@huggingface/tasks 0.19.2 → 0.19.4

package/src/model-libraries-snippets.ts

@@ -237,6 +237,20 @@ output = model.generate(text)
 sf.write("simple.mp3", output, 44100)`,
 ];
 
+export const describe_anything = (model: ModelData): string[] => [
+	`# pip install git+https://github.com/NVlabs/describe-anything
+from huggingface_hub import snapshot_download
+from dam import DescribeAnythingModel
+
+snapshot_download(${model.id}, local_dir="checkpoints")
+
+dam = DescribeAnythingModel(
+	model_path="checkpoints",
+	conv_mode="v1",
+	prompt_mode="focal_prompt",
+)`,
+];
+
 const diffusersDefaultPrompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k";
 
 const diffusers_default = (model: ModelData) => [
@@ -727,6 +741,17 @@ model = pe.VisionTransformer.from_config("${model.id}", pretrained=True)`;
 		return [vision_encoder];
 	}
 };
+export const phantom_wan = (model: ModelData): string[] => [
+	`from huggingface_hub import snapshot_download
+from phantom_wan import WANI2V, configs
+
+checkpoint_dir = snapshot_download("${model.id}")
+wan_i2v = WanI2V(
+            config=configs.WAN_CONFIGS['i2v-14B'],
+            checkpoint_dir=checkpoint_dir,
+        )
+ video = wan_i2v.generate(text_prompt, image_prompt)`,
+];
 
 export const pyannote_audio_pipeline = (model: ModelData): string[] => [
 	`from pyannote.audio import Pipeline
@@ -1117,13 +1142,31 @@ export const transformers = (model: ModelData): string[] => {
 	}
 
 	if (model.pipeline_tag && LIBRARY_TASK_MAPPING.transformers?.includes(model.pipeline_tag)) {
-		const pipelineSnippet = ["# Use a pipeline as a high-level helper", "from transformers import pipeline", ""];
+		const pipelineSnippet = [
+			"# Use a pipeline as a high-level helper",
+			"from transformers import pipeline",
+			"",
+			`pipe = pipeline("${model.pipeline_tag}", model="${model.id}"` + remote_code_snippet + ")",
+		];
 
-		if (model.tags.includes("conversational") && model.config?.tokenizer_config?.chat_template) {
-			pipelineSnippet.push("messages = [", '    {"role": "user", "content": "Who are you?"},', "]");
-		}
-		pipelineSnippet.push(`pipe = pipeline("${model.pipeline_tag}", model="${model.id}"` + remote_code_snippet + ")");
-		if (model.tags.includes("conversational") && model.config?.tokenizer_config?.chat_template) {
+		if (model.tags.includes("conversational")) {
+			if (model.tags.includes("image-text-to-text")) {
+				pipelineSnippet.push(
+					"messages = [",
+					[
+						"    {",
+						'        "role": "user",',
+						'        "content": [',
+						'            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},',
+						'            {"type": "text", "text": "What animal is on the candy?"}',
+						"        ]",
+						"    },",
+					].join("\n"),
+					"]"
+				);
+			} else {
+				pipelineSnippet.push("messages = [", '    {"role": "user", "content": "Who are you?"},', "]");
+			}
 			pipelineSnippet.push("pipe(messages)");
 		}
 
@@ -1309,19 +1352,53 @@ model = SwarmFormerModel.from_pretrained("${model.id}")
 `,
 ];
 
-export const mlx = (model: ModelData): string[] => [
+const mlx_unknown = (model: ModelData): string[] => [
 	`pip install huggingface_hub hf_transfer
 
 export HF_HUB_ENABLE_HF_TRANSFER=1
 huggingface-cli download --local-dir ${nameWithoutNamespace(model.id)} ${model.id}`,
 ];
 
+const mlxlm = (model: ModelData): string[] => [
+	`pip install --upgrade mlx-lm
+
+mlx_lm.generate --model ${model.id} --prompt "Hello"`,
+];
+
+const mlxchat = (model: ModelData): string[] => [
+	`pip install --upgrade mlx-lm
+
+mlx_lm.chat --model ${model.id}`,
+];
+
+const mlxvlm = (model: ModelData): string[] => [
+	`pip install --upgrade mlx-vlm
+
+mlx_vlm.generate --model ${model.id} \\
+	--prompt "Describe this image." \\
+	--image "https://huggingface.co/datasets/huggingface/documentation-images/resolve/0052a70beed5bf71b92610a43a52df6d286cd5f3/diffusers/rabbit.jpg"`,
+];
+
 export const mlxim = (model: ModelData): string[] => [
 	`from mlxim.model import create_model
 
 model = create_model(${model.id})`,
 ];
 
