Cooking method based on built-in ai large model, and smart cooking apparatus therefor

Abstract

Provided in the present invention are a cooking method based on a built-in AI large model, and a smart cooking apparatus therefor. The smart cooking apparatus is provided with an information collection module, a storage module, an AI large model module, a processor and a wireless communication module, wherein the AI large model module is located in the storage module; the information collection module, the storage module, the AI large model module and the wireless communication module are electrically connected to the processor; the information collection module is used for receiving cooking requirement information input by a user, and pushing the cooking requirement information to the AI large model module located in the storage module; and the AI large model module analyzes the cooking requirement information and generates a new cooking method. By means of the present invention, the application of an AI large model in the field of cooking is realized, and the AI large model is built in the smart cooking apparatus, such that the smart cooking apparatus can autonomously generate a cooking method that meets personalized requirements proposed by a user, and food that meets different requirements are cooked.

Description

A cooking method based on built-in AI large model and intelligent cooking device thereof Technical Field

[0001] The present invention relates to a cooking device and method, and in particular to a cooking method based on a built-in AI large model and an intelligent cooking device thereof. Background Art

[0002] With the development of science and technology and the accelerated pace of life, smart cooking devices have emerged, which can realize automated cooking.

[0003] Regarding the combination of AI big models and smart cooking devices, the latest existing technologies include patent application number CN 116843510 A, published on October 3, 2023, entitled "Smart Cooking Machine Cloud Platform Data Management System and Method Based on AI Big Model," and application number 202310784823.7;

[0004] This invention application discloses a cloud platform data management system and method for intelligent cooking machines based on an AI big model, which relates to the technical field of data management systems. The management method includes the following steps: various parameter data of the cooking process are collected through the collection end, and the parameter data are transmitted to the cloud platform through the network to ensure real-time and stability. The processing end analyzes the stored parameter data through the AI ​​big model to analyze whether there is any abnormality in the intelligent cooking machine during the cooking process. When the intelligent cooking machine is analyzed to have an abnormality, it is judged that there is a deviation in the taste of the dish based on the analysis result, and the management system sends a warning signal. At this time, other intelligent cooking machines are replaced for cooking. The invention can judge whether there is a deviation in the taste of the dish cooked this time based on the operating status of the intelligent cooking machine, so as to be able to issue a timely warning to ensure the quality and taste of the dish.

[0005] The role of applying AI big models to smart cooking machines in the existing technology is only for the processing end to analyze the stored parameter data through the AI ​​big model to analyze whether there is any abnormality in the smart cooking machine during the cooking process. When the smart cooking machine is analyzed to have an abnormality, it is judged based on the analysis results that there is a deviation in the taste of the dish, and it is necessary to replace other smart cooking machines to ensure the quality and taste of the dish. In fact, replacing other smart cooking machines for cooking may not necessarily ensure the quality and taste of the dish, because the other replaced smart cooking machines may also have abnormalities. In this way, users need to constantly replace smart cooking machines to ensure the quality and taste of the dish.

[0006] The present invention provides a cooking method based on a built-in AI big model and an intelligent cooking device thereof. The method uses a third-party AI big model to fine-tune a large amount of cooking data sets. After the training is completed, a cooking AI big model with all relevant data of the cooking method is generated. The user inputs cooking requirement information through voice, video, text, picture or 3D model format, and generates a new cooking method by calling the built-in cooking AI big model. The intelligent cooking device executes the new cooking method generated by the user calling the built-in cooking AI big model to cook food that meets the user's needs. Technical issues

[0007] The role of applying AI big models to smart cooking machines in the existing technology is only for the processing end to analyze the stored parameter data through the AI ​​big model to analyze whether there is any abnormality in the smart cooking machine during the cooking process. When the smart cooking machine is analyzed to have an abnormality, it is judged based on the analysis results that there is a deviation in the taste of the dish, and it is necessary to replace other smart cooking machines to ensure the quality and taste of the dish. In fact, replacing other smart cooking machines for cooking may not necessarily ensure the quality and taste of the dish, because the other replaced smart cooking machines may also have abnormalities. In this way, users need to constantly replace smart cooking machines to ensure the quality and taste of the dish. Technical Solutions

