From efaff340a4bbdf12b16d6352a286b335c481f147 Mon Sep 17 00:00:00 2001
From: Hermes Agent service account
 <hermes@carousel-of-progress.local.mk-labs.cloud>
Date: Sat, 15 Aug 2026 00:12:40 -0500
Subject: [PATCH] openviking: fix VLM model alias and lower max_input_tokens to
 1024

- vlm.model was 'llama3.1-8b' which doesn't exist on astro-orbiter's
  /v1/models, causing every summarization call to 400 and endless
  circuit-breaker retries. Correct id: Meta-Llama-3.1-8B-Instruct-Q4_K_M.
- embedding.max_input_tokens=1536 still let chunks through that actually
  tokenized to 2000-2860 real tokens (estimator undercounts vs llama.cpp's
  tokenizer by 1.35x-1.86x on this corpus). Lowered to 1024 for real margin
  under the 2048 n_ctx ceiling.
---
 cluster/platform/openviking/values.yaml | 21 +++++++++++++++++++--
 1 file changed, 19 insertions(+), 2 deletions(-)

diff --git a/cluster/platform/openviking/values.yaml b/cluster/platform/openviking/values.yaml
index 6f89988..3913b37 100644
--- a/cluster/platform/openviking/values.yaml
+++ b/cluster/platform/openviking/values.yaml
@@ -156,7 +156,19 @@ config:
     # estimator is not the same tokenizer llama.cpp uses to count context, so token
     # counts won't match 1:1 between the two. Approved by Ryan as lowest-risk fix
     # (option 1 of 3) vs. touching the astro-orbiter serving stack further.
-    max_input_tokens: 1536
+    #
+    # ROOT CAUSE (2026-08-15 incident): 1536 was still not low enough. Observed
+    # llama.cpp actual n_prompt_tokens vs. OpenViking's own max_input_tokens=1536
+    # estimate ratio ranged 1.35x-1.86x across real ingested chunks (see homelab
+    # re-ingest circuit-breaker errors, e.g. estimate 1536 -> actual 2860 tokens,
+    # 2124, 2088, 2066... all > 2048 n_ctx ceiling). OpenViking's estimator
+    # (likely a chars/4 or similar heuristic) undercounts vs. llama.cpp's real
+    # BPE/wordpiece tokenizer for this corpus's content (dense code/config
+    # snippets tokenize denser than the estimator assumes). Lowering to 1536 alone
+    # does not hold for all chunks; using worst-observed ratio (1.86x) with margin,
+    # 2048 / 1.86 ~= 1100, rounded down further for safety across untested
+    # corpora -> 1024.
+    max_input_tokens: 1024
 
   # ============================================================================
   # VLM / Summarization configuration (L0/L1/L2 generation)
@@ -169,7 +181,12 @@ config:
   vlm:
     api_base: "http://astro-orbiter:8002/v1"
     api_key: "${OPENVIKING_VLM_API_KEY}"  # Placeholder: "local-llama" or similar
-    model: "llama3.1-8b"  # Confirm exact alias from astro-orbiter /v1/models before applying
+    # Fixed 2026-08-15: "llama3.1-8b" does not exist on astro-orbiter's /v1/models
+    # (caused every summarization call to fail with 400 model not found, endless
+    # circuit-breaker retries). Actual served model id/alias confirmed via
+    # /home/hermes/git/homelab/ansible/playbooks/day2_add_nomic_embed.yml and
+    # day2_per_model_ctx_size.yml: "Meta-Llama-3.1-8B-Instruct-Q4_K_M".
+    model: "Meta-Llama-3.1-8B-Instruct-Q4_K_M"
     provider: "openai"
     temperature: 0.0
     max_retries: 2