@huggingface/transformers 3.4.2 → 3.5.0

package/dist/transformers.node.mjs

@@ -152,6 +152,7 @@ var TOKEN_TYPES = Object.freeze({
   Is: "Is",
   NotIn: "NotIn",
   Else: "Else",
+  EndSet: "EndSet",
   EndIf: "EndIf",
   ElseIf: "ElseIf",
   EndFor: "EndFor",
@@ -168,6 +169,7 @@ var KEYWORDS = Object.freeze({
   is: TOKEN_TYPES.Is,
   if: TOKEN_TYPES.If,
   else: TOKEN_TYPES.Else,
+  endset: TOKEN_TYPES.EndSet,
   endif: TOKEN_TYPES.EndIf,
   elif: TOKEN_TYPES.ElseIf,
   endfor: TOKEN_TYPES.EndFor,
@@ -404,10 +406,11 @@ var For = class extends Statement {
   type = "For";
 };
 var SetStatement = class extends Statement {
-  constructor(assignee, value) {
+  constructor(assignee, value, body) {
     super();
     this.assignee = assignee;
     this.value = value;
+    this.body = body;
   }
   type = "Set";
 };
@@ -615,10 +618,19 @@ function parse(tokens) {
     const left = parseExpression();
     if (is(TOKEN_TYPES.Equals)) {
       ++current;
-      const value = parseSetStatement();
-      return new SetStatement(left, value);
+      const value = parseExpression();
+      return new SetStatement(left, value, []);
+    } else {
+      const body = [];
+      expect(TOKEN_TYPES.CloseStatement, "Expected %} token");
+      while (!(tokens[current]?.type === TOKEN_TYPES.OpenStatement && tokens[current + 1]?.type === TOKEN_TYPES.EndSet)) {
+        const another = parseAny();
+        body.push(another);
+      }
+      expect(TOKEN_TYPES.OpenStatement, "Expected {% token");
+      expect(TOKEN_TYPES.EndSet, "Expected endset token");
+      return new SetStatement(left, null, body);
     }
-    return left;
   }
   function parseIfStatement() {
     const test = parseExpression();
@@ -1425,6 +1437,8 @@ var Interpreter = class {
             );
           case "join":
             return new StringValue(operand.value.map((x) => x.value).join(""));
+          case "string":
+            return new StringValue(toJSON(operand));
           default:
             throw new Error(`Unknown ArrayValue filter: ${filter.value}`);
         }
@@ -1692,7 +1706,7 @@ var Interpreter = class {
     return value instanceof RuntimeValue ? value : new UndefinedValue();
   }
   evaluateSet(node, environment) {
-    const rhs = this.evaluate(node.value, environment);
+    const rhs = node.value ? this.evaluate(node.value, environment) : this.evaluateBlock(node.body, environment);
     if (node.assignee.type === "Identifier") {
       const variableName = node.assignee.value;
       environment.setVariable(variableName, rhs);
@@ -4013,6 +4027,7 @@ class AutoConfig {
 /**
  * Transformers.js-specific configuration, possibly present in config.json under the key `transformers.js_config`.
  * @typedef {Object} TransformersJSConfig
+ * @property {Record<import('./utils/devices.js').DeviceType, DeviceConfig>} [device_config] Device-specific configurations.
  * @property {import('./utils/tensor.js').DataType|Record<import('./utils/dtypes.js').DataType, import('./utils/tensor.js').DataType>} [kv_cache_dtype] The data type of the key-value cache.
  * @property {Record<string, number>} [free_dimension_overrides] Override the free dimensions of the model.
  * See https://onnxruntime.ai/docs/tutorials/web/env-flags-and-session-options.html#freedimensionoverrides
@@ -4022,6 +4037,11 @@ class AutoConfig {
  * @property {import('./utils/hub.js').ExternalData|Record<string, import('./utils/hub.js').ExternalData>} [use_external_data_format=false] Whether to load the model using the external data format (used for models >= 2GB in size).
  */
 
+/**
+ * Device-specific configuration options.
+ * @typedef {Omit<TransformersJSConfig, "device" | "device_config">} DeviceConfig
+ */
+
 
 /***/ }),
 
@@ -4067,7 +4087,7 @@ __webpack_require__.r(__webpack_exports__);
 
