lucasjin
lucasjin
AI & ML interests
None yet
Organizations
lucasjin's activity
How does the attention_mask contribute to the projector preformance?
5
#45 opened 12 days ago
by
lucasjin
What processor to use?
5
#4 opened 19 days ago
by
floschne
V100 32GB can not inference
1
#3 opened 7 days ago
by
lucasjin
is this the llama-3-8b model clone?
10
#1 opened about 1 month ago
by
malhajar
About the token / per image in report
2
#13 opened about 1 month ago
by
lucasjin
Where is source code?
1
#12 opened about 1 month ago
by
lucasjin
The performance is very bad
#2 opened about 1 month ago
by
lucasjin
What's the chat template used here?
3
#1 opened about 2 months ago
by
lucasjin
How to download the images?
#25 opened about 2 months ago
by
lucasjin
GeoQA+ missing images
3
#5 opened 3 months ago
by
lucasjin
Many assistant response are None.
#2 opened 4 months ago
by
lucasjin
Does the pretrain data open?
1
#1 opened 4 months ago
by
lucasjin
Why a 2B model size up to 80% of 7B?
1
#2 opened 4 months ago
by
lucasjin
I didn't say how does clear and sublimt claimed in code?
#2 opened 4 months ago
by
lucasjin
Can't load model&inference with LLava's latest inference code
3
#3 opened 4 months ago
by
kldsaid
Unable to load
1
#4 opened 4 months ago
by
lucasjin
Int4 version maybe?
2
#2 opened 4 months ago
by
lucasjin
yi-34B VL's int4 please
#1 opened 4 months ago
by
lucasjin
source code and paper?
8
#6 opened 4 months ago
by
josephykwang
is this base model or sft model?
1
#3 opened 4 months ago
by
lucasjin
Add some multi language data
1
#10 opened 5 months ago
by
lucasjin
Is there a combined model?
#16 opened 5 months ago
by
lucasjin
Compare with official Yi-34B-Chat?
5
#6 opened 5 months ago
by
lucasjin
Nevermind. This was probably an Arc test issue.
2
#7 opened 5 months ago
by
Phil337
Use in ComfyUI?
5
#9 opened 9 months ago
by
eyesss
the inpainted result seems blury, why?
2
#13 opened 7 months ago
by
lvZic
the inpainting result by this repo is false, the resolution degrade and noise increase!
1
#15 opened 7 months ago
by
lvZic
sd xl refiner for 1.0 version
#15 opened 6 months ago
by
lucasjin
生成decode会有乱码
3
#1 opened 6 months ago
by
lucasjin
thank s for the model, I got some bad bytes when decoding Chinese
#2 opened 6 months ago
by
lucasjin
KeyError: 'model.embed_tokens.weight'
#1 opened 6 months ago
by
lucasjin
When will a 1024x1024 turbo model will be out?
5
#8 opened 6 months ago
by
lucasjin
Can use chatml format?
1
#15 opened 6 months ago
by
lucasjin
Result not good
3
#4 opened 6 months ago
by
lucasjin
the int8 speed are very slow
1
#1 opened 6 months ago
by
lucasjin
It can not even answer this question: 一公斤的棉花和一公斤的铁,哪一个更重?
10
#16 opened 6 months ago
by
lucasjin
LCDM support
1
#33 opened 6 months ago
by
lucasjin
training code
#2 opened 6 months ago
by
lucasjin
About sft got unstoppable \n when output
1
#10 opened 6 months ago
by
lucasjin
When will disl_large_v3 comes out?
2
#16 opened 6 months ago
by
lucasjin
Any int4 version ?
1
#9 opened 6 months ago
by
lucasjin
How is this even possible?
5
#2 opened 6 months ago
by
pig4431
Small error in text-to-image example
3
#3 opened 6 months ago
by
Quasimondo
Does the weights is fp16?
1
#6 opened 6 months ago
by
lucasjin
What base model does it based?
2
#14 opened 6 months ago
by
lucasjin
Inference error
1
#3 opened 6 months ago
by
lucasjin
will chat model opensource?
6
#6 opened 7 months ago
by
lucasjin
when will the model update?
1
#1 opened 7 months ago
by
lucasjin
AttributeError: 'BaichuanTokenizer' object has no attribute 'sp_model'
1
#3 opened 7 months ago
by
lucasjin
AttributeError: 'BaichuanTokenizer' object has no attribute 'sp_model'
7
#18 opened 7 months ago
by
lucasjin
How many Chinese data in this dataset?
1
#3 opened 7 months ago
by
lucasjin
Add some Chinese data
#1 opened 7 months ago
by
lucasjin
When will base model release?
#2 opened 7 months ago
by
lucasjin
上下文长度是多少?
3
#9 opened 7 months ago
by
lucasjin