HFQuantizer implementation for compressed-tensors library

docs/source/en/main_classes/quantization.md

@@ -61,3 +61,6 @@ Learn how to quantize models in the [Quantization](../quantization) guide.
 
 [[autodoc]] FbgemmFp8Config
 
+## CompressedTensorsConfig
+
+[[autodoc]] CompressedTensorsConfig

docs/source/en/quantization/compressed_tensors.md

@@ -0,0 +1,47 @@
+<!--Copyright 2024 The HuggingFace Team. 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.
+
+⚠️ Note that this file is in Markdown but contain specific syntax for our doc-builder (similar to MDX) that may not be
+rendered properly in your Markdown viewer.
+
+-->
+
+# Compressed Tensors
+
+Compressed tensors supports the quantization of models to a variety of formats and provides an extensible
+framework for adding new formats and strategies.
+
+Compressed models can be easily created using [llm-compressor](https://github.com/vllm-project/llm-compressor).
+Alternatively models can be created indepedenty and serialized with a compressed tensors config.
+
+Supported formats include:
+
+ - FP8, INT4, INT8 (for Q/DQ arbitrary precision is allowed for INT)
+ - Activation quantization (static)
+ - Dynamic per-token activation quantization
+ - Supports quantization of arbitrary layer types
+ - Targeted support or ignoring of layers by name or class
+
+## Installation
+
+```bash
+pip install compressed-tensors
+```
+
+
+## Sample Model Load
+```python
+from transformers import AutoModelForCausalLM
+compressed_tensors_model = AutoModelForCausalLM.from_pretrained("nm-testing/tinyllama-oneshot-w4a16-group128-v3")
+```
+
+
+## More Coming Soon!

src/transformers/__init__.py

@@ -934,6 +934,7 @@
         "AqlmConfig",
         "AwqConfig",
         "BitsAndBytesConfig",
+        "CompressedTensorsConfig",
         "EetqConfig",
         "FbgemmFp8Config",
         "GPTQConfig",
@@ -5668,6 +5669,7 @@
         AqlmConfig,
         AwqConfig,
         BitsAndBytesConfig,
+        CompressedTensorsConfig,
         EetqConfig,
         FbgemmFp8Config,
         GPTQConfig,

src/transformers/modeling_utils.py

@@ -3990,7 +3990,7 @@ def from_pretrained(
                 dispatch_model(model, **device_map_kwargs)
 
         if hf_quantizer is not None:
-            hf_quantizer.postprocess_model(model)
+            hf_quantizer.postprocess_model(model, resolved_archive_file=resolved_archive_file)
             model.hf_quantizer = hf_quantizer
 
         if _adapter_model_path is not None:

src/transformers/quantizers/auto.py

@@ -19,6 +19,7 @@
     AqlmConfig,
     AwqConfig,
     BitsAndBytesConfig,
+    CompressedTensorsConfig,
     EetqConfig,
     FbgemmFp8Config,
     GPTQConfig,
@@ -31,6 +32,7 @@
 from .quantizer_awq import AwqQuantizer
 from .quantizer_bnb_4bit import Bnb4BitHfQuantizer
 from .quantizer_bnb_8bit import Bnb8BitHfQuantizer
+from .quantizer_compressed_tensors import CompressedTensorsHfQuantizer
 from .quantizer_eetq import EetqHfQuantizer
 from .quantizer_fbgemm_fp8 import FbgemmFp8HfQuantizer
 from .quantizer_gptq import GptqHfQuantizer
@@ -47,6 +49,7 @@
     "quanto": QuantoHfQuantizer,
     "eetq": EetqHfQuantizer,
     "hqq": HqqHfQuantizer,
+    "compressed-tensors": CompressedTensorsHfQuantizer,
     "fbgemm_fp8": FbgemmFp8HfQuantizer,
 }
 
@@ -59,6 +62,7 @@
     "aqlm": AqlmConfig,
     "quanto": QuantoConfig,
     "hqq": HqqConfig,
+    "compressed-tensors": CompressedTensorsConfig,
     "fbgemm_fp8": FbgemmFp8Config,
 }
 

