[Core] Update dtype detection and defaults (#14858)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
2025-03-19 13:49:33 +08:00
parent 8b3e94a357
commit f690372b68
22 changed files with 175 additions and 227 deletions
--- a/tests/entrypoints/openai/test_vision_embedding.py
+++ b/tests/entrypoints/openai/test_vision_embedding.py
@@ -28,8 +28,6 @@ def server():
    args = [
        "--task",
        "embed",
-        "--dtype",
-        "bfloat16",
        "--max-model-len",
        "2048",
        "--max-num-seqs",