Add accuracy and latency tests of eagle into CI (#3027)

This commit is contained in:
Lianmin Zheng
2025-01-21 02:55:14 -08:00
committed by GitHub
parent ec1c21cdc4
commit a4331cd260
7 changed files with 186 additions and 123 deletions

View File

@@ -37,8 +37,7 @@ class TestQwen2(unittest.TestCase):
port=int(self.base_url.split(":")[-1]),
)
metrics = run_eval(args)
print(metrics)
print(f"{metrics=}")
self.assertGreater(metrics["accuracy"], 0.81)
@@ -69,8 +68,7 @@ class TestQwen2FP8(unittest.TestCase):
port=int(self.base_url.split(":")[-1]),
)
metrics = run_eval(args)
print(metrics)
print(f"{metrics=}")
self.assertGreater(metrics["accuracy"], 0.79)