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 13, 09:30 UTC
[BugFix]: --enable-lora with model granite-4.0-micro crash (#27733)
Signed-off-by: zxw <1020938856@qq.com>
yyzxw committed
Nov 12, 2025 at 17:03 UTC
1761dea1a8567fc143b7bfbe61cb1e00cc081c7f
1 file changed
+3
vllm/model_executor/models/granitemoehybrid.py
+3
@@ -597,6 +597,9 @@ class GraniteMoeHybridForCausalLM(
597
"k_proj",
598
"v_proj",
599
],
600
+ "conv1d": ["conv1d"],
601
+ "in_proj": ["in_proj"],
602
+ "input_linear": ["input_linear"],
603
}
604
embedding_modules = {
605
"embed_tokens": "input_embeddings",