Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions .github/workflows/auto-update-dev.yml
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@ on:
push:
branches:
- master # Trigger workflow on commits to 'master' branch


jobs:
update-dev:
Expand Down
17 changes: 12 additions & 5 deletions language/llama3.1-8b/ref_eval.py
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,8 @@ def rouge(label, pred):


def niah_em(label, pred):
label_uuids = re.findall(r'[\w]{8}-[\w]{4}-[\w]{4}-[\w]{4}-[\w]{12}', label)
label_uuids = re.findall(
r'[\w]{8}-[\w]{4}-[\w]{4}-[\w]{4}-[\w]{12}', label)
pred_uuids = re.findall(r'[\w]{8}-[\w]{4}-[\w]{4}-[\w]{4}-[\w]{12}', pred)

# https://github.com/hsiehjackson/RULER/blob/main/scripts/eval/synthetic/constants.py#L28
Expand All @@ -43,7 +44,8 @@ def qa_em(label, pred):
return {'exact_match': 100.0}

normalized_answer = re.sub(r'\s+', '', answer_substring).lower()
label_entries = [re.sub(r'\s+', '', entry).lower() for entry in label.split('|')]
label_entries = [re.sub(r'\s+', '', entry).lower()
for entry in label.split('|')]

match_found = any(entry in normalized_answer for entry in label_entries)
return {'exact_match': 100.0 if match_found else 0.0}
Expand All @@ -63,7 +65,12 @@ def process_row(row):

def run_evaluation(df):
with Pool(cpu_count()) as pool:
accuracies = list(tqdm(pool.imap(process_row, df.to_dict('records')), total=len(df)))
accuracies = list(
tqdm(
pool.imap(
process_row,
df.to_dict('records')),
total=len(df)))

df['accuracy'] = accuracies
return df
Expand All @@ -74,10 +81,10 @@ def run_evaluation(df):
df = pd.read_pickle(fname)

df = run_evaluation(df)
#df.to_pickle(str(fname).replace(".pkl", "_eval.pkl"))
# df.to_pickle(str(fname).replace(".pkl", "_eval.pkl"))
print(f"WROTE: {str(fname).replace('.pkl', '_eval.pkl')}")

accuracy = df.accuracy.apply(pd.Series)
print(df.dataset.value_counts())
print(accuracy.describe())
print(df.describe())
print(df.describe())
6 changes: 3 additions & 3 deletions loadgen/mlperf.conf
Original file line number Diff line number Diff line change
Expand Up @@ -27,9 +27,9 @@ whisper.*.performance_sample_count_override = 1633
3d-unet.*.performance_sample_count_override = 0

# Set seeds.
*.*.qsl_rng_seed = 6023615788873153749
*.*.sample_index_rng_seed = 15036839855038426416
*.*.schedule_rng_seed = 9933818062894767841
*.*.qsl_rng_seed = 1780908523862526354
*.*.sample_index_rng_seed = 14771362308971278857
*.*.schedule_rng_seed = 18209322760996052031

# Set seeds for TEST_05 (not needed from v5.0 onwards)
*.*.test05_qsl_rng_seed = 7975553102935885558
Expand Down
2 changes: 1 addition & 1 deletion text_to_image/tools/sample_ids.py
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@ def get_args():
default=10,
help="Dataset download location")
parser.add_argument(
"--seed", "-s", type=int, default=265673710, help="Dataset download location"
"--seed", "-s", type=int, default=927722784, help="Dataset download location"
)
args = parser.parse_args()
return args
Expand Down
20 changes: 10 additions & 10 deletions text_to_image/tools/sample_ids.txt
Original file line number Diff line number Diff line change
@@ -1,10 +1,10 @@
716
2387
3694
273
1221
2942
2047
1691
1510
4399
2747
2235
2165
1515
1538
1367
2419
4629
3657
4532
61 changes: 33 additions & 28 deletions tools/submission/generate_final_report.py
Original file line number Diff line number Diff line change
Expand Up @@ -135,26 +135,24 @@ def main():
[
"resnet",
"retinanet",
"3d-unet-99",
"3d-unet-99.9",
"rnnt",
"bert-99",
"bert-99.9",
"dlrm-v2-99",
"dlrm-v2-99.9",
"gptj-99",
"gptj-99.9",
"stable-diffusion-xl",
"3d-unet-99",
"3d-unet-99.9",
"llama3.1-8b",
"llama2-70b-99",
"llama2-70b-99.9",
"llama2-70b-interactive-99",
"llama2-70b-interactive-99.9",
"llama3.1-405b",
"stable-diffusion-xl",
"mixtral-8x7b",
"pointpainting",
"llama3.1-405b",
"rgat",
"pointpainting",
"deepseek-r1",
"whisper"
],
["SingleStream", "MultiStream", "Server", "Offline"],
["SingleStream", "MultiStream", "Server", "Offline", "Interactive"],
[
"Latency (ms)",
"Samples/s",
Expand All @@ -165,14 +163,13 @@ def main():
],
]

