[trainer] add tf32-mode control

docs/source/performance.md

@@ -358,8 +358,12 @@ Like all cases with reduced precision this may or may not be satisfactory for yo
 
 If you're already using fp16 or bf16 mixed precision it may help with the throughput as well.
 
+The 🤗 Trainer has this mode enabled by default but can be disabled automatically by passing `--tf32 0`.
+
 Note: tf32 mode is internal to CUDA and can't be accessed directly via `tensor.to(dtype=torch.tf32)` as `torch.tf32` doesn't exit.
 
+Note: you need `torch>=1.7` to enjoy this feature.
+
 
 ### Gradient Checkpointing
 

src/transformers/file_utils.py

@@ -341,7 +341,7 @@ def is_torch_bf16_available():
             return False
         if int(torch.version.cuda.split(".")[0]) < 11:
             return False
-        if not version.parse(torch.__version__) >= version.parse("1.10"):
+        if version.parse(torch.__version__) < version.parse("1.10"):
             return False
         if not hasattr(torch, "autocast"):
             return False
@@ -351,6 +351,24 @@ def is_torch_bf16_available():
         return False
 
 
+def is_torch_tf32_available():
+    if is_torch_available():
+        import torch
+
+        if not torch.cuda.is_available() or torch.version.cuda is None:
+            return False
+        if torch.cuda.get_device_properties(torch.cuda.current_device()).major < 8:
+            return False
+        if int(torch.version.cuda.split(".")[0]) < 11:
+            return False
+        if version.parse(torch.__version__) < version.parse("1.7"):
+            return False
+
+        return True
+    else:
+        return False
+
+
 _torch_fx_available = _torch_onnx_dict_inputs_support_available = False
 if _torch_available:
     torch_version = version.parse(importlib_metadata.version("torch"))

src/transformers/testing_utils.py

@@ -50,6 +50,7 @@
     is_tokenizers_available,
     is_torch_available,
     is_torch_bf16_available,
+    is_torch_tf32_available,
     is_torch_tpu_available,
     is_torchaudio_available,
     is_vision_available,
@@ -495,9 +496,17 @@ def require_torch_gpu(test_case):
 
 
 def require_torch_bf16(test_case):
-    """Decorator marking a test that requires CUDA hardware supporting bf16 and PyTorch >= 1.10."""
+    """Decorator marking a test that requires Ampere or a newer GPU arch, cuda>=11 and torch>=1.10."""
     if not is_torch_bf16_available():
-        return unittest.skip("test requires CUDA hardware supporting bf16 and PyTorch >= 1.10")(test_case)
+        return unittest.skip("test requires Ampere or a newer GPU arch, cuda>=11 and torch>=1.10")(test_case)
+    else:
+        return test_case
+
+
+def require_torch_tf32(test_case):
+    """Decorator marking a test that requires Ampere or a newer GPU arch, cuda>=11 and torch>=1.7."""
+    if not is_torch_tf32_available():
+        return unittest.skip("test requires Ampere or a newer GPU arch, cuda>=11 and torch>=1.7")(test_case)
     else:
         return test_case
 

src/transformers/training_args.py

@@ -29,6 +29,7 @@
     is_sagemaker_dp_enabled,
     is_sagemaker_mp_enabled,
     is_torch_available,
+    is_torch_tf32_available,
     is_torch_tpu_available,
     torch_required,
 )
@@ -227,6 +228,9 @@ class TrainingArguments:
         fp16_full_eval (:obj:`bool`, `optional`, defaults to :obj:`False`):
             Whether to use full float16 evaluation instead of 32-bit. This will be faster and save memory but can harm
             metric values.
+        tf32 (:obj:`bool`, `optional`, defaults to :obj:`True`):
+            Whether to enable tf32 mode, available in Ampere and newer GPU architectures. This is an experimental API
+            and it may change.
         local_rank (:obj:`int`, `optional`, defaults to -1):
             Rank of the process during distributed training.
         xpu_backend (:obj:`str`, `optional`):
@@ -548,6 +552,12 @@ class TrainingArguments:
         default=False,
         metadata={"help": "Whether to use full float16 evaluation instead of 32-bit"},
     )
+    tf32: bool = field(
+        default=True,
+        metadata={
+            "help": "Whether to enable tf32 mode, available in Ampere and newer GPU architectures. This is an experimental API and it may change."
+        },
+    )
     local_rank: int = field(default=-1, metadata={"help": "For distributed training: local_rank"})
     xpu_backend: str = field(
         default=None,
@@ -802,6 +812,9 @@ def __post_init__(self):
                 "Mixed precision training with AMP or APEX (`--fp16` or `--bf16`) and half precision evaluation (`--fp16_full_eval` or `--bf16_full_eval`) can only be used on CUDA devices."
             )
 
+        if is_torch_available() and is_torch_tf32_available():
+            torch.backends.cuda.matmul.allow_tf32 = True if self.tf32 else False
+
         if self.report_to is None:
             logger.info(
                 "The default value for the training argument `--report_to` will change in v5 (from all installed "

tests/test_trainer.py

@@ -57,6 +57,7 @@
     require_torch_gpu,
     require_torch_multi_gpu,
     require_torch_non_multi_gpu,
+    require_torch_tf32,
     require_torch_up_to_2_gpus,
     slow,
 )
@@ -492,6 +493,15 @@ def test_mixed_bf16(self):
 
         # will add more specific tests once there are some bugs to fix
 
+    @require_torch_gpu
+    @require_torch_tf32
+    def test_tf32(self):
+
+        # very basic test
+        trainer = get_regression_trainer(learning_rate=0.1, tf32=True)
+        trainer.train()
+        self.check_trained_model(trainer.model)
+
 
 @require_torch
 @require_sentencepiece