Issue in resuming finetuning Llama 3.1 Instruct Model #36035
Description
When I run the below script, I am not able to resume it.
trainer.train(resume_from_checkpoint = "/content/drive/MyDrive/001_projects/exigent/Contract Management and Extraction (CME)/main_data/001_CML_3/CMS_models/phase_6/llama_models/tuning_Llama_3-1_8B_instruct_phase_6_exp_with_sys_prompt/checkpoint-21450/") # tuning_Llama_3-1_8B_instruct_phase_6_exp_with_sys_prompt
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
[<ipython-input-61-9126eb228050>](https://localhost:8080/#) in <cell line: 0>()
----> 1 trainer.train(resume_from_checkpoint = "/content/drive/MyDrive/001_projects/exigent/Contract Management and Extraction (CME)/main_data/001_CML_3/CMS_models/phase_6/llama_models/tuning_Llama_3-1_8B_instruct_phase_6_exp_with_sys_prompt/checkpoint-21450/") # tuning_Llama_3-1_8B_instruct_phase_6_exp_with_sys_prompt
2 frames
[/usr/local/lib/python3.11/dist-packages/trl/trainer/sft_trainer.py](https://localhost:8080/#) in train(self, *args, **kwargs)
449 self.model = self._trl_activate_neftune(self.model)
450
--> 451 output = super().train(*args, **kwargs)
452
453 # After training we make sure to retrieve back the original forward pass method
[/usr/local/lib/python3.11/dist-packages/transformers/trainer.py](https://localhost:8080/#) in train(self, resume_from_checkpoint, trial, ignore_keys_for_eval, **kwargs)
2134 if resume_from_checkpoint is not None:
2135 if not is_sagemaker_mp_enabled() and not self.is_deepspeed_enabled and not self.is_fsdp_enabled:
-> 2136 self._load_from_checkpoint(resume_from_checkpoint)
2137 # In case of repeating the find_executable_batch_size, set `self._train_batch_size` properly
2138 state = TrainerState.load_from_json(os.path.join(resume_from_checkpoint, TRAINER_STATE_NAME))
[/usr/local/lib/python3.11/dist-packages/transformers/trainer.py](https://localhost:8080/#) in _load_from_checkpoint(self, resume_from_checkpoint, model)
2842 if hasattr(model, "active_adapters"):
2843 active_adapters = model.active_adapters
-> 2844 if len(active_adapters) > 1:
2845 logger.warning("Multiple active adapters detected will only consider the first adapter")
2846 active_adapter = active_adapters[0]
TypeError: object of type 'method' has no len()