Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -3,10 +3,16 @@ from huggingface_hub import hf_hub_download
|
|
| 3 |
import json
|
| 4 |
import gzip
|
| 5 |
|
|
|
|
|
|
|
|
|
|
| 6 |
filepath = hf_hub_download(repo_id="bigcode/the-stack-username-to-repo", filename="username_to_repo.json.gz", repo_type="dataset", revision="v1.1")
|
|
|
|
|
|
|
| 7 |
|
|
|
|
| 8 |
with gzip.open(filepath, 'r') as f:
|
| 9 |
-
usernames = json.loads(f.read().decode('utf-8'))
|
| 10 |
|
| 11 |
text = """\
|
| 12 |
**_The Stack is an open governance interface between the AI community and the open source community._**
|
|
@@ -19,10 +25,10 @@ As part of the BigCode project, we released and maintain [The Stack](https://hug
|
|
| 19 |
This tool lets you check if a repository under a given username is part of The Stack dataset. Would you like to have your data removed from future versions of The Stack? You can opt-out following the instructions [here](https://www.bigcode-project.org/docs/about/the-stack/#how-can-i-request-that-my-data-be-removed-from-the-stack).
|
| 20 |
"""
|
| 21 |
|
| 22 |
-
def check_username(username):
|
| 23 |
output_md = ""
|
| 24 |
-
if username in usernames and len(usernames[username])>0:
|
| 25 |
-
repos = usernames[username]
|
| 26 |
repo_word = "repository" if len(repos)==1 else "repositories"
|
| 27 |
output_md += f"**Yes**, there is code from **{len(repos)} {repo_word}** in The Stack:\n\n"
|
| 28 |
for repo in repos:
|
|
@@ -33,12 +39,13 @@ def check_username(username):
|
|
| 33 |
|
| 34 |
with gr.Blocks() as demo:
|
| 35 |
gr.Markdown(text)
|
|
|
|
| 36 |
username = gr.Text("", label="Your GitHub username:")
|
| 37 |
check_button = gr.Button("Check!")
|
| 38 |
|
| 39 |
repos = gr.Markdown()
|
| 40 |
|
| 41 |
-
check_button.click(check_username, [username], repos)
|
| 42 |
|
| 43 |
|
| 44 |
demo.launch()
|
|
|
|
| 3 |
import json
|
| 4 |
import gzip
|
| 5 |
|
| 6 |
+
|
| 7 |
+
usernames = {}
|
| 8 |
+
|
| 9 |
filepath = hf_hub_download(repo_id="bigcode/the-stack-username-to-repo", filename="username_to_repo.json.gz", repo_type="dataset", revision="v1.1")
|
| 10 |
+
with gzip.open(filepath, 'r') as f:
|
| 11 |
+
usernames["v1.1"] = json.loads(f.read().decode('utf-8'))
|
| 12 |
|
| 13 |
+
filepath = hf_hub_download(repo_id="bigcode/the-stack-username-to-repo", filename="username_to_repo.json.gz", repo_type="dataset")
|
| 14 |
with gzip.open(filepath, 'r') as f:
|
| 15 |
+
usernames["v1.0"] = json.loads(f.read().decode('utf-8'))
|
| 16 |
|
| 17 |
text = """\
|
| 18 |
**_The Stack is an open governance interface between the AI community and the open source community._**
|
|
|
|
| 25 |
This tool lets you check if a repository under a given username is part of The Stack dataset. Would you like to have your data removed from future versions of The Stack? You can opt-out following the instructions [here](https://www.bigcode-project.org/docs/about/the-stack/#how-can-i-request-that-my-data-be-removed-from-the-stack).
|
| 26 |
"""
|
| 27 |
|
| 28 |
+
def check_username(username, version):
|
| 29 |
output_md = ""
|
| 30 |
+
if username in usernames[version] and len(usernames[version][username])>0:
|
| 31 |
+
repos = usernames[version][username]
|
| 32 |
repo_word = "repository" if len(repos)==1 else "repositories"
|
| 33 |
output_md += f"**Yes**, there is code from **{len(repos)} {repo_word}** in The Stack:\n\n"
|
| 34 |
for repo in repos:
|
|
|
|
| 39 |
|
| 40 |
with gr.Blocks() as demo:
|
| 41 |
gr.Markdown(text)
|
| 42 |
+
version = gr.Dropdown(["v1.1", "v1.0"], label="The Stack version:", value="v1.1")
|
| 43 |
username = gr.Text("", label="Your GitHub username:")
|
| 44 |
check_button = gr.Button("Check!")
|
| 45 |
|
| 46 |
repos = gr.Markdown()
|
| 47 |
|
| 48 |
+
check_button.click(check_username, [username, version], repos)
|
| 49 |
|
| 50 |
|
| 51 |
demo.launch()
|