 
 
-const VERSION = '3.4.2';
+const VERSION = '3.5.0';
 
 // Check if various APIs are available (depends on environment)
 const IS_BROWSER_ENV = typeof window !== "undefined" && typeof window.document !== "undefined";
@@ -4159,7 +4179,8 @@ const localModelPath = RUNNING_LOCALLY
  * @property {string} cacheDir The directory to use for caching files with the file system. By default, it is `./.cache`.
  * @property {boolean} useCustomCache Whether to use a custom cache system (defined by `customCache`), defaults to `false`.
  * @property {Object} customCache The custom cache to use. Defaults to `null`. Note: this must be an object which
- * implements the `match` and `put` functions of the Web Cache API. For more information, see https://developer.mozilla.org/en-US/docs/Web/API/Cache
+ * implements the `match` and `put` functions of the Web Cache API. For more information, see https://developer.mozilla.org/en-US/docs/Web/API/Cache.
+ * If you wish, you may also return a `Promise<string>` from the `match` function if you'd like to use a file path instead of `Promise<Response>`.
  */
 
 /** @type {TransformersEnvironment} */
@@ -5186,13 +5207,15 @@ class NoBadWordsLogitsProcessor extends LogitsProcessor {
             const batch_logits_data = /** @type {Float32Array} */(logits[i].data);
             const ids = input_ids[i];
             for (const bad_word_ids of this.bad_words_ids) {
+                // There aren't enough tokens to match the banned sequence
+                if (ids.length < bad_word_ids.length - 1) continue;
+
                 // Whether to modify the logits of the last token in the bad word id sequence
                 let mark = true;
 
                 // For each bad word in the list, if the current sequence of input ids ends with this sequence (excluding the last),
                 // then we set the logits of the last bad word id to -Infinity.
-                for (let j = 1; j <= bad_word_ids.length - 1 && bad_word_ids.length < ids.length; ++j) {
-
+                for (let j = 1; j <= bad_word_ids.length - 1; ++j) {
                     // NOTE: We use != instead of !== to compare bigint and number
                     // @ts-ignore
                     if (bad_word_ids.at(-j - 1) != ids.at(-j)) {
@@ -6664,7 +6687,8 @@ const MODEL_CLASS_TO_NAME_MAPPING = new Map();
  * @private
  */
 async function getSession(pretrained_model_name_or_path, fileName, options) {
-    const custom_config = options.config?.['transformers.js_config'] ?? {};
+    let custom_config = options.config?.['transformers.js_config'] ?? {};
+
     let device = options.device ?? custom_config.device;
     if (device && typeof device !== 'string') {
         if (device.hasOwnProperty(fileName)) {
@@ -6679,8 +6703,18 @@ async function getSession(pretrained_model_name_or_path, fileName, options) {
     const selectedDevice = /** @type {import("./utils/devices.js").DeviceType} */(
         device ?? (_env_js__WEBPACK_IMPORTED_MODULE_14__.apis.IS_NODE_ENV ? 'cpu' : 'wasm')
     );
+
     const executionProviders = (0,_backends_onnx_js__WEBPACK_IMPORTED_MODULE_1__.deviceToExecutionProviders)(selectedDevice);
 
+    // Update custom config with the selected device's config, if it exists
+    const device_config = custom_config.device_config ?? {};
+    if (device_config.hasOwnProperty(selectedDevice)) {
+        custom_config = {
+            ...custom_config,
+            ...device_config[selectedDevice],
+        };
+    }
+
     // If options.dtype is specified, we use it to choose the suffix for the model file.
     // Otherwise, we use the default dtype for the device.
     let dtype = options.dtype ?? custom_config.dtype;
@@ -6697,11 +6731,11 @@ async function getSession(pretrained_model_name_or_path, fileName, options) {
         // Try to choose the auto dtype based on the custom config
         let config_dtype = custom_config.dtype;
         if (typeof config_dtype !== 'string') {
-            config_dtype = config_dtype[fileName];
+            config_dtype = config_dtype?.[fileName];
         }
 
         if (config_dtype && config_dtype !== _utils_dtypes_js__WEBPACK_IMPORTED_MODULE_2__.DATA_TYPES.auto && _utils_dtypes_js__WEBPACK_IMPORTED_MODULE_2__.DATA_TYPES.hasOwnProperty(config_dtype)) {
-            // Defined by the custom config, and is not "auto"
+            // Defined by the config, and is not "auto"
             dtype = config_dtype;
         } else {
             // Choose default dtype based on device, falling back to fp32
@@ -6718,10 +6752,11 @@ async function getSession(pretrained_model_name_or_path, fileName, options) {
     }
 
