Remove images

README.md

@@ -55,7 +55,7 @@ To calculate the score, pass a list of samples and a similarity function or a st
 - **k**: a pairwise similarity function, or a string identifying a predefined 
        similarity function. If k is a pairwise similarity function, it should
        be symmetric and k(x, x) = 1.
-       Options: ngram_overlap, text_embeddings, pixels, image_embeddings.
+       Options: ngram_overlap, text_embeddings.
 - **score_K**: if true, samples is an n x n similarity matrix K.
 - **score_X**: if true, samples is an n x d feature matrix X.
 - **score_dual**: if true,  samples is an n x d feature matrix X and we will
@@ -63,20 +63,15 @@ To calculate the score, pass a list of samples and a similarity function or a st
 - **normalize**: if true, normalize the similarity scores.
 - **model (optional)**: if k is "text_embeddings", a model mapping sentences to
        embeddings (output should be an object with an attribute called
-       `pooler_output` or `last_hidden_state`). If k is "image_embeddings", a
-       model mapping images to embeddings.
+       `pooler_output` or `last_hidden_state`).
 - **tokenizer (optional)**: if k is "text_embeddings" or "ngram_overlap", a
        tokenizer mapping strings to lists.
-- **transform (optional)**: if k is "image_embeddings", a torchvision transform
-       to apply to the samples.
 - **model_path (optional)**: if k is "text_embeddings", the name of a model on
        the HuggingFace hub.
 - **ns (optional)**: if k is "ngram_overlap", the values of n to calculate.
-- **batch_size (optional)**: batch size to use if k is "text_embedding" or
-       "image_embedding".
+- **batch_size (optional)**: batch size to use if k is "text_embedding".
 - **device (optional)**: a string (e.g. "cuda", "cpu") or torch.device 
-       identifying the device to use if k is "text_embedding"
-       or "image_embedding".
+       identifying the device to use if k is "text_embedding".
 
 
 ### Output Values
@@ -116,27 +111,6 @@ to compute the Vendi Score using the covariance matrix, `X @ X.T`.
 {'VS': 1.99989...}
 ```
 
-Image similarity can be calculated using inner products between pixel vectors or between embeddings from a neural network.
-The default embeddings are from the pool-2048 layer of the torchvision version of the Inception v3 model; other embedding functions can be passed to the `model` argument.
-```
->>> from torchvision import datasets
->>> mnist = datasets.MNIST("data/mnist", train=False, download=True)
->>> digits = [[x for x, y in mnist if y == c] for c in range(10)]
->>> pixel_vs = [vendiscore.compute(samples=imgs, k="pixels") for imgs in digits]
->>> inception_vs = [vendiscore.compute(samples=imgs, k="image_embeddings", batch_size=64, device="cuda") for imgs in digits]
->>> for y, (pvs, ivs) in enumerate(zip(pixel_vs, inception_vs)): print(f"{y}\t{pvs:.02f}\t{ivs:02f}")
-0       7.68    3.45
-1       5.31    3.50
-2       12.18   3.62
-3       9.97    2.97
-4       11.10   3.75
-5       13.51   3.16
-6       9.06    3.63
-7       9.58    4.07
-8       9.69    3.74
-9       8.56    3.43
-```
-
 Text similarity can be calculated using n-gram overlap or using inner products between embeddings from a neural network.
 ```
 >>> vendiscore = evaluate.load("danf0/vendiscore", "text")

vendiscore.py

@@ -14,10 +14,8 @@
 import evaluate
 import datasets
 import numpy as np
-import PIL
-from PIL import Image
 
-from vendi_score import vendi, image_utils, text_utils
+from vendi_score import vendi, text_utils
 
 # TODO: Add BibTeX citation
 _CITATION = ""
@@ -36,30 +34,26 @@ Args:
        matrix K, or an n x d feature matrix X.
    k: a pairwise similarity function, or a string identifying a predefined 
        similarity function.
-       Options: ngram_overlap, text_embeddings, pixels, image_embeddings.
+       Options: ngram_overlap, text_embeddings.
    score_K: if true, samples is an n x n similarity matrix K.
    score_X: if true, samples is an n x d feature matrix X.
    score_dual: if true, compute diversity score of X @ X.T.
    normalize: if true, normalize the similarity scores.
    model (optional): if k is "text_embeddings", a model mapping sentences to
        embeddings (output should be an object with an attribute called
-       `pooler_output` or `last_hidden_state`). If k is "image_embeddings", a
-       model mapping images to embeddings.
+       `pooler_output` or `last_hidden_state`).
    tokenizer (optional): if k is "text_embeddings" or "ngram_overlap", a
        tokenizer mapping strings to lists.
-   transform (optional): if k is "image_embeddings", a torchvision transform
-       to apply to the samples.
    model_path (optional): if k is "text_embeddings", the name of a model on the
        HuggingFace hub.
    ns (optional): if k is "ngram_overlap", the values of n to calculate.
-   batch_size (optional): batch size to use if k is "text_embedding" or
-       "image_embedding".
+   batch_size (optional): batch size to use if k is "text_embedding".
    device (optional): a string (e.g. "cuda", "cpu") or torch.device identifying
-       the device to use if k is "text_embedding or "image_embedding".
+       the device to use if k is "text_embedding".
 Returns:
     VS: The Vendi Score. 
 Examples:
-    >>> vendiscore = evaluate.load("danf0/vendiscore")
+    >>> vendiscore = evaluate.load("danf0/vendiscore", "text")
     >>> samples = ["Look, Jane.",
                    "See Spot.",
                    "See Spot run.",
@@ -74,11 +68,8 @@ Examples:
 def get_features(config_name):
     if config_name in ("text", "default"):
         return datasets.Features({"samples": datasets.Value("string")})
-    if config_name == "image":
-        return [
-            datasets.Features({"samples": datasets.Array2D}),
-            datasets.Features({"samples": datasets.Array3D}),
-        ]
+    # if config_name == "image":
+    #     return datasets.Features({"samples": datasets.Image})
     if config_name in ("K", "X"):
         return [
             datasets.Features(
@@ -130,7 +121,6 @@ class VendiScore(evaluate.Metric):
         normalize=False,
         model=None,
         tokenizer=None,
-        transform=None,
         model_path=None,
         ns=[1, 2],
         batch_size=16,
@@ -155,18 +145,18 @@ class VendiScore(evaluate.Metric):
                 device=device,
                 model_path=model_path,
             )
-        elif type(k) == str and k == "pixels":
-            vs = image_utils.pixel_vendi_score(
-                [Image.fromarray(x) for x in samples]
-            )
-        elif type(k) == str and k == "image_embeddings":
-            vs = image_utils.embedding_vendi_score(
-                [Image.fromarray(x) for x in samples],
-                batch_size=batch_size,
-                device=device,
-                model=model,
-                transform=transform,
-            )
+        # elif type(k) == str and k == "pixels":
+        #     vs = image_utils.pixel_vendi_score(
+        #         [Image.fromarray(x) for x in samples]
+        #     )
+        # elif type(k) == str and k == "image_embeddings":
+        #     vs = image_utils.embedding_vendi_score(
+        #         [Image.fromarray(x) for x in samples],
+        #         batch_size=batch_size,
+        #         device=device,
+        #         model=model,
+        #         transform=transform,
+        #     )
         else:
             vs = vendi.score(samples, k)
         return {"VS": vs}