Veyra2-Apricot-50M-Tetris-Verbose

Note: AI was used in the creation of this project. But that's why you're here, isn't it?

This is a Tetris finetune of Veyra2-Apricot...

A new small finetuning corpus was constructed with data from https://huggingface.co/spaces/DedeProGames/SLM-Tetris-Arena, and this model is intended to dominate. And, oh hell yes, it dominates.

It is HILARIOUSLY good compared to all other models at playing Tetris, and it handily beats every model in the arena, no questions asked.

Compared to other Tetris-Verbose models, it's not as good, for some reason. I'm not sure why... but I've done some experimenting and I believe Rose-1.5-Medium by GODELEV is next for FTing

Update: I believe the training data may be contaminated, as the Rose checkpoint is also quite bad.

Benchmarks

Good at Tetris.

Model specs

• Layers: 16
• Hidden size: 512
• Attention heads: 8
• KV heads: 2
• Context length: 1K
• Intermediate size: 1408
• Vocabulary: 8,192
• Parameters: 49,303,040
• OG Creator: veyra-ai

Thanks veyra-ai for this awesome model! This is a finetuned model of SupraLabs/SupraNeo-4M.