Spaces:
Runtime error
Runtime error
Aaron Mueller
commited on
Commit
·
1ab72b5
1
Parent(s):
e7e9a2c
testing
Browse files- app.py +1 -1
- src/leaderboard/read_evals.py +1 -0
- src/submission/submit.py +6 -9
app.py
CHANGED
|
@@ -191,7 +191,7 @@ with demo:
|
|
| 191 |
model_name_textbox,
|
| 192 |
revision_name_textbox,
|
| 193 |
track_name,
|
| 194 |
-
|
| 195 |
],
|
| 196 |
submission_result,
|
| 197 |
)
|
|
|
|
| 191 |
model_name_textbox,
|
| 192 |
revision_name_textbox,
|
| 193 |
track_name,
|
| 194 |
+
submitted_predictions,
|
| 195 |
],
|
| 196 |
submission_result,
|
| 197 |
)
|
src/leaderboard/read_evals.py
CHANGED
|
@@ -54,6 +54,7 @@ class EvalResult:
|
|
| 54 |
|
| 55 |
def _get_task_results(task):
|
| 56 |
# We average all scores of a given metric (not all metrics are present in all files)
|
|
|
|
| 57 |
accs = np.array([v.get(task.metric, None) for k, v in data["results"].items() if task.benchmark == k])
|
| 58 |
if accs.size == 0 or any([acc is None for acc in accs]):
|
| 59 |
return None
|
|
|
|
| 54 |
|
| 55 |
def _get_task_results(task):
|
| 56 |
# We average all scores of a given metric (not all metrics are present in all files)
|
| 57 |
+
print(data["results"].items())
|
| 58 |
accs = np.array([v.get(task.metric, None) for k, v in data["results"].items() if task.benchmark == k])
|
| 59 |
if accs.size == 0 or any([acc is None for acc in accs]):
|
| 60 |
return None
|
src/submission/submit.py
CHANGED
|
@@ -16,9 +16,9 @@ USERS_TO_SUBMISSION_DATES = None
|
|
| 16 |
|
| 17 |
def add_new_eval(
|
| 18 |
model_name: str,
|
| 19 |
-
preds_path: str,
|
| 20 |
-
track: str,
|
| 21 |
revision: str,
|
|
|
|
|
|
|
| 22 |
):
|
| 23 |
global REQUESTED_MODELS
|
| 24 |
global USERS_TO_SUBMISSION_DATES
|
|
@@ -32,9 +32,6 @@ def add_new_eval(
|
|
| 32 |
model_path = model_name.split("/")[1]
|
| 33 |
|
| 34 |
current_time = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
|
| 35 |
-
|
| 36 |
-
if preds_path is None or preds_path == "":
|
| 37 |
-
return styled_error("Please enter a URL where your predictions file can be downloaded.")
|
| 38 |
|
| 39 |
if track is None:
|
| 40 |
return styled_error("Please select a track.")
|
|
@@ -58,9 +55,9 @@ def add_new_eval(
|
|
| 58 |
|
| 59 |
eval_entry = {
|
| 60 |
"model_name": model_name,
|
| 61 |
-
"preds_path": preds_path,
|
| 62 |
-
"track": track,
|
| 63 |
"revision": revision,
|
|
|
|
|
|
|
| 64 |
"status": "PENDING",
|
| 65 |
"submitted_time": current_time,
|
| 66 |
"private": False,
|
|
@@ -91,5 +88,5 @@ def add_new_eval(
|
|
| 91 |
os.remove(out_path)
|
| 92 |
|
| 93 |
return styled_message(
|
| 94 |
-
"Your request has been submitted to the evaluation queue!\nPlease wait for up to an hour for the
|
| 95 |
-
)
|
|
|
|
| 16 |
|
| 17 |
def add_new_eval(
|
| 18 |
model_name: str,
|
|
|
|
|
|
|
| 19 |
revision: str,
|
| 20 |
+
track: str,
|
| 21 |
+
predictions: dict,
|
| 22 |
):
|
| 23 |
global REQUESTED_MODELS
|
| 24 |
global USERS_TO_SUBMISSION_DATES
|
|
|
|
| 32 |
model_path = model_name.split("/")[1]
|
| 33 |
|
| 34 |
current_time = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%SZ")
|
|
|
|
|
|
|
|
|
|
| 35 |
|
| 36 |
if track is None:
|
| 37 |
return styled_error("Please select a track.")
|
|
|
|
| 55 |
|
| 56 |
eval_entry = {
|
| 57 |
"model_name": model_name,
|
|
|
|
|
|
|
| 58 |
"revision": revision,
|
| 59 |
+
"track": track,
|
| 60 |
+
"predictions": predictions,
|
| 61 |
"status": "PENDING",
|
| 62 |
"submitted_time": current_time,
|
| 63 |
"private": False,
|
|
|
|
| 88 |
os.remove(out_path)
|
| 89 |
|
| 90 |
return styled_message(
|
| 91 |
+
"Your request has been submitted to the evaluation queue!\nPlease wait for up to an hour for the request to show in the PENDING list."
|
| 92 |
+
)
|