Update base to transformers v4.30.2 (#81)

* Add recipe_name to default file names

* Upgrade to transformers release V4.30.2 (#62)

* Update trainer and model flows to accommodate sparseml

Disable FP16 on QAT start (#12)

* Override LRScheduler when using LRModifiers

* Disable FP16 on QAT start

* keep wrapped scaler object for training after disabling

Using QATMatMul in DistilBERT model class (#41)

Removed double quantization of output of context layer. (#45)

Fix DataParallel validation forward signatures (#47)

* Fix: DataParallel validation forward signatures

* Update: generalize forward_fn selection

Best model after epoch (#46)

fix sclaer check for non fp16 mode in trainer (#38)

Mobilebert QAT (#55)

* Remove duplicate quantization of vocabulary.

enable a QATWrapper for non-parameterized matmuls in BERT self attention (#9)

* Utils and auxillary changes

update Zoo stub loading for SparseZoo 1.1 refactor (#54)

add flag to signal NM integration is active (#32)

Add recipe_name to file names

* Fix errors introduced in manual cherry-pick upgrade

Co-authored-by: Benjamin Fineran <[email protected]>

* update build versions for NM fork pypi push (#74)

* fix nightly package name (#75)

* add make build command (#76)

* add GHA workflow files to build nightly and release packages (#77)

* add GHA workflow files to build nightly and release packages

* fix name

---------

Co-authored-by: dhuang <[email protected]>

* bump up version to 1.6.0 (#79)

Co-authored-by: dhuang <[email protected]>

---------

Co-authored-by: Konstantin <[email protected]>
Co-authored-by: Konstantin Gulin <[email protected]>
Co-authored-by: dhuangnm <[email protected]>
Co-authored-by: dhuang <[email protected]>

Author: 4 people

.github/workflows/build-nightly.yml

@@ -0,0 +1,33 @@
+name: build-nightly
+run-name: ${{ github.workflow }} is to create nightly wheel file for pypi
+on:
+  push:
+    branches:
+      - 'main'
+  schedule:
+    - cron: '0 0 * * *'
+jobs:
+  build-nightly:
+    runs-on: ubuntu-22.04
+    permissions:
+      id-token: write
+      contents: read
+    steps:
+      - uses: aws-actions/configure-aws-credentials@v2
+        with:
+          role-to-assume: arn:aws:iam::498127099666:role/WebIdentity-nm-github-actions-only
+          aws-region: us-east-1
+      - uses: actions/checkout@v3
+      - run: | 
+             pwd
+             sudo apt-get install python3-pip
+             pip3 --version
+             sudo pip3 install virtualenv
+             virtualenv venv
+             source venv/bin/activate
+             pip install -e .
+             make -B build
+             deactivate
+             ls dist/
+             aws s3 ls s3://nm-github-actions/
+             aws s3 cp dist/*nightly*.whl s3://nm-github-actions/transformers/

.github/workflows/build-release.yml

@@ -0,0 +1,33 @@
+name: build-release
+run-name: ${{ github.workflow }} is to create release wheel file for pypi
+on:
+  push:
+    branches:
+      - 'release/[0-9]+.[0-9]+'
+      
+jobs:
+  build-release:
+    runs-on: ubuntu-22.04
+    permissions:
+      id-token: write
+      contents: read
+    steps:
+      - uses: aws-actions/configure-aws-credentials@v2
+        with:
+          role-to-assume: arn:aws:iam::498127099666:role/WebIdentity-nm-github-actions-only
+          aws-region: us-east-1
+      - uses: actions/checkout@v3
+      - run: | 
+             pwd
+             sudo apt-get install python3-pip
+             pip3 --version
+             sudo pip3 install virtualenv
+             virtualenv venv
+             source venv/bin/activate
+             pip install -e .
+             sed -i 's/is_release = False/is_release = True/g' src/transformers/version.py
+             make -B build
+             deactivate
+             ls dist/
+             aws s3 ls s3://nm-github-actions/
+             aws s3 cp dist/*.whl s3://nm-github-actions/transformers/

Makefile

@@ -1,4 +1,4 @@
-.PHONY: deps_table_update modified_only_fixup extra_style_checks quality style fixup fix-copies test test-examples
+.PHONY: deps_table_update modified_only_fixup extra_style_checks quality style fixup fix-copies test test-examples build
 
