- Did you update?
pip install --upgrade unsloth unsloth_zoo Yes
Colab or Kaggle or local / cloud Local- Number GPUs used, use
nvidia-smi 1
- Which notebook? Please link! None
- Which Unsloth version, TRL version, transformers version, PyTorch version?
unsloth: 2025.7.9
transofmers: 4.53.3
torch: 2.7.1
- Which trainer?
SFTTrainer, GRPOTrainer etc SFTTrainer
I get this error:
TypeError: Unsloth: quantization_method can only be a string or a list of strings
It is happening here in save.py:
if quantization_method is not None:
# Convert quantization_method to list
if isinstance(quantization_method, list): pass
elif isinstance(quantization_method, str): quantization_method = [ quantization_method, ]
elif isinstance(quantization_method, tuple): quantization_method = list(quantization_method)
else:
raise TypeError("Unsloth: quantization_method can only be a string or a list of strings")
passHowever my quantization_method is an object:
quantization_method = PixtralProcessor:
- image_processor: PixtralImageProcessorFast
My code is:
# Load using Unsloth
model, tokenizer = FastLanguageModel.from_pretrained(
model_name="outputs/checkpoint-120",
max_seq_length=2048,
dtype=torch.float16,
load_in_4bit=True,
)
# Push to hub
model.save_pretrained_gguf(
"Mistral-Small-3.2",
tokenizer,)I use unsloth/Mistral-Small-3.2-24B-Instruct-2506
pip install --upgrade unsloth unsloth_zooYesColaborKaggleor local / cloud Localnvidia-smi1unsloth: 2025.7.9
transofmers: 4.53.3
torch: 2.7.1
SFTTrainer,GRPOTraineretc SFTTrainerI get this error:
TypeError: Unsloth: quantization_method can only be a string or a list of strings
It is happening here in save.py:
However my quantization_method is an object:
quantization_method = PixtralProcessor:
- image_processor: PixtralImageProcessorFast
My code is:
I use unsloth/Mistral-Small-3.2-24B-Instruct-2506