Spaces:
Running
Running
Adding contribute tutorial
Browse files- CONTRIBUTING.md +43 -0
- app.py +30 -45
- notebooks/eda.json +80 -0
- notebooks/embeddings.json +67 -0
- notebooks/finetuning.json +6 -0
- notebooks/rag.json +6 -0
- utils/notebook_utils.py +19 -1
CONTRIBUTING.md
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# How to contribute
|
2 |
+
|
3 |
+
Thanks for your interest in contributing! 🙌 This project helps users quickly create notebooks to showcase how they use datasets. The generated code can be added to their repositories or used in research projects.
|
4 |
+
|
5 |
+
## Ways to Contribute
|
6 |
+
There are a few ways you can help:
|
7 |
+
|
8 |
+
- 💡**Share ideas**: Got a cool feature in mind? Let us know!
|
9 |
+
- 🐞**Report bugs**: If something isn’t working, we’d love to fix it.
|
10 |
+
- 🚀**Suggest improvements**: Any suggestions to make the tool better are welcome.
|
11 |
+
- 📓**Add new notebook types**: This is one of the most exciting ways to contribute!
|
12 |
+
|
13 |
+
## For Ideas, Bugs, or Suggestions:
|
14 |
+
- Start a new discussion [here](https://huggingface.co/spaces/asoria/auto-notebook-creator/discussions/new).
|
15 |
+
- Tell me what’s on your mind and include any details that might help.
|
16 |
+
|
17 |
+
## For Adding New Notebook Types:
|
18 |
+
- Open a pull request (PR) [here](https://huggingface.co/spaces/asoria/auto-notebook-creator/discussions?new_pr=true).
|
19 |
+
- Add a new `.json` file in the notebooks folder. There’s a sample file you can copy and tweak.
|
20 |
+
- Submit your PR! 🎉
|
21 |
+
|
22 |
+
## Running the Space Application
|
23 |
+
To execute the space, follow these steps:
|
24 |
+
|
25 |
+
1. Set Required Environment Variables:
|
26 |
+
- `NOTEBOOKS_REPOSITORY`: The name of the repository where the generated notebooks will be stored. Ensure that you have **write** permissions for this repository. For example, I use [asoria/dataset-notebook-creator-content](https://huggingface.co/datasets/asoria/dataset-notebook-creator-content) repository.
|
27 |
+
- `HF_TOKEN`: Your Hugging Face token, used for authentication to push changes to the repository.
|
28 |
+
|
29 |
+
Example setup:
|
30 |
+
|
31 |
+
```bash
|
32 |
+
export HF_TOKEN=your_huggingface_token
|
33 |
+
export NOTEBOOKS_REPOSITORY=your_repository_name
|
34 |
+
```
|
35 |
+
|
36 |
+
2. Execute the following command to start the application:
|
37 |
+
|
38 |
+
```bash
|
39 |
+
python app.py
|
40 |
+
```
|
41 |
+
|
42 |
+
I am excited to see what you come up with. Thanks for helping make this project even better! 💖
|
43 |
+
|
app.py
CHANGED
@@ -6,25 +6,23 @@ from httpx import Client
|
|
6 |
import logging
|
7 |
import pandas as pd
|
8 |
from utils.notebook_utils import (
|
9 |
-
eda_cells,
|
10 |
replace_wildcards,
|
11 |
-
|
12 |
-
embeggins_cells,
|
13 |
)
|
14 |
from dotenv import load_dotenv
|
15 |
import os
|
16 |
from nbconvert import HTMLExporter
|
17 |
|
18 |
-
# TODOs:
|
19 |
-
# Improve UI code preview
|
20 |
-
# Add template for training
|
21 |
-
|
22 |
load_dotenv()
|
23 |
|
24 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
25 |
assert HF_TOKEN is not None, "You need to set HF_TOKEN in your environment variables"
|
26 |
|
27 |
-
NOTEBOOKS_REPOSITORY = "
|
|
|
|
|
|
|
|
|
28 |
|
29 |
BASE_DATASETS_SERVER_URL = "https://datasets-server.huggingface.co"
|
30 |
HEADERS = {"Accept": "application/json", "Content-Type": "application/json"}
|
@@ -96,18 +94,6 @@ def longest_string_column(df):
|
