univa-JASON

Finetuning

2

Hi, How Can I replace this model's language model with another model and fine-tune it? and can't i fine tune this model without shift?

In Technical Report

2

How did you proceed with DPO learning? Using CPMTrainer, or HF DPOTrainer? Does CPM Trainer support DPO finetuning?

Can I apply wsd scheduler in This repo's finetuning code? When i set as follows, ``` --lr_scheduler_type warmup_stable_decay \ ``` i got this error ``` Traceback (most recent call last):...

about multi-image finetune

4

when i finetune v2_5, can i compose dataset like this? ``` [ { "id": "0", "image": { "": "path/to/image_0.jpg", "": "path/to/image_1.jpg", "": "path/to/image_2.jpg" }, "conversations": [ { "role": "user", "content":...