@huggingface/transformers 3.0.0-alpha.13 → 3.0.0-alpha.14

package/README.md

@@ -101,7 +101,7 @@ npm i @huggingface/transformers
 Alternatively, you can use it in vanilla JS, without any bundler, by using a CDN or static hosting. For example, using [ES Modules](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Modules), you can import the library with:
 ```html
 <script type="module">
-    import { pipeline } from 'https://cdn.jsdelivr.net/npm/@huggingface/transformers@3.0.0-alpha.13';
+    import { pipeline } from 'https://cdn.jsdelivr.net/npm/@huggingface/transformers@3.0.0-alpha.14';
 </script>
 ```
 
@@ -134,7 +134,7 @@ Check out the Transformers.js [template](https://huggingface.co/new-space?templa
 
 
 
-By default, Transformers.js uses [hosted pretrained models](https://huggingface.co/models?library=transformers.js) and [precompiled WASM binaries](https://cdn.jsdelivr.net/npm/@huggingface/transformers@3.0.0-alpha.13/dist/), which should work out-of-the-box. You can customize this as follows:
+By default, Transformers.js uses [hosted pretrained models](https://huggingface.co/models?library=transformers.js) and [precompiled WASM binaries](https://cdn.jsdelivr.net/npm/@huggingface/transformers@3.0.0-alpha.14/dist/), which should work out-of-the-box. You can customize this as follows:
 
 ### Settings
 
@@ -310,6 +310,7 @@ You can refine your search by selecting the task you're interested in (e.g., [te
 1. **[GPT-J](https://huggingface.co/docs/transformers/model_doc/gptj)** (from EleutherAI) released in the repository [kingoflolz/mesh-transformer-jax](https://github.com/kingoflolz/mesh-transformer-jax/) by Ben Wang and Aran Komatsuzaki.
 1. **[GPTBigCode](https://huggingface.co/docs/transformers/model_doc/gpt_bigcode)** (from BigCode) released with the paper [SantaCoder: don't reach for the stars!](https://arxiv.org/abs/2301.03988) by Loubna Ben Allal, Raymond Li, Denis Kocetkov, Chenghao Mou, Christopher Akiki, Carlos Munoz Ferrandis, Niklas Muennighoff, Mayank Mishra, Alex Gu, Manan Dey, Logesh Kumar Umapathi, Carolyn Jane Anderson, Yangtian Zi, Joel Lamy Poirier, Hailey Schoelkopf, Sergey Troshin, Dmitry Abulkhanov, Manuel Romero, Michael Lappert, Francesco De Toni, Bernardo García del Río, Qian Liu, Shamik Bose, Urvashi Bhattacharyya, Terry Yue Zhuo, Ian Yu, Paulo Villegas, Marco Zocca, Sourab Mangrulkar, David Lansky, Huu Nguyen, Danish Contractor, Luis Villa, Jia Li, Dzmitry Bahdanau, Yacine Jernite, Sean Hughes, Daniel Fried, Arjun Guha, Harm de Vries, Leandro von Werra.
 1. **[HerBERT](https://huggingface.co/docs/transformers/model_doc/herbert)** (from Allegro.pl, AGH University of Science and Technology) released with the paper [KLEJ: Comprehensive Benchmark for Polish Language Understanding](https://www.aclweb.org/anthology/2020.acl-main.111.pdf) by Piotr Rybak, Robert Mroczkowski, Janusz Tracz, Ireneusz Gawlik.
+1. **[Hiera](https://huggingface.co/docs/transformers/model_doc/hiera)** (from Meta) released with the paper [Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles](https://arxiv.org/pdf/2306.00989) by Chaitanya Ryali, Yuan-Ting Hu, Daniel Bolya, Chen Wei, Haoqi Fan, Po-Yao Huang, Vaibhav Aggarwal, Arkabandhu Chowdhury, Omid Poursaeed, Judy Hoffman, Jitendra Malik, Yanghao Li, Christoph Feichtenhofer.
 1. **[Hubert](https://huggingface.co/docs/transformers/model_doc/hubert)** (from Facebook) released with the paper [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://arxiv.org/abs/2106.07447) by Wei-Ning Hsu, Benjamin Bolte, Yao-Hung Hubert Tsai, Kushal Lakhotia, Ruslan Salakhutdinov, Abdelrahman Mohamed.
 1. **JAIS** (from Core42) released with the paper [Jais and Jais-chat: Arabic-Centric Foundation and Instruction-Tuned Open Generative Large Language Models](https://arxiv.org/pdf/2308.16149) by Neha Sengupta, Sunil Kumar Sahu, Bokang Jia, Satheesh Katipomu, Haonan Li, Fajri Koto, William Marshall, Gurpreet Gosal, Cynthia Liu, Zhiming Chen, Osama Mohammed Afzal, Samta Kamboj, Onkar Pandit, Rahul Pal, Lalit Pradhan, Zain Muhammad Mujahid, Massa Baali, Xudong Han, Sondos Mahmoud Bsharat, Alham Fikri Aji, Zhiqiang Shen, Zhengzhong Liu, Natalia Vassilieva, Joel Hestness, Andy Hock, Andrew Feldman, Jonathan Lee, Andrew Jackson, Hector Xuguang Ren, Preslav Nakov, Timothy Baldwin, Eric Xing.
 1. **[LongT5](https://huggingface.co/docs/transformers/model_doc/longt5)** (from Google AI) released with the paper [LongT5: Efficient Text-To-Text Transformer for Long Sequences](https://arxiv.org/abs/2112.07916) by Mandy Guo, Joshua Ainslie, David Uthus, Santiago Ontanon, Jianmo Ni, Yun-Hsuan Sung, Yinfei Yang.

