inspect-ai 0.3.81__py3-none-any.whl → 0.3.83__py3-none-any.whl

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 (297) hide show
  1. inspect_ai/__init__.py +2 -1
  2. inspect_ai/_cli/eval.py +35 -2
  3. inspect_ai/_cli/util.py +44 -1
  4. inspect_ai/_display/core/config.py +1 -1
  5. inspect_ai/_display/core/display.py +13 -4
  6. inspect_ai/_display/core/results.py +1 -1
  7. inspect_ai/_display/textual/app.py +14 -3
  8. inspect_ai/_display/textual/display.py +4 -0
  9. inspect_ai/_display/textual/widgets/samples.py +9 -3
  10. inspect_ai/_display/textual/widgets/task_detail.py +8 -8
  11. inspect_ai/_display/textual/widgets/tasks.py +17 -1
  12. inspect_ai/_display/textual/widgets/vscode.py +44 -0
  13. inspect_ai/_eval/eval.py +74 -25
  14. inspect_ai/_eval/evalset.py +22 -18
  15. inspect_ai/_eval/loader.py +34 -11
  16. inspect_ai/_eval/run.py +13 -15
  17. inspect_ai/_eval/score.py +13 -3
  18. inspect_ai/_eval/task/generate.py +8 -9
  19. inspect_ai/_eval/task/log.py +55 -6
  20. inspect_ai/_eval/task/run.py +51 -10
  21. inspect_ai/_eval/task/task.py +23 -9
  22. inspect_ai/_util/constants.py +2 -0
  23. inspect_ai/_util/file.py +30 -1
  24. inspect_ai/_util/json.py +37 -1
  25. inspect_ai/_util/registry.py +1 -0
  26. inspect_ai/_util/vscode.py +37 -0
  27. inspect_ai/_view/server.py +113 -1
  28. inspect_ai/_view/www/App.css +7 -1
  29. inspect_ai/_view/www/dist/assets/index.css +813 -415
  30. inspect_ai/_view/www/dist/assets/index.js +54475 -32003
  31. inspect_ai/_view/www/eslint.config.mjs +1 -1
  32. inspect_ai/_view/www/log-schema.json +137 -31
  33. inspect_ai/_view/www/node_modules/flatted/python/flatted.py +149 -0
  34. inspect_ai/_view/www/package.json +11 -2
  35. inspect_ai/_view/www/src/App.tsx +161 -853
  36. inspect_ai/_view/www/src/api/api-browser.ts +176 -5
  37. inspect_ai/_view/www/src/api/api-vscode.ts +75 -1
  38. inspect_ai/_view/www/src/api/client-api.ts +66 -10
  39. inspect_ai/_view/www/src/api/jsonrpc.ts +2 -0
  40. inspect_ai/_view/www/src/api/types.ts +107 -2
  41. inspect_ai/_view/www/src/appearance/icons.ts +2 -0
  42. inspect_ai/_view/www/src/components/AsciinemaPlayer.tsx +3 -3
  43. inspect_ai/_view/www/src/components/Card.tsx +6 -4
  44. inspect_ai/_view/www/src/components/DownloadPanel.tsx +2 -2
  45. inspect_ai/_view/www/src/components/ExpandablePanel.tsx +56 -61
  46. inspect_ai/_view/www/src/components/FindBand.tsx +17 -9
  47. inspect_ai/_view/www/src/components/HumanBaselineView.tsx +1 -1
  48. inspect_ai/_view/www/src/components/JsonPanel.tsx +14 -24
  49. inspect_ai/_view/www/src/components/LargeModal.tsx +2 -35
  50. inspect_ai/_view/www/src/components/LightboxCarousel.tsx +27 -11
  51. inspect_ai/_view/www/src/components/LinkButton.module.css +16 -0
  52. inspect_ai/_view/www/src/components/LinkButton.tsx +33 -0
  53. inspect_ai/_view/www/src/components/LiveVirtualList.module.css +11 -0
  54. inspect_ai/_view/www/src/components/LiveVirtualList.tsx +177 -0
  55. inspect_ai/_view/www/src/components/MarkdownDiv.tsx +116 -26
  56. inspect_ai/_view/www/src/components/MessageBand.tsx +14 -9
  57. inspect_ai/_view/www/src/components/Modal.module.css +38 -0
  58. inspect_ai/_view/www/src/components/Modal.tsx +77 -0
  59. inspect_ai/_view/www/src/components/MorePopOver.tsx +3 -3
  60. inspect_ai/_view/www/src/components/NavPills.tsx +20 -8
  61. inspect_ai/_view/www/src/components/NoContentsPanel.module.css +12 -0
  62. inspect_ai/_view/www/src/components/NoContentsPanel.tsx +20 -0
  63. inspect_ai/_view/www/src/components/ProgressBar.module.css +5 -4
  64. inspect_ai/_view/www/src/components/ProgressBar.tsx +3 -2
  65. inspect_ai/_view/www/src/components/PulsingDots.module.css +81 -0
  66. inspect_ai/_view/www/src/components/PulsingDots.tsx +45 -0
  67. inspect_ai/_view/www/src/components/TabSet.tsx +4 -37
  68. inspect_ai/_view/www/src/components/ToolButton.tsx +3 -4
  69. inspect_ai/_view/www/src/index.tsx +26 -94
  70. inspect_ai/_view/www/src/logfile/remoteLogFile.ts +9 -1
  71. inspect_ai/_view/www/src/logfile/remoteZipFile.ts +30 -4
  72. inspect_ai/_view/www/src/metadata/RenderedContent.tsx +4 -6
  73. inspect_ai/_view/www/src/plan/DetailStep.module.css +4 -0
  74. inspect_ai/_view/www/src/plan/DetailStep.tsx +6 -3
  75. inspect_ai/_view/www/src/plan/ScorerDetailView.tsx +1 -1
  76. inspect_ai/_view/www/src/plan/SolverDetailView.module.css +2 -1
  77. inspect_ai/_view/www/src/samples/InlineSampleDisplay.module.css +9 -1
  78. inspect_ai/_view/www/src/samples/InlineSampleDisplay.tsx +74 -28
  79. inspect_ai/_view/www/src/samples/SampleDialog.tsx +58 -22
  80. inspect_ai/_view/www/src/samples/SampleDisplay.module.css +4 -0
