[Feature] Add Muse Glimmer model support (#34262)

Co-authored-by: sglang-bot <232288953+sglang-bot@users.noreply.github.com>
Co-authored-by: Brayden Zhong <brayden.zhong@radixark.ai>
Co-authored-by: Jimmy Shong <69131491+Jiminator@users.noreply.github.com>
Co-authored-by: hnyls2002 <lsyincs@gmail.com>
Co-authored-by: Alex Nails <alex.nails@radixark.ai>
Co-authored-by: Liangsheng Yin <hnyls2002@gmail.com>
This commit is contained in:
sglang-bot
2026-08-11 15:41:52 -07:00
committed by GitHub
co-authored by sglang-bot Brayden Zhong Jimmy Shong hnyls2002 Alex Nails Liangsheng Yin
parent 9c1517df4a
commit fde9ad2531
47 changed files with 5009 additions and 50 deletions
@@ -75,7 +75,7 @@ class TestKVCacheQuantRegistry(CustomTestCase):
from sglang.srt.runtime_context import get_context
runner = object.__new__(ModelRunner)
runner.server_args = SimpleNamespace()
runner.server_args = SimpleNamespace(speculative_draft_kv_cache_dtype=None)
runner.draft_attention_backend = None
# The runner reads the requested dtype off the model bag, so the double
# publishes it rather than carrying it on a stand-in config.