From f3cbcadfa221363a30089de72d2c428006baad03 Mon Sep 17 00:00:00 2001
From: Zhenzhong1 <zhenzhongxu@zhenzhongxu-vm-u22.habana-labs.com>
Date: Tue, 22 Oct 2024 10:26:44 +0300
Subject: [PATCH] fixed visualqna image issues & tgi params issues

---
 ChatQnA/benchmark/performance/helm_charts/deployment.py | 2 +-
 VisualQnA/benchmark/performance/helm_charts/values.yaml | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/ChatQnA/benchmark/performance/helm_charts/deployment.py b/ChatQnA/benchmark/performance/helm_charts/deployment.py
index 1bee3db2a..bdb84cdde 100644
--- a/ChatQnA/benchmark/performance/helm_charts/deployment.py
+++ b/ChatQnA/benchmark/performance/helm_charts/deployment.py
@@ -69,7 +69,7 @@ def generate_yaml(num_nodes, mode="oob", with_rerank="True"):
         {
             "name": "llm-dependency-deploy",
             "args": [
-                {"name": "--max-input-length", "values": 1280},
+                {"name": "--max-input-length", "value": 1280},
                 {"name": "--max-total-tokens", "value": 2048},
                 {"name": "--max-batch-total-tokens", "value": 35536},
                 {"name": "--max-batch-prefill-tokens", "value": 4096},
diff --git a/VisualQnA/benchmark/performance/helm_charts/values.yaml b/VisualQnA/benchmark/performance/helm_charts/values.yaml
index 503c57fb5..40b41e469 100644
--- a/VisualQnA/benchmark/performance/helm_charts/values.yaml
+++ b/VisualQnA/benchmark/performance/helm_charts/values.yaml
@@ -30,7 +30,7 @@ microservices:
       - containerPort: 8399
 
   - name: visualqna-tgi
-    image: opea/llava-tgi:latest
+    image: ghcr.io/huggingface/tgi-gaudi:2.0.5
     replicas: 1
     ports:
       - containerPort: 8399