From 9f9160985a4f4848fa3f89d83a9f859958bd8e3c Mon Sep 17 00:00:00 2001
From: Bjørn Christian Seime <bjorncs@vespa.ai>
Date: Thu, 4 Apr 2024 11:35:52 +0200
Subject: Add equivalent to `Map.computeIfAbsent()` to simplify typical usage
 of the cache

Current interface requires a lot of boilerplate code.
---
 .../src/main/java/ai/vespa/embedding/ColBertEmbedder.java    | 11 ++---------
 .../ai/vespa/embedding/huggingface/HuggingFaceEmbedder.java  | 12 +-----------
 2 files changed, 3 insertions(+), 20 deletions(-)

(limited to 'model-integration')

diff --git a/model-integration/src/main/java/ai/vespa/embedding/ColBertEmbedder.java b/model-integration/src/main/java/ai/vespa/embedding/ColBertEmbedder.java
index 2f4c0343bf6..a9d6d308df8 100644
--- a/model-integration/src/main/java/ai/vespa/embedding/ColBertEmbedder.java
+++ b/model-integration/src/main/java/ai/vespa/embedding/ColBertEmbedder.java
@@ -196,7 +196,7 @@ public class ColBertEmbedder extends AbstractComponent implements Embedder {
         runtime.sampleEmbeddingLatency((System.nanoTime() - start) / 1_000_000d, context);
         return resultTensor;
     }
-    @SuppressWarnings("unchecked")
+
     protected Tensor embedDocument(String text, Context context, TensorType tensorType) {
         var start = System.nanoTime();
 
@@ -228,15 +228,8 @@ public class ColBertEmbedder extends AbstractComponent implements Embedder {
      * @param hashKey the key to the cached value
      * @return the model output
      */
-    @SuppressWarnings("unchecked")
     protected Map<String, Tensor> evaluateIfNotPresent(Map<String, Tensor> inputs, Context context, String hashKey) {
-        if (context.getCachedValue(hashKey) == null) {
-            Map<String, Tensor> outputs = evaluator.evaluate(inputs);
-            context.putCachedValue(hashKey, outputs);
-            return outputs;
-        } else {
-            return (Map<String, Tensor>) context.getCachedValue(hashKey);
-        }
+        return context.computeCachedValueIfAbsent(hashKey, () -> evaluator.evaluate(inputs));
     }
 
     public static Tensor toFloatTensor(IndexedTensor result, TensorType type, int nTokens) {
diff --git a/model-integration/src/main/java/ai/vespa/embedding/huggingface/HuggingFaceEmbedder.java b/model-integration/src/main/java/ai/vespa/embedding/huggingface/HuggingFaceEmbedder.java
index 08c98fedf3e..1b9f9dd2fe3 100644
--- a/model-integration/src/main/java/ai/vespa/embedding/huggingface/HuggingFaceEmbedder.java
+++ b/model-integration/src/main/java/ai/vespa/embedding/huggingface/HuggingFaceEmbedder.java
@@ -182,22 +182,12 @@ public class HuggingFaceEmbedder extends AbstractComponent implements Embedder {
      * @param hashKey the key to the cached value
      * @return the model output
      */
-    @SuppressWarnings("unchecked")
     protected Map<String, Tensor> evaluateIfNotPresent(Map<String, Tensor> inputs, Context context, String hashKey) {
-        if (context.getCachedValue(hashKey) == null) {
-            Map<String, Tensor> outputs = evaluator.evaluate(inputs);
-            context.putCachedValue(hashKey, outputs);
-            return outputs;
-        } else {
-            return (Map<String, Tensor>) context.getCachedValue(hashKey);
-        }
+        return context.computeCachedValueIfAbsent(hashKey, () -> evaluator.evaluate(inputs));
     }
 
     /**
      * Binary quantization of the embedding into a tensor of type int8 with the specified dimensions.
-     * @param embedding
-     * @param tensorType
-     * @return
      */
     static public Tensor binarize(IndexedTensor embedding, TensorType tensorType) {
         Tensor.Builder builder = Tensor.Builder.of(tensorType);
-- 
cgit v1.2.3