convert : fix gemma v1 tokenizer convert

[ggerganov]

Follow up on #8244

It seems that Gemma v1 tokenization has always been broken due to missing add_space_prefix == false flag. Also, add tokenizer tests for both Gemma and Gemma-2

# get tokenizers
python3 convert-hf-to-gguf-update.py <hf_token>

# generate ggml vocabs and tests
python3 convert-hf-to-gguf.py models/tokenizers/gemma/   --outfile models/ggml-vocab-gemma.gguf   --vocab-only
python3 convert-hf-to-gguf.py models/tokenizers/gemma-2/ --outfile models/ggml-vocab-gemma-2.gguf --vocab-only

# run the tests
make -j tests
./tests/test-tokenizer-0 models/ggml-vocab-gemma.gguf
./tests/test-tokenizer-0 models/ggml-vocab-gemma-2.gguf

• I have read the contributing guidelines
• Self-reported review complexity:
  • Low
  • Medium
  • High