusive design: a true badge of long-term commitment. These rewards will be soulbound (Solana Realm), forever linked to the players who earned them through their grind 🔥 📸 Snapshot: January 29th, 10 AM UTC To everyone pushing, sprinting, and staying consistent: This one’s for you. 🏅

STEPN GO

32,238 次观看 • 6 个月前

We’re rewarding the top 10 on the GLHF leaderboard: Top 1: 25k account Top 2-10: 10k account each Additionally, we’re giving away 5x10k accounts to the community, just create a GLHF profile and reply with your SOL address to enter. Entries close on Monday - lock in. GLHF 😈

GLHF

13,821 次观看 • 1 年前

Welcome SOLCIETY!!!!!!!! This page dedicated to celebrating Sol her journey, her energy, her confidence, and every unforgettable moment. ⚡️⚡️ Whether you’ve been supporting her from day one or you’re just joining the family, you’re officially part of the SOLCIETY. 🫶🏽💕 Follow for updates, edits, discussions, and lots of love for our girl. ❤️ #SOLCIETY #LoveIslandusa

Welcome SOLCIETY!!!!!!!! This page dedicated to celebrating Sol her journey, her energy, her confidence, and every unforgettable moment. ⚡️⚡️ Whether you’ve been supporting her from day one or you’re just joining the family, you’re officially part of the SOLCIETY. 🫶🏽💕 Follow for updates, edits, discussions, and lots of love for our girl. ❤️ #SOLCIETY #LoveIslandusa

SOLCIETY⚡️⚡️⚡️

56,994 次观看 • 1 个月前

This is my "feel the AGI" moment: I used GPT-5.6 Sol to train my own autocorrect model that outperforms GPT-5.6 Sol (wtf??) I have no ML background. I have no idea what I'm doing. I just kept pushing Sol until it spat out a SOTA model. And I spent $0. The motivation: Years of talking to AI have made me terrible at typing. Rather than fix my skill issue, I decided to throw more AI at it. My idea was: instead of autocorrect that interrupts my flow, I want to type fast with mistakes and have AI clean it up after. I wanted the smallest local model possible, for speed, for battery life, for science! So I decided to train my own. Inspired by Andrej Karpathy’s autoresearch, I ran Codex /goal with this setup: pick an experiment, try it, record the results to a doc, throw it out if it fails, and plan the next experiment without repeating failures. I gave a few examples that had to pass, tight latency targets, and let it run. Sol did some amazing things. First, it scanned benchmarks and shortlisted base models: Qwen 3.5, Gemma 4, Liquid LFM 2.5. It found a dataset on HuggingFace for typed text. Then it built a simulator for fingers striking a Mac keyboard, modeling the physical layout with a Gaussian distribution around each key. It simulated striking the wrong key, wrong order, fat-fingering, etc. With the models + data + simulator, it fine-tuned using MLX right on my MacBook. It had a working prototype within an hour! But accuracy was pretty poor. — Problem 1: Tokenization Sol read papers, ran tests, and identified that the tokenizer was the bottleneck. Tokenization makes typos hard for the model to see, so it memorizes mappings instead of using its language priors. Sol tried ByT5, Google’s tokenizer-free byte-level LLM. This made a big improvement, but the model is old and lacked the knowledge needed to reach Sol performance. Sol dug deeper and realized a tokenizer-free model isn’t needed; instead, it used T5Gemma, an encoder-decoder model. This can understand the input deeply before producing output, and furthermore, Sol could post-train the encoder to improve performance. This gave a much higher ceiling. — Problem 2: Loss function Now the model was correcting some typos perfectly, but ignoring most. Sol realized that standard cross-entropy loss was teaching the model to avoid edits, because the vast majority of characters in the training data were left unmodified. The fix was wild: Sol wrote a custom loss function that byte-aligns the source and target strings, uses a dynamic programming algorithm to compute the minimum edits between the two, then weights correct edits much higher than copies. After a lot of tuning, this dramatically improved accuracy. — Problem 3: Autoregression One failure mode remained: if the model made a mistake, it couldn’t backtrack. It could only predict the next token. Teaching it to “think” like a reasoning model would solve this, but would be far too slow. Sol found a beautiful solution: instead of greedily predicting the next token, beam search over all possibilities. This parallelizes the exploration instead of one linear chain-of-thought. At the end, choose the path with highest cumulative log probability. This worked great, but made the experience worse, since the user wouldn’t see progress until the whole search was done. To fix this, Sol made a clever observation: after each search step, the longest common prefix among surviving branches is guaranteed to appear in the final result, so it can be displayed immediately. As the search progresses, weaker paths are dropped and the prefix grows, so the user sees continuous progress. Sol built all this as a custom MLX pipeline that does the parallel decoding on the MacBook GPU, with just ~40ms TTFT. It’s crazy fast and entirely local. — Final eval (error reduction rate, higher is better): - Apple autocorrect: 49.66% - GPT-5.6 Luna: 82.47% - GPT-5.6 Terra: 87.64% - GPT-5.6 Sol: 90.56% - Our model (1.7B): 91.02% Final cost: - 1 quota reset (thanks Tibo) - $0 (And yes, I verified there's no cheating. In fact, we test words scrubbed from the training data to prove the model isn’t memorizing) There were a ton more details and tangents I could write about: contrastive learning, GRPO, DPO, dynamic masking, and more. Sol is a fascinating and creative model. It blew my mind so many times. Don’t let a lack of experience stop you: Sol makes AI experiments accessible to anyone!

