Methods

Domain Types

Dpo Hyperparameters

• DpoHyperparameters object { batch_size, beta, learning_rate_multiplier, n_epochs }

  The hyperparameters used for the DPO fine-tuning job.

  • batch_size: optional "auto" or number

    Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.

    • "auto"

      • "auto"

    • number

  • beta: optional "auto" or number

    The beta value for the DPO method. A higher beta value will increase the weight of the penalty between the policy and reference model.

    • "auto"

      • "auto"

    • number

  • learning_rate_multiplier: optional "auto" or number

    Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.

    • "auto"

      • "auto"

    • number

  • n_epochs: optional "auto" or number

    The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.

    • "auto"

      • "auto"

    • number

Dpo Method

• DpoMethod object { hyperparameters }

  Configuration for the DPO fine-tuning method.

  • hyperparameters: optional DpoHyperparameters

    The hyperparameters used for the DPO fine-tuning job.

    • batch_size: optional "auto" or number

      Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.

      • "auto"

        • "auto"

      • number

    • beta: optional "auto" or number

      The beta value for the DPO method. A higher beta value will increase the weight of the penalty between the policy and reference model.

      • "auto"

        • "auto"

      • number

    • learning_rate_multiplier: optional "auto" or number

      Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.

      • "auto"

        • "auto"

      • number

    • n_epochs: optional "auto" or number

      The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.

      • "auto"

        • "auto"

      • number

Reinforcement Hyperparameters

• ReinforcementHyperparameters object { batch_size, compute_multiplier, eval_interval, 4 more }

  The hyperparameters used for the reinforcement fine-tuning job.

  • batch_size: optional "auto" or number

    Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.

    • "auto"

      • "auto"

    • number

  • compute_multiplier: optional "auto" or number

    Multiplier on amount of compute used for exploring search space during training.

    • "auto"

      • "auto"

    • number

  • eval_interval: optional "auto" or number

    The number of training steps between evaluation runs.

    • "auto"

      • "auto"

    • number

  • eval_samples: optional "auto" or number

    Number of evaluation samples to generate per training step.

    • "auto"

      • "auto"

    • number

  • learning_rate_multiplier: optional "auto" or number

    Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.

    • "auto"

      • "auto"

    • number

  • n_epochs: optional "auto" or number

    The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.

    • "auto"

      • "auto"

    • number

  • reasoning_effort: optional "default" or "low" or "medium" or "high"

    Level of reasoning effort.

    • "default"

    • "low"

    • "medium"

    • "high"

Reinforcement Method

• ReinforcementMethod object { grader, hyperparameters }

  Configuration for the reinforcement fine-tuning method.

  • grader: StringCheckGrader or TextSimilarityGrader or PythonGrader or 2 more

    The grader used for the fine-tuning job.

    • StringCheckGrader object { input, name, operation, 2 more }

      A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.

      • input: string

        The input text. This may include template strings.

      • name: string

        The name of the grader.

      • operation: "eq" or "ne" or "like" or "ilike"

        The string check operation to perform. One of eq, ne, like, or ilike.

        • "eq"

        • "ne"

        • "like"

        • "ilike"

      • reference: string

        The reference text. This may include template strings.

      • type: "string_check"

        The object type, which is always string_check.

        • "string_check"

    • TextSimilarityGrader object { evaluation_metric, input, name, 2 more }

      A TextSimilarityGrader object which grades text based on similarity metrics.

      • evaluation_metric: "cosine" or "fuzzy_match" or "bleu" or 8 more

        The evaluation metric to use. One of cosine, fuzzy_match, bleu, gleu, meteor, rouge_1, rouge_2, rouge_3, rouge_4, rouge_5, or rouge_l.

        • "cosine"

        • "fuzzy_match"

        • "bleu"

        • "gleu"

        • "meteor"

        • "rouge_1"

        • "rouge_2"

        • "rouge_3"

        • "rouge_4"

        • "rouge_5"

        • "rouge_l"

      • input: string

        The text being graded.

      • name: string

        The name of the grader.

      • reference: string

        The text being graded against.

      • type: "text_similarity"

        The type of grader.

        • "text_similarity"

    • PythonGrader object { name, source, type, image_tag }

      A PythonGrader object that runs a python script on the input.

      • name: string

        The name of the grader.

      • source: string

        The source code of the python script.

      • type: "python"

        The object type, which is always python.

        • "python"

      • image_tag: optional string

        The image tag to use for the python script.

    • ScoreModelGrader object { input, model, name, 3 more }

      A ScoreModelGrader object that uses a model to assign a score to the input.

