com.volcengine.model.maas.api

接口 Api.ParametersOrBuilder

所有超级接口:

com.google.protobuf.MessageLiteOrBuilder, com.google.protobuf.MessageOrBuilder

所有已知实现类:

Api.Parameters, Api.Parameters.Builder

封闭类:

Api

public static interface Api.ParametersOrBuilder
extends com.google.protobuf.MessageOrBuilder

方法概要

限定符和类型	方法和说明
boolean	getDoSample() Whether or not to use sampling, use greedy decoding otherwise.
float	getFrequencyPenalty() Number between -2.0 and 2.0.
long	getLogprobs() optional int64 logprobs = 12;
long	getMaxNewTokens() The maximum number of tokens to generate, ignoring the number of tokens in the prompt optional int64 max_new_tokens = 6;
long	getMaxPromptTokens() the maximum number of prompt tokens, if prompt tokens length over this limit, it will be truncated as prompt[-max_prompt_tokens:] optional int64 max_prompt_tokens = 11;
long	getMaxTokens() The maximum number of tokens to generate in the char completion.
long	getMinNewTokens() the minimum number of tokens to generate optional int64 min_new_tokens = 10;
float	getPresencePenalty() Number between -2.0 and 2.0.
float	getRepetitionPenalty() The parameter for repetition penalty, from [1.0, 2.0] optional float repetition_penalty = 7;
String	getStop(int index) repeated string stop = 13;
com.google.protobuf.ByteString	getStopBytes(int index) repeated string stop = 13;
int	getStopCount() repeated string stop = 13;
List<String>	getStopList() repeated string stop = 13;
float	getTemperature() Exponential scaling output probability distribution optional float temperature = 1;
long	getTopK() The number of highest probability vocabulary tokens to keep for top-k-filtering.
float	getTopP() An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass optional float top_p = 3;
boolean	hasDoSample() Whether or not to use sampling, use greedy decoding otherwise.
boolean	hasFrequencyPenalty() Number between -2.0 and 2.0.
boolean	hasLogprobs() optional int64 logprobs = 12;
boolean	hasMaxNewTokens() The maximum number of tokens to generate, ignoring the number of tokens in the prompt optional int64 max_new_tokens = 6;
boolean	hasMaxPromptTokens() the maximum number of prompt tokens, if prompt tokens length over this limit, it will be truncated as prompt[-max_prompt_tokens:] optional int64 max_prompt_tokens = 11;
boolean	hasMaxTokens() The maximum number of tokens to generate in the char completion.
boolean	hasMinNewTokens() the minimum number of tokens to generate optional int64 min_new_tokens = 10;
boolean	hasPresencePenalty() Number between -2.0 and 2.0.
boolean	hasRepetitionPenalty() The parameter for repetition penalty, from [1.0, 2.0] optional float repetition_penalty = 7;
boolean	hasTemperature() Exponential scaling output probability distribution optional float temperature = 1;
boolean	hasTopK() The number of highest probability vocabulary tokens to keep for top-k-filtering.
boolean	hasTopP() An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass optional float top_p = 3;

从接口继承的方法 com.google.protobuf.MessageOrBuilder

findInitializationErrors, getAllFields, getDefaultInstanceForType, getDescriptorForType, getField, getInitializationErrorString, getOneofFieldDescriptor, getRepeatedField, getRepeatedFieldCount, getUnknownFields, hasField, hasOneof

从接口继承的方法 com.google.protobuf.MessageLiteOrBuilder

isInitialized

方法详细资料

hasTemperature

boolean hasTemperature()

 Exponential scaling output probability distribution
 

optional float temperature = 1;

返回:

Whether the temperature field is set.

getTemperature

float getTemperature()

 Exponential scaling output probability distribution
 

optional float temperature = 1;

返回:

The temperature.

hasMaxTokens

boolean hasMaxTokens()

 The maximum number of tokens to generate in the char completion.
 

optional int64 max_tokens = 2;

返回:

Whether the maxTokens field is set.

getMaxTokens

long getMaxTokens()

 The maximum number of tokens to generate in the char completion.
 

optional int64 max_tokens = 2;

返回:

The maxTokens.

hasTopP

boolean hasTopP()

 An alternative to sampling with temperature, called nucleus sampling,
 where the model considers the results of the tokens with top_p probability
 mass
 

optional float top_p = 3;

返回:

Whether the topP field is set.

getTopP

float getTopP()

 An alternative to sampling with temperature, called nucleus sampling,
 where the model considers the results of the tokens with top_p probability
 mass
 

optional float top_p = 3;

返回:

The topP.

hasPresencePenalty

boolean hasPresencePenalty()

 Number between -2.0 and 2.0. Positive values penalize new tokens based
 on whether they appear in the text so far, increasing the model's
 likelihood to talk about new topics.
 

optional float presence_penalty = 4;

返回:

Whether the presencePenalty field is set.

getPresencePenalty

float getPresencePenalty()

 Number between -2.0 and 2.0. Positive values penalize new tokens based
 on whether they appear in the text so far, increasing the model's
 likelihood to talk about new topics.
 

optional float presence_penalty = 4;

返回:

The presencePenalty.

hasFrequencyPenalty

boolean hasFrequencyPenalty()

 Number between -2.0 and 2.0. Positive values penalize new tokens based on
 their existing frequency in the text so far, decreasing the model's
 likelihood to repeat the same line verbatim.
 

optional float frequency_penalty = 5;

返回:

Whether the frequencyPenalty field is set.

getFrequencyPenalty

float getFrequencyPenalty()

 Number between -2.0 and 2.0. Positive values penalize new tokens based on
 their existing frequency in the text so far, decreasing the model's
 likelihood to repeat the same line verbatim.
 

optional float frequency_penalty = 5;

返回:

The frequencyPenalty.

hasMaxNewTokens

boolean hasMaxNewTokens()

 The maximum number of tokens to generate, ignoring the number of tokens in
 the prompt
 

optional int64 max_new_tokens = 6;

返回:

Whether the maxNewTokens field is set.

getMaxNewTokens

long getMaxNewTokens()

 The maximum number of tokens to generate, ignoring the number of tokens in
 the prompt
 

optional int64 max_new_tokens = 6;

返回:

The maxNewTokens.

hasRepetitionPenalty

boolean hasRepetitionPenalty()

 The parameter for repetition penalty, from [1.0, 2.0]
 

optional float repetition_penalty = 7;

返回:

Whether the repetitionPenalty field is set.

getRepetitionPenalty

float getRepetitionPenalty()

 The parameter for repetition penalty, from [1.0, 2.0]
 

optional float repetition_penalty = 7;

返回:

The repetitionPenalty.

hasDoSample

boolean hasDoSample()

 Whether or not to use sampling, use greedy decoding otherwise. Default to
 false
 

optional bool do_sample = 8;

返回:

Whether the doSample field is set.

getDoSample

boolean getDoSample()

 Whether or not to use sampling, use greedy decoding otherwise. Default to
 false
 

optional bool do_sample = 8;

返回:

The doSample.

hasTopK

boolean hasTopK()

 The number of highest probability vocabulary tokens to keep for
 top-k-filtering.
 

optional int64 top_k = 9;

返回:

Whether the topK field is set.

getTopK

long getTopK()

 The number of highest probability vocabulary tokens to keep for
 top-k-filtering.
 

optional int64 top_k = 9;

返回:

The topK.

hasMinNewTokens

boolean hasMinNewTokens()

 the minimum number of tokens to generate
 

optional int64 min_new_tokens = 10;

返回:

Whether the minNewTokens field is set.

getMinNewTokens

long getMinNewTokens()

 the minimum number of tokens to generate
 

optional int64 min_new_tokens = 10;

返回:

The minNewTokens.

hasMaxPromptTokens

boolean hasMaxPromptTokens()

 the maximum number of prompt tokens, if prompt tokens length over this
 limit, it will be truncated as prompt[-max_prompt_tokens:]
 

optional int64 max_prompt_tokens = 11;

返回:

Whether the maxPromptTokens field is set.

getMaxPromptTokens

long getMaxPromptTokens()

 the maximum number of prompt tokens, if prompt tokens length over this
 limit, it will be truncated as prompt[-max_prompt_tokens:]
 

optional int64 max_prompt_tokens = 11;

返回:

The maxPromptTokens.

hasLogprobs

boolean hasLogprobs()

optional int64 logprobs = 12;

返回:

Whether the logprobs field is set.

getLogprobs

long getLogprobs()

optional int64 logprobs = 12;

返回:

The logprobs.

getStopList

List<String> getStopList()

repeated string stop = 13;

返回:

A list containing the stop.

getStopCount

int getStopCount()

repeated string stop = 13;

返回:

The count of stop.

getStop

String getStop(int index)

repeated string stop = 13;

参数:

index - The index of the element to return.

返回:

The stop at the given index.

getStopBytes

com.google.protobuf.ByteString getStopBytes(int index)

repeated string stop = 13;

参数:

index - The index of the value to return.

返回:

The bytes of the stop at the given index.