|
96 |
return longest_col
|
97 |
|
98 |
|
99 |
-
def generate_eda_cells(dataset_id):
|
100 |
-
return generate_cells(dataset_id, eda_cells, "eda")
|
101 |
-
|
102 |
-
|
103 |
-
def generate_rag_cells(dataset_id):
|
104 |
-
return generate_cells(dataset_id, rag_cells, "rag")
|
105 |
-
|
106 |
-
|
107 |
-
def generate_embedding_cells(dataset_id):
|
108 |
-
return generate_cells(dataset_id, embeggins_cells, "embeddings")
|
109 |
-
|
110 |
-
|
111 |
def _push_to_hub(
|
112 |
dataset_id,
|
113 |
notebook_file,
|
@@ -129,8 +115,15 @@ def _push_to_hub(
|
|
129 |
raise
|
130 |
|
131 |
|
132 |
-
|
133 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
134 |
try:
|
135 |
libraries = get_compatible_libraries(dataset_id)
|
136 |
except Exception as err:
|
@@ -161,6 +154,7 @@ def generate_cells(dataset_id, cells, notebook_type="eda"):
|
|
161 |
has_numeric_columns = len(df.select_dtypes(include=["number"]).columns) > 0
|
162 |
has_categoric_columns = len(df.select_dtypes(include=["object"]).columns) > 0
|
163 |
|
|
|
164 |
if notebook_type in ("rag", "embeddings") and not has_categoric_columns:
|
165 |
logging.error(
|
166 |
"Dataset does not have categorical columns, which are required for RAG generation."
|
@@ -250,34 +244,25 @@ with gr.Blocks(
|
|
250 |
|
251 |
gr.Markdown("## 2. Select the type of notebook you want to generate")
|
252 |
with gr.Row():
|
253 |
-
|
254 |
-
|
255 |
-
|
256 |
-
|
257 |
-
|
258 |
-
|
259 |
-
|
260 |
-
|
|
|
|
|
|
|
261 |
|
262 |
gr.Markdown("## 3. Notebook code result")
|
263 |
code_component = gr.HTML(elem_id="box")
|
264 |
go_to_notebook = gr.Markdown("", visible=True)
|
265 |
|
266 |
-
|
267 |
-
|
268 |
-
inputs=[dataset_name],
|
269 |
-
outputs=[code_component, go_to_notebook],
|
270 |
-
)
|
271 |
-
|
272 |
-
generate_embedding_btn.click(
|
273 |
-
generate_embedding_cells,
|
274 |
-
inputs=[dataset_name],
|
275 |
-
outputs=[code_component, go_to_notebook],
|
276 |
-
)
|
277 |
-
|
278 |
-
generate_rag_btn.click(
|
279 |
-
generate_rag_cells,
|
280 |
-
inputs=[dataset_name],
|
281 |
outputs=[code_component, go_to_notebook],
|
282 |
)
|
283 |
|
|
|
6 |
import logging
|
7 |
import pandas as pd
|
8 |
from utils.notebook_utils import (
|
|
|
9 |
replace_wildcards,
|
10 |
+
load_json_files_from_folder,
|
|
|
11 |
)
|
12 |
from dotenv import load_dotenv
|
13 |
import os
|
14 |
from nbconvert import HTMLExporter
|
15 |
|
|
|
|
|
|
|
|
|
16 |
load_dotenv()
|
17 |
|
18 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
19 |
assert HF_TOKEN is not None, "You need to set HF_TOKEN in your environment variables"
|
20 |
|
21 |
+
NOTEBOOKS_REPOSITORY = os.getenv("NOTEBOOKS_REPOSITORY")
|
22 |
+
assert (
|
23 |
+
NOTEBOOKS_REPOSITORY is not None
|
24 |
+
), "You need to set NOTEBOOKS_REPOSITORY in your environment variables"
|
25 |
+
|
26 |
|
27 |
BASE_DATASETS_SERVER_URL = "https://datasets-server.huggingface.co"
|
28 |
HEADERS = {"Accept": "application/json", "Content-Type": "application/json"}
|
|
|
94 |
return longest_col
|
95 |
|
96 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
97 |
def _push_to_hub(
|
98 |
dataset_id,
|
99 |
notebook_file,
|
|
|
115 |
raise
|
116 |
|
117 |
|
118 |
+
folder_path = "notebooks"
|
119 |
+
notebook_templates = load_json_files_from_folder(folder_path)
|
120 |
+
logging.info(f"Available notebooks {notebook_templates.keys()}")
|
121 |
+
|
122 |
+
|
123 |
+
def generate_cells(dataset_id, notebook_title):
|
124 |
+
logging.info(f"Generating {notebook_title} notebook for dataset {dataset_id}")
