📝 update docstrings for create_model

pull/1643/head
nateraw 2 years ago
parent 3aa31f537d
commit f0dc8a8267

@ -46,27 +46,59 @@ def create_model(
no_jit: Optional[bool] = None, no_jit: Optional[bool] = None,
**kwargs, **kwargs,
): ):
"""Create a model """Create a model.
Lookup model's entrypoint function and pass relevant args to create a new model. Lookup model's entrypoint function and pass relevant args to create a new model.
**kwargs will be passed through entrypoint fn to timm.models.build_model_with_cfg() <Tip>
**kwargs will be passed through entrypoint fn to ``timm.models.build_model_with_cfg()``
and then the model class __init__(). kwargs values set to None are pruned before passing. and then the model class __init__(). kwargs values set to None are pruned before passing.
</Tip>
Args: Args:
model_name (str): name of model to instantiate model_name (str):
pretrained (bool): load pretrained ImageNet-1k weights if true Name of model to instantiate.
pretrained_cfg (Union[str, dict, PretrainedCfg]): pass in external pretrained_cfg for model pretrained (`bool`, *optional*, defaults to `False`):
pretrained_cfg_overlay (dict): replace key-values in base pretrained_cfg with these If set to `True`, load pretrained ImageNet-1k weights.
checkpoint_path (str): path of checkpoint to load _after_ the model is initialized pretrained_cfg (Union[str, dict, PretrainedCfg], *optional*):
scriptable (bool): set layer config so that model is jit scriptable (not working for all models yet) Pass in an external pretrained_cfg for model.
exportable (bool): set layer config so that model is traceable / ONNX exportable (not fully impl/obeyed yet) pretrained_cfg_overlay (dict, *optional*):
no_jit (bool): set layer config so that model doesn't utilize jit scripted layers (so far activations only) Replace key-values in base pretrained_cfg with these.
checkpoint_path (str, *optional*):
Keyword Args: Path of checkpoint to load _after_ the model is initialized.
drop_rate (float): dropout rate for training (default: 0.0) scriptable (bool, *optional*):
global_pool (str): global pool type (default: 'avg') Set layer config so that model is jit scriptable (not working for all models yet).
**: other kwargs are consumed by builder or model __init__() exportable (bool, *optional*):
Set layer config so that model is traceable / ONNX exportable (not fully impl/obeyed yet).
no_jit (bool, *optional*):
Set layer config so that model doesn't utilize jit scripted layers (so far activations only).
**Keyword Args**:
- **drop_rate** (float, *optional*, defaults to `0.0`):
Dropout rate for training.
- **global_pool** (str, *optional*, defaults to `'avg'`):
Global pooling type.
- All other kwargs are consumed by builder or model ``__init__()``.
Example:
```py
>>> from timm import create_model
>>> # Create a MobileNetV3-Large model with no pretrained weights.
>>> model = create_model('mobilenetv3_large_100')
>>> # Create a MobileNetV3-Large model with pretrained weights.
>>> model = create_model('mobilenetv3_large_100', pretrained=True)
>>> model.num_classes
1000
>>> # Create a MobileNetV3-Large model with pretrained weights and a new head with 10 classes.
>>> model = create_model('mobilenetv3_large_100', pretrained=True, num_classes=10)
>>> model.num_classes
10
```
""" """
# Parameters that aren't supported by all models or are intended to only override model defaults if set # Parameters that aren't supported by all models or are intended to only override model defaults if set
# should default to None in command line args/cfg. Remove them if they are present and not set so that # should default to None in command line args/cfg. Remove them if they are present and not set so that

Loading…
Cancel
Save