-
### Feature Description
Since the `CompleteAccessor` stores the metadata, and `Access::info` return the `Arc`, we can move logic from `fn metadata(&self) -> Arc` to `impl Layer for CompleteLayer` to …
-
- get `dcurl` for `github` [DLTcollab/dcurl](https://github.com/DLTcollab/dcurl)
- compile by bitbake
- install `dcurl` into target image
-
I'm running post-training on a pruning model. After post-training, I get degraded performance - eg. mmlu goes down to 24%. is this expected?
```
MODEL=meta-llama/Llama-2-7b-hf
prune_ckpt_path=…
-
The functions
- da.linalg.cholesky
- da.linalg.lu
- da.linalg.solve (it's a wrapper around the above two)
contain the same keys in multiple layers of the HighLevelGraph.
Since #7274 (2021.03.…
-
### What happened?
Hi there.
I am trying to use the `np` parameter to serve multiple requests in parallel. However, the generated tokens are garbled when I set the `np` parameter to a relatively lar…
-
Hi,
There are many license issues by filtered by SPDX license policy which would be added by scarthgap branch. I modified them manually but, I am not sure it is correct approach.
License issues …
-
### What happened?
llama.cpp is running slow on NVIDIA A100 80GB GPU
Steps to reproduce:
1. git clone https://github.com/ggerganov/llama.cpp && cd llama.cpp
2. mkdir build && cd build
3. cmak…
-
Hi @mirzak can you create the relevant branch for kirkstone? Such as all Yocto related layers do, so as to have it clearer which Yocto release your board is supporting, aside from https://github.com/m…
-
### What happened?
Hi, im trying to use Google [Madlad400 in GGUF version,](https://huggingface.co/NikolayKozloff/madlad400-10b-mt-Q8_0-GGUF) but I'm unable to work it with `llama-server` but it work…
-
### What happened?
Hi, recently, I'm trying to learn the gguf-py lib and use the gruff-py and write a script to make a gguf file, after I made the file, I tried to load it using llama-cli, but it sai…