!241 修正文档中关于evaluation_strategy更名为eval_strategy的问题

Merge pull request !241 from 幽若/master-0616

docs/zh/basic_tutorial/metrics.md

@@ -55,7 +55,8 @@ small_eval_dataset = tokenized_datasets["validation"].shuffle(seed=42).select(ra
 from openmind import TrainingArguments, Trainer, metrics
 import numpy as np
 
-training_args = TrainingArguments(output_dir="test_trainer", evaluation_strategy="epoch")
+# 在4.51.3版本的transformers中，evaluation_strategy参数已更名为eval_strategy, 参见https://github.com/huggingface/transformers/blob/v4.51.3/src/transformers/training_args.py#L239
+training_args = TrainingArguments(output_dir="test_trainer", eval_strategy="epoch")
 
 def compute_metrics(eval_pred):
     logits, labels = eval_pred

docs/zh/quick_start.md

@@ -259,7 +259,7 @@ openMind Library提供了一个`Trainer`类来实现训练模型所需功能。
         per_device_train_batch_size=4,
         per_device_eval_batch_size=4,
         num_train_epochs=10,
-        evaluation_strategy="epoch",
+        eval_strategy="epoch",
     )
     ```
 

src/openmind/integrations/transformers/bitsandbytes.py

@@ -15,6 +15,7 @@
 import importlib
 import importlib.metadata
 import sys
+from functools import wraps
 from typing import Any, Dict, List, Optional
 
 import torch
@@ -30,6 +31,7 @@ from transformers.utils import (
     is_torch_npu_available,
     is_torch_xpu_available,
 )
+from transformers.utils.quantization_config import QuantizationMethod
 
 
 def create_quantized_param_patch(
@@ -187,7 +189,6 @@ def is_bitsandbytes_available_patch():
 
 
 def validate_environment_patch(self, *args, **kwargs):
-
     if not is_accelerate_available():
         raise ImportError(
             f"Using `bitsandbytes` 4-bit quantization requires Accelerate: `pip install 'accelerate>={ACCELERATE_MIN_VERSION}'`"
@@ -235,6 +236,57 @@ def validate_environment_patch(self, *args, **kwargs):
         )
 
 
+@wraps(torch.nn.Module.to)
+def to(self, *args, **kwargs):
+    # For BNB/GPTQ models, we prevent users from casting the model to another dtype to restrict unwanted behaviours.
+    # the correct API should be to load the model with the desired dtype directly through `from_pretrained`.
+    dtype_present_in_args = "dtype" in kwargs
+
+    if not dtype_present_in_args:
+        for arg in args:
+            if isinstance(arg, torch.dtype):
+                dtype_present_in_args = True
+                break
+
+    if getattr(self, "quantization_method", None) == QuantizationMethod.HQQ:
+        raise ValueError("`.to` is not supported for HQQ-quantized models.")
+
+    if dtype_present_in_args and getattr(self, "quantization_method", None) == QuantizationMethod.QUARK:
+        raise ValueError("Casting a Quark quantized model to a new `dtype` is not supported.")
+
+    # Checks if the model has been loaded in 4-bit or 8-bit with BNB
+    if getattr(self, "quantization_method", None) == QuantizationMethod.BITS_AND_BYTES:
+        if dtype_present_in_args:
+            raise ValueError(
+                "You cannot cast a bitsandbytes model in a new `dtype`. Make sure to load the model using `from_pretrained` using the"
+                " desired `dtype` by passing the correct `torch_dtype` argument."
+            )
+
+        if getattr(self, "is_loaded_in_8bit", False):
+            raise ValueError(
+                "`.to` is not supported for `8-bit` bitsandbytes models. Please use the model as it is, since the"
+                " model has already been set to the correct devices and casted to the correct `dtype`."
+            )
+        elif is_torch_npu_available():
+            if version.parse(importlib.metadata.version("bitsandbytes-npu-beta")) < version.parse("0.43.1"):
+                raise ValueError(
+                    "Calling `to()` is not supported for `4-bit` quantized models with the installed version of bitsandbytes. "
+                    f"The current device is `{self.device}`. If you intended to move the model, please install bitsandbytes >= 0.43.2."
+                )
+        elif version.parse(importlib.metadata.version("bitsandbytes")) < version.parse("0.43.2"):
+            raise ValueError(
+                "Calling `to()` is not supported for `4-bit` quantized models with the installed version of bitsandbytes. "
+                f"The current device is `{self.device}`. If you intended to move the model, please install bitsandbytes >= 0.43.2."
+            )
+    elif getattr(self, "quantization_method", None) == QuantizationMethod.GPTQ:
+        if dtype_present_in_args:
+            raise ValueError(
+                "You cannot cast a GPTQ model in a new `dtype`. Make sure to load the model using `from_pretrained` using the desired"
+                " `dtype` by passing the correct `torch_dtype` argument."
+            )
+    return super(PreTrainedModel, self).to(*args, **kwargs)
+
+
 def is_serializable_patch(self, safe_serialization=None):
     _is_4bit_serializable = version.parse(importlib.metadata.version("bitsandbytes-npu-beta")) >= version.parse(
         "0.41.3"
@@ -249,7 +301,6 @@ def is_serializable_patch(self, safe_serialization=None):
 
 
 def patch_bnb():
-
     for module in sys.modules.values():
         if hasattr(module, "is_bitsandbytes_available"):
             module.is_bitsandbytes_available = is_bitsandbytes_available_patch
@@ -267,3 +318,4 @@ def patch_bnb():
         validate_environment_patch,
     )
     setattr(transformers.quantizers.quantizer_bnb_4bit.Bnb4BitHfQuantizer, "is_serializable", is_serializable_patch)
+    setattr(transformers.modeling_utils.PreTrainedModel, "to", to)