Repository Issue Activity (beta)

#1928 [Question] Why does LLaVA evaluation assert batch_size == 1 for benchmark inference?

#1927 [Question] What are the dependencies like what is the minimal Flashattention verison, Pytorch version any other ?

#1913 [Question] Changing the vision tower.

#602 [Question] Is the response format "Answer the question using a single word or phrase. " used during training?

#1916 [Usage] EvalAI challenge server for vizwiz is no longer open for testing

#1914 [Discussion] How to finetune_task_lora.sh in google collab pro which is already python 3.12?

#1926 [Question] the number of finetuning epochs mismatch

#1164 [Usage] Missing "trainer_state.json" when resuming training from saved checkpoints

#408 [Usage] LoRA finetuned weights provided for vicuna-13b-v1.3 gives NaN / inf error when performing inference on COCO-2014 questions after merging LoRA weights

#1923 📋 Documentation Enhancement Suggestion

#1313 [Usage] can not apply inference with a merged mistral model after finetune it

#1922 LLaVA-v1.5-7B fails to parse images and returns no valid response in Gradio UI on Windows+RTX 5060

#1906 [Question] Token indices sequence length is longer than the specified maximum sequence length for this model (2071 > 2048).

#772 [Question] Why VQAv2 always "Execution time (sec.): None"

#927 [Question] Unable to reproduce GQA, SQA, and MME results on LLaVA-v1.5-7b

#1607 [Question] Unable to submit VQAv2 result file to the evaluation server

#1792 [Question] Why does text-only data use the empty image token?

#1919 Help with "assert self.weight.shape[1] == 1" Error in LLaVA After Fine-Tuning

#1789 [Question] Fintune llava-v1.5-7b just use all textvqa data, but get Accuracy: 0.00%

#1918 Tokenizer behavior difference when adding custom token "[SEG]" between LLaVA v1.5 and LLaVA-LLaMA2

#1175 [Question] llava-1.6 在自有数据上输出不符合预期

#677 [Usage] Getting IndexErrors finetuning on a custom dataset

#947 [Usage] How can I use the LoRA checkpoints for finetuning on custom data ? I tried but ran into some issues

#673 [Usage] Some Weights not used, when loaded llava-v1.5-7b to test

#1470 [Question] FT use 1.5 face a issue that tensor mismatch