  81. inspect_ai/_view/www/src/samples/SampleDisplay.tsx +135 -104
  82. inspect_ai/_view/www/src/samples/SampleSummaryView.module.css +10 -0
  83. inspect_ai/_view/www/src/samples/SampleSummaryView.tsx +83 -36
  84. inspect_ai/_view/www/src/samples/SamplesTools.tsx +35 -30
  85. inspect_ai/_view/www/src/samples/chat/ChatMessage.tsx +2 -1
  86. inspect_ai/_view/www/src/samples/chat/ChatMessageRenderer.tsx +1 -1
  87. inspect_ai/_view/www/src/samples/chat/ChatViewVirtualList.tsx +45 -53
  88. inspect_ai/_view/www/src/samples/chat/MessageContent.tsx +6 -1
  89. inspect_ai/_view/www/src/samples/chat/MessageContents.tsx +5 -0
  90. inspect_ai/_view/www/src/samples/chat/messages.ts +36 -0
  91. inspect_ai/_view/www/src/samples/chat/tools/ToolCallView.module.css +3 -0
  92. inspect_ai/_view/www/src/samples/chat/tools/ToolCallView.tsx +11 -1
  93. inspect_ai/_view/www/src/samples/chat/tools/ToolInput.tsx +22 -46
  94. inspect_ai/_view/www/src/samples/descriptor/samplesDescriptor.tsx +34 -20
  95. inspect_ai/_view/www/src/samples/descriptor/score/BooleanScoreDescriptor.module.css +3 -3
  96. inspect_ai/_view/www/src/samples/descriptor/score/BooleanScoreDescriptor.tsx +1 -1
  97. inspect_ai/_view/www/src/samples/descriptor/score/ObjectScoreDescriptor.module.css +4 -4
  98. inspect_ai/_view/www/src/samples/descriptor/score/ObjectScoreDescriptor.tsx +10 -10
  99. inspect_ai/_view/www/src/samples/descriptor/types.ts +6 -5
  100. inspect_ai/_view/www/src/samples/list/SampleFooter.module.css +22 -3
  101. inspect_ai/_view/www/src/samples/list/SampleFooter.tsx +27 -2
  102. inspect_ai/_view/www/src/samples/list/SampleList.tsx +122 -85
  103. inspect_ai/_view/www/src/samples/list/SampleRow.module.css +6 -0
  104. inspect_ai/_view/www/src/samples/list/SampleRow.tsx +28 -15
  105. inspect_ai/_view/www/src/samples/sample-tools/SelectScorer.tsx +29 -18
  106. inspect_ai/_view/www/src/samples/sample-tools/SortFilter.tsx +28 -28
  107. inspect_ai/_view/www/src/samples/sample-tools/sample-filter/SampleFilter.tsx +19 -9
  108. inspect_ai/_view/www/src/samples/sampleDataAdapter.ts +33 -0
  109. inspect_ai/_view/www/src/samples/sampleLimit.ts +2 -2
  110. inspect_ai/_view/www/src/samples/scores/SampleScores.tsx +12 -27
  111. inspect_ai/_view/www/src/samples/scores/SampleScoresGrid.module.css +38 -0
  112. inspect_ai/_view/www/src/samples/scores/SampleScoresGrid.tsx +118 -0
  113. inspect_ai/_view/www/src/samples/scores/{SampleScoreView.module.css → SampleScoresView.module.css} +10 -1
  114. inspect_ai/_view/www/src/samples/scores/SampleScoresView.tsx +78 -0
  115. inspect_ai/_view/www/src/samples/transcript/ErrorEventView.tsx +0 -13
  116. inspect_ai/_view/www/src/samples/transcript/InfoEventView.tsx +0 -13
  117. inspect_ai/_view/www/src/samples/transcript/InputEventView.tsx +0 -13
  118. inspect_ai/_view/www/src/samples/transcript/ModelEventView.module.css +4 -0
  119. inspect_ai/_view/www/src/samples/transcript/ModelEventView.tsx +10 -24
  120. inspect_ai/_view/www/src/samples/transcript/SampleInitEventView.tsx +0 -13
  121. inspect_ai/_view/www/src/samples/transcript/SampleLimitEventView.tsx +4 -22
  122. inspect_ai/_view/www/src/samples/transcript/SandboxEventView.tsx +15 -24
  123. inspect_ai/_view/www/src/samples/transcript/ScoreEventView.tsx +0 -13
  124. inspect_ai/_view/www/src/samples/transcript/StepEventView.tsx +6 -28
  125. inspect_ai/_view/www/src/samples/transcript/SubtaskEventView.tsx +24 -34
  126. inspect_ai/_view/www/src/samples/transcript/ToolEventView.module.css +4 -0
  127. inspect_ai/_view/www/src/samples/transcript/ToolEventView.tsx +33 -17
  128. inspect_ai/_view/www/src/samples/transcript/TranscriptView.tsx +197 -338
  129. inspect_ai/_view/www/src/samples/transcript/TranscriptVirtualListComponent.module.css +16 -0
  130. inspect_ai/_view/www/src/samples/transcript/TranscriptVirtualListComponent.tsx +44 -0
  131. inspect_ai/_view/www/src/samples/transcript/event/EventNav.tsx +7 -4
  132. inspect_ai/_view/www/src/samples/transcript/event/EventPanel.tsx +81 -60
  133. inspect_ai/_view/www/src/samples/transcript/event/EventProgressPanel.module.css +23 -0
  134. inspect_ai/_view/www/src/samples/transcript/event/EventProgressPanel.tsx +27 -0
  135. inspect_ai/_view/www/src/samples/transcript/state/StateEventRenderers.tsx +29 -1
  136. inspect_ai/_view/www/src/samples/transcript/state/StateEventView.tsx +102 -72
  137. inspect_ai/_view/www/src/scoring/utils.ts +87 -0
  138. inspect_ai/_view/www/src/state/appSlice.ts +244 -0
  139. inspect_ai/_view/www/src/state/hooks.ts +399 -0
  140. inspect_ai/_view/www/src/state/logPolling.ts +200 -0
  141. inspect_ai/_view/www/src/state/logSlice.ts +224 -0
