vebaev Wrote:Upload it to MEGA which is highspeed most of the time.

p.s
what are updates of 2.5 from 2.0?

It is already uploaded to MEGA. The differences between 2.0 and 2.5 are minor. Some minor errors with the voice were fixed up.

I uploaded v2.5 to mediafire:Link
Let me know if the file has to be uploaded anywhere else, I would be happy to help.

Also wanted to give my thanks for all your hard work. These decks look awesome and am looking forward to the new core900 in the future.

Rivvie Wrote:I uploaded v2.5 to mediafire:Link
Let me know if the file has to be uploaded anywhere else, I would be happy to help.

Also wanted to give my thanks for all your hard work. These decks look awesome and am looking forward to the new core900 in the future.

You are a gentleman(woman?) and a scholar. Thank you. I tried 4 different computers over 3 different networks and kept getting errors from Mega.

Rivvie Wrote:I uploaded v2.5 to mediafire:Link
Let me know if the file has to be uploaded anywhere else, I would be happy to help.

Also wanted to give my thanks for all your hard work. These decks look awesome and am looking forward to the new core900 in the future.

Thanks for that, I have made a link on the main post.

@Rivvie: You are a godsend. Thank you for the link!

@Nayr: And you sir, are a saint. I'm looking forward to the new deck. I'm curious about something you mentioned though.

Quote:All cards will come already pre-arranged in n+1 using morphman, so no need to stuff around with complicated settings.

I was under the impression that morphman could only rearranged cards based on the user's knowledge. I guess this is not the case.

How exactly would I go about doing this with your 2.5 deck? I saw you post this earlier:

Quote:All I did to get it to work was change this part of the config.py:


# Model overrides can only override the entries marked above. 2nd priority
model_overrides = {
'subs2srs': { 'enabled':True },
'JtMW': { 'enabled':True },
'Nayrs Japanese Core5000': { 'enabled':True },
}


Not sure if there is a better way or not.

After I installed morphman, I followed this same advice, only changing that part of the config.py. Now whenever I hit morphman recalc, it actually goes through a process and rearranges the deck. But I'm unsure as to whether it rearranged it to n+1. Is that really all it takes?

I spent the better part of a day yesterday trying in vain wrap my head around morphman. I came to the conclusion that morphman requires some sort of knowledge about what you know in order to optimize decks. Something about creating a morpheme database. I'm sure I'm wrong though.

aurelius Wrote:@Rivvie: You are a godsend. Thank you for the link!

@Nayr: And you sir, are a saint. I'm looking forward to the new deck. I'm curious about something you mentioned though.

Quote:All cards will come already pre-arranged in n+1 using morphman, so no need to stuff around with complicated settings.

I was under the impression that morphman could only rearranged cards based on the user's knowledge. I guess this is not the case.

My guess, is that he set up a blank profile, or reset the morphman database, and made sure that all of the cards in 3.0 are unstudied. Doing this and running morphman would act as if you had no knowledge; that is i=0.

aurelius Wrote:@Rivvie: You are a godsend. Thank you for the link!

@Nayr: And you sir, are a saint. I'm looking forward to the new deck. I'm curious about something you mentioned though.

Quote:All cards will come already pre-arranged in n+1 using morphman, so no need to stuff around with complicated settings.

I was under the impression that morphman could only rearranged cards based on the user's knowledge. I guess this is not the case.

How exactly would I go about doing this with your 2.5 deck? I saw you post this earlier:

Quote:All I did to get it to work was change this part of the config.py:


# Model overrides can only override the entries marked above. 2nd priority
model_overrides = {
'subs2srs': { 'enabled':True },
'JtMW': { 'enabled':True },
'Nayrs Japanese Core5000': { 'enabled':True },
}


Not sure if there is a better way or not.

After I installed morphman, I followed this same advice, only changing that part of the config.py. Now whenever I hit morphman recalc, it actually goes through a process and rearranges the deck. But I'm unsure as to whether it rearranged it to n+1. Is that really all it takes?

I spent the better part of a day yesterday trying in vain wrap my head around morphman. I came to the conclusion that morphman requires some sort of knowledge about what you know in order to optimize decks. Something about creating a morpheme database. I'm sure I'm wrong though.

