llama : suppress conversion from 'size_t' to 'int'

This commit updates llm_tokenizer_spm.tokenize to suppress/remove the
following warnings that are generated on Windows when using MSVC:

```console
src\llama-vocab.cpp(211,1): warning C4267: 'argument':
    conversion from 'size_t' to 'int', possible loss of data
src\llama-vocab.cpp(517,1): warning C4267: 'argument':
    conversion from 'size_t' to 'int', possible loss of data
```

This is done by adding a cast for the size_t returned from
symbols.size(). I believe this is safe as it seems unlikely that
symbols, which stores an entry for each UTF8 character, would become
larger than INT_MAX.

The motivation for this change is to reduce the number of warnings that
are currently generated when building on Windows.

src/llama-vocab.cpp

@@ -207,7 +207,8 @@ struct llm_tokenizer_spm {
         }
 
         // seed the work queue with all possible 2-character tokens.
-        for (size_t i = 1; i < symbols.size(); ++i) {
+        int symbols_size = static_cast<int>(symbols.size());
+        for (int i = 1; i < symbols_size; ++i) {
             try_add_bigram(i - 1, i);
         }
 
@@ -511,7 +512,8 @@ struct llm_tokenizer_bpe {
                 index++;
                 symbols.emplace_back(sym);
             }
-            for (size_t i = 1; i < symbols.size(); ++i) {
+            int symbols_size = static_cast<int>(symbols.size());
+            for (int i = 1; i < symbols_size; ++i) {
                 add_new_bigram(i - 1, i);
             }