ZeroDivisionError: division by zero

[ei-grad]

Python 3.10.11 (main, Apr 20 2023, 19:02:41) [GCC 11.2.0]
Type 'copyright', 'credits' or 'license' for more information
IPython 8.17.2 -- An enhanced Interactive Python. Type '?' for help.

In [1]: from transformers import AutoTokenizer, TextStreamer
   ...: from intel_extension_for_transformers.transformers import AutoModelForCausalLM
   ...: 
   ...: model_name = "Intel/neural-chat-7b-v1-1"     # Hugging Face model_id or local 
   ...: model
   ...: prompt = "Once upon a time, there existed a little girl,"
   ...: 
   ...: tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
   ...: inputs = tokenizer(prompt, return_tensors="pt").input_ids
   ...: streamer = TextStreamer(tokenizer)
   ...: 
   ...: model = AutoModelForCausalLM.from_pretrained(model_name, load_in_4bit=True)
   ...: outputs = model.generate(inputs, streamer=streamer, max_new_tokens=300)
None of PyTorch, TensorFlow >= 2.0, or Flax have been found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.

---------------------------------------------------------------------------
ZeroDivisionError                         Traceback (most recent call last)
Cell In[1], line 2
      1 from transformers import AutoTokenizer, TextStreamer
----> 2 from intel_extension_for_transformers.transformers import AutoModelForCausalLM
      3 model_name = "Intel/neural-chat-7b-v1-1"     # Hugging Face model_id or local model
      4 prompt = "Once upon a time, there existed a little girl,"

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/intel_extension_for_transformers/transformers/__init__.py:19
      1 #!/usr/bin/env python
      2 # -*- coding: utf-8 -*-
      3 #
   (...)
     15 # See the License for the specific language governing permissions and
     16 # limitations under the License.
---> 19 from .config import (
     20     WEIGHTS_NAME,
     21     AutoDistillationConfig,
     22     BenchmarkConfig,
     23     DistillationConfig,
     24     DynamicLengthConfig,
     25     FlashDistillationConfig,
     26     NASConfig,
     27     Provider,
     28     PrunerV2,
     29     PruningConfig,
     30     QuantizationConfig,
     31     TFDistillationConfig,
     32 )
     33 from .distillation import (
     34     SUPPORTED_DISTILLATION_CRITERION_MODE,
     35     DistillationCriterionMode,
     36 )
     37 from .mixture.auto_distillation import AutoDistillation

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/intel_extension_for_transformers/transformers/config.py:22
     20 import yaml
     21 from enum import Enum
---> 22 from neural_compressor.conf.config import (
     23     Distillation_Conf, Pruner, Pruning_Conf, Quantization_Conf
     24 )
     25 from neural_compressor.conf.dotdict import DotDict, deep_set
     26 from .utils.metrics import Metric

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/__init__.py:28
     20 # we need to set a global 'NA' backend, or Model can't be used
     21 from .config import (
     22     DistillationConfig,
     23     PostTrainingQuantConfig,
   (...)
     26     MixedPrecisionConfig,
     27 )
---> 28 from .contrib import *
     29 from .model import *
     30 from .metric import *

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/contrib/__init__.py:18
      1 #!/usr/bin/env python
      2 # -*- coding: utf-8 -*-
      3 #
   (...)
     15 # See the License for the specific language governing permissions and
     16 # limitations under the License.
     17 """Built-in strategy for multiple framework backends."""
---> 18 from .strategy import *

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/contrib/strategy/__init__.py:25
     23 for f in modules:
     24     if isfile(f) and not f.startswith("__") and not f.endswith("__init__.py"):
---> 25         __import__(basename(f)[:-3], globals(), locals(), level=1)

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/contrib/strategy/tpe.py:26
     22 from pathlib import Path
     24 import numpy as np
---> 26 from neural_compressor.strategy.strategy import TuneStrategy, strategy_registry
     27 from neural_compressor.strategy.utils.tuning_sampler import OpWiseTuningSampler
     28 from neural_compressor.strategy.utils.tuning_structs import OpTuningConfig

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/strategy/__init__.py:19
      1 #!/usr/bin/env python
      2 # -*- coding: utf-8 -*-
      3 #
   (...)
     15 # See the License for the specific language governing permissions and
     16 # limitations under the License.
     17 """Intel Neural Compressor Strategy."""
---> 19 from .strategy import STRATEGIES
     20 from os.path import dirname, basename, isfile, join
     21 import glob

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/strategy/strategy.py:35
     32 import numpy as np
     33 import yaml
---> 35 from neural_compressor.adaptor.tensorflow import TensorFlowAdaptor
     37 from ..adaptor import FRAMEWORKS
     38 from ..algorithm import ALGORITHMS, AlgorithmScheduler

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/adaptor/__init__.py:26
     24 for f in modules:
     25     if isfile(f) and not f.startswith("__") and not f.endswith("__init__.py"):
---> 26         __import__(basename(f)[:-3], globals(), locals(), level=1)
     28 __all__ = ["FRAMEWORKS"]

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/adaptor/pytorch.py:43
     40 torch_utils = LazyImport("neural_compressor.adaptor.torch_utils")
     41 ipex = LazyImport("intel_extension_for_pytorch")
---> 43 REDUCE_RANGE = False if CpuInfo().vnni else True
     44 logger.debug("Reduce range is {}".format(str(REDUCE_RANGE)))
     47 def get_torch_version():

