Quantize mistral-7b-instruct-v0.2 with llamafile-0.8 BF16

mistral-7b-instruct-v0.2.BF16.llamafile

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ee7927b3a15d8758692b831b464fc876c837965f3b2153d2ec3b77718cb84986
-size 14509085167
+oid sha256:4c60e4778c07fdd1cded0e4d7cba956e5e17831c06e0c24da81d81a70ebb8097
+size 14511837925