orchestrated 0.1.2 → 0.1.4
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +16 -17
- package/index.d.ts +1517 -243
- package/index.js +13977 -13485
- package/index.js.map +108 -26
- package/package.json +3 -2
package/README.md
CHANGED
|
@@ -12,7 +12,6 @@ Orchestrated is a comprehensive evaluation framework for LLM applications. It pr
|
|
|
12
12
|
- 📡 **Data Sources** - Built-in support for various data sources
|
|
13
13
|
- 🔄 **Batch Processing** - Efficient batch evaluation with resume support
|
|
14
14
|
- 📈 **Progress Reporting** - Beautiful terminal UI with progress bars
|
|
15
|
-
- 🔍 **OpenTelemetry** - Built-in tracing and telemetry support
|
|
16
15
|
|
|
17
16
|
## Installation
|
|
18
17
|
|
|
@@ -27,19 +26,23 @@ yarn add orchestrated
|
|
|
27
26
|
## Quick Start
|
|
28
27
|
|
|
29
28
|
```typescript
|
|
29
|
+
import { Levenshtein } from "autoevals";
|
|
30
30
|
import { Eval } from "orchestrated";
|
|
31
31
|
|
|
32
|
-
|
|
33
|
-
Eval("My First Eval", {
|
|
32
|
+
Eval("Example 2: Standard Eval", {
|
|
34
33
|
data: [
|
|
35
|
-
{ input: "What is 2+2?", output: "4", expected: "4" },
|
|
36
34
|
{
|
|
37
|
-
input: "What is
|
|
38
|
-
|
|
39
|
-
|
|
35
|
+
input: "What is a good name for a child?",
|
|
36
|
+
expected: "John",
|
|
37
|
+
output: "Nurse",
|
|
38
|
+
},
|
|
39
|
+
{
|
|
40
|
+
input: "What is a good name for a child?",
|
|
41
|
+
expected: "John",
|
|
42
|
+
output: "John",
|
|
40
43
|
},
|
|
41
44
|
],
|
|
42
|
-
scores: [
|
|
45
|
+
scores: [Levenshtein],
|
|
43
46
|
});
|
|
44
47
|
```
|
|
45
48
|
|
|
@@ -48,7 +51,7 @@ Eval("My First Eval", {
|
|
|
48
51
|
### With Custom Task Function
|
|
49
52
|
|
|
50
53
|
```typescript
|
|
51
|
-
import {
|
|
54
|
+
import { Factuality } from "autoevals";
|
|
52
55
|
import { generateAI } from "./my-ai-service";
|
|
53
56
|
|
|
54
57
|
Eval("LLM Evaluation", {
|
|
@@ -60,22 +63,18 @@ Eval("LLM Evaluation", {
|
|
|
60
63
|
const output = await generateAI(input);
|
|
61
64
|
return output;
|
|
62
65
|
},
|
|
63
|
-
scores: [
|
|
66
|
+
scores: [Factuality],
|
|
64
67
|
});
|
|
65
68
|
```
|
|
66
69
|
|
|
67
|
-
### With Data
|
|
70
|
+
### With Orchestrated Console Data Source
|
|
68
71
|
|
|
69
72
|
```typescript
|
|
70
73
|
import { Eval, interactions } from "orchestrated";
|
|
71
74
|
|
|
72
75
|
Eval("Production Eval", {
|
|
73
|
-
data: interactions(
|
|
74
|
-
|
|
75
|
-
serviceName: "my-service",
|
|
76
|
-
environment: "production",
|
|
77
|
-
}),
|
|
78
|
-
scores: ["Effectiveness", "GuardrailAdherence"],
|
|
76
|
+
data: interactions(),
|
|
77
|
+
scores: ["Effectiveness"],
|
|
79
78
|
});
|
|
80
79
|
```
|
|
81
80
|
|