HRM-Text-1B
A PrefixLM-architecture language model developed by Sapient AI with 1.18B parameters, serving as the foundation model for ZY-S1 fine-tuning.
Timeline (last 90 days)
ZY-S1 Preview was not trained from scratch, but is built on top of HRM-Text-1B with adaptations and adjustments made for the ZY-S1 project.
HRM-Text-1B is a 1-billion-parameter-scale language model that has attracted much attention in the open-source community recently; 'HRM' in the name usually refers to Hierarchical Recurrent Memory or a similar efficient architecture design approach.
ZY-S1 takes Sapient AI's HRM-Text-1B (a PrefixLM architecture) as its base and performs 1000 steps of supervised fine-tuning (SFT), plus 120 steps of error-correction training.
All Facts (3)
ZY-S1 Preview was not trained from scratch, but is built on top of HRM-Text-1B with adaptations and adjustments made for the ZY-S1 project.
50%UnverifiedHRM-Text-1B is a 1-billion-parameter-scale language model that has attracted much attention in the open-source community recently; 'HRM' in the name usually refers to Hierarchical Recurrent Memory or a similar efficient architecture design approach.
50%UnverifiedZY-S1 takes Sapient AI's HRM-Text-1B (a PrefixLM architecture) as its base and performs 1000 steps of supervised fine-tuning (SFT), plus 120 steps of error-correction training.
50%