Small, tokenizer-free language models trained entirely from scratch. No BPE, no subword vocabulary — these models read and write raw bytes or characte
❓ How small can it be?
Ömür Bera Işık
omurberaisik
AI & ML interests
Hey! I'm Ömür Bera Işık. Big dreams, tiny budget :D but I'm doing my best to keep my projects going anyway. I live in Turkey. No team, no backers yet (besides my family), just grinding away at whatever I can. Current balance: $0 :)
Recent Activity
updated a Space about 7 hours ago
VeroLM/Sandbox updated a model 3 days ago
VeroLM/NoTokenLM-Gen-3.5 updated a model 3 days ago
VeroLM/NoTokenLM-Gen-2.5Organizations
models 8
omurberaisik/MateMistral-7B-base
Updated • 2
omurberaisik/tiny-LLM
Text Generation • Updated • 35 • 1
omurberaisik/Lossantos-image
Text-to-Image • 0.9B • Updated • 8 • 1
omurberaisik/holocomn13b
Updated
omurberaisik/holocom-lite-v8_test_ai
Text Generation • Updated • 5 • 1
omurberaisik/holocomnb7-merged
Text Generation • 7B • Updated • 9 • 1
omurberaisik/holocomnb7
Text Generation • Updated • 10
omurberaisik/biotransformer-1b-v1
Text Generation • Updated
datasets 0
None public yet