You'll need multiple simple tools to refine the raw arrow files and NDJSON files.
A lightning-fast search engine API bringing AI-powered hybrid search to your sites and applications. But also the most stared search engine on GitHub just being Elastic Search.
You can download it from the releases page.
| // This is set by the engine for every invocation of this function. | |
| let context = #{ group_size: 3, group_id: 1 }; | |
| // The _vectors field doesn't even exists. | |
| let doc = #{ id: "abc1234", metadata: #{ title: "super title" } }; | |
| print(doc); | |
| // ---- start ----- | |
| fn hash(s) { |
| import OpenAI from "https://deno.land/x/[email protected]/mod.ts"; | |
| const client = new OpenAI({ | |
| apiKey: process.env['MEILI_API_KEY'], | |
| baseURL: 'http://localhost:7700/chats/main', | |
| }); | |
| // Can be called by Meilisearch to inform the front-end about its progression | |
| const meili_search_progress = { | |
| "type": "function", |
Those are the redb benchmarks with different key value stores and the two different versions of LMDB living on two different branches: mdb.master and mdb.master3. The benchmarks were done on my iMac M3 (2023).
| redb | lmdb (mdb.master) | lmdb (mdb.master3) | rocksdb | sled | sanakirja | |
|---|---|---|---|---|---|---|
| bulk load | 2.97s | 982.90ms | 987.29ms | 5.40s | 4.49s | 865.30ms |
| individual writes | 443.95ms | 7.30ms | 7.37ms | 2.49ms | 403.99ms | 52.14ms |
| batch writes | 4.07s | 2.09s | 2.10s | 222.74ms | 1.38s | 3.63s |
| len() | 2.96µs | **125. |
| // This script can literally replace the whole changes I made on this Rust program to alterate | |
| // my dataset to be able to compute scores. You can find the function by follozing this link: | |
| // <https://github.com/Kerollmops/musicbrainz-artist-scoring/blob/901ea2d35707477880f328211c76945a15b2410b/src/main.rs> | |
| // The file extension is .rhai.rs just to be able to highlight the code. | |
| // You can run this Rhai script on the Rhai Playground: | |
| // <https://rhai.rs/playground/stable> | |
| // Here is the context and document Meilisearch provides. | |
| // It's just the setup part of the script |
| // Meilisearch is setting the document for you, this way. | |
| // Note we use the rhai.rs file extension to benefit from | |
| // the Rust Syntax Highlighting only, it's Rhai nothing more. | |
| let doc = #{ title: "star wars" }; | |
| print(doc); | |
| // -- script starts here -- |
| [ | |
| { | |
| "bindings": { | |
| "cmd-shift-[": "pane::ActivatePrevItem", | |
| "cmd-shift-]": "pane::ActivateNextItem", | |
| "ctrl-pagedown": "pane::ActivatePrevItem", | |
| "ctrl-pageup": "pane::ActivateNextItem", | |
| "ctrl-shift-tab": "pane::ActivateNextItem", | |
| "ctrl-tab": "pane::ActivatePrevItem", | |
| "cmd-+": "zed::IncreaseBufferFontSize" |
| # Create an API KEY there: https://pushover.net/apps/build | |
| # Find your user TOKEN here: https://pushover.net | |
| # | |
| # Note that the `title` field is optional and don't forget to put double quotes | |
| # around your argument when calling this function. | |
| # | |
| # Here is a small example of pipping logs through pushover: | |
| # | |
| # nohup tail -f nohup.out | while read l; do pushover "$l"; done & | |
| # |
| /* | |
| * par_build_tree.cpp | |
| * | |
| * Move this in the examples folder of Annoy and compile it like the other c++ example. | |
| * | |
| * Created on: Dec 6, 2023 | |
| * Author: Clément Renault | |
| */ |
| {} |