     // Only valid for models with a decoder
-    const kv_cache_dtype = custom_config.kv_cache_dtype
-        ? (typeof custom_config.kv_cache_dtype === 'string'
-            ? custom_config.kv_cache_dtype
-            : custom_config.kv_cache_dtype[selectedDtype] ?? 'float32')
+    const kv_cache_dtype_config = custom_config.kv_cache_dtype;
+    const kv_cache_dtype = kv_cache_dtype_config
+        ? (typeof kv_cache_dtype_config === 'string'
+            ? kv_cache_dtype_config
+            : kv_cache_dtype_config[selectedDtype] ?? 'float32')
         : undefined;
 
     if (kv_cache_dtype && !['float32', 'float16'].includes(kv_cache_dtype)) {
@@ -6749,14 +6784,15 @@ async function getSession(pretrained_model_name_or_path, fileName, options) {
         session_options.freeDimensionOverrides ??= free_dimension_overrides;
     } else if (selectedDevice.startsWith('webnn') && !session_options.freeDimensionOverrides) {
         console.warn(
-            'WebNN does not currently support dynamic shapes and requires `free_dimension_overrides` to be set in config.json as a field within "transformers.js_config". ' +
-            'When `free_dimension_overrides` is not set, you may experience significant performance degradation.'
+            `WebNN does not currently support dynamic shapes and requires 'free_dimension_overrides' to be set in config.json, preferably as a field within config["transformers.js_config"]["device_config"]["${selectedDevice}"]. ` +
+            `When 'free_dimension_overrides' is not set, you may experience significant performance degradation.`
         );
     }
 
-    const bufferOrPathPromise = (0,_utils_hub_js__WEBPACK_IMPORTED_MODULE_5__.getModelFile)(pretrained_model_name_or_path, modelFileName, true, options, _env_js__WEBPACK_IMPORTED_MODULE_14__.apis.IS_NODE_ENV);
+    const return_path = _env_js__WEBPACK_IMPORTED_MODULE_14__.apis.IS_NODE_ENV && _env_js__WEBPACK_IMPORTED_MODULE_14__.env.useFSCache;
+    const bufferOrPathPromise = (0,_utils_hub_js__WEBPACK_IMPORTED_MODULE_5__.getModelFile)(pretrained_model_name_or_path, modelFileName, true, options, return_path);
 
-    // handle onnx external data files
+    // Handle onnx external data files
     const use_external_data_format = options.use_external_data_format ?? custom_config.use_external_data_format;
     /** @type {Promise<string|{path: string, data: Uint8Array}>[]} */
     let externalDataPromises = [];
@@ -6782,7 +6818,7 @@ async function getSession(pretrained_model_name_or_path, fileName, options) {
             const path = `${baseName}_data${i === 0 ? '' : '_' + i}`;
             const fullPath = `${options.subfolder ?? ''}/${path}`;
             externalDataPromises.push(new Promise(async (resolve, reject) => {
-                const data = await (0,_utils_hub_js__WEBPACK_IMPORTED_MODULE_5__.getModelFile)(pretrained_model_name_or_path, fullPath, true, options, _env_js__WEBPACK_IMPORTED_MODULE_14__.apis.IS_NODE_ENV);
+                const data = await (0,_utils_hub_js__WEBPACK_IMPORTED_MODULE_5__.getModelFile)(pretrained_model_name_or_path, fullPath, true, options, return_path);
                 resolve(data instanceof Uint8Array ? { path, data } : path);
             }));
         }
@@ -14961,7 +14997,7 @@ class ASTFeatureExtractor extends _base_feature_extraction_utils_js__WEBPACK_IMP
 
