Skip to content

Add a --json mode to evals #55

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
sgoedecke merged 6 commits into from
Jun 5, 2025
Merged

Add a --json mode to evals #55

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
0ae63f6
Add docs link and remove unneeded type
sgoedecke Jun 5, 2025
0b65416
JSON mode for evals
sgoedecke Jun 5, 2025
53887d6
go fmt
sgoedecke Jun 5, 2025
eecd38f
Refactor out BuildChatCompletionOptions
sgoedecke Jun 5, 2025
b2cb120
Refactor out message role function
sgoedecke Jun 5, 2025
41f4662
Add docs to readme and fixup pass rate logic
sgoedecke Jun 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Add docs to readme and fixup pass rate logic
  • Loading branch information
@sgoedecke
sgoedecke committed Jun 5, 2025
commit 41f46629efc50313d8d951f0956ad13d85932119
14 changes: 14 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,6 +66,20 @@ Run the extension with output from a command. This uses single-shot mode.
cat README.md | gh models run openai/gpt-4o-mini "summarize this text"
```

#### Evaluating prompts

Run evaluation tests against a model using a `.prompt.yml` file:
```shell
gh models eval my_prompt.prompt.yml
```

The evaluation will run test cases defined in the prompt file and display results in a human-readable format. For programmatic use, you can output results in JSON format:
```shell
gh models eval my_prompt.prompt.yml --json
```

The JSON output includes detailed test results, evaluation scores, and summary statistics that can be processed by other tools or CI/CD pipelines.

## Notice

Remember when interacting with a model you are experimenting with AI, so content mistakes are possible. The feature is
Expand Down
9 changes: 6 additions & 3 deletions cmd/eval/eval.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,6 +76,9 @@ func NewEvalCommand(cfg *command.Config) *cobra.Command {
string:
contains: "hello"

By default, results are displayed in a human-readable format. Use the --json flag
to output structured JSON data for programmatic use or integration with CI/CD pipelines.

See https://docs.github.com/github-models/use-github-models/storing-prompts-in-github-repositories#supported-file-format for more information.
`),
Example: "gh models eval my_prompt.prompt.yml",
Expand Down Expand Up @@ -172,7 +175,7 @@ func (h *evalCommandHandler) runEvaluation(ctx context.Context) error {
}

// Calculate pass rate
passRate := 0.0
passRate := 100.0
if totalTests > 0 {
passRate = float64(passedTests) / float64(totalTests) * 100
}
Expand Down Expand Up @@ -238,9 +241,9 @@ func (h *evalCommandHandler) printSummary(passedTests, totalTests int, passRate
// Summary
h.cfg.WriteToOut("Evaluation Summary:\n")
if totalTests == 0 {
h.cfg.WriteToOut("Passed: 0/0 (0.0%)\n")
h.cfg.WriteToOut("Passed: 0/0 (0.00%)\n")
} else {
h.cfg.WriteToOut(fmt.Sprintf("Passed: %d/%d (%.1f%%)\n",
h.cfg.WriteToOut(fmt.Sprintf("Passed: %d/%d (%.2f%%)\n",
passedTests, totalTests, passRate))
}

Expand Down
Loading