cuongngm commented 3 years ago

Tks for your repo, the code is very clean. I trained your model with my dataset, the loss is converged good. But when predicted, a few the value of item['text'] was different format with the value of file .tsv. Example: "ÁO MOA TT CÀI CÚC" vs "ÁO MƯA TT CÀI CÚC",..vv Can you help me explain?

cuongngm commented 3 years ago

i solved problem by created my vocab in my languague and retrained model

babyhockey commented 3 years ago

Hi Could you explain how you created the vocab in your language? Did you list all possible words (or characters) in keys.txt? Looking forward to your reply. Thanks!

cuongngm commented 3 years ago

that's right

babyhockey commented 3 years ago

Thank you!

On Mon, May 24, 2021, 10:09 PM cuongng.m @.***> wrote:

that's right

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/wenwenyu/PICK-pytorch/issues/68#issuecomment-847476434, or unsubscribe https://github.com/notifications/unsubscribe-auth/AQORHKZL42NJRBPEIZOZ5FLTPMBELANCNFSM4VG2RBCQ .

khangnguyenhuu commented 2 years ago

hi there, can u give me an example in vietnamese vocab :( thanks a lot

cuongngm commented 2 years ago

give me your email and i will send for you example of Vietnamese vocab

khangnguyenhuu commented 2 years ago

give me your email and i will send for you example of Vietnamese vocab nskhang1@gmail.com, thanks a lot

htdung167 commented 1 year ago

@cuongngm I'm having the same problem. Can you give me Vietnamese vocab? I am using vocab:

ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz| 0123456789ĂÂÊÔƠƯÁẮẤÉẾÍÓỐỚÚỨÝÀẰẦÈỀÌÒỒỜÙỪỲẢẲẨĐẺỂỈỎỔỞỦỬỶÃẴẪẼỄĨÕỖỠŨỮỸẠẶẬẸỆỊỌỘỢỤỰỴăâêôơưáắấéếíóốớúứýàằầèềìòồờùừỳảẳẩđẻểỉỏổởủửỷãẵẫẽễĩõỗỡũữỹạặậẹệịọộợụựỵ\'*:,@.-(#%")/~!^&_+={}[]\;<>?※”$€£¥₫°²™ā–

aAàÀảẢãÃáÁạẠăĂằẰẳẲẵẴắẮặẶâÂầẦẩẨẫẪấẤậẬbBcCdDđĐeEèÈẻẺẽẼéÉẹẸêÊềỀểỂễỄếẾệỆfFgGhHiIìÌỉỈĩĨíÍịỊjJkKlLmMnNoOòÒỏỎõÕóÓọỌôÔồỒổỔỗỖốỐộỘơƠờỜởỞỡỠớỚợỢpPqQrRsStTuUùÙủỦũŨúÚụỤưƯừỪửỬữỮứỨựỰvVwWxXyYỳỲỷỶỹỸýÝỵỴzZ0123456789!"#$%&''()*+,-./:;<=>?@[]^_`{|}~

wenwenyu / PICK-pytorch

result when run test.py #68