We propose Index Backpropagation Quantization (IBQ), a new vector quantization method for the joint optimization of all codebook embeddings and the visual encoder, ensuring the consistent latent space. IBQ enables scalable training of visual tokenizers and, for the first time, achieves a large-scale codebook (2^18) with high dimension (256) and high utilization.
Runs of TencentARC IBQ-Tokenizer-262144 on huggingface.co
7
Total runs
-3
24-hour runs
-3
3-day runs
-4
7-day runs
-5
30-day runs
More Information About IBQ-Tokenizer-262144 huggingface.co Model
IBQ-Tokenizer-262144 huggingface.co is an AI model on huggingface.co that provides IBQ-Tokenizer-262144's model effect (), which can be used instantly with this TencentARC IBQ-Tokenizer-262144 model. huggingface.co supports a free trial of the IBQ-Tokenizer-262144 model, and also provides paid use of the IBQ-Tokenizer-262144. Support call IBQ-Tokenizer-262144 model through api, including Node.js, Python, http.
IBQ-Tokenizer-262144 huggingface.co is an online trial and call api platform, which integrates IBQ-Tokenizer-262144's modeling effects, including api services, and provides a free online trial of IBQ-Tokenizer-262144, you can try IBQ-Tokenizer-262144 online for free by clicking the link below.
TencentARC IBQ-Tokenizer-262144 online free url in huggingface.co:
IBQ-Tokenizer-262144 is an open source model from GitHub that offers a free installation service, and any user can find IBQ-Tokenizer-262144 on GitHub to install. At the same time, huggingface.co provides the effect of IBQ-Tokenizer-262144 install, users can directly use IBQ-Tokenizer-262144 installed effect in huggingface.co for debugging and trial. It also supports api for free installation.
IBQ-Tokenizer-262144 install url in huggingface.co: