-
I am running evaluations on `mmmu_val` with W&B logging enabled. And have the following issues:
```
Traceback (most recent call last):
File "/root/miniconda3/envs/py3.10/lib/python3.10/site-packa…
-
Thank you for this nice tool.
I was wondering if susie_rss() could be safely applied to GWAS summary statistic (SS) derived from a linear mixed effect model, the SS being the fixed effects of SNPs …
-
Error from Rust:
```zsh
error[E0432]: unresolved imports `libc::posix_fadvise`, `libc::POSIX_FADV_DONTNEED`
--> crates/rooch-common/src/fs/file_cache.rs:4:12
|
4 | use libc::{posix_fadvise,…
-
Using the command: `CUDA_VISIBLE_DEVICES=0,1,2,3 python -m sglang.launch_server --model-path lmms-lab/llava-next-72b --tokenizer-path lmms-lab/llavanext-qwen-tokenizer --port=8000 --host="0.0.0.0" --t…
-
when I run sglang/examples/usage/llava/srt_llava_next_test.py,
change to "lmms-lab/llava-next-72b" instead of "lmms-lab/llama3-llava-next-8b", it reports OOM as below.
could anyone take a time to gi…
-
Hi,
this is quite high level but I believe it would be useful to share how specific software with timelines examples.
I have in mind Kdenlive but also LMMS, Audacity, etc. Being able to use the …
-
### Question
Hi, great work! I want to download the llava-next stronger checkpoint, but the website (https://huggingface.co/collections/lmms-lab/llava-next-6623288e2d61edba3ddbf5ff) meets 404 Error.
-
训练过程中很快出现loss跳变为0的现象,降低学习率无法解决该问题。
配置文件如下:
model:
arch: st_llm_hf
model_type: instructblip_vicuna0
use_grad_checkpoint: True
max_txt_len: 256
end_sym: "###"
#prompt_path: "prompt…
-
This appears to be an upstream issue, but I am adding it here to inform should the issue, should it come up with another user. I am experiencing the same problem. It is a shame that there doesn't appe…
-
the output of output_ids is tensor([[1, 2]], device='cuda:0')
Other output of the demo script is:
Question: A chat between a curious user and an artificial intelligence assistant. The assistant gi…