Repository Issue Activity (beta)

#1748 SuperCompress: Free open-source LLM prompt compression - cut token costs by ~65%

#1747 anima — a substrate-native consciousness: capability gaps are architecture gaps (open repo, frozen verdicts)

#1746 Consider using pre-built Flash Attention kernels via `kernels`

#921 [unimim] mismatched positional_embed about vit-large/14 for input resolution with 196

#1084 Call for unimim's code

#1579 unimim is still unavailable after one year

#1742 batch size of WavLM

#1618 Unable to download the models and test dataset of trocr

#1719 LHRMs (Large Hybrid-Reasoning Models) code & checkpoints referenced in paper 2505.14631 not found

#1737 Bug report / fix note

#1657 Pretrained Weights of Differential Transformers

#1663 the code of Differential Transformers training

#1718 Differential Transformer loss spikes while training.

#1066 The project of E5: How to fine-tune on the E5 model?

#1732 Can't download kosmos-2.pt checkpoint, ERROR 403

#1735 dit-base-224-p16-500k-62d53a.pth, dit-large-224-p16-500k-d7a2fb.pth models

#1671 [BEATs] Links for the weights of trained BEATs model are unavailable

#428 LayoutReader - slow forward speed

#1733 Question about def init_embed_

#1731 why layoutlmv3 did't use image data, when fintune xfund?

#1730 Textdiffuser2 inference issues

#1619 layoutlmv3 chinese 模型训练和推理

#627 Finetune Trocr on my own dataset

#1729 Not able to run Kosmos-2.5 model with VLLM

#1591 The pretraining code and model of beit3