@openpond/evals 0.5.0 → 0.7.0

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.
Files changed (118) hide show
  1. package/CONTRACT.md +13 -1
  2. package/LEARNING.md +106 -0
  3. package/README.md +28 -2
  4. package/conformance/telemetry/v1/invalid-batch.json +17 -0
  5. package/conformance/telemetry/v1/valid-batch.json +58 -0
  6. package/dist/graders.js +26 -5
  7. package/dist/index.js +8 -0
  8. package/dist/javascript-verifier-contract.js +7 -0
  9. package/dist/javascript-verifier-node.js +56 -0
  10. package/dist/javascript-verifier-worker-source.js +1 -0
  11. package/dist/javascript-verifier-worker.js +6 -0
  12. package/dist/javascript-verifier.js +64 -0
  13. package/dist/learned-preference.js +334 -0
  14. package/dist/learning/admission.js +209 -0
  15. package/dist/learning/assets.js +37 -0
  16. package/dist/learning/contracts.js +267 -0
  17. package/dist/learning/errors.js +12 -0
  18. package/dist/learning/grade-worker.js +99 -0
  19. package/dist/learning/index.js +9 -0
  20. package/dist/learning/operations.js +33 -0
  21. package/dist/learning/repository.js +48 -0
  22. package/dist/learning/service.js +367 -0
  23. package/dist/learning/transport.js +25 -0
  24. package/dist/preferences.js +77 -11
  25. package/dist/rewards.js +230 -0
  26. package/dist/task-schema.js +142 -0
  27. package/dist/tasksets.js +3 -0
  28. package/dist/telemetry/index.js +4 -0
  29. package/dist/telemetry-analysis.js +183 -0
  30. package/dist/telemetry-bundle.js +60 -0
  31. package/dist/telemetry-catalog.js +50 -0
  32. package/dist/telemetry.js +112 -0
  33. package/dist/types/builtin-benchmarks/harness-refiner.d.ts +7 -0
  34. package/dist/types/builtin-benchmarks/harness-refiner.d.ts.map +1 -1
  35. package/dist/types/conformance.d.ts +14 -0
  36. package/dist/types/conformance.d.ts.map +1 -1
  37. package/dist/types/execution-contracts.d.ts +14 -0
  38. package/dist/types/execution-contracts.d.ts.map +1 -1
  39. package/dist/types/graders.d.ts.map +1 -1
  40. package/dist/types/index.d.ts +8 -0
  41. package/dist/types/index.d.ts.map +1 -1
  42. package/dist/types/javascript-verifier-contract.d.ts +9 -0
  43. package/dist/types/javascript-verifier-contract.d.ts.map +1 -0
  44. package/dist/types/javascript-verifier-node.d.ts +5 -0
  45. package/dist/types/javascript-verifier-node.d.ts.map +1 -0
  46. package/dist/types/javascript-verifier-worker-source.d.ts +2 -0
  47. package/dist/types/javascript-verifier-worker.d.ts +2 -0
  48. package/dist/types/javascript-verifier-worker.d.ts.map +1 -0
  49. package/dist/types/javascript-verifier.d.ts +15 -0
  50. package/dist/types/javascript-verifier.d.ts.map +1 -0
  51. package/dist/types/learned-preference.d.ts +282 -0
  52. package/dist/types/learned-preference.d.ts.map +1 -0
  53. package/dist/types/learning/admission.d.ts +335 -0
  54. package/dist/types/learning/admission.d.ts.map +1 -0
  55. package/dist/types/learning/assets.d.ts +49 -0
  56. package/dist/types/learning/assets.d.ts.map +1 -0
  57. package/dist/types/learning/contracts.d.ts +2149 -0
  58. package/dist/types/learning/contracts.d.ts.map +1 -0
