RenShuhuai-Andy/TimeChat issues and pull requests

#33 - Why the result of temporal video grounding is always the multiple of 5?

Issue - State: open - Opened by zhengrongz 6 months ago - 5 comments

#32 - 用自己的数据集finetune，如何在train的过程中进行eval？

Issue - State: open - Opened by changqinyao 6 months ago

#31 - Based transformers version needed for modifying models/modeling_llama.py

Issue - State: open - Opened by yeahjack 6 months ago - 3 comments

#30 - Questions about the provided fine-tuning model parameters

Issue - State: open - Opened by LanXingXuan 7 months ago - 1 comment

#29 - Inference with audio

Issue - State: open - Opened by lakshya-frontera 7 months ago - 2 comments

#28 - Question about the output of the time-aware frame encoder

Issue - State: open - Opened by Mingxiao-Li 7 months ago - 2 comments

#27 - When conducting SFT experiments, setting batch_size_train to 1 or 2 has the same memory usage.

Issue - State: open - Opened by tiesanguaixia 7 months ago

#26 - Can this model do qa tasks？

Issue - State: open - Opened by leexinhao 7 months ago - 1 comment

#25 - Question about fune-tune

Issue - State: closed - Opened by zhengxingmao 7 months ago - 7 comments

#24 - Subset of YT-Temporal

Issue - State: closed - Opened by patrick-tssn 7 months ago - 1 comment

#23 - Question about batch size

Issue - State: closed - Opened by gyxxyg 7 months ago - 1 comment

#22 - the performance is very low on my own dataset.

Issue - State: closed - Opened by onlyonewater 7 months ago - 5 comments

#21 - Experiment-related question

Issue - State: closed - Opened by zhaodongliang678 7 months ago - 3 comments

#20 - Question about prompt

Issue - State: closed - Opened by Ironieser 7 months ago - 5 comments

#19 - Question about the tokenizer

Issue - State: closed - Opened by gyxxyg 8 months ago - 5 comments

#18 - Question about prompts.

Issue - State: closed - Opened by gyxxyg 8 months ago - 2 comments

#17 - What is the relationship between segment and timetoken？

Issue - State: closed - Opened by sunwhw 8 months ago - 3 comments

#16 - Inquiry on training cost

Issue - State: closed - Opened by HenryHZY 8 months ago - 2 comments

#15 - Demo can‘t show the same desult

Issue - State: closed - Opened by xiaoxiaoli666 8 months ago - 1 comment

#14 - Bad performance of Charades

Issue - State: closed - Opened by soyeonhong 8 months ago - 1 comment

#13 - RAM and VRAM requirement

Issue - State: closed - Opened by Coronal-Halo 9 months ago - 2 comments

#12 - Seeking Clarification about Fine-tuning Datasets

Issue - State: closed - Opened by ShramanPramanick 9 months ago - 2 comments

#11 - Details of sliding qformer operation

Issue - State: closed - Opened by jihwanp 9 months ago - 1 comment

#10 - Do we need to crop the HiREST videos?

Issue - State: closed - Opened by yeliudev 10 months ago - 14 comments

#9 - torch.load raise TypeError: 'strict' is an invalid keyword argument for Unpickler()

Issue - State: closed - Opened by wwq66 10 months ago - 4 comments

#8 - the generalization performance is bad when testing on custom videos.

Issue - State: closed - Opened by dragen1860 10 months ago - 1 comment

#7 - Error in loading Video-LLaMA-2-7b_Finetuned