[0008] In order to solve the above-mentioned problems of the prior art, the present invention provides a cooking method based on a built-in AI big model, which is characterized in that it is applied to an intelligent cooking device, and the intelligent cooking device is provided with an information collection module, a storage module, an AI big model module, a processor and a wireless communication module, and the AI ​​big model module is located inside the storage module, and the information collection module, the storage module, the AI ​​big model module, and the wireless communication module are electrically connected to the processor, and the information collection module is used to receive cooking requirement information input by the user in voice, video, text, picture or 3D model format, and push this cooking requirement information to the AI ​​big model module located in the storage module, the AI ​​big model module analyzes the cooking requirement information and generates a new cooking method, and sends the new cooking method to the processor through the storage module, and the new cooking method includes the cooking method of the intelligent cooking device station , cooking work steps, cooking parameters or cooking operation instructions that require user cooperation, the AI ​​big model module generates a new cooking method including: building a cooking AI big model and calling a built-in AI big model for cooking, wherein building a cooking AI big model includes collecting cooking data, preprocessing cooking data, selecting an AI big model that can be applied to cooking, training a cooking AI big model, verifying and testing the cooking AI big model, and deploying and maintaining the cooking AI big model in the AI ​​big model module, the cooking requirement information input by the user in voice, video, text, picture or 3D model format is called by the AI ​​big model module to generate the new cooking method through the built-in AI big model for cooking, the calling of the built-in AI big model for cooking includes assembling query statements, the built-in AI big model performing reasoning operations and the built-in AI big model returning results, and the new cooking method is sent to the processor through the storage module after the reasoning operation of the built-in AI big model for cooking is completed.

[0009] As an improvement to the cooking method based on the built-in AI big model of the present invention, the construction of the cooking AI big model includes:

[0010] Step 1: Collect cooking data: Collect all information about cooking methods of ingredients presented through voice, video, text, pictures or 3D models;

[0011] Step 2: Preprocessing cooking data: All collected information about cooking methods is processed to ensure its integrity and usability. This includes converting information in different formats into text and editing the text according to a specific format to facilitate subsequent training of the AI ​​model.

[0012] Step 3: Select AI models applicable to cooking: Select domestic and international third-party AI models, and measure them using accuracy, response speed, and diversity indicators;

[0013] Step 4: Train the cooking AI model: After organizing the cooking dataset in step 2, fine-tune the cooking dataset using a third-party AI model. After training, a cooking AI model containing all relevant data about cooking methods is generated.

[0014] Step 5: Verify and test the cooking AI model: Perform a task-specific performance test on the cooking AI model generated in step 4. If the performance fails, repeat steps 1, 2, 3, and 4, retraining the model until the performance passes. Generate a cooking AI model and store it on the cloud platform.

[0015] Step 6. Deploy and maintain the cooking AI big model: deploy the newly generated AI big model to the AI ​​big model module of the intelligent cooking device, and perform continuous maintenance and updates, and update data regularly to ensure the timeliness and accuracy of the data.

[0016] As an improvement to the cooking method based on the built-in AI big model of the present invention, the third step selects an AI big model applicable to cooking. The selected AI big model is the Baichuan2-13B AI big model. The parameters of the AI ​​big model are as follows: hidden layer dimension: 5,120, number of layers: 40, number of attention heads: 40, vocabulary size: 64,000, total number of parameters: 13,264,901,120, training data (tokens): 1.4 trillion, position encoding: ALiBi, maximum length: 4,096;

[0017] As an improvement to the cooking method based on the built-in AI large model of the present invention, the training process of the fourth step of training the cooking AI large model is as follows: first, download the model weights of baichuan13b from huggingface, then download the belle dataset train_0.5M_CN to the local computer and put it in the dataset folder under the project directory, and finally run the sft_lora.py script. Then, quantize Baichuan LLM using qlora's nf4 and double quantization methods. Finally, use lora to fine-tune the instructions.

[0018] As an improvement to the cooking method based on the built-in AI big model of the present invention, calling the built-in AI big model for cooking specifically includes the following steps:

[0019] The first step is to assemble the query statement;

[0020] The second step is to use the built-in AI big model for inference calculations: The query statement is passed to the AI ​​big model, which then performs inference calculations. This involves the following steps:

[0021] 1) Understanding the input: Distributed semantic parsing first receives a text sequence and converts it into word vectors. This process is based on the distributed semantics assumption that the meaning of a word is determined by its context.