It will create the database based off the cards you already know, have already seen or your already mature cards. Lets say for example you have already studied 500 or so cards from Core2000 or a Genki deck or something, and now you want to study the Core5000 deck.

If you just change the note type of the other cards you studied to 'Nayrs Japanese Core5000' it will take those cards into account when you do the calculation.

Alternatively, you can manually tell morphman what cards you already know by simply adding the tag 'alreadyKnown' to a card.

Nayr182 Wrote:All I did to get it to work was change this part of the config.py:


# Model overrides can only override the entries marked above. 2nd priority
model_overrides = {
'subs2srs': { 'enabled':True },
'JtMW': { 'enabled':True },
'Nayrs Japanese Core5000': { 'enabled':True },
}


Not sure if there is a better way or not.

Thank you very much for your help, that's exactly what I needed.

And of course thanks a bunch for creating this awesome deck.

aurelius Wrote:I was under the impression that morphman could only rearranged cards based on the user's knowledge. I guess this is not the case.

…snip…

After I installed morphman, I followed this same advice, only changing that part of the config.py. Now whenever I hit morphman recalc, it actually goes through a process and rearranges the deck. But I'm unsure as to whether it rearranged it to n+1. Is that really all it takes?

I spent the better part of a day yesterday trying in vain wrap my head around morphman. I came to the conclusion that morphman requires some sort of knowledge about what you know in order to optimize decks. Something about creating a morpheme database. I'm sure I'm wrong though.

RandomQuotes Wrote:My guess, is that he set up a blank profile, or reset the morphman database, and made sure that all of the cards in 3.0 are unstudied. Doing this and running morphman would act as if you had no knowledge; that is i=0.

Elaborating on what the man Nayr182 already said. RandomQuotes is right. To run morph man without considering the set of words you already know, set up a new user account, add the decks in question, modify the source code (hah!) if necessary, mark one sentence as "known" via a tag I thick, and run it.

About your other question: it rearranges the due dates on the deck! So if you browse the deck, you'll see the same "sort order" as before, but if you sort by due date you'll see the optimized order. That really confused me too.

Hope this helps…?

Hello. Thanks for uploading the data of the Frequency Dictionary of Japanese.

I would like to use the data of that dictionary with GoldenDict. Therefore, is it possible to convert the .apkg file to plain text?

After obtaining a plain text file, it would be possible to create a version in a format readable by GoldenDict (e.g. dsl, bgl, mdx, etc).

Looking forward to your response,

tovaremeterio Wrote:I would like to use the data of that dictionary with GoldenDict. Therefore, is it possible to convert the .apkg file to plain text?

When using Ankis export function you can set the export format to plain text format, which will give you a tab separated text file.

Hi Nayr,

I want to start using the deck but if the version 3.0 will be out soon in 1-2 months maybe it is better to wait as for now I have things to occupy me. Can you mention when you plan to release it and what will be the new features/differences

Many thanks for the great work that is helping us all!!!

[RawToast]

Another thing to look out for is tags on your notes. Morphman will tag any i+0 cards with "comprehension", i+1 cards with "vocab", and i+2 or greater as "not ready".

vebaev Wrote:Hi Nayr,

I want to start using the deck but if the version 3.0 will be out soon in 1-2 months maybe it is better to wait as for now I have things to occupy me. Can you mention when you plan to release it and what will be the new features/differences

Many thanks for the great work that is helping us all!!!

I would suggest you just crack into the version 2.5 as I am not working on 3.0 anymore.


What I am working on at the moment (with the help of my colleague) is something completely original and not sourced from textbooks etc.

The small project which was going to be Core5000 3.0 has evolved into a much much bigger and ambitious co-collaboration. Not really going to reveal much more than that. I will release more information about it when I am closer to a finished product.

Wow, this just gets more and more exciting

Thanks for this fenomenal work. I am working very hard in this deck. Thanks again.

Nayr182 Wrote:Nayr's Core5000 comprises of the 5,000 most commonly used words in the Japanese language.

**this is completely original and in no way related to Core2k, 6k or 10k**

All sentences have been sourced from "A Frequency Dictionary of Japanese (Routledge Frequency Dictionaries) – March 9, 2013" which is based on a 100 million word corpus, composed of spoken, fiction, non-fiction and news texts in current use.

