LWM · Issues· 59 open

DreamX-World([https://github.com/AMAP-ML/DreamX-World](https://github.com/AMAP-ML/DreamX-World)): using Wan2.2-5B-TI2V for controllable world simulation

Getting memory error while running run_vision.sh

A question on your implementation of decoder phase of llama

Question about loading tokenizer

bash run_vision_chat.sh -- cause flax.errors.ScopeParamNotFoundError: Could not find parameter named "embedding" in scope "/transformer/wte"

Generate video Only First frame has img, other frames are random pixel

Any Plan for Including Torch Models for Vision-Language

Asking to pad but the tokenizer does not have a padding token

Removing Truncation/Padding during Prefix Tokenization

Error while running bash command: run_sample_video.sh | Error: "TypeError: missing a required argument: 'segment_ids'"

checkpoints for run_sample_video.sh

I wonder if you will release the tokenized dataset.

ValueError: bytes is too large when running scripts/run_train_text.sh

Quantize model weights

Great work! Any plan to train a smaller version, e.g. around 3B?