The ZodObject at ... and contains an input transformation (ZodDefault) defined at ... which is also used in an output schema.

[RonB]

Hi, sorry to bother you again,

but I get this error if a schema is used as a body parameter and as a response schema:

The ZodObject at /entities/Model/search > post > responses > 200 > content > application/json > schema is used within a registered compoment schema (ModelSearch) and contains an input transformation (ZodDefault) defined at ....: parameters which is also used in an output schema.

Your explanation in the error is clear but not if the transformation is z.default.

I can solve it only by adding optional() but that is not what i want. Can i use some other zod effect to prevent this error?

Thanks for your info...

[samchungy]

https://github.com/samchungy/zod-openapi#zod-effects

G'day, I think I explain it in that section

[RonB]

Thank you,
i have read those docs and understand the problem, but just cannot get to a proper solution. Here is my schema:

import 'zod-openapi/extend';
import { z } from "zod";
import { BaseSchema } from "./Base";

export const ModelParameterSchema = z
  .object({
    temperature: z
      .number()
      .min(0)
      .max(1)
      .step(0.1)
      .default(0.1)
      .describe(
        "“Temperature” is a setting that controls randomness when picking words during text creation. Low values of temperature make the text more predictable and consistent, while high values let more freedom and creativity into the mix, but can also make things less consistent."
      ).optional(),
    top_p: z
      .number()
      .min(1)
      .max(10)
      .step(1)
      .describe(
        "“Top_p” (nucleus sampling) is a setting that decides how many possible words to consider. A high “top_p” value means the model looks at more possible words, even the less likely ones, which makes the generated text more diverse."
      ),
    top_k: z
      .number()
      .min(5)
      .max(100)
      .step(5)
      .default(50)
      .optional()
      .describe(
        "“Top_k“ (Top-k Sampling): It restricts the selection of tokens to the “k” most likely options, based on their probabilities. This prevents the model from considering tokens with very low probabilities, making the output more focused and coherent."
      ),
    repetition_penalty: z
      .number()
      .min(0.1)
      .max(2)
      .step(0.1)
      .default(1.2)
      .optional()
      .describe(
        "Repetition penalty is a factor applied to discourage the model from generating repetitive text or phrases. By adjusting this penalty, users can influence the model's output, reducing the likelihood of it producing redundant or repeated content."
      ),
    max_tokens: z
      .number()
      .int()
      .min(4096)
      .step(4096)
      .default(4096)
      .optional()
      .describe("The maximum number of tokens that a model can generate. This includes the input and output text. As a rule of thumb a token is approximately 3/4 of a word."),
    stop: z.array(z.string()).default([]).optional(),
  })
  .default({
    temperature: 0.2,
    top_p: 5,
    top_k: 50,
    repetition_penalty: 1.2,
    max_tokens: 4096,
    stop: [],
  })
  .describe("ModelParameters");

export type ModelParameterType = z.infer<typeof ModelParameterSchema>;

export const ModelSchema = BaseSchema.extend({
  name: z.string().describe("The full name of thet model.").optional(),
  description: z.string().describe("The description of the model").optional(),
  system: z.string().default("").optional(),
  provider: z
    .enum(["ollama", "openai", "anthropic", "huggingface", "google"])
    .default("ollama")
    .optional(),
  status: z.enum(["", "PULLING", "VERIFYING", "READY", "ERROR", "REMOVED", "SAVED"]).default("").optional(),
  size: z.number().default(0).optional(),
  tag: z.string().optional(),
  tags: z
    .object({
      name: z.string().nullable().optional(),
      info: z.string().nullable().optional(),
      size: z.number().nullable().optional(),
    })
    .array()
    .default([])
    .optional(),
  parameters: ModelParameterSchema,
}).openapi({ effectType: "input" }).describe("Model");

The issue here lies in the parameters object. I have tried to put the openapi({effectType: "input"}) to the toplevel and also to the parameters object itself but keep getting the same error.

Please advise if you have the time

[RonB]

Ok, the problem ws that i cannot add the openapi({effectType: "same"}) to the base schema.
After adding the same schema with .openapi({effectType: "same"}) not to the base but to the schema components it works.