===================================BUG REPORT===================================
Welcome to bitsandbytes. For bug reports, please submit your error trace to: https://github.com/TimDettmers/bitsandbytes/issues
For effortless bug reporting copy-paste your error into this form: https://docs.google.com/forms/d/e/1FAIpQLScPB8emS3Thkp66nvqwmjTEgxp8Y9ufuWTzFyr9kJ5AoI47dQ/viewform?usp=sf_link
================================================================================
CUDA SETUP: Loading binary d:\anaconda\envs\roboflow\lib\site-packages\bitsandbytes\libbitsandbytes_cuda116.dll...
{
"name": "AttributeError",
"message": "function 'cadam32bit_grad_fp32' not found",
"stack": "---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
d:\\github_projet\\transformers-code\\01-Getting Started\\02-pipeline\\pipeline.ipynb 单元格 2 line 1
----> <a href='vscode-notebook-cell:/d%3A/github_projet/transformers-code/01-Getting%20Started/02-pipeline/pipeline.ipynb#W1sZmlsZQ%3D%3D?line=0'>1</a> from transformers.pipelines import SUPPORTED_TASKS
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\transformers\\pipelines\\__init__.py:44
34 from ..tokenization_utils import PreTrainedTokenizer
35 from ..utils import (
36 HUGGINGFACE_CO_RESOLVE_ENDPOINT,
37 is_kenlm_available,
(...)
42 logging,
43 )
---> 44 from .audio_classification import AudioClassificationPipeline
45 from .automatic_speech_recognition import AutomaticSpeechRecognitionPipeline
46 from .base import (
47 ArgumentHandler,
48 CsvPipelineDataFormat,
(...)
56 infer_framework_load_model,
57 )
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\transformers\\pipelines\\audio_classification.py:21
18 import requests
20 from ..utils import add_end_docstrings, is_torch_available, logging
---> 21 from .base import PIPELINE_INIT_ARGS, Pipeline
24 if is_torch_available():
25 from ..models.auto.modeling_auto import MODEL_FOR_AUDIO_CLASSIFICATION_MAPPING
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\transformers\\pipelines\\base.py:35
33 from ..feature_extraction_utils import PreTrainedFeatureExtractor
34 from ..image_processing_utils import BaseImageProcessor
---> 35 from ..modelcard import ModelCard
36 from ..models.auto.configuration_auto import AutoConfig
37 from ..tokenization_utils import PreTrainedTokenizer
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\transformers\\modelcard.py:48
31 from . import __version__
32 from .models.auto.modeling_auto import (
33 MODEL_FOR_AUDIO_CLASSIFICATION_MAPPING_NAMES,
34 MODEL_FOR_CAUSAL_LM_MAPPING_NAMES,
(...)
46 MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING_NAMES,
47 )
---> 48 from .training_args import ParallelMode
49 from .utils import (
50 MODEL_CARD_NAME,
51 cached_file,
(...)
57 logging,
58 )
61 TASK_MAPPING = {
62 \"text-generation\": MODEL_FOR_CAUSAL_LM_MAPPING_NAMES,
63 \"image-classification\": MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING_NAMES,
(...)
74 \"zero-shot-image-classification\": MODEL_FOR_ZERO_SHOT_IMAGE_CLASSIFICATION_MAPPING_NAMES,
75 }
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\transformers\\training_args.py:67
64 import torch.distributed as dist
66 if is_accelerate_available():
---> 67 from accelerate.state import AcceleratorState, PartialState
68 from accelerate.utils import DistributedType
70 if is_torch_tpu_available(check_device=False):
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\accelerate\\__init__.py:3
1 __version__ = \"0.21.0\"
----> 3 from .accelerator import Accelerator
4 from .big_modeling import (
5 cpu_offload,
6 cpu_offload_with_hook,
(...)
11 load_checkpoint_and_dispatch,
12 )
13 from .data_loader import skip_first_batches
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\accelerate\\accelerator.py:35
32 import torch
33 import torch.utils.hooks as hooks
---> 35 from .checkpointing import load_accelerator_state, load_custom_state, save_accelerator_state, save_custom_state
36 from .data_loader import DataLoaderDispatcher, prepare_data_loader, skip_first_batches
37 from .logging import get_logger
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\accelerate\\checkpointing.py:24
21 import torch
22 from torch.cuda.amp import GradScaler
---> 24 from .utils import (
25 MODEL_NAME,
26 OPTIMIZER_NAME,
27 RNG_STATE_NAME,
28 SCALER_NAME,
29 SCHEDULER_NAME,
30 get_pretty_name,
31 is_tpu_available,
32 is_xpu_available,
33 save,
34 )
37 if is_tpu_available(check_device=False):
38 import torch_xla.core.xla_model as xm
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\accelerate\\utils\\__init__.py:131
121 if is_deepspeed_available():
122 from .deepspeed import (
123 DeepSpeedEngineWrapper,
124 DeepSpeedOptimizerWrapper,
(...)
