[NPU]support model MiniCPM3-4B for npu (#16866)
This commit is contained in:
@@ -406,7 +406,9 @@ class ModelRunnerKVCacheMixin:
|
||||
dtype=self.kv_cache_dtype,
|
||||
kv_lora_rank=self.model_config.kv_lora_rank,
|
||||
qk_rope_head_dim=self.model_config.qk_rope_head_dim,
|
||||
index_head_dim=self.model_config.index_head_dim,
|
||||
index_head_dim=(
|
||||
self.model_config.index_head_dim if is_nsa_model else None
|
||||
),
|
||||
layer_num=self.num_effective_layers,
|
||||
device=self.device,
|
||||
enable_memory_saver=self.server_args.enable_memory_saver,
|
||||
|
||||
Reference in New Issue
Block a user