hephaes7us

7 ms·

61.
by hephaes7us 10mo ago
Intuitively I feel like it's something like the light bulb. For a while, bulbs had to meet efficiency standards. These standards were configured such that they didn't technically exclude incandescent bulbs, however, for an incand
62.
by hephaes7us 10mo ago
I like this kind of thinking. I wonder though if you're paying as much attention to see who's "struggling with output" on Tues, Thurs, and Fri? As in, are you skeptical of WFH and looking for downsides, or does your par
63.
by hephaes7us 1y ago
Tokens are expensive, sure, but I don't even _want_ Ollama to run inference for me. Ollama gives me, essentially, a wrapper for llama.cpp and convenient hosting where I can download models. I'm happy to pay for the bandwidth, plus
64.
by hephaes7us 1y ago
In this case, it's not about whether it fits on my physical hardware or not. It's about what seems like an arbitrary restriction designed to start pushing users to their cloud offering.
65.
by hephaes7us 1y ago
Have you tried Whisper itself? It's open-weights. One of the features of the project posted above is "transformations" that you can run on transcripts. They feed the text into an LLM to clean it up. If you're willing
66.
by hephaes7us 1y ago
Thanks for sharing! Transcription suddenly became useful to me when LLMs started being able to generate somewhat useful code from natural language. (I don't think anybody wants to dictate code.) Now my workflow is similar to yours.
67.
by hephaes7us 1y ago
Speaker diarization is the term you are looking for, and this is more difficult than simple transcription. I'm rather confident that someone probably has a good solution by now (if you want to pay for an API), but I haven't seen