From 2200006c419dd9f8fb1ec1ff61bc9f301f1d142f Mon Sep 17 00:00:00 2001 From: sirius0xdev Date: Wed, 6 May 2026 01:31:45 +0000 Subject: [PATCH] fix vllm error --- apps/base/customer1/hermes-agent/rays-deployment.yaml | 2 +- infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/apps/base/customer1/hermes-agent/rays-deployment.yaml b/apps/base/customer1/hermes-agent/rays-deployment.yaml index 8900fbe..f67b58f 100644 --- a/apps/base/customer1/hermes-agent/rays-deployment.yaml +++ b/apps/base/customer1/hermes-agent/rays-deployment.yaml @@ -42,7 +42,7 @@ spec: key: XAI_API_KEY - name: TELEGRAM_ALLOWED_USERS - value: "" + value: "7105451284" # === Local vLLM (OpenAI-compatible) === - name: OPENAI_BASE_URL diff --git a/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml b/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml index b032cbd..addcb57 100644 --- a/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml +++ b/infrastructure/gpus/base/vllm-servers/rtx6000-vllm.yaml @@ -40,7 +40,7 @@ spec: - --gpu-memory-utilization=0.98 - --max-model-len=131072 - --enable-auto-tool-choice - - --kv-cache=dtype=fp8 + - --kv-cache-dtype=fp8 - --max-num-batched-tokens=32768 - --max-num-seqs=16 - --trust-remote-code