[0022] 2) Parameter association: context-focus chaining, inputting these word vectors into the Transformer Encoder to generate context representation;

[0023] 3) Generate Answers: Generative probabilistic modeling: The model initializes the Transformer's decoder and feeds the encoder output and the current output sequence into the decoder. The decoder generates a probability distribution for the next word and selects the word with the highest probability or another set probability distribution as the output. This word is then appended to the output sequence.

[0024] 4) Select the most appropriate answer: Dynamic word string evolution, repeat the above steps, adding new words to the output sequence each time, until a complete output sequence is generated;

[0025] Step 3: The built-in AI big model returns the result: After the built-in AI big model completes the inference operation, it returns the information related to the new cooking method and the new cooking method operation instructions, including content provided to the user in text, image, audio, video or 3D model format;

[0026] The present invention provides an intelligent cooking device, characterized in that it is provided with an information collection module, a storage module, an AI big model module, a processor and a wireless communication module. The AI ​​big model module is located inside the storage module, and the information collection module, the storage module, the AI ​​big model module, and the wireless communication module are electrically connected to the processor. The information collection module is used to collect all information about cooking ingredients presented by the user in voice, video, text, picture or 3D model format. The wireless communication module can be used to connect the processor with a cloud platform and a user mobile terminal. The processor is used to execute the cooking method based on the built-in AI big model deployed in the AI ​​big model module in claim 1. After the new cooking method is inferred by the cooking built-in AI big model, it is sent to the intelligent cooking device through the storage module and prompts the user to cook.

[0027] As an improvement to the intelligent cooking device of the present invention, there are intelligent cooking device stations for frying, baking, boiling, air frying, braising, stewing, baking, steaming or stir-frying. The intelligent cooking device receives the new cooking method and prompts the user to cook and perform cooking work steps at the corresponding intelligent cooking device station.

[0028] As an improvement to the intelligent cooking device of the present invention, the intelligent cooking device workstations for frying, baking, boiling, air frying, braising, stewing, baking, steaming or stir-frying are provided with corresponding operation detection feedback systems, which are used to detect whether the cooking operations performed by the user on the intelligent cooking device meet the requirements of the new cooking method.

[0029] As an improvement to the intelligent cooking device of the present invention, a human-computer interaction system is provided, which is used for information exchange between the intelligent cooking device and the user, including the user operating the user mobile terminal to confirm cooking parameters and start cooking operation instructions.

[0030] As an improvement to the intelligent cooking device of the present invention, the new cooking method is sent to the user's mobile terminal through the wireless communication module. The user's mobile terminal receives the new cooking method and prompts the user to cook and perform cooking work steps at the corresponding intelligent cooking device workstation, and the user operates the mobile terminal to confirm cooking parameters and start cooking operation instructions.

[0031] The present invention provides a cooking method based on a built-in AI big model and an intelligent cooking device thereof, which have the following beneficial effects: the present invention provides a cooking method based on a built-in AI big model and an intelligent cooking device thereof, which collects all information on cooking methods of ingredients from people of all countries and regions, all races, and all ages around the world presented through voice, video, text, pictures or 3D models, even including the cooking methods of various foods that a specific individual has been accustomed to or fond of in the past, and uses a third-party AI big model for fine-tuning training. After the training is completed, a cooking AI big model with all relevant data of massive cooking methods is generated. The user inputs cooking requirement information through voice, video, text, pictures or 3D model format, and the intelligent cooking device executes the new cooking method generated by the built-in cooking AI big model by the user to cook food that meets the different needs of different users, and even cooks food that meets the different needs of the same user. In this way, not only the different needs of different users can be met, but also the different needs of the same user at different times and in different environments can be met. Moreover, the AI ​​big model is built into the intelligent cooking device, and the use of the AI ​​big model to cook food by the intelligent cooking device is not affected by the Internet. Beneficial effects

[0032] The present invention provides a cooking method based on a built-in AI big model and an intelligent cooking device thereof. The method uses a third-party AI big model to fine-tune a large amount of cooking data sets. After the training is completed, a cooking AI big model with all relevant data of the cooking method is generated. The user inputs cooking requirement information through voice, video, text, picture or 3D model format, and generates a new cooking method by calling the built-in cooking AI big model. The intelligent cooking device executes the new cooking method generated by the user calling the built-in cooking AI big model to cook food that meets the user's needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] FIG1 is a flowchart of constructing a built-in AI big model for cooking according to a preferred embodiment of a cooking method based on a built-in AI big model and an intelligent cooking device thereof of the present invention.

