nemo_automodel.components.models.qwen3_omni_moe.state_dict_adapter

Module Contents

Classes

Qwen3OmniMoeStateDictAdapter

Converts between HF Qwen3OmniMoe checkpoints and grouped-experts native format.

Data

logger

API

nemo_automodel.components.models.qwen3_omni_moe.state_dict_adapter.logger

‘getLogger(…)’

class nemo_automodel.components.models.qwen3_omni_moe.state_dict_adapter.Qwen3OmniMoeStateDictAdapter(

config: Any,

moe_config: nemo_automodel.components.moe.layers.MoEConfig,

backend: nemo_automodel.components.moe.utils.BackendConfig,

dtype: torch.dtype = torch.float32,

)

Bases: nemo_automodel.components.moe.state_dict_mixin.MoESplitExpertsStateDictMixin, nemo_automodel.components.checkpoint.state_dict_adapter.StateDictAdapter

Converts between HF Qwen3OmniMoe checkpoints and grouped-experts native format.

Initialization

to_hf(

state_dict: dict[str, Any],

exclude_key_regex: Optional[str] = None,

quantization: bool = False,

**kwargs,

) → dict[str, Any]

from_hf(

hf_state_dict: dict[str, Any],

device_mesh: Optional[torch.distributed.device_mesh.DeviceMesh] = None,

**kwargs,

) → dict[str, Any]

convert_single_tensor_to_hf(

fqn: str,

tensor: Any,

**kwargs,

) → list[tuple[str, Any]]