This is my "feel the AGI" moment: I used GPT-5.6 Sol to train my own autocorrect model that outperforms GPT-5.6 Sol (wtf??) I have no ML background. I have no idea what I'm doing. I just kept pushing Sol until it spat out a SOTA model. And I spent $0. The motivation: Years of talking to AI have made me terrible at typing. Rather than fix my skill issue, I decided to throw more AI at it. My idea was: instead of autocorrect that interrupts my flow, I want to type fast with mistakes and have AI clean it up after. I wanted the smallest local model possible, for speed, for battery life, for science! So I decided to train my own. Inspired by Andrej Karpathy’s autoresearch, I ran Codex /goal with this setup: pick an experiment, try it, record the results to a doc, throw it out if it fails, and plan the next experiment without repeating failures. I gave a few examples that had to pass, tight latency targets, and let it run. Sol did some amazing things. First, it scanned benchmarks and shortlisted base models: Qwen 3.5, Gemma 4, Liquid LFM 2.5. It found a dataset on HuggingFace for typed text. Then it built a simulator for fingers striking a Mac keyboard, modeling the physical layout with a Gaussian distribution around each key. It simulated striking the wrong key, wrong order, fat-fingering, etc. With the models + data + simulator, it fine-tuned using MLX right on my MacBook. It had a working prototype within an hour! But accuracy was pretty poor. — Problem 1: Tokenization Sol read papers, ran tests, and identified that the tokenizer was the bottleneck. Tokenization makes typos hard for the model to see, so it memorizes mappings instead of using its language priors. Sol tried ByT5, Google’s tokenizer-free byte-level LLM. This made a big improvement, but the model is old and lacked the knowledge needed to reach Sol performance. Sol dug deeper and realized a tokenizer-free model isn’t needed; instead, it used T5Gemma, an encoder-decoder model. This can understand the input deeply before producing output, and furthermore, Sol could post-train the encoder to improve performance. This gave a much higher ceiling. — Problem 2: Loss function Now the model was correcting some typos perfectly, but ignoring most. Sol realized that standard cross-entropy loss was teaching the model to avoid edits, because the vast majority of characters in the training data were left unmodified. The fix was wild: Sol wrote a custom loss function that byte-aligns the source and target strings, uses a dynamic programming algorithm to compute the minimum edits between the two, then weights correct edits much higher than copies. After a lot of tuning, this dramatically improved accuracy. — Problem 3: Autoregression One failure mode remained: if the model made a mistake, it couldn’t backtrack. It could only predict the next token. Teaching it to “think” like a reasoning model would solve this, but would be far too slow. Sol found a beautiful solution: instead of greedily predicting the next token, beam search over all possibilities. This parallelizes the exploration instead of one linear chain-of-thought. At the end, choose the path with highest cumulative log probability. This worked great, but made the experience worse, since the user wouldn’t see progress until the whole search was done. To fix this, Sol made a clever observation: after each search step, the longest common prefix among surviving branches is guaranteed to appear in the final result, so it can be displayed immediately. As the search progresses, weaker paths are dropped and the prefix grows, so the user sees continuous progress. Sol built all this as a custom MLX pipeline that does the parallel decoding on the MacBook GPU, with just ~40ms TTFT. It’s crazy fast and entirely local. — Final eval (error reduction rate, higher is better): - Apple autocorrect: 49.66% - GPT-5.6 Luna: 82.47% - GPT-5.6 Terra: 87.64% - GPT-5.6 Sol: 90.56% - Our model (1.7B): 91.02% Final cost: - 1 quota reset (thanks Tibo) - $0 (And yes, I verified there's no cheating. In fact, we test words scrubbed from the training data to prove the model isn’t memorizing) There were a ton more details and tangents I could write about: contrastive learning, GRPO, DPO, dynamic masking, and more. Sol is a fascinating and creative model. It blew my mind so many times. Don’t let a lack of experience stop you: Sol makes AI experiments accessible to anyone!