[0034] FIG2 is a flowchart of calling the built-in AI big model for cooking according to a preferred embodiment of the cooking method based on the built-in AI big model and the intelligent cooking device thereof of the present invention.

[0035] FIG3 is a workflow diagram of a preferred embodiment of the cooking method and intelligent cooking device based on a built-in AI large model of the present invention.

[0036] FIG4 is a workflow diagram of one of other embodiments of the cooking method and intelligent cooking device based on the built-in AI large model of the present invention.

[0037] FIG5 is a workflow diagram of another embodiment of the cooking method and intelligent cooking device based on the built-in AI large model of the present invention.

[0038] FIG6 is a workflow diagram of another embodiment of the cooking method and intelligent cooking device based on the built-in AI large model of the present invention.

[0039] FIG7 is a workflow diagram of another fourth embodiment of the cooking method and intelligent cooking device based on the built-in AI large model of the present invention.

[0040] FIG8 is a workflow diagram of another fifth embodiment of the cooking method and intelligent cooking device based on the built-in AI large model of the present invention. Best Mode for Carrying Out the Invention

[0041] The present invention is further described below in conjunction with Figures 1-8 and specific implementation methods and other embodiments. It should be noted that, under the premise of no conflict, the various technical features described below can be arbitrarily combined to form new embodiments.

[0042] In a preferred embodiment, referring to Figures 1-3, the present invention provides a cooking method based on a built-in AI large model, characterized in that it is applied to an intelligent cooking device, wherein the intelligent cooking device is provided with an information collection module 6, a storage module 4, an AI large model module 3, a processor 5, and a wireless communication module 7. The AI ​​large model module 3 is located inside the storage module 4, and the information collection module 6, the storage module 4, the AI ​​large model module 3, and the wireless communication module 7 are electrically connected to the processor 5.

[0043] In a preferred embodiment, the new cooking method includes the intelligent cooking device station, cooking work steps, cooking parameters or cooking operation instructions requiring user cooperation;

[0044] Referring to FIG. 2 , in a preferred embodiment, the present invention is based on an improvement of a cooking method with a built-in AI big model, wherein the construction of the cooking AI big model includes:

[0045] 201. Collect cooking data: Collect all information about cooking methods of ingredients presented through voice, video, text, pictures or 3D models;

[0046] 202. Preprocessing cooking data: Process all collected information about cooking methods to ensure its integrity and usability. This includes converting information in different formats into text and editing the text according to a specific format to facilitate subsequent training of the AI ​​model.

[0047] 203. Select AI big models that can be applied to cooking: Select domestic and foreign third-party AI big models and measure them using accuracy, response speed, and diversity indicators;

[0048] 204. Training the cooking AI model: After the cooking dataset is organized in step 2, fine-tuning the cooking dataset using a third-party AI model is performed. After training, a cooking AI model containing all relevant data on cooking methods is generated.

[0049] 205. Verify and test the cooking AI model: Perform task-specific performance evaluation on the cooking AI model generated in step 4. If the performance fails, repeat steps 1, 2, 3, and 4, retraining until the performance evaluation passes. Generate a cooking AI model and store it on the cloud platform.

[0050] 206. Deploy and maintain the cooking AI big model: deploy the newly generated AI big model to the AI ​​big model module 3 of the intelligent cooking device, and perform continuous maintenance and updates, and update data regularly to ensure the timeliness and accuracy of the data.

