Open zlh1992 opened 1 year ago
confusion with the setting:
parser_training.add_argument("--groupsize", type=int, default=-1, help="Groupsize of v2 model") parser_training.add_argument("--v1", action="store_true", help="Use V1 model")
what is the difference between v1 model and v2 model? and where is the different weights of them in huggingface models?
v1 model is the model without group size (the first type of gptq model)
confusion with the setting:
V2 model support
parser_training.add_argument("--groupsize", type=int, default=-1, help="Groupsize of v2 model") parser_training.add_argument("--v1", action="store_true", help="Use V1 model")
what is the difference between v1 model and v2 model? and where is the different weights of them in huggingface models?