NoMIRACL [EMNLP'24 Findings] is a multilingual relevance assessment dataset for evaluating query & passage relevancy in large language models (LLMs). This is extremely useful in RAG settings, i.e., when a retrieval systems retrieves a subset of passages or documents which either can or cannot be relevant to the user query. The LLM (as the generator) should assess the relevancy and only answer -- if a relevant passage is found within the subset, else abstain from answering.
This repository provides starter code to evaluate diverse multilingual LLMs using our prompt template on NoMIRACL.
For more information, checkout out our publication:
- “Knowing When You Don’t Know”: A Multilingual Relevance Assessment Dataset for Robust Retrieval-Augmented Generation (Thakur et al., ⭐ EMNLP 2024 Findings)
You can install NoMIRACL code repository via pip:
pip install nomiraclIf you want to build from source, use:
$ git clone https://github.com/project-miracl/nomiracl.git
$ cd nomiracl
$ pip install -e .- 50% of relevant examples, 50% of non-relevant, both maximum capped at 250.
- Full example available in sample_load_no_miracl.py.
from nomiracl.dataset import NoMIRACLDataLoader
data_loader = NoMIRACLDataLoader(language = "english",
split = "test", # or 'dev'
hf_dataset_name="miracl/nomiracl",
load_from_huggingface=True)
corpus, queries, qrels = data_loader.load_data_sample(
relevant_ratio = 0.5, non_relevant_ratio = 0.5, max_sample_pool = 250)- Full example available in sample_model_generation.py.
from nomiracl.generation.utils import load_model
model_name = "meta-llama/Meta-Llama-3-8B-Instruct"
# List of techniques supported in nomiracl:
# huggingface (GPUs), vllm (GPUs), cohere (API), openai (API), nvidia (API), azure (API), anyscale (API)
# `cohere` requires COHERE_API_KEY, `openai` requires OPENAI_API_KEY, `nvidia` requires NVIDIA_API_KEY
# `azure` requires AZURE_OPENAI_API_BASE, AZURE_OPENAI_API_VERSION and AZURE_OPENAI_API_KEY
# `anyscale` requires ANYSCALE_BASE_URL and ANYSCALE_API_KEY.
technique = "vllm" # or huggingface or nvidia, anyscale etc.
model = load_model(
technique, # technique
model_name, # model_name
cache_dir="<your-cache-dir>", # extra kwargs
batch_size=2, # extra kwargs
num_gpus=1, # extra kwargs
concurrency=2 # extra kwargs
)
# Sample prompts
prompts = [
"What is the capital of France?",
"What is the capital of Germany?",
"What is the capital of Italy?",
]
model_results = model.call(prompts)
for prompt, result in zip(prompts, model_results):
print("Prompt: {}".format(prompt))
print("{} result: {}".format(model_name, result))- Full example available in sample_vanilla_prompt_exploration.py.
from nomiracl.prompts.utils import load_prompt_template
# Options include: vanilla, role, repeat, explanation
prompt_cls = load_prompt_template("vanilla", count = 10) # as we include 10 passages
query = "Which is the best programming language?"
passages = [
"Python is the best programming language.",
"Javascript is the best programming language.",
"Go is the best programming language.",
"Java is the best programming language.",
"C# is the best programming language.",
"Ruby is the best programming language.",
"R is the best programming language.",
"C++ is the best programming language.",
"C is the best programming language.",
"Rust is the best programming language.",
]
prompt = prompt_cls(query=query, passages=passages)Or you can provide your own custom prompt template by modifying the self.template in nomiracl.VanillaTemplate.
from nomiracl.prompts import VanillaTemplate
class CustomTemplate(VanillaTemplate):
def __init__(self, count: int = 1):
super().__init__(count)
self.template = (
"This is a pairwise prompt template. Respond as either "{self.answer}" or "{self.no_answer}".'
+ "\n\nQUESTION:\n{query}\n\n"
+ "CONTEXT:\n"
+ "\n\n".join(
[
"[{}] {}".format(i, "{" + passage + "}")
for i, passage in enumerate(self.passage_variables, 1)
]
)
+ "\n\nOUTPUT:\n"
)The NoMIRACL dataset is available in HuggingFace under: miracl/nomiracl.
Languages covered: Arabic (ar), Bengali (bn), German (de), English (en), Spanish (es), Persian (fa), Finnish (fi), French (fr), Hindi (hi), Indonesian (id), Japanese (ja), Korean (ko), Russian (ru), Swahili (sw), Thai (th), Yoruba (yo), Chinese (zh).