 # make sure to test the local checkout in scripts and not the pre-installed one (don't use quotes!)
 export PYTHONPATH = src
@@ -119,3 +119,7 @@ build-release:
 	python setup.py bdist_wheel
 	python setup.py sdist
 	python utils/check_build.py
+
+# neuralmagic: creates wheel file
+build:
+	python3 setup.py sdist bdist_wheel

setup.py

@@ -423,17 +423,22 @@ def run(self):
     deps["tqdm"],  # progress bars in model download and training scripts
 ]
 
+# default variable to be overwritten by the version.py file
+version = "unknown"
+# load and overwrite version and release info from version.py
+exec(open(os.path.join("src", "transformers", "version.py")).read())
+
 setup(
-    name="transformers",
-    version="4.34.1",  # expected format is one of x.y.z.dev0, or x.y.z.rc1 or x.y.z (no to dashes, yes to dots)
+    name="nm-transformers" if is_release else "nm-transformers-nightly",
+    version=version,  # major.minor.patch to match NM repos, fourth entry is either transformers base version or nightly date
     author="The Hugging Face team (past and future) with the help of all our contributors (https://github.com/huggingface/transformers/graphs/contributors)",
     author_email="[email protected]",
     description="State-of-the-art Machine Learning for JAX, PyTorch and TensorFlow",
     long_description=open("README.md", "r", encoding="utf-8").read(),
     long_description_content_type="text/markdown",
     keywords="NLP vision speech deep learning transformer pytorch tensorflow jax BERT GPT-2 Wav2Vec2 ViT",
     license="Apache 2.0 License",
-    url="https://github.com/huggingface/transformers",
+    url="https://github.com/neuralmagic/transformers",
     package_dir={"": "src"},
     packages=find_packages("src"),
     include_package_data=True,

src/transformers/__init__.py

@@ -18,7 +18,7 @@
 # to defer the actual importing for when the objects are requested. This way `import transformers` provides the names
 # in the namespace without actually importing anything (and especially none of the backends).
 
-__version__ = "4.34.1"
+from .version import *
 
 from typing import TYPE_CHECKING
 

src/transformers/hf_argparser.py

@@ -23,8 +23,16 @@
 from pathlib import Path
 from typing import Any, Callable, Dict, Iterable, List, Literal, NewType, Optional, Tuple, Union, get_type_hints
 
+import os
 import yaml
 
+from sparsezoo import Model
+
+from .utils.logging import get_logger
+
+
+logger = get_logger(__name__)
+
 
 DataClass = NewType("DataClass", Any)
 DataClassType = NewType("DataClassType", Any)
@@ -341,12 +349,17 @@ def parse_args_into_dataclasses(
             # additional namespace.
             outputs.append(namespace)
         if return_remaining_strings:
-            return (*outputs, remaining_args)
+            return tuple(
+                *[_download_dataclass_zoo_stub_files(output) for output in outputs],
+                remaining_args,
+            )
         else:
             if remaining_args:
                 raise ValueError(f"Some specified arguments are not used by the HfArgumentParser: {remaining_args}")
 
-            return (*outputs,)
+            return tuple(
+                [_download_dataclass_zoo_stub_files(output) for output in outputs]
+            )
 
     def parse_dict(self, args: Dict[str, Any], allow_extra_keys: bool = False) -> Tuple[DataClass, ...]:
         """
@@ -374,7 +387,9 @@ def parse_dict(self, args: Dict[str, Any], allow_extra_keys: bool = False) -> Tu
             outputs.append(obj)
         if not allow_extra_keys and unused_keys:
             raise ValueError(f"Some keys are not used by the HfArgumentParser: {sorted(unused_keys)}")
-        return tuple(outputs)
+        return tuple(
+            [_download_dataclass_zoo_stub_files(output) for output in outputs]
+        )
 
