IT 기술 따라잡기

The input text to classify

분류할 입력 텍스트

Two content moderations models are available: text-moderation-stable and text-moderation-latest.

두 가지 콘텐츠 조정 모델(text-moderation-stable 및 text-moderation-latest)을 사용할 수 있습니다.

The default is text-moderation-latest which will be automatically upgraded over time. This ensures you are always using our most accurate model. If you use text-moderation-stable, we will provide advanced notice before updating the model. Accuracy of text-moderation-stable may be slightly lower than for text-moderation-latest.

기본값은 시간이 지남에 따라 자동으로 업그레이드되는 text-moderation-latest입니다. 이렇게 하면 항상 가장 정확한 모델을 사용할 수 있습니다. text-moderation-stable을 사용할 경우 모델 업데이트 전 사전 공지를 드립니다. text-moderation-stable의 정확도는 text-moderation-latest보다 약간 낮을 수 있습니다.

The ID of an uploaded file that contains training data.

교육 데이터가 포함된 업로드된 파일의 ID입니다.

See upload file for how to upload a file.

파일 업로드 방법은 파일 업로드를 참조하세요.

Your dataset must be formatted as a JSONL file, where each training example is a JSON object with the keys "prompt" and "completion". Additionally, you must upload your file with the purpose fine-tune.

데이터 세트는 JSONL 파일로 형식화되어야 합니다. 여기서 각 교육 예제는 "prompt" 및 "completion" 키가 있는 JSON 개체입니다. 또한 미세 조정 목적으로 파일을 업로드해야 합니다.

See the fine-tuning guide for more details.

자세한 내용은 미세 조정 가이드를 참조하세요.

The ID of an uploaded file that contains validation data.

유효성 검사 데이터가 포함된 업로드된 파일의 ID입니다.

If you provide this file, the data is used to generate validation metrics periodically during fine-tuning. These metrics can be viewed in the fine-tuning results file. Your train and validation data should be mutually exclusive.

이 파일을 제공하면 미세 조정 중에 정기적으로 유효성 검사 지표를 생성하는 데 데이터가 사용됩니다. 이러한 메트릭은 미세 조정 결과 파일에서 볼 수 있습니다. 학습 및 검증 데이터는 상호 배타적이어야 합니다.

Your dataset must be formatted as a JSONL file, where each validation example is a JSON object with the keys "prompt" and "completion". Additionally, you must upload your file with the purpose fine-tune.

데이터 세트는 JSONL 파일로 형식화되어야 합니다. 여기서 각 검증 예제는 "prompt" 및 "completion" 키가 있는 JSON 개체입니다. 또한 미세 조정 목적으로 파일을 업로드해야 합니다.

See the fine-tuning guide for more details.

자세한 내용은 미세 조정 가이드를 참조하세요.

The name of the base model to fine-tune. You can select one of "ada", "babbage", "curie", "davinci", or a fine-tuned model created after 2022-04-21. To learn more about these models, see the Models documentation.

미세 조정할 기본 모델의 이름입니다. "ada", "babbage", "curie", "davinci" 또는 2022-04-21 이후 생성된 미세 조정된 모델 중 하나를 선택할 수 있습니다. 이러한 모델에 대한 자세한 내용은 모델 설명서를 참조하십시오.

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

모델을 훈련할 에포크 수입니다. 에포크는 교육 데이터 세트를 통한 하나의 전체 주기를 나타냅니다.

The batch size to use for training. The batch size is the number of training examples used to train a single forward and backward pass.

교육에 사용할 배치 크기입니다. 배치 크기는 단일 정방향 및 역방향 패스를 훈련하는 데 사용되는 훈련 예제의 수입니다.

By default, the batch size will be dynamically configured to be ~0.2% of the number of examples in the training set, capped at 256 - in general, we've found that larger batch sizes tend to work better for larger datasets.

기본적으로 배치 크기는 훈련 세트에 있는 예제 수의 ~0.2%로 동적으로 구성되며 최대 256개로 제한됩니다. 일반적으로 배치 크기가 클수록 데이터 세트가 더 잘 작동하는 경향이 있습니다.

The learning rate multiplier to use for training. The fine-tuning learning rate is the original learning rate used for pretraining multiplied by this value.

훈련에 사용할 학습률 승수입니다. 미세 조정 학습률은 사전 훈련에 사용된 원래 학습률에 이 값을 곱한 것입니다.

By default, the learning rate multiplier is the 0.05, 0.1, or 0.2 depending on final batch_size (larger learning rates tend to perform better with larger batch sizes). We recommend experimenting with values in the range 0.02 to 0.2 to see what produces the best results.

기본적으로 학습률 승수는 최종 batch_size에 따라 0.05, 0.1 또는 0.2입니다(배치 크기가 클수록 학습률이 높을수록 더 잘 수행되는 경향이 있음). 0.02에서 0.2 범위의 값으로 실험하여 최상의 결과를 생성하는 것이 무엇인지 확인하는 것이 좋습니다.