128 HfDeepSpeedConfig,
129 )
--> 131 from .bnb import has_4bit_bnb_layers, load_and_quantize_model
132 from .fsdp_utils import load_fsdp_model, load_fsdp_optimizer, save_fsdp_model, save_fsdp_optimizer
133 from .launch import (
134 PrepareForLaunch,
135 _filter_args,
(...)
140 prepare_tpu,
141 )
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\accelerate\\utils\\bnb.py:42
31 from .modeling import (
32 find_tied_parameters,
33 get_balanced_memory,
(...)
37 set_module_tensor_to_device,
38 )
41 if is_bnb_available():
---> 42 import bitsandbytes as bnb
44 from copy import deepcopy
47 logger = logging.getLogger(__name__)
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\__init__.py:6
1 # Copyright (c) Facebook, Inc. and its affiliates.
2 #
3 # This source code is licensed under the MIT license found in the
4 # LICENSE file in the root directory of this source tree.
----> 6 from . import cuda_setup, utils, research
7 from .autograd._functions import (
8 MatmulLtState,
9 bmm_cublas,
(...)
13 matmul_4bit
14 )
15 from .cextension import COMPILED_WITH_CUDA
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\research\\__init__.py:1
----> 1 from . import nn
2 from .autograd._functions import (
3 switchback_bnb,
4 matmul_fp8_global,
5 matmul_fp8_mixed,
6 )
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\research\
n\\__init__.py:1
----> 1 from .modules import LinearFP8Mixed, LinearFP8Global
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\research\
n\\modules.py:8
5 from torch import Tensor, device, dtype, nn
7 import bitsandbytes as bnb
----> 8 from bitsandbytes.optim import GlobalOptimManager
9 from bitsandbytes.utils import OutlierTracer, find_outlier_dims
11 T = TypeVar(\"T\", bound=\"torch.nn.Module\")
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\optim\\__init__.py:8
1 # Copyright (c) Facebook, Inc. and its affiliates.
2 #
3 # This source code is licensed under the MIT license found in the
4 # LICENSE file in the root directory of this source tree.
6 from bitsandbytes.cextension import COMPILED_WITH_CUDA
----> 8 from .adagrad import Adagrad, Adagrad8bit, Adagrad32bit
9 from .adam import Adam, Adam8bit, Adam32bit, PagedAdam, PagedAdam8bit, PagedAdam32bit
10 from .adamw import AdamW, AdamW8bit, AdamW32bit, PagedAdamW, PagedAdamW8bit, PagedAdamW32bit
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\optim\\adagrad.py:5
1 # Copyright (c) Facebook, Inc. and its affiliates.
2 #
3 # This source code is licensed under the MIT license found in the
4 # LICENSE file in the root directory of this source tree.
----> 5 from bitsandbytes.optim.optimizer import Optimizer1State
8 class Adagrad(Optimizer1State):
9 def __init__(
10 self,
11 params,
(...)
21 block_wise=True,
22 ):
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\optim\\optimizer.py:12
8 from itertools import chain
10 import torch
---> 12 import bitsandbytes.functional as F
15 class MockArgs:
16 def __init__(self, initial_data):
File d:\\anaconda\\envs\\roboflow\\lib\\site-packages\\bitsandbytes\\functional.py:31
29 \"\"\"C FUNCTIONS FOR OPTIMIZERS\"\"\"
30 str2optimizer32bit = {}
---> 31 str2optimizer32bit[\"adam\"] = (lib.cadam32bit_grad_fp32, lib.cadam32bit_grad_fp16, lib.cadam32bit_grad_bf16)
32 str2optimizer32bit[\"momentum\"] = (
33 lib.cmomentum32bit_grad_32,
34 lib.cmomentum32bit_grad_16,
35 )
36 str2optimizer32bit[\"rmsprop\"] = (
37 lib.crmsprop32bit_grad_32,
38 lib.crmsprop32bit_grad_16,
39 )
File d:\\anaconda\\envs\\roboflow\\lib\\ctypes\\__init__.py:386, in CDLL.__getattr__(self, name)
384 if name.startswith('__') and name.endswith('__'):
385 raise AttributeError(name)
--> 386 func = self.__getitem__(name)
387 setattr(self, name, func)
388 return func
File d:\\anaconda\\envs\\roboflow\\lib\\ctypes\\__init__.py:391, in CDLL.__getitem__(self, name_or_ordinal)
390 def __getitem__(self, name_or_ordinal):
--> 391 func = self._FuncPtr((name_or_ordinal, self))
392 if not isinstance(name_or_ordinal, int):
393 func.__name__ = name_or_ordinal
AttributeError: function 'cadam32bit_grad_fp32' not found"
}