Updated with correct pre-tokenizer

In preparation for full requantization with correct pre-tokenization and gpt2 tokenizer, see the following PRs:
[lama : improve BPE pre-processing + LLaMA 3 and Deepseek support](https://github.com/ggerganov/llama.cpp/pull/6920) - Sets correct pre-tokenizer, applied to new fp16.gguf
[lama3 custom regex split](https://github.com/ggerganov/llama.cpp/pull/6965) - Fixes gpt2 tokenization

.gitattributes

@@ -45,3 +45,4 @@ OpenCodeInterpreter-DS-6.7B.IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text
 OpenCodeInterpreter-DS-6.7B.IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text
 OpenCodeInterpreter-DS-6.7B.imatrix-4096.dat filter=lfs diff=lfs merge=lfs -text
 OpenCodeInterpreter-DS-6.7B.IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text
+OpenCodeInterpreter-DS-6.7B.fp16.gguf filter=lfs diff=lfs merge=lfs -text

OpenCodeInterpreter-DS-6.7B.fp16.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d227d02983842c1025a37b47bbf12f22ac0582f0769b551a7a8d6398c00a1891
+size 13482732384