Orr Zohar

Hi, Is it possible to add support for running eval without training given the pretrained attention weights? Best, Orr

### 是否已有关于该错误的issue或讨论？ | Is there an existing issue / discussion for this? - [x] 我已经搜索过已有的issues和讨论 | I have searched the existing issues / discussions ### 该问题是否在FAQ中有解答？ | Is there an...

Can you release the code for video generation?

Hi, Why is this functionality not implemented? https://github.com/LLaVA-VL/LLaVA-NeXT/blob/1654303050a429f5c0dbb97033bf24a2263531cf/llava/model/llava_arch.py#L233-L235 What is needed to implement this? Best, Orr

what size are the resized videos? are they also temporally resampled?

LongVILA - compatibility with other LLMs

1

Hi! very impressed by your work with LongVILA. I would like to do long context with Qwen/LLaMA3.1, but currently, i only see support for Mixtral. Any chance you plan on...

Question re. LanguageModel vs LanguageModelForCausalLM functionalies

2

Hi, Why did you refactor such that the model is of type 'LanguageModel' and not 'LanguageModelForCausalLM'? and why did you move 'get_vision_tower'/etc from 'LlavaMetaForCausalLM' to 'LlavaMetaModel'? Best, Orr

### System Info / 系統信息 Hi I noticed that the learning rate for the instruction tuning phase (1e-7) is 100 times smaller than what was reported in the COG-VLM technical...

functioning?

5

I tried using your code, I reached: when I navigated to http://localhost:3000 How am I supposed to run? Best, Orr

Hi, Is it possible to install the autoresume SDK somehow? from userlib.auto_resume import AutoResume I could not find a python library `userlib`. https://github.com/NVIDIA/Megatron-LM/blob/9ec2337648b31eaac81678e0b8c333decbfcbdf6/megatron/training/global_vars.py#L201 Best, Orr