Skip to content

fix: warn on invalid context_length format in config.yaml#10067

Merged
teknium1 merged 1 commit into
mainfrom
hermes/hermes-1887e8cf
Apr 15, 2026
Merged

fix: warn on invalid context_length format in config.yaml#10067
teknium1 merged 1 commit into
mainfrom
hermes/hermes-1887e8cf

Conversation

@teknium1

@teknium1 teknium1 commented Apr 15, 2026

Copy link
Copy Markdown
Contributor

Summary

Non-integer context_length values (e.g. 256K, 1M) in config.yaml were silently ignored — the int() conversion threw ValueError, the except block set it to None, and the agent fell back to 128K auto-detection with zero user feedback.

This was reported by a community user (ChFarhan) who had a LiteLLM custom endpoint proxying GPT-5.4 (256K context) and Qwen 3.6 Plus (1M context). Both showed as 128K because their context_length config values couldn't be parsed.

Changes

run_agent.py — Both silent-failure paths now:

  • Log a WARNING with the invalid value and expected format
  • Print a clear ⚠ message to stderr on CLI launch
  • Still fall back gracefully to auto-detection

Two paths covered:

  1. model.context_length in the model config section
  2. custom_providers[].models.<model>.context_length per-model override

tests/run_agent/test_invalid_context_length_warning.py — 5 tests:

  • Valid integer works silently
  • String '256K' triggers warning + sets None
  • Numeric string '256000' parses fine (no warning)
  • Invalid value in custom_providers triggers warning
  • Valid value in custom_providers works silently

What it looks like

⚠ Invalid model.context_length in config.yaml: '256K'
  Must be a plain integer (e.g. 256000, not '256K').
  Falling back to auto-detected context window.

Test plan

python -m pytest tests/run_agent/test_invalid_context_length_warning.py -n0 -v

@teknium1
fix: warn on invalid context_length format in config.yaml
d232e27 
Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
@teknium1
teknium1 merged commit 93fe4ea into main Apr 15, 2026
6 of 7 checks passed
@teknium1
teknium1 deleted the hermes/hermes-1887e8cf branch April 15, 2026 05:14
@charleneleong-ai charleneleong-ai mentioned this pull request Apr 18, 2026
Closed
ulasbilgen pushed a commit to ulasbilgen/hermes-adhd-agent that referenced this pull request May 1, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
ae9f21c 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
aj-nt pushed a commit to aj-nt/hermes-agent that referenced this pull request May 1, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
038f8bf 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
02356abc pushed a commit to 02356abc/hermes-agent that referenced this pull request May 14, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
757e711 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
gweeteve pushed a commit to gweeteve/hermes-agent that referenced this pull request Jun 2, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
475af56 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
Egavasyug pushed a commit to Egavasyug/hermes-agent that referenced this pull request Jun 10, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
b1d4761 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
waefrebeorn pushed a commit to waefrebeorn/slermes that referenced this pull request Jul 2, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
6e3a982 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
liuchanchen pushed a commit to liuchanchen/hermes-agent that referenced this pull request Jul 3, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
d8eb952 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
kulikman pushed a commit to kulikman/hermes-agent that referenced this pull request Jul 16, 2026
@teknium1
fix: warn on invalid context_length format in config.yaml (NousResear…
1799a07 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
waym0reom3ga added a commit to waym0reom3ga/autolycus-agent that referenced this pull request Jul 21, 2026
@waym0reom3ga
fix: warn on invalid context_length format in config.yaml (NousResear…
6a51d92 
…ch#10067)

Previously, non-integer context_length values (e.g. '256K') in
config.yaml were silently ignored, causing the agent to fall back
to 128K auto-detection with no user feedback. This was confusing
for users with custom LiteLLM endpoints expecting larger context.

Now prints a clear stderr warning and logs at WARNING level when
model.context_length or custom_providers[].models.<model>.context_length
cannot be parsed as an integer, telling users to use plain integers
(e.g. 256000 instead of '256K').

Reported by community user ChFarhan via Discord.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@teknium1