     def parse_json_file(self, json_file: str, allow_extra_keys: bool = False) -> Tuple[DataClass, ...]:
         """
@@ -417,3 +432,28 @@ def parse_yaml_file(self, yaml_file: str, allow_extra_keys: bool = False) -> Tup
         """
         outputs = self.parse_dict(yaml.safe_load(Path(yaml_file).read_text()), allow_extra_keys=allow_extra_keys)
         return tuple(outputs)
+
+def _download_dataclass_zoo_stub_files(data_class: DataClass):
+    for name, val in data_class.__dict__.items():
+        if not isinstance(val, str) or "recipe" in name or not val.startswith("zoo:"):
+            continue
+
+        logger.info(f"Downloading framework files for SparseZoo stub: {val}")
+
+        zoo_model = Model(val)
+        framework_file_paths = [file.path for file in zoo_model.training.default.files]
+        assert framework_file_paths, "Unable to download any framework files for SparseZoo stub {val}"
+        framework_file_names = [os.path.basename(path) for path in framework_file_paths]
+        if "pytorch_model.bin" not in framework_file_names or ("config.json" not in framework_file_names):
+            raise RuntimeError(
+                "Unable to find 'pytorch_model.bin' and 'config.json' in framework "
+                f"files downloaded from {val}. Found {framework_file_names}. Check "
+                "if the given stub is for a transformers repo model"
+            )
+        framework_dir_path = Path(framework_file_paths[0]).parent.absolute()
+
+        logger.info(f"Overwriting argument {name} to downloaded {framework_dir_path}")
+
+        data_class.__dict__[name] = str(framework_dir_path)
+
+    return data_class

src/transformers/models/bert/modeling_bert.py

@@ -241,6 +241,22 @@ def forward(
         return embeddings
 
 
+class QATMatMul(nn.Module):
+    def __init__(self):
+        super().__init__()
+
+        # behaves like normal torch.matmul unless a SparseML QuantizationModifier
+        # is initialized
+        self.wrap_qat = True
+        self.qat_wrapper_kwargs = {
+            "num_inputs": 2,
+            "input_qconfigs": ["asymmetric", "symmetric"],
+        }
+
+    def forward(self, a: torch.Tensor, b: torch.Tensor):
+        return torch.matmul(a, b)
+
+
 class BertSelfAttention(nn.Module):
     def __init__(self, config, position_embedding_type=None):
         super().__init__()
@@ -258,6 +274,11 @@ def __init__(self, config, position_embedding_type=None):
         self.key = nn.Linear(config.hidden_size, self.all_head_size)
         self.value = nn.Linear(config.hidden_size, self.all_head_size)
 
+        # non-parameterized matmuls will behave as normal torch.matmul ops unless
+        # Quantization-Aware-Training is invoked
+        self.attention_scores_matmul = QATMatMul()
+        self.context_layer_matmul = QATMatMul()
+
         self.dropout = nn.Dropout(config.attention_probs_dropout_prob)
         self.position_embedding_type = position_embedding_type or getattr(
             config, "position_embedding_type", "absolute"
@@ -322,7 +343,7 @@ def forward(
             past_key_value = (key_layer, value_layer)
 
         # Take the dot product between "query" and "key" to get the raw attention scores.
-        attention_scores = torch.matmul(query_layer, key_layer.transpose(-1, -2))
+        attention_scores = self.attention_scores_matmul(query_layer, key_layer.transpose(-1, -2))
 
         if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
             query_length, key_length = query_layer.shape[2], key_layer.shape[2]
@@ -362,7 +383,7 @@ def forward(
         if head_mask is not None:
             attention_probs = attention_probs * head_mask
 
-        context_layer = torch.matmul(attention_probs, value_layer)
+        context_layer = self.context_layer_matmul(attention_probs, value_layer)
 
         context_layer = context_layer.permute(0, 2, 1, 3).contiguous()
         new_context_layer_shape = context_layer.size()[:-2] + (self.all_head_size,)

src/transformers/models/distilbert/modeling_distilbert.py

@@ -88,6 +88,38 @@ def _create_sinusoidal_embeddings(n_pos: int, dim: int, out: torch.Tensor):
     out.detach_()
 
