thunlp / LLaVA-UHD

LLaVA-UHD: an LMM Perceiving Any Aspect Ratio and High-Resolution Images

thunlp/LLaVA-UHD Issues

RuntimeError: Given groups=1, weight of size [1024, 3, 14, 14], expected input xxx to have 3 channels, but got xx channels instead
Updated 20 days ago1
The Vicuna LLM is not frozen during pretraining
Updated a month ago
RuntimeError: Given groups=1, weight of size [1024, 3, 14, 14], expected input[12, 15, 336, 336] to have 3 channels, but got 15 channels instead
Updated a month ago
Trained Resample with Siglip Got inconvergence loss
Updated a month ago
Inconsistent Calculation of Patch Numbers in Image Processing and Encoding
Updated 2 months ago
Can we get the weights? :)
Updated 2 months ago
About the adaptive size part question
Updated 2 months ago
Loss Convergence and whether ViT is Trained
Updated 2 months ago
batch size matters in training
Updated 2 months ago
Isn't 2048 max lenght would out of context if 6 images put in?
Updated 2 months ago
[Question] Proof about Range of Slice Aspect Ratios
Updated 2 months ago1
Can I input 2688×672 images?
Updated 2 months ago
Meaning of '8' and '4'
Updated a month ago1
Runtime error
Updated 2 months ago1
[Question] `image_features` not matched to input text
Updated 2 months ago
position embedding reshape error
Updated 2 months ago2
Weird pre-train loss
Updated 2 months ago1
a problem configuring the environment：ERROR: llava@ file://
Updated 2 months ago3
About the slice_logic
Closed 2 months ago2
RuntimeError: Given groups=1, weight of size [1024, 3, 14, 14], expected input[16, 9, 336, 336] to have 3 channels, but got 9 channels instead
Updated a month ago11
How do you train your ViTs?
Closed 2 months ago3
update vit
Closed 2 months ago1
Weird slice_logic test output
Closed 2 months ago1
Comparing with LLaVA 1.6 Next
Updated 3 months ago1