[BUG] Bug when apply tokenizer

[devlancer412]

Prerequisites

• I have read the documentation.
• I have checked other issues for similar problems.

Backend

Local

Interface Used

CLI

CLI Command

params = LLMTrainingParams(
model=model,
data_path="HuggingFaceFW/fineweb-edu-score-2",
chat_template="tokenizer",
text_column="text",
train_split=f"{sub_dataset}:train",
trainer="sft",
epochs=3,
lr=1e-5,
peft=True,
quantization="int4",
target_modules=target_modules,
padding="right",
optimizer="adamw_torch",
scheduler="cosine",
gradient_accumulation=8,
mixed_precision="bf16",
merge_adapter=True,
project_name=model_name,
log="tensorboard",
push_to_hub=True,
username=hf_username,
token=hf_token,
model_max_length=4096,
auto_find_batch_size=True,
)

backend = "local"
project = AutoTrainProject(params=params, backend=backend, process=True)
project.create()

UI Screenshots & Parameters

No response

Error Logs

File "/workspace/global_venv/lib/python3.11/site-packages/autotrain/trainers/clm/train_clm_sft.py", line 16, in train
train_data, valid_data = utils.process_data_with_chat_template(config, tokenizer, train_data, valid_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/autotrain/trainers/clm/utils.py", line 639, in process_data_with_chat_template
train_data = train_data.map(
^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/datasets/arrow_dataset.py", line 560, in wrapper
out: Union["Dataset", "DatasetDict"] = func(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/datasets/arrow_dataset.py", line 3055, in map
for rank, done, content in Dataset._map_single(**dataset_kwargs):
File "/workspace/global_venv/lib/python3.11/site-packages/datasets/arrow_dataset.py", line 3428, in _map_single
example = apply_function_on_filtered_inputs(example, i, offset=offset)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/datasets/arrow_dataset.py", line 3320, in apply_function_on_filtered_inputs
processed_inputs = function(*fn_args, *additional_args, **fn_kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/autotrain/trainers/clm/utils.py", line 346, in apply_chat_template
example[config.text_column] = tokenizer.apply_chat_template(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 1871, in apply_chat_template
rendered_chat = compiled_template.render(
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/workspace/global_venv/lib/python3.11/site-packages/jinja2/environment.py", line 1304, in render
self.environment.handle_exception()
File "/workspace/global_venv/lib/python3.11/site-packages/jinja2/environment.py", line 939, in handle_exception
raise rewrite_traceback_stack(source=source)
File "", line 2, in top-level template code
jinja2.exceptions.UndefinedError: 'str object' has no attribute 'role'

ERROR | 2024-12-13 15:54:50 | autotrain.trainers.common:wrapper:216 - 'str object' has no attribute 'role'

Additional Information

I tried to finetune model with fineweb-2 dataset. but got this error and not able to fix it with parameters.
Please help me