GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
while-ai
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
while-ai
Model Tree
Qwen/Qwen2.5-1.5B-Instruct
Input Modalities
Output Modalities
Supported Functionality
| Arm | pass@1 | 95% CI | pass@4 | Steps | GPU min |
|---|---|---|---|---|---|
| Base, no training | 0.40 | [0.33, 0.47] | 0.58 | 0 | 0 |
| Baseline (uncorrected stale REINFORCE) | 0.47 | [0.41, 0.55] | 0.68 | 40 | 22.9 |
Recipe (wai.FlashReinforce) | 0.41 | [0.34, 0.48] | 0.63 | 40 | 21.3 |
Recipe vs baseline: -0.069 [-0.115, -0.025] over 120 paired tasks. Verdict: unresolved, one seed per arm. The rollouts were barely stale at this step size: mean sampler-to-learner KL was 3e-4 and the trust gate masked one of 64 trajectories on 5 of 40 steps. The correction had almost nothing to correct. Only the recipe arm's adapter was kept; history.json is its per-step log.
The root holds the arm the recipe README's headline number reports. Every other arm is a subfolder named after it. checkpoints/ never ships.
| folder | arm |
|---|---|
. | recipe arm: wai.FlashReinforce update, 2026-09-21 run |
python
from peft import PeftModelfrom transformers import AutoModelForCausalLM base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-1.5B-Instruct")model = PeftModel.from_pretrained(base, "while-ai/paper-flash-reinforce-1.5b") # the headline armbash
git clone https://github.com/whilehq/whileai-sdk && cd whileai-sdk/recipes/papers/flash-reinforcepython recipe.pyThe recipe README pins the seed, the library versions and the GPU, and its Checks table says what the eval verified. Read the Learned section before quoting a number from this card.
from peft import PeftModelfrom transformers import AutoModelForCausalLM base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-1.5B-Instruct")model = PeftModel.from_pretrained(base, "while-ai/paper-flash-reinforce-1.5b") # the headline arm