GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
distil-labs
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
distil-labs
Model Tree
Qwen/Qwen3.5-4B
Input Modalities
Output Modalities
Supported Functionality
| Model | Correct |
|---|---|
| Fine-tuned, thinking on | 98 (median 272.5 reasoning tokens) |
| Fine-tuned, thinking off | 84 |
| Qwen3.5-4B untuned, thinking on | 37 (63 of 100 cut off at 2,048 tokens) |
| Qwen3.5-4B untuned, thinking off | 53 |
Data, configs, predictions and the scoring script: https://github.com/distil-labs/reasoning-blogpost-benchmark.
Serve the base model with this adapter:
bash
vllm serve Qwen/Qwen3.5-4B --enable-lora --max-lora-rank 64 --lora-modules distil-qwen3.5-4b-invoice-decision-no-thinking=distil-labs/distil-qwen3.5-4b-invoice-decision-no-thinking --port 8001python
from openai import OpenAI client = OpenAI(base_url="http://127.0.0.1:8001/v1", api_key="EMPTY")response = client.chat.completions.create( model="distil-qwen3.5-4b-invoice-decision-no-thinking", messages=[{"role": "system", "content": SYSTEM_PROMPT}, {"role": "user", "content": text}], temperature=0, extra_body={"chat_template_kwargs": {"enable_thinking": False}},)Example answer: {"decision": "hold_price"}. Keep the system prompt exactly as below and serve with thinking off, as trained.
text
You check vendor invoices for the accounts payable team of Northwind before they are paid. You receive three documents: the invoice message as it arrived by email, the purchase order (PO) from the ERP, and the goods receipt from the ERP. Invoice lines can be in a different order than the PO lines and can use the vendor's own item names or abbreviations, so match each invoice line to its PO line by item. Amounts in the message that are not part of this invoice (a previous balance, a payment received, a quote) are ignored. Run these checks in order and stop at the first one that fails:1. The PO number referenced on the invoice is exactly the number of the PO. If not: hold_no_po.2. On every invoice line, the invoiced quantity is not more than the received quantity of that item. If not: hold_quantity.3. On every invoice line, the invoiced unit price is not more than 2% above the PO unit price of that item. A lower price is fine. If not: hold_price.4. The total stated on the invoice equals the sum of the invoice line totals, plus the freight charge only if the PO says freight may be added by the vendor. If the PO says freight is not allowed, a freight charge must not be part of the total. If not: hold_total.If every check passes: approve. Answer with a JSON object and nothing else: {"decision": "<label>"}, where <label> is one of approve, hold_no_po, hold_quantity, hold_price, hold_total.| Base model | Qwen/Qwen3.5-4B |
| Teacher | GLM 5.3, reasoning effort high |
| Thinking | off (enable_thinking: false) |
| Seed examples | 40 |
| Synthetic examples | 4,156 |
| Method | LoRA (rank 64), 4 epochs; this repo holds the adapter only |
The data is synthetic and in English. The model is trained for this one task and policy; it is not a general assistant.
distil labs · GitHub · Hugging Face · X
vllm serve Qwen/Qwen3.5-4B --enable-lora --max-lora-rank 64 --lora-modules distil-qwen3.5-4b-invoice-decision-no-thinking=distil-labs/distil-qwen3.5-4b-invoice-decision-no-thinking --port 8001from openai import OpenAI client = OpenAI(base_url="http://127.0.0.1:8001/v1", api_key="EMPTY")response = client.chat.completions.create( model="distil-qwen3.5-4b-invoice-decision-no-thinking", messages=[{"role": "system", "content": SYSTEM_PROMPT}, {"role": "user", "content": text}], temperature=0, extra_body={"chat_template_kwargs": {"enable_thinking": False}},)You check vendor invoices for the accounts payable team of Northwind before they are paid.You receive three documents: the invoice message as it arrived by email, the purchase order (PO) from the ERP, and the goods receipt from the ERP. Invoice lines can be in a different order than the PO lines and can use the vendor's own item names or abbreviations, so match each invoice line to its PO line by item. Amounts in the message that are not part of this invoice (a previous balance, a payment received, a quote) are ignored.Run these checks in order and stop at the first one that fails:1. The PO number referenced on the invoice is exactly the number of the PO. If not: hold_no_po.2. On every invoice line, the invoiced quantity is not more than the received quantity of that item. If not: hold_quantity.3. On every invoice line, the invoiced unit price is not more than 2% above the PO unit price of that item. A lower price is fine. If not: hold_price.4. The total stated on the invoice equals the sum of the invoice line totals, plus the freight charge only if the PO says freight may be added by the vendor. If the PO says freight is not allowed, a freight charge must not be part of the total. If not: hold_total.If every check passes: approve.Answer with a JSON object and nothing else: {"decision": "<label>"}, where <label> is one of approve, hold_no_po, hold_quantity, hold_price, hold_total.