From 95d0b38d0acc1d65ddbfb3d5af4e8bace97950f4 Mon Sep 17 00:00:00 2001 From: Vuong Hoang Date: Sat, 27 Jun 2026 08:33:43 -0700 Subject: [PATCH] litellm: canonical defaults for granite + GLM (completes fleet sweep) - granite-4.1-8b (+ summarizer/classifier): temperature 0 (IBM vendor-canonical "temp 0 for inferencing"; top_p/top_k no-ops at temp 0, omitted). Deterministic baseline for summ/classify; creative callers override. - GLM family (z.ai cloud): temperature + top_p 0.95 only (the ONLY params z.ai chat API accepts per its OpenAPI schema; top_k/min_p/penalties absent -> not set). temp 1.0 for glm-5.1/5.2/5-turbo/4.7 + gen-frontier; temp 0.6 for glm-4.5-air. Matches z.ai API defaults -> explicit-over-implicit, future-proofs vs vendor drift. Round-2 dvalin-researched (provenance-labeled), verified live, granite+glm smoked 200. Embeddings/rerankers excluded (no sampling). Fleet-wide canonical-defaults sweep complete. --- stacks/litellm/conf/config.yaml | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/stacks/litellm/conf/config.yaml b/stacks/litellm/conf/config.yaml index d1d85c8..c5e0888 100644 --- a/stacks/litellm/conf/config.yaml +++ b/stacks/litellm/conf/config.yaml @@ -27,6 +27,7 @@ model_list: model: hosted_vllm/granite-4.1-8b api_base: http://10.250.50.54:8004/v1 api_key: os.environ/VLLM_API_KEY + temperature: 0 model_info: mode: chat @@ -38,6 +39,7 @@ model_list: model: hosted_vllm/granite-4.1-8b api_base: http://10.250.50.54:8004/v1 api_key: os.environ/VLLM_API_KEY + temperature: 0 model_info: mode: chat @@ -48,6 +50,7 @@ model_list: model: hosted_vllm/granite-4.1-8b api_base: http://10.250.50.54:8004/v1 api_key: os.environ/VLLM_API_KEY + temperature: 0 model_info: mode: chat @@ -272,6 +275,8 @@ model_list: model: openai/glm-5.1 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 extra_body: thinking: type: disabled @@ -281,6 +286,8 @@ model_list: model: openai/glm-5.1 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 extra_body: thinking: type: enabled @@ -293,6 +300,8 @@ model_list: model: openai/glm-5.2 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 extra_body: thinking: type: disabled @@ -301,6 +310,8 @@ model_list: model: openai/glm-5.2 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 extra_body: thinking: type: enabled @@ -315,6 +326,8 @@ model_list: model: openai/glm-5.2 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 extra_body: thinking: type: disabled @@ -323,6 +336,8 @@ model_list: model: openai/glm-5.2 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 extra_body: thinking: type: enabled @@ -331,16 +346,22 @@ model_list: model: openai/glm-5-turbo api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 - model_name: glm-4.7 litellm_params: model: openai/glm-4.7 api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 1.0 + top_p: 0.95 - model_name: glm-4.5-air litellm_params: model: openai/glm-4.5-air api_base: https://api.z.ai/api/coding/paas/v4 api_key: os.environ/Z_AI_API_KEY + temperature: 0.6 + top_p: 0.95 # --- (removed 2026-06-20, operator call) the `*` wildcard → llama-swap # (ana-ml2:9292). llama-swap is decommissioned (:9292 confirmed down), so