Issues
- 3
FVD on UCF101
#123 opened by huangyf530 - 2
The resolution of T2V
#118 opened by jang-1996 - 1
- 3
RuntimeError: CUDA error: CUBLAS_STATUS_NOT_SUPPORTED when calling `cublasGemmEx( handle, opa, opb, m, n, k, &falpha, a, CUDA_R_16F, lda, b, CUDA_R_16F, ldb, &fbeta, c, CUDA_R_16F, ldc, CUDA_R_32F, CUBLAS_GEMM_DEFAULT_TENSOR_OP)`
#125 opened by LukeLIN-web - 11
fvd evaluation error on the sky dataset
#117 opened by yizenghan - 5
模型在ucf101上无法收敛
#84 opened by renyuanzhe - 7
Can we use batch_size>1 in sample_t2x.py
#93 opened by foreverpiano - 2
Error once speed up training
#81 opened by moeinheidari7829 - 1
the code of variant 4
#79 opened by pangpangjy - 3
Is autoregression possible?
#71 opened by zhaohm14 - 3
FaceForensics数据集
#67 opened by likeatingcake - 7
how to place and preprocess these datasets
#78 opened by renyuanzhe - 2
Question: model code and design choices
#77 opened by mrartemevmorphic - 2
- 2
question on t2v model training
#75 opened by diffusion-lover - 4
有关时长的问题。
#74 opened by LinQianhe02grey - 2
About video VAE
#73 opened by Darius-H - 3
Some weights of AutoencoderKL were not initialized from the model checkpoint at /path/to/Latte/t2v_required_models/ and are newly initialized because the shapes did not match:
#66 opened by likeatingcake - 2
Dataset of maxin-cn/Latte-1
#122 opened by jinhaaya - 2
huggingface_hub.errors.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': 'maxin-cn/Latte-1/'. Use `repo_type` argument if needed.
#124 opened by LukeLIN-web - 7
训练模型后的效果评估
#102 opened by hejiashuuu - 2
Works. But question: Attributes passed, but not expected (LatteTransformer3DModel)
#111 opened by MarsThunder - 3
About training details
#110 opened by zhang-haojie - 4
Results & ckpts of different sized Latte on UCF-101
#101 opened by yizenghan - 2
How can i train Video VAE?
#120 opened by whikwon - 21
- 3
ImportError: cannot import name 'LattePipeline'
#109 opened by lxrswdd - 5
Invalid JSON file
#116 opened by lxrswdd - 10
FVD of the Taichi checkpoint
#115 opened by yizenghan - 2
About dataset of pretrained T2V model
#113 opened by jang-1996 - 7
T2V training and evaluation
#100 opened by yizenghan - 5
- 2
- 6
CUDA out of memory
#64 opened by likeatingcake - 5
FFS evaluation
#103 opened by yizenghan - 15
FVD on UCF-101
#97 opened by yizenghan - 2
- 4
How to get preprocessed_ffs
#82 opened by zhang-haojie - 4
image_size = [256,512]
#62 opened by likeatingcake - 2
Can Latte train for I2V tasks?
#90 opened by 0Godness - 2
How can I utilize the weights of pre-trained PixArt-α to initialize the parameters of the spatial Transformer block in the Latte T2V model?
#95 opened by g-makerr - 3
- 4
Evaluate the `FVD` on FFS
#94 opened by zhang-haojie - 1
Batch Size Ablations
#91 opened by fan23j - 2
what the param <input_sq_size> stands for?
#92 opened by leonardodora - 6
Question: evaluate the FVD
#80 opened by Alienge - 2
- 0
No positional embeddings in LatteT2V?
#70 opened by DanielSHKao - 5
Evaluate the FVD？
#65 opened by huangjch526 - 2