Bartowski

InternLM2 math model breaks with exllamav2_HF loader (works with non-HF)

12

### Describe the bug Trying out https://huggingface.co/bartowski/internlm2-math-20b-llama-exl2 and whenever I use HF loader to generate i get a bunch of errors. When using the non HF loader I get no...

Currently with some wireless/battery powered cameras (in my case, Eufy cameras), they are implemented in a clever way such that they run a local motion sensor, and on motion detection,...

Recreating a docker container causes errors looking for old one until reboot

2

When I bring a docker container down and back up, I see my logs spammed with: 2023-04-21 16:16:03.098 ERROR (Thread-2 (RunDocker)) [custom_components.monitor_docker.helpers] [Docker] teslamate_teslamate_1: Container not available anymore (3) (DockerError(404,...

I run my LLMs on text-generation-webui which also provides an openai-compatible endpoint, it would be great if we could provide our own generation parameters since it's unhappy with the max_length...

Bug: Quantizing Llama 3.1 70B to Q4_K_S with imatrix gives NaN

3

### What happened? Trying to quantize Llama 3.1 70B to Q4_K_S with imatrix gives NaN for block 48 Tagging @slaren because you always seem to solve these Didn't see it...

### System Info When using HF_HUB_ENABLE_HF_TRANSFER, I still see my python huggingface_hub uploading as LFS file ### Reproduction import os os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1" from huggingface_hub import HfApi, logging api.upload_file(/information here/)...

Any way to limit upload speeds?

3

hf_transfer is SO efficient that it actually completely clogs my bandwidth and makes me unable to maintain SSH connections.. Is there any way we can add a cap on up/download...