[0051] In this embodiment, in the process of building a cooking AI big model based on the cooking method with a built-in AI big model of the present invention, step 203 selects the AI ​​big model for cooking, specifically the Baichuan2-13B AI big model. The AI ​​big model parameters are as follows: hidden layer dimension: 5,120, number of layers: 40, number of attention heads: 40, vocabulary size: 64,000, total number of parameters: 13,264,901,120, training data (tokens): 1.4 trillion, position encoding: ALiBi, maximum length: 4,096;

[0052] In this embodiment, the training process of the cooking AI big model in 204 of the process of constructing the cooking AI big model of the cooking method based on the built-in AI big model of the present invention is as follows: first, download the model weights of baichuan13b from huggingface, then download the belle dataset train_0.5M_CN to the local and put it in the dataset folder under the project directory, and finally run the sft_lora.py script. Then, quantize Baichuan LLM using qlora's nf4 and double quantization methods, and finally, use lora to fine-tune the instructions.

[0053] In a preferred embodiment, when a user uses the APP of the mobile terminal 2 or the information collection module 6 of the intelligent cooking device to send the user's cooking requirement information in the form of voice, video, text, picture or 3D model, the cooking requirement information is transmitted to the processor 5. The processor 5 transmits the cooking requirement information to the AI ​​large model module 3 of the storage module 4. The AI ​​large model module 3 calls the built-in large model of the cooking AI to analyze the cooking requirement information and generate a new cooking method, which is then sent to the processor 5. The new cooking method has related intelligent cooking device stations, cooking steps, cooking parameters or cooking operation instructions that require user cooperation. The processor 5 first determines the intelligent cooking device station based on the new cooking method, then sets the relevant parameters for the intelligent cooking device station, and then controls the intelligent cooking device station to cook. During this cooking process, if some steps require user cooperation, the processor 5 will send information to the user 1. When cooking is completed, the processor 5 will control the intelligent cooking device station to perform a self-cleaning operation.

[0054] In a preferred embodiment, when the user needs to cook the dish "Hot and Sour Potato Shreds", the user uses the APP of the user mobile terminal 2 to send the cooking requirement information, which includes the following contents: the spiciness of the dish, such as mild, medium, and strong; the saltiness of the dish, such as light, medium, and salty; the acidity of the dish, such as mild, medium, and strong; the portion of the dish, such as large, medium, and small; if the user wants to add some of his own innovations, he can also add some side dishes, such as sausages, eggs, and meatballs. The user can freely match the existing ingredients on the APP interface and then submit this requirement on the APP. This cooking requirement information is received by the wireless communication module 7 and transmitted to the AI ​​large model module 3 via the processor 5. The AI ​​large model module 3 will analyze the cooking requirement information and generate a new cooking method, and then send the new cooking method to the user's APP on the user mobile terminal 2 via the wireless communication module 7 via the processor 5. When the user confirms to cook on the APP, the processor 5 will perform relevant operations according to the new cooking method. The new cooking method includes the following contents: cooking device station, cooking work steps, cooking parameters, including cooking work steps; each step requires the use of the corresponding intelligent cooking device station; cooking parameters include working time setting, power setting, temperature setting and water volume setting and other parameters; cooking operation instructions that require user cooperation.

[0055] The specific implementation is as follows:

[0056] The first step is to heat the pot: the processor 5 sets the working parameters of the intelligent cooking device station stir-fry 11 to 3000W power and heat to 200 degrees and maintain it according to the new cooking method. After the parameters are set, the intelligent cooking device station stir-fry 11 is started to work;

[0057] Step 2: Add oil with the user's cooperation: While the first step is being carried out, the processor 7 has synchronously sent a message to the APP of the user's mobile terminal 2 according to the new cooking method, reminding the user to remove the oil package and add a certain amount of oil to the pot after hearing the prompt tone. At the same time, the processor 5 uses the operation detection feedback system 14 to detect whether the user has performed the required operation;

[0058] Step 3: The user cooperates in adding the food: After the processor 5 detects that the pot temperature has reached 200 degrees, the processor 5 sends a message to the APP of the user's mobile terminal 2 according to the new cooking method, reminding the user to remove the packaging of the ingredients. After hearing the prompt tone, the user pours the predetermined amount of food into the pot. At the same time, the processor 5 detects whether the user has performed the required operation through the operation detection feedback system 14;

