Bojun Feng
Bojun Feng
Rewrote deprecated Gradio UI with RESTful Client and added it to examples as `gradio_arena.py`. Instructions should be available by running the file with the `--help` tag. e.g. `python path/to/inference/examples/gradio_arena.py --help`...
Add `--model-uid -i` in the terminal as an optional argument that overrides the default randomly generated uid. Resolves #254 example: Run the following two commands in two separate terminal windows....
Resolve #483 at frontend level by filtering options on render if machine is Mac-like. Tested locally, successfully removed 4-bit and 8-bit quantization on MacBook
Resolves #504 - Refactor theme to implement translations - Translate the page names in dashboard Effect:  This PR does not complete translating everything, but hopefully provides a framework to...
Resolve #533 Add support for rwkv-4-pile models. Successfully tested the 169m variant on a local Mac setup. I would like some guidance on setting the model size. Historically, we've rounded...
### Anything you want to discuss about vllm. I was testing out vLLM on Colab and notices something weird. It seems from the code that vLLM is using first come...
Updated comment in [toy_example/point-estimation.ipynb](https://github.com/litian96/TERM/blob/adb12e31a3f063ac9f01d84e75871423f6e14c17/toy_example/point-estimation.ipynb) to match the function of code. Resolve #13
In the following screenshot from [toy_example/point-estimation.ipynb](https://github.com/litian96/TERM/blob/adb12e31a3f063ac9f01d84e75871423f6e14c17/toy_example/point-estimation.ipynb), it seems that the re-centered data is centered around (-0.6, -0.6) instead of (3,3) as mentioned in the comments.
### Is your feature request related to a problem? Please describe Similar to OpenAI's JSON mode, we can use custom grammar files to ask Llama.cpp to generate a valid JSON...