package/dist/transformers.cjs

@@ -4437,7 +4437,7 @@ __webpack_require__.r(__webpack_exports__);
 
 
 
-const VERSION = '3.0.0-alpha.13';
+const VERSION = '3.0.0-alpha.14';
 
 // Check if various APIs are available (depends on environment)
 const IS_BROWSER_ENV = typeof self !== 'undefined';
@@ -5533,18 +5533,18 @@ class NoBadWordsLogitsProcessor extends LogitsProcessor {
     _call(input_ids, logits) {
         for (let i = 0; i < input_ids.length; ++i) {
             const batch_logits_data = /** @type {Float32Array} */(logits[i].data);
-
+            const ids = input_ids[i];
             for (const bad_word_ids of this.bad_words_ids) {
                 // 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 i = 1; i <= bad_word_ids.length - 1 && bad_word_ids.length < input_ids[i].length; ++i) {
+                for (let j = 1; j <= bad_word_ids.length - 1 && bad_word_ids.length < ids.length; ++j) {
 
                     // NOTE: We use != instead of !== to compare bigint and number
                     // @ts-ignore
-                    if (bad_word_ids.at(-i - 1) != input_ids[i].at(-i)) {
+                    if (bad_word_ids.at(-j - 1) != ids.at(-j)) {
                         // We have found a mismatch
                         mark = false;
                         break;
@@ -6530,6 +6530,9 @@ __webpack_require__.r(__webpack_exports__);
 /* harmony export */   GemmaForCausalLM: () => (/* binding */ GemmaForCausalLM),
 /* harmony export */   GemmaModel: () => (/* binding */ GemmaModel),
 /* harmony export */   GemmaPreTrainedModel: () => (/* binding */ GemmaPreTrainedModel),
+/* harmony export */   HieraForImageClassification: () => (/* binding */ HieraForImageClassification),
+/* harmony export */   HieraModel: () => (/* binding */ HieraModel),
+/* harmony export */   HieraPreTrainedModel: () => (/* binding */ HieraPreTrainedModel),
 /* harmony export */   HubertForCTC: () => (/* binding */ HubertForCTC),
 /* harmony export */   HubertForSequenceClassification: () => (/* binding */ HubertForSequenceClassification),
 /* harmony export */   HubertModel: () => (/* binding */ HubertModel),
@@ -6958,6 +6961,7 @@ async function getSession(pretrained_model_name_or_path, fileName, options) {
         });
         if (Object.keys(shapes).length > 0 && !(0,_backends_onnx_js__WEBPACK_IMPORTED_MODULE_1__.isONNXProxy)()) {
             // Only set preferredOutputLocation if shapes are present and we aren't proxying ONNX
+            /** @type {Record<string, import('onnxruntime-common').Tensor.DataLocation>} */
             const preferredOutputLocation = {};
             for (const key in shapes) {
                 preferredOutputLocation[key] = 'gpu-buffer';
@@ -11178,6 +11182,19 @@ class DeiTForImageClassification extends DeiTPreTrainedModel {
 }
 //////////////////////////////////////////////////
 
+//////////////////////////////////////////////////
+class HieraPreTrainedModel extends PreTrainedModel { }
+class HieraModel extends HieraPreTrainedModel { }
+class HieraForImageClassification extends HieraPreTrainedModel {
+    /**
+     * @param {any} model_inputs
+     */
+    async _call(model_inputs) {
+        return new SequenceClassifierOutput(await super._call(model_inputs));
+    }
+}
+//////////////////////////////////////////////////
+
 
