👑 Easy-to-use and powerful NLP and LLM library with 🤗 Awesome model zoo, supporting wide-range of NLP tasks from research to industrial applications, including 🗂Text Classification, 🔍 Neural Search, ❓ Question Answering, ℹ️ Information Extraction, 📄 Document Intelligence, 💌 Sentiment Analysis etc.
使用paddleslim 量化unimo时报错:Operator (fusion_unified_decoding) is not registered.
转为静态图后好像不支持’fusion_unified_decoding‘算子,有什么办法可以支持该算子(例如如何register)?
Preparation stage, Run batch:| | 0/10
Traceback (most recent call last):
File "ptq.py", line 28, in
batch_nums=10)
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddleslim/quant/quanter.py", line 627, in quant_post_static
post_training_quantization.quantize()
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/contrib/slim/quantization/post_training_quantization.py", line 401, in quantize
scope=self._scope)
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/executor.py", line 1463, in run
six.reraise(*sys.exc_info())
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/six.py", line 719, in reraise
raise value
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/executor.py", line 1459, in run
return_merged=return_merged)
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/executor.py", line 1641, in _run_impl
self.place, scope)
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/executor.py", line 745, in get_program_and_executor
fetch_var_name, place, scope))
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/executor.py", line 808, in _get_program_and_executor
skip_var_names)
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/executor.py", line 411, in _apply_inplace_addto_pass
attr_types)
File "/home/cciip/Application/miniconda3/envs/aigc/lib/python3.7/site-packages/paddle/fluid/framework.py", line 4359, in _apply_pass
attrs = core.apply_pass(
RuntimeError: (NotFound) Operator (fusion_unified_decoding) is not registered.
[Hint: op_info_ptr should not be null.] (at /paddle/paddle/fluid/framework/op_info.h:156)
请提出你的问题
使用paddleslim 量化unimo时报错:Operator (fusion_unified_decoding) is not registered. 转为静态图后好像不支持’fusion_unified_decoding‘算子,有什么办法可以支持该算子(例如如何register)?