Skip to content

Commit 458d7e7

Browse files
CopilotXunzhuo
andcommitted
Add documentation for category-level jailbreak settings
Co-authored-by: Xunzhuo <[email protected]>
1 parent 1e384ef commit 458d7e7

File tree

4 files changed

+5
-4
lines changed
  • config
  • src/training/training_lora
    • classifier_model_fine_tuning_lora
    • pii_model_fine_tuning_lora
    • prompt_guard_fine_tuning_lora

4 files changed

+5
-4
lines changed

config/config.yaml

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -19,7 +19,7 @@ tools:
1919
fallback_to_empty: true
2020

2121
prompt_guard:
22-
enabled: true
22+
enabled: true # Global default - can be overridden per category with jailbreak_enabled
2323
use_modernbert: true
2424
model_id: "models/jailbreak_classifier_modernbert-base_model"
2525
threshold: 0.7
@@ -62,6 +62,7 @@ classifier:
6262
categories:
6363
- name: business
6464
system_prompt: "You are a senior business consultant and strategic advisor with expertise in corporate strategy, operations management, financial analysis, marketing, and organizational development. Provide practical, actionable business advice backed by proven methodologies and industry best practices. Consider market dynamics, competitive landscape, and stakeholder interests in your recommendations."
65+
# jailbreak_enabled: true # Optional: Override global jailbreak detection per category
6566
model_scores:
6667
- model: qwen3
6768
score: 0.7

src/training/training_lora/classifier_model_fine_tuning_lora/go.mod

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,4 +4,4 @@ go 1.24.1
44

55
replace github.com/vllm-project/semantic-router/candle-binding => ../../../../candle-binding
66

7-
require github.com/vllm-project/semantic-router/candle-binding v0.0.0-00010101000000-000000000000
7+
require github.com/vllm-project/semantic-router/candle-binding v0.0.0-00010101000000-000000000000

src/training/training_lora/pii_model_fine_tuning_lora/go.mod

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,4 +4,4 @@ go 1.24.1
44

55
replace github.com/vllm-project/semantic-router/candle-binding => ../../../../candle-binding
66

7-
require github.com/vllm-project/semantic-router/candle-binding v0.0.0-00010101000000-000000000000
7+
require github.com/vllm-project/semantic-router/candle-binding v0.0.0-00010101000000-000000000000

src/training/training_lora/prompt_guard_fine_tuning_lora/go.mod

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,4 +4,4 @@ go 1.24.1
44

55
replace github.com/vllm-project/semantic-router/candle-binding => ../../../../candle-binding
66

7-
require github.com/vllm-project/semantic-router/candle-binding v0.0.0-00010101000000-000000000000
7+
require github.com/vllm-project/semantic-router/candle-binding v0.0.0-00010101000000-000000000000

0 commit comments

Comments
 (0)