src/transformers/quantizers/quantizer_compressed_tensors.py

@@ -0,0 +1,71 @@
+# Copyright 2024 The HuggingFace Inc. team. 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.
+from ..utils import is_torch_available, logging
+from ..utils.quantization_config import QuantizationConfigMixin
+from .base import HfQuantizer
+
+
+if is_torch_available():
+    import torch
+
+logger = logging.get_logger(__name__)
+
+
+class CompressedTensorsHfQuantizer(HfQuantizer):
+    """
+    Quantizer for the compressed_tensors package.  Loads and restores models to
+    quantized state with compressed_tensors
+    """
+
+    requires_calibration = False
+    # requires_parameters_quantization = True
+    required_packages = ["compressed_tensors"]
+
+    def __init__(self, quantization_config: QuantizationConfigMixin, **kwargs):
+        super().__init__(quantization_config, **kwargs)
+
+        from compressed_tensors.compressors import ModelCompressor
+
+        self.compressor = ModelCompressor.from_compression_config(quantization_config)
+
+    def validate_environment(self, *args, **kwargs):
+        # check torch and compressed_tensors are available, let ImportError raise otherwise
+        import torch  # noqa
+        from compressed_tensors.compressors import ModelCompressor  # noqa
+
+    def update_torch_dtype(self, torch_dtype: "torch.dtype") -> "torch.dtype":
+        if torch_dtype is None:
+            torch_dtype = torch.float16
+        elif torch_dtype != torch.float16:
+            logger.info(
+                "We suggest you to set `torch_dtype=torch.float16` for better efficiency with compressed_tensors."
+            )
+        return torch_dtype
+
+    def _process_model_before_weight_loading(self, model, **kwargs):
+        if self.quantization_config.quantization_config is not None:
+            from compressed_tensors.quantization import apply_quantization_config
+
+            apply_quantization_config(model, self.quantization_config.quantization_config)
+
+    def _process_model_after_weight_loading(self, model, resolved_archive_file, **kwargs):
+        self.compressor.decompress(model_path=resolved_archive_file, model=model)
+
+    @property
+    def is_trainable(self):
+        return False
+
+    @property
+    def is_serializable(self):
+        return False

src/transformers/testing_utils.py

@@ -61,6 +61,7 @@
     is_av_available,
     is_bitsandbytes_available,
     is_bs4_available,
+    is_compressed_tensors_available,
     is_cv2_available,
     is_cython_available,
     is_decord_available,
@@ -1117,6 +1118,13 @@ def require_quanto(test_case):
     return unittest.skipUnless(is_quanto_available(), "test requires quanto")(test_case)
 
 
+def require_compressed_tensors(test_case):
+    """
+    Decorator for compressed_tensors dependency
+    """
+    return unittest.skipUnless(is_compressed_tensors_available(), "test requires compressed_tensors")(test_case)
+
+
 def require_fbgemm_gpu(test_case):
     """
     Decorator for fbgemm_gpu dependency

src/transformers/utils/__init__.py

@@ -119,6 +119,7 @@
     is_bitsandbytes_available,
     is_bs4_available,
     is_coloredlogs_available,
+    is_compressed_tensors_available,
     is_cv2_available,
     is_cython_available,
     is_datasets_available,

src/transformers/utils/import_utils.py

@@ -138,6 +138,7 @@ def _is_package_available(pkg_name: str, return_version: bool = False) -> Union[
 # `importlib.metadata.version` doesn't work with `awq`
 _auto_awq_available = importlib.util.find_spec("awq") is not None
 _quanto_available = _is_package_available("quanto")
+_compressed_tensors_available = _is_package_available("compressed_tensors")
 _pandas_available = _is_package_available("pandas")
 _peft_available = _is_package_available("peft")
 _phonemizer_available = _is_package_available("phonemizer")
@@ -888,6 +889,10 @@ def is_quanto_available():
     return _quanto_available
 
 
+def is_compressed_tensors_available():
+    return _compressed_tensors_available
+
+
 def is_auto_gptq_available():
     return _auto_gptq_available
 

src/transformers/utils/quantization_config.py

@@ -42,6 +42,7 @@ class QuantizationMethod(str, Enum):
     QUANTO = "quanto"
     EETQ = "eetq"
     HQQ = "hqq"
+    COMPRESSED_TENSORS = "compressed-tensors"
     FBGEMM_FP8 = "fbgemm_fp8"
 
 
@@ -1050,7 +1051,75 @@ def post_init(self):
             raise ValueError(f"Only support weights in {accepted_weights} but found {self.weights}")
 
