-
Hi there, thanks for all the hard work.
My system has 2x4090.
First, the FP16 model works if using bitsandbytes 4bit, with decent speeds.
`Output generated in 81.62 seconds (4.47 tokens/s, 36…
-
There is a build issue for libCEED that is missing the compiler flag -fPIC.
After adding the flag in ExternalLibCEED.cmake file, the whole building passes.
Platform: ubuntu 22.04
-
10 t/s vs. 6 t/s on text-generation-webui.
Great project.
-
##### System information (version)
- OpenVINO => openvino2tensorflow-1.34.0
- Operating System / Platform => colab.sandbox default runtime
- Compiler => Not relevant
- Problem classification: Mo…
-
Hi! While 3-bit and 2-bit quantisations are obviously less popular than 4-bit quantisations, I'm looking into the possibility of loading 13B models with 8 GB of VRAM. So far, loading a 3-bit 13B model…
-
starting the server by hand and choose one of the vulkan GPUs works well – actually very well! kudos 4 u!
in auto mode, it tries to load the model to the first reported (integrated) GPU (here: has…
-
Hi! It seems like I can't get the output files using `avworkflow_files`. Below, I tried to extract the output file locations from both aborted and successfully-done workflows using Terra API directly …
-
https://results.finngen.fi/region/WEIGHT_IRN/5:123116412-123516412
The same thing also in gene page https://results.finngen.fi/gene/IL6R
-
### Describe the bug
Hello, I'm new to Oobabooga. fresh install using windows installer.
Crashes with Vicunia and also Alapaca when I deleted the model folder
Here is the full log. Hoping it he…
-
微博内容精选