[STG-570] add file extract for reducto #927
Conversation
|
59b36ba to
a986d79
Compare
| /** | ||
| * Your Reducto API key for PDF extraction | ||
| */ | ||
| reductoApiKey?: string; |
There was a problem hiding this comment.
can also make this more ambiguous like ocr model key
| } | ||
| } | ||
|
|
||
| private zodToJsonSchema(schema: z.AnyZodObject): Record<string, unknown> { |
There was a problem hiding this comment.
can move this and zodtypetojsonschema to a utils
| return `${timestamp}::[stagehand:${logLine.category}] ${logLine.message}\n ${logLine.auxiliary.error.value}\n ${logLine.auxiliary.trace.value}`; | ||
| const errorValue = logLine.auxiliary.error?.value || ""; | ||
| const traceValue = logLine.auxiliary.trace?.value || ""; | ||
| return `${timestamp}::[stagehand:${logLine.category}] ${logLine.message}\n ${errorValue}${traceValue ? "\n " + traceValue : ""}`; |
There was a problem hiding this comment.
this was causing a whole bunch of errors in the evals testing, so i added this fix
| @@ -0,0 +1,219 @@ | |||
| import { EvalFunction } from "@/types/evals"; | |||
There was a problem hiding this comment.
i still need to add a test to make sure it works with file path, need to find or create a website with a pdf download on it.
why
Adding OCR extractions on PDFs
what changed
Added a reducto client, and added an eval for PDF extraction.
test plan