Skip to content

Commit c51dea8

Browse files
merveenoyanosanseviero
merveenoyan
and
osanseviero
authored
Update tasks with new models (#460)
--------- Co-authored-by: Omar Sanseviero <[email protected]>
1 parent 226c4b5 commit c51dea8

File tree

6 files changed

+92
-55
lines changed

6 files changed

+92
-55
lines changed

‎packages/tasks/src/tasks/depth-estimation/data.ts

Lines changed: 8 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -28,8 +28,8 @@ const taskData: TaskDataCustom = {
2828
id: "Intel/dpt-large",
2929
},
3030
{
31-
description: "Strong Depth Estimation model trained on the KITTI dataset.",
32-
id: "facebook/dpt-dinov2-large-kitti",
31+
description: "Strong Depth Estimation model trained on a big compilation of datasets.",
32+
id: "LiheYoung/depth-anything-large-hf",
3333
},
3434
{
3535
description: "A strong monocular depth estimation model.",
@@ -42,8 +42,12 @@ const taskData: TaskDataCustom = {
4242
id: "radames/dpt-depth-estimation-3d-voxels",
4343
},
4444
{
45-
description: "An application that can estimate the depth in a given image.",
46-
id: "keras-io/Monocular-Depth-Estimation",
45+
description: "An application to compare the outputs of different depth estimation models.",
46+
id: "LiheYoung/Depth-Anything",
47+
},
48+
{
49+
description: "An application to try state-of-the-art depth estimation.",
50+
id: "merve/compare_depth_models",
4751
},
4852
],
4953
summary: "Depth estimation is the task of predicting depth of the objects present in an image.",

‎packages/tasks/src/tasks/mask-generation/data.ts

Lines changed: 42 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -3,14 +3,51 @@ import type { TaskDataCustom } from "..";
33
const taskData: TaskDataCustom = {
44
datasets: [],
55
demo: {
6-
inputs: [],
7-
outputs: [],
6+
inputs: [
7+
{
8+
filename: "mask-generation-input.png",
9+
type: "img",
10+
},
11+
],
12+
outputs: [
13+
{
14+
filename: "mask-generation-output.png",
15+
type: "img",
16+
},
17+
],
818
},
919
metrics: [],
10-
models: [],
11-
spaces: [],
20+
models: [
21+
{
22+
description: "Small yet powerful mask generation model.",
23+
id: "Zigeng/SlimSAM-uniform-50",
24+
},
25+
{
26+
description: "Very strong mask generation model.",
27+
id: "facebook/sam-vit-huge",
28+
},
29+
],
30+
spaces: [
31+
{
32+
description:
33+
"An application that combines a mask generation model with an image embedding model for open-vocabulary image segmentation.",
34+
id: "SkalskiP/SAM_and_MetaCLIP",
35+
},
36+
{
37+
description: "An application that compares the performance of a large and a small mask generation model.",
38+
id: "merve/slimsam",
39+
},
40+
{
41+
description: "An application based on an improved mask generation model.",
42+
id: "linfanluntan/Grounded-SAM",
43+
},
44+
{
45+
description: "An application to remove objects from videos using mask generation models.",
46+
id: "SkalskiP/SAM_and_ProPainter",
47+
},
48+
],
1249
summary:
13-
"Mask generation is creating a binary image that identifies a specific object or region of interest in an input image. Masks are often used in segmentation tasks, where they provide a precise way to isolate the object of interest for further processing or analysis.",
50+
"Mask generation is the task of generating masks that identify a specific object or region of interest in a given image. Masks are often used in segmentation tasks, where they provide a precise way to isolate the object of interest for further processing or analysis.",
1451
widgetModels: [],
1552
youtubeId: "",
1653
};

