[Bugfix] Fix mamba state dtype setting for Qwen3-Next and Qwen3.5 (#34489)
Signed-off-by: Roger Wang <hey@rogerw.io>
This commit is contained in:
@@ -341,7 +341,9 @@ class Qwen3NextGatedDeltaNet(nn.Module, MambaBase):
|
||||
|
||||
def get_state_dtype(self) -> tuple[torch.dtype, torch.dtype]:
|
||||
return MambaStateDtypeCalculator.gated_delta_net_state_dtype(
|
||||
self.model_config.dtype, self.cache_config.mamba_cache_dtype
|
||||
self.model_config.dtype,
|
||||
self.cache_config.mamba_cache_dtype,
|
||||
self.cache_config.mamba_ssm_cache_dtype,
|
||||
)
|
||||
|
||||
def get_state_shape(self) -> tuple[tuple[int, ...], tuple[int, ...]]:
|
||||
@@ -1372,7 +1374,9 @@ class Qwen3NextForCausalLM(
|
||||
vllm_config: "VllmConfig",
|
||||
) -> tuple[torch.dtype, torch.dtype]:
|
||||
return MambaStateDtypeCalculator.gated_delta_net_state_dtype(
|
||||
vllm_config.model_config.dtype, vllm_config.cache_config.mamba_cache_dtype
|
||||
vllm_config.model_config.dtype,
|
||||
vllm_config.cache_config.mamba_cache_dtype,
|
||||
vllm_config.cache_config.mamba_ssm_cache_dtype,
|
||||
)
|
||||
|
||||
@classmethod
|
||||
|
||||
Reference in New Issue
Block a user