  142. inspect_ai/_view/www/src/state/logsPolling.ts +118 -0
  143. inspect_ai/_view/www/src/state/logsSlice.ts +181 -0
  144. inspect_ai/_view/www/src/state/samplePolling.ts +314 -0
  145. inspect_ai/_view/www/src/state/sampleSlice.ts +140 -0
  146. inspect_ai/_view/www/src/state/sampleUtils.ts +21 -0
  147. inspect_ai/_view/www/src/state/scrolling.ts +206 -0
  148. inspect_ai/_view/www/src/state/store.ts +168 -0
  149. inspect_ai/_view/www/src/state/store_filter.ts +84 -0
  150. inspect_ai/_view/www/src/state/utils.ts +23 -0
  151. inspect_ai/_view/www/src/storage/index.ts +26 -0
  152. inspect_ai/_view/www/src/types/log.d.ts +36 -26
  153. inspect_ai/_view/www/src/types/markdown-it-katex.d.ts +21 -0
  154. inspect_ai/_view/www/src/types.ts +94 -32
  155. inspect_ai/_view/www/src/utils/attachments.ts +58 -23
  156. inspect_ai/_view/www/src/utils/json-worker.ts +79 -12
  157. inspect_ai/_view/www/src/utils/logger.ts +52 -0
  158. inspect_ai/_view/www/src/utils/polling.ts +100 -0
  159. inspect_ai/_view/www/src/utils/react.ts +30 -0
  160. inspect_ai/_view/www/src/utils/vscode.ts +1 -1
  161. inspect_ai/_view/www/src/workspace/WorkSpace.tsx +184 -217
  162. inspect_ai/_view/www/src/workspace/WorkSpaceView.tsx +11 -53
  163. inspect_ai/_view/www/src/workspace/navbar/Navbar.tsx +8 -18
  164. inspect_ai/_view/www/src/workspace/navbar/PrimaryBar.module.css +1 -0
  165. inspect_ai/_view/www/src/workspace/navbar/PrimaryBar.tsx +40 -22
  166. inspect_ai/_view/www/src/workspace/navbar/ResultsPanel.module.css +16 -1
  167. inspect_ai/_view/www/src/workspace/navbar/ResultsPanel.tsx +159 -103
  168. inspect_ai/_view/www/src/workspace/navbar/RunningStatusPanel.module.css +32 -0
  169. inspect_ai/_view/www/src/workspace/navbar/RunningStatusPanel.tsx +32 -0
  170. inspect_ai/_view/www/src/workspace/navbar/ScoreGrid.module.css +35 -0
  171. inspect_ai/_view/www/src/workspace/navbar/ScoreGrid.tsx +117 -0
  172. inspect_ai/_view/www/src/workspace/navbar/SecondaryBar.tsx +12 -14
  173. inspect_ai/_view/www/src/workspace/navbar/StatusPanel.tsx +6 -2
  174. inspect_ai/_view/www/src/workspace/sidebar/LogDirectoryTitleView.tsx +4 -4
  175. inspect_ai/_view/www/src/workspace/sidebar/Sidebar.module.css +3 -2
  176. inspect_ai/_view/www/src/workspace/sidebar/Sidebar.tsx +28 -13
  177. inspect_ai/_view/www/src/workspace/tabs/InfoTab.tsx +5 -10
  178. inspect_ai/_view/www/src/workspace/tabs/JsonTab.tsx +4 -4
  179. inspect_ai/_view/www/src/workspace/tabs/RunningNoSamples.module.css +22 -0
  180. inspect_ai/_view/www/src/workspace/tabs/RunningNoSamples.tsx +19 -0
  181. inspect_ai/_view/www/src/workspace/tabs/SamplesTab.tsx +128 -115
  182. inspect_ai/_view/www/src/workspace/tabs/grouping.ts +37 -5
  183. inspect_ai/_view/www/src/workspace/tabs/types.ts +4 -0
  184. inspect_ai/_view/www/src/workspace/types.ts +4 -3
  185. inspect_ai/_view/www/src/workspace/utils.ts +4 -4
  186. inspect_ai/_view/www/vite.config.js +6 -0
  187. inspect_ai/_view/www/yarn.lock +464 -355
  188. inspect_ai/agent/__init__.py +36 -0
  189. inspect_ai/agent/_agent.py +268 -0
  190. inspect_ai/agent/_as_solver.py +72 -0
  191. inspect_ai/agent/_as_tool.py +122 -0
  192. inspect_ai/{solver → agent}/_bridge/bridge.py +23 -37
  193. inspect_ai/{solver → agent}/_bridge/patch.py +9 -8
  194. inspect_ai/agent/_filter.py +46 -0
  195. inspect_ai/agent/_handoff.py +93 -0
  196. inspect_ai/{solver/_human_agent → agent/_human}/agent.py +11 -12
  197. inspect_ai/{solver/_human_agent → agent/_human}/commands/__init__.py +2 -3
  198. inspect_ai/{solver/_human_agent → agent/_human}/commands/clock.py +3 -1
  199. inspect_ai/{solver/_human_agent → agent/_human}/commands/score.py +5 -5
  200. inspect_ai/{solver/_human_agent → agent/_human}/install.py +6 -3
  201. inspect_ai/{solver/_human_agent → agent/_human}/service.py +7 -3
  202. inspect_ai/{solver/_human_agent → agent/_human}/state.py +5 -5
  203. inspect_ai/agent/_react.py +241 -0
  204. inspect_ai/agent/_run.py +36 -0
  205. inspect_ai/agent/_types.py +81 -0
  206. inspect_ai/log/_condense.py +26 -0
  207. inspect_ai/log/_log.py +17 -5
  208. inspect_ai/log/_recorders/buffer/__init__.py +14 -0
  209. inspect_ai/log/_recorders/buffer/buffer.py +30 -0
  210. inspect_ai/log/_recorders/buffer/database.py +685 -0
  211. inspect_ai/log/_recorders/buffer/filestore.py +259 -0
  212. inspect_ai/log/_recorders/buffer/types.py +84 -0
  213. inspect_ai/log/_recorders/eval.py +2 -11
  214. inspect_ai/log/_recorders/types.py +30 -0
  215. inspect_ai/log/_transcript.py +32 -2
  216. inspect_ai/model/__init__.py +7 -1
  217. inspect_ai/model/_call_tools.py +257 -52
  218. inspect_ai/model/_chat_message.py +7 -4
