lewtun HF staff commited on
Commit
aa6287d
1 Parent(s): 3f9b5a9

Speed up Space loading

Browse files
Files changed (1) hide show
  1. app.py +14 -16
app.py CHANGED
@@ -4,16 +4,15 @@ from pathlib import Path
4
 
5
  import numpy as np
6
  import pandas as pd
7
- import requests
8
  import streamlit as st
9
  from datasets import get_dataset_config_names
10
  from dotenv import load_dotenv
 
11
 
12
  if Path(".env").is_file():
13
  load_dotenv(".env")
14
 
15
  auth_token = os.getenv("HF_HUB_TOKEN")
16
- header = {"Authorization": "Bearer " + auth_token}
17
 
18
  TASKS = sorted(get_dataset_config_names("ought/raft"))
19
  # Split and capitalize the task names, e.g. banking_77 => Banking 77
@@ -22,37 +21,36 @@ FORMATTED_TASK_NAMES = sorted([" ".join(t.capitalize() for t in task.split("_"))
22
 
23
  def extract_tags(dataset):
24
  tags = {}
25
- for tag in dataset["tags"]:
26
  k, v = tuple(tag.split(":", 1))
27
  tags[k] = v
28
  return tags
29
 
30
 
31
  def download_submissions():
32
- response = requests.get("http://huggingface.co/api/datasets", headers=header)
33
- all_datasets = response.json()
34
-
35
  submissions = []
36
 
37
- for dataset in all_datasets:
38
  tags = extract_tags(dataset)
39
- if tags.get("benchmark") == "raft" and tags.get("type") == "evaluation":
40
  submissions.append(dataset)
41
  return submissions
42
 
43
 
44
  def format_submissions(submissions):
45
- submission_data = {**{"Submitter": []}, **{"Submission Name": []}, **{"Submission Date": []}, **{t: [] for t in TASKS}}
 
 
 
 
 
46
 
47
  # The following picks the latest submissions which adhere to the model card schema
48
  for submission in submissions:
49
- submission_id = submission["id"]
50
- response = requests.get(
51
- f"http://huggingface.co/api/datasets/{submission_id}?full=true",
52
- headers=header,
53
- )
54
- data = response.json()
55
- card_data = data["cardData"]
56
  username = card_data["submission_dataset"].split("/")[0]
57
  submission_data["Submitter"].append(username)
58
  submission_id = card_data["submission_id"]
 
4
 
5
  import numpy as np
6
  import pandas as pd
 
7
  import streamlit as st
8
  from datasets import get_dataset_config_names
9
  from dotenv import load_dotenv
10
+ from huggingface_hub import DatasetFilter, list_datasets
11
 
12
  if Path(".env").is_file():
13
  load_dotenv(".env")
14
 
15
  auth_token = os.getenv("HF_HUB_TOKEN")
 
16
 
17
  TASKS = sorted(get_dataset_config_names("ought/raft"))
18
  # Split and capitalize the task names, e.g. banking_77 => Banking 77
 
21
 
22
  def extract_tags(dataset):
23
  tags = {}
24
+ for tag in dataset.tags:
25
  k, v = tuple(tag.split(":", 1))
26
  tags[k] = v
27
  return tags
28
 
29
 
30
  def download_submissions():
31
+ filt = DatasetFilter(benchmark="raft")
32
+ all_submissions = list_datasets(filter=filt, cardData=True, use_auth_token=auth_token)
 
33
  submissions = []
34
 
35
+ for dataset in all_submissions:
36
  tags = extract_tags(dataset)
37
+ if tags.get("type") == "evaluation":
38
  submissions.append(dataset)
39
  return submissions
40
 
41
 
42
  def format_submissions(submissions):
43
+ submission_data = {
44
+ **{"Submitter": []},
45
+ **{"Submission Name": []},
46
+ **{"Submission Date": []},
47
+ **{t: [] for t in TASKS},
48
+ }
49
 
50
  # The following picks the latest submissions which adhere to the model card schema
51
  for submission in submissions:
52
+ submission_id = submission.id
53
+ card_data = submission.cardData
 
 
 
 
 
54
  username = card_data["submission_dataset"].split("/")[0]
55
  submission_data["Submitter"].append(username)
56
  submission_id = card_data["submission_id"]