@huggingface/transformers 3.0.0 → 3.0.1

package/README.md

@@ -11,25 +11,19 @@
 </p>
 
 <p align="center">
-    <a href="https://www.npmjs.com/package/@huggingface/transformers">
-        <img alt="NPM" src="https://img.shields.io/npm/v/@huggingface/transformers">
-    </a>
-    <a href="https://www.npmjs.com/package/@huggingface/transformers">
-        <img alt="NPM Downloads" src="https://img.shields.io/npm/dw/@huggingface/transformers">
-    </a>
-    <a href="https://www.jsdelivr.com/package/npm/@huggingface/transformers">
-        <img alt="jsDelivr Hits" src="https://img.shields.io/jsdelivr/npm/hw/@huggingface/transformers">
-    </a>
-    <a href="https://github.com/huggingface/transformers.js/blob/main/LICENSE">
-        <img alt="License" src="https://img.shields.io/github/license/huggingface/transformers.js?color=blue">
-    </a>
-    <a href="https://huggingface.co/docs/transformers.js/index">
-        <img alt="Documentation" src="https://img.shields.io/website/http/huggingface.co/docs/transformers.js/index.svg?down_color=red&down_message=offline&up_message=online">
-    </a>
+    <a href="https://www.npmjs.com/package/@huggingface/transformers"><img alt="NPM" src="https://img.shields.io/npm/v/@huggingface/transformers"></a>
+    <a href="https://www.npmjs.com/package/@huggingface/transformers"><img alt="NPM Downloads" src="https://img.shields.io/npm/dw/@huggingface/transformers"></a>
+    <a href="https://www.jsdelivr.com/package/npm/@huggingface/transformers"><img alt="jsDelivr Hits" src="https://img.shields.io/jsdelivr/npm/hw/@huggingface/transformers"></a>
+    <a href="https://github.com/huggingface/transformers.js/blob/main/LICENSE"><img alt="License" src="https://img.shields.io/github/license/huggingface/transformers.js?color=blue"></a>
+    <a href="https://huggingface.co/docs/transformers.js/index"><img alt="Documentation" src="https://img.shields.io/website/http/huggingface.co/docs/transformers.js/index.svg?down_color=red&down_message=offline&up_message=online"></a>
 </p>
 
 
-State-of-the-art Machine Learning for the web. Run 🤗 Transformers directly in your browser, with no need for a server!
+<h3 align="center">
+  <p>State-of-the-art Machine Learning for the Web</p>
+</h3>
+
+Run 🤗 Transformers directly in your browser, with no need for a server!
 
 Transformers.js is designed to be functionally equivalent to Hugging Face's [transformers](https://github.com/huggingface/transformers) python library, meaning you can run the same pretrained models using a very similar API. These models support common tasks in different modalities, such as:
   - 📝 **Natural Language Processing**: text classification, named entity recognition, question answering, language modeling, summarization, translation, multiple choice, and text generation.
@@ -42,6 +36,22 @@ Transformers.js uses [ONNX Runtime](https://onnxruntime.ai/) to run models in th
 For more information, check out the full [documentation](https://huggingface.co/docs/transformers.js).
 
 
+## Installation
+
+
+To install via [NPM](https://www.npmjs.com/package/@huggingface/transformers), run:
+```bash
+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@3.0.1';
+</script>
+```
+
+
 ## Quick tour
 
 
@@ -72,9 +82,9 @@ out = pipe('I love transformers!')
 import { pipeline } from '@huggingface/transformers';
 
 // Allocate a pipeline for sentiment-analysis
-let pipe = await pipeline('sentiment-analysis');
+const pipe = await pipeline('sentiment-analysis');
 
-let out = await pipe('I love transformers!');
+const out = await pipe('I love transformers!');
 // [{'label': 'POSITIVE', 'score': 0.999817686}]
 ```
 
@@ -86,29 +96,40 @@ let out = await pipe('I love transformers!');
 You can also use a different model by specifying the model id or path as the second argument to the `pipeline` function. For example:
 ```javascript
 // Use a different model for sentiment-analysis
-let pipe = await pipeline('sentiment-analysis', 'Xenova/bert-base-multilingual-uncased-sentiment');
+const pipe = await pipeline('sentiment-analysis', 'Xenova/bert-base-multilingual-uncased-sentiment');
 ```
 
+By default, when running in the browser, the model will be run on your CPU (via WASM). If you would like
+to run the model on your GPU (via WebGPU), you can do this by setting `device: 'webgpu'`, for example:
+```javascript
+// Run the model on WebGPU
+const pipe = await pipeline('sentiment-analysis', 'Xenova/distilbert-base-uncased-finetuned-sst-2-english', {
+  device: 'webgpu',
+});
+```
 
-## Installation
-
+For more information, check out the [WebGPU guide](https://huggingface.co/docs/transformers.js/guides/webgpu).
 
-To install via [NPM](https://www.npmjs.com/package/@huggingface/transformers), run:
-```bash
-npm i @huggingface/transformers
-```
+> [!WARNING]
+> The WebGPU API is still experimental in many browsers, so if you run into any issues,
+> please file a [bug report](https://github.com/huggingface/transformers.js/issues/new?title=%5BWebGPU%5D%20Error%20running%20MODEL_ID_GOES_HERE&assignees=&labels=bug,webgpu&projects=&template=1_bug-report.yml).
 
-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';
-</script>
+In resource-constrained environments, such as web browsers, it is advisable to use a quantized version of
+the model to lower bandwidth and optimize performance. This can be achieved by adjusting the `dtype` option,
+which allows you to select the appropriate data type for your model. While the available options may vary
+depending on the specific model, typical choices include `"fp32"` (default for WebGPU), `"fp16"`, `"q8"`
+(default for WASM), and `"q4"`. For more information, check out the [quantization guide](https://huggingface.co/docs/transformers.js/guides/dtypes).
+```javascript
+// Run the model at 4-bit quantization
+const pipe = await pipeline('sentiment-analysis', 'Xenova/distilbert-base-uncased-finetuned-sst-2-english', {
+  dtype: 'q4',
+});
 ```
 
