@xenova/transformers - npm Package Compare versions

Comparing version 2.13.4 to 2.14.0

package.json

 {
   "name": "@xenova/transformers",
-  "version": "2.13.4",
+  "version": "2.14.0",
   "description": "State-of-the-art Machine Learning for the web. Run 🤗 Transformers directly in your browser, with no need for a server!",
@@ -5,0 +5,0 @@ "main": "./src/transformers.js",

README.md

@@ -104,3 +104,3 @@
 <script type="module">
-    import { pipeline } from 'https://cdn.jsdelivr.net/npm/@xenova/transformers@2.13.4';
+    import { pipeline } from 'https://cdn.jsdelivr.net/npm/@xenova/transformers@2.14.0';
 </script>
@@ -138,3 +138,3 @@ ```
 
-By default, Transformers.js uses [hosted pretrained models](https://huggingface.co/models?library=transformers.js) and [precompiled WASM binaries](https://cdn.jsdelivr.net/npm/@xenova/transformers@2.13.4/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/@xenova/transformers@2.14.0/dist/), which should work out-of-the-box. You can customize this as follows:
 
@@ -333,2 +333,3 @@
 1. **[SegFormer](https://huggingface.co/docs/transformers/model_doc/segformer)** (from NVIDIA) released with the paper [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://arxiv.org/abs/2105.15203) by Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M. Alvarez, Ping Luo.
+1. **[Segment Anything](https://huggingface.co/docs/transformers/model_doc/sam)** (from Meta AI) released with the paper [Segment Anything](https://arxiv.org/pdf/2304.02643v1.pdf) by Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alex Berg, Wan-Yen Lo, Piotr Dollar, Ross Girshick.
 1. **[SigLIP](https://huggingface.co/docs/transformers/main/model_doc/siglip)** (from Google AI) released with the paper [Sigmoid Loss for Language Image Pre-Training](https://arxiv.org/abs/2303.15343) by Xiaohua Zhai, Basil Mustafa, Alexander Kolesnikov, Lucas Beyer.
@@ -335,0 +336,0 @@ 1. **[SpeechT5](https://huggingface.co/docs/transformers/model_doc/speecht5)** (from Microsoft Research) released with the paper [SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing](https://arxiv.org/abs/2110.07205) by Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, Zhihua Wei, Yao Qian, Jinyu Li, Furu Wei.

src/env.js

@@ -32,3 +32,3 @@ /**
 
-const VERSION = '2.13.4';
+const VERSION = '2.14.0';
 
@@ -35,0 +35,0 @@ // Check if various APIs are available (depends on environment)

src/utils/core.js

@@ -112,4 +112,4 @@
  *
- * @param {Array} arr The nested array to calculate dimensions for.
- * @returns {Array} An array containing the dimensions of the input array.
+ * @param {any[]} arr The nested array to calculate dimensions for.
+ * @returns {number[]} An array containing the dimensions of the input array.
  */
@@ -116,0 +116,0 @@ export function calculateDimensions(arr) {

src/utils/hub.js

@@ -401,3 +401,3 @@
             .replaceAll('{model}', path_or_repo_id)
-            .replaceAll('{revision}', revision),
+            .replaceAll('{revision}', encodeURIComponent(revision)),
         filename
@@ -404,0 +404,0 @@ );

src/utils/image.js

@@ -93,2 +93,6 @@
 
+    /** 
+     * Returns the size of the image (width, height).
+     * @returns {[number, number]} The size of the image (width, height).
+     */
     get size() {
@@ -169,2 +173,6 @@ return [this.width, this.height];
     static fromTensor(tensor, channel_format = 'CHW') {
+        if (tensor.dims.length !== 3) {
+            throw new Error(`Tensor should have 3 dimensions, but has ${tensor.dims.length} dimensions.`);
+        }
+
         if (channel_format === 'CHW') {
@@ -171,0 +179,0 @@ tensor = tensor.transpose(1, 2, 0);

types/env.d.ts

@@ -22,3 +22,3 @@ export namespace env {
 declare const __dirname: any;
-declare const VERSION: "2.13.4";
+declare const VERSION: "2.14.0";
 declare const localModelPath: any;
@@ -25,0 +25,0 @@ declare const FS_AVAILABLE: boolean;

