fix gguf-py: Conversion error when multiple licenses are configured

[mmngays]

In the write_kv_data_to_file function of GGUFWriter, when there are multiple licenses, they were being passed as a list, which was causing an error. Therefore, I have joined them into a string.

general.license ['llama3.1', 'gemma']

Traceback (most recent call last):
  File "/home/mmnga/llm-data/./llama.cpp/convert_hf_to_gguf.py", line 4433, in <module>
    main()
  File "/home/mmnga/llm-data/./llama.cpp/convert_hf_to_gguf.py", line 4427, in main
    model_instance.write()
  File "/home/mmnga/llm-data/./llama.cpp/convert_hf_to_gguf.py", line 436, in write
    self.gguf_writer.write_kv_data_to_file()
  File "/home/mmnga/llm-data/llama.cpp/gguf-py/gguf/gguf_writer.py", line 243, in write_kv_data_to_file
    kv_bytes += self._pack_val(val.value, val.type, add_vtype=True)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/mmnga/llm-data/llama.cpp/gguf-py/gguf/gguf_writer.py", line 882, in _pack_val
    kv_data += encoded_val
TypeError: can't concat list to bytearray

I have confirmed this error in this model.
license is llama3.1 and gemma.
https://huggingface.co/tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.1

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

[mmngays]

Sorry for the delay in making the corrections. As you pointed out, I moved it to the location where add_license in metadata.py is being called.