GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
RSRS64
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
RSRS64
Model Tree
Qwen/Qwen3-4B-Instruct-2507
Input Modalities
Output Modalities
Supported Functionality
Held-out eval, all 9 ctf-lite tasks, same mini_swe_agent harness, n=16 rollouts/task, Fisher's exact:
| base Qwen3-4B | + this adapter | Δ | p | |
|---|---|---|---|---|
| overall (9 tasks) | 51/144 (0.35) | 55/144 (0.38) | +0.03 | 0.71 |
| trained subset (3 tasks) | 35/48 (0.73) | 39/48 (0.81) | +0.08 | 0.47 |
No result reaches p<0.05 — the difference is within noise. Training reward rose (~0.75→0.82) but did not translate to a significant single-shot held-out gain. An earlier n=3 eval suggested +0.11; that was a small-sample artifact that vanished at n=16. The adapter is published for reproducibility, not as a capability claim. Full writeup + logs + the lever that would actually move it (a procedural difficulty generator): https://github.com/rohseh303/prime-rl-environments (RESULTS.md)
python
from peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizerbase = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B-Instruct-2507", torch_dtype="bfloat16")model = PeftModel.from_pretrained(base, "RSRS64/qwen3-4b-ctf-lite-grpo")tok = AutoTokenizer.from_pretrained("Qwen/Qwen3-4B-Instruct-2507")
GRPO + LoRA, lr 1e-5, group 16, batch 16, ~70 steps, 2×H100. Curriculum = the 3 tasks the base model
partially solves (git-leaked-secret, xor-vault, hash-crack) — the only ones with reward variance,
hence gradient. Training reward climbed ~0.75 → ~0.82.