onnx_diagnostic.tasks.mixture_of_expert

onnx_diagnostic.tasks.mixture_of_expert.get_inputs(model: Module, config: Any | None, dummy_max_token_id: int, num_key_value_heads: int, num_hidden_layers: int, head_dim: int, width: int, height: int, num_channels: int, batch_size: int = 2, sequence_length: int = 30, sequence_length2: int = 3, n_images: int = 2, dynamic_rope: bool = False, add_second_input: bool = False, **kwargs)

Generates input for task MoE.

Parameters:

• model – model to get the missing information

• config – configuration used to generate the model

• head_dim – last dimension of the cache

• dummy_max_token_id – dummy max token id

• batch_size – batch size

• sequence_length – sequence length

• sequence_length2 – new sequence length

• n_images – number of images

• width – width of the image

• height – height of the image

• num_channels – number of channels

• dynamic_rope – use dynamic rope (see transformers.LlamaConfig)

Returns:

dictionary

onnx_diagnostic.tasks.mixture_of_expert.random_input_kwargs(config: Any) → Tuple[Dict[str, Any], Callable]

Inputs kwargs.

If the configuration is None, the function selects typical dimensions.

onnx_diagnostic.tasks.mixture_of_expert.reduce_model_config(config: Any) → Dict[str, Any]

Reduces a model size.