feat: switch models — ollama qwen3.5:4b-fast, vllm numind/NuExtract3

- Ollama: qwen3.5:4b-fast with NUM_PARALLEL=8, num_ctx=16384 (~6GB VRAM)
- vLLM: numind/NuExtract3 at 192.168.42.254:2701, gpu-memory-util=0.45
- All 3 pipelines (beta/paper/live) updated with correct endpoints
- Ollama proxy at 10.1.1.12:2701 (K8s ollama-metrics)
- vLLM proxy at 192.168.42.254:2701 (NixOS nginx)
This commit is contained in:
Celes Renata
2026-07-03 11:15:28 +00:00
parent d1533ed31d
commit 3996ec6ef5
3 changed files with 9 additions and 6 deletions
+3 -3
View File
@@ -40,9 +40,9 @@ config:
BROKER_MODE: "paper" BROKER_MODE: "paper"
BROKER_PROVIDER: "alpaca" BROKER_PROVIDER: "alpaca"
OLLAMA_BASE_URL: "http://10.1.1.12:2701" OLLAMA_BASE_URL: "http://10.1.1.12:2701"
OLLAMA_MODEL: "qwen3.5:9b-fast" OLLAMA_MODEL: "qwen3.5:4b-fast"
VLLM_BASE_URL: "http://10.1.1.12:2701" VLLM_BASE_URL: "http://192.168.42.254:2701"
VLLM_MODEL: "AxionML/Qwen3.5-9B-NVFP4" VLLM_MODEL: "numind/NuExtract3"
VLLM_TIMEOUT: "120" VLLM_TIMEOUT: "120"
VLLM_MAX_RETRIES: "2" VLLM_MAX_RETRIES: "2"
VLLM_TEMPERATURE: "0.7" VLLM_TEMPERATURE: "0.7"
@@ -17,6 +17,9 @@ config:
DEPLOY_STAGE: "paper" DEPLOY_STAGE: "paper"
POSTGRES_USER: "stonks_paper" POSTGRES_USER: "stonks_paper"
OLLAMA_BASE_URL: "http://10.1.1.12:2701" OLLAMA_BASE_URL: "http://10.1.1.12:2701"
OLLAMA_MODEL: "qwen3.5:4b-fast"
VLLM_BASE_URL: "http://192.168.42.254:2701"
VLLM_MODEL: "numind/NuExtract3"
MARKET_DATA_BASE_URL: "https://api.polygon.io" MARKET_DATA_BASE_URL: "https://api.polygon.io"
## Secrets — all credentials for the paper stage ## Secrets — all credentials for the paper stage
+3 -3
View File
@@ -186,14 +186,14 @@ config:
MINIO_ENDPOINT: "minio.minio-service.svc.cluster.local:80" MINIO_ENDPOINT: "minio.minio-service.svc.cluster.local:80"
MINIO_SECURE: "false" MINIO_SECURE: "false"
OLLAMA_BASE_URL: "http://10.1.1.12:2701" OLLAMA_BASE_URL: "http://10.1.1.12:2701"
OLLAMA_MODEL: "qwen3.5:9b-fast" OLLAMA_MODEL: "qwen3.5:4b-fast"
OLLAMA_TIMEOUT: "240" OLLAMA_TIMEOUT: "240"
OLLAMA_MAX_RETRIES: "2" OLLAMA_MAX_RETRIES: "2"
OLLAMA_RETRY_BASE_DELAY: "1.0" OLLAMA_RETRY_BASE_DELAY: "1.0"
OLLAMA_RETRY_MAX_DELAY: "10.0" OLLAMA_RETRY_MAX_DELAY: "10.0"
OLLAMA_RETRY_BACKOFF_MULTIPLIER: "2.0" OLLAMA_RETRY_BACKOFF_MULTIPLIER: "2.0"
VLLM_BASE_URL: "http://10.1.1.12:2701" VLLM_BASE_URL: "http://192.168.42.254:2701"
VLLM_MODEL: "qwen3.5:9b-fast" VLLM_MODEL: "numind/NuExtract3"
VLLM_TIMEOUT: "120" VLLM_TIMEOUT: "120"
VLLM_MAX_RETRIES: "2" VLLM_MAX_RETRIES: "2"
VLLM_TEMPERATURE: "0.7" VLLM_TEMPERATURE: "0.7"