| --- |
| base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B |
| tags: |
| - rlve |
| - grpo |
| - opd-teacher |
| --- |
| |
| # RLVE OPD teacher: Sorting |
|
|
| DeepSeek-R1-Distill-Qwen-1.5B trained for 150 steps on difficulty-0 |
| `Sorting` prompts, with four prompts and 16 rollouts per step. Training |
| did not use DAPO prompt filtering. These are the final model weights at step |
| 149 for environment-specific on-policy distillation. |
|
|
| Training project: `david-heineman/rl-data-opd-teachers-r1-distil`. |
| Training group: `opd-teachers-r1-nofilter16-20260929-231458`. |
|
|
