Merge branch 'main' of https://github.com/stanford-crfm/benchmarking …

…into testvlm
stanford-crfm · May 5, 2024 · 6f4b7a5 · 6f4b7a5
2 parents edce2dc + 1ce9e82
commit 6f4b7a5
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 1 deletion.
diff --git a/helm-frontend/project_metadata.json b/helm-frontend/project_metadata.json
@@ -29,6 +29,12 @@
 		"id": "mmlu",
 		"releases": ["v1.1.0", "v1.0.0"]
 	},
+	{
+		"title": "VHELM",
+		"description": "Holistic Evaluation of Vision-Language Models",
+		"id": "vhelm",
+		"releases": ["v1.0.0"]
+	},
 	{
 		"title": "All Leaderboards",
 		"description": "Home page for all HELM leaderboards",

diff --git a/helm-frontend/src/components/VHELMLanding.tsx b/helm-frontend/src/components/VHELMLanding.tsx
@@ -33,7 +33,7 @@ export default function VHELMLanding() {
         <em>Holistic Evaluation of Vision-Language Models (VHELM)</em> by
         extending the <a href="https://arxiv.org/abs/2211.09110">HELM</a>{" "}
         framework with the necessary adaptation methods to assess the
-        performance of 11 prominent VLMs on 3 standard VLM benchmarks.
+        performance of 6 prominent VLMs on 3 standard VLM benchmarks.
       </p>
       <p className="my-4 font-bold">
         This is ongoing work to achieve holistic evaluation for vision-language