[0059] Step 4: Cooking: The processor 5 adjusts the working parameters of the intelligent cooking device station 11 to 3000W power, heats to 200 degrees and maintains it, starts flipping and cooking for 5 minutes, and simultaneously sends a message to the APP of the user's mobile terminal 2, reminding the user to remove the vinegar material package and pour a certain amount of vinegar into the pot after hearing the prompt tone. At the same time, the processor 5 detects whether the user has performed the required operation through the operation detection feedback system 14;

[0060] Step 4: Add seasoning: The processor 5 adjusts the working parameters of the stir-frying station 11 of the intelligent cooking device to 2000W power according to the new cooking method, heats to 200 degrees and maintains it, starts flipping and cooking for 1 minute, and simultaneously sends a message to the APP of the user's mobile terminal 2, reminding the user to remove the seasoning package and pour the predetermined amount of seasoning into the pot after hearing the prompt tone. At the same time, the processor 5 detects whether the user has performed the required operation through the operation detection feedback system 14;

[0061] Step 5: Serving the dish: The processor 5 sends a message to the APP of the user's mobile terminal 2 according to the new cooking method, reminding the user to put the dish plate on the designated position of the wok 11 of the intelligent cooking device. After the operation detection feedback system 14 detects that the user has operated as required, the processor controls the wok 11 of the intelligent cooking device to flip over and pour the dish onto the dish plate, and reminds the user to take the dish.

[0062] Step 6: Cleaning: The processor 5 controls the intelligent cooking device station frying 11 to perform a self-cleaning operation according to the new cooking method, and then enters a standby state.

[0063] The above is just an example of the operation process of one dish. When it is implemented, it can be more intelligent and personalized according to user needs. The built-in AI large model can customize a variety of cooking methods according to user needs for user selection.

[0064] In this embodiment, the cooking method based on the built-in AI big model of the present invention specifically includes the following steps:

[0065] 301. Assemble query statements;

[0066] 302. The built-in AI big model performs reasoning operations: The query statement is passed to the AI ​​big model module 3, and the AI ​​big model module 3 performs reasoning operations, which includes the following four processes:

[0067] 302-1. Understanding Input: Distributed semantic parsing first receives a text sequence and converts it into a word vector. This process is based on the distributed semantics assumption that the meaning of a word is determined by its use in context.

[0068] 302-2. Parameter association: context-focus chaining, inputting these word vectors into the Transformer Encoder to generate context representation;

[0069] 302-3. Generate answers: Generative probabilistic modeling. The model initializes the decoder part of the Transformer and inputs the encoder output and the current output sequence into the decoder. The decoder generates a probability distribution for the next word and selects the word with the highest probability or other set probability distribution as the output. This word will be added to the output sequence.

[0070] 302-4. Choose the most appropriate answer: Dynamic word string evolution, repeating the above steps, adding new words to the output sequence each time, until a complete output sequence is generated;

[0071] 303. Return results of the built-in AI big model: After the cooking built-in AI big model completes the inference operation, it returns information related to the new cooking method and the new cooking method operation instructions, including content provided to the user in text, image, audio, video or 3D model format;

[0072] The present invention provides an intelligent cooking device, characterized in that the intelligent cooking device is provided with an information collection module 6, a storage module 4, an AI large model module 3, a processor 5 and a wireless communication module 7. The AI ​​large model module 3 is located inside the storage module 4. The information collection module 6, the storage module 4, the AI ​​large model module 3, and the wireless communication module 7 are electrically connected to the processor 5. The information collection module 6 is used to collect all information about cooking ingredients presented by the user in the format of voice, video, text, picture or 3D model. The processor 5 is connected to the cloud platform 16 and the user mobile terminal 2 through the wireless communication module 7. The processor 5 is used to execute the cooking method based on the built-in AI large model deployed in the AI ​​large model module 3 in claim 1. After the new cooking method is inferred and calculated by the cooking built-in AI large model, it is sent to the intelligent cooking device through the storage module 4 and prompts the user to cook. The information collection module 6 has information input ports for voice, video, text, picture and 3D model, and the user can use these information input ports to Cooking demand information is input through some information input ports, or the information collection module 6 is connected to the user mobile terminal 2 through the wireless communication module 7 to receive cooking demand information. The cooking AI big model is deployed in the AI ​​big model module 3 of the intelligent cooking device. The intelligent cooking device is provided with the processor 5. The processor 5 can set working parameters for each intelligent cooking device station of the intelligent cooking device and control each intelligent cooking device station to execute relevant instructions, such as standby, start, stop, cleaning and other working instructions. When the user's cooking demand information is input through the information collection module 6 and then transmitted to the AI ​​big model module 3 through the processor 5, the AI ​​big model module 3 analyzes and generates a cooking method and transmits it to the processor 5. This cooking method includes specific cooking data information. The processor 5 will set parameters for the relevant intelligent cooking device stations based on these cooking data information and control them to execute relevant instructions. During the period of use of the intelligent cooking device, the data of the built-in cooking AI big model built into the intelligent cooking device will be updated and maintained regularly through the cloud platform 16 or on-site.

