cache-management (#951)
Browse files- Disable cache (93c865a9a70cb708c7759e76b32802e7320cfc8e)
- Correct uv dependency (29a7485638a84d6d3f65766b278e5374b30921a8)
- app.py +7 -10
- pyproject.toml +5 -5
app.py
CHANGED
|
@@ -1,4 +1,3 @@
|
|
| 1 |
-
import os
|
| 2 |
import logging
|
| 3 |
import time
|
| 4 |
import schedule
|
|
@@ -100,29 +99,27 @@ def download_dataset(repo_id, local_dir, repo_type="dataset", max_attempts=3, ba
|
|
| 100 |
attempt += 1
|
| 101 |
raise Exception(f"Failed to download {repo_id} after {max_attempts} attempts")
|
| 102 |
|
| 103 |
-
def get_latest_data_leaderboard(leaderboard_initial_df
|
| 104 |
global NEW_DATA_ON_LEADERBOARD
|
| 105 |
global LEADERBOARD_DF
|
| 106 |
if NEW_DATA_ON_LEADERBOARD:
|
| 107 |
print("Leaderboard updated at reload!")
|
| 108 |
leaderboard_dataset = datasets.load_dataset(
|
| 109 |
-
AGGREGATED_REPO,
|
| 110 |
-
"default",
|
| 111 |
-
split="train",
|
| 112 |
-
cache_dir=
|
| 113 |
-
download_mode=datasets.DownloadMode.
|
| 114 |
verification_mode="no_checks"
|
| 115 |
)
|
| 116 |
LEADERBOARD_DF = get_leaderboard_df(
|
| 117 |
-
leaderboard_dataset=leaderboard_dataset,
|
| 118 |
cols=COLS,
|
| 119 |
benchmark_cols=BENCHMARK_COLS,
|
| 120 |
)
|
| 121 |
NEW_DATA_ON_LEADERBOARD = False
|
| 122 |
-
|
| 123 |
else:
|
| 124 |
LEADERBOARD_DF = leaderboard_initial_df
|
| 125 |
-
|
| 126 |
return LEADERBOARD_DF
|
| 127 |
|
| 128 |
|
|
|
|
|
|
|
| 1 |
import logging
|
| 2 |
import time
|
| 3 |
import schedule
|
|
|
|
| 99 |
attempt += 1
|
| 100 |
raise Exception(f"Failed to download {repo_id} after {max_attempts} attempts")
|
| 101 |
|
| 102 |
+
def get_latest_data_leaderboard(leaderboard_initial_df=None):
|
| 103 |
global NEW_DATA_ON_LEADERBOARD
|
| 104 |
global LEADERBOARD_DF
|
| 105 |
if NEW_DATA_ON_LEADERBOARD:
|
| 106 |
print("Leaderboard updated at reload!")
|
| 107 |
leaderboard_dataset = datasets.load_dataset(
|
| 108 |
+
AGGREGATED_REPO,
|
| 109 |
+
"default",
|
| 110 |
+
split="train",
|
| 111 |
+
cache_dir=None, # Disable cache directory usage
|
| 112 |
+
download_mode=datasets.DownloadMode.FORCE_REDOWNLOAD, # Always download fresh data
|
| 113 |
verification_mode="no_checks"
|
| 114 |
)
|
| 115 |
LEADERBOARD_DF = get_leaderboard_df(
|
| 116 |
+
leaderboard_dataset=leaderboard_dataset,
|
| 117 |
cols=COLS,
|
| 118 |
benchmark_cols=BENCHMARK_COLS,
|
| 119 |
)
|
| 120 |
NEW_DATA_ON_LEADERBOARD = False
|
|
|
|
| 121 |
else:
|
| 122 |
LEADERBOARD_DF = leaderboard_initial_df
|
|
|
|
| 123 |
return LEADERBOARD_DF
|
| 124 |
|
| 125 |
|
pyproject.toml
CHANGED
|
@@ -17,7 +17,7 @@ dependencies = [
|
|
| 17 |
"sentencepiece>=0.2.0",
|
| 18 |
"transformers==4.44.2",
|
| 19 |
"tokenizers>=0.19.0",
|
| 20 |
-
"gradio-space-ci
|
| 21 |
"isort>=5.13.2",
|
| 22 |
"ruff>=0.6.4",
|
| 23 |
"gradio-leaderboard==0.0.11",
|
|
@@ -34,16 +34,16 @@ ignore=["I","EM","FBT","TRY003","S101","D101","D102","D103","D104","D105","G004"
|
|
| 34 |
fixable=["ALL"]
|
| 35 |
select=["ALL"]
|
| 36 |
|
| 37 |
-
|
| 38 |
select = ["E", "F"]
|
| 39 |
fixable = ["ALL"]
|
| 40 |
ignore = ["E501"] # line too long (black is taking care of this)
|
| 41 |
|
| 42 |
-
|
| 43 |
profile = "black"
|
| 44 |
|
| 45 |
[tool.black]
|
| 46 |
line-length = 119
|
| 47 |
|
| 48 |
-
[tool.
|
| 49 |
-
|
|
|
|
| 17 |
"sentencepiece>=0.2.0",
|
| 18 |
"transformers==4.44.2",
|
| 19 |
"tokenizers>=0.19.0",
|
| 20 |
+
"gradio-space-ci",
|
| 21 |
"isort>=5.13.2",
|
| 22 |
"ruff>=0.6.4",
|
| 23 |
"gradio-leaderboard==0.0.11",
|
|
|
|
| 34 |
fixable=["ALL"]
|
| 35 |
select=["ALL"]
|
| 36 |
|
| 37 |
+
[tool.ruff.lint]
|
| 38 |
select = ["E", "F"]
|
| 39 |
fixable = ["ALL"]
|
| 40 |
ignore = ["E501"] # line too long (black is taking care of this)
|
| 41 |
|
| 42 |
+
[tool.isort]
|
| 43 |
profile = "black"
|
| 44 |
|
| 45 |
[tool.black]
|
| 46 |
line-length = 119
|
| 47 |
|
| 48 |
+
[tool.uv.sources]
|
| 49 |
+
gradio-space-ci = { git = "https://huggingface.co/spaces/Wauplin/gradio-space-ci" }
|