GONDARCENGIZ Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 (düzenlendi) şöyle bir thread var, millet niye büyük olay yarattı diye düşünürken okuduktan sonra anlaşıldı. günün sonunda gaming gpu'lara çökmezler inş dfgdfg 2021 gibi olursa sıçtık mk. (bilgi sahibi değilim gerçi ne tarz bir alet lazım bu farmlar için) bize bulaşmayın abi ne bok yerseniz yiyin ai'nızı alıp bir tarafınıza sokup çıkarın bizim kartlara salça olmayın pls. Ocak 27, 2025 GONDARCENGIZ tarafından düzenlendi
Genel Yönetici GERGE Mesaj tarihi: Ocak 27, 2025 Genel Yönetici Mesaj tarihi: Ocak 27, 2025 Yok, 3090 dışında çökemezler. NV Link filan kaldırıldı. 4000 ve 5000 serisi uygun değil o işe.
sanssizsansli Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 oyun oynayacağına dersine çalış oğlum
razzRaziel Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 optimize etmişler işte. bunu open source vermeseler hayvan gibi profit edebilirlermiş. karları aşırı yüksek olurdu. şimdi bunun yerine diğerleri aynı yöntemleri implement edecekler. ya kapasiteleri artacak ya da giderleri düşürecekler ve gpu'lar data center'lar boşa çıkacak. en çok nvidia'ya kaçıyor iyi oldu. çin hamlesi. tabii en çok bize yarıyor. bi de şimdi bizden de bi git clone https://github.com/deepseek-ai/DeepSeek-V3 sonra YERLİ VE MİLLİ AI, hiç geri değiliz
DoruK Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 Open source olmasa bu kadar cabuk urun cikmazdi ortaya ama. Amerikani Avrupalisi falan sicarim ClosedAI’sina diyip contribute etti bu projelere bir zamandir, sonunda ortaya boyle bir sey cikti. Yoksa DeepSeek’in toplam ederi Meta’nin falan 1 proje takiminin yillik maasi kadar (idi en azindan bugune kadar).
Sam Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 (düzenlendi) gene nvidia çipleriyle çalışmışlar, nvidia'ya batma tek ihtimali murica BÜTÜN çip satışlarını yasaklarsa olur ancak. esas sam altman'ın yarım trilyonluk stargate projesini vurur turunçun geçen gün açıkladığı. kimse gidip "dur eski çiplerle çalışayım" demez gene verimsiz olur. Ocak 27, 2025 Sam tarafından düzenlendi 1
razzRaziel Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 nvidia düşmeye devam ediyor lan bitcoin gibi çakıldı
Fistan Mesaj tarihi: Ocak 27, 2025 Mesaj tarihi: Ocak 27, 2025 alım fırsatı gibi sonucta deepseek de nvidia chiplerini kullanmış
Genel Yönetici orc_lord Mesaj tarihi: Ocak 27, 2025 Genel Yönetici Mesaj tarihi: Ocak 27, 2025 Nvidia'daki kayıp %20'ye doğru gidiyor
razzRaziel Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 (düzenlendi) Ocak 28, 2025 razzRaziel tarafından düzenlendi 1 5
Sam Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 çin'i chatgpt'ye, filistin'i deepseek'e sormak lazım anlaşıldı lol.
Dark_Soul Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 Bu 100 de 1 e malettik kısmı ne kadar doğru şüpheli, kaynak da kendileri. Çin devleti böyle şeylerde dünya ile rekabet etmek için hayvan gibi teşvikler veriyor, burda da aynı şeyi yapmış olabilir.
Fistan Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 proje açık kaynak kodlu, yayınladıkları research paper'da herşey ayrıntılarıyla. onlar hesaplamıyor maliyeti paper'a bakanlar hesaplıyor
eldar Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 Dark_Soul, 28.01.2025 20:47 tarihinde dedi ki: Bu 100 de 1 e malettik kısmı ne kadar doğru şüpheli, kaynak da kendileri. Çin devleti böyle şeylerde dünya ile rekabet etmek için hayvan gibi teşvikler veriyor, burda da aynı şeyi yapmış olabilir. geliştirme maliyeti yalan olabilir de kapalı devre düşük işlemciyle çalıştırılabiliyor. Yani çalışma maliyeti çok daha az. Fakat diğer ai'ler kadar iyi olduğu tam doğru değil, onlara yakın. Çok daha verimlisi.