  219. inspect_ai/model/_conversation.py +13 -62
  220. inspect_ai/model/_display.py +85 -0
  221. inspect_ai/model/_generate_config.py +2 -2
  222. inspect_ai/model/_model.py +114 -14
  223. inspect_ai/model/_model_output.py +14 -9
  224. inspect_ai/model/_openai.py +16 -4
  225. inspect_ai/model/_openai_computer_use.py +162 -0
  226. inspect_ai/model/_openai_responses.py +319 -165
  227. inspect_ai/model/_providers/anthropic.py +20 -21
  228. inspect_ai/model/_providers/azureai.py +24 -13
  229. inspect_ai/model/_providers/bedrock.py +1 -7
  230. inspect_ai/model/_providers/cloudflare.py +3 -3
  231. inspect_ai/model/_providers/goodfire.py +2 -6
  232. inspect_ai/model/_providers/google.py +11 -10
  233. inspect_ai/model/_providers/groq.py +6 -3
  234. inspect_ai/model/_providers/hf.py +7 -3
  235. inspect_ai/model/_providers/mistral.py +7 -10
  236. inspect_ai/model/_providers/openai.py +47 -17
  237. inspect_ai/model/_providers/openai_o1.py +11 -4
  238. inspect_ai/model/_providers/openai_responses.py +12 -14
  239. inspect_ai/model/_providers/providers.py +2 -2
  240. inspect_ai/model/_providers/together.py +12 -2
  241. inspect_ai/model/_providers/util/chatapi.py +7 -2
  242. inspect_ai/model/_providers/util/hf_handler.py +4 -2
  243. inspect_ai/model/_providers/util/llama31.py +4 -2
  244. inspect_ai/model/_providers/vertex.py +11 -9
  245. inspect_ai/model/_providers/vllm.py +4 -4
  246. inspect_ai/scorer/__init__.py +2 -0
  247. inspect_ai/scorer/_metrics/__init__.py +2 -0
  248. inspect_ai/scorer/_metrics/grouped.py +84 -0
  249. inspect_ai/scorer/_score.py +26 -6
  250. inspect_ai/solver/__init__.py +2 -2
  251. inspect_ai/solver/_basic_agent.py +22 -9
  252. inspect_ai/solver/_bridge.py +31 -0
  253. inspect_ai/solver/_chain.py +20 -12
  254. inspect_ai/solver/_fork.py +5 -1
  255. inspect_ai/solver/_human_agent.py +52 -0
  256. inspect_ai/solver/_prompt.py +3 -1
  257. inspect_ai/solver/_run.py +59 -0
  258. inspect_ai/solver/_solver.py +14 -4
  259. inspect_ai/solver/_task_state.py +5 -3
  260. inspect_ai/tool/_tool_call.py +15 -8
  261. inspect_ai/tool/_tool_def.py +17 -12
  262. inspect_ai/tool/_tool_support_helpers.py +4 -4
  263. inspect_ai/tool/_tool_with.py +14 -11
  264. inspect_ai/tool/_tools/_bash_session.py +11 -2
  265. inspect_ai/tool/_tools/_computer/_common.py +18 -2
  266. inspect_ai/tool/_tools/_computer/_computer.py +18 -2
  267. inspect_ai/tool/_tools/_computer/_resources/tool/_constants.py +2 -0
  268. inspect_ai/tool/_tools/_computer/_resources/tool/_x11_client.py +17 -0
  269. inspect_ai/tool/_tools/_think.py +1 -1
  270. inspect_ai/tool/_tools/_web_browser/_web_browser.py +103 -62
  271. inspect_ai/util/__init__.py +2 -0
  272. inspect_ai/util/_anyio.py +27 -0
  273. inspect_ai/util/_sandbox/__init__.py +2 -1
  274. inspect_ai/util/_sandbox/context.py +32 -7
  275. inspect_ai/util/_sandbox/docker/cleanup.py +4 -0
  276. inspect_ai/util/_sandbox/docker/compose.py +2 -2
  277. inspect_ai/util/_sandbox/docker/docker.py +12 -1
  278. inspect_ai/util/_store_model.py +30 -7
  279. inspect_ai/util/_subprocess.py +13 -3
  280. inspect_ai/util/_subtask.py +1 -0
  281. {inspect_ai-0.3.81.dist-info → inspect_ai-0.3.83.dist-info}/METADATA +1 -1
  282. {inspect_ai-0.3.81.dist-info → inspect_ai-0.3.83.dist-info}/RECORD +295 -229
  283. inspect_ai/_view/www/src/samples/scores/SampleScoreView.tsx +0 -169
  284. inspect_ai/_view/www/src/samples/transcript/SampleTranscript.tsx +0 -22
  285. /inspect_ai/{solver → agent}/_bridge/__init__.py +0 -0
  286. /inspect_ai/{solver/_human_agent → agent/_human}/__init__.py +0 -0
  287. /inspect_ai/{solver/_human_agent → agent/_human}/commands/command.py +0 -0
  288. /inspect_ai/{solver/_human_agent → agent/_human}/commands/instructions.py +0 -0
  289. /inspect_ai/{solver/_human_agent → agent/_human}/commands/note.py +0 -0
  290. /inspect_ai/{solver/_human_agent → agent/_human}/commands/status.py +0 -0
  291. /inspect_ai/{solver/_human_agent → agent/_human}/commands/submit.py +0 -0
  292. /inspect_ai/{solver/_human_agent → agent/_human}/panel.py +0 -0
  293. /inspect_ai/{solver/_human_agent → agent/_human}/view.py +0 -0
  294. {inspect_ai-0.3.81.dist-info → inspect_ai-0.3.83.dist-info}/WHEEL +0 -0
  295. {inspect_ai-0.3.81.dist-info → inspect_ai-0.3.83.dist-info}/entry_points.txt +0 -0
  296. {inspect_ai-0.3.81.dist-info → inspect_ai-0.3.83.dist-info}/licenses/LICENSE +0 -0
  297. {inspect_ai-0.3.81.dist-info → inspect_ai-0.3.83.dist-info}/top_level.txt +0 -0
@@ -1,27 +1,22 @@
1
1
  import { ApplicationIcons } from "../appearance/icons";
2
2
  import { LargeModal, ModalTool, ModalTools } from "../components/LargeModal";
3
3
 
