rblundon/homelab
0
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

hindsight: swap LLM stepfun/step-3.7-flash:free -> upstage/solar-pro4:free (t_8516dba2)

Browse Source 
stepfun rejects Hindsight tagged fact-extraction (400 'missing tags');
solar-pro4:free is the previously-verified-good Nous model for retain/recall
(t_e3375410 / t_d0dffc3d). base_url + provider unchanged.
This commit is contained in:
Hermes Agent service account
2026-08-28 23:35:54 -05:00
parent aee61d4511
commit 7f8ba8b859
1 changed files with 10 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
cluster/applications/hindsight/values.yaml
View File

@@ -49,19 +49,22 @@ existingSecret: hindsight-credentials
# ----------------------------------------------------------------------------
# API container environment (explicit env entries; the chart renders this map
# to individual env vars). LLM points at the LOCAL astro-orbiter Qwen node
# (OpenAI-compatible) — moved OFF the Nous free tier (solar-pro4 was 404/dead).
# to individual env vars). LLM points at the Nous free-tier inference endpoint
# (https://inference-api.nousresearch.com/v1), model upstage/solar-pro4:free
# (tool-calling, verified retain/recall). stepfun/step-3.7-flash was swapped out
# because it rejects Hindsight's tagged fact-extraction (BadRequestError 400
# 'missing tags'); solar-pro4 is the previously-verified-good Nous model for the
# extract/retain path (tasks t_e3375410 / t_d0dffc3d).
# HINDSIGHT_API_LLM_API_KEY is NOT set here — it comes from the existingSecret
# via envFrom (1Password `nous` item); the local node ignores it but the secret
# still satisfies the chart's envFrom contract.
# via envFrom (1Password `nous` item).
# ----------------------------------------------------------------------------
api:
env:
HINDSIGHT_API_LLM_BASE_URL: "https://inference-api.nousresearch.com/v1"
HINDSIGHT_API_LLM_PROVIDER: "openai"
# Interim swap (2026-08-28, Ryan directive): local astro-orbiter Qwen is DOWN,
# so back Hindsight with Nous free stepfun/step-3.7-flash until the local node recovers.
HINDSIGHT_API_LLM_MODEL: "stepfun/step-3.7-flash:free"
# Swap (2026-08-28, t_8516dba2): stepfun failed retain w/ 400 'missing tags';
# return to verified-good Nous solar-pro4:free.
HINDSIGHT_API_LLM_MODEL: "upstage/solar-pro4:free"
# ----------------------------------------------------------------------------
# Ingress via the chart's native template.