if args.version == "4.1":
if args.version == "5.0":
filter_scenarios = {
"datacenter": {
"resnet": ["Server", "Offline"],
"retinanet": ["Server", "Offline"],
"rnnt": ["Server", "Offline"],
"bert-99": ["Server", "Offline"],
"bert-99.9": ["Server", "Offline"],
"bert-99": [],
"bert-99.9": [],
"dlrm-v2-99": ["Server", "Offline"],
"dlrm-v2-99.9": ["Server", "Offline"],
"3d-unet-99": ["Offline"],
Expand All @@ -182,21 +179,32 @@ def main():
"stable-diffusion-xl": ["Server", "Offline"],
"llama2-70b-99": ["Server", "Offline"],
"llama2-70b-99.9": ["Server", "Offline"],
"llama2-70b-interactive-99": ["Server", "Offline"],
"llama2-70b-interactive-99.9": ["Server", "Offline"],
"mixtral-8x7b": ["Server", "Offline"],
"rgat": ["Offline"],
"llama3.1-405b": ["Offline", "Server"],
"pointpainting": []
},
"edge": {
"resnet": ["SingleStream", "MultiStream", "Offline"],
"retinanet": ["SingleStream", "MultiStream", "Offline"],
"rnnt": ["SingleStream", "Offline"],
"bert-99": ["SingleStream", "Offline"],
"bert-99.9": [],
"bert-99.9": ["SingleStream", "Offline"],
"dlrm-v2-99": [],
"dlrm-v2-99.9": [],
"3d-unet-99": ["SingleStream", "Offline"],
"3d-unet-99.9": ["SingleStream", "Offline"],
"llama2-70b-99": [],
"llama2-70b-99.9": [],
"llama2-70b-interactive-99": [],
"llama2-70b-interactive-99.9": [],
"llama3.1-405b": [],
"gptj-99": ["SingleStream", "Offline"],
"gptj-99.9": ["SingleStream", "Offline"],
"rgat": [],
"stable-diffusion-xl": ["SingleStream", "Offline"],
"pointpainting": ["SingleStream"],
},
}
else:
Expand All @@ -210,17 +218,16 @@ def main():
"dlrm-v2-99.9": ["Server", "Offline"],
"3d-unet-99": ["Offline"],
"3d-unet-99.9": ["Offline"],
"gptj-99": ["Server", "Offline"],
"gptj-99.9": ["Server", "Offline"],
"stable-diffusion-xl": ["Server", "Offline"],
"llama2-70b-99": ["Server", "Offline"],
"llama2-70b-99.9": ["Server", "Offline"],
"llama2-70b-interactive-99": ["Server", "Offline"],
"llama2-70b-interactive-99.9": ["Server", "Offline"],
"llama2-70b-99": ["Server", "Offline", "Interactive"],
"llama2-70b-99.9": ["Server", "Offline", "Interactive"],
"mixtral-8x7b": ["Server", "Offline"],
"rgat": ["Offline"],
"llama3.1-405b": ["Offline", "Server"],
"pointpainting": []
"pointpainting": [],
"llama3.1-8b": ["Server", "Offline"],
"deepseek-r1": ["Server", "Offline"],
"whisper": ["Offline"],
},
"edge": {
"resnet": ["SingleStream", "MultiStream", "Offline"],
Expand All @@ -233,14 +240,12 @@ def main():
"3d-unet-99.9": ["SingleStream", "Offline"],
"llama2-70b-99": [],
"llama2-70b-99.9": [],
"llama2-70b-interactive-99": [],
"llama2-70b-interactive-99.9": [],
"llama3.1-405b": [],
"gptj-99": ["SingleStream", "Offline"],
"gptj-99.9": ["SingleStream", "Offline"],
"rgat": [],
"stable-diffusion-xl": ["SingleStream", "Offline"],
"pointpainting": ["SingleStream"],
"llama3.1-8b": ["SingleStream", "Offline"],
"whisper": ["Offline"],
},
}

Expand Down
Loading