The model returns constantly </s>
#1
by
simonesartoni1
- opened
The model returns to most of the answers. I have tried the GPTQ quantized version and it works, while this one does not. Could there have been an error in the quantization process or in respecting the templates?