-
我完成了tinybert模型的torch->onnx->mnn的转换和用MNNPythonOfflineQuant进行量化,但量化后模型大小仅减少了约1MB
该torch模型的word embedding占了90%参数量,推测embedding没有被量化,请问mnn int8量化是否支持embedding相关算子?
-
**Describe the bug**
It seems to be the case, when the question have more than 6 answers it can (and do) cause the "index out of bounds error" in some scenarios
**Error message**
----------------…
-
tinybert蒸馏的方法适用于macbert蒸馏吗
-
Hi, thanks for your work. I noticed that you did not use pre-trained weights for the student model (ImageNet-1K pre-trained, as I understand it). In Tab. 2, the performance after distillation is worse…
-
现在发现ERNIE 模型🉑️,但是RoBERTa和TinyBERT模型跑起来报错了,看起来是缺少类似ernie下面的`faster_tokenizer.py`来支持fastertokenizer功能
-
### 请提出你的问题
![微信图片_20230105134604](https://user-images.githubusercontent.com/107381937/210711000-99be6807-495c-4440-96de-86327151339a.png)
![微信图片_20230105134554](https://user-images.githubuserconten…
-
作者您好:
想请教您些问题,
本次试验环境:torch1.7.1 cuda 11 Geforce RTX3090(2块)
我在执行task_distillation.py的第一步:
Step 1: use task_distill.py to run the intermediate layer distillation.
# ${FT_BERT_BASE_DIR}$ contain…
-
**Question**
In the example code of the FARMRanker, the given example is:
```
ranker = FARMRanker(model_name_or_path="saved_models/roberta-base-asnq-binary")
```
However, it throws me an error th…
-
使用PaddleNLP-MIniLMv2 的训练脚本,进行模型压缩,但模型无法直接部署到paddleServing。怎么才能将model_state.pdparams的模型转化成paddleServing部署所需要的格式?
-
欢迎您反馈PaddleNLP使用问题,非常感谢您对PaddleNLP的贡献!
在留下您的问题时,辛苦您同步提供如下信息:
- 版本、环境信息
1)PaddleNLP和PaddlePaddle版本:请提供您的PaddleNLP和PaddlePaddle版本号,例如PaddleNLP 2.0.4,PaddlePaddle2.1.1
2)系统环境:请您描述系统类型,例如Linux/Windows…