alder

Simplified/Traditional support is now deployed. Really glad you like it, thanks!

I deployed a new release with a Simplified/Traditional switch at the top of the workspace. Thanks for the OpenCC link. Please let me know if anything's off.

There's a related feature already. The mic icon in the gap control plays the chunk, records you, and plays it back (the recording stays local). But if you mean a tool that automatically grades your pronunciation, I never thought about that. Interesting area to explore though.

Yes, you're right, public speaking is not really my thing. I'm a tech nerd. I should probably rely on partnerships with other people to help it grow. But the first thing I wanted to find out was if the app is actually useful to anyone other than myself. And the response from this community is encouraging. I also thought about letting content producers publish their content/courses as collections here with restricted access. The app already supports that. It is also possible to realign the transcript to their "gold" transcript. They would have instant Anki cards in multiple languages, and the shadowing practice for their subscribers. I would have users, win-win :) But again I have no idea how to do marketing.

Thanks so much for testing Japanese! I'll review the starter deck and what goes into it. At the moment it's a bit random, and the nonsensical LibriVox intro spills in. I'll also try to filter out the function words. I have no way to judge that myself, my Japanese is zero.

Glad the mobile interface works. There's an iPhone app too, basically the web app but with native audio support. It can play in the background. Though it's not ready for release yet.

The built-in flashcards use FSRS-5, and I believe it's O(1) so it should scale but I haven't actually done any performance testing yet, to be honest.

The infinite loop is a bug, I'll dig into that too. And I'm not sure if you noticed, but I also put up an FSI Japanese course here: https://lingochunk.com/c/fsi-japanese. Again, I can't judge the quality myself, so I'd really welcome any feedback.

And thank you for the kind review!

Thanks! As far as I understand your idea is to starts from the word and pulls examples from some huge data source. My approach is the other way round: I start from a source (the audio that you want to learn), and the tool extracts only the words that appear in it, with their meaning in that context. I think that hugely simplifies the implementation, and it is more useful for learners. They learn the meaning in a particular context.

As for the stack: STT with Soniox (word level timestamps), then spaCy for segmentation, POS and lemmas, then AI enrichment, correcting the lemma when spaCy is wrong. Some languages have no spaCy model at all and others are unreliable. I am trying to do spaCy thing in LLM then. Plus some extra magic for Japanese and Chinese.

Thanks, really useful extension link. Proper traditional support probably needs a context aware layer, not a plain lookup. I will experiment with additional LLM enrichment. Appreciate you digging into this!

Thanks, just to clarify "they" is actually only me :) I'm a contractor and run this through my own company. I try to collect as little as possible. And you're right about the UI fonts it's clearly something I need to fix ASAP. Appreciate the feedback!

No TTS at all in my app :) that was a deliberate choice, only STT. I experimented with many STT options, even self hosting Whisper, but ended up with Soniox. A bit expensive, but reliable. For the AI enrichment I went with Gemini Flash. I also tried Gemma 31B, which is really cheap and surprisingly good, on par with Gemini Flash, but extremely slow everywhere I tried. So you can make your own calculations :) And thanks for the congrats!

Thanks, I hope it will be helpful! If anything looks off, please let me know.

Yes, the transcriber API I use (Soniox) actually supports more than 60 languages. I just didn't have any automated testing for them. The way I tested was to find audio with a reliable reference transcription and put it through my pipeline. Then compare the results. Also some languages don't have reliable libraries to get part of speech and lemmas, something that flashcard needs.

For segmentation and POS I rely on spaCy zh_core_web_sm, pinyin from pypinyin library. Also the small correction level on top. But I am not a Chinese language expert to judge if it really works and I'll rely on feedback from the users to improve it.

Thanks! Yes, it's getting better for Greek but still not on par with other languages. I completed the only 2 Greek levels on Duolingo and they are really boring compared to the German one I am doing now. Easy Greek is a bit above my level, and the number of YouTubers in Greek is tiny compared to German.

Thanks! Chinese and Japanese as source languages are still experimental, I did my best to support them but I have to rely on people who actually know the language and this kind of feedback is really useful. I'll look into adding traditional characters and fixing the pinyin.

To be honest I haven't tested it on a 4K monitor yet, so I am not surprised. There are two controls above the transcript that change the font size and the line spacing, which should help a bit for now. Something to fix, thanks!

Thanks! I am glad you like it! I essentially mine the source audio, and all examples have cloze style gaps (blurring, in my case) that are revealed on the back of the card. I also beep the word in the sentence when you try to play it on the front card in built-in SRS system. Unfortunately that is not implemented in the Anki export, but it is technically possible.