|
125 |
+
cells = notebook_templates[notebook_title]["notebook_template"]
|
126 |
+
notebook_type = notebook_templates[notebook_title]["notebook_type"]
|
127 |
try:
|
128 |
libraries = get_compatible_libraries(dataset_id)
|
129 |
except Exception as err:
|
|
|
154 |
has_numeric_columns = len(df.select_dtypes(include=["number"]).columns) > 0
|
155 |
has_categoric_columns = len(df.select_dtypes(include=["object"]).columns) > 0
|
156 |
|
157 |
+
# TODO: Validate by notebook type
|
158 |
if notebook_type in ("rag", "embeddings") and not has_categoric_columns:
|
159 |
logging.error(
|
160 |
"Dataset does not have categorical columns, which are required for RAG generation."
|
|
|
244 |
|
245 |
gr.Markdown("## 2. Select the type of notebook you want to generate")
|
246 |
with gr.Row():
|
247 |
+
notebook_type = gr.Dropdown(
|
248 |
+
choices=notebook_templates.keys(), label="Notebook type"
|
249 |
+
)
|
250 |
+
generate_button = gr.Button("Generate Notebook", variant="primary")
|
251 |
+
contribute_btn = gr.Button(
|
252 |
+
"Or Contribute",
|
253 |
+
visible=True,
|
254 |
+
variant="secondary",
|
255 |
+
size="sm",
|
256 |
+
link="https://huggingface.co/spaces/asoria/auto-notebook-creator/blob/main/CONTRIBUTING.md",
|
257 |
+
)
|
258 |
|
259 |
gr.Markdown("## 3. Notebook code result")
|
260 |
code_component = gr.HTML(elem_id="box")
|
261 |
go_to_notebook = gr.Markdown("", visible=True)
|
262 |
|
263 |
+
generate_button.click(
|
264 |
+
generate_cells,
|
265 |
+
inputs=[dataset_name, notebook_type],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
266 |
outputs=[code_component, go_to_notebook],
|
267 |
)
|
268 |
|
notebooks/eda.json
ADDED
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"notebook_title": "Exploratory data analysis (EDA)",
|
3 |
+
"notebook_type": "eda",
|
4 |
+
"dataset_type": "numeric",
|
5 |
+
"notebook_template": [
|
6 |
+
{
|
7 |
+
"cell_type": "markdown",
|
8 |
+
"source": "\n---\n# **Exploratory Data Analysis (EDA) Notebook for {dataset_name} dataset**\n---\n"
|
9 |
+
},
|
10 |
+
{
|
11 |
+
"cell_type": "markdown",
|
12 |
+
"source": "## 1. Setup necessary libraries and load the dataset"
|
13 |
+
},
|
14 |
+
{
|
15 |
+
"cell_type": "code",
|
16 |
+
"source": "\n# Install and import necessary libraries.\n!pip install pandas matplotlib seaborn\n"
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"cell_type": "code",
|
20 |
+
"source": "\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n"
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"cell_type": "code",
|
24 |
+
"source": "\n# Load the dataset as a DataFrame\n{first_code}\n"
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"cell_type": "markdown",
|
28 |
+
"source": "## 2. Understanding the Dataset"
|
29 |
+
},
|
30 |
+
{
|
31 |
+
"cell_type": "code",
|
32 |
+
"source": "\n# First rows of the dataset and info\nprint(df.head())\nprint(df.info())\n"
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"cell_type": "code",
|
36 |
+
"source": "\n# Check for missing values\nprint(df.isnull().sum())\n"
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"cell_type": "code",
|
40 |
+
"source": "\n# Identify data types of each column\nprint(df.dtypes)\n"
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"cell_type": "code",
|
44 |
+
"source": "\n# Detect duplicated rows\nprint(df.duplicated().sum())\n"
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"cell_type": "code",
|
48 |
+