All cards have native Japanese voice.
All cards are numbered in order of word frequency.
All cards have been checked by native Japanese for furigana errors.

Nayr's Core5000 [Version 2.5] can be found here: https://mega.co.nz/#!M0RSmTBb!Z1OiBONGOs...zS1q784Vt0

Alternatively try here:
http://www.mediafire.com/download/7e3450..._v2.5.apkg

Or here:
https://www.dropbox.com/s/srgy6alqsqb52d....apkg?dl=0

Please feel free to make any suggestions.

**Special thanks to aldebrn for helping me get these decks up and going so quickly!**

Just wondering (because I don't know)...do you need permission from Routledge to produce this? The actual book published by Routledge "Frequency Dictionary of Japanese" is fairly expensive to purchase. Someone might say that using these "decks" is a backdoor way of avoiding shelling out the money to buy the book.

john555 Wrote:Just wondering (because I don't know)...do you need permission from Routledge to produce this? The actual book published by Routledge "Frequency Dictionary of Japanese" is fairly expensive to purchase. Someone might say that using these "decks" is a backdoor way of avoiding shelling out the money to buy the book.

Hmm.... And someone might say that making a words deck and sharing it is a backdoor way for avoiding shelling money for a dictionary. I think that's ridiculous concept.

Robik Wrote:Hmm.... And someone might say that making a words deck and sharing it is a backdoor way for avoiding shelling money for a dictionary. I think that's ridiculous concept.

That (your false analogy) would be a ridiculous concept. But original content, like the sentences in that book, being protected by copyright, isn't.

Hopefully the publisher is smart enough to realize this isn't costing them book sales (and in fact is probably helping sell a few extra), but if they're not, they do have the right to have it pulled from the Anki page.

Stansfield123 Wrote:

Robik Wrote:Hmm.... And someone might say that making a words deck and sharing it is a backdoor way for avoiding shelling money for a dictionary. I think that's ridiculous concept.

That (your false analogy) would be a ridiculous concept. But original content, like the sentences in that book, being protected by copyright, isn't.

Hopefully the publisher is smart enough to realize this isn't costing them book sales (and in fact is probably helping sell a few extra), but if they're not, they do have the right to have it pulled from the Anki page.

Ah, i thought only the frequency of words was taken out of it, not entire sentences...

[RawToast]

The sentences are slightly different to those in the book: they have been 'kanjified', mistakes fixed, etc. There's a bit more to the book than just the sentences.

If the book came with audio and it had been ripped to the deck, this would definitely be up for removal. It might be worth removing the frequency data from the next version of the deck -- I could see that being up for debate.

RawToast Wrote:The sentences are slightly different to those in the book: they have been 'kanjified', mistakes fixed, etc. There's a bit more to the book than just the sentences.

If the book came with audio and it had been ripped to the deck, this would definitely be up for removal. It might be worth removing the frequency data from the next version of the deck -- I could see that being up for debate.

Well its no longer uploaded to anki now anyhow so it's not really an issue. As I am working on something else, I haven't got the time to further edit core5000 to eliminate any possible copyright infringements. If push comes to shove I'll just take it offline.

The new project isn't going to have any of these problems as it will be an original work, and this is my main focus at the moment.

[RawToast]

Nayr182 Wrote:The new project isn't going to have any of these problems as it will be an original work, and this is my main focus at the moment.

I am interested I love your decks, decks with good native audio are hard to find -- especially if you're looking for more challenging sentences. One of my favourite decks was the "Listen and Learn" deck, which despite being an original creation no-longer exists on Anki web.

I guess this will not be the Grammar Dictionaries? Unless you've modified the sentences.

I will love to see mixed vocabulary and grammar project with native audio

RawToast Wrote:

Nayr182 Wrote:The new project isn't going to have any of these problems as it will be an original work, and this is my main focus at the moment.

I am interested I love your decks, decks with good native audio are hard to find -- especially if you're looking for more challenging sentences. One of my favourite decks was the "Listen and Learn" deck, which despite being an original creation no-longer exists on Anki web.

I guess this will not be the Grammar Dictionaries? Unless you've modified the sentences.

The lesson I take away from reading all these posts is that, if you like an Anki deck, download it to your own computer right away so you've got it forever.