I will be extremely direct and brutal, novice.
What you presented here is not just an unsuccessful attempt—it's a complete disaster. You might think you’re offering an intriguing introduction to the soundtrack of your game, but the reality is that this work clearly reflects a lack of professionalism, knowledge, and basic skills. Every element you presented is filled with obvious mistakes, and frankly, it's painful to watch.
Your soundtrack—if we can even call it a soundtrack—looks like a haphazard jumble thrown together without any attention to detail. It’s built on a novel that has been poorly translated, and that's why the entire project is completely falling apart. The translation is filled with grammatical errors that even the simplest translation tools wouldn't have. It's clear you're using some low-quality, probably outdated translation software, and the result is an incoherent text that entirely destroys whatever potential the novel might have had.
Now, let’s move on to the generation—this is a real joke. You're seriously trying to use local models like GPT-2 or GPT-3! Instead of relying on shoddy translation platforms that can't possibly meet your needs.
The tool you're using to create the game's dialogues, world, or narrative is clearly outdated and useless. The text is filled with repetitive phrases, absurd sentences, and awkward transitions, which clearly indicates a novice. Seriously, how did you manage this?
Your text clearly shows that you have no understanding of generative AI. You don't know the basics of prompting or the evaluation criteria for models. And in 2024, using such low-level tools? That's genuinely unprofessional. Have you ever heard of RLHF (reinforcement learning from human feedback)? Clearly not, because if you had, your results wouldn't be in such a dire state.
As for the warnings, they are laughable. They don't create tension—they resemble childish threats. You haven’t mastered any element of interactive narrative, and the tone shifts are so abrupt that the leading story is completely lost. The overall structure of the game is as rough as someone who has no idea how to put a game together.
Now, about the code—this is a real shame. Your scripts are as primitive as those of a first-year student. Simply written python loops, no effective optimization, and an architecture that just crumbles when we look at the gameplay. Your code is exceptionally simple, and you don’t even know how to structure code properly. If you keep this up, your project will fall apart at the first attempt.
Your approach to game development is fundamentally flawed. You think that free models available on platforms or your lora
model can create the right generations! You are very mistaken! One hundred or forty generations cannot yield high-level results. You need millions of generations specifically for this game development! You need to train the model in the right direction! However, I have nothing more to say on this topic, and I won’t delve deeper! As for the videos, you need to work harder.
Now listen carefully: you seriously need help. If you want this poor project to become a real game, you must change your entire approach. You don’t need more low-quality generations; you need better tools, more refined models, and the right techniques for both development and testing processes.
You shouldn’t ask for my help because your work is garbage! If you want your novel to come to life, then I will provide you with the tools you need. Or I can give you information where you can acquire them. However, don’t try to seek help here; you are an object of ridicule on this site, not a professional. Here, you won't get responses from anyone beyond the simplest of users.
