Merge pull request #1255 from 545999961/master

update examples

research/llm_dense_retriever/examples/bge-en-icl/MTEB/AmazonReviewsClassification.json

@@ -1,22 +1,22 @@
 [
     {
-        "query": "DO NOT ORDER THIS\n\nThis isn't what's described at all. Taking it out of the package lace was cut upon arrival, wig was cut to like 14 inch, not curly, and smelled like cigarettes. I obviously was sent what someone returned, disgusting.Not what I ordered at all, not pleased at all. I want my money back DO NOT ORDER",
+        "query": "DO NOT ORDER THIS\\n\\nThis isn't what's described at all. Taking it out of the package lace was cut upon arrival, wig was cut to like 14 inch, not curly, and smelled like cigarettes. I obviously was sent what someone returned, disgusting.Not what I ordered at all, not pleased at all. I want my money back DO NOT ORDER",
         "response": "1 star"
     },
     {
-        "query": "And I can\u2019t return it\n\nThis product seemed like good quality but it does not stay stuck to the soles at all. You walk a few steps and then you find the black shoe grip somewhere on the floor.",
+        "query": "And I can\u2019t return it\\n\\nThis product seemed like good quality but it does not stay stuck to the soles at all. You walk a few steps and then you find the black shoe grip somewhere on the floor.",
         "response": "2 star"
     },
     {
-        "query": "Three Stars\n\nnew yearly subscription plan is horrible, but the product still works as it did in the past",
+        "query": "Three Stars\\n\\nnew yearly subscription plan is horrible, but the product still works as it did in the past",
         "response": "3 star"
     },
     {
-        "query": "I like how it has lots of pockets to put stuff ...\n\nI like how it has lots of pockets to put stuff in. I would have liked to have a shorter securing strap so it would not slide around so much. Good product.",
+        "query": "I like how it has lots of pockets to put stuff ...\\n\\nI like how it has lots of pockets to put stuff in. I would have liked to have a shorter securing strap so it would not slide around so much. Good product.",
         "response": "4 star"
     },
     {
-        "query": "Great\n\nIt is really good. That's my favorite. THANK YOU",
+        "query": "Great\\n\\nIt is really good. That's my favorite. THANK YOU",
         "response": "5 star"
     }
 ]

research/llm_dense_retriever/examples/bge-en-icl/MTEB/BiorxivClusteringP2P.json

@@ -8,7 +8,7 @@
         "response": "microbiology"
     },
     {
-        "query": "Deep Learning Approaches for Predicting Protein-Protein Interactions from Sequence Data\nProtein-protein interactions (PPIs) are fundamental to numerous biological processes, and understanding these interactions is critical for uncovering cellular mechanisms and developing therapeutic strategies. Traditional experimental methods for identifying PPIs are labor-intensive and time-consuming, highlighting the need for computational approaches. In this study, we present DeepPPI, a deep learning-based framework designed to predict PPIs directly from protein sequence data. DeepPPI employs a combination of convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to capture both local and global sequence features. We trained DeepPPI on a comprehensive dataset of known PPIs and benchmarked its performance against existing methods, demonstrating superior accuracy and generalizability. Additionally, we applied DeepPPI to predict novel interactions in the human proteome and validated a subset of these predictions experimentally. Our results indicate that DeepPPI not only achieves high prediction accuracy but also provides insights into the structural and functional basis of protein interactions, making it a valuable tool for the bioinformatics community.",
+        "query": "Deep Learning Approaches for Predicting Protein-Protein Interactions from Sequence Data\\nProtein-protein interactions (PPIs) are fundamental to numerous biological processes, and understanding these interactions is critical for uncovering cellular mechanisms and developing therapeutic strategies. Traditional experimental methods for identifying PPIs are labor-intensive and time-consuming, highlighting the need for computational approaches. In this study, we present DeepPPI, a deep learning-based framework designed to predict PPIs directly from protein sequence data. DeepPPI employs a combination of convolutional neural networks (CNNs) and recurrent neural networks (RNNs) to capture both local and global sequence features. We trained DeepPPI on a comprehensive dataset of known PPIs and benchmarked its performance against existing methods, demonstrating superior accuracy and generalizability. Additionally, we applied DeepPPI to predict novel interactions in the human proteome and validated a subset of these predictions experimentally. Our results indicate that DeepPPI not only achieves high prediction accuracy but also provides insights into the structural and functional basis of protein interactions, making it a valuable tool for the bioinformatics community.",
         "response": "bioinformatics"
     },
     {