         const sampling_rate = this.config.sampling_rate;
         const mel_filters = (0,_utils_audio_js__WEBPACK_IMPORTED_MODULE_2__.mel_filter_bank)(
-            256, // num_frequency_bins
+            257, // num_frequency_bins
             this.config.num_mel_bins, // num_mel_filters
             20, // min_frequency
             Math.floor(sampling_rate / 2), // max_frequency
@@ -14970,11 +15006,6 @@ class ASTFeatureExtractor extends _base_feature_extraction_utils_js__WEBPACK_IMP
             "kaldi", // mel_scale
             true, // triangularize_in_mel_space
         );
-
-        // Do padding:
-        for (let i = 0; i < mel_filters.length; ++i) {
-            mel_filters[i].push(0);
-        }
         this.mel_filters = mel_filters;
 
         this.window = (0,_utils_audio_js__WEBPACK_IMPORTED_MODULE_2__.window_function)(400, 'hann', {
@@ -18564,7 +18595,7 @@ class SeamlessM4TFeatureExtractor extends _base_feature_extraction_utils_js__WEB
 
         const sampling_rate = this.config.sampling_rate;
         const mel_filters = (0,_utils_audio_js__WEBPACK_IMPORTED_MODULE_2__.mel_filter_bank)(
-            256, // num_frequency_bins
+            257, // num_frequency_bins
             this.config.num_mel_bins, // num_mel_filters
             20, // min_frequency
             Math.floor(sampling_rate / 2), // max_frequency
@@ -18573,11 +18604,6 @@ class SeamlessM4TFeatureExtractor extends _base_feature_extraction_utils_js__WEB
             "kaldi", // mel_scale
             true, // triangularize_in_mel_space
         );
-
-        // Do padding:
-        for (let i = 0; i < mel_filters.length; ++i) {
-            mel_filters[i].push(0);
-        }
         this.mel_filters = mel_filters;
 
         this.window = (0,_utils_audio_js__WEBPACK_IMPORTED_MODULE_2__.window_function)(400, 'povey', {
@@ -19330,7 +19356,7 @@ class WeSpeakerFeatureExtractor extends _base_feature_extraction_utils_js__WEBPA
 
         const sampling_rate = this.config.sampling_rate;
         const mel_filters = (0,_utils_audio_js__WEBPACK_IMPORTED_MODULE_2__.mel_filter_bank)(
-            256, // num_frequency_bins
+            257, // num_frequency_bins
             this.config.num_mel_bins, // num_mel_filters
             20, // min_frequency
             Math.floor(sampling_rate / 2), // max_frequency
@@ -19339,11 +19365,6 @@ class WeSpeakerFeatureExtractor extends _base_feature_extraction_utils_js__WEBPA
             "kaldi", // mel_scale
             true, // triangularize_in_mel_space
         );
-
-        // Do padding:
-        for (let i = 0; i < mel_filters.length; ++i) {
-            mel_filters[i].push(0);
-        }
         this.mel_filters = mel_filters;
 
         this.window = (0,_utils_audio_js__WEBPACK_IMPORTED_MODULE_2__.window_function)(400, 'hamming', {
@@ -24330,8 +24351,24 @@ class BPE extends TokenizerModel {
 
         this.ignore_merges = this.config.ignore_merges ?? false;
 
-        /** @type {Map<string, string[]>} */
-        this.cache = new Map();
+        /**
+         * The maximum length we should cache in a model.
+         * Strings that are too long have minimal chances to cache hit anyway
+         */
+        this.max_length_to_cache = 256;
+
+        /**
+         * The default capacity for a `BPE`'s internal cache.
+         */
+        this.cache_capacity = 10000;
+        this.cache = new _utils_data_structures_js__WEBPACK_IMPORTED_MODULE_5__.LRUCache(this.cache_capacity);
+    }
+
+    /**
+     * Clears the cache.
+     */
+    clear_cache() {
+        this.cache.clear();
     }
 
     /**
@@ -24458,8 +24495,10 @@ class BPE extends TokenizerModel {
             }
         }
 
