YAML parsing fails with cryptic error "Fatal Error: Failed to parse assistant: models.0: Invalid inputmodels.1: Invalid input"

[sammyvoncheese]

Before submitting your bug report

• I believe this is a bug. I'll try to join the Continue Discord for questions
• I'm not able to find an open issue that reports the same bug
• I've seen the troubleshooting guide on the Continue Docs

Relevant environment info

- OS: Windows 11
- Continue version: 1.1.18
- IDE version: vscode Version: 1.98.2 (user setup)
Commit: ddc367ed5c8936efe395cffeec279b04ffd7db78
Date: 2025-03-12T13:32:45.399Z
Electron: 34.2.0
ElectronBuildId: 11161602
Chromium: 132.0.6834.196
Node.js: 20.18.2
V8: 13.2.152.36-electron.0
OS: Windows_NT x64 10.0.26100
- Model: all
- config:
  : v1
version: 1.0.0


models:

  - name: gemma3:27b
    provider: ollama
    model: gemma3:27b
    apiBase: http://127.0.0.1:11434
    defaultCompletionOptions:
        maxtokens: 4056
        temperature: 0.2
        topP: .2
    capabilities:
      - tools_use
    roles:
      - chat

  - name: qwen2.5:32b
    provider: ollama
    model: qwen2.5:32b
    apiBase: http://127.0.0.1:11434
    defaultCompletionOptions:
        maxtokens: 8000
        temperature: 0.2
        topP: .2
    capabilities:
      - tools_use
    roles:
      - chat
     

  
  OR link to assistant in Continue hub:

Description

After updating to version 1.0.5 I got an error message when loading. I have a large config.yaml which has grown over time, with lots of old/unnecessary or possibly outdated stuff in it. Because of this, every model had tools enabled (I never checked each one to see if they actually supported tools)

The error is caused by values this value in the yaml. In the cases I showed, those 2 models will fail because of
capabilities: - tools_use

Removing that from the yaml will allow it to load.

Here are 2 examples where it fails.

schema: v1
version: 1.0.0

models:

• name: gemma3:27b
  provider: ollama
  model: gemma3:27b
  apiBase: http://127.0.0.1:11434
  defaultCompletionOptions:
  maxtokens: 4056
  temperature: 0.2
  topP: .2
  capabilities:

  • tools_use
    roles:
  • chat

• name: qwen2.5:32b
  provider: ollama
  model: qwen2.5:32b
  apiBase: http://127.0.0.1:11434
  defaultCompletionOptions:
  maxtokens: 8000
  temperature: 0.2
  topP: .2
  capabilities:

  • tools_use
    roles:
  • chat

To reproduce

No response

Log output

[Patrick-Erichsen]

Thanks for narrowing this down and letting us know @sammyvoncheese 👍 we could definitely use some better error message naming in our YAML parsing.

Not sure how soon we'll be able to prioritize this, so I added a good first issue label if anyone is interested in picking it up. Relevant code is here: https://github.com/continuedev/continue/tree/main/packages/config-yaml

[sammyvoncheese]

I can take a look.