HuggingFace Page: https://huggingface.co/datasets/miracl/nomiracl
import datasets
language = 'german' # or any of the 18 languages
subset = 'relevant' # or 'non_relevant'
split = 'test' # or 'dev' for development split
# four combinations available: 'dev.relevant', 'dev.non_relevant', 'test.relevant' and 'test.non_relevant'
nomiracl = datasets.load_dataset('miracl/nomiracl', language, split=f'{split}.{subset}')Baseline results (250 queries) are available within the repository under ./results/baselines/non_relevant.
An example datapoint under ./results/baselines/non_relevant/en.test.vanilla_prompt.jsonl
{
"query_id": "842558#0",
"docids": ["2842207#5", "7004944#45", "3310762#14", "47220460#1", "36451733#7", "3310762#20", "4724576#4", "22373402#0", "52203230#0", "23126218#4"],
"prompt": "I will give you a question and several contexts containing information about the question. [ ... ] \n\nOUTPUT:\n",
"template": "vanilla",
"results": {"gpt-4-azure": "Yes, answer is present.",
"llama-2-13b-chat": "\nYes, answer is present in [6].\n\nNo answers found in the other contexts.",
[...]
"aya-101": "Wales"}
}
Baseline results (250 queries) are available within the repository under ./results/baselines/relevant.
An example datapoint under ./results/baselines/relevant/en.test.vanilla_prompt.jsonl
{
"query_id": "8706103#0",
"docids": ["42057469#2", "4998067#1", "29247933#0", "162619#81", "422315#13", "26790310#4", "41298602#18", "22816#16", "123427#61", "23576525#0"],
"prompt": "I will give you a question and several contexts containing information about the question. [ ... ] \n\nQUESTION:\nWhat is the course that will be discontinued as defined by the National Education Policy? [ ... ] \n\nOUTPUT:\n",
"template": "vanilla",
"results": {"gpt-4-azure": "I don't know.",
"llama-2-13b-chat": "Please answer the question based on the given contexts.",
[...]
"aya-101": "I don't know"}
}
NoMIRACL is a multilingual dataset designed to evaluate LLM robustness in relevance assessment to help avoid errors in first-stage retrieval. The dataset covers 18 typologically diverse languages and includes two subsets: non-relevant and relevant.
- Queries with no-known answers within the retrieved oracle passages.
- All top-k passages manually judged as non-relevant (relevancy score = 0).
- Queries with known answers within the retrieved oracle passages.
- At least one of the top-k passages manually judged as relevant (relevancy score = 1).
We conduct a robustness evaluation using a binary classification task, comparing LLM predictions against the ground truth provided in NoMIRACL. The metrics used are hallucination rate and error rate.
-
Hallucination Rate:
FP/(FP + TN)Measures the model's tendency to hallucinate an answer when no answer is present in the non-relevant subset. -
Error Rate:
FN/(FN + TP)Measures the model's inaccuracy in recognizing relevant passages in the relevant subset.
The NoMIRACL dataset has been made possible due to a collaborative effort of the following universities and organizations:
- University of Waterloo
- Huawei Noah's Ark Lab
Parts of the NoMIRACL code structure has been inspired by:
If you use NoMIRACL or parts in a research paper, please cite our work as follows:
@article{thakur:2024,
author = {Nandan Thakur and
Luiz Bonifacio and
Xinyu Zhang and
Odunayo Ogundepo and
Ehsan Kamalloo and
David Alfonso{-}Hermelo and
Xiaoguang Li and
Qun Liu and
Boxing Chen and
Mehdi Rezagholizadeh and
Jimmy Lin},
title = {NoMIRACL: Knowing When You Don't Know for Robust Multilingual Retrieval-Augmented
Generation},
journal = {CoRR},
volume = {abs/2312.11361},
year = {2023},
url = {https://doi.org/10.48550/arXiv.2312.11361},
doi = {10.48550/ARXIV.2312.11361},
eprinttype = {arXiv},
eprint = {2312.11361},
timestamp = {Tue, 16 Jan 2024 11:57:42 +0100},
biburl = {https://dblp.org/rec/journals/corr/abs-2312-11361.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
Contact person: Nandan Thakur, [email protected]
This repository contains experimental software and is published for the sole purpose of giving additional background details on the respective publication.