GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
srpone
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
srpone
Model Tree
Qwen/Qwen3-Reranker-4B
Input Modalities
Output Modalities
Supported Functionality
Pairwise accuracy in percent on ShopRank-Bench (10,511 judged preference pairs from 2,991 queries); intervals are 95% query-clustered bootstrap CIs. Tiers are the number of judge families (of three) that committed to the label.
| Model | Product text | Overall [95% CI] | Gold (3/3) | Silver (2/3) | Bronze (1/3) |
|---|---|---|---|---|---|
| ZooWork-ShopRanker-4B | Structured | 81.2 [80.3, 82.0] | 96.3 | 84.3 | 71.4 |
Qwen3-Reranker-4B (base) | Structured | 77.5 [76.5, 78.5] | 92.7 | 81.4 | 66.7 |
| ZooWork-ShopRanker-4B | Natural language | 79.5 [78.6, 80.4] | 95.4 | 82.1 | 69.8 |
Qwen3-Reranker-4B (base) | Natural language | 77.9 [76.9, 78.9] | 92.1 | 81.0 | 68.5 |
Significantly beats the strongest open baseline (Jina-reranker-m0: 79.2 structured, 78.1 natural language) in both formats and its own base by 3.7 points on structured text.
Report per tier: every reranker loses 20–25 points from gold to bronze, so an aggregate score is dominated by pairs nothing gets wrong.
One NVIDIA H200, fp16, 1024-token limit, 256 real (query, product) pairs, adapter merged into the weights (LoRA adds no inference cost once merged).
| Params | Batch-1 p50 | Batch-1 p95 | Docs/s (batch 32) | Peak memory |
|---|---|---|---|---|
| 4.02B | 25.2 ms | 33.3 ms | 154.5 | 11.2 GB |
Requires pip install transformers peft. Load the base weights and the adapter from this
repo as shown: wrapping with PeftModel keeps the adapter in fp32, which reproduces the paper
numbers exactly (loading the repo id with from_pretrained alone also attaches the adapter,
but in bf16, and drifts by up to ~0.3 points).
The score is P(yes) = softmax over the bare no / yes logits at the last position of
the native Qwen3-Reranker prompt -- exactly the scorer behind the numbers above.
python
import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer model_id = "srpone/zoowork-shopranker-4b" # base weights + LoRA adapter, both in this repotok = AutoTokenizer.from_pretrained(model_id, padding_side="left")model = AutoModelForCausalLM.from_pretrained(model_id, dtype=torch.bfloat16)# Re-wrapping keeps the adapter in fp32, as in the paper. peft may warn "Already found a# `peft_config` attribute"; that is expected here -- exactly one adapter ends up loaded.model = PeftModel.from_pretrained(model, model_id).cuda().eval() PREFIX = ("<|im_start|>system\nJudge whether the Document meets the requirements based on " "the Query and the Instruct provided. Note that the answer can only be \"yes\" " "or \"no\".<|im_end|>\n<|im_start|>user\n")SUFFIX = "<|im_end|>\n<|im_start|>assistant\n<think>\n\n</think>\n\n"INSTRUCTION = "Given a web search query, retrieve relevant passages that answer the query"YES, NO = tok.convert_tokens_to_ids("yes"), tok.convert_tokens_to_ids("no") @torch.no_grad()def score(query: str, products: list[str]) -> list[float]: texts = [PREFIX + f"<Instruct>: {INSTRUCTION}\n<Query>: {query}\n<Document>: {p}" + SUFFIX for p in products] enc = tok(texts, return_tensors="pt", padding=True, truncation=True, max_length=4096, add_special_tokens=False).to(model.device) logits = model(**enc).logits[:, -1, :] return torch.softmax(logits[:, [NO, YES]].float(), dim=-1)[:, 1].tolist() print(score("jumpsuit under $100 boho", [ "title: Boho Wide-Leg Jumpsuit | price: $44 | color: rust", "title: Bohemian Silk Jumpsuit | price: $239 | color: ivory",]))Product text may be the structured attribute schema or natural-language prose. The model was trained on the structured schema and carries its gain to prose (see the table).
