@danielhanchen Hi Daniel, thanks for your work!

having an error just like in the issue #275, but this time while trying to save tuned version of unsloth/gemma-2-9b-it-bnb-4bit.

model.save_pretrained_gguf("gemma_gguf_4bit", tokenizer, quantization_method = "q4_k_m")

error:

Traceback (most recent call last):
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 3668, in
main()
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 3661, in main
model_instance.write()
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 400, in write
self.prepare_metadata(vocab_only=False)
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 393, in prepare_metadata
self.set_vocab()
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 2514, in set_vocab
self._set_vocab_sentencepiece()
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 683, in _set_vocab_sentencepiece
tokens, scores, toktypes = self._create_vocab_sentencepiece()
File "/content/gemma_gguf_4bit/llama.cpp/convert_hf_to_gguf.py", line 700, in _create_vocab_sentencepiece
raise FileNotFoundError(f"File not found: {tokenizer_path}")
FileNotFoundError: File not found: gemma_gguf_4bit/tokenizer.model

RuntimeError Traceback (most recent call last)
in <cell line: 1>()
----> 1 model.save_pretrained_gguf("gemma_gguf_4bit", tokenizer, quantization_method = "q4_k_m")

1 frames
/usr/local/lib/python3.10/dist-packages/unsloth/save.py in save_to_gguf(model_type, model_dtype, is_sentencepiece, model_directory, quantization_method, first_conversion, _run_installer)
1110 )
1111 else:
-> 1112 raise RuntimeError(
1113 f"Unsloth: Quantization failed for {final_location}\n"
1114 "You might have to compile llama.cpp yourself, then run this again.\n"\