 //////////////////////////////////////////////////
 /**
@@ -13058,6 +13075,7 @@ const MODEL_MAPPING_NAMES_ENCODER_ONLY = new Map([
     ['owlv2', ['Owlv2Model', Owlv2Model]],
     ['beit', ['BeitModel', BeitModel]],
     ['deit', ['DeiTModel', DeiTModel]],
+    ['hiera', ['HieraModel', HieraModel]],
     ['convnext', ['ConvNextModel', ConvNextModel]],
     ['convnextv2', ['ConvNextV2Model', ConvNextV2Model]],
     ['dinov2', ['Dinov2Model', Dinov2Model]],
@@ -13265,6 +13283,7 @@ const MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING_NAMES = new Map([
     ['mobilevitv2', ['MobileViTV2ForImageClassification', MobileViTV2ForImageClassification]],
     ['beit', ['BeitForImageClassification', BeitForImageClassification]],
     ['deit', ['DeiTForImageClassification', DeiTForImageClassification]],
+    ['hiera', ['HieraForImageClassification', HieraForImageClassification]],
     ['convnext', ['ConvNextForImageClassification', ConvNextForImageClassification]],
     ['convnextv2', ['ConvNextV2ForImageClassification', ConvNextV2ForImageClassification]],
     ['dinov2', ['Dinov2ForImageClassification', Dinov2ForImageClassification]],
@@ -14092,20 +14111,31 @@ __webpack_require__.r(__webpack_exports__);
 
 
 
+/**
+ * Asynchronously creates a wrapper function for running an ONNX inference session.
+ *
+ * @param {number[]} session_bytes The session data in bytes.
+ * @param {import('onnxruntime-common').InferenceSession.SessionOptions} session_options The options for the ONNX session.
+ * @template {string | [string] | string[]} T
+ * @param {T} names The name(s) of the output tensor(s).
+ * 
+ * @returns {Promise<function(Record<string, Tensor>): Promise<T extends string ? Tensor : T extends string[] ? { [K in keyof T]: Tensor } : never>>}
+ * The wrapper function for running the ONNX inference session.
+ */
 const wrap = async (session_bytes, session_options, names) => {
     const session = await (0,_backends_onnx_js__WEBPACK_IMPORTED_MODULE_0__.createInferenceSession)(
         new Uint8Array(session_bytes), session_options,
     );
-    return async (inputs) => {
+    return /** @type {any} */(async (/** @type {Record<string, Tensor>} */ inputs) => {
         const ortFeed = Object.fromEntries(Object.entries(inputs).map(([k, v]) => [k, v.ort_tensor]));
         const outputs = await session.run(ortFeed);
 
         if (Array.isArray(names)) {
             return names.map((n) => new _utils_tensor_js__WEBPACK_IMPORTED_MODULE_1__.Tensor(outputs[n]));
         } else {
-            return new _utils_tensor_js__WEBPACK_IMPORTED_MODULE_1__.Tensor(outputs[names]);
+            return new _utils_tensor_js__WEBPACK_IMPORTED_MODULE_1__.Tensor(outputs[/** @type {string} */(names)]);
         }
-    }
+    })
 }
 
 // In-memory registry of initialized ONNX operators
@@ -17783,9 +17813,8 @@ function post_process_semantic_segmentation(outputs, target_sizes = null) {
         // Store which objects have labels
         // This is much more efficient that creating a set of the final values
         const hasLabel = new Array(data.dims[0]);
-        const out = segmentation.data;
-        for (let j = 0; j < out.length; ++j) {
-            const index = out[j];
+        for (let j = 0; j < segmentation_data.length; ++j) {
+            const index = segmentation_data[j];
             hasLabel[index] = index;
         }
         /** @type {number[]} The unique list of labels that were detected */
@@ -27901,7 +27930,7 @@ function magnitude(arr) {
 /**
  * Returns the value and index of the minimum element in an array.
  * @param {number[]|TypedArray} arr array of numbers.
- * @returns {number[]} the value and index of the minimum element, of the form: [valueOfMin, indexOfMin]
+ * @returns {[number, number]} the value and index of the minimum element, of the form: [valueOfMin, indexOfMin]
  * @throws {Error} If array is empty.
  */
 function min(arr) {
@@ -30497,6 +30526,9 @@ __webpack_require__.r(__webpack_exports__);
 /* harmony export */   GemmaTokenizer: () => (/* reexport safe */ _tokenizers_js__WEBPACK_IMPORTED_MODULE_3__.GemmaTokenizer),
 /* harmony export */   Grok1Tokenizer: () => (/* reexport safe */ _tokenizers_js__WEBPACK_IMPORTED_MODULE_3__.Grok1Tokenizer),
 /* harmony export */   HerbertTokenizer: () => (/* reexport safe */ _tokenizers_js__WEBPACK_IMPORTED_MODULE_3__.HerbertTokenizer),
+/* harmony export */   HieraForImageClassification: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.HieraForImageClassification),
+/* harmony export */   HieraModel: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.HieraModel),
+/* harmony export */   HieraPreTrainedModel: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.HieraPreTrainedModel),
 /* harmony export */   HubertForCTC: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.HubertForCTC),
 /* harmony export */   HubertForSequenceClassification: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.HubertForSequenceClassification),
 /* harmony export */   HubertModel: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.HubertModel),