chris-aeviator

It might some of the points I raised really belong to a streaming compatible format, so much reminds me of tiling used e.g. in Vector Maps. If gltf supports additional...

same. I can get around the 'flickering' re acquiring of the microphone also by increasing the config from 300 to 5000 but the result is no output or gibberish. The...

@louis030195 so it's working for you? I'm using the standard model as in the example code and see gibberish output.

I will try to dig the notebook up

It would be nice to know (and see in the zap UI) how many confirmations an opening transaction has received. I can see for example in mempool, that I have...

the VM will not boot without the `ks.cfg` file provided via the `http_directory` file. Unfort. the `ks.cfg` appears in the docs, but is not part of this repository.

> I skimmed to the paper and realized that the `regex`-like generation will not work, but I played around with the `json` generation, which proved useful for this usecase. Do...

the host setting seems to be configured with viser https://viser.studio/server/

while I can patch the model it won't work with the standard HF tools. Transformers assumes past_key_values to be subscriptable but past_key_values is a ContextManager ``` model_inputs = tokenizer([prompt], return_tensors="pt").to("cuda")...

Does this have any real consequences? I realized inconsistent ingests (ingest works fine, Get queries return only few results from many) and also see this message.