+export const mlx = (model: ModelData): string[] => {
+	if (model.tags.includes("image-text-to-text")) {
+		return mlxvlm(model);
+	}
+	if (model.tags.includes("conversational")) {
+		if (model.config?.tokenizer_config?.chat_template) {
+			return mlxchat(model);
+		} else {
+			return mlxlm(model);
+		}
+	}
+	return mlx_unknown(model);
+};
+
 export const model2vec = (model: ModelData): string[] => [
 	`from model2vec import StaticModel
 
@@ -1448,4 +1525,23 @@ export const hezar = (model: ModelData): string[] => [
 
 model = Model.load("${model.id}")`,
 ];
+
+export const zonos = (model: ModelData): string[] => [
+	`# pip install git+https://github.com/Zyphra/Zonos.git
+import torchaudio
+from zonos.model import Zonos
+from zonos.conditioning import make_cond_dict
+
+model = Zonos.from_pretrained("${model.id}", device="cuda")
+
+wav, sr = torchaudio.load("speaker.wav")           # 5-10s reference clip
+speaker = model.make_speaker_embedding(wav, sr)
+
+cond  = make_cond_dict(text="Hello, world!", speaker=speaker, language="en-us")
+codes = model.generate(model.prepare_conditioning(cond))
+
+audio = model.autoencoder.decode(codes)[0].cpu()
+torchaudio.save("sample.wav", audio, model.autoencoder.sampling_rate)
+`,
+];
 //#endregion

package/src/model-libraries.ts

@@ -230,6 +230,13 @@ export const MODEL_LIBRARIES_UI_ELEMENTS = {
 		filter: false,
 		countDownloads: `path:"scin_dataset_precomputed_embeddings.npz" OR path:"saved_model.pb"`,
 	},
+	"describe-anything": {
+		prettyLabel: "Describe Anything",
+		repoName: "Describe Anything",
+		repoUrl: "https://github.com/NVlabs/describe-anything",
+		snippets: snippets.describe_anything,
+		filter: false,
+	},
 	"dia-tts": {
 		prettyLabel: "Dia",
 		repoName: "Dia",
@@ -662,6 +669,14 @@ export const MODEL_LIBRARIES_UI_ELEMENTS = {
 		snippets: snippets.perception_encoder,
 		countDownloads: `path_extension:"pt"`,
 	},
+	"phantom-wan": {
+		prettyLabel: "Phantom",
+		repoName: "Phantom",
+		repoUrl: "https://github.com/Phantom-video/Phantom",
+		snippets: snippets.phantom_wan,
+		filter: false,
+		countDownloads: `path_extension:"pth"`,
+	},
 	pxia: {
 		prettyLabel: "pxia",
 		repoName: "pxia",
@@ -1009,6 +1024,14 @@ export const MODEL_LIBRARIES_UI_ELEMENTS = {
 		countDownloads: `path_extension:"pt" OR path_extension:"safetensors"`,
 		snippets: snippets.ultralytics,
 	},
+	zonos: {
+		prettyLabel: "Zonos",
+		repoName: "Zonos",
+		repoUrl: "https://github.com/Zyphra/Zonos",
+		docsUrl: "https://github.com/Zyphra/Zonos",
+		snippets: snippets.zonos,
+		filter: false,
+	},
 	"3dtopia-xl": {
 		prettyLabel: "3DTopia-XL",
 		repoName: "3DTopia-XL",

package/src/tasks/index.ts

@@ -184,7 +184,7 @@ export const TASKS_MODEL_LIBRARIES: Record<PipelineType, ModelLibraryKey[]> = {
 /**
  * Return the whole TaskData object for a certain task.
  * If the partialTaskData argument is left undefined,
- * the default placholder data will be used.
+ * the default placeholder data will be used.
  */
 function getData(type: PipelineType, partialTaskData: TaskDataCustom = placeholder): TaskData {
 	return {

package/src/tasks/reinforcement-learning/about.md

@@ -111,7 +111,7 @@ env = gym.make("LunarLander-v2")
 # Loading the saved model
 model = PPO.load("PPO-LunarLander-v2",env=env)
 
-# Initializating the evaluation environment
+# Initializing the evaluation environment
 eval_env = gym.make("LunarLander-v2")
 
 # Running the trained agent on eval_env for 10 time steps and getting the mean reward