Error Converting Hugging Face Llama 3 finetuned Model to GUFF

[MontassarTn]

• I encountered an issue while attempting to convert my Hugging Face Llama 3 model to GUFF format using the provided command:

!python /content/llama.cpp/convert.py vicuna-hf \
  --outfile llama3_ocr_to_xml_A1.gguf \
  --outtype q8_0

• Error Message:
  Loading model file vicuna-hf/model-00001-of-00004.safetensors
  Loading model file vicuna-hf/model-00001-of-00004.safetensors
  Loading model file vicuna-hf/model-00002-of-00004.safetensors
  Loading model file vicuna-hf/model-00003-of-00004.safetensors
  Loading model file vicuna-hf/model-00004-of-00004.safetensors
  params = Params(n_vocab=128256, n_embd=4096, n_layer=32, n_ctx=8192, n_ff=14336, n_head=32, n_head_kv=8, n_experts=None, n_experts_used=None, f_norm_eps=1e-05, rope_scaling_type=None, f_rope_freq_base=500000.0, f_rope_scale=None, n_orig_ctx=None, rope_finetuned=None, ftype=<GGMLFileType.MostlyF16: 1>, path_model=PosixPath('vicuna-hf'))
  Traceback (most recent call last):
  File "/content/llama.cpp/convert.py", line 1548, in
  main()
  File "/content/llama.cpp/convert.py", line 1515, in main
  vocab, special_vocab = vocab_factory.load_vocab(vocab_types, model_parent_path)
  File "/content/llama.cpp/convert.py", line 1417, in load_vocab
  vocab = self._create_vocab_by_path(vocab_types)
  File "/content/llama.cpp/convert.py", line 1407, in _create_vocab_by_path
  raise FileNotFoundError(f"Could not find a tokenizer matching any of {vocab_types}")
  FileNotFoundError: Could not find a tokenizer matching any of ['spm', 'hfft']

[LostRuins]

I believe #6745 needs to be merged first.

[MontassarTn]

run with "--vocab-type bpe"

[MontassarTn]

I add "--vocab-type bpe" but when I try to use model on vsCode the Kernel crashed and on LM Studio Error loading model
show up.

[MrVolts]

Same issue here, I have not solved it

[phymbert]

Please wait for:

• Support Llama 3 conversion #6745
• Added llama-3 chat template #6751