added override for hparams in load_from_ckpt (#1797)

* added override for hparams in load_from_ckpt

* override hparams

* override hparams

* Apply suggestions from code review

Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>

* update doctest

* typo

* chlog

Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Jirka <jirka.borovec@seznam.cz>

CHANGELOG.md

@@ -22,6 +22,8 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 - Added option to provide seed to random generators to ensure reproducibility ([#1572](https://github.com/PyTorchLightning/pytorch-lightning/pull/1572))
 
+- Added override for hparams in `load_from_ckpt` ([#1797](https://github.com/PyTorchLightning/pytorch-lightning/pull/1797))
+
 ### Changed
 
 - Reduction when `batch_size < num_gpus` ([#1609](https://github.com/PyTorchLightning/pytorch-lightning/pull/1609))

pytorch_lightning/core/lightning.py

@@ -16,8 +16,8 @@
 from pytorch_lightning.core.grads import GradInformation
 from pytorch_lightning.core.hooks import ModelHooks
 from pytorch_lightning.core.memory import ModelSummary
+from pytorch_lightning.core.saving import ModelIO, load_hparams_from_tags_csv, update_hparams
 from pytorch_lightning.core.properties import DeviceDtypeModuleMixin
-from pytorch_lightning.core.saving import ModelIO, load_hparams_from_tags_csv
 from pytorch_lightning.overrides.data_parallel import LightningDistributedDataParallel
 from pytorch_lightning.utilities.exceptions import MisconfigurationException
 from pytorch_lightning.utilities import rank_zero_warn
@@ -1439,6 +1439,7 @@ def load_from_checkpoint(
             checkpoint_path: str,
             map_location: Optional[Union[Dict[str, str], str, torch.device, int, Callable]] = None,
             tags_csv: Optional[str] = None,
+            hparam_overrides: Optional[Dict] = None,
             *args, **kwargs
     ) -> 'LightningModule':
         r"""
@@ -1480,6 +1481,7 @@ def __init__(self, hparams):
                 use this method to pass in a .csv file with the hparams you'd like to use.
                 These will be converted into a :class:`~argparse.Namespace` and passed into your
                 :class:`LightningModule` for use.
+            hparam_overrides: A dictionary with keys to override in the hparams
 
         Return:
             :class:`LightningModule` with loaded weights and hyperparameters (if available).
@@ -1503,6 +1505,12 @@ def __init__(self, hparams):
                     tags_csv='/path/to/hparams_file.csv'
                 )
 
+                # override some of the params with new values
+                MyLightningModule.load_from_checkpoint(
+                    PATH,
+                    hparam_overrides={'num_layers': 128, 'pretrained_ckpt_path': NEW_PATH}
+                )
+
                 # or load passing whatever args the model takes to load
                 MyLightningModule.load_from_checkpoint(
                     'path/to/checkpoint.ckpt',
@@ -1521,12 +1529,16 @@ def __init__(self, hparams):
         else:
             checkpoint = torch.load(checkpoint_path, map_location=lambda storage, loc: storage)
 
+        # add the hparams from csv file to checkpoint
         if tags_csv is not None:
-            # add the hparams from csv file to checkpoint
             hparams = load_hparams_from_tags_csv(tags_csv)
             hparams.__setattr__('on_gpu', False)
             checkpoint['hparams'] = vars(hparams)
 
+        # override the hparam keys that were passed in
+        if hparam_overrides is not None:
+            update_hparams(hparams, hparam_overrides)
+
         model = cls._load_model_state(checkpoint, *args, **kwargs)
         return model
 

pytorch_lightning/core/saving.py

@@ -48,6 +48,37 @@ def on_hpc_load(self, checkpoint: Dict[str, Any]) -> None:
         """
 
 
+def update_hparams(hparams: dict, updates: dict) -> None:
+    """
+    Overrides hparams with new values
+
+    >>> hparams = {'c': 4}
+    >>> update_hparams(hparams, {'a': {'b': 2}, 'c': 1})
+    >>> hparams['a']['b'], hparams['c']
+    (2, 1)
+    >>> update_hparams(hparams, {'a': {'b': 4}, 'c': 7})
+    >>> hparams['a']['b'], hparams['c']
+    (4, 7)
+
+    Args:
+        hparams: the original params and also target object
+        updates: new params to be used as update
+
+    """
+    for k, v in updates.items():
+        # if missing, add the key
+        if k not in hparams:
+            hparams[k] = v
+            continue
+
+        # recurse if dictionary
+        if isinstance(v, dict):
+            update_hparams(hparams[k], updates[k])
+        else:
+            # update the value
+            hparams.update({k: v})
+
+
 def load_hparams_from_tags_csv(tags_csv: str) -> Namespace:
     if not os.path.isfile(tags_csv):
         log.warning(f'Missing Tags: {tags_csv}.')