Merge commit '0fac87b157305eb82a70902327abffbbce25bd3e' into concedo_experimental

# Conflicts:
#	.github/workflows/build-android.yml
#	.github/workflows/hip-quality-check.yml
#	docs/multimodal.md
#	scripts/hip/gcn-cdna-vgpr-check.py
#	scripts/snapdragon/windows/run-bench.ps1
#	scripts/snapdragon/windows/run-cli.ps1
#	scripts/snapdragon/windows/run-tool.ps1
#	tests/test-backend-ops.cpp
#	tests/test-llama-archs.cpp
#	tools/imatrix/imatrix.cpp
#	tools/mtmd/CMakeLists.txt

src/llama-vocab.cpp

@@ -2188,7 +2188,8 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
             } else if (
                     tokenizer_pre == "qwen2" ||
                     tokenizer_pre == "deepseek-r1-qwen" ||
-                    tokenizer_pre == "kormo") {
+                    tokenizer_pre == "kormo" ||
+                    tokenizer_pre == "f2llmv2") {
                 pre_type = LLAMA_VOCAB_PRE_TYPE_QWEN2;
                 clean_spaces = false;
             } else if (
@@ -2728,6 +2729,7 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
                     || t.first == "[EOS]" // Kimi-K2
                     || t.first == "<|end_of_text|>"
                     || t.first == "<end_of_utterance>" // smoldocling
+                    || t.first == "<｜end▁of▁sentence｜>" // deepseek-ocr
                ) {
                 special_eog_ids.insert(t.second);
                 if ((attr & LLAMA_TOKEN_ATTR_CONTROL) == 0) {