4
- import { FC, Ref, RefObject, useCallback, useMemo, useRef } from "react";
4
+ import { FC, Ref, useCallback, useEffect, useMemo, useRef } from "react";
5
5
  import { ErrorPanel } from "../components/ErrorPanel";
6
- import { EvalSample } from "../types/log";
6
+ import { useLogSelection, usePrevious, useSampleData } from "../state/hooks";
7
+ import { useStatefulScrollPosition } from "../state/scrolling";
8
+ import { useStore } from "../state/store";
7
9
  import { SampleDisplay } from "./SampleDisplay";
8
- import { SamplesDescriptor } from "./descriptor/samplesDescriptor";
9
10
 
10
11
  interface SampleDialogProps {
11
12
  id: string;
12
13
  title: string;
13
- sampleStatus: string;
14
- sampleError?: Error;
15
- sample?: EvalSample;
16
- sampleDescriptor: SamplesDescriptor;
17
14
  selectedTab?: string;
18
15
  setSelectedTab: (tab: string) => void;
19
16
  showingSampleDialog: boolean;
20
17
  setShowingSampleDialog: (showing: boolean) => void;
21
18
  nextSample: () => void;
22
19
  prevSample: () => void;
23
- sampleScrollPositionRef: RefObject<number>;
24
- setSampleScrollPosition: (position: number) => void;
25
20
  }
26
21
 
27
22
  /**
@@ -30,21 +25,62 @@ interface SampleDialogProps {
30
25
  export const SampleDialog: FC<SampleDialogProps> = ({
31
26
  id,
32
27
  title,
33
- sample,
34
- sampleDescriptor,
35
28
  nextSample,
36
29
  prevSample,
37
- sampleStatus,
38
- sampleError,
39
30
  showingSampleDialog,
40
31
  setShowingSampleDialog,
41
32
  selectedTab,
42
33
  setSelectedTab,
43
- sampleScrollPositionRef,
44
- setSampleScrollPosition,
45
34
  }) => {
35
+ // Scroll referernce (attach stateful trackign)
46
36
  const scrollRef: Ref<HTMLDivElement> = useRef(null);
37
+ useStatefulScrollPosition(scrollRef, "sample-dialog");
47
38
 
