EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

Add torchao quant (int4/int8/fp8) to llama models (#1341)

Browse Source 
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
This commit is contained in:
Jerry Zhang
2024-09-09 05:32:41 -07:00
committed by GitHub
parent e4d68afcf0
commit a7c47e0f02
10 changed files with 151 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
test/srt/test_moe_eval_accuracy_large.py
View File

@@ -42,7 +42,7 @@ class TestEvalAccuracyLarge(unittest.TestCase):
)
metrics = run_eval(args)
assert metrics["score"] >= 0.62, f"{metrics}"
assert metrics["score"] >= 0.625, f"{metrics}"
def test_human_eval(self):
args = SimpleNamespace(
@@ -54,7 +54,7 @@ class TestEvalAccuracyLarge(unittest.TestCase):
)
metrics = run_eval(args)
assert metrics["score"] >= 0.42, f"{metrics}"
assert metrics["score"] >= 0.425, f"{metrics}"
def test_mgsm_en(self):
args = SimpleNamespace(
@@ -66,7 +66,7 @@ class TestEvalAccuracyLarge(unittest.TestCase):
)
metrics = run_eval(args)
assert metrics["score"] >= 0.62, f"{metrics}"
assert metrics["score"] >= 0.625, f"{metrics}"
if __name__ == "__main__":