sprappcom

possible to do one that can fit into 7GB vram?

2

7b is > 12gb ram use, can u do one that is maybe 3b parameters or have one 7b whose quantisation is 4_0 gguf or something?

what tokenizer to use for mistral? stories110M is working fine (a lot of nonsense text generated) but how to use mixtral gguf etc?

1. how many cpu core counts is pingora designed for? 2. is this the current one being used on cloudflare?

is there any possibility to support like cloudflare ES stuff? https://developers.cloudflare.com/workers/examples/return-html/ export default { async fetch(request) { const html = ` Hello World This markup was generated by a Cloudflare...

📝 [Proposal]: Reduce memory usage for "prefork" mode of the master process

2

### Feature Proposal Description the master process is using a lot of memory doing nothing ### Alignment with Express API none ### HTTP RFC Standards Compliance none ### API Stability...

as titled

how to use it in client browser?

1

1. is this production ready? 2. who are the users of it? thx

any benchmark on moonscript usage?

1

as titled benchmark against vanilla luajit etc

is the phonenumbers package needed? looks buggy to me. ./vendor/github.com/cloudwego/hertz/pkg/app/server/binding/default.go:73: exprValidator "github.com/bytedance/go-tagexpr/v2/validator" ./vendor/github.com/cloudwego/hertz/pkg/app/server/binding/config.go:25: exprValidator "github.com/bytedance/go-tagexpr/v2/validator" uses github.com/nyaruka/phonenumbers/phonenumbers.go referenced here as well: https://github.com/cloudwego/hertz/issues/1137  

The Fastest Zero Allocation Golang LRU Cache - AccelruX

1

@cristaloleg https://github.com/cloudxaas/gocache X version - super fast https://github.com/cloudxaas/gocache/tree/main/lrux/bytes Normal fast version https://github.com/cloudxaas/gocache/tree/main/lru/bytes