39
+ // Sample hooks
40
+ const sampleData = useSampleData();
41
+ const loadSample = useStore((state) => state.sampleActions.loadSample);
42
+ const pollSample = useStore((state) => state.sampleActions.pollSample);
43
+ const logSelection = useLogSelection();
44
+
45
+ useEffect(() => {
46
+ if (sampleData.running && logSelection.logFile && logSelection.sample) {
47
+ pollSample(logSelection.logFile, logSelection.sample);
48
+ }
49
+ }, []);
50
+
51
+ // Load sample
52
+ const prevCompleted = usePrevious(
53
+ logSelection.sample?.completed !== undefined
54
+ ? logSelection.sample.completed
55
+ : true,
56
+ );
57
+ const prevLogFile = usePrevious<string | undefined>(logSelection.logFile);
58
+ useEffect(() => {
59
+ if (logSelection.logFile && logSelection.sample) {
60
+ const currentSampleCompleted =
61
+ logSelection.sample.completed !== undefined
62
+ ? logSelection.sample.completed
63
+ : true;
64
+
65
+ if (
66
+ prevLogFile !== logSelection.logFile ||
67
+ sampleData.sample?.id !== logSelection.sample.id ||
68
+ sampleData.sample?.epoch !== logSelection.sample.epoch ||
69
+ currentSampleCompleted !== prevCompleted
70
+ ) {
71
+ loadSample(logSelection.logFile, logSelection.sample);
72
+ }
73
+ }
74
+ }, [
75
+ logSelection.logFile,
76
+ logSelection.sample?.id,
77
+ logSelection.sample?.epoch,
78
+ logSelection.sample?.completed,
79
+ sampleData.sample?.id,
80
+ sampleData.sample?.epoch,
81
+ ]);
82
+
83
+ // Tools
48
84
  const tools = useMemo<ModalTools>(() => {
49
85
  const nextTool: ModalTool = {
50
86
  label: "Next Sample",
@@ -100,18 +136,18 @@ export const SampleDialog: FC<SampleDialogProps> = ({
100
136
  onkeyup={handleKeyUp}
101
137
  visible={showingSampleDialog}
102
138
  onHide={onHide}
103
- showProgress={sampleStatus === "loading"}
104
- initialScrollPositionRef={sampleScrollPositionRef}
105
- setInitialScrollPosition={setSampleScrollPosition}
139
+ showProgress={
140
+ sampleData.status === "loading" || sampleData.status === "streaming"
141
+ }
106
142
  scrollRef={scrollRef}
107
143
  >
108
- {sampleError ? (
109
- <ErrorPanel title="Sample Error" error={sampleError} />
144
+ {sampleData.error ? (
145
+ <ErrorPanel title="Sample Error" error={sampleData.error} />
110
146
  ) : (
111
147
  <SampleDisplay
112
148
  id={id}
113
- sample={sample}
114
- sampleDescriptor={sampleDescriptor}
149
+ sample={sampleData.sample}
150
+ runningEvents={sampleData.running}
115
151
  selectedTab={selectedTab}
116
152
  setSelectedTab={setSelectedTab}
117
153
  scrollRef={scrollRef}
@@ -35,3 +35,7 @@
35
35
  column-gap: 0.5em;
36
36
  min-width: 200px;
37
37
  }
38
+
39
+ .chat {
40
+ padding-top: 1em;
41
+ }
@@ -7,13 +7,20 @@ import { isVscode } from "../utils/vscode";
7
7
  import { ApplicationIcons } from "../appearance/icons";
8
8
  import { ANSIDisplay } from "../components/AnsiDisplay";
9
9
  import { ToolButton } from "../components/ToolButton";
10
- import { SampleScoreView } from "./scores/SampleScoreView";
11
10
 
12
11
  import clsx from "clsx";
13
- import { FC, Fragment, MouseEvent, RefObject } from "react";
12
+ import {
13
+ FC,
14
+ Fragment,
15
+ MouseEvent,
16
+ RefObject,
17
+ useCallback,
18
+ useMemo,
19
+ } from "react";
20
+ import { SampleSummary } from "../api/types";
14
21
  import { Card, CardBody, CardHeader } from "../components/Card";
15
- import { EmptyPanel } from "../components/EmptyPanel";
16
22
  import { JSONPanel } from "../components/JsonPanel";
23
+ import { NoContentsPanel } from "../components/NoContentsPanel";
17
24
  import {
18
25
  kSampleErrorTabId,
19
26
  kSampleJsonTabId,
@@ -22,23 +29,26 @@ import {
22
29
  kSampleScoringTabId,
23
30
  kSampleTranscriptTabId,
24
31
  } from "../constants";
25
- import { EvalSample } from "../types/log";
32
+ import { useSampleSummaries } from "../state/hooks";
33
+ import { useStore } from "../state/store";
34
+ import { EvalSample, Events } from "../types/log";
26
35
  import { ModelTokenTable } from "../usage/ModelTokenTable";
27
36
  import { formatTime } from "../utils/format";
28
37
  import { printHeadingHtml, printHtml } from "../utils/print";
29
38
  import { ChatViewVirtualList } from "./chat/ChatViewVirtualList";
30
- import { SamplesDescriptor } from "./descriptor/samplesDescriptor";
39
+ import { messagesFromEvents } from "./chat/messages";
31
40
  import styles from "./SampleDisplay.module.css";
32
41
  import { SampleSummaryView } from "./SampleSummaryView";
33
- import { SampleTranscript } from "./transcript/SampleTranscript";
42
+ import { SampleScoresView } from "./scores/SampleScoresView";
43
+ import { TranscriptVirtualList } from "./transcript/TranscriptView";
34
44
 
35
45
  interface SampleDisplayProps {
36
46
  id: string;
37
47
  sample?: EvalSample;
38
- sampleDescriptor: SamplesDescriptor;
39
48
  selectedTab?: string;
40
49
  setSelectedTab: (tab: string) => void;
41
50
  scrollRef: RefObject<HTMLDivElement | null>;
51
+ runningEvents?: Events;
42
52
  }
43
53
 
44
54
  /**
@@ -47,18 +57,32 @@ interface SampleDisplayProps {
47
57
  export const SampleDisplay: FC<SampleDisplayProps> = ({
48
58
  id,
49
59
  sample,
50
- sampleDescriptor,
51
60
  selectedTab,
52
61
  setSelectedTab,
53
62
  scrollRef,
63
+ runningEvents: runningSampleData,
54
64
  }) => {
55
65
  // Tab ids
56
66
  const baseId = `sample-dialog`;
67
+ const sampleSummaries = useSampleSummaries();
68
+ const selectedSampleIndex = useStore(
69
+ (state) => state.log.selectedSampleIndex,
70
+ );
57
71
 
