Type Alias: LLamaContextualRepeatPenalty

type LLamaContextualRepeatPenalty = {
  lastTokens: number;
  punishTokensFilter: (tokens: Token[]) => Token[];
  penalizeNewLine: boolean;
  penalty: number;
  frequencyPenalty: number;
  presencePenalty: number;
};

Defined in: types.ts:214

Properties

lastTokens?

optional lastTokens: number;

Defined in: types.ts:219

Number of recent tokens generated by the model to apply penalties to repetition of. Defaults to 64.

---

punishTokensFilter()?

optional punishTokensFilter: (tokens: Token[]) => Token[];

Defined in: types.ts:221

Parameters

Parameter	Type
tokens	Token[]

Returns

Token[]

---

penalizeNewLine?

optional penalizeNewLine: boolean;

Defined in: types.ts:227

Penalize new line tokens. Enabled by default.

---

penalty?

optional penalty: number;

Defined in: types.ts:234

The relative amount to lower the probability of the tokens in punishTokens by Defaults to 1.1. Set to 1 to disable.

---

frequencyPenalty?

optional frequencyPenalty: number;

Defined in: types.ts:241

For n time a token is in the punishTokens array, lower its probability by n * frequencyPenalty Disabled by default (0). Set to a value between 0 and 1 to enable.

---

presencePenalty?

optional presencePenalty: number;

Defined in: types.ts:248

Lower the probability of all the tokens in the punishTokens array by presencePenalty Disabled by default (0). Set to a value between 0 and 1 to enable.