[Model] Use upstream GLM-5.3 and Qwen4-Exp configs and processor (#57387)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
co-authored by
Claude Opus 5
parent
866f02db8a
commit
e3cae8d2ac
@@ -43,7 +43,7 @@ def test_glm_mtp_defers_lm_head(default_vllm_config):
|
||||
index_kpool=4,
|
||||
)
|
||||
vllm_config = mock.MagicMock()
|
||||
vllm_config.speculative_config.draft_model_config.hf_config = config
|
||||
vllm_config.speculative_config.draft_model_config.hf_text_config = config
|
||||
vllm_config.scheduler_config.max_num_batched_tokens = 4
|
||||
|
||||
with (
|
||||
|
||||
Reference in New Issue
Block a user