"source": "\n# Generate descriptive statistics\nprint(df.describe())\n"
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"type": "categoric",
|
52 |
+
"cell_type": "code",
|
53 |
+
"source": "\n# Unique values in categorical columns\ndf.select_dtypes(include=['object']).nunique()\n"
|
54 |
+
},
|
55 |
+
{
|
56 |
+
"cell_type": "markdown",
|
57 |
+
"source": "## 3. Data Visualization"
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"type": "numeric",
|
61 |
+
"cell_type": "code",
|
62 |
+
"source": "\n# Correlation matrix for numerical columns\ncorr_matrix = df.corr(numeric_only=True)\nplt.figure(figsize=(10, 8))\nsns.heatmap(corr_matrix, annot=True, fmt='.2f', cmap='coolwarm', square=True)\nplt.title('Correlation Matrix')\nplt.show()\n"
|
63 |
+
},
|
64 |
+
{
|
65 |
+
"type": "numeric",
|
66 |
+
"cell_type": "code",
|
67 |
+
"source": "\n# Distribution plots for numerical columns\nfor column in df.select_dtypes(include=['int64', 'float64']).columns:\n plt.figure(figsize=(8, 4))\n sns.histplot(df[column], kde=True)\n plt.title(f'Distribution of {column}')\n plt.xlabel(column)\n plt.ylabel('Frequency')\n plt.show()\n"
|
68 |
+
},
|
69 |
+
{
|
70 |
+
"type": "categoric",
|
71 |
+
"cell_type": "code",
|
72 |
+
"source": "\n# Count plots for categorical columns\nfor column in df.select_dtypes(include=['object']).columns:\n plt.figure(figsize=(8, 4))\n sns.countplot(x=column, data=df)\n plt.title(f'Count Plot of {column}')\n plt.xlabel(column)\n plt.ylabel('Count')\n plt.show()\n"
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"type": "numeric",
|
76 |
+
"cell_type": "code",
|
77 |
+
"source": "\n# Box plots for detecting outliers in numerical columns\nfor column in df.select_dtypes(include=['int64', 'float64']).columns:\n plt.figure(figsize=(8, 4))\n sns.boxplot(df[column])\n plt.title(f'Box Plot of {column}')\n plt.xlabel(column)\n plt.show()\n"
|
78 |
+
}
|
79 |
+
]
|
80 |
+
}
|
notebooks/embeddings.json
ADDED
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"notebook_title": "Text Embeddings",
|
3 |
+
"notebook_type": "embeddings",
|
4 |
+
"dataset_type": "text",
|
5 |
+
"notebook_template": [
|
6 |
+
{
|
7 |
+
"cell_type": "markdown",
|
8 |
+
"source": "---\n# **Embeddings Notebook for {dataset_name} dataset**\n---"
|
9 |
+
},
|
10 |
+
{
|
11 |
+
"cell_type": "markdown",
|
12 |
+
"source": "## 1. Setup necessary libraries and load the dataset"
|
13 |
+
},
|
14 |
+
{
|
15 |
+
"cell_type": "code",
|
16 |
+
"source": "# Install and import necessary libraries.\n!pip install pandas sentence-transformers faiss-cpu "
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"cell_type": "code",
|
20 |
+
"source": "from sentence_transformers import SentenceTransformer\nimport faiss"
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"cell_type": "code",
|
24 |
+
"source": "# Load the dataset as a DataFrame\n{first_code}"
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"cell_type": "code",
|
28 |
+
"source": "# Specify the column name that contains the text data to generate embeddings\ncolumn_to_generate_embeddings = '{longest_col}'"
|
29 |
+
},
|
30 |
+
{
|
31 |
+
"cell_type": "markdown",
|
32 |
+
"source": "## 2. Loading embedding model and creating FAISS index"
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"cell_type": "code",
|
36 |
+
"source": "# Remove duplicate entries based on the specified column\ndf = df.drop_duplicates(subset=column_to_generate_embeddings)"
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"cell_type": "code",
