feat: Add support for minP, repeatLastN, and repeatPenalty model settings
This commit is contained in:
@@ -60,11 +60,11 @@
|
||||
"keepAlive": {
|
||||
"label": "Keep Alive",
|
||||
"help": "controls how long the model will stay loaded into memory following the request (default: 5m)",
|
||||
"placeholder": "Enter Keep Alive duration (e.g. 5m, 10m, 1h)"
|
||||
"placeholder": "e.g. 5m, 10m, 1h"
|
||||
},
|
||||
"temperature": {
|
||||
"label": "Temperature",
|
||||
"placeholder": "Enter Temperature value (e.g. 0.7, 1.0)"
|
||||
"placeholder": "e.g. 0.7, 1.0"
|
||||
},
|
||||
"numCtx": {
|
||||
"label": "Number of Contexts",
|
||||
@@ -72,24 +72,36 @@
|
||||
},
|
||||
"numPredict": {
|
||||
"label": "Max Tokens (num_predict)",
|
||||
"placeholder": "Enter Max Tokens value (e.g. 2048, 4096)"
|
||||
"placeholder": "e.g. 2048, 4096"
|
||||
},
|
||||
"seed": {
|
||||
"label": "Seed",
|
||||
"placeholder": "Enter Seed value (e.g. 1234)",
|
||||
"placeholder": "e.g. 1234",
|
||||
"help": "Reproducibility of the model output"
|
||||
},
|
||||
"topK": {
|
||||
"label": "Top K",
|
||||
"placeholder": "Enter Top K value (e.g. 40, 100)"
|
||||
"placeholder": "e.g. 40, 100"
|
||||
},
|
||||
"topP": {
|
||||
"label": "Top P",
|
||||
"placeholder": "Enter Top P value (e.g. 0.9, 0.95)"
|
||||
"placeholder": "e.g. 0.9, 0.95"
|
||||
},
|
||||
"useMMap": {
|
||||
"label": "useMmap"
|
||||
},
|
||||
"minP": {
|
||||
"label": "Min P",
|
||||
"placeholder": "e.g. 0.05"
|
||||
},
|
||||
"repeatPenalty": {
|
||||
"label": "Repeat Penalty",
|
||||
"placeholder": "e.g. 1.1, 1.2"
|
||||
},
|
||||
"repeatLastN": {
|
||||
"label": "Repeat Last N",
|
||||
"placeholder": "e.g. 64, 128"
|
||||
},
|
||||
"numGpu": {
|
||||
"label": "Num GPU",
|
||||
"placeholder": "Enter number of layers to send to GPU(s)"
|
||||
@@ -125,4 +137,4 @@
|
||||
"pin": "Pin",
|
||||
"unpin": "Unpin",
|
||||
"generationInfo": "Generation Info"
|
||||
}
|
||||
}
|
||||
Reference in New Issue
Block a user