  59. package/dist/types/learning/errors.d.ts +8 -0
  60. package/dist/types/learning/errors.d.ts.map +1 -0
  61. package/dist/types/learning/grade-worker.d.ts +31 -0
  62. package/dist/types/learning/grade-worker.d.ts.map +1 -0
  63. package/dist/types/learning/index.d.ts +10 -0
  64. package/dist/types/learning/index.d.ts.map +1 -0
  65. package/dist/types/learning/operations.d.ts +2061 -0
  66. package/dist/types/learning/operations.d.ts.map +1 -0
  67. package/dist/types/learning/repository.d.ts +1143 -0
  68. package/dist/types/learning/repository.d.ts.map +1 -0
  69. package/dist/types/learning/service.d.ts +22 -0
  70. package/dist/types/learning/service.d.ts.map +1 -0
  71. package/dist/types/learning/transport.d.ts +2139 -0
  72. package/dist/types/learning/transport.d.ts.map +1 -0
  73. package/dist/types/preferences.d.ts +55 -7
  74. package/dist/types/preferences.d.ts.map +1 -1
  75. package/dist/types/review-conformance.d.ts +5 -5
  76. package/dist/types/rewards.d.ts +611 -0
  77. package/dist/types/rewards.d.ts.map +1 -0
  78. package/dist/types/task-schema.d.ts +18 -0
  79. package/dist/types/task-schema.d.ts.map +1 -0
  80. package/dist/types/tasksets.d.ts +28 -0
  81. package/dist/types/tasksets.d.ts.map +1 -1
  82. package/dist/types/telemetry/index.d.ts +5 -0
  83. package/dist/types/telemetry/index.d.ts.map +1 -0
  84. package/dist/types/telemetry-analysis.d.ts +116 -0
  85. package/dist/types/telemetry-analysis.d.ts.map +1 -0
  86. package/dist/types/telemetry-bundle.d.ts +309 -0
  87. package/dist/types/telemetry-bundle.d.ts.map +1 -0
  88. package/dist/types/telemetry-catalog.d.ts +269 -0
  89. package/dist/types/telemetry-catalog.d.ts.map +1 -0
  90. package/dist/types/telemetry.d.ts +266 -0
  91. package/dist/types/telemetry.d.ts.map +1 -0
  92. package/package.json +40 -3
  93. package/schemas/learning/v1/asset.schema.json +89 -0
  94. package/schemas/learning/v1/batch.schema.json +132 -0
  95. package/schemas/learning/v1/binding.schema.json +207 -0
  96. package/schemas/learning/v1/command-request.schema.json +2414 -0
  97. package/schemas/learning/v1/decision.schema.json +440 -0
  98. package/schemas/learning/v1/definition.schema.json +397 -0
  99. package/schemas/learning/v1/evidence.schema.json +317 -0
  100. package/schemas/learning/v1/example-submission.schema.json +253 -0
  101. package/schemas/learning/v1/feedback-submission.schema.json +114 -0
  102. package/schemas/learning/v1/feedback.schema.json +233 -0
  103. package/schemas/learning/v1/grade.schema.json +412 -0
  104. package/schemas/learning/v1/iteration.schema.json +236 -0
  105. package/schemas/learning/v1/package.schema.json +776 -0
  106. package/schemas/learning/v1/policy.schema.json +373 -0
  107. package/schemas/learning/v1/read-request.schema.json +101 -0
  108. package/schemas/learning/v1/reward.schema.json +302 -0
  109. package/schemas/learning/v1/source.schema.json +198 -0
  110. package/schemas/telemetry/v1/evidence-completeness.schema.json +82 -0
  111. package/schemas/telemetry/v1/evidence-reference.schema.json +41 -0
  112. package/schemas/telemetry/v1/metric-definition.schema.json +96 -0