58
- if (!sample) {
59
- // Placeholder
60
- return <EmptyPanel />;
61
- }
72
+ const sampleSummary = sampleSummaries[selectedSampleIndex];
73
+
74
+ // Consolidate the events and messages into the proper list
75
+ // whether running or not
76
+ const sampleEvents = sample?.events || runningSampleData;
77
+ const sampleMessages = useMemo(() => {
78
+ if (sample?.messages) {
79
+ return sample.messages;
80
+ } else if (runningSampleData) {
81
+ return messagesFromEvents(runningSampleData);
82
+ } else {
83
+ return [];
84
+ }
85
+ }, [sample?.messages, runningSampleData]);
62
86
 
63
87
  // Tab selection
64
88
  const onSelectedTab = (e: MouseEvent<HTMLElement>) => {
@@ -68,12 +92,15 @@ export const SampleDisplay: FC<SampleDisplayProps> = ({
68
92
  return false;
69
93
  };
70
94
 
71
- const scorerNames = Object.keys(sample.scores || {});
72
95
  const sampleMetadatas = metadataViewsForSample(`${baseId}-${id}`, sample);
73
96
 
74
97
  const tabsetId = `task-sample-details-tab-${id}`;
75
98
  const targetId = `${tabsetId}-content`;
76
99
 
100
+ const handlePrintClick = useCallback(() => {
101
+ printSample(id, targetId);
102
+ }, [printSample, id, targetId]);
103
+
77
104
  const tools = [];
78
105
  if (!isVscode()) {
79
106
  tools.push(
@@ -81,51 +108,48 @@ export const SampleDisplay: FC<SampleDisplayProps> = ({
81
108
  key="sample-print-tool"
82
109
  label="Print"
83
110
  icon={ApplicationIcons.copy}
84
- onClick={() => {
85
- printSample(id, targetId);
86
- }}
111
+ onClick={handlePrintClick}
87
112
  />,
88
113
  );
89
114
  }
90
115
 
