tri.promptfx.apps.ImageDescribeView.kt Maven / Gradle / Ivy

Go to download
/*-
 * #%L
 * tri.promptfx:promptfx
 * %%
 * Copyright (C) 2023 - 2024 Johns Hopkins University Applied Physics Laboratory
 * %%
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 * 
 *      http://www.apache.org/licenses/LICENSE-2.0
 * 
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 * #L%
 */
package tri.promptfx.apps

import javafx.beans.property.SimpleObjectProperty
import javafx.scene.image.Image
import tornadofx.*
import tri.ai.core.TextChatRole
import tri.ai.core.VisionLanguageChatMessage
import tri.ai.pips.task
import tri.ai.prompt.AiPromptLibrary
import tri.promptfx.AiPlanTaskView
import tri.promptfx.PromptFxModels
import tri.promptfx.ui.PromptSelectionModel
import tri.promptfx.ui.promptfield
import tri.util.ui.NavigableWorkspaceViewImpl
import tri.util.ui.toUri
import java.net.URI

/** Plugin for the [ImageDescribeView]. */
class ImageDescribePlugin : NavigableWorkspaceViewImpl("Vision", "Image Description", type = ImageDescribeView::class)

/** View designed to describe an image. */
class ImageDescribeView: AiPlanTaskView("Image Description (beta)", "Drop an image to describe into the box on the left.") {

    companion object {
        private const val PROMPT_PREFIX = "image-describe"
    }

    private val image = SimpleObjectProperty(null)
    private val model = SimpleObjectProperty(PromptFxModels.visionLanguageModelDefault())

    private val prompt = PromptSelectionModel("$PROMPT_PREFIX-basic")

    init {
        addInputImageArea(image)
        parameters("Vision Language Model") {
            field("Model") {
                combobox(model, PromptFxModels.visionLanguageModels())
            }
        }
        parameters("Prompt") {
            tooltip("Loads from prompts.yaml with prefix $PROMPT_PREFIX")
            promptfield("Prompt", prompt, AiPromptLibrary.withPrefix(PROMPT_PREFIX), workspace)
        }
        parameters("Model Parameters") {
            with (common) {
                temperature()
                maxTokens()
            }
        }
    }

    override fun plan() = task("Describe Image") {
        describeImage(prompt.text.value)
    }.planner

    fun setImage(image: Image) {
        this.image.value = image
    }

    private suspend fun describeImage(prompt: String): String? {
        val res = model.value.chat(
            listOf(
                VisionLanguageChatMessage(TextChatRole.User, prompt, URI.create(image.value.toUri()))
            ),
            common.temp.value,
            common.maxTokens.value,
            null,
            false
        )
        return res.firstValue!!.content!!
    }

}