  113. package/schemas/telemetry/v1/metric-observation.schema.json +182 -0
  114. package/schemas/telemetry/v1/run-metric-summary.schema.json +98 -0
  115. package/schemas/telemetry/v1/run-telemetry-batch.schema.json +405 -0
  116. package/schemas/telemetry/v1/run-telemetry-event.schema.json +201 -0
  117. package/schemas/telemetry/v1/telemetry-cohort.schema.json +100 -0
  118. package/schemas/telemetry/v1/telemetry-export-bundle.schema.json +655 -0
@@ -0,0 +1,89 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "type": "object",
4
+ "properties": {
5
+ "schemaVersion": {
6
+ "type": "string",
7
+ "const": "openpond.learningTextAsset.v1"
8
+ },
9
+ "id": {
10
+ "$ref": "#/$defs/__schema0"
11
+ },
12
+ "revision": {
13
+ "type": "number",
14
+ "const": 1
15
+ },
16
+ "asset": {
17
+ "type": "object",
18
+ "properties": {
19
+ "id": {
20
+ "$ref": "#/$defs/__schema0"
21
+ },
22
+ "path": {
23
+ "type": "string",
24
+ "minLength": 1,
25
+ "maxLength": 2000
26
+ },
27
+ "contentHash": {
28
+ "$ref": "#/$defs/__schema1"
29
+ },
30
+ "sizeBytes": {
31
+ "type": "integer",
32
+ "minimum": 0,
33
+ "maximum": 250000000
34
+ },
35
+ "mediaType": {
36
+ "type": "string",
37
+ "minLength": 1,
38
+ "maxLength": 200
39
+ },
40
+ "visibility": {
41
+ "type": "string",
42
+ "enum": [
43
+ "policy",
44
+ "verifier",
45
+ "host_private"
46
+ ]
47
+ }
48
+ },
49
+ "required": [
50
+ "id",
51
+ "path",
52
+ "contentHash",
53
+ "sizeBytes",
54
+ "mediaType",
55
+ "visibility"
56
+ ],
57
+ "additionalProperties": false
58
+ },
59
+ "text": {
60
+ "type": "string",
61
+ "maxLength": 524288
62
+ },
63
+ "contentHash": {
64
+ "$ref": "#/$defs/__schema1"
65
+ }
66
+ },
67
+ "required": [
68
+ "schemaVersion",
69
+ "id",
70
+ "revision",
71
+ "asset",
72
+ "text",
73
+ "contentHash"
74
+ ],
75
+ "additionalProperties": false,
76
+ "$defs": {
77
+ "__schema0": {
78
+ "type": "string",
79
+ "minLength": 1,
80
+ "maxLength": 240
81
+ },
82
+ "__schema1": {
83
+ "type": "string",
84
+ "pattern": "^[a-f0-9]{64}$"
85
+ }
86
+ },
87
+ "$id": "urn:openpond:learning:v1:asset",
88
+ "$comment": "Generated from @openpond/evals. Cross-resource identity, authorization, idempotency, admission, grading and split constraints are enforced by the shared domain service in addition to this structural schema."
89
+ }
@@ -0,0 +1,132 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "type": "object",
4
+ "properties": {
5
+ "schemaVersion": {
6
+ "type": "string",
7
+ "const": "openpond.taskBatch.v1"
8
+ },
9
+ "id": {
10
+ "$ref": "#/$defs/__schema0"
11
+ },
12
+ "revision": {
13
+ "type": "number",
14
+ "const": 1
15
+ },
16
+ "taskDefinition": {
17
+ "$ref": "#/$defs/__schema1"
18
+ },
19
+ "rewardBinding": {
20
+ "$ref": "#/$defs/__schema1"
21
+ },
22
+ "purpose": {
23
+ "type": "string",
24
+ "enum": [
25
+ "supervised_training",
26
+ "reward_training",
27
+ "evaluation"
28
+ ]
29
+ },
30
+ "examples": {
31
+ "minItems": 1,
32
+ "maxItems": 10000,
33
+ "type": "array",
34
+ "items": {
35
+ "$ref": "#/$defs/__schema3"
36
+ }
37
+ },
38
+ "sealedAt": {
39
+ "type": "string",
40
+ "format": "date-time",
41
