EvelynRYW comments

Results 15 comments of


                                            EvelynRYW

Increase size of "file" field in General Tab of Entry-Editor

Hello, @[[ThiloteE](https://github.com/ThiloteE)] , our group are interested in your problem and would like to work on this issue. Can we have a try? ——SE 2022 group haha, SUSTech

Increase size of "file" field in General Tab of Entry-Editor

And for result, simply show! ![G7XGBAA`YQYNEG{H3V%5$6C](https://user-images.githubusercontent.com/74488290/170689766-c5c469bd-3588-4584-970d-ef83f631da52.JPG) ![4`_022PM)2R69)ZBHX1 U)T](https://user-images.githubusercontent.com/74488290/170689768-5f159bcd-3691-49c1-bbe3-dc1510e3846e.JPG)

Increase size of "file" field in General Tab of Entry-Editor

Thanks, but also, I just take a example, not complete!

Inconsistent Handling Of Duplicate Attributes

Hello, @[alycecil](https://github.com/alycecil) , our group are interested in your problem and would like to work on this issue. Can we have a try? ——SE 2022 group haha, SUSTech

Merge some (or all) of the "new entry" related buttons

Hello, @[[claell](https://github.com/claell)] , our group are interested in your problem and would like to work on this issue. Can we have a try? ——SE 2022 group haha, SUSTech

Grounding微调的prompt模版

想问下目前1.5版本的VG任务微调，prompt版本还是一样么？在上传的playground中，sharegpt4v_mix665k_cap23k_coco-ap9k_lcs3k_sam9k_div2k.jsonl文件里面涉及的coco定位任务prompt似乎并没有的tag？

Using Swift to perform inference and fine-tune InternVL-Chat-V1.5

@hjh0119 想问下按照您的教程里，我想对int8版本进行微调，替换掉了--model_type internvl-chat-v1_5-int8 模型名称，加载了从huggingface下载的模型权重，但报了如下的错误，似乎和int8有些关系，对与1.5的版本则不会报错，可以正常微调，想问下这是需要加什么特殊的参数影响么？ ``` /data2/renyw/InstallationPackage/anaconda3/envs/swift/lib/python3.10/site-packages/bitsandbytes/autograd/_functions.py:316: UserWarning: MatMul8bitLt: inputs will be cast from torch.bfloat16 to float16 during quantization warnings.warn(f"MatMul8bitLt: inputs will be cast from {A.dtype} to float16 during quantization")...

Using Swift to perform inference and fine-tune InternVL-Chat-V1.5

> @MVP-D77 不需要替换model_type，使用本地模型文件在--model_id_or_path指定路径，再指定--model_type internvl-chat-v1_5-int8. @hjh0119 您好，这是我的微调命令，请您看一下有没有问题，它还是报错，和上面报错信息一样，同样的数据集，internvl-1.5就不会报错，另外想问下，flash_attn如果想开启的话，是加一个设置为true么，我看打印的log文件里面显示use_flash_attn 为null ``` CUDA_VISIBLE_DEVICES=0,1 swift sft --model_type internvl-chat-v1_5-int8 --dataset coco-mini-en-2 --model_id_or_path xxxxxxxxx/InternVL/pretrained/InternVL-Chat-V1-5-Int8 ```

Using Swift to perform inference and fine-tune InternVL-Chat-V1.5

> > @hjh0119 您好，这是我的微调命令，请您看一下有没有问题，它还是报错，和上面报错信息一样，同样的数据集，internvl-1.5就不会报错，另外想问下，flash_attn如果想开启的话，是加一个设置为true么，我看打印的log文件里面显示use_flash_attn 为null > > 可以复现，修复中开启flash-attention： --use_flash_attn true 关于swift的问题欢迎到swift提issue @hjh0119 谢谢您的回复，也感谢您的工作，我还有两个新的有关internvl微调的问题，已在swift仓库提出issue [https://github.com/modelscope/swift/issues/925](https://github.com/modelscope/swift/issues/925) 非常期待您的回复