Hopefully there will be an official model for the gptq method.
#6 opened 2 days ago
by
warlock-edward
Amazing, the only local model which repaired code in my test.
#5 opened 3 days ago
by
krustik
fp8
#3 opened 9 days ago
by
erichartford
Can you publish evaluated results on benchmarks?
#2 opened 15 days ago
by
barty