types/processors.d.ts

@@ -115,2 +115,16 @@ declare const FeatureExtractor_base: new () => {
     /**
+     * Find the target (width, height) dimension of the output image after
+     * resizing given the input image and the desired size.
+     * @param {RawImage} image The image to resize.
+     * @param {any} size The size to use for resizing the image.
+     * @returns {[number, number]} The target (width, height) dimension of the output image after resizing.
+     */
+    get_resize_output_image_size(image: RawImage, size: any): [number, number];
+    /**
+     * Resizes the image.
+     * @param {RawImage} image The image to resize.
+     * @returns {Promise<RawImage>} The resized image.
+     */
+    resize(image: RawImage): Promise<RawImage>;
+    /**
      * @typedef {object} PreprocessedImage
@@ -178,2 +192,8 @@ * @property {HeightWidth} original_size The original size of the image.
 export class ConvNextFeatureExtractor extends ImageFeatureExtractor {
+    constructor(config: any);
+    /**
+     * Percentage of the image to crop. Only has an effect if this.size < 384.
+     */
+    crop_pct: any;
+    resize(image: any): Promise<any>;
 }
@@ -324,13 +344,32 @@ export class ConvNextImageProcessor extends ConvNextFeatureExtractor {
  * @property {HeightWidth[]} reshaped_input_sizes
- * @property {Tensor} input_points
+ * @property {Tensor} [input_points]
+ * @property {Tensor} [input_labels]
  */
 export class SamImageProcessor extends ImageFeatureExtractor {
     /**
-     * @param {RawImage[]} images The image(s) to extract features from.
-     * @param {*} input_points A 3D or 4D array, representing the input points provided by the user.
+     *
+     * @param {any} input_points
+     * @param {HeightWidth[]} original_sizes
+     * @param {HeightWidth[]} reshaped_input_sizes
+     * @returns {Tensor}
+     */
+    reshape_input_points(input_points: any, original_sizes: HeightWidth[], reshaped_input_sizes: HeightWidth[]): Tensor;
+    /**
+     *
+     * @param {any} input_labels
+     * @param {Tensor} input_points
+     * @returns {Tensor}
+     */
+    add_input_labels(input_labels: any, input_points: Tensor): Tensor;
+    /**
+     * @param {any[]} images The URL(s) of the image(s) to extract features from.
+     * @param {any} [input_points] A 3D or 4D array, representing the input points provided by the user.
      * - 3D: `[point_batch_size, nb_points_per_image, 2]`. In this case, `batch_size` is assumed to be 1.
      * - 4D: `[batch_size, point_batch_size, nb_points_per_image, 2]`.
+     * @param {any} [input_labels] A 2D or 3D array, representing the input labels for the points, used by the prompt encoder to encode the prompt.
+     * - 2D: `[point_batch_size, nb_points_per_image]`. In this case, `batch_size` is assumed to be 1.
+     * - 3D: `[batch_size, point_batch_size, nb_points_per_image]`.
      * @returns {Promise<SamImageProcessorResult>}
      */
-    _call(images: RawImage[], input_points: any): Promise<SamImageProcessorResult>;
+    _call(images: any[], input_points?: any, input_labels?: any): Promise<SamImageProcessorResult>;
     /**
@@ -520,7 +559,5 @@ * Remove padding and upscale masks to the original image size.
     /**
-     * @param {*} images
-     * @param {*} input_points
-     * @returns {Promise<any>}
+     * @borrows SamImageProcessor#_call as _call
      */
-    _call(images: any, input_points: any): Promise<any>;
+    _call(...args: any[]): Promise<any>;
     /**
@@ -530,2 +567,6 @@ * @borrows SamImageProcessor#post_process_masks as post_process_masks
     post_process_masks(...args: any[]): any;
+    /**
+     * @borrows SamImageProcessor#reshape_input_points as reshape_input_points
+     */
+    reshape_input_points(...args: any[]): any;
 }
@@ -673,3 +714,4 @@ /**
     reshaped_input_sizes: HeightWidth[];
-    input_points: Tensor;
+    input_points?: Tensor;
+    input_labels?: Tensor;
 };
@@ -676,0 +718,0 @@ import { RawImage } from './utils/image.js';

types/utils/core.d.ts

@@ -56,6 +56,6 @@ /**
  *
- * @param {Array} arr The nested array to calculate dimensions for.
- * @returns {Array} An array containing the dimensions of the input array.
+ * @param {any[]} arr The nested array to calculate dimensions for.
+ * @returns {number[]} An array containing the dimensions of the input array.
  */
-export function calculateDimensions(arr: any[]): any[];
+export function calculateDimensions(arr: any[]): number[];
 /**
@@ -62,0 +62,0 @@ * Replicate python's .pop() method for objects.

types/utils/image.d.ts

@@ -48,4 +48,8 @@ export class RawImage {
     channels: 2 | 1 | 3 | 4;
-    get size(): number[];
     /**
+     * Returns the size of the image (width, height).
+     * @returns {[number, number]} The size of the image (width, height).
+     */
+    get size(): [number, number];
+    /**
      * Convert the image to grayscale format.
@@ -52,0 +56,0 @@ * @returns {RawImage} `this` to support chaining.

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

46250212

increased by0.11%

Lines of code

55394

increased by1.08%

Number of lines in readme file

352

increased by0.28%

No dependency changes