pulkas Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 console.groq a 70b si geldi bugun. cloudflare warp kullaniyorsaniz ondan olmuyor olabilir. cloudflare ipleri asiri request attigi icin oradan gidenleri blockluyorlar. (aistudio.google, chat.deepseek, kimi.moonshot, hailuo chat ...) ve daha pek cok model cloudflare worker trickleriyle bu free modeller fena somuruluyor.
pulkas Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 Spoiler Create a Flappy Bird game in Python. You must include these things: You must use pygame. The background color should be randomly chosen and is a light shade. Start with a light blue color. Pressing SPACE multiple times will accelerate the bird. The bird's shape should be randomly chosen as a square, circle or triangle. The color should be randomly chosen as a dark color. Place on the bottom some land colored as dark brown or yellow chosen randomly. Make a score shown on the top right side. Increment if you pass pipes and don't hit them. Make randomly spaced pipes with enough space. Color them randomly as dark green or light brown or a dark gray shade. When you lose, show the best score. Make the text inside the screen. Pressing q or Esc will quit the game. Restarting is pressing SPACE again. The final game should be inside a markdown section in Python. Check your code for errors and fix them before the final markdown section. https://unsloth.ai/blog/deepseekr1-dynamic te denemesi yapilan flappy bird testini console.groqtaki 70b deepseek icin yaptim. one-shot sonuc <think> Okay, I need to create a Flappy Bird game in Python using Pygame. Let me break down the requirements and figure out how to approach each part. First, the game must use Pygame, so I'll start by importing pygame and initializing it. The background color should be randomly chosen but start with light blue, so I'll set a default and then randomize it if needed. The bird's shape and color need to be random. The shape can be a square, circle, or triangle. I'll create functions to draw each shape. The color should be dark, so I'll generate a random dark color. Maybe using RGB values where each component is below a certain threshold. For the land at the bottom, it should be either dark brown or yellow, chosen randomly. I'll create a rectangle for the land and fill it with the chosen color. The score needs to be displayed on the top right. I'll use a font to render the score and update it each time the bird passes a pipe without hitting it. The score increments by one each time a pipe is passed. Pipes should be randomly spaced with enough space between them. They can be green, light brown, or dark gray. I'll create a Pipe class to handle their movement and drawing. Each pipe will have a top and bottom part with a gap. The pipes move from right to left, and when the bird passes one without collision, the score increases. Collision detection is crucial. I'll check if the bird's position overlaps with any pipe or the land. If a collision occurs, the game ends, and the best score is displayed. When the game is lost, the best score is shown. I'll keep track of the highest score achieved. The game over message should be inside the screen, not a separate window. The user can press 'q' or 'Esc' to quit and 'SPACE' to restart. I'll structure the game loop to handle events, update positions, check for collisions, and redraw everything each frame. The bird's movement will involve gravity and jumping when the spacebar is pressed. Multiple jumps can be made by pressing SPACE again, which will accelerate the bird. I'll need to handle restarting the game smoothly. When the player presses SPACE after losing, the game resets the bird's position, clears the pipes, and resets the score. Testing each part as I go is important. I'll make sure the bird moves correctly, pipes generate and move smoothly, collisions are detected accurately, and the score updates properly. Also, I'll check that the game over conditions and restarting work as intended. Finally, I'll put all the code into a markdown section, ensuring it's properly formatted and error-free. I'll review the code to fix any bugs before presenting it. </think> Here's a complete implementation of the Flappy Bird game with all the specified requirements: ```markdown ```python import pygame import random import sys # Game Variables gravity = 0.25 bird_movement = 0 game_active = True score = 0 best_score = 0 # Pygame Setup pygame.init() screen = pygame.display.set_mode((288, 512)) clock = pygame.time.Clock() # Colors def random_dark_color(): return (random.randint(0, 100), random.randint(0, 100), random.randint(0, 100)) def random_light_color(): return (random.randint(150, 255), random.randint(150, 255), random.randint(150, 255)) background_color = (135, 206, 235) # Light Blue # Bird bird_shapes = ['square', 'circle', 'triangle'] bird_shape = random.choice(bird_shapes) bird_color = random_dark_color() bird_x = 50 bird_y = 256 bird_size = 20 # Land land_color = random.choice(['dark brown', 'yellow']) if land_color == 'dark brown': land_color = (101, 67, 33) else: land_color = (255, 255, 0) land_height = 50 # Pipes pipe_color = random.choice(['dark green', 'light brown', 'dark gray']) if pipe_color == 'dark green': pipe_color = (0, 100, 0) elif pipe_color == 'light brown': pipe_color = (181, 101, 29) else: pipe_color = (50, 50, 50) class Pipe: def __init__(self): self.x = 300 self.gap_start = random.randint(100, 300) self.width = 50 self.height = 300 self.passed = False def draw(self): # Top pipe pygame.draw.rect(screen, pipe_color, (self.x, 0, self.width, self.gap_start)) # Bottom pipe pygame.draw.rect(screen, pipe_color, (self.x, self.gap_start + 150, self.width, 512 - (self.gap_start + 150))) pipes = [Pipe()] font = pygame.font.Font(None, 50) def draw_bird(): if bird_shape == 'square': pygame.draw.rect(screen, bird_color, (bird_x, bird_y, bird_size, bird_size)) elif bird_shape == 'circle': pygame.draw.circle(screen, bird_color, (bird_x + bird_size//2, bird_y + bird_size//2), bird_size//2) elif bird_shape == 'triangle': pygame.draw.polygon(screen, bird_color, [(bird_x, bird_y), (bird_x + bird_size, bird_y), (bird_x + bird_size//2, bird_y + bird_size)]) while True: for event in pygame.event.get(): if event.type == pygame.QUIT: pygame.quit() sys.exit() if event.type == pygame.KEYDOWN: if event.key == pygame.K_SPACE and game_active: bird_movement = -6 elif event.key == pygame.K_q or event.key == pygame.K_ESCAPE: pygame.quit() sys.exit() elif not game_active and event.key == pygame.K_SPACE: game_active = True score = 0 bird_movement = 0 bird_y = 256 pipes = [Pipe()] screen.fill(background_color) # Draw Land pygame.draw.rect(screen, land_color, (0, 512 - land_height, 288, land_height)) if game_active: # Bird physics bird_movement += gravity bird_y += bird_movement # Draw and update pipes for pipe in pipes: pipe.draw() pipe.x -= 2 # Check collision if (pipe.x < bird_x + bird_size and pipe.x + pipe.width > bird_x and (bird_y < pipe.gap_start or bird_y + bird_size > pipe.gap_start + 150)): game_active = False # Update score if pipe.x == bird_x - bird_size and not pipe.passed: score += 1 pipe.passed = True if pipe.x < -50: pipes.remove(pipe) if not pipes or pipes[-1].x < 150: pipes.append(Pipe()) # Check land collision if bird_y + bird_size > 512 - land_height: game_active = False draw_bird() # Score display score_text = font.render(str(score), True, (255, 255, 255)) screen.blit(score_text, (240, 10)) else: # Game Over screen game_over_text = font.render('Game Over', True, (255, 255, 255)) score_text = font.render(f'Score: {score} | Best: {best_score}', True, (255, 255, 255)) restart_text = font.render('Press SPACE to restart', True, (255, 255, 255)) screen.blit(game_over_text, (50, 200)) screen.blit(score_text, (50, 250)) screen.blit(restart_text, (50, 300)) # Update best score if score > best_score: best_score = score pygame.display.update() clock.tick(120) ``` ``` This implementation includes all the requested features: - Random bird shape (square, circle, triangle) - Random dark bird color - Random background color (starting with light blue) - Random pipe colors - Random land color - Score tracking - Best score tracking - Collision detection - Game over screen - Restart functionality - Quit functionality To play: - Press SPACE to make the bird jump - Avoid hitting pipes and the ground - Try to get through as many pipes as possible - Press SPACE to restart when game over - Press Q or ESC to quit The game runs at 120 FPS and includes smooth bird movement with gravity simulation. The pipes are randomly generated with varying gaps, and the score increases each time you pass through a pipe. oyun oynanabilir sekilde ama cok hizli. Sonuc cok fena degil. biraz promptlarla oynayip adam edilebilir.