|
40 |
+
"source": "# Convert the column data to a list of text entries\ntext_list = df[column_to_generate_embeddings].tolist()"
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"cell_type": "code",
|
44 |
+
"source": "# Specify the embedding model you want to use\nmodel = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')"
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"cell_type": "code",
|
48 |
+
"source": "vectors = model.encode(text_list)\nvector_dimension = vectors.shape[1]\n\n# Initialize the FAISS index with the appropriate dimension (384 for this model)\nindex = faiss.IndexFlatL2(vector_dimension)\n\n# Encode the text list into embeddings and add them to the FAISS index\nindex.add(vectors)"
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"cell_type": "markdown",
|
52 |
+
"source": "## 3. Perform a text search"
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"cell_type": "code",
|
56 |
+
"source": "# Specify the text you want to search for in the list\ntext_to_search = text_list[0]\nprint(f\"Text to search: {text_to_search}\")"
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"cell_type": "code",
|
60 |
+
"source": "# Generate the embedding for the search query\nquery_embedding = model.encode([text_to_search])"
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"cell_type": "code",
|
64 |
+
"source": "# Perform the search to find the 'k' nearest neighbors (adjust 'k' as needed)\nD, I = index.search(query_embedding, k=10)\n\n# Print the similar documents\nprint(f\"Similar documents: {[text_list[i] for i in I[0]]}\")"
|
65 |
+
}
|
66 |
+
]
|
67 |
+
}
|
notebooks/finetuning.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"notebook_title": "Supervised fine-tuning (SFT)",
|
3 |
+
"notebook_type": "sft",
|
4 |
+
"dataset_type": "numeric",
|
5 |
+
"notebook_template": []
|
6 |
+
}
|
notebooks/rag.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"notebook_title": "Retrieval-augmented generation (RAG)",
|
3 |
+
"notebook_type": "rag",
|
4 |
+
"dataset_type": "text",
|
5 |
+
"notebook_template": []
|
6 |
+
}
|
utils/notebook_utils.py
CHANGED
@@ -1,3 +1,7 @@
|
|
|
|
|
|
|
|
|
|
1 |
def replace_wildcards(
|
2 |
templates, wildcards, replacements, has_numeric_columns, has_categoric_columns
|
3 |
):
|
@@ -20,7 +24,7 @@ def replace_wildcards(
|
|
20 |
return new_templates
|
21 |
|
22 |
|
23 |
-
|
24 |
{
|
25 |
"cell_type": "markdown",
|
26 |
"source": """
|
@@ -475,3 +479,17 @@ def generate_rag_system_prompt():
|
|
475 |
|
476 |
Use the provided code to load the dataset; do not use any other method.
|
477 |
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import json
|
3 |
+
|
4 |
+
|
5 |
def replace_wildcards(
|
6 |
templates, wildcards, replacements, has_numeric_columns, has_categoric_columns
|
7 |
):
|
|
|
24 |
return new_templates
|
25 |
|
26 |
|
27 |
+
embeddings_cells = [
|
28 |
{
|
29 |
"cell_type": "markdown",
|
30 |
"source": """
|
|
|
479 |
|
480 |
Use the provided code to load the dataset; do not use any other method.
|
481 |
"""
|
482 |
+
|
483 |
+
|
484 |
+
def load_json_files_from_folder(folder_path):
|
485 |
+
components = {}
|
486 |
+
|
487 |
+
for filename in os.listdir(folder_path):
|
488 |
+
if filename.endswith(".json"):
|
489 |
+
file_path = os.path.join(folder_path, filename)
|
490 |
+
|
491 |
+
with open(file_path, "r") as json_file:
|
492 |
+
data = json.load(json_file)
|
493 |
+
components[data["notebook_title"]] = data
|
494 |
+
|
495 |
+
return components
|