+ "pattern": "^(?:(?:\\d\\d[2468][048]|\\d\\d[13579][26]|\\d\\d0[48]|[02468][048]00|[13579][26]00)-02-29|\\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\\d|30)|(?:02)-(?:0[1-9]|1\\d|2[0-8])))T(?:(?:[01]\\d|2[0-3]):[0-5]\\d(?::[0-5]\\d(?:\\.\\d+)?)?(?:Z|([+-](?:[01]\\d|2[0-3]):[0-5]\\d)))$"
42
+ },
43
+ "sealedBy": {
44
+ "$ref": "#/$defs/__schema0"
45
+ },
46
+ "contentHash": {
47
+ "$ref": "#/$defs/__schema2"
48
+ }
49
+ },
50
+ "required": [
51
+ "schemaVersion",
52
+ "id",
53
+ "revision",
54
+ "taskDefinition",
55
+ "rewardBinding",
56
+ "purpose",
57
+ "examples",
58
+ "sealedAt",
59
+ "sealedBy",
60
+ "contentHash"
61
+ ],
62
+ "additionalProperties": false,
63
+ "$defs": {
64
+ "__schema0": {
65
+ "type": "string",
66
+ "minLength": 1,
67
+ "maxLength": 240
68
+ },
69
+ "__schema1": {
70
+ "type": "object",
71
+ "properties": {
72
+ "id": {
73
+ "$ref": "#/$defs/__schema0"
74
+ },
75
+ "contentHash": {
76
+ "$ref": "#/$defs/__schema2"
77
+ },
78
+ "revision": {
79
+ "type": "integer",
80
+ "exclusiveMinimum": 0,
81
+ "maximum": 9007199254740991
82
+ }
83
+ },
84
+ "required": [
85
+ "id",
86
+ "contentHash",
87
+ "revision"
88
+ ],
89
+ "additionalProperties": false
90
+ },
91
+ "__schema2": {
92
+ "type": "string",
93
+ "pattern": "^[a-f0-9]{64}$"
94
+ },
95
+ "__schema3": {
96
+ "type": "object",
97
+ "properties": {
98
+ "evidence": {
99
+ "$ref": "#/$defs/__schema1"
100
+ },
101
+ "decision": {
102
+ "$ref": "#/$defs/__schema1"
103
+ },
104
+ "familyKey": {
105
+ "$ref": "#/$defs/__schema0"
106
+ },
107
+ "inputHash": {
108
+ "$ref": "#/$defs/__schema2"
109
+ },
110
+ "split": {
111
+ "type": "string",
112
+ "enum": [
113
+ "train",
114
+ "validation",
115
+ "test",
116
+ "frozen_eval"
117
+ ]
118
+ }
119
+ },
120
+ "required": [
121
+ "evidence",
122
+ "decision",
123
+ "familyKey",
124
+ "inputHash",
125
+ "split"
126
+ ],
127
+ "additionalProperties": false
128
+ }
129
+ },
130
+ "$id": "urn:openpond:learning:v1:batch",
131
+ "$comment": "Generated from @openpond/evals. Cross-resource identity, authorization, idempotency, admission, grading and split constraints are enforced by the shared domain service in addition to this structural schema."
132
+ }
@@ -0,0 +1,207 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "type": "object",
4
+ "properties": {
5
+ "schemaVersion": {
6
+ "type": "string",
7
+ "const": "openpond.rewardBinding.v1"
8
+ },
9
+ "id": {
10
+ "$ref": "#/$defs/__schema0"
11
+ },
12
+ "revision": {
13
+ "type": "integer",
14
+ "exclusiveMinimum": 0,
15
+ "maximum": 9007199254740991
16
+ },
17
+ "name": {
18
+ "type": "string",
19
+ "minLength": 1,
20
+ "maxLength": 500
21
+ },
22
+ "description": {
23
+ "type": "string",
24
+ "maxLength": 10000
25
+ },
26
+ "recipeRef": {
27
+ "$ref": "#/$defs/__schema1"
28
+ },
29
+ "sources": {
30
+ "minItems": 1,
31
+ "maxItems": 100,
32
+ "type": "array",
33
+ "items": {
34
+ "$ref": "#/$defs/__schema3"
35
+ }
36
+ },
37
+ "aggregation": {
38
+ "type": "string",
39
+ "const": "weighted_mean"
40
+ },
41
+ "unscorable": {
42
+ "type": "string",
43
+ "const": "exclude_optional_require_all_required"
44
+ },
45
+ "contentHash": {
