yavuz-ai/oghuz-tokenizer
Updated
Emre Yavuz PRO
yavuz-ai
AI & ML interests
ships ai in places where it has to behave · post-training research in the open · oghuz, a 2.5b turkish lm from scratch · regulated-evals for regulated finance · rlhf, rlvr, constitutional ai · basel
Organizations
yavuz-ai/qwen2.5-3b-dpo-finqa
Text Generation • Updated • 10
yavuz-ai/qwen2.5-3b-grpo-finqa
Text Generation • Updated • 12
yavuz-ai/qwen2.5-1.5b-grpo-gsm8k
Text Generation • Updated
yavuz-ai/grpo
Updated
yavuz-ai/qwen2.5-1.5b-ppo
Text Generation • Updated
yavuz-ai/qwen2.5-1.5b-rm-ultrafeedback
Text Classification • 2B • Updated • 640
yavuz-ai/ppo
Updated
yavuz-ai/qwen3-4b-cai-rlaif-v2
Text Generation • Updated • 4
yavuz-ai/qwen3-4b-cai-rlaif
Text Generation • Updated • 10
yavuz-ai/qwen3-4b-cai-sft
Text Generation • 4B • Updated • 8
yavuz-ai/qwen3-4b-cai-rlaif-v2-merged
Text Generation • 4B • Updated • 2
yavuz-ai/qwen3-4b-cai-rlaif-merged
Text Generation • 4B • Updated • 4
yavuz-ai/ministral-8b-dpo-ultrafeedback-lora
Text Generation • Updated • 1
yavuz-ai/qwen3-4b-dpo-ultrafeedback-lora
Text Generation • Updated • 13