Repository contents. config.json, model*.safetensors and the tokenizer files are
the unmodified Qwen3-Reranker-4B release; adapter_config.json and
adapter_model.safetensors are the ZooWork-ShopRanker LoRA adapter. The adapter is kept
separate rather than merged so the published model reproduces the paper numbers exactly.
If your serving stack needs a single set of weights, merging the adapter into the base in
bf16 stays within 0.4 points of the reported accuracy.
LoRA (rank 16, α=32, dropout 0.05, on q/k/v/o projections) over judge-labelled preference pairs from private search traffic, labelled by a three-family LLM judge panel in both presentation orders, mixed with general retrieval replay data. This model was distilled from the aligned 8B, then sharpened on judged pairs. The benchmark comes from a disjoint slice of traffic (zero query and zero exact-pair overlap).
price and color while degrading style and text
product typeApache License 2.0 (see LICENSE), inherited from Qwen/Qwen3-Reranker-4B
(Apache-2.0, by the Qwen team). This repository redistributes the base weights
and tokenizer unmodified; the ZooWork-ShopRanker modification is the LoRA adapter,
trained on judged preference data.
bibtex
@article{xue2026zooworkshopranker, title = {ZooWork-ShopRanker: An Open, Preference-Aligned E-Commerce Reranker}, author = {Xue, Siqiao and Liu, Shuxuan and Hu, Ning}, journal = {arXiv preprint arXiv:2609.31002}, year = {2026}, url = {https://arxiv.org/abs/2609.31002}}import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer model_id = "srpone/zoowork-shopranker-4b" # base weights + LoRA adapter, both in this repotok = AutoTokenizer.from_pretrained(model_id, padding_side="left")model = AutoModelForCausalLM.from_pretrained(model_id, dtype=torch.bfloat16)# Re-wrapping keeps the adapter in fp32, as in the paper. peft may warn "Already found a# `peft_config` attribute"; that is expected here -- exactly one adapter ends up loaded.model = PeftModel.from_pretrained(model, model_id).cuda().eval() PREFIX = ("<|im_start|>system\nJudge whether the Document meets the requirements based on " "the Query and the Instruct provided. Note that the answer can only be \"yes\" " "or \"no\".<|im_end|>\n<|im_start|>user\n")SUFFIX = "<|im_end|>\n<|im_start|>assistant\n<think>\n\n</think>\n\n"INSTRUCTION = "Given a web search query, retrieve relevant passages that answer the query"YES, NO = tok.convert_tokens_to_ids("yes"), tok.convert_tokens_to_ids("no") @torch.no_grad()def score(query: str, products: list[str]) -> list[float]: texts = [PREFIX + f"<Instruct>: {INSTRUCTION}\n<Query>: {query}\n<Document>: {p}" + SUFFIX for p in products] enc = tok(texts, return_tensors="pt", padding=True, truncation=True, max_length=4096, add_special_tokens=False).to(model.device) logits = model(**enc).logits[:, -1, :] return torch.softmax(logits[:, [NO, YES]].float(), dim=-1)[:, 1].tolist() print(score("jumpsuit under $100 boho", [ "title: Boho Wide-Leg Jumpsuit | price: $44 | color: rust", "title: Bohemian Silk Jumpsuit | price: $239 | color: ivory",]))@article{xue2026zooworkshopranker, title = {ZooWork-ShopRanker: An Open, Preference-Aligned E-Commerce Reranker}, author = {Xue, Siqiao and Liu, Shuxuan and Hu, Ning}, journal = {arXiv preprint arXiv:2609.31002}, year = {2026}, url = {https://arxiv.org/abs/2609.31002}}