abhishek thakur

yes, hopefully!

seems like a transformer issue. could you merge the model and then try the same code? here is code for merging: ``` def merge_adapter(base_model_path, target_model_path, adapter_path): logger.info("Loading adapter...") model =...

can you remove the old adapter files when you reload?

``` model = AutoModelForCausalLM.from_pretrained( base_model_path, torch_dtype=torch.float16, low_cpu_mem_usage=True, trust_remote_code=True, device_map="auto", ) ```

glad to hear it worked. is there still an issue?

its a bit difficult to answer that question without looking at the training data and the training parameters used. does it happen with all datasets?

hi, if you have more upcoming fixes like this, please add them in this pr

lets have a talk about this before we start so that we are all on the same page. :)

please see this example and convert your code to new Model class and get rid of the warning: https://github.com/abhishekkrthakur/tez/blob/main/examples/image/digit_recognizer.py :)

fwiw, i ended up reverting my username :D