Quantize F16 with llamafile-0.8.5

mistral-7b-instruct-v0.2.F16.llamafile

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4ae69b232fa1576a88527aa4306e73f0d63f6dbbd824993b7d8c5301933affb7
-size 14514075819
+oid sha256:7d00c898a6ada381a14ebdfa5452b81491136af805fd44607926a5f5bb06e905
+size 14520236124