fix: set gpus as strings in config files (#1123)

examples/llm/configs/agg.yaml

@@ -34,7 +34,7 @@ VllmWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 1
+      gpu: '1'
   common-configs: [model, block-size, max-model-len]
 
 Planner:

examples/llm/configs/agg_router.yaml

@@ -39,7 +39,7 @@ VllmWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 1
+      gpu: '1'
   common-configs: [model, block-size, max-model-len, router, kv-transfer-config]
 
 Planner:

examples/llm/configs/disagg_router.yaml

@@ -42,7 +42,7 @@ VllmWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 1
+      gpu: '1'
   common-configs: [model, block-size, max-model-len, router, kv-transfer-config]
 
 PrefillWorker:
@@ -50,7 +50,7 @@ PrefillWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 1
+      gpu: '1'
   common-configs: [model, block-size, max-model-len, kv-transfer-config]
 
 Planner:

examples/llm/configs/multinode-405b.yaml

@@ -49,7 +49,7 @@ VllmWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 8
+      gpu: '8'
 
 PrefillWorker:
   model: nvidia/Llama-3.1-405B-Instruct-FP8
@@ -63,4 +63,4 @@ PrefillWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 8
+      gpu: '8'
\ No newline at end of file

examples/llm/configs/multinode_agg_r1.yaml

@@ -35,5 +35,5 @@ VllmWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 1
+      gpu: '1'
   common-configs: [model, block-size, max-model-len]

examples/llm/configs/mutinode_disagg_r1.yaml

@@ -34,7 +34,7 @@ VllmWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 16
+      gpu: '16'
   common-configs: [model, block-size, max-model-len, kv-transfer-config, tensor-parallel-size]
 
 PrefillWorker:
@@ -42,5 +42,5 @@ PrefillWorker:
   ServiceArgs:
     workers: 1
     resources:
-      gpu: 16
+      gpu: '16'
   common-configs: [model, block-size, max-model-len, kv-transfer-config, tensor-parallel-size]