GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
Egbertjing
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
Egbertjing
Model Tree
Qwen/Qwen3-32B
Input Modalities
Output Modalities
Supported Functionality
Qwen/Qwen3-32Bq_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_projRLPF evaluates generated code by extracting the solution, compiling or loading it, running correctness checks, and measuring runtime. The default reward is:
text
R = max(stage_shape, CGRE + alpha_F * FBR + alpha_R * RBR)
where:
CGRE: correctness-gated relative efficiencyFBR: faster-than-baseline indicatorRBR: reference-or-better indicatoralpha_F = 0.16alpha_R = 0.10-0.05 / 0.00 / 0.05 / 0.10On the PerfCodeBench test split with 306 family-disjoint tasks:
| Model | CRR | FBR | RBR | CGRE |
|---|---|---|---|---|
| Qwen3-32B base | 11.11 | 8.17 | 5.56 | 8.11 |
| RLPF adapter, checkpoint-1500 | 54.58 | 46.41 | 25.82 | 38.58 |
All values are percentages. CRR is correct-and-runnable rate; FBR is faster-than-baseline rate; RBR is reference-or-better rate; CGRE is correctness-gated relative efficiency.
python
from transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import PeftModelbase = "Qwen/Qwen3-32B"adapter = "Egbertjing/RLPF-Qwen3-32B-PerfCodeBench"tokenizer = AutoTokenizer.from_pretrained(base, trust_remote_code=True)model = AutoModelForCausalLM.from_pretrained(base,torch_dtype="auto",device_map="auto",trust_remote_code=True,)model = PeftModel.from_pretrained(model, adapter)
For best compatibility with the training distribution, use the chat template with enable_thinking=False when supported.
This adapter is intended for research on code generation, executable feedback, and performance optimization. It is not a general software-engineering assistant and should not be used without correctness tests, sandboxing, and runtime validation.
Generated code may be incorrect, unsafe, non-compiling, or slower than the baseline. Performance measurements depend on compiler versions, hardware, system load, and benchmark harness settings. Always evaluate generated code under the target environment.