Class GoogleCloudAiplatformV1beta1SchemaModelevaluationMetricsPairwiseTextGenerationEvaluationMetrics
Metrics for general pairwise text generation evaluation results.
Inheritance
Implements
Inherited Members
Namespace: Google.Apis.Aiplatform.v1beta1.Data
Assembly: Google.Apis.Aiplatform.v1beta1.dll
Syntax
public class GoogleCloudAiplatformV1beta1SchemaModelevaluationMetricsPairwiseTextGenerationEvaluationMetrics : IDirectResponseSchema
Properties
Accuracy
Fraction of cases where the autorater agreed with the human raters.
Declaration
[JsonProperty("accuracy")]
public virtual float? Accuracy { get; set; }
Property Value
Type | Description |
---|---|
float? |
BaselineModelWinRate
Percentage of time the autorater decided the baseline model had the better response.
Declaration
[JsonProperty("baselineModelWinRate")]
public virtual float? BaselineModelWinRate { get; set; }
Property Value
Type | Description |
---|---|
float? |
CohensKappa
A measurement of agreement between the autorater and human raters that takes the likelihood of random agreement into account.
Declaration
[JsonProperty("cohensKappa")]
public virtual float? CohensKappa { get; set; }
Property Value
Type | Description |
---|---|
float? |
ETag
The ETag of the item.
Declaration
public virtual string ETag { get; set; }
Property Value
Type | Description |
---|---|
string |
F1Score
Harmonic mean of precision and recall.
Declaration
[JsonProperty("f1Score")]
public virtual float? F1Score { get; set; }
Property Value
Type | Description |
---|---|
float? |
FalseNegativeCount
Number of examples where the autorater chose the baseline model, but humans preferred the model.
Declaration
[JsonProperty("falseNegativeCount")]
public virtual long? FalseNegativeCount { get; set; }
Property Value
Type | Description |
---|---|
long? |
FalsePositiveCount
Number of examples where the autorater chose the model, but humans preferred the baseline model.
Declaration
[JsonProperty("falsePositiveCount")]
public virtual long? FalsePositiveCount { get; set; }
Property Value
Type | Description |
---|---|
long? |
HumanPreferenceBaselineModelWinRate
Percentage of time humans decided the baseline model had the better response.
Declaration
[JsonProperty("humanPreferenceBaselineModelWinRate")]
public virtual float? HumanPreferenceBaselineModelWinRate { get; set; }
Property Value
Type | Description |
---|---|
float? |
HumanPreferenceModelWinRate
Percentage of time humans decided the model had the better response.
Declaration
[JsonProperty("humanPreferenceModelWinRate")]
public virtual float? HumanPreferenceModelWinRate { get; set; }
Property Value
Type | Description |
---|---|
float? |
ModelWinRate
Percentage of time the autorater decided the model had the better response.
Declaration
[JsonProperty("modelWinRate")]
public virtual float? ModelWinRate { get; set; }
Property Value
Type | Description |
---|---|
float? |
Precision
Fraction of cases where the autorater and humans thought the model had a better response out of all cases where the autorater thought the model had a better response. True positive divided by all positive.
Declaration
[JsonProperty("precision")]
public virtual float? Precision { get; set; }
Property Value
Type | Description |
---|---|
float? |
Recall
Fraction of cases where the autorater and humans thought the model had a better response out of all cases where the humans thought the model had a better response.
Declaration
[JsonProperty("recall")]
public virtual float? Recall { get; set; }
Property Value
Type | Description |
---|---|
float? |
TrueNegativeCount
Number of examples where both the autorater and humans decided that the model had the worse response.
Declaration
[JsonProperty("trueNegativeCount")]
public virtual long? TrueNegativeCount { get; set; }
Property Value
Type | Description |
---|---|
long? |
TruePositiveCount
Number of examples where both the autorater and humans decided that the model had the better response.
Declaration
[JsonProperty("truePositiveCount")]
public virtual long? TruePositiveCount { get; set; }
Property Value
Type | Description |
---|---|
long? |