A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 3, 13:00 UTC
Fix: Propagate pinned model revisions into Ultravox secondary weight loading (#42830)
weizhoublue committed
May 17, 2026 at 01:02 UTC
d1586e1a1242754d2f6ac51f4f16680f7d4b129b
1 file changed
+2
-2
vllm/model_executor/models/ultravox.py
+2
-2
@@ -566,7 +566,7 @@ class UltravoxModel(nn.Module, SupportsMultiModal, SupportsPP, SupportsLoRA):
566
self.secondary_weights.append(
567
DefaultModelLoader.Source(
568
model_or_path=config.audio_model_id,
569
- revision=None,
569
+ revision=vllm_config.model_config.revision,
570
prefix="audio_tower.",
571
)
572
)
@@ -576,7 +576,7 @@ class UltravoxModel(nn.Module, SupportsMultiModal, SupportsPP, SupportsLoRA):
576
self.secondary_weights.append(
577
DefaultModelLoader.Source(
578
model_or_path=config.text_model_id,
579
- revision=None,
579
+ revision=vllm_config.model_config.revision,
580
prefix="language_model.",
581
)
582
)