 
 ## Examples
 
-Want to jump straight in? Get started with one of our sample applications/templates:
+Want to jump straight in? Get started with one of our sample applications/templates, which can be found [here](https://github.com/huggingface/transformers.js-examples).
 
 | Name              | Description                      | Links                   |
 |-------------------|----------------------------------|-------------------------------|
@@ -134,7 +155,7 @@ 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/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@3.0.1/dist/), which should work out-of-the-box. You can customize this as follows:
 
 ### Settings
 

package/dist/transformers.cjs

@@ -4460,7 +4460,7 @@ __webpack_require__.r(__webpack_exports__);
 
 
 
-const VERSION = '3.0.0';
+const VERSION = '3.0.1';
 
 // Check if various APIs are available (depends on environment)
 const IS_BROWSER_ENV = typeof self !== 'undefined';
@@ -4508,7 +4508,7 @@ const apis = Object.freeze({
 
 const RUNNING_LOCALLY = IS_FS_AVAILABLE && IS_PATH_AVAILABLE;
 const dirname__ = RUNNING_LOCALLY
-    ? path__WEBPACK_IMPORTED_MODULE_1__.dirname(path__WEBPACK_IMPORTED_MODULE_1__.dirname(url__WEBPACK_IMPORTED_MODULE_2__.fileURLToPath("file:///workspaces/transformers.js/src/env.js")))
+    ? path__WEBPACK_IMPORTED_MODULE_1__.dirname(path__WEBPACK_IMPORTED_MODULE_1__.dirname(url__WEBPACK_IMPORTED_MODULE_2__.fileURLToPath("file:///home/runner/work/transformers.js/transformers.js/src/env.js")))
     : './';
 
 // Only used for environments with access to file system
@@ -7181,7 +7181,7 @@ function replaceTensors(obj) {
 
 /**
  * Converts an array or Tensor of integers to an int64 Tensor.
- * @param {Array|Tensor} items The input integers to be converted.
+ * @param {any[]|Tensor} items The input integers to be converted.
  * @returns {Tensor} The int64 Tensor with the converted values.
  * @throws {Error} If the input array is empty or the input is a batched Tensor and not all sequences have the same length.
  * @private
@@ -8104,35 +8104,37 @@ class PreTrainedModel extends _utils_generic_js__WEBPACK_IMPORTED_MODULE_3__.Cal
         let { decoder_input_ids, ...model_inputs } = model_kwargs;
 
         // Prepare input ids if the user has not defined `decoder_input_ids` manually.
-        if (!decoder_input_ids) {
-            decoder_start_token_id ??= bos_token_id;
-
-            if (this.config.model_type === 'musicgen') {
-                // Custom logic (TODO: move to Musicgen class)
-                decoder_input_ids = Array.from({
-                    length: batch_size * this.config.decoder.num_codebooks
-                }, () => [decoder_start_token_id]);
-
-            } else if (Array.isArray(decoder_start_token_id)) {
-                if (decoder_start_token_id.length !== batch_size) {
-                    throw new Error(
-                        `\`decoder_start_token_id\` expcted to have length ${batch_size} but got ${decoder_start_token_id.length}`
-                    )
+        if (!(decoder_input_ids instanceof _utils_tensor_js__WEBPACK_IMPORTED_MODULE_9__.Tensor)) {
+            if (!decoder_input_ids) {
+                decoder_start_token_id ??= bos_token_id;
+
+                if (this.config.model_type === 'musicgen') {
+                    // Custom logic (TODO: move to Musicgen class)
+                    decoder_input_ids = Array.from({
+                        length: batch_size * this.config.decoder.num_codebooks
+                    }, () => [decoder_start_token_id]);
+
+                } else if (Array.isArray(decoder_start_token_id)) {
+                    if (decoder_start_token_id.length !== batch_size) {
+                        throw new Error(
+                            `\`decoder_start_token_id\` expcted to have length ${batch_size} but got ${decoder_start_token_id.length}`
+                        )
+                    }
+                    decoder_input_ids = decoder_start_token_id;
+                } else {
+                    decoder_input_ids = Array.from({
+                        length: batch_size,
+                    }, () => [decoder_start_token_id]);
                 }
-                decoder_input_ids = decoder_start_token_id;
-            } else {
+            } else if (!Array.isArray(decoder_input_ids[0])) {
+                // Correct batch size
                 decoder_input_ids = Array.from({
                     length: batch_size,
-                }, () => [decoder_start_token_id]);
+                }, () => decoder_input_ids);
             }
-        } else if (!Array.isArray(decoder_input_ids[0])) {
-            // Correct batch size
-            decoder_input_ids = Array.from({
-                length: batch_size,
-            }, () => decoder_input_ids);
+            decoder_input_ids = toI64Tensor(decoder_input_ids);
         }
 