[0073] In this embodiment, the intelligent cooking device of the present invention is provided with intelligent cooking device stations for frying 8, baking 9, boiling 10, air frying 12, steaming 13 and stir-frying 11. The intelligent cooking device stations for frying 8, baking 9, boiling 10, air frying 12, steaming 13 and stir-frying 11 are provided with a common operation detection feedback system 14. The operation detection feedback system 14 is used to detect whether the cooking operation performed by the user in the intelligent cooking device meets the requirements of the cooking method based on the built-in AI large model. The operation detection feedback system 14 includes detection devices such as cameras, infrared detection, radar detection, magnetic detection, and weight detection, which are used to detect whether the cooking operation performed by the user in the intelligent cooking device meets the requirements of the cooking method based on the built-in AI large model. For example, In the frying intelligent cooking device station, the intelligent cooking device requires the user to place the ingredients to be fried on the station, but the user does not place them or places them in the wrong position. At this time, the operation detection and feedback system 14 detects it through the camera and feeds this information back to the processor 5 of the intelligent cooking device. The processor 5 controls the intelligent cooking device station to pause the next operation and sends information to the user for correction. When the user corrects it, the operation detection and feedback system 14 detects it through the camera and feeds this information back to the processor 5. The processor 5 controls the intelligent cooking device station to perform the next operation. The above is just an example. In addition, infrared detection, radar detection, magnetic detection, weight detection, or a combination thereof can also be used for detection.

[0074] In another embodiment 1, referring to FIG. 4 , the frying 8 , baking 9 , boiling 10 , air frying 12 , and steaming 13 stations of the intelligent cooking device of the present invention are provided with a common operation detection and feedback system 14 , and the rest are the same as those of the preferred embodiment of the present invention;

[0075] In another embodiment 2, referring to FIG. 5 , the baking 9, boiling 10, air frying 12 and steaming 13 intelligent cooking device stations of the intelligent cooking device of the present invention are provided with a common operation detection feedback system 14 , and the rest are the same as the preferred embodiment of the present invention;

[0076] In another third embodiment, referring to FIG. 6 , the cooking 10 and stir-fry 11 intelligent cooking device stations of the intelligent cooking device of the present invention are provided with a common operation detection feedback system 14 , and the rest are the same as the preferred embodiment of the present invention;

[0077] In another fourth embodiment, referring to FIG. 7 , the intelligent cooking device station 11 of the intelligent cooking device of the present invention is provided with an operation detection feedback system 14 , and the rest is the same as the preferred embodiment of the present invention;

[0078] In another embodiment 5, referring to FIG8 , the intelligent cooking device of the present invention is not provided with the user mobile terminal 2 , and the rest is the same as the preferred embodiment of the present invention;

[0079] In other embodiments, the smart cooking device is one or a combination of a cooker, an air fryer, an induction cooker, a microwave oven, an oven, a steam oven, an electric rice cooker, an electric pressure cooker, an electric stew pot, an integrated stove or a frying pan, and one or more smart cooking devices include one or more cooking functions of stir-frying, air frying, baking, frying, braising, stewing, steaming, boiling or baking to cook ingredients.

