mtmd, model : merge HunyuanOCR into HunyuanVL and fix OCR vision precision (#23329)

- HunyuanOCR shares the same HF arch and vision layout as HunyuanVL butwas split into a separate path that skipped the +0.1 bilinear sampler used by the HF reference.
- Collapse OCR into the HUNYUANVL projector + HUNYUAN_VL text arch

tools/mtmd/clip-impl.h

@@ -170,7 +170,7 @@
 #define TN_TOK_BOI         "v.boi"
 #define TN_TOK_EOI         "v.eoi"
 
-// hunyuanocr / hunyuanvl (shared GGUF tensor names)
+// hunyuanvl (shared GGUF tensor names)
 #define TN_MM_PRE_NORM     "mm.pre_norm.%s"
 #define TN_TOK_IMG_BEGIN   "mm.image_begin"
 #define TN_TOK_IMG_END     "mm.image_end"
@@ -343,7 +343,6 @@ enum projector_type {
     PROJECTOR_TYPE_YASA2,
     PROJECTOR_TYPE_KIMIK25,
     PROJECTOR_TYPE_NEMOTRON_V2_VL,
-    PROJECTOR_TYPE_HUNYUANOCR,
     PROJECTOR_TYPE_HUNYUANVL,
     PROJECTOR_TYPE_MINICPMV4_6,
     PROJECTOR_TYPE_GRANITE_SPEECH,
@@ -393,7 +392,6 @@ static std::map<projector_type, std::string> PROJECTOR_TYPE_NAMES = {
     { PROJECTOR_TYPE_YASA2,     "yasa2"},
     { PROJECTOR_TYPE_KIMIK25,   "kimik25"},
     { PROJECTOR_TYPE_NEMOTRON_V2_VL, "nemotron_v2_vl"},
-    { PROJECTOR_TYPE_HUNYUANOCR, "hunyuanocr"},
     { PROJECTOR_TYPE_HUNYUANVL,  "hunyuanvl"},
     { PROJECTOR_TYPE_MINICPMV4_6, "minicpmv4_6"},
     { PROJECTOR_TYPE_GRANITE_SPEECH, "granite_speech"},