-
### Describe the bug
Last week after updating it stopped working opted for a reinstall.
At the end of the installation there are warnings and errors.
```Installing collected packages: sentencepi…
-
Hi,
I'm using agx xavier, L4T: 35.4.1, JP:5.1.2, started: riva_quickstart_arm64_v2.12.0. (IP: 192.168.0.40)
and I have one orin agx, JP 6. running docker: dustynv/local_llm:r36.2.0 (IP: 192.168.…
-
### Question
I cloned the latest code and I have tried various versions of the LLaMA-13B weights (including the original one). However, I can not reproduce the results on ScienceQA following the auth…
-
### When did you clone our code?
I cloned the code base after 5/1/23
### Describe the issue
when I running
> `llava.serve.model_worker` with `liuhaotian/llava-v1-0719-336px-lora-merge-vicuna-13b-…
-
### 🐛 Describe the bug
When I try to train model using torch.distributed.FullyShardedDataParallel, I found that :
when training using single-node multi-gpu (1x8A100), the training speed is normal.…
-
### When did you clone our code?
I cloned the code base after 5/1/23
### Describe the issue
I manually download the pre-trained model at my path, here, which click the download button for each.
![…
-
### Question
1, In my understanding, the first pretraining stage uses either the CC-3M Concept-balanced 595K dataset or the LAION/CC/SBU BLIP-Caption Concept-balanced 558K dataset. The second stage u…
-
### System Info
https://github.com/open-mmlab/Multimodal-GPT
Are there any good ways to quantize open-flamingo?
I found after using prepare_model_for_kbit_training the flamingo_init() is reve…
-
## Idea 💡
The **ULTIMATE** achievement for this project would be if Auto-GPT was able to recursively improve itself. That, after-all, is how AGI is predicted by many to come about.
## Suggestion …
-
![image](https://user-images.githubusercontent.com/22076188/233404474-9d0977c7-c374-4aae-b673-06fabccb0466.png)