Merge branch 'huggingface:main' into main

packages/tasks/src/tasks/text-generation/about.md

@@ -32,6 +32,16 @@ The most popular models for this task are GPT-based models, [Mistral](mistralai/
 
 These models are trained to learn the mapping between a pair of texts (e.g. translation from one language to another). The most popular variants of these models are [NLLB](facebook/nllb-200-distilled-600M), [FLAN-T5](https://huggingface.co/google/flan-t5-xxl), and [BART](https://huggingface.co/docs/transformers/model_doc/bart). Text-to-Text models are trained with multi-tasking capabilities, they can accomplish a wide range of tasks, including summarization, translation, and text classification.
 
+## Language Model Variants
+
+When it comes to text generation, the underlying language model can come in several types:
+
+- **Base models:** refers to plain language models like [Mistral 7B](mistralai/Mistral-7B-v0.1) and [Llama-2-70b](https://huggingface.co/meta-llama/Llama-2-70b-hf). These models are good for fine-tuning and few-shot prompting.
+
+- **Instruction-trained models:** these models are trained in a multi-task manner to follow a broad range of instructions like "Write me a recipe for chocolate cake". Models like [Flan-T5](https://huggingface.co/google/flan-t5-xl), [Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1), and [falcon-40b-instruct](https://huggingface.co/tiiuae/falcon-40b-instruct) are examples of instruction-trained models. In general, instruction-trained models will produce better responses to instructions than base models.
+
+- **Human feedback models:** these models extend base and instruction-trained models by incorporating human feedback that rates the quality of the generated text according to criteria like [helpfulness, honesty, and harmlessness](https://arxiv.org/abs/2112.00861). The human feedback is then combined with an optimization technique like reinforcement learning to align the original model to be closer with human preferences. The overall methodology is often called [Reinforcement Learning from Human Feedback](https://huggingface.co/blog/rlhf), or RLHF for short. [Llama2-Chat](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf) is an open-source model aligned through human feedback.
+
 ## Inference
 
 You can use the 🤗 Transformers library `text-generation` pipeline to do inference with Text Generation models. It takes an incomplete text and returns multiple outputs with which the text can be completed.

packages/widgets/src/lib/components/InferenceWidget/shared/WidgetOutputChart/WidgetOutputChart.svelte

@@ -16,13 +16,13 @@ from-yellow-400 to-yellow-200 dark:from-yellow-400 dark:to-yellow-600
 	type LabelField = "label" | "answer";
 	export let labelField: LabelField = "label";
 	export let output: Array<
-		{ label: string; score: number; color?: string } | { answer: string; score: number; color?: string }
+		{ label: string; score?: number; color?: string } | { answer: string; score?: number; color?: string }
 	> = [];
 	export let highlightIndex = -1;
 	export let mouseover: (index: number) => void = () => {};
 	export let mouseout: () => void = () => {};
 
-	$: scoreMax = Math.max(0, ...output.map((x) => x.score));
+	$: scoreMax = Math.max(0, ...output.map((x) => x.score ?? 0));
 
 	function text(outputItem: (typeof output)[0]) {
 		if (labelField in outputItem) {
@@ -54,11 +54,13 @@ from-yellow-400 to-yellow-200 dark:from-yellow-400 dark:to-yellow-600
 							to-{color ?? defaultBarColor}-200
 							dark:from-{color ?? defaultBarColor}-400
 							dark:to-{color ?? defaultBarColor}-600"
-						style={`width: ${Math.ceil((score / scoreMax) * 100 * 0.8)}%;`}
+						style={`width: ${score ? Math.ceil((score / scoreMax) * 100 * 0.8) : 0}%;`}
 					/>
 					<span class="leading-snug">{text(output[index])}</span>
 				</div>
-				<span class="pl-2">{score.toFixed(3)}</span>
+				{#if typeof score === "number"}
+					<span class="pl-2">{score.toFixed(3)}</span>
+				{/if}
 			</div>
 		{/each}
 	</div>

packages/widgets/src/lib/components/InferenceWidget/widgets/VisualQuestionAnsweringWidget/VisualQuestionAnsweringWidget.svelte

@@ -25,7 +25,7 @@
 		isLoading: false,
 		estimatedTime: 0,
 	};
-	let output: Array<{ answer: string; score: number }> | null = [];
+	let output: Array<{ answer: string; score?: number }> | null = [];
 	let outputJson: string;
 	let question = "";
 	let imgSrc = "";
@@ -55,15 +55,18 @@
 		});
 	}
 
-	function isValidOutput(arg: any): arg is { answer: string; score: number }[] {
-		return Array.isArray(arg) && arg.every((x) => typeof x.answer === "string" && typeof x.score === "number");
+	function isValidOutput(arg: any): arg is { answer: string; score?: number }[] {
+		return (
+			Array.isArray(arg) &&
+			arg.every((x) => typeof x.answer === "string" && (typeof x.score === "number" || x.score === undefined))
+		);
 	}
 
-	function parseOutput(body: unknown): Array<{ answer: string; score: number }> {
+	function parseOutput(body: unknown): Array<{ answer: string; score?: number }> {
 		if (isValidOutput(body)) {
 			return body;
 		}
-		throw new TypeError("Invalid output: output must be of type Array<answer: string, score:number>");
+		throw new TypeError("Invalid output: output must be of type Array<{ answer: string, score?: number }>");
 	}
 
 	async function applyInputSample(sample: WidgetExampleAssetAndTextInput, opts: ExampleRunOpts = {}) {