‎packages/tasks/src/tasks/text-generation/data.ts

Lines changed: 22 additions & 38 deletions
Original file line numberDiff line numberDiff line change
@@ -12,12 +12,12 @@ const taskData: TaskDataCustom = {
1212
id: "the_pile",
1313
},
1414
{
15-
description: "A crowd-sourced instruction dataset to develop an AI assistant.",
16-
id: "OpenAssistant/oasst1",
15+
description: "Truly open-source, curated and cleaned dialogue dataset.",
16+
id: "HuggingFaceH4/ultrachat_200k",
1717
},
1818
{
19-
description: "A crowd-sourced instruction dataset created by Databricks employees.",
20-
id: "databricks/databricks-dolly-15k",
19+
description: "An instruction dataset with preference ratings on responses.",
20+
id: "openbmb/UltraFeedback",
2121
},
2222
],
2323
demo: {
@@ -59,66 +59,50 @@ const taskData: TaskDataCustom = {
5959
id: "bigcode/starcoder",
6060
},
6161
{
62-
description: "A model trained to follow instructions, uses Pythia-12b as base model.",
63-
id: "databricks/dolly-v2-12b",
62+
description: "A very powerful text generation model.",
63+
id: "mistralai/Mixtral-8x7B-Instruct-v0.1",
6464
},
6565
{
66-
description: "A model trained to follow instructions curated by community, uses Pythia-12b as base model.",
67-
id: "OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5",
66+
description: "Small yet powerful text generation model.",
67+
id: "microsoft/phi-2",
6868
},
6969
{
70-
description: "A large language model trained to generate text in English.",
71-
id: "stabilityai/stablelm-tuned-alpha-7b",
70+
description: "A very powerful model that can chat, do mathematical reasoning and write code.",
71+
id: "openchat/openchat-3.5-0106",
7272
},
7373
{
74-
description: "A model trained to follow instructions, based on mosaicml/mpt-7b.",
75-
id: "mosaicml/mpt-7b-instruct",
74+
description: "Very strong yet small assistant model.",
75+
id: "HuggingFaceH4/zephyr-7b-beta",
7676
},
7777
{
78-
description: "A large language model trained to generate text in English.",
79-
id: "EleutherAI/pythia-12b",
80-
},
81-
{
82-
description: "A large text-to-text model trained to follow instructions.",
83-
id: "google/flan-ul2",
84-
},
85-
{
86-
description: "A large and powerful text generation model.",
87-
id: "tiiuae/falcon-40b",
88-
},
89-
{
90-
description: "State-of-the-art open-source large language model.",
78+
description: "Very strong open-source large language model.",
9179
id: "meta-llama/Llama-2-70b-hf",
9280
},
9381
],
9482
spaces: [
9583
{
96-
description: "A robust text generation model that can perform various tasks through natural language prompting.",
97-
id: "huggingface/bloom_demo",
84+
description: "A leaderboard to compare different open-source text generation models based on various benchmarks.",
85+
id: "HuggingFaceH4/open_llm_leaderboard",
9886
},
9987
{
100-
description: "An text generation based application that can write code for 80+ languages.",
101-
id: "bigcode/bigcode-playground",
88+
description: "An text generation based application based on a very powerful LLaMA2 model.",
89+
id: "ysharma/Explore_llamav2_with_TGI",
10290
},
10391
{
104-
description: "An text generation based application for conversations.",
105-
id: "h2oai/h2ogpt-chatbot",
92+
description: "An text generation based application to converse with Zephyr model.",
93+
id: "HuggingFaceH4/zephyr-chat",
10694
},
10795
{
10896
description: "An text generation application that combines OpenAI and Hugging Face models.",
10997
id: "microsoft/HuggingGPT",
11098
},
11199
{
112-
description: "An text generation application that uses StableLM-tuned-alpha-7b.",
113-
id: "stabilityai/stablelm-tuned-alpha-chat",
114-
},
115-
{
116-
description: "An UI that uses StableLM-tuned-alpha-7b.",
117-
id: "togethercomputer/OpenChatKit",
100+
description: "An chatbot to converse with a very powerful text generation model.",
101+
id: "mlabonne/phixtral-chat",
118102
},
119103
],
120104
summary:
121-
"Generating text is the task of producing new text. These models can, for example, fill in incomplete text or paraphrase.",
105+
"Generating text is the task of generating new text given another text. These models can, for example, fill in incomplete text or paraphrase.",
122106
widgetModels: ["HuggingFaceH4/zephyr-7b-beta"],
123107
youtubeId: "Vpjb1lu0MDk",
124108
};

‎packages/tasks/src/tasks/text-to-image/data.ts

Lines changed: 6 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -79,13 +79,17 @@ const taskData: TaskDataCustom = {
7979
id: "latent-consistency/lcm-lora-for-sdxl",
8080
},
8181
{
82-
description: "A powerful text-to-image application that can generate 3D representations.",
83-
id: "hysts/Shap-E",
82+
description: "A gallery to explore various text-to-image models.",
83+
id: "multimodalart/LoraTheExplorer",
8484
},
8585
{
8686
description: "An application for `text-to-image`, `image-to-image` and image inpainting.",
8787
id: "ArtGAN/Stable-Diffusion-ControlNet-WebUI",
8888
},
89+
{
90+
description: "An application to generate realistic images given photos of a person and a prompt.",
91+
id: "InstantX/InstantID",
92+
},
8993
],
9094
summary:
9195
"Generates images from input text. These models can be used to generate and modify images based on text prompts.",

‎packages/tasks/src/tasks/zero-shot-image-classification/data.ts

Lines changed: 8 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -52,9 +52,8 @@ const taskData: TaskDataCustom = {
5252
id: "openai/clip-vit-base-patch16",
5353
},
5454
{
55-
description:
56-
"Robust image classification model trained on publicly available image-caption data trained on additional high pixel data for better performance.",
57-
id: "openai/clip-vit-large-patch14-336",
55+
description: "Strong zero-shot image classification model.",
56+
id: "google/siglip-base-patch16-224",
5857
},
5958
{
6059
description: "Strong image classification model for biomedical domain.",
@@ -64,12 +63,16 @@ const taskData: TaskDataCustom = {
6463
spaces: [
6564
{
6665
description:
67-
"An application that leverages zero shot image classification to find best captions to generate an image. ",
66+
"An application that leverages zero-shot image classification to find best captions to generate an image. ",
6867
id: "pharma/CLIP-Interrogator",
6968
},
69+
{
70+
description: "An application to compare different zero-shot image classification models. ",
71+
id: "merve/compare_clip_siglip",
72+
},
7073
],
7174
summary:
72-
"Zero shot image classification is the task of classifying previously unseen classes during training of a model.",
75+
"Zero-shot image classification is the task of classifying previously unseen classes during training of a model.",
7376
widgetModels: ["openai/clip-vit-large-patch14-336"],
7477
youtubeId: "",
7578
};

‎packages/tasks/src/tasks/zero-shot-object-detection/data.ts

Lines changed: 6 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -47,7 +47,12 @@ const taskData: TaskDataCustom = {
4747
id: "google/owlv2-base-patch16-ensemble",
4848
},
4949
],
50-
spaces: [],
50+
spaces: [
51+
{
52+
description: "A demo to try the state-of-the-art zero-shot object detection model, OWLv2.",
53+
id: "merve/owlv2",
54+
},
55+
],
5156
summary:
5257
"Zero-shot object detection is a computer vision task to detect objects and their classes in images, without any prior training or knowledge of the classes. Zero-shot object detection models receive an image as input, as well as a list of candidate classes, and output the bounding boxes and labels where the objects have been detected.",
5358
widgetModels: [],

0 commit comments

Comments
 (0)