@huggingface/transformers 3.0.0-alpha.0

package/types/models/whisper/generation_whisper.d.ts

@@ -0,0 +1,76 @@
+export class WhisperGenerationConfig extends GenerationConfig {
+    /**
+     * Whether to return the timestamps with the text. This enables the `WhisperTimestampsLogitsProcessor`.
+     * @type {boolean}
+     */
+    return_timestamps: boolean;
+    /**
+     * Whether to return token-level timestamps
+     * with the text. This can be used with or without the `return_timestamps` option. To get word-level
+     * timestamps, use the tokenizer to group the tokens into words.
+     * @type {boolean}
+     */
+    return_token_timestamps: boolean;
+    /**
+     * The number of audio frames available in this chunk. This is only used generating word-level timestamps.
+     * @type {number}
+     */
+    num_frames: number;
+    /**
+     * Alignment heads to predict word-level timestamps. This is a list of [layer, head] pairs that
+     * select the cross-attention heads that are highly correlated to word-level timing.
+     * @type {[number, number][]}
+     */
+    alignment_heads: [number, number][];
+    /**
+     * Task to use for generation, either "translate" or "transcribe".
+     * @type {string}
+     */
+    task: string;
+    /**
+     * Language token to use for generation, can be either in the form of `<|en|>`, `en` or `english`.
+     * You can find all the possible language tokens in the `model.generation_config.lang_to_id` dictionary.
+     * @type {string}
+     */
+    language: string;
+    /**
+     * The id of the `"<|notimestamps|>"` token.
+     * @type {number}
+     */
+    no_timestamps_token_id: number;
+    /**
+     * Rank-1 list of token IDs created by passing text to [`~WhisperProcessor.get_prompt_ids`] that is
+     * provided as a prompt to each chunk. This can be used to provide or "prompt-engineer" a context for
+     * transcription, e.g. custom vocabularies or proper nouns to make it more likely to predict those words
+     * correctly. It cannot be used in conjunction with `decoder_start_token_id` as it overwrites this value.
+     * @type {number[]}
+     */
+    prompt_ids: number[];
+    /**
+     * Whether the model is multilingual or not.
+     * @type {boolean}
+     */
+    is_multilingual: boolean;
+    /**
+     * (Optional) A mapping from language tokens to their corresponding IDs.
+     * Only required if the model is multilingual.
+     * @type {Record<string, number>|null}
+     */
+    lang_to_id: Record<string, number> | null;
+    /**
+     * (Optional) A mapping from task tokens to their corresponding IDs.
+     * @type {Record<string, number>|null}
+     */
+    task_to_id: Record<string, number> | null;
+    /**
+     * Used to set the maximum value of the initial timestamp. This is used to prevent the model from
+     * predicting timestamps that are too far in the future.
+     * @type {number}
+     */
+    max_initial_timestamp_index: number;
+}
+export type WhisperGenerationFunctionParameters = any & {
+    generation_config: WhisperGenerationConfig;
+} & WhisperGenerationConfig;
+import { GenerationConfig } from "../../generation/configuration_utils.js";
+//# sourceMappingURL=generation_whisper.d.ts.map

package/types/models/whisper/generation_whisper.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"generation_whisper.d.ts","sourceRoot":"","sources":["../../../src/models/whisper/generation_whisper.js"],"names":[],"mappings":"AAEA;IAEI;;;OAGG;IACH,mBAFU,OAAO,CAEQ;IAEzB;;;;;OAKG;IACH,yBAFU,OAAO,CAEc;IAE/B;;;OAGG;IACH,YAFU,MAAM,CAEE;IAElB;;;;OAIG;IACH,iBAFU,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,CAEL;IAEvB;;;OAGG;IACH,MAFU,MAAM,CAEJ;IAEZ;;;;OAIG;IACH,UAFU,MAAM,CAEA;IAEhB;;;OAGG;IACH,wBAFU,MAAM,CAEc;IAE9B;;;;;;OAMG;IACH,YAFU,MAAM,EAAE,CAEA;IAElB;;;OAGG;IACH,iBAFU,OAAO,CAEM;IAEvB;;;;OAIG;IACH,YAFU,OAAO,MAAM,EAAE,MAAM,CAAC,GAAC,IAAI,CAEnB;IAElB;;;OAGG;IACH,YAFU,OAAO,MAAM,EAAE,MAAM,CAAC,GAAC,IAAI,CAEnB;IAElB;;;;OAIG;IACH,6BAFU,MAAM,CAEgB;CACnC;kDAGY,MAAwE;IAAC,iBAAiB,EAAE,uBAAuB,CAAA;CAAC,GAAG,uBAAuB;iCAvF1H,yCAAyC"}