 
-@dataclass
+class CompressedTensorsConfig(QuantizationConfigMixin):
+    """
+    This is a wrapper class that handles compressed-tensors quantization config options.
+    It is a wrapper around `compressed_tensors.QuantizationConfig`
+    Args:
+        config_groups (`typing.Dict[str, typing.Union[ForwardRef('QuantizationScheme'), typing.List[str]]]`, *optional*):
+            dictionary mapping group name to a quantization scheme definition
+        format (`str`, *optional*, defaults to `"dense"`):
+            format the model is represented as
+        quantization_status (`QuantizationStatus`, *optional*, defaults to `"initialized"`):
+            status of model in the quantization lifecycle, ie 'initialized', 'calibration', 'frozen'
+        global_compression_ratio (`typing.Union[float, NoneType]`, *optional*):
+            0-1 float percentage of model compression
+        ignore (`typing.Union[typing.List[str], NoneType]`, *optional*):
+            layer names or types to not quantize, supports regex prefixed by 're:'
+        sparsity_config (`typing.Dict[str, typing.Any]`, *optional*):
+            configuration for sparsity compression
+        quant_method (`str`, *optional*, defaults to `"compressed-tensors"`):
+            do not override, should be compressed-tensors
+    """
+
+    def __init__(
+        self,
+        config_groups: Dict[str, Union["QuantizationScheme", List[str]]] = None,  # noqa: F821
+        format: str = "dense",
+        quantization_status: "QuantizationStatus" = "initialized",  # noqa: F821
+        global_compression_ratio: Optional[float] = None,
+        ignore: Optional[List[str]] = None,
+        sparsity_config: Dict[str, Any] = None,
+        quant_method: str = "compressed-tensors",
+        **kwargs,
+    ):
+        from compressed_tensors import QuantizationConfig
+        from compressed_tensors.config import SparsityCompressionConfig
+
+        self.quantization_config = None
+        self.sparsity_config = None
+
+        # parse from dict to load nested QuantizationScheme objects
+        if config_groups:
+            self.quantization_config = QuantizationConfig.parse_obj(
+                {
+                    "config_groups": config_groups,
+                    "quant_method": quant_method,
+                    "format": format,
+                    "quantization_status": quantization_status,
+                    "global_compression_ratio": global_compression_ratio,
+                    "ignore": ignore,
+                }
+            )
+
+        if sparsity_config:
+            self.sparsity_config = SparsityCompressionConfig.load_from_registry(
+                sparsity_config.get("format"), **sparsity_config
+            )
+
+        super().__init__(quant_method=QuantizationMethod.COMPRESSED_TENSORS)
+
+    def to_dict(self) -> Dict[str, Any]:
+        """
+        Serializes this instance to a Python dictionary. Returns:
+            `Dict[str, Any]`: Dictionary of all the attributes that make up this configuration instance.
+        """
+        return {
+            "quantization_config": self.quantization_config.dict(),
+            "sparsity_config": self.sparsity_config.dict(),
+        }
+
+
 class FbgemmFp8Config(QuantizationConfigMixin):
     """
     This is a wrapper class about all possible attributes and features that you can play with a model that has been

tests/quantization/compressed_tensor/test_compressed_tensors.py

@@ -0,0 +1,53 @@
+import gc
+import unittest
+
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from transformers.testing_utils import require_compressed_tensors, require_torch
+from transformers.utils import is_torch_available
+
+
+if is_torch_available():
+    import torch
+
+
+@require_compressed_tensors
+@require_torch
+class CompressedTensorsTest(unittest.TestCase):
+    quantized_model_name = "nm-testing/tinyllama-oneshot-w8a8-test-static-shape-change-v3"
+
+    prompt = "Paris is the capital of which country?"
+
+    def tear_down(self):
+        gc.collect()
+        torch.cuda.empty_cache()
+        gc.collect()
+
+    @classmethod
+    def setUpClass(self):
+        """
+        Setup quantized model
+        """
+        self.tokenizer = AutoTokenizer.from_pretrained(self.quantized_model_name)
+        self.quantized_model = AutoModelForCausalLM.from_pretrained(self.quantized_model_name)
+        self.device = self.quantized_model.device
+
+    def test_quantized_model(self):
+        """Carry out generation"""
+        self.assertIsNotNone(
+            self.quantized_model.config.quantization_config,
+            "quantization_config should not be None",
+        )
+        self.assertTrue(
+            any(
+                key
+                for key, tensor in self.quantized_model.state_dict().items()
+                if "scale" in key and not torch.all(tensor == 1.0)
+            ),
+            "quantized model should load a non-trivail scale into the state dict",
+        )
+        inputs = self.tokenizer(self.prompt, return_tensors="pt").to(self.device)
+        generated_ids = self.quantized_model.generate(**inputs, max_length=50)
+        outputs = self.tokenizer.batch_decode(generated_ids)
+
+        self.assertIsNotNone(outputs)
+        self.tear_down()