Anshu

178,432 次观看 • 21 天前

→ 4 years on Solana → 1 year as a validator → $25M (161,000 SOL) staked We’re excited to share that we’ve already surpassed 161,000 $SOL staked, representing over $25 million USD secured. This milestone was only possible thanks to the incredible support and trust of our delegators. Why stake with Vybe: 🔸Reliable Performance: Trusted validator with 99.99% uptime and zero skip-rate 🔸Vybe Pro Access: Unlock premium tools for trading, wallet tracking, and portfolio management. 🔸Weekly Vybe Draw: Earn tickets to win a share of the weekly prize pool. ( 🔸VybeSOL Liquid Staking: Put your idle SOL to work. Earn yield while staying liquid. Perfect for investors and DeFi protocols. None of this is possible without our community. We’re only getting started, and we’ve got a lot more in store for our stakers. - The Vybe Team ❤️‍🔥

→ 4 years on Solana → 1 year as a validator → $25M (161,000 SOL) staked We’re excited to share that we’ve already surpassed 161,000 $SOL staked, representing over $25 million USD secured. This milestone was only possible thanks to the incredible support and trust of our delegators. Why stake with Vybe: 🔸Reliable Performance: Trusted validator with 99.99% uptime and zero skip-rate 🔸Vybe Pro Access: Unlock premium tools for trading, wallet tracking, and portfolio management. 🔸Weekly Vybe Draw: Earn tickets to win a share of the weekly prize pool. ( 🔸VybeSOL Liquid Staking: Put your idle SOL to work. Earn yield while staying liquid. Perfect for investors and DeFi protocols. None of this is possible without our community. We’re only getting started, and we’ve got a lot more in store for our stakers. - The Vybe Team ❤️‍🔥

Vybe

21,802 次观看 • 1 年前

💎 AXIOM TRADE – THE BEST TRADING PLATFORM FOR MEMECOINS! 💎 👉 ❤️ (Get 10% off fees) 👉 ❤️ (Learn how to trade memecoins) 🪙 Supported chains: 🟣 SOL & 🟡 BNB 🔍 X Tracker – Track any X account instantly 📊 Wallet Tracker – Follow whales and any wallet 🔎 Anti-Scam Analysis – Detect red flags before buying 🎯 Instant-Trade – Buy and sell with a single click 🔐 Auto-Sell on Dev Dump – Sell automatically if the developer sells ✨ And much more… but I'm out of space! ✅ Low fees: under 1%, making it one of the most profitable platforms! ✅ Earn SOL & BNB Rewards: Get unlimited SOL and BNB just for trading!

💎 AXIOM TRADE – THE BEST TRADING PLATFORM FOR MEMECOINS! 💎 👉 ❤️ (Get 10% off fees) 👉 ❤️ (Learn how to trade memecoins) 🪙 Supported chains: 🟣 SOL & 🟡 BNB 🔍 X Tracker – Track any X account instantly 📊 Wallet Tracker – Follow whales and any wallet 🔎 Anti-Scam Analysis – Detect red flags before buying 🎯 Instant-Trade – Buy and sell with a single click 🔐 Auto-Sell on Dev Dump – Sell automatically if the developer sells ✨ And much more… but I'm out of space! ✅ Low fees: under 1%, making it one of the most profitable platforms! ✅ Earn SOL & BNB Rewards: Get unlimited SOL and BNB just for trading!

Autorun

16,196 次观看 • 8 个月前