The weight to use for loss on the prompt tokens. This controls how much the model tries to learn to generate the prompt (as compared to the completion which always has a weight of 1.0), and can add a stabilizing effect to training when completions are short.

프롬프트 토큰에서 손실에 사용할 가중치입니다. 이것은 모델이 프롬프트를 생성하기 위해 학습을 시도하는 정도를 제어하고(항상 가중치가 1.0인 완료와 비교하여) 완료가 짧을 때 훈련에 안정화 효과를 추가할 수 있습니다.

If prompts are extremely long (relative to completions), it may make sense to reduce this weight so as to avoid over-prioritizing learning the prompt.

프롬프트가 매우 긴 경우(완료에 비해) 프롬프트 학습에 과도한 우선순위를 두지 않도록 이 가중치를 줄이는 것이 좋습니다.

If set, we calculate classification-specific metrics such as accuracy and F-1 score using the validation set at the end of every epoch. These metrics can be viewed in the results file.

설정된 경우 매 에포크가 끝날 때마다 검증 세트를 사용하여 정확도 및 F-1 점수와 같은 분류별 메트릭을 계산합니다. 이러한 메트릭은 결과 파일에서 볼 수 있습니다.

In order to compute classification metrics, you must provide a validation_file. Additionally, you must specify classification_n_classes for multiclass classification or classification_positive_class for binary classification.

분류 메트릭을 계산하려면 validation_file을 제공해야 합니다. 또한 다중 클래스 분류의 경우 classification_n_classes를 지정하고 이진 분류의 경우 classification_positive_class를 지정해야 합니다.

The positive class in binary classification.

이진 분류의 포지티브 클래스입니다.

This parameter is needed to generate precision, recall, and F1 metrics when doing binary classification.

이 매개변수는 이진 분류를 수행할 때 정밀도, 재현율 및 F1 메트릭을 생성하는 데 필요합니다.

If this is provided, we calculate F-beta scores at the specified beta values. The F-beta score is a generalization of F-1 score. This is only used for binary classification.

이것이 제공되면 지정된 베타 값에서 F-베타 점수를 계산합니다. F-베타 점수는 F-1 점수를 일반화한 것입니다. 이진 분류에만 사용됩니다.

With a beta of 1 (i.e. the F-1 score), precision and recall are given the same weight. A larger beta score puts more weight on recall and less on precision. A smaller beta score puts more weight on precision and less on recall.

베타 1(즉, F-1 점수)에서는 정밀도와 재현율에 동일한 가중치가 부여됩니다. 베타 점수가 클수록 재현율에 더 많은 가중치를 부여하고 정밀도에는 덜 적용합니다. 베타 점수가 작을수록 정밀도에 더 많은 가중치를 부여하고 재현율에 더 적은 가중치를 둡니다.

A string of up to 40 characters that will be added to your fine-tuned model name.

미세 조정된 모델 이름에 추가될 최대 40자의 문자열입니다.

For example, a suffix of "custom-model-name" would produce a model name like ada:ft-your-org:custom-model-name-2022-02-15-04-21-04.

예를 들어 "custom-model-name" 접미사는 ada:ft-your-org:custom-model-name-2022-02-15-04-21-04와 같은 모델 이름을 생성합니다.

ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.

사용할 모델의 ID입니다. 모델 목록 API를 사용하여 사용 가능한 모든 모델을 보거나 모델 개요에서 설명을 볼 수 있습니다.

Input text to get embeddings for, encoded as a string or array of tokens. To get embeddings for multiple inputs in a single request, pass an array of strings or array of token arrays. Each input must not exceed 8192 tokens in length.

문자열 또는 토큰 배열로 인코딩된 임베딩을 가져올 입력 텍스트입니다. 단일 요청에서 여러 입력에 대한 임베딩을 가져오려면 문자열 배열 또는 토큰 배열 배열을 전달합니다. 각 입력은 길이가 8192 토큰을 초과할 수 없습니다.

A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. Learn more.

A text description of the desired image(s). The maximum length is 1000 characters.

원하는 이미지에 대한 텍스트 설명입니다. 최대 길이는 1000자입니다.

The number of images to generate. Must be between 1 and 10.

생성할 이미지 수입니다. 1에서 10 사이여야 합니다.

The size of the generated images. Must be one of 256x256, 512x512, or 1024x1024.

생성된 이미지의 크기입니다. 256x256, 512x512 또는 1024x1024 중 하나여야 합니다.

The format in which the generated images are returned. Must be one of url or b64_json.

생성된 이미지가 반환되는 형식입니다. url 또는 b64_json 중 하나여야 합니다.

A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. Learn more.

OpenAI가 남용을 모니터링하고 탐지하는 데 도움이 될 수 있는 최종 사용자를 나타내는 고유 식별자입니다. 더 알아보기.

https://beta.openai.com/docs/api-reference/images/create-edit

Create image edit

POST https://api.openai.com/v1/images/edits

Creates an edited or extended image given an original image and a prompt.

원본 이미지와 프롬프트가 주어지면 편집되거나 확장된 이미지를 생성합니다.