Shawn
> Have you solve your problem? I searched `image_size` in the whole project, but `image_szie` only exists in `config.yaml`. > > In the default training config, the size was 256,but...
> Have you solve your problem? I searched `image_size` in the whole project, but `image_szie` only exists in `config.yaml`. > > In the default training config, the size was 256,but...
> @leoxiaobin , can u release demo code,when there is one person in img,but result has two,how to deal with it? Hi, you can find demo code here: https://github.com/HRNet/HigherHRNet-Human-Pose-Estimation/pull/28
when I change transformer into 4.38.0, it shows ``` File "/home/user1/workspace/colossal-ai/ColossalAI/examples/language/llama2/attn.py", line 133, in attention_forward cos, sin = self.rotary_emb(v, seq_len=kv_len) File "/home/user1/anaconda3/envs/colossalai/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl return forward_call(*args, **kwargs) TypeError:...
> Transformers v4.37 is OK. But just as I said, you can use v4.39 and still enjoy the speedup from flash_attn by setting `use_flash_attn ` to `False`. Because flash attention...
> > > @DingQiang2018Hello, I noticed that the author modified LOMOTrainer and LOMOLoRaTrainer according to your suggestion. LOMOTrainer runs without problems, but LOMOLoRaTrainer will report an error at self.model.optimizer.partition_all_parameters(). Have...
> > @DingQiang2018 您好，我注意到作者按照您的建议修改了LOMOTrainer和LOMOLoRaTrainer，LOMOTrainer运行没有问题，但LOMOLoRaTrainer会在self.model.optimizer.partition_all_parameters()处报错，您是否遇到了同样的问题呢？谢谢！ > > Hi, lomo_lora_trainer中因为多了lora的optimizer所以不能通过model.optimizer来调用DeepSpeedZeRoOffload。我目前把lomo_lora_trainer.py中的save_model()回退到之前版本了。 Hi, 我注意到了，但是我目前还是没办法做到 merge 之后的 model 有相同的eval resutls。。。。
> I use mmengine==0.8.4, but I don't think this is the reason... Thanks for your patience! Yes... I just tried the mmengine==0.8.4, still raises the same error...
Hi, does this error `ConnectionError: Couldn't reach 'openai_humaneval' on the Hub (ConnectTimeout)` means that I did not connect the API successfully? Thanks Xuan
> Hi @shawnricecake, LLM-Pruner is a structural method and thus produces a dense model after pruning. Hi, thanks for your reply, so, the model weights after merge the lora weights...