TensorParallelLayerScenario
Training scenarios · Implementation: mlsysim/engine/scenario_definitions/training.py.
Construct this scenario with keyword inputs. The fields below show the types and defaults declared by this class.
One layer of workload split tp_size ways: compute against the exchange on each link.
Notes
Equation: Compose mlsysim.solvers.TensorParallelLayerModel.
Assumptions: The stated fields are inputs to mlsysim.solvers.TensorParallelLayerModel; its documented validity regime, units, and modeling choices apply unchanged.
Excludes: This scenario adds no measurement, empirical calibration, or mechanisms beyond mlsysim.solvers.TensorParallelLayerModel.
References
Derivation: this scenario binds its input fields to the referenced implementation contracts: mlsysim.solvers.TensorParallelLayerModel. The methods below identify any additional arithmetic.
Examples
>>> from mlsysim import Scenarios
>>> case = Scenarios.Gpt3TensorParallelLayerTp8
>>> type(case.result).__name__
'TensorParallelLayerResult'Attributes
| Name | Description |
|---|---|
| accelerator | |
| collectives_per_layer | |
| links | |
| local_efficiency | |
| precision | |
| result | Evaluate this scenario through mlsysim.solvers.TensorParallelLayerModel.solve. |
| tokens | |
| tp_size | |
| workload |
workload
Type: Workload.
Declared default: required.
tp_size
Type: int.
Declared default: required.
tokens
Type: int.
Declared default: required.
accelerator
Type: HardwareNode.
Declared default: required.
precision
Type: str.
Declared default: 'fp16_dense'.
local_efficiency
Type: float.
Declared default: required.
links
Type: Dict[str, Quantity].
Declared default: required.
collectives_per_layer
Type: int.
Declared default: 2.
result
engine.scenario_types.TensorParallelLayerScenario.result
Evaluate this scenario through mlsysim.solvers.TensorParallelLayerModel.solve. Equation: pass the stated fields to that implementation; its contract defines the calculation and feasibility tests. Units: each returned field follows the referenced result schema; no independent equation is evaluated by this wrapper.