-
# Issue
In https://github.com/kinu-garage/hut_10sqft/issues/862#issuecomment-1631509730 after a few days of try & error, it is finally getting clear that p16s is not powerful enough for Unreal 5.2 …
130s updated
3 months ago
-
### Your current environment
```text
PyTorch version: 2.2.1+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
OS: Ubuntu 20.04.3 LTS (x86_64)
GCC ve…
-
When I boot the API server:
`python3 -m vllm.entrypoints.openai.api_server --port 9999 --dtype half --trust-remote-code --host 127.0.0.1
`
The API spins up in 10 seconds.
However when I downlo…
jpeig updated
5 months ago
-
As a Windows user, I tried to compile this and found the problem was on these two files "```flash_fwd_launch_template.h```" and "```flash_bwd_launch_template.h```". below "```./flash-attention/csrc/fl…
-
How long does the training run take on Cifar100 dataset, with a mismatch rate of 20% and other options as default?
I took 20 hours up to 5 cycles with RTX3090. Is it right for it to take this long?
-
Hi, Thank you for your reply. I have one more question related to configurations. Can you elaborate a bit on " --json_file JSON_FILE path to json file". What kind of data does it contain? Is it the fo…
-
Hi,
Thanks for this repo. It works great. I am just not able to make Flash Attention works. I have a RTX 3090 so it's supposed to work based on their documentation.
I followed the instructions a…
-
Hi!
I'm using rtx 3090 ti 20GB gpu.
I used 32 for `dataset-config.batch.image_width` and `dataset-config.batch.image_height` in `train_small.sh` (also for `validation` ones) but still gets the OOM…
-
### System Info
peft installed from main
c4c826c0c0dfe06620db8b8adcfa55f22acfed61
- `transformers` version: 4.39.1
- Platform: Linux-6.5.0-26-generic-x86_64-with-glibc2.35
- Python version: 3…
-
Dear Developers,
Hi. I enjoyed a lot using FBPIC. But when I tried to run my simulation across several GPUs, I got weird, unphysical results. I have performed the same simulation with one, two and fo…