116
+ // Is the sample running?
117
+ const running = isRunning(sampleSummary, runningSampleData);
118
+
91
119
  return (
92
120
  <Fragment>
93
- <SampleSummaryView
94
- parent_id={id}
95
- sample={sample}
96
- sampleDescriptor={sampleDescriptor}
97
- />
121
+ {sample || sampleSummary ? (
122
+ <SampleSummaryView parent_id={id} sample={sample || sampleSummary} />
123
+ ) : undefined}
98
124
  <TabSet
99
125
  id={tabsetId}
100
126
  tabControlsClassName={clsx("text-size-base")}
101
127
  tabPanelsClassName={clsx(styles.tabPanel)}
102
128
  tools={tools}
103
129
  >
104
- {sample.events && sample.events.length > 0 ? (
105
- <TabPanel
106
- key={kSampleTranscriptTabId}
107
- id={kSampleTranscriptTabId}
108
- className="sample-tab"
109
- title="Transcript"
110
- onSelected={onSelectedTab}
111
- selected={
112
- selectedTab === kSampleTranscriptTabId ||
113
- selectedTab === undefined
114
- }
115
- scrollable={false}
116
- >
117
- <SampleTranscript
118
- key={`${baseId}-transcript-display-${id}`}
119
- id={`${baseId}-transcript-display-${id}`}
120
- evalEvents={sample.events}
121
- scrollRef={scrollRef}
122
- />
123
- </TabPanel>
124
- ) : null}
130
+ <TabPanel
131
+ key={kSampleTranscriptTabId}
132
+ id={kSampleTranscriptTabId}
133
+ className="sample-tab"
134
+ title="Transcript"
135
+ onSelected={onSelectedTab}
136
+ selected={
137
+ selectedTab === kSampleTranscriptTabId || selectedTab === undefined
138
+ }
139
+ scrollable={false}
140
+ >
141
+ <TranscriptVirtualList
142
+ key={`${baseId}-transcript-display-${id}`}
143
+ id={`${baseId}-transcript-display-${id}`}
144
+ events={sampleEvents || []}
145
+ running={running}
146
+ scrollRef={scrollRef}
147
+ />
148
+ </TabPanel>
125
149
  <TabPanel
126
150
  key={kSampleMessagesTabId}
127
151
  id={kSampleMessagesTabId}
128
- className={clsx("sample-tab", styles.fullWidth)}
152
+ className={clsx("sample-tab", styles.fullWidth, styles.chat)}
129
153
  title="Messages"
130
154
  onSelected={onSelectedTab}
131
155
  selected={selectedTab === kSampleMessagesTabId}
@@ -134,62 +158,37 @@ export const SampleDisplay: FC<SampleDisplayProps> = ({
134
158
  <ChatViewVirtualList
135
159
  key={`${baseId}-chat-${id}`}
136
160
  id={`${baseId}-chat-${id}`}
137
- messages={sample.messages}
161
+ messages={sampleMessages}
138
162
  indented={true}
139
163
  scrollRef={scrollRef}
140
164
  toolCallStyle="complete"
165
+ running={running}
141
166
  />
142
167
  </TabPanel>
143
- {scorerNames.length === 1 ? (
144
- <TabPanel
145
- key={kSampleScoringTabId}
146
- id={kSampleScoringTabId}
147
- className="sample-tab"
148
- title="Scoring"
149
- onSelected={onSelectedTab}
150
- selected={selectedTab === kSampleScoringTabId}
151
- >
152
- <SampleScoreView
153
- sample={sample}
154
- sampleDescriptor={sampleDescriptor}
155
- scorer={scorerNames[0]}
156
- />
157
- </TabPanel>
158
- ) : (
159
- <>
160
- {Object.keys(sample.scores || {}).map((scorer) => {
161
- const tabId = `score-${scorer}`;
162
- return (
163
- <TabPanel
164
- key={tabId}
165
- id={tabId}
166
- className="sample-tab"
167
- title={scorer}
168
- onSelected={onSelectedTab}
169
- selected={selectedTab === tabId}
170
- >
171
- <SampleScoreView
172
- sample={sample}
173
- sampleDescriptor={sampleDescriptor}
174
- scorer={scorer}
175
- />
176
- </TabPanel>
177
- );
178
- })}
179
- </>
180
- )}
181
- {sampleMetadatas.length > 0 ? (
182
- <TabPanel
183
- id={kSampleMetdataTabId}
184
- className={clsx("sample-tab")}
185
- title="Metadata"
186
- onSelected={onSelectedTab}
187
- selected={selectedTab === kSampleMetdataTabId}
188
- >
168
+ <TabPanel
169
+ key={kSampleScoringTabId}
170
+ id={kSampleScoringTabId}
171
+ className="sample-tab"
172
+ title="Scoring"
173
+ onSelected={onSelectedTab}
174
+ selected={selectedTab === kSampleScoringTabId}
175
+ >
176
+ <SampleScoresView sample={sample} />
177
+ </TabPanel>
178
+ <TabPanel
179
+ id={kSampleMetdataTabId}
180
+ className={clsx("sample-tab")}
181
+ title="Metadata"
182
+ onSelected={onSelectedTab}
183
+ selected={selectedTab === kSampleMetdataTabId}
184
+ >
185
+ {sampleMetadatas.length > 0 ? (
189
186
  <div className={clsx(styles.metadataPanel)}>{sampleMetadatas}</div>
190
- </TabPanel>
191
- ) : null}
192
- {sample.error ? (
187
+ ) : (
188
+ <NoContentsPanel text="No metadata" />
189
+ )}
190
+ </TabPanel>
191
+ {sample?.error ? (
193
192
  <TabPanel
194
193
  id={kSampleErrorTabId}
195
194
  className="sample-tab"
@@ -205,14 +204,18 @@ export const SampleDisplay: FC<SampleDisplayProps> = ({
205
204
  </div>
206
205
  </TabPanel>
207
206
  ) : null}
208
- {sample.messages.length < 100 ? (
209
- <TabPanel
210
- id={kSampleJsonTabId}
211
- className={"sample-tab"}
212
- title="JSON"
213
- onSelected={onSelectedTab}
214
- selected={selectedTab === kSampleJsonTabId}
215
- >
207
+ <TabPanel
208
+ id={kSampleJsonTabId}
209
+ className={"sample-tab"}
210
+ title="JSON"
211
+ onSelected={onSelectedTab}
212
+ selected={selectedTab === kSampleJsonTabId}
213
+ >
214
+ {!sample ? (
215
+ <NoContentsPanel text="JSON not available" />
216
+ ) : sample.messages.length > 100 ? (
217
+ <NoContentsPanel text="JSON too large too display" />
218
+ ) : (
216
219
  <div className={clsx(styles.padded, styles.fullWidth)}>
217
220
  <JSONPanel
218
221
  data={sample}
@@ -220,14 +223,17 @@ export const SampleDisplay: FC<SampleDisplayProps> = ({
220
223
  className={clsx("text-size-small")}
221
224
  />
222
225
  </div>
223
- </TabPanel>
224
- ) : null}
226
+ )}
227
+ </TabPanel>
225
228
  </TabSet>
226
229
  </Fragment>
227
230
  );
228
231
  };
229
232
 
230
- const metadataViewsForSample = (id: string, sample: EvalSample) => {
233
+ const metadataViewsForSample = (id: string, sample?: EvalSample) => {
234
+ if (!sample) {
235
+ return [];
236
+ }
231
237
  const sampleMetadatas = [];
232
238
 
233
239
  if (sample.model_usage && Object.keys(sample.model_usage).length > 0) {
@@ -360,3 +366,28 @@ const printSample = (id: string, targetId: string) => {
360
366
  }
361
367
  }
362
368
  };
369
+
370
+ const isRunning = (
371
+ sampleSummary?: SampleSummary,
372
+ runningSampleData?: Events,
373
+ ): boolean => {
374
+ if (sampleSummary && sampleSummary.completed === false) {
375
+ // An explicitly incomplete sample summary
376
+ return true;
377
+ }
378
+
379
+ if (
380
+ !sampleSummary &&
381
+ (!runningSampleData || runningSampleData.length === 0)
382
+ ) {
383
+ // No sample summary yet and no running samples, must've just started
384
+ return true;
385
+ }
386
+
387
+ if (runningSampleData && runningSampleData.length > 0) {
388
+ // There are running samples
389
+ return true;
390
+ }
391
+
392
+ return false;
393
+ };
@@ -19,6 +19,11 @@
19
19
  justify-content: center;
20
20
  }
21
21
 
22
+ .centerValue {
23
+ display: flex;
24
+ align-items: center;
25
+ }
26
+
22
27
  .wrap {
23
28
  word-wrap: anywhere;
24
29
  }
@@ -26,3 +31,8 @@
26
31
  .titled:hover {
27
32
  cursor: pointer;
28
33
  }
34
+
35
+ .value {
36
+ flex-direction: column;
37
+ padding-top: 0.1em;
38
+ }