Issue while loading file-tuned gemma2

[amanpreetsingh459]

Getting this warning while loading fine-tuned Gemma2-2b model. And getting weird outputs too when go with this
Some weights of the model checkpoint at <model_dir> were not used when initializing Gemma2ForCausalLM

[John6666]

In the HF library, this is a message that often appears in both image generation AI and LLM, but the fact that it is only a warning means that the warning itself probably does not interfere with operation.
I think the cause of the strange output is something else.

Unable to load checkpoint after finetuning

Hi, So I’m running into trouble when trying to load the model weights after finetuning a pretrained base model, steps I’m taking: model= BertForSequenceClassification.from_pretrained("model/repo") model.classifier = torch.nn.Sequential( torch.nn.Dropout(0.4), torch.nn.Linear(model.config.hidden_size, 256), torch.nn.ReLU(), torch.nn.Dropout(0.4), torch.nn.Linear(256, 2), ) getting a warning here which I guess is fine for the moment: Some weights of the model checkpoint at…

[Sidakdeep22]

@amanpreetsingh459 I’m facing the same issue. How did you resolve it?

[amanpreetsingh459]

I resolved it by following below steps:-

1. push adapter weights(from trainer)
   trainer.push_to_hub(“your_huggingface_dir”)

2. push the base model as well
   base_model.push_to_hub(“your_huggingface_dir”)

3. then loading the model with:
   model_finetuned = AutoModelForCausalLM.from_pretrained(
   finetuned_model_name,
   device_map=“auto”,
   torch_dtype=torch.bfloat16
   )