-
**Describe the bug**
I tried running deepspeed zero 3 on a new huggingface model and got the following error:
[2023-12-13 04:12:18,837] [WARNING] [parameter_offload.py:86:_apply_to_tenso…
-
### [What anime is this?](https://whatanime.ga/about)
Search engine that helps users trace back the original anime by screenshot.
Request url: `https://whatanime.ga/search`
Form: `data=(image)`…
koteq updated
4 years ago
-
When having multiple commits or when executing MOE with git author, which is different from the one who made the commits to sync, the magic directive fails with the error as stated in the title of thi…
-
**Describe the bug**
I'm trying to use the Llama2 model saved with `--use-dist-ckpt` after SFT (Supervised Fine-Tuning) to train a reward model. The reward model does not require the original checkpo…
-
### Your current environment
Not applicable -- Dockerfile.
### 🐛 Describe the bug
Steps to reproduce:
- Clone the `vllm` repo
- run `docker build . --target vllm-base`
- Build fails
```shel…
-
The vLLM [fused moe kernel](https://github.com/vllm-project/vllm/blob/main/vllm/model_executor/layers/fused_moe.py) used for Mixtral uses the standard data parallel parallelization which works well wi…
-
**Describe the bug**
I am trying to create a minimal run-able example of Smart Scheduling proposed by the FasterMoE paper. However, when I profile the example using Nsight Systems, it seems that ther…
-
### Your current environment
```text
PyTorch version: 2.4.0+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
OS: Ubuntu 22.04.4 LTS (x86_64)
GCC vers…
-
Hi there,
i am trying to run a unit test that i wrote to test a certain functionality.
It works fine if let it execute through "sudo docker build -t moe_container ."
However i would like to be abl…
-
Hi,
I like your app and I was wondering if we could add search engine as "mods" or something like that.
And can we add search engine now easily? (I don't know kotlin well)
Thanks for this app