File size: 6,839 Bytes
37277ef cd46688 37277ef e63dba6 37277ef e63dba6 d0fb77a 77958ab d0fb77a e63dba6 37277ef 364dfd0 37277ef 9d4c33b 37277ef 8d8faa8 a7047a4 37277ef cd46688 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 | ---
task_categories:
- question-answering
language:
- en
tags:
- medical
pretty_name: Med_data
size_categories:
- 100K<n<1M
---
# Complete Dataset
Access the complete dataset using the link below:
[Download Dataset](https://fp2s.short.gy/dataset)
# Connect with Me on Happenstance
Join me on Happenstance!
[Click here to add me as a friend](https://happenstance.ai/invite/friend/y5OCIMc4sLNjSuMCFyyVtLxAoYU)
Looking forward to connecting!
For more information or assistance, feel free to contact us at **harryjosh242@gmail.com**.
short_description: Medical datasets for healthcare model training.
---
# **Medical Datasets**
This Medical dataset is crafted as a versatile resource for enthusiasts of data science, machine learning, and data analysis. It replicates the characteristics of real-world healthcare data, offering users a platform to practice, refine, and showcase their data manipulation and analytical skills within the healthcare domain.
## **Potential Uses**
- Building and testing predictive models specific to healthcare.
- Practicing techniques for data cleaning, transformation, and analysis.
- Designing visualizations to uncover insights into healthcare trends.
- Learning and teaching data science and machine learning concepts in a healthcare setting.
## **Acknowledgments**
- This dataset is entirely synthetic, created with a focus on respecting healthcare data privacy and security. It contains no real patient information and complies with privacy regulations.
- The goal is to support advancements in data science and healthcare analytics while inspiring innovative ideas.
## Directory Structure
├── evaluation-medical-instruction-datasets/
│ ├── evaluation-medical-instruction-dataset.json
│ ├── medmcqa-train-instruction-dataset.json
│ ├── medqa-train-instruction-dataset.json
│ └── pubmedqa-train-instruction-train.json
├── general-medical-instruction-datasets/
│ ├── general-medical-instruction-dataset.json
│ ├── GenMedGPT-5k.json
│ ├── HealthCareMagic-100k.json
│ ├── medical_meadow_wikidoc_medical_flashcards.json
│ ├── medical_meadow_wikidoc_patient_info.json
│ └── medicationqa.json
├── medical-preference-data.json
└── medical-pretraining-datasets/
## **Dataset Contents**
### **Evaluation Medical Instruction Datasets**
Contains datasets used for evaluating medical instruction models:
- `evaluation-medical-instruction-dataset.json`
- `medmcqa-train-instruction-dataset.json`
- `medial-train-instruction-dataset.json`
- `pubmedqa-train-instruction-train.json`
### **General Medical Instruction Datasets**
Contains general medical instruction datasets:
- `general-medical-instruction-dataset.json`
- `GenMedGPT-5k.json`
- `HealthCareMagic-100k.json`
- `medical_meadow_wikidoc_medical_flashcards.json`
- `medical_meadow_wikidoc_patient_info.json`
- `medicationqa.json`
### **Medical Preference Data**
- `medical-preference-data.json`: Contains data related to medical preferences.
### **Medical Pretraining Datasets**
Contains datasets used for pretraining medical models.
### **quality_report**
| Total Rows | Missing Data (%) | Duplicate Rows (%) | Duplicate Rate (%) | Outlier Count | File Name | Error |
|--------------|------------------|--------------------|--------------------|---------------|-----------------------------------------------|-------|
| 2,000,000 | 0 | 114 | 0.03 | 0 | evaluation-medical-instruction-dataset.json | NaN |
| 1,400,000 | 0 | 379 | 1.3 | 0 | general-medical-instruction-dataset.json | NaN |
| 27,000 | 0 | 0 | 0 | 0 | GenMedGPT-5k.json | NaN |
| 560,000 | 0 | 0 | 0 | 0 | HealthCareMagic-100k.json | NaN |
| 169,000 | 0 | 427 | 1.26 | 0 | medical_meadow_wikidoc_medical_flashcards.json | NaN |
| 29,000 | 0 | 92 | 1.55 | 0 | medical_meadow_wikidoc_patient_info.json | NaN |
| 50,000 | 0 | 0 | 0 | 0 | medical_meadow_wikidoc.json | NaN |
| 120,000 | 0 | 0 | 0 | 0 | medical-preference-data.json | NaN |
| 2,700 | 0 | 0 | 0 | 0 | medicationqa.json | NaN |
| 910,000 | 0 | 0 | 0 | 0 | medmcqa-train-instruction-dataset.json | NaN |
| 50,000 | 0 | 0 | 0 | 0 | medqa-train-instruction-dataset.json | NaN |
| 120,000 | 0 | 0 | 0 | 0 | open-platypus.json | NaN |
| 400,000 | NaN | NaN | NaN | NaN | pmc_and_guidelines_and_pubmedqa_train.txt | NaN |
| 200,000 | NaN | NaN | NaN | NaN | pmc_and_guidelines.txt | NaN |
| 1,000,000 | 0 | 114 | 0.5 | 0 | pubmedqa-train-instruction-train.json | NaN |
| 250,000 | 0 | 320 | 6.55 | 0 | umls_relation.json | NaN |
| 240,000 | 0 | 0 | 0 | 0 | umls.json | NaN |
## **Usage**
These datasets can be used to train and evaluate models for medical question answering and other related tasks. Each dataset is provided in **JSON** and **TXT** formats and contains various medical instructions, inputs, and outputs.
## Example
Here is an example entry from the `medicationqa.json` dataset:
```json
{
"instruction": "Your role as a doctor requires you to answer the medical questions taking into account the patient's description.",
"input": "how fast does the medication take to work amlodipine",
"output": "After oral administration of therapeutic doses of amlodipine besylate, absorption produces peak plasma concentrations between 6 and 12 hours."
} |