-
When I try to use the method in llm-awq to quantize models by seeting w_bit = 3,4 and 8 bit, I found that it only worked for 4-bit.
So, I want to clarify the way you compressed the models.
Should …
-
Presently it seems that when connecting to Mail Slurper, it does not respond with server capabilities. Most importantly it doesn't reply with "8BITMIME". MailKit.Net checks to see if a server respon…
-
I'm getting the following error when using channel.put and channel.get to pass simple Ruby Hash objects around (1 level of nesting in the hash).
It works for a little while, but as the communication …
-
This may not be possible, so feel free to close as necessary. It really depends at what bit-depth the new image processing pipeline on the Pi5 handles the image processing - raw conversion/debayering …
-
### Describe the bug
Hi Team,
We observed that when Fluentd agent is overwhelmed with "pattern not matched:" warning after some time it gets stuck and stopped processing. Both tailing and flushing…
-
**Is your feature request related to a problem? Please describe.**
I know that I can use libcc to instrument a C program but I am not sure how to instrument a Rust program and get the coverage data.
…
-
**Is your feature request related to a problem? Please describe.**
When changing the brightness of my keyboard i get difrent levels of green.
The lower the brightness the more green it gets where …
-
Dear Sir
I have managed to work with the screen thank you for that, but I think something in the library is not performing well.
It seems that the library (even when it set to - #define TFT_…
-
Hi all, I've set up Starcoder as follows:
```
gen_checkpoint = "bigcode/starcoder"
gen_device = "cuda"
gen_tokenizer, gen_model = setup_model_tokenizer(
gen_checkpoint, bit_4=False, device=…
-
### Question
gpu信息:
![image](https://github.com/haotian-liu/LLaVA/assets/46987480/51363fcd-1559-42d3-a69c-7a0dbc45060c)
运行命令如下:
CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6 python -m llava.serve.model…