      • input: array of object { content, role, type }

        The input messages evaluated by the grader. Supports text, output text, input image, and input audio content blocks, and may include template strings.

        • content: string or ResponseInputText or object { text, type } or 3 more

          Inputs to the model - can contain template strings. Supports text, output text, input images, and input audio, either as a single item or an array of items.

          • TextInput = string

            A text input to the model.

          • ResponseInputText object { text, type, prompt_cache_breakpoint }

            A text input to the model.

            • text: string

              The text input to the model.

            • type: "input_text"

              The type of the input item. Always input_text.

              • "input_text"

            • prompt_cache_breakpoint: optional object { mode }

              Marks the exact end of a reusable prompt prefix. The breakpoint inherits its TTL from the request's prompt_cache_options.ttl; the boundary is not rounded to a token block.

              • mode: "explicit"

                The breakpoint mode. Always explicit.

                • "explicit"

          • OutputText object { text, type }

            A text output from the model.

            • text: string

              The text output from the model.

            • type: "output_text"

              The type of the output text. Always output_text.

              • "output_text"

          • InputImage object { image_url, type, detail }

            An image input block used within EvalItem content arrays.

            • image_url: string

              The URL of the image input.

            • type: "input_image"

              The type of the image input. Always input_image.

              • "input_image"

            • detail: optional string

              The detail level of the image to be sent to the model. One of high, low, or auto. Defaults to auto.

          • ResponseInputAudio object { input_audio, type }

            An audio input to the model.

            • input_audio: object { data, format }

              • data: string

                Base64-encoded audio data.

              • format: "mp3" or "wav"

                The format of the audio data. Currently supported formats are mp3 and wav.

                • "mp3"

                • "wav"

            • type: "input_audio"

              The type of the input item. Always input_audio.

              • "input_audio"

          • GraderInputs = array of string or ResponseInputText or object { text, type } or 2 more

            A list of inputs, each of which may be either an input text, output text, input image, or input audio object.

            • TextInput = string

              A text input to the model.

            • ResponseInputText object { text, type, prompt_cache_breakpoint }

              A text input to the model.

            • OutputText object { text, type }

              A text output from the model.

              • text: string

                The text output from the model.

              • type: "output_text"

                The type of the output text. Always output_text.

                • "output_text"

            • InputImage object { image_url, type, detail }

              An image input block used within EvalItem content arrays.

              • image_url: string

                The URL of the image input.

              • type: "input_image"

                The type of the image input. Always input_image.

                • "input_image"

              • detail: optional string

                The detail level of the image to be sent to the model. One of high, low, or auto. Defaults to auto.

            • ResponseInputAudio object { input_audio, type }

              An audio input to the model.

        • role: "user" or "assistant" or "system" or "developer"

          The role of the message input. One of user, assistant, system, or developer.

          • "user"

          • "assistant"

          • "system"

          • "developer"

        • type: optional "message"

          The type of the message input. Always message.

          • "message"

      • model: string

        The model to use for the evaluation.

      • name: string

        The name of the grader.

      • type: "score_model"

        The object type, which is always score_model.

        • "score_model"

      • range: optional array of number

        The range of the score. Defaults to [0, 1].

      • sampling_params: optional object { max_completions_tokens, reasoning_effort, seed, 2 more }

        The sampling parameters for the model.

        • max_completions_tokens: optional number or null

          The maximum number of tokens the grader model may generate in its response.

        • reasoning_effort: optional ReasoningEffort or null

          Constrains effort on reasoning for reasoning models. Currently supported values are none, minimal, low, medium, high, xhigh, and max. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response. Not all reasoning models support every value. See the reasoning guide for model-specific support.

          • "none"

          • "minimal"

          • "low"

          • "medium"

          • "high"

          • "xhigh"

          • "max"

        • seed: optional number or null

          A seed value to initialize the randomness, during sampling.

        • temperature: optional number or null

          A higher temperature increases randomness in the outputs.

        • top_p: optional number or null

          An alternative to temperature for nucleus sampling; 1.0 includes all tokens.

    • MultiGrader object { calculate_output, graders, name, type }

      A MultiGrader object combines the output of multiple graders to produce a single score.

      • calculate_output: string

        A formula to calculate the output based on grader results.

      • graders: StringCheckGrader or TextSimilarityGrader or PythonGrader or 2 more

        A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.

        • StringCheckGrader object { input, name, operation, 2 more }

          A StringCheckGrader object that performs a string comparison between input and reference using a specified operation.