-        // Save the result to the cache
-        this.cache.set(token, result);
+        if (token.length < this.max_length_to_cache) {
+            // Save the result to the cache
+            this.cache.put(token, result);
+        }
 
         return result;
     }
@@ -28259,7 +28298,8 @@ function linspace(start, end, num) {
  * various implementation exist, which differ in the number of filters, the shape of the filters, the way the filters
  * are spaced, the bandwidth of the filters, and the manner in which the spectrum is warped. The goal of these
  * features is to approximate the non-linear human perception of the variation in pitch with respect to the frequency.
- * @param {number} num_frequency_bins Number of frequencies used to compute the spectrogram (should be the same as in `stft`).
+ * @param {number} num_frequency_bins Number of frequency bins (should be the same as `n_fft // 2 + 1`
+ * where `n_fft` is the size of the Fourier Transform used to compute the spectrogram).
  * @param {number} num_mel_filters Number of mel filters to generate.
  * @param {number} min_frequency Lowest frequency of interest in Hz.
  * @param {number} max_frequency Highest frequency of interest in Hz. This should not exceed `sampling_rate / 2`.
@@ -28285,6 +28325,14 @@ function mel_filter_bank(
         throw new Error('norm must be one of null or "slaney"');
     }
 
+    if (num_frequency_bins < 2) {
+        throw new Error(`Require num_frequency_bins: ${num_frequency_bins} >= 2`);
+    }
+
+    if (min_frequency > max_frequency) {
+        throw new Error(`Require min_frequency: ${min_frequency} <= max_frequency: ${max_frequency}`);
+    }
+
     const mel_min = hertz_to_mel(min_frequency, mel_scale);
     const mel_max = hertz_to_mel(max_frequency, mel_scale);
     const mel_freqs = linspace(mel_min, mel_max, num_mel_filters + 2);
@@ -28293,7 +28341,7 @@ function mel_filter_bank(
     let fft_freqs; // frequencies of FFT bins in Hz
 
     if (triangularize_in_mel_space) {
-        const fft_bin_width = sampling_rate / (num_frequency_bins * 2);
+        const fft_bin_width = sampling_rate / ((num_frequency_bins - 1) * 2);
         fft_freqs = hertz_to_mel(Float64Array.from({ length: num_frequency_bins }, (_, i) => i * fft_bin_width), mel_scale);
         filter_freqs = mel_freqs;
     } else {
@@ -29170,6 +29218,7 @@ __webpack_require__.r(__webpack_exports__);
 /* harmony export */ __webpack_require__.d(__webpack_exports__, {
 /* harmony export */   CharTrie: () => (/* binding */ CharTrie),
 /* harmony export */   DictionarySplitter: () => (/* binding */ DictionarySplitter),
+/* harmony export */   LRUCache: () => (/* binding */ LRUCache),
 /* harmony export */   PriorityQueue: () => (/* binding */ PriorityQueue),
 /* harmony export */   TokenLattice: () => (/* binding */ TokenLattice)
 /* harmony export */ });
@@ -29695,6 +29744,59 @@ class DictionarySplitter {
     }
 }
 
+/**
+* A simple Least Recently Used (LRU) cache implementation in JavaScript.
+* This cache stores key-value pairs and evicts the least recently used item
+* when the capacity is exceeded.
+*/
+class LRUCache {
+    /**
+     * Creates an LRUCache instance.
+     * @param {number} capacity The maximum number of items the cache can hold.
+     */
+    constructor(capacity) {
+        this.capacity = capacity;
+        this.cache = new Map();
+    }
+
+    /**
+     * Retrieves the value associated with the given key and marks the key as recently used.
+     * @param {any} key The key to retrieve.
+     * @returns {any} The value associated with the key, or undefined if the key does not exist.
+     */
+    get(key) {
+        if (!this.cache.has(key)) return undefined;
+        const value = this.cache.get(key);
+        this.cache.delete(key);
+        this.cache.set(key, value);
+        return value;
+    }
+
+    /**
+     * Inserts or updates the key-value pair in the cache.
+     * If the key already exists, it is updated and marked as recently used.
+     * If the cache exceeds its capacity, the least recently used item is evicted.
+     * @param {any} key The key to add or update.
+     * @param {any} value The value to associate with the key.
+     */
+    put(key, value) {
+        if (this.cache.has(key)) {
+            this.cache.delete(key);
+        }
+        this.cache.set(key, value);
+        if (this.cache.size > this.capacity) {
+            this.cache.delete(this.cache.keys().next().value);
+        }
+    }
+
+    /**
+     * Clears the cache.
+     */
+    clear() {
+        this.cache.clear();
+    }
+}
+
 
