Datasets:
Upload folder using huggingface_hub
Browse files
README.md
CHANGED
@@ -88,11 +88,27 @@ Langchain's RecursiveCharacterTextSplitter is used for chunking and the most com
|
|
88 |
are replaced with readable equivalents. The textbooks are then broken into separate subsets, indicated below along with
|
89 |
the textbooks they comprise:
|
90 |
|
91 |
-
-
|
92 |
- Anatomy_Gray.txt, First_Aid_Step1.txt, First_Aid_Step2.txt, Immunology_Janeway.txt, InternalMed_Harrison.txt, Neurology_Adams.txt, Obstentrics_Williams.txt, Pathoma_Husain.txt, Pediatrics_Nelson.txt, and Surgery_Schwartz.txt
|
93 |
-
-
|
94 |
- Biochemistry_Lippincott.txt, Cell_Biology_Alberts.txt, Histology_Ross.txt, Pathology_Robbins.txt, and Physiology_Levy.txt
|
95 |
-
- _*
|
96 |
- Pharmacology_Katzung.txt
|
97 |
-
- _*
|
98 |
-
- Psichiatry_DSM-5.txt
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
88 |
are replaced with readable equivalents. The textbooks are then broken into separate subsets, indicated below along with
|
89 |
the textbooks they comprise:
|
90 |
|
91 |
+
- Core Clinical Medicine (_*core_clinical*_)
|
92 |
- Anatomy_Gray.txt, First_Aid_Step1.txt, First_Aid_Step2.txt, Immunology_Janeway.txt, InternalMed_Harrison.txt, Neurology_Adams.txt, Obstentrics_Williams.txt, Pathoma_Husain.txt, Pediatrics_Nelson.txt, and Surgery_Schwartz.txt
|
93 |
+
- Basic Biology (_*basic_biology*_)
|
94 |
- Biochemistry_Lippincott.txt, Cell_Biology_Alberts.txt, Histology_Ross.txt, Pathology_Robbins.txt, and Physiology_Levy.txt
|
95 |
+
- Pharmacology (_*pharmacology*_)
|
96 |
- Pharmacology_Katzung.txt
|
97 |
+
- Psychiatry (_*psychiatry*_)
|
98 |
+
- Psichiatry_DSM-5.txt
|
99 |
+
|
100 |
+
So, you can load the basic biology subset of the corpus via:
|
101 |
+
|
102 |
+
```python
|
103 |
+
In [1]: import datasets
|
104 |
+
In [2]: ds = datasets.load_dataset('cogbuji/medqa_corpus_en', 'basic_biology')
|
105 |
+
In [3]: ds
|
106 |
+
Generating train split: 50386 examples [00:00, 92862.56 examples/s]
|
107 |
+
Out[3]:
|
108 |
+
DatasetDict({
|
109 |
+
train: Dataset({
|
110 |
+
features: ['text', 'source'],
|
111 |
+
num_rows: 50386
|
112 |
+
})
|
113 |
+
})
|
114 |
+
```
|