GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Aeroalrza
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
Aeroalrza
Model Tree
Qwen/Qwen2.5-0.5B-Instruct
Input Modalities
Output Modalities
Supported Functionality
Educational DPO pilot trained for 50 optimizer steps on Intel/orca_dpo_pairs.
This repository contains the LoRA adapter, not the full base model. The validation preference accuracy was about 97.8%; this is a pair-ranking metric, not a measure of factual correctness. The pilot had mixed generation results and has not been thoroughly evaluated.
Load the adapter with the base model:
python
from transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import PeftModelimport torch base_id = "Qwen/Qwen2.5-0.5B-Instruct"adapter_id = "Aeroalrza/qwen25-05b-orca-dpo-pilot"tokenizer = AutoTokenizer.from_pretrained(adapter_id)base = AutoModelForCausalLM.from_pretrained( base_id, dtype=torch.float16, device_map="auto")model = PeftModel.from_pretrained(base, adapter_id)