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/utils/utility.py:129, in singleton.<locals>._singleton(*args, **kw)
    127 """Create a singleton object."""
    128 if cls not in instances:
--> 129     instances[cls] = cls(*args, **kw)
    130 return instances[cls]

File ~/.virtualenvs/intel-transformers/lib/python3.10/site-packages/neural_compressor/utils/utility.py:254, in CpuInfo.__init__(self)
    252 self._sockets = self.get_number_of_sockets()
    253 self._cores = psutil.cpu_count(logical=False)
--> 254 self._cores_per_socket = int(self._cores / self._sockets)

ZeroDivisionError: division by zero

[Zhenzhong1]

Hi, the error you met, I thought it was probably related to the neural compressor modification recently.
https://github.com/intel/neural-compressor/blob/master/neural_compressor/utils/utility.py#L255C34-L255C55

Please install the ITREX from the source code~
I have run the whole installation process. It's ok if you follow the commands.

git clone https://github.com/intel/intel-extension-for-transformers
cd intel-extension-for-transformers
pip install -r requirements.txt
pip install transformers==4.33.1
python setup.py install

Please put your script outside of the intel-extension-for-transformers root directory to make sure the script calls API from the conda env you installed rather than local ITREX directories.

from transformers import AutoTokenizer, TextStreamer
from intel_extension_for_transformers.transformers import AutoModelForCausalLM
model_name = "Intel/neural-chat-7b-v1-1"     # Hugging Face model_id or local model
prompt = "Once upon a time, there existed a little girl,"

tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
inputs = tokenizer(prompt, return_tensors="pt").input_ids
streamer = TextStreamer(tokenizer)

model = AutoModelForCausalLM.from_pretrained(model_name, load_in_4bit=True)
outputs = model.generate(inputs, streamer=streamer, max_new_tokens=300)

Inference screenshot:

[Spycsh]

Hi @ei-grad , could you please give us further information on which chip or which platform are you using and paste the info here? Specifically, could you please run "lscpu" on your platform if it is Linux or 'wmic cpu get DeviceID | C:\Windows\System32\find.exe /C "CPU"' if your platform is windows?

[KangkangStu]

回溯（最近一次调用最后一次）：
文件“/home/kangyuhan/finetuning/intel-transformers/intel-extension-for-transformers/demo.py”，第 2 行，来自
intel_extension_for_transformers.transformers import AutoModelForCausalLM
文件“/home/kangyuhan /finetuning/intel-transformers/intel-extension-for-transformers/intel_extension_for_transformers/transformers/ init .py"，第 19 行，来自
.config import (
文件 "/home/kangyuhan/finetuning/intel-transformers/intel-extension- for-transformers/intel_extension_for_transformers/transformers/config.py"，第 22 行，来自
neural_compressor.conf.config import (
File "/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor / init .py”，第 28 行，来自
.contrib import *
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/contrib/ init .py”，第 18 行，在
from .strategy import *
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/contrib/strategy/ init .py”，第 25 行，在
import (basename( f)[:-3], globals(), locals(), level=1)
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/contrib/strategy/ sigopt.py”，第 21 行，来自
neural_compressor.strategy.strategy导入TuneStrategy，strategy_registry
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/strategy/ init .py ”，第 19 行，来自
.strategy import STRATEGIES
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/strategy/strategy.py”，第 35 行，来自
neural_compressor .adaptor.tensorflow import TensorFlowAdaptor
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/adaptor/ init .py”，第 26 行，导入
( basename(f)[ :-3], globals(), locals(), level=1)
文件 "/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/adaptor/pytorch.py​​",第 43 行，在
REDUCE_RANGE = False if CpuInfo().vnni else True
File "/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/utils/utility.py"，第 129 行，在 _singleton
实例中[cls] = cls(*args, **kw)
文件“/home/kangyuhan/.conda/envs/llama_factory/lib/python3.10/site-packages/neural_compressor/utils/utility.py”，第 257 行，在init
self._cores_per_socket = int(self._cores / self._sockets)
ZeroDivisionError：除以零

我遇到了同样的问题。我的CPU是Intel Xeon Platinum 8358 操作系统为Linux
lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 128
On-line CPU(s) list: 0-127
Thread(s) per core: 2
Core(s) per socket: 32
座： 2
NUMA 节点： 2
厂商 ID： GenuineIntel
CPU 系列： 6
型号： 106
型号名称： Intel(R) Xeon(R) Platinum 8358 CPU @ 2.60GHz
步进： 6
CPU MHz： 800.000
CPU max MHz: 3400.0000
CPU min MHz: 800.0000
BogoMIPS： 5200.00
虚拟化： VT-x
L1d 缓存： 48K
L1i 缓存： 32K
L2 缓存： 1280K
L3 缓存： 49152K
NUMA 节点0 CPU： 0-31,64-95
NUMA 节点1 CPU： 32-63,96-127

[Spycsh]

@KangkangStu Thanks for detailed information! I believe this is because your system is non-english. I will let you know after fixing this problem.

[Spycsh]

@ei-grad , @KangkangStu , this should work after intel/neural-compressor#1514 is merged. After the merge, please re-install neural compressor from source (https://github.com/intel/neural-compressor/blob/master/docs/source/installation_guide.md#install-from-source) and check. Let me know if you meet any further problems!