Alex Cannan

Receivers broken?

2

The app has worked beautifully for me over the past couple years. But suddenly, starting May 29th, the app no longer collects scrobbles into the cache from any receiver I...

Building diffs for patching lm binaries?

1

I'm curious if it would be possible to build a storage-efficient lm.diff file to patch an older lm.binary file into a newer one. I've experimented with some existing binary diff...

Repeated scrobbles if jellyfin is left idle without stopping playback

8

I've noticed an issue with duplicated scrobbles when my jellyfin client is left idle without explicitly stopping the playback, just pausing. Repeated scrobbles seem to happen every 1 to 2...

Syncing favorites even though the option is disabled

1

Hello, I have noticed recently that even though I have the "Sync favorites for this user?" option NOT selected, my favorited tracks are being (very quickly) synced to lastfm. Jellyfin...

feat(python-openai): Set or reveal generationId during chat completion

2

### Describe the bug The SDK supports passing in langfuse data alongside openai (and other integrations I assume): https://langfuse.com/docs/integrations/openai/python/get-started#custom-trace-properties I have a specific use case in my app where multiple...

I've noticed some cases of the `large-v3-turbo` model outputting text like `Richard Jr.'s favorite toy`. In my testing this has caused a mismatch between the sentence spans (punkt considers `Jr.`...

Punkt Sentence Tokenizer incorrectly splits on multi-period abbreviations

1

When using the Punkt sentence tokenizer, sentences containing abbreviations with multiple periods (e.g., "M.D.") are improperly split. For example, I've prepared a snippet: ``` from nltk.tokenize.punkt import PunktSentenceTokenizer, PunktParameters #...