BeNN
СтатистикаFrom simple ML algorithms to Neural Networks and Transformers — and from Number Theory to Topology, Cosmology to QED — dive into the world where code meets the cosmos.👨💻🌌 For Any Questions @benasphy
- Последний пост
- 5 июн.
- Последнее чтение
- 13 авг.
- Постов за неделю
- 0
- Всего постов
- 24
- Тип
- открытый
- Язык
- английский
- В каталоге с
- 13 авг.
- 1/24сутки в ленте
- —
- 1/48двое суток
- —
- 1/72трое суток
- —
Оценка по просмотрам недавних постов: пост набирает почти всё за первые сутки.
Посты
I'll be out for long time and wish y'all the best. Looking forward for cool stuff when I come back.
Lol😂
One way or another the AI bubble gonna hit same graph as dotcom Bubble, but that doesn't mean we won't reach our goal.
📊 AI Bubble hits same concentration level that resulted in the bursting of previous bubbles, including the Dot Com. @Fin_Watch
видео или голосовое, без подписи
Lol There are Levels to this game 😂😂
What a Blessed Day! Slot Sacked and We have UCL Final🔥
Tg is also gonna stop P2P in Ethiopia! what other options is left guys👀
Lol😂😂TokenMaxxing at its peak
Remember It's Better to fail at something you Love than to Succeed at something you don't Love!
Unpopular Opinion: Use X instead of YouTube for Tutorial!
Eid al-Adha Mubarak to everyone 🌙
Final Output
The Second project is Cliff Walking and I used Proximal Policy Optimization(PPO) Gymnasium: https://gymnasium.farama.org/environments/toy_text/cliff_walking/ Github: https://github.com/benasphy/OpenAI-s_Gym_library_Project/tree/main/Cliff_Walking
Fun Fact: Incase you have watched Viking Series or Heard about it, Ivar The Boneless and Bjorn IronSide has the same Mother in Real History unlike the series. @BeNN_Pi
What a sad day for Live and city fans😔What an Era it was. CityPool was sth else! Hope it continues that way.
The Final Output
8x8 after tweaking the epsilon greedy parameters and others
8x8 grid and obviously more noise
4x4 Grid output, as you can see there is a less noise and the model is effectively using Q-Learning Table to find optimal path after like 2000 episodes.