Package software.amazon.awssdk.services.sagemakertrainingsessionruntime.model
package software.amazon.awssdk.services.sagemakertrainingsessionruntime.model
-
ClassDescriptionThe Adam optimizer hyperparameters used during an optimizer step.The configuration for computing log-probabilities of prompt tokens without generating new text.The result of a compute-logprobs operation.The training session is in a state that cannot accept new operations.A single training datum containing model input and optional loss function inputs.A chunk of pre-tokenized text represented as a list of token IDs.The configuration for a forward and backward pass through the model.The result of a forward and backward pass through the model, including computed gradients.The configuration for a forward-only pass through the model.The result of a forward-only pass through the model.An internal service error occurred.The configuration for loading training state from a previously saved checkpoint.The result of loading training state.The configuration for loading model weights for inference and sampling without restoring optimizer state.The result of loading weights for sampling.The tokenized model input composed of ordered encoded text chunks.The configuration for a training operation.The result of a completed training operation.The configuration for an optimizer step that updates model weights using gradients accumulated from previous forward-backward passes.The result of an optimizer step.The specified training session or operation does not exist, or the operation does not belong to the specified training session.The configuration for generating text samples from the model.A single generated sequence from the model.The configuration for a sampling operation.The result of a completed sampling operation.The response containing generated samples and optional log-probabilities.The token usage statistics for a sample generation.The parameters controlling the text generation sampling strategy.The configuration for saving full training state, including model weights and optimizer state.The result of saving full training state.The configuration for saving model weights as a checkpoint.The result of saving model weights.A tensor with data, data type, and shape metadata.The request was throttled.The log-probability for a single token.The token usage statistics for training operations.The request is not valid.