GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
PrimeIntellect
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
PrimeIntellect
Model Tree
PrimeIntellect/Qwen3.5-0.8B-Reverse-Text-SFT
Input Modalities
Output Modalities
Supported Functionality
21814b401.examples/basic/reverse-text/rl.toml at that commit with model.name = the SFT checkpoint and orchestrator.renderer.name = "qwen3.5": 20 steps, 8 prompts x 16 rollouts (batch 128), 128 max tokens, lr 3e-6, 1 inference + 1 trainer H200, NCCL weight broadcast.reverse-text eval reward (256 prompts, temperature 1): step 0: 0.3086, step 5: 0.3543, step 10: 0.3967, step 15: 0.4027, step 20: 0.4077.On Qwen3.5-0.8B, prime-rl RL currently shows a much higher trainer/inference mismatch KL (0.01-0.1 per step) than on Qwen3-0.6B (about 0.002) with the same recipe, and learning stalls after about 20 steps. This is under investigation.