rank1-32b: Test-Time Compute for Reranking in Information Retrieval
📄 Paper | 🚀 GitHub Repository
rank1 is a reasoning reranker model that "thinks" before making relevance judgments. This 32B parameter model is trained from the Qwen2.5-32B base model and leverages test-time compute to generate reasoning chains before deciding if a document is relevant to a query.
Model Description
rank1 introduces a novel approach to information retrieval by generating explicit reasoning chains before making relevance judgments. Unlike traditional rerankers that directly output scores, rank1:
- Receives a query and document pair
- Generates a reasoning chain within a
<think>...</think>section - Makes a binary relevance judgment (
trueorfalse) - Returns a confidence score based on the logits of the true/false tokens
This approach helps the model break down complex relevance decisions into logical steps, improving performance across diverse retrieval tasks.
Model Family
| Model | Base | Description |
|---|---|---|
| rank1-0.5b | Qwen2.5-0.5B | Smallest variant (0.5B parameters) |
| rank1-1.5b | Qwen2.5-1.5B | Smaller variant (1.5B parameters) |
| rank1-3b | Qwen2.5-3B | Smaller variant (3B parameters) |
| rank1-7b | Qwen2.5-7B | Smaller variant (7B parameters) |
| rank1-14b | Qwen2.5-14B | Larger variant (14B parameters) |
| rank1-32b | Qwen2.5-32B | Current model (32B parameters) |
| rank1-mistral-2501-24b | Mistral-Small 2501 24B | Trained from Mistral base |
| rank1-llama3-8b | Llama 3.1 8B | Trained from Llama 3.1 base |
Quantized Variants
| Model | Description |
|---|---|
| rank1-7b-awq | Quantized version of rank1-7b |
| rank1-14b-awq | Quantized version of rank1-14b |
| rank1-32b-awq | Quantized version of rank1-32b |
| rank1-mistral-2501-24b-awq | Quantized version of rank1-mistral-24b |
| rank1-llama3-8b-awq | Quantized version of rank1-llama3-8b |
Associated Data and Resources
| Resource | Description |
|---|---|
| rank1-r1-msmarco | All R1 output examples from MS MARCO |
| rank1-training-data | Training data used for rank1 models |
| rank1-run-files | Pre-computed run files for use in top 100 doc reranking |
| GitHub Repository | Official rank1 repository |
Usage
Note that official usage is found on the Github and accounts for edge cases. But for simple use cases the minimal example below works.
from vllm import LLM, SamplingParams
import math
# Initialize the model with vLLM
model = LLM(
model="jhu-clsp/rank1-32b",
tensor_parallel_size=1, # Number of GPUs
trust_remote_code=True,
max_model_len=16000, # Context length
gpu_memory_utilization=0.9,
dtype="float16",
)
# Set up sampling parameters
sampling_params = SamplingParams(
temperature=0,
max_tokens=8192,
logprobs=20,
stop=["</think> true", "</think> false"],
skip_special_tokens=False
)
# Prepare the prompt
def create_prompt(query, document):
return (
"Determine if the following passage is relevant to the query. "
"Answer only with 'true' or 'false'.\n"
f"Query: {query}\n"
f"Passage: {document}\n"
"<think>"
)
# Example usage
query = "What are the effects of climate change?"
document = "Climate change leads to rising sea levels, extreme weather events, and disruptions to ecosystems. These effects are caused by increasing greenhouse gas concentrations in the atmosphere due to human activities."
# Generate prediction
prompt = create_prompt(query, document)
outputs = model.generate([prompt], sampling_params)
# Extract score
output = outputs[0].outputs[0]
text = output.text
final_logits = output.logprobs[-1]
# Get token IDs for "true" and "false" tokens
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("jhu-clsp/rank1-32b")
true_token = tokenizer(" true", add_special_tokens=False).input_ids[0]
false_token = tokenizer(" false", add_special_tokens=False).input_ids[0]
# Calculate relevance score (probability of "true")
true_logit = final_logits[true_token].logprob
false_logit = final_logits[false_token].logprob
true_score = math.exp(true_logit)
false_score = math.exp(false_logit)
relevance_score = true_score / (true_score + false_score)
print(f"Reasoning chain: {text}")
print(f"Relevance score: {relevance_score}")
Performance
rank1-32b demonstrates strong performance on retrieval benchmarks, particularly on tasks requiring complex reasoning. The model's ability to "think through" relevance decisions makes it especially effective for nuanced topics.
For specific benchmark results and comparisons with other models, please refer to the paper and the official GitHub repository.
Installation
Please see the Github for detailed installation instructions.
MTEB Integration
rank1 is compatible with the MTEB benchmarking framework:\
from mteb import MTEB
from rank1 import rank1 # From the official repo
# Initialize the model
model = rank1(
model_name_or_path="jhu-clsp/rank1-7b",
num_gpus=1,
device="cuda"
)
# Run evaluation on specific tasks
evaluation = MTEB(tasks=["NevIR"])
results = evaluation.run(model)
Citation
If you use rank1 in your research, please cite our work:
@misc{weller2025rank1testtimecomputereranking,
title={Rank1: Test-Time Compute for Reranking in Information Retrieval},
author={Orion Weller and Kathryn Ricci and Eugene Yang and Andrew Yates and Dawn Lawrie and Benjamin Van Durme},
year={2025},
eprint={2502.18418},
archivePrefix={arXiv},
primaryClass={cs.IR},
url={https://arxiv.org/abs/2502.18418},
}