--enable_scope_fusion_passes

Applicability

Product	Supported
Atlas A3 training products/Atlas A3 inference products	√
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	√
Atlas inference products	√
Atlas training products	√

Description

Sets the scope fusion pattern (or scope fusion patterns separated by commas) to take effect during compilation.

Scope fusion patterns (either built-in or custom) are classified into the following two types:

General: applicable to all networks. They are enabled by default and cannot be manually disabled.
Non-general: scope fusion patterns applicable to specific networks. By default, they do not take effect. You can use --enable_scope_fusion_passes to specify the fusion pattern list that takes effect.

For the full list of supported fusion patterns, see the TensorFlow Parser Scope Fusion Pattern Reference.

Arguments

Argument: registered fusion pattern names.

Restrictions: Multiple arguments are allowed. Separate them by commas (,), for example, ScopePass1,ScopePass2, ....

Suggestions and Benefits

None

Examples

--enable_scope_fusion_passes=ScopePass1,ScopePass2

Dependencies and Restrictions

This option applies only to TensorFlow networks. To view logs related to fusion patterns during model tuning, set --log to at least the warning level.

Parent topic: Advanced Functionality