Add vLLM v0.18.1 source tree with KV transfer abort fix

third_party/vllm/ now tracked in git for direct patch management.
Based on vLLM v0.18.1 release with one patch applied:

  vllm/v1/core/sched/scheduler.py:
    Replace fatal assert with graceful skip when KV transfer callback
    arrives for an already-aborted request during PD disaggregated serving.

Future vLLM modifications should be made directly in third_party/vllm/
and committed normally. The patches/ directory is kept as documentation
of what changed from upstream.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

third_party/vllm/examples/pooling/embed/template/vlm2vec_phi3v.jinja

@@ -0,0 +1,16 @@
+{%- if messages | length > 1 -%}
+    {{ raise_exception('Embedding models should only embed one message at a time') }}
+{%- endif -%}
+
+{% set vars = namespace(parts=[], next_image_id=1) %}
+{%- for message in messages -%}
+    {%- for content in message['content'] -%}
+        {%- if content['type'] == 'text' -%}
+            {%- set vars.parts = vars.parts + [content['text']] %}
+        {%- elif content['type'] == 'image' -%}
+            {%- set vars.parts = vars.parts + ['<|image_{i:d}|>'.format(i=vars.next_image_id)] %}
+            {%- set vars.next_image_id = vars.next_image_id + 1 %}
+        {%- endif -%}
+    {%- endfor -%}
+{%- endfor -%}
+{{ vars.parts | join(' ') }}