vtempest

Results 10 comments of vtempest
trafficstars

I am having this issue trying to run cheerio :( `import * as cheerio from "cheerio"; `

Can it get an image data file inside of a pdf - export it as base 64 https://stackoverflow.com/questions/18680261/extract-images-from-pdf-file-with-javascript This is needed to do pdf-to-html in cloudflare workers, and to keep...

> How about @napi-rs/canvas https://github.com/Brooooooklyn/canvas/blob/main/example/image-data.js This might work as polyfil. I'm using this pdf npm to build airesearch.js.org and qwksearch alt to google that shows pdf text alongside result list....

@xenova I am interested in Gliner to ONNX and have extensive exp with Transformers.js v3 in CFW. I am using it in ai-research-agent See here: https://airesearch.js.org/functions/getEmbeddingModel.html https://airesearch.js.org/functions/addEmbeddingVectorsToIndex.html

Sounds good. It can also have a cite like airesearch.js.org

We want a tool that can run in serverless Node.js not something that needs to have a python server running. It could be based on pdfjs like this https://github.com/vtempest/ai-research-agent/blob/master/src/extractor/url-to-content/pdf-to-content.js#L21

I would like to help build the fastest Simd accelerated vector search. If usearch uses simsimd and compiles to wasm, I'd love to improve upon it and build my [vsearch](https://airesearch.js.org/functions/addEmbeddingVectorsToIndex.html)...

Yes I was working on top of the original hnswlib ported to wasm here: https://github.com/kaiobarb/hnswlib-wasm?tab=readme-ov-file I am wondering how to integrate usearch instead as the base lib and then add...

Would that approach (pyodide, etc) work in cloudflare workers? Check out my usearch / vsearch demo. It has getEmbeddingModel, convertTextToEmbedding, addEmbeddingVectorsToIndex, and exports/imports the vector bin for a specific file...

https://github.com/vtempest/ai-research-agent/blob/master/data/misspelled-typos-8k.json I've been working on this issue for 3 wks and compiled this list of typos from 3 well updated sources into one that's 8k total pairs. I've been messaging...