GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
AnVu10
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
AnVu10
Model Tree
Qwen/Qwen3.5-9B
Input Modalities
Output Modalities
Supported Functionality
LoRA adapter fine-tuned for Vietnamese customer-support ticket triage → 4-field
JSON (intent, urgency, product, sentiment). Lab 21 (AICB-P2T3 Day 21,
Track 3), student submission.
Qwen/Qwen3.5-9B (bf16) · placement text-linear (12 module types, vision
tower excluded) · r=16 · alpha=32 · LR 1e-4 · 30 optimizer steps · loss mask
assistant-only (verified by decoding the supervised span).python
from peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizerbase = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3.5-9B", dtype="bfloat16", device_map="auto")tok = AutoTokenizer.from_pretrained("Qwen/Qwen3.5-9B")model = PeftModel.from_pretrained(base, "AnVu10/lab21-qwen3.5-9b-triage-vi-lora")