-        decoder_input_ids = toI64Tensor(decoder_input_ids);
         model_kwargs['decoder_attention_mask'] = (0,_utils_tensor_js__WEBPACK_IMPORTED_MODULE_9__.ones_like)(decoder_input_ids);
 
         return { input_ids: decoder_input_ids, model_inputs };
@@ -9955,8 +9957,11 @@ class WhisperForConditionalGeneration extends WhisperPreTrainedModel {
 class VisionEncoderDecoderModel extends PreTrainedModel {
     main_input_name = 'pixel_values';
     forward_params = [
+        // Encoder inputs
         'pixel_values',
-        'input_ids',
+
+        // Decoder inpputs
+        'decoder_input_ids',
         'encoder_hidden_states',
         'past_key_values',
     ];
@@ -16572,7 +16577,6 @@ class DocumentQuestionAnsweringPipeline extends (/** @type {new (options: TextIm
 
     /** @type {DocumentQuestionAnsweringPipelineCallback} */
     async _call(image, question, generate_kwargs = {}) {
-        throw new Error('This pipeline is not yet supported in Transformers.js v3.'); // TODO: Remove when implemented
 
         // NOTE: For now, we only support a batch size of 1
 
@@ -17373,6 +17377,7 @@ __webpack_require__.r(__webpack_exports__);
 /* harmony export */   DeiTFeatureExtractor: () => (/* binding */ DeiTFeatureExtractor),
 /* harmony export */   DetrFeatureExtractor: () => (/* binding */ DetrFeatureExtractor),
 /* harmony export */   DonutFeatureExtractor: () => (/* binding */ DonutFeatureExtractor),
+/* harmony export */   DonutImageProcessor: () => (/* binding */ DonutImageProcessor),
 /* harmony export */   EfficientNetImageProcessor: () => (/* binding */ EfficientNetImageProcessor),
 /* harmony export */   FeatureExtractor: () => (/* binding */ FeatureExtractor),
 /* harmony export */   Florence2Processor: () => (/* binding */ Florence2Processor),
@@ -18615,6 +18620,7 @@ class DonutFeatureExtractor extends ImageFeatureExtractor {
         });
     }
 }
+class DonutImageProcessor extends DonutFeatureExtractor { } // NOTE extends DonutFeatureExtractor
 class NougatImageProcessor extends DonutFeatureExtractor { } // NOTE extends DonutFeatureExtractor
 
 /**
@@ -19975,6 +19981,7 @@ class AutoProcessor {
         MaskFormerFeatureExtractor,
         YolosFeatureExtractor,
         DonutFeatureExtractor,
+        DonutImageProcessor,
         NougatImageProcessor,
         EfficientNetImageProcessor,
 
@@ -30304,6 +30311,7 @@ __webpack_require__.r(__webpack_exports__);
 /* harmony export */   DistilBertTokenizer: () => (/* reexport safe */ _tokenizers_js__WEBPACK_IMPORTED_MODULE_3__.DistilBertTokenizer),
 /* harmony export */   DocumentQuestionAnsweringPipeline: () => (/* reexport safe */ _pipelines_js__WEBPACK_IMPORTED_MODULE_1__.DocumentQuestionAnsweringPipeline),
 /* harmony export */   DonutFeatureExtractor: () => (/* reexport safe */ _processors_js__WEBPACK_IMPORTED_MODULE_4__.DonutFeatureExtractor),
+/* harmony export */   DonutImageProcessor: () => (/* reexport safe */ _processors_js__WEBPACK_IMPORTED_MODULE_4__.DonutImageProcessor),
 /* harmony export */   DonutSwinModel: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.DonutSwinModel),
 /* harmony export */   DonutSwinPreTrainedModel: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.DonutSwinPreTrainedModel),
 /* harmony export */   EfficientNetForImageClassification: () => (/* reexport safe */ _models_js__WEBPACK_IMPORTED_MODULE_2__.EfficientNetForImageClassification),