Proper error when using only model-id self-hosted

2 files changed, 53 insertions, 0 deletions

diff --git a/config-model/src/test/cfg/application/embed_cloud_only/configdefinitions/embedding.bert-base-embedder.def b/config-model/src/test/cfg/application/embed_cloud_only/configdefinitions/embedding.bert-base-embedder.def
new file mode 100644
index 00000000000..144dfbd0001
--- /dev/null
+++ b/config-model/src/test/cfg/application/embed_cloud_only/configdefinitions/embedding.bert-base-embedder.def
@@ -0,0 +1,30 @@
+# Copy of this Vespa config stored here because Vespa config definitions are not
+# available in unit tests, and are needed (by DomConfigPayloadBuilder.parseLeaf)
+# Alternatively, we could make that not need it as it is not strictly necessaery.
+
+namespace=embedding
+
+# Wordpiece tokenizer
+tokenizerVocab model
+
+transformerModel  model
+
+# Max length of token sequence model can handle
+transformerMaxTokens int default=384
+
+# Pooling strategy
+poolingStrategy enum { cls, mean } default=mean
+
+# Input names
+transformerInputIds      string default=input_ids
+transformerAttentionMask string default=attention_mask
+transformerTokenTypeIds  string default=token_type_ids
+
+# Output name
+transformerOutput string default=output_0
+
+# Settings for ONNX model evaluation
+onnxExecutionMode enum { parallel, sequential } default=sequential
+onnxInterOpThreads int default=1
+onnxIntraOpThreads int default=-4  # n=number of threads -> n<0: CPUs/(-n), n==0: CPUs, n>0: n
+
diff --git a/config-model/src/test/cfg/application/embed_cloud_only/services.xml b/config-model/src/test/cfg/application/embed_cloud_only/services.xml
new file mode 100644
index 00000000000..57db4f5bfae
--- /dev/null
+++ b/config-model/src/test/cfg/application/embed_cloud_only/services.xml
@@ -0,0 +1,23 @@
+<?xml version="1.0" encoding="utf-8" ?>
+<!-- Copyright Yahoo. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root. -->
+<services version="1.0">
+
+  <container version="1.0">
+
+    <component id="transformer" class="ai.vespa.embedding.BertBaseEmbedder" bundle="model-integration">
+      <config name="embedding.bert-base-embedder">
+        <!-- No fallback to url or path when deploying outside cloud -->
+        <transformerModel model-id="minilm-l6-v2"/>
+        <tokenizerVocab path="files/vocab.txt"/>
+
+        <!-- tunable parameters: number of threads etc -->
+        <onnxIntraOpThreads>4</onnxIntraOpThreads>
+      </config>
+    </component>
+
+    <nodes>
+      <node hostalias="node1" />
+    </nodes>
+  </container>
+
+</services>