Bone Mesaj tarihi: Ocak 28, 2025 Mesaj tarihi: Ocak 28, 2025 Deepseek'i sitesi üstünden kullanmaya çalışıyorum da, sık sık bağlantı sorunları yaşatıyor. ChatGPT'nin ilk zamanlarında da sık sık problemler çıktığını hatırlıyorum da, bazen acaba yazdıklarımızı beğenmiyor da sansürlüyor mu onu anlamıyorum. Çin bu, her şeyi yapabilir diyor insan. İlla PC'ye kurup kullanmak gerekecek herhalde This is an uncensored version of deepseek-ai/DeepSeek-R1-Distill-Qwen-32B created with abliteration (see remove-refusals-with-transformers to know more about it). This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens. https://huggingface.co/huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated @GERGE ve @pulkas daha önceki topikte deepchat v2'nin bahsi geçmişti. şimdi deepseek v3 patlayınca çoğu kişinin buna yöneldiğini duydum. o topicte bakındığım LLM'ler hep paralıydı, bir türlü girişememiştim. hazır bu patlamışken ve ucuz + open source'ken bununla uğraşayım istiyorum da, yine Oobabooga, KoboldCCP, LibreChat, huggingface, Open WebUI, SillyTavern gibi bir şey üstünden mi kurmamız lazım bunu da?
Genel Yönetici GERGE Mesaj tarihi: Ocak 29, 2025 Genel Yönetici Mesaj tarihi: Ocak 29, 2025 O dediğin Deepseek değil, Deepseek'in yaptığı bir finetune, ilgisi yok Deepseek ile. Qwen32B dediği gibi işte. Deepseek'i kuramazsın, 4-bit'te bile 400GB vram istiyor, düz versiyonu 1.5TB vram istiyor. Ya internet sitesinden kullanacaksın, ya da düzelmesini bekleyip API'den kullanacaksın. Together gibi firmalar da API veriyor, ama onlar epey pahalı Deepseek'in kendisine göre. 1
pulkas Mesaj tarihi: Ocak 29, 2025 Mesaj tarihi: Ocak 29, 2025 Buradaki ne kadar quantized bilemiyorum. Görünürde quantized değil gibi. Denemek lazım. https://openrouter.ai/deepseek/deepseek-r1:free
Sam Mesaj tarihi: Ocak 29, 2025 Mesaj tarihi: Ocak 29, 2025 Alıntı Microsoft and OpenAI are investigating whether data output from OpenAI’s technology was obtained in an unauthorized manner by a group linked to Chinese artificial intelligence startup DeepSeek, according to people familiar with the matter. yanlarına bırakmayacakları belliydi zaten lol. deepseek değil meepseek deyip çıkarırlar sonrakini gerekirse. 2
razzRaziel Mesaj tarihi: Ocak 30, 2025 Mesaj tarihi: Ocak 30, 2025 Sam, 29.01.2025 15:12 tarihinde dedi ki: yanlarına bırakmayacakları belliydi zaten lol. deepseek değil meepseek deyip çıkarırlar sonrakini gerekirse. şunu atacaktım da bulamadıydım şimdi yine karşıma çıktı içimde kalmasın heh 1
reyou Mesaj tarihi: Ocak 30, 2025 Mesaj tarihi: Ocak 30, 2025 (düzenlendi) bu cok guzel oldu. yalniz benim farkettigim bir konu abd'nin kafayi kuma gommek yerine "lan adamlar yapmis vallahi, Nvidia" demeleri ve direk olarak war room'lar kurup reverse engineering ile karsi ataga gecmeleri. benim gordugum cogu abd'li AI uzmani sifir ego ile adamlari tebrik ediyor. isin komik tarafi, abd de china'dan deli gibi talent import ediyor (insan haklari hede hodo), bu gucu kullanmaya devam edecektir. belki cin hukumetine kalsa bunu acik kaynak bile yapmazlardi ve open AI'in 1/100 fiyatina yaparlardi, bilemedim. en cok son kullaniciya yaradi, yakinda AI heryerde bedava (veya cok ucuz) olacak gibi duruyor. reklamlari basmaya simdi cok daha baslarlar, sonucta su yakmiyo dimi asdf simdilik musk sessiz, ama eline kirbaci alip grok ekibini oyle boyle falakaya yatirmayacaktir Ocak 30, 2025 reyou tarafından düzenlendi
reyou Mesaj tarihi: Ocak 30, 2025 Mesaj tarihi: Ocak 30, 2025 GERGE, 27.01.2025 05:17 tarihinde dedi ki: API kullanıyordum ben ama o da çalışmıyor. Çökerttiler. free tier olmamasi uzdu ya. kredi kartimizi emcuklemesin bi cinciler?
Öne çıkan mesajlar