lonePatient albert_pytorch issues

lonePatient / albert_pytorch

A Lite Bert For Self-Supervised Learning Language Representations

Apache License 2.0

709 stars 152 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

tf模型转换pt失败

#71 wqw547243068 closed 11 months ago
2
请教一下，您这边给出来的pytorch的albert预训练模型是谷歌经过大量语料训练出来的吗

#70 hfy-only opened 1 year ago
2
请问optimizer.bin 这个文件要去哪里下载？谢谢

#69 lsx0930 opened 2 years ago
0
run_pretraining.py可能缺少调用main函数的地方

#68 Freeshman opened 2 years ago
0
BertLayer第二个子层为何有两个nn.Linear

#67 aibeiyun opened 2 years ago
0
训练时显存占用问题

#66 Qiu-dot opened 3 years ago
0
create_masked_lm_predictions（）

#65 flyerjia closed 3 years ago
0
请问可以使用pytorch的版本重新进行预训练吗？如果可以，应该怎么做？

#64 yysirs closed 3 years ago
0
tf模型转pytorch模型报错？

#63 ghost closed 3 years ago
1
readme中fine-tune结果的超参数

#62 Sisi0518 opened 3 years ago
0
No such file or directory

#61 ShaoaAllen opened 3 years ago
0
基于albert模型继续在自己的数据集上pretraining albert的问题

#60 liuxiaohao-xn closed 3 years ago
2
转换模型时出错

#59 llxbush opened 4 years ago
0
AlbertForNextSentencePredicition

#58 DeepakDhanasekar opened 4 years ago
2
如何加入重计算机制（torch.utils.checkpoint)

#57 ReyonRen opened 4 years ago
0
TF_WEIGHTS_NAME = 'model.ckpt'里面的model.ckpt文件在哪里？

#56 yexingxin opened 4 years ago
0
What is the relationship between learning rate and BERT model size (especially the depth)

#55 wsn555 opened 4 years ago
0
将brightmart中预训练的TensorFlow模型转换为Pytorch模型报如下错误，'Embedding' object has no attribute 'shape'

#54 wyqnumber opened 4 years ago
0
ALBERT-small训练效果问题

#53 Hanlard opened 4 years ago
0
关于继续在Classification问题上finetune的问题

#52 zipzou closed 4 years ago
0
生成GLUE提交版本文件

#51 Lillian9707 closed 4 years ago
0
global_step的位置导致多次执行无谓的evaluate()

#50 illusions-LYY opened 4 years ago
0
not initialized from pretrained model: ['bert.albert.embeddings.word_embeddings.weight', 'bert.albert.embeddings.position_embeddings.weight', 'bert.albert.embeddings.token_type_embeddings.weight',

#49 yc0815024 closed 4 years ago
4
No Factorization？

#48 SCNUJackyChen closed 4 years ago
1
How can I do MLM task on new dataset with new vocab.txt utilizing AlbertForPreTraining with pretrained weights?

#47 ChineseYjh opened 4 years ago
0
加载Brightmart模型时，AlbertForMaskedLM中的cls.predictions.project_layer参数无法正确加载

#46 hong-feng opened 4 years ago
5
“pooled_output = self.dropout(pooled_output+0.1)” is confusing

#45 matricer closed 4 years ago
1
关于继续pretrain的问题

#44 SCNUJackyChen closed 4 years ago
5
How can I do when there is only 1 sentence in a document?

#43 SCNUJackyChen closed 4 years ago
2
sentence-order prediction

#42 qiunlp opened 4 years ago
4
AlbertForNextSentencePrediction

#41 qiunlp closed 4 years ago
3
delete unnecessary assertion

#40 SCNUJackyChen closed 4 years ago
0
How can i use albert to finish CLOTH test? I need some suggestion. Thank you.

#39 jiangjiaqi6 opened 4 years ago
1
Couldn't find 'checkpoint' file or checkpoints in given directory

#38 jiangjiaqi6 closed 4 years ago
7
tensorflow到pytorch版本模型转换错误

#37 Zessay closed 4 years ago
4
咨询albert在huggingface/transformers下的使用

#36 renjunxiang closed 4 years ago
6
与当前huggingface/transformers的代码不兼容

#35 ghost closed 4 years ago
3
AttributeError: 'AlbertForSequenceClassification' object has no attribute 'keys'

#34 chiragsanghvi10 opened 4 years ago
0
[Question] 为什么在中文版本使用bert标准分词，英文版本使用sentencepiece分词啊？

#33 sakuranew closed 4 years ago
3
albert_tiny_google_zh convert to pytorch

#32 miangangzhen closed 4 years ago
2
能不能用albert_chinese_pytorch代码做英文语料的预训练

#31 shenlanyilang closed 4 years ago
2
请问能否支持一下Lamb optimizer

#30 lxl910915 closed 4 years ago
2
跑seq=1024，batch_szie=2，1万步左右loss会变成nan，是不是因为batch_size太小了

#29 lxl910915 closed 4 years ago
0
训练过程中albert占用的显存很大

#28 fatmelon opened 4 years ago
3
CUDA error: no kernel image is available for execution on the device

#27 chiragsanghvi10 closed 4 years ago
1
这个断言可以修改一下，前面内循环判断。 if len(masked_token_labels) >= num_to_mask:

#26 moonblue333 opened 4 years ago
2
RuntimeError: Error(s) in loading state_dict for BertModel

#25 ZiteHe opened 4 years ago
1
使用albert.base（英文）finetuning的时候，--gradient_accumulation_steps设置为大于1时直接进入evaluating而不training

#24 YuxiangLu opened 4 years ago
1
请教下share_type和embedding_size的选择上有什么建议吗

#23 BaeSeulki closed 4 years ago
1
请问如何使用微调好的模型进行预测？

#22 zhu1090093659 opened 4 years ago
1