46
+ "$ref": "#/$defs/__schema2"
47
+ }
48
+ },
49
+ "required": [
50
+ "schemaVersion",
51
+ "id",
52
+ "revision",
53
+ "sources",
54
+ "aggregation",
55
+ "unscorable",
56
+ "contentHash"
57
+ ],
58
+ "additionalProperties": false,
59
+ "$defs": {
60
+ "__schema0": {
61
+ "type": "string",
62
+ "minLength": 1,
63
+ "maxLength": 240
64
+ },
65
+ "__schema1": {
66
+ "type": "object",
67
+ "properties": {
68
+ "id": {
69
+ "$ref": "#/$defs/__schema0"
70
+ },
71
+ "contentHash": {
72
+ "$ref": "#/$defs/__schema2"
73
+ },
74
+ "revision": {
75
+ "type": "integer",
76
+ "exclusiveMinimum": 0,
77
+ "maximum": 9007199254740991
78
+ }
79
+ },
80
+ "required": [
81
+ "id",
82
+ "contentHash",
83
+ "revision"
84
+ ],
85
+ "additionalProperties": false
86
+ },
87
+ "__schema2": {
88
+ "type": "string",
89
+ "pattern": "^[a-f0-9]{64}$"
90
+ },
91
+ "__schema3": {
92
+ "type": "object",
93
+ "properties": {
94
+ "graderId": {
95
+ "$ref": "#/$defs/__schema0"
96
+ },
97
+ "reward": {
98
+ "$ref": "#/$defs/__schema1"
99
+ },
100
+ "role": {
101
+ "type": "string",
102
+ "enum": [
103
+ "training",
104
+ "evaluation"
105
+ ]
106
+ },
107
+ "normalization": {
108
+ "oneOf": [
109
+ {
110
+ "type": "object",
111
+ "properties": {
112
+ "kind": {
113
+ "type": "string",
114
+ "const": "identity"
115
+ }
116
+ },
117
+ "required": [
118
+ "kind"
119
+ ],
120
+ "additionalProperties": false
121
+ },
122
+ {
123
+ "type": "object",
124
+ "properties": {
125
+ "kind": {
126
+ "type": "string",
127
+ "const": "linear"
128
+ },
129
+ "minimum": {
130
+ "type": "number"
131
+ },
132
+ "maximum": {
133
+ "type": "number"
134
+ },
135
+ "direction": {
136
+ "type": "string",
137
+ "enum": [
138
+ "higher",
139
+ "lower"
140
+ ]
141
+ }
142
+ },
143
+ "required": [
144
+ "kind",
145
+ "minimum",
146
+ "maximum",
147
+ "direction"
148
+ ],
149
+ "additionalProperties": false
150
+ }
151
+ ]
152
+ },
153
+ "weight": {
154
+ "type": "number",
155
+ "minimum": 0,
156
+ "maximum": 1000
157
+ },
158
+ "required": {
159
+ "type": "boolean"
160
+ },
161
+ "hardGate": {
162
+ "type": "boolean"
163
+ },
164
+ "privileged": {
165
+ "type": "boolean"
166
+ },
167
+ "fixtureRefs": {
168
+ "maxItems": 1000,
169
+ "type": "array",
170
+ "items": {
171
+ "$ref": "#/$defs/__schema4"
172
+ }
173
+ }
174
+ },
175
+ "required": [
176
+ "graderId",
177
+ "reward",
178
+ "role",
179
+ "normalization",
180
+ "weight",
181
+ "required",
182
+ "hardGate",
183
+ "privileged",
184
+ "fixtureRefs"
185
+ ],
186
+ "additionalProperties": false
187
+ },
188
+ "__schema4": {
189
+ "type": "object",
190
+ "properties": {
191
+ "id": {
192
+ "$ref": "#/$defs/__schema0"
193
+ },
194
+ "contentHash": {
195
+ "$ref": "#/$defs/__schema2"
196
+ }
197
+ },
198
+ "required": [
199
+ "id",
200
+ "contentHash"
201
+ ],
202
+ "additionalProperties": false
203
+ }
204
+ },
205
+ "$id": "urn:openpond:learning:v1:binding",
206
+ "$comment": "Generated from @openpond/evals. Cross-resource identity, authorization, idempotency, admission, grading and split constraints are enforced by the shared domain service in addition to this structural schema."
207
+ }