 /***/ }),
 
@@ -30302,6 +30404,22 @@ async function getModelFile(path_or_repo_id, filename, fatal = true, options = {
     // First, check if the a caching backend is available
     // If no caching mechanism available, will download the file every time
     let cache;
+    if (!cache && _env_js__WEBPACK_IMPORTED_MODULE_2__.env.useCustomCache) {
+        // Allow the user to specify a custom cache system.
+        if (!_env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache) {
+            throw Error('`env.useCustomCache=true`, but `env.customCache` is not defined.')
+        }
+
+        // Check that the required methods are defined:
+        if (!_env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache.match || !_env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache.put) {
+            throw new Error(
+                "`env.customCache` must be an object which implements the `match` and `put` functions of the Web Cache API. " +
+                "For more information, see https://developer.mozilla.org/en-US/docs/Web/API/Cache"
+            )
+        }
+        cache = _env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache;
+    }
+
     if (!cache && _env_js__WEBPACK_IMPORTED_MODULE_2__.env.useBrowserCache) {
         if (typeof caches === 'undefined') {
             throw Error('Browser cache is not available in this environment.')
@@ -30319,28 +30437,14 @@ async function getModelFile(path_or_repo_id, filename, fatal = true, options = {
     }
 
     if (!cache && _env_js__WEBPACK_IMPORTED_MODULE_2__.env.useFSCache) {
-        // TODO throw error if not available
+        if (!_env_js__WEBPACK_IMPORTED_MODULE_2__.apis.IS_FS_AVAILABLE) {
+            throw Error('File System Cache is not available in this environment.');
+        }
 
         // If `cache_dir` is not specified, use the default cache directory
         cache = new FileCache(options.cache_dir ?? _env_js__WEBPACK_IMPORTED_MODULE_2__.env.cacheDir);
     }
 
-    if (!cache && _env_js__WEBPACK_IMPORTED_MODULE_2__.env.useCustomCache) {
-        // Allow the user to specify a custom cache system.
-        if (!_env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache) {
-            throw Error('`env.useCustomCache=true`, but `env.customCache` is not defined.')
-        }
-
-        // Check that the required methods are defined:
-        if (!_env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache.match || !_env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache.put) {
-            throw new Error(
-                "`env.customCache` must be an object which implements the `match` and `put` functions of the Web Cache API. " +
-                "For more information, see https://developer.mozilla.org/en-US/docs/Web/API/Cache"
-            )
-        }
-        cache = _env_js__WEBPACK_IMPORTED_MODULE_2__.env.customCache;
-    }
-
     const revision = options.revision ?? 'main';
     const requestURL = pathJoin(path_or_repo_id, filename);
 
@@ -30522,7 +30626,7 @@ async function getModelFile(path_or_repo_id, filename, fatal = true, options = {
     });
 
     if (result) {
-        if (return_path) {
+        if (!_env_js__WEBPACK_IMPORTED_MODULE_2__.apis.IS_NODE_ENV && return_path) {
             throw new Error("Cannot return path in a browser environment.")
         }
         return result;
@@ -30531,12 +30635,18 @@ async function getModelFile(path_or_repo_id, filename, fatal = true, options = {
         return response.filePath;
     }
 
-    const path = await cache.match(cacheKey);
-    if (path instanceof FileResponse) {
-        return path.filePath;
+    // Otherwise, return the cached response (most likely a `FileResponse`).
+    // NOTE: A custom cache may return a Response, or a string (file path)
+    const cachedResponse = await cache?.match(cacheKey);
+    if (cachedResponse instanceof FileResponse) {
+        return cachedResponse.filePath;
+    } else if (cachedResponse instanceof Response) {
+        return new Uint8Array(await cachedResponse.arrayBuffer());
+    } else if (typeof cachedResponse === 'string') {
+        return cachedResponse;
     }
-    throw new Error("Unable to return path for response.");
 
+    throw new Error("Unable to get model file path or buffer.");
 }
 
 /**