 
+class QATAttentionScores(nn.Module):
+    def __init__(self):
+        super().__init__()
+
+        # behaves like normal torch.matmul unless a SparseML QuantizationModifier
+        # is initialized
+        self.wrap_qat = True
+        self.qat_wrapper_kwargs = {
+            "num_inputs": 2,
+            "input_qconfigs": ["asymmetric", "symmetric"],
+        }
+
+    def forward(self, a: torch.Tensor, b: torch.Tensor):
+        return torch.matmul(a, b)
+
+class QATContextLayer(nn.Module):
+    def __init__(self):
+        super().__init__()
+
+        # behaves like normal torch.matmul unless a SparseML QuantizationModifier
+        # is initialized
+        self.wrap_qat = True
+        self.qat_wrapper_kwargs = {
+            "num_inputs": 2,
+            "num_outputs": 0,
+            "input_qconfigs": ["asymmetric", "symmetric"],
+        }
+
+    def forward(self, a: torch.Tensor, b: torch.Tensor):
+        return torch.matmul(a, b)
+
+
 class Embeddings(nn.Module):
     def __init__(self, config: PretrainedConfig):
         super().__init__()
@@ -159,6 +191,11 @@ def __init__(self, config: PretrainedConfig):
         self.pruned_heads: Set[int] = set()
         self.attention_head_size = self.dim // self.n_heads
 
+        # non-parameterized matmuls will behave as normal torch.matmul ops unless
+        # Quantization-Aware-Training is invoked
+        self.attention_scores_matmul = QATAttentionScores()
+        self.context_layer_matmul = QATContextLayer()
+
     def prune_heads(self, heads: List[int]):
         if len(heads) == 0:
             return
@@ -217,7 +254,7 @@ def unshape(x: torch.Tensor) -> torch.Tensor:
         v = shape(self.v_lin(value))  # (bs, n_heads, k_length, dim_per_head)
 
         q = q / math.sqrt(dim_per_head)  # (bs, n_heads, q_length, dim_per_head)
-        scores = torch.matmul(q, k.transpose(2, 3))  # (bs, n_heads, q_length, k_length)
+        scores = self.attention_scores_matmul(q, k.transpose(2, 3))  # (bs, n_heads, q_length, k_length)
         mask = (mask == 0).view(mask_reshp).expand_as(scores)  # (bs, n_heads, q_length, k_length)
         scores = scores.masked_fill(
             mask, torch.tensor(torch.finfo(scores.dtype).min)
@@ -230,7 +267,7 @@ def unshape(x: torch.Tensor) -> torch.Tensor:
         if head_mask is not None:
             weights = weights * head_mask
 
-        context = torch.matmul(weights, v)  # (bs, n_heads, q_length, dim_per_head)
+        context = self.context_layer_matmul(weights, v)  # (bs, n_heads, q_length, dim_per_head)
         context = unshape(context)  # (bs, q_length, dim)
         context = self.out_lin(context)  # (bs, q_length, dim)
 
@@ -688,7 +725,6 @@ def forward(
             loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
         """
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
-
         dlbrt_output = self.distilbert(
             input_ids=input_ids,
             attention_mask=attention_mask,

src/transformers/models/mobilebert/modeling_mobilebert.py

@@ -169,6 +169,23 @@ def forward(self, input_tensor: torch.Tensor) -> torch.Tensor:
 
 NORM2FN = {"layer_norm": nn.LayerNorm, "no_norm": NoNorm}
 
+class QATEmbeddingTransformation(nn.Module):
+    def __init__(self, embedded_input_size, hidden_size):
+        super().__init__()
+
+        # Behaves like normal Linear module unless a SparseML QuantizationModifier
+        # is initialized.
+        # When initialized, does not quantize inputs.
+        # Only weights are quantized (inputs come quantized from embeddings)
+        self.linear = nn.Linear(embedded_input_size, hidden_size)
+        self.wrap_qat = True
+        self.qat_wrapper_kwargs = {
+            "num_inputs": 0,
+            "num_outputs": 1,
+        }
+
+    def forward(self, x: torch.Tensor):
+        return self.linear(x)
 
 class MobileBertEmbeddings(nn.Module):
     """Construct the embeddings from word, position and token_type embeddings."""
@@ -185,7 +202,7 @@ def __init__(self, config):
 
         embed_dim_multiplier = 3 if self.trigram_input else 1
         embedded_input_size = self.embedding_size * embed_dim_multiplier
-        self.embedding_transformation = nn.Linear(embedded_input_size, config.hidden_size)
+        self.embedding_transformation = QATEmbeddingTransformation(embedded_input_size, config.hidden_size)
 
         self.LayerNorm = NORM2FN[config.normalization_type](config.hidden_size)
         self.dropout = nn.Dropout(config.hidden_dropout_prob)