feat: add quantizer algorithm

src/pruna/algorithms/torchao_autoquant.py

@@ -0,0 +1,164 @@
+# Copyright 2025 - Pruna AI GmbH. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import inspect
+from typing import Any, Callable, Dict
+
+import torch
+
+from pruna.algorithms.base.pruna_base import PrunaAlgorithmBase
+from pruna.algorithms.base.tags import AlgorithmTag
+from pruna.config.hyperparameters import Boolean
+from pruna.config.smash_config import SmashConfigPrefixWrapper
+from pruna.engine.model_checks import get_diffusers_transformer_models
+from pruna.engine.save import SAVE_FUNCTIONS
+
+
+class Autoquant(PrunaAlgorithmBase):
+    """
+    Implement autoquantization using the torchao library.
+
+    This algorithm compiles, quantizes and sparsifies weights, gradients, and activations for inference.
+    This algorithm is specifically adapted for Image-Gen models.
+    """
+
+    algorithm_name: str = "torchao_autoquant"
+    group_tags: list[AlgorithmTag] = [AlgorithmTag.QUANTIZER]
+    references: dict[str, str] = {"GitHub": "https://huggingface.co/docs/diffusers/quantization/torchao"}
+    save_fn: SAVE_FUNCTIONS = SAVE_FUNCTIONS.save_before_apply
+    tokenizer_required: bool = False
+    processor_required: bool = False
+    runs_on: list[str] = ["cpu", "cuda"]
+    dataset_required: bool = False
+
+    def get_hyperparameters(self) -> list:
+        """
+        Configure all algorithm-specific hyperparameters with ConfigSpace.
+
+        Returns
+        -------
+        list
+            The hyperparameters.
+        """
+        return [
+            Boolean("compile", default=True, meta=dict(desc="Whether to compile the model after quantization or not.")),
+        ]
+
+    def model_check_fn(self, model: Any) -> bool:
+        """
+        Check if the model is a torch.nn.Module.
+
+        Parameters
+        ----------
+        model : Any
+            The model to check.
+
+        Returns
+        -------
+        bool
+            True if the model is a causal language model, False otherwise.
+        """
+        transformer_models = get_diffusers_transformer_models()
+
+        if isinstance(model, tuple(transformer_models)):
+            return True
+
+        for _, attr_value in inspect.getmembers(model):
+            if isinstance(attr_value, tuple(transformer_models)):
+                return True
+        return isinstance(model, torch.nn.Module)
+
+    def _apply(self, model: Any, smash_config: SmashConfigPrefixWrapper) -> Any:
+        """
+        Quantize the model.
+
+        Parameters
+        ----------
+        model : Any
+            The model to quantize.
+        smash_config : SmashConfigPrefixWrapper
+            The configuration for the quantization.
+
+        Returns
+        -------
+        Any
+            The quantized model.
+        """
+        transformer_models = get_diffusers_transformer_models()
+
+        working_model = model.transformer if isinstance(model.transformer, tuple(transformer_models)) else model
+
+        if smash_config["compile"]:
+            working_model.torch_compiler = TorchCompiler(working_model)
+
+            # Now we can compile the model
+            working_model = working_model.torch_compiler.compile()
+        working_model = self.import_algorithm_packages()["autoquant"](working_model, error_on_unseen=False)
+
+        if isinstance(model.transformer, tuple(transformer_models)):
+            model.transformer = working_model
+        else:
+            model = working_model
+
+        return model
+
+    def import_algorithm_packages(self) -> Dict[str, Any]:
+        """
+        Provide a algorithm packages for the algorithm.
+
+        Returns
+        -------
+        Dict[str, Any]
+            The algorithm packages.
+        """
+        from torchao.quantization import autoquant
+
+        return dict(autoquant=autoquant)
+
+
+class TorchCompiler(object):
+    """
+    A class that compiles a PyTorch model using the pre-defined compilation options.
+
+    Parameters
+    ----------
+    model : torch.nn.Module
+        The PyTorch model to be compiled.
+    """
+
+    def __init__(self, model: Callable[..., Any]) -> None:
+        """
+        Initialize the TorchCompiler with a model and a configuration.
+
+        Parameters
+        ----------
+        model : Callable[..., Any]
+            The PyTorch model to be compiled.
+        smash_config : dict
+            A configuration dictionary that contains the settings for the compilation process.
+        """
+        self.model = model
+
+    def compile(self) -> Callable[..., Any]:
+        """
+        Compile the PyTorch model using options provided in the `smash_config`.
+
+        Returns
+        -------
+        torch.nn.Module
+            The compiled PyTorch model.
+        """
+        self.model = torch.compile(self.model, mode="max-autotune-no-cudagraphs", fullgraph=True)
+
+        return self.model

tests/algorithms/testers/autoquant.py

@@ -0,0 +1,18 @@
+import pytest
+
+from pruna.algorithms.torchao_autoquant import Autoquant
+
+from .base_tester import AlgorithmTesterBase
+
+
+# This is classified as high because autoquant requires INT8 tensor cores which require GPUs with
+# Turing architecture and above
+@pytest.mark.high
+class TestAutoquant(AlgorithmTesterBase):
+    """Test the Autoquant quantizer."""
+
+    models = ["flux_tiny_random"]
+    reject_models = ["dummy_lambda"]
+    allow_pickle_files = False
+    algorithm_class = Autoquant
+    metrics = ["cmmd"]