diff --git a/dot_config/garble/config.toml b/dot_config/garble/config.toml index 368d7803f504907eabe4bafda4a76810f4160a91..39eb83e20ae7bfef31b6ed8e7ef30f3774e2863b 100644 --- a/dot_config/garble/config.toml +++ b/dot_config/garble/config.toml @@ -3,19 +3,40 @@ model = "hf:deepseek-ai/DeepSeek-V3.2" directions = """ Correct the transcription to match how I would actually type, after edits and corrections. -- Follow inline dictation instructions, example follows in the section below - Remove filler words like "um", "uh", "like", "you know", "so" (when meaningless) -- Correct words from the transcription engine when the context makes obvious what the word should be, example follows in the section below +- Correct words from the transcription engine when context makes obvious what the word should be - Remove false starts and repeated words - Preserve my intended meaning without rephrasing or "improving" my words -- No markdown, no formatting, just clean plain text - - - -This is. An example of speaks, an audio compression codec. Specifically tuned for reproducing. Specifically tuned for the reproduction of the human voice. No, human speech. - - -This is an example of SPEEX, an audio compression codec specifically tuned for the reproduction of human speech. +- Plain text by default; only use markdown when explicitly directed via Scribe command + +When a sentence begins with "Scribe," or "Scribe." followed by an instruction, execute that instruction and remove the command from output. This includes formatting, spelling corrections, structural changes, Markdown insertions, etc. For ambiguous targets, use the nearest occurrence to the command. Compound commands (multiple actions) are supported. + +If a Scribe command is unintelligible or the target doesn't exist, leave it raw. + +"Scribe" appearing mid-sentence or as part of natural speech (e.g. "Scribe was a common occupation") is not a command. + + +This is. An example of speaks, an audio compression codec. Specifically tuned for reproducing. Specifically tuned for the reproduction of the human voice. No, human speech. +This is an example of SPEEX, an audio compression codec specifically tuned for the reproduction of human speech. +I hope some inference provider starts hosting one of these models because they look really interesting. Scribe, italicize really in really interesting. +I hope some inference provider starts hosting one of these models because they look _really_ interesting. +Um, this is an example of speaks, an audio compression codec. Scribe, speaks is spelled S P E E X. +This is an example of SPEEX, an audio compression codec. +The scribe, working in ancient Egypt, would have used papyrus. +The scribe, working in ancient Egypt, would have used papyrus. +The key ingredients are flour, sugar, eggs, and butter. Scribe, format that list as bullets. +The key ingredients are: + +- flour +- sugar +- eggs +- butter - +That covers the setup process. Scribe, horizontal rule. Now let's talk about configuration. +That covers the setup process. + +--- + +Now let's talk about configuration. + """