Upload Phi-3-mini-128k-instruct.Q4_K_M.gguf

via QuantFactory

It is the only Phi-3-mini model I've found that since release that:
- naturally ends an assistant turn without heavy prompting
- generates non-gibberish with over 2K input tokens
- works naturally well with RAG and existing flows

.gitattributes

@@ -42,3 +42,4 @@ llava-v1.6-mistral-7b.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
 tinymistral-248m-sft-v4.q2_k.gguf filter=lfs diff=lfs merge=lfs -text
 stablelm-zephyr-3b.Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
 Hermes-2-Pro-Mistral-7B.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
+Phi-3-mini-128k-instruct.Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text

Phi-3-mini-128k-instruct.Q4_K_M.gguf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b27c1a245243b3eadf6db453ddefd419a31e388820824beeba1c60eee17d05e
+size 2393231712