research/llm_dense_retriever/examples/bge-en-icl/MTEB/MedrxivClusteringP2P.json

@@ -1,10 +1,10 @@
 [
     {
-        "query": "Socioeconomic Disparities in COVID-19 Transmission Risk: A Population-Based Study from Norway\nObjective: Explore socioeconomic disparities in COVID-19 transmission risk across occupational categories in Norway.\nMethods: Analyzed data from 3,559,694 residents aged 20-70 using the International Standard Classification of Occupations (ISCO-08). Logistic regression models adjusted for various factors examined the association between occupation and SARS-CoV-2 infection risk and hospitalization during different pandemic phases.\nResults: Occupations with varying socioeconomic statuses showed different COVID-19 infection risks. Healthcare professionals had higher odds during the initial wave, while service workers had increased odds during later waves. Teachers and administrative personnel also had moderate risk increases. Occupation had limited association with hospitalization after adjusting for confounders.\nConclusion: Socioeconomic factors significantly influence COVID-19 transmission in occupational settings. Targeted public health interventions addressing workplace conditions, testing accessibility, and socioeconomic vulnerability are essential for mitigating future pandemic impacts and developing equitable pandemic preparedness strategies.\nKeywords: COVID-19, Socioeconomic Disparities, Occupational Risk, Pandemic Preparedness, Public Health, Norway, ISCO-08, SARS-CoV-2",
+        "query": "Socioeconomic Disparities in COVID-19 Transmission Risk: A Population-Based Study from Norway\\nObjective: Explore socioeconomic disparities in COVID-19 transmission risk across occupational categories in Norway.\\nMethods: Analyzed data from 3,559,694 residents aged 20-70 using the International Standard Classification of Occupations (ISCO-08). Logistic regression models adjusted for various factors examined the association between occupation and SARS-CoV-2 infection risk and hospitalization during different pandemic phases.\\nResults: Occupations with varying socioeconomic statuses showed different COVID-19 infection risks. Healthcare professionals had higher odds during the initial wave, while service workers had increased odds during later waves. Teachers and administrative personnel also had moderate risk increases. Occupation had limited association with hospitalization after adjusting for confounders.\\nConclusion: Socioeconomic factors significantly influence COVID-19 transmission in occupational settings. Targeted public health interventions addressing workplace conditions, testing accessibility, and socioeconomic vulnerability are essential for mitigating future pandemic impacts and developing equitable pandemic preparedness strategies.\\nKeywords: COVID-19, Socioeconomic Disparities, Occupational Risk, Pandemic Preparedness, Public Health, Norway, ISCO-08, SARS-CoV-2",
         "response": "infectious diseases"
     },
     {
-        "query": "Assessing Socioeconomic Determinants of Infectious Disease Spread: A Cross-National Analysis Using Machine Learning Approaches\nBackground: Understanding socioeconomic factors influencing infectious disease transmission is crucial for targeted public health interventions.\nMethods: This study uses machine learning techniques and Bayesian optimization to analyze the impact of socioeconomic variables such as income, education, and healthcare access on disease dynamics. It integrates datasets on disease transmission and socio-demographic characteristics.\nResults: Significant associations between socioeconomic indicators and infectious disease spread were found, highlighting disparities in vulnerability and transmission rates.\nConclusion: Advanced analytical techniques provide nuanced insights into the socioeconomic determinants of disease transmission, aiding evidence-based policymaking to reduce health disparities and enhance epidemic preparedness.\nKeywords: Socioeconomic Determinants, Infectious Disease, Machine Learning, Public Health, Epidemiology, Health Disparities, Bayesian Optimization",
+        "query": "Assessing Socioeconomic Determinants of Infectious Disease Spread: A Cross-National Analysis Using Machine Learning Approaches\\nBackground: Understanding socioeconomic factors influencing infectious disease transmission is crucial for targeted public health interventions.\\nMethods: This study uses machine learning techniques and Bayesian optimization to analyze the impact of socioeconomic variables such as income, education, and healthcare access on disease dynamics. It integrates datasets on disease transmission and socio-demographic characteristics.\\nResults: Significant associations between socioeconomic indicators and infectious disease spread were found, highlighting disparities in vulnerability and transmission rates.\\nConclusion: Advanced analytical techniques provide nuanced insights into the socioeconomic determinants of disease transmission, aiding evidence-based policymaking to reduce health disparities and enhance epidemic preparedness.\\nKeywords: Socioeconomic Determinants, Infectious Disease, Machine Learning, Public Health, Epidemiology, Health Disparities, Bayesian Optimization",
         "response": "epidemiology"
     },
     {
@@ -16,7 +16,7 @@
         "response": "genetic and genomic medicine"
     },
     {
-        "query": "Longitudinal Analysis of Sleep Disturbances and Cognitive Decline in Older Adults: A 5-Year Prospective Cohort Study Background: Sleep disturbances in older adults are a recognized risk factor for cognitive decline. This study examines their impact on cognitive function over five years.\nMethods: 3,200 participants aged 60+ from Karnataka, India, were assessed annually using sleep questionnaires and cognitive tests. Exclusions included major neuropsychiatric disorders.\nResults: 25% reported sleep disturbances at baseline; 30% developed mild cognitive impairment, and 15% progressed to dementia. Insomnia and sleep apnea significantly accelerated cognitive decline. CPAP for sleep apnea showed modest protective effects.\nConclusion: Addressing sleep disturbances is crucial for mitigating cognitive decline in older adults.",
+        "query": "Longitudinal Analysis of Sleep Disturbances and Cognitive Decline in Older Adults: A 5-Year Prospective Cohort Study Background: Sleep disturbances in older adults are a recognized risk factor for cognitive decline. This study examines their impact on cognitive function over five years.\\nMethods: 3,200 participants aged 60+ from Karnataka, India, were assessed annually using sleep questionnaires and cognitive tests. Exclusions included major neuropsychiatric disorders.\\nResults: 25% reported sleep disturbances at baseline; 30% developed mild cognitive impairment, and 15% progressed to dementia. Insomnia and sleep apnea significantly accelerated cognitive decline. CPAP for sleep apnea showed modest protective effects.\\nConclusion: Addressing sleep disturbances is crucial for mitigating cognitive decline in older adults.",
         "response": "neurology"
     }
 ]