Diwank Singh Tomer

@chuanqisun @Volland @simonwh - have you guys tried integrating with IndexedDB? If not, it could be worthwhile implementing an indexeddb storage engine. Relevant: - https://github.com/cozodb/cozo/blob/main/cozo-core/src/storage/mem.rs - https://developer.mozilla.org/en-US/docs/Web/API/IndexedDB_API - https://github.com/devashishdxt/rexie

@zh217 thoughts on this? wasi-sqlite and indexeddb could be both promising directions

@lloyd: **Fix is working. Merge?**

`pipenv lock` also fails with the following: ```Could not find a version that matches grpcio=1.0.2,>=1.7.0; extra == "grpc"``` Running: - `python 3.6.3` - ~`pipenv 8.2.6`~ `pipenv 8.3.2`

@zh217 Ran `cargo test` and `cargo bench` for all features. All tests pass.

> Hey @creatorrr, founder @ ellipsis-dev here. You can disable code reviews from certain branch names, such as into `dev` or from `f/*` by adding a config file: https://docs.ellipsis.dev/config#ignore-certain-basehead-branches. Thanks...

We need a flag for adaptive context, maybe an environment variable like `EXPERIMENTAL_ADAPTIVE_CONTEXT_ON`

multimodal-support is now merged in to dev and the main change is that the content field of chatml messages can now be a list of `[{type: "image_url", image_url: {...}}, {type:...

I used a few different platforms (replicate, huggingface, openai) for a project recently and noticed that they returned quite a few transient errors on API calls mostly around models loading...

+1 @JohnnyC08 It'd also be interesting to see if including context (conversation history / last utterances) improves the accuracy of predictions.