HNHacker News
TopNewBestAskShowJobs

hint23

29 karma · joined March 7, 2018

submissionscomments
hint23··on AI language models can exceed PNG and FLAC in lossless compression, says study
ts_zip ( https://bellard.org/ts_server/ts_zip.html ) already implements lossless compression with language models and is fast enough to be usable for some applications.
hint23··on Llama 2 on ONNX runs locally
For best performance on x86 CPU and Nvidia GPU, ts_server is interesting ( https://bellard.org/ts_server ).
hint23··on GPT-J-6B – A 6 billion parameter, autoregressive text generation model
GPT-J can now be used at https://bellard.org/textsynth
hint23··on Show HN: CheerpX – x86 Node.js/python3/Ruby REPL in browser via WebAssembly JIT
You can already do all that with https://bellard.org/jslinux
hint23··on Lossless compression of English messages using GPT-2
FYI, the corresponding standalone Linux command line version is available at https://bellard.org/nncp/gpt2tc.html . It also does text completion and file compression.
hint23··on OpenAI Releases Largest GPT-2 Text Generation Model
You can try it at: http://textsynth.org
hint23··on Lessons Learned from Building an AI Writing App
To give some perspective, the setup of http://textsynth.org consists in a single C Linux executable of 250 KB on the server and in 150 lines of Javascript code on the client without any dependency on other libraries...
hint23··on Natural Language Processing: The Age of Transformers
You can also look at http://textsynth.org .
hint23··on LibBF – a small library to handle arbitrary precision floating point numbers
You can already try LibBF online at http://numcalc.com