        • TextSimilarityGrader object { evaluation_metric, input, name, 2 more }

          A TextSimilarityGrader object which grades text based on similarity metrics.

        • PythonGrader object { name, source, type, image_tag }

          A PythonGrader object that runs a python script on the input.

        • ScoreModelGrader object { input, model, name, 3 more }

          A ScoreModelGrader object that uses a model to assign a score to the input.

        • LabelModelGrader object { input, labels, model, 3 more }

          A LabelModelGrader object which uses a model to assign labels to each item in the evaluation.

          • input: array of object { content, role, type }

            • content: string or ResponseInputText or object { text, type } or 3 more

              Inputs to the model - can contain template strings. Supports text, output text, input images, and input audio, either as a single item or an array of items.

              • TextInput = string

                A text input to the model.

              • ResponseInputText object { text, type, prompt_cache_breakpoint }

                A text input to the model.

              • OutputText object { text, type }

                A text output from the model.

                • text: string

                  The text output from the model.

                • type: "output_text"

                  The type of the output text. Always output_text.

                  • "output_text"

              • InputImage object { image_url, type, detail }

                An image input block used within EvalItem content arrays.

                • image_url: string

                  The URL of the image input.

                • type: "input_image"

                  The type of the image input. Always input_image.

                  • "input_image"

                • detail: optional string

                  The detail level of the image to be sent to the model. One of high, low, or auto. Defaults to auto.

              • ResponseInputAudio object { input_audio, type }

                An audio input to the model.

              • GraderInputs = array of string or ResponseInputText or object { text, type } or 2 more

                A list of inputs, each of which may be either an input text, output text, input image, or input audio object.

            • role: "user" or "assistant" or "system" or "developer"

              The role of the message input. One of user, assistant, system, or developer.

              • "user"

              • "assistant"

              • "system"

              • "developer"

            • type: optional "message"

              The type of the message input. Always message.

              • "message"

          • labels: array of string

            The labels to assign to each item in the evaluation.

          • model: string

            The model to use for the evaluation. Must support structured outputs.

          • name: string

            The name of the grader.

          • passing_labels: array of string

            The labels that indicate a passing result. Must be a subset of labels.

          • type: "label_model"

            The object type, which is always label_model.

            • "label_model"

      • name: string

        The name of the grader.

      • type: "multi"

        The object type, which is always multi.

        • "multi"

  • hyperparameters: optional ReinforcementHyperparameters

    The hyperparameters used for the reinforcement fine-tuning job.

    • batch_size: optional "auto" or number

      Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.

      • "auto"

        • "auto"

      • number

    • compute_multiplier: optional "auto" or number

      Multiplier on amount of compute used for exploring search space during training.

      • "auto"

        • "auto"

      • number

    • eval_interval: optional "auto" or number

      The number of training steps between evaluation runs.

      • "auto"

        • "auto"

      • number

    • eval_samples: optional "auto" or number

      Number of evaluation samples to generate per training step.

      • "auto"

        • "auto"

      • number

    • learning_rate_multiplier: optional "auto" or number

      Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.

      • "auto"

        • "auto"

      • number

    • n_epochs: optional "auto" or number

      The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.

      • "auto"

        • "auto"

      • number

    • reasoning_effort: optional "default" or "low" or "medium" or "high"

      Level of reasoning effort.

      • "default"

      • "low"

      • "medium"

      • "high"

Supervised Hyperparameters

• SupervisedHyperparameters object { batch_size, learning_rate_multiplier, n_epochs }

  The hyperparameters used for the fine-tuning job.

  • batch_size: optional "auto" or number

    Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.

    • "auto"

      • "auto"

    • number

  • learning_rate_multiplier: optional "auto" or number

    Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.

    • "auto"

      • "auto"

    • number

  • n_epochs: optional "auto" or number

    The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.

    • "auto"

      • "auto"

    • number

Supervised Method

• SupervisedMethod object { hyperparameters }

  Configuration for the supervised fine-tuning method.

  • hyperparameters: optional SupervisedHyperparameters

    The hyperparameters used for the fine-tuning job.

    • batch_size: optional "auto" or number

      Number of examples in each batch. A larger batch size means that model parameters are updated less frequently, but with lower variance.

      • "auto"

        • "auto"

      • number

    • learning_rate_multiplier: optional "auto" or number

      Scaling factor for the learning rate. A smaller learning rate may be useful to avoid overfitting.

      • "auto"

        • "auto"

      • number

    • n_epochs: optional "auto" or number

      The number of epochs to train the model for. An epoch refers to one full cycle through the training dataset.

      • "auto"

        • "auto"

      • number