[0080] In other embodiments, the smart cooking device prompts the user to place different ingredients on different workstations on the smart cooking device for cooking. For example, the cooking position of stir-frying is different from that of steaming, air frying, and frying in the smart cooking device, and the user needs to be prompted to cook at the corresponding workstation of the smart cooking device; the cooking work steps include the cooking steps that the smart cooking device prompts the user to perform when cooking different ingredients; the cooking parameters include the cooking time, cooking temperature, and cooking power of the ingredients; the cooking operation instructions that require user cooperation include confirming the cooking parameters on the mobile terminal or confirming that the operation of the smart cooking device workstation and cooking work steps has been completed.

[0081] In other embodiments, the intelligent cooking device is an integrated body of multiple cooking functions such as stir-frying, air frying, baking, frying, braising, stewing, steaming, boiling or baking. For example, the intelligent cooking device is a cooking utensil that integrates functions such as air frying, baking, frying and steaming, or a cooking utensil that integrates functions such as stir-frying, air frying, braising, stewing and boiling, etc.

[0082] In other embodiments, the intelligent cooking device is provided with multiple cooking functions including one or more cooking functions of stir-frying, sautéing, deep-frying, cooking, frying, sticking, roasting, braising, stewing, steaming, blanching, boiling, stewing, sautéing, mixing, marinating, roasting, braising, freezing, candied food, honey sauce, smoking, rolling, sliding or baking. For example, the intelligent cooking device is a cooking utensil that integrates functions such as honey sauce, roasting, frying and steaming, or a cooking utensil that integrates functions such as stir-frying, air frying, braising, stewing and boiling, etc.

[0083] In this embodiment, the intelligent cooking device of the present invention is provided with a human-computer interaction system 15, which is used for information interaction between the intelligent cooking device and the user 1, including the user operating the user mobile terminal 2 to confirm cooking parameters and start cooking operation instructions. The human-computer interaction system 15 includes a touch screen, a sound input and output unit, an image input unit and a camera device. The human-computer interaction system 15 is electrically connected to the processor 5 for information interaction between the intelligent cooking device and the user 1. The user 1 can input cooking requirement information in the form of voice, video, text, picture or 3D model through the human-computer interaction system 15, and the human-computer interaction system 15 sends this cooking requirement information to the processor 5, and the processor 5 transmits this cooking requirement information to AI large model module 3, AI large model module 3 will analyze the cooking demand information and generate a new cooking method, and then send the new cooking method to the processor 5. The processor 5 will send this information to the user 1 for confirmation through the human-computer interaction system 15 in the form of one or a combination of images, voice and text. After the user 1 confirms through gestures or voice, the processor 5 will first determine the intelligent cooking device station according to the new cooking method, and then set the relevant parameters for the intelligent cooking device station, and then control the intelligent cooking device station to cook. During this cooking process, when some steps require the cooperation of the user 1, the processor 5 will send the information to the user 1 through the human-computer interaction system 15 in the form of one or a combination of images, voice and text.

[0084] The cooking method based on the built-in AI large model and the intelligent cooking device thereof are used for cooking in the following specific implementation steps:

[0085] The specific implementation of process 202 for preprocessing cooking data in the cooking method based on a built-in AI model is as follows: all collected information is processed to ensure its integrity and usability. This includes converting information in different formats into text and editing the text information according to a specific format to facilitate subsequent training of the AI ​​model.

[0086] The data format is as follows:

[0087] instruction: Task instruction, cannot be empty.

[0088] input: Task input, can be empty. If not empty, when processing training data within the project, instruction and input will be concatenated together as the input of the task.

[0089] output: task output, cannot be empty

[0090] The data example is as follows:

[0091] {

[0092] "instruction": "Sweet and Sour Spare Ribs Pre-prepared Dish Cooking Method",

[0093] "input": "",

[0094] "output": "1. Pour the pre-cooked spareribs into the pan and fry them over medium heat. Heat the oil to 60% and fry slowly. 2. Fry for 3 minutes until the spareribs are slightly browned and drain the oil. 3. Leave a small amount of oil in the pan, add the sweet and sour sauce, and bring to a boil over high heat. 4. Add the fried spareribs to the sweet and sour sauce and stir-fry evenly.

[0095] }

[0096] The specific implementation process of step 303 of the AI ​​big model-based cooking method of the present invention, in which the built-in AI big model is called, is as follows: After the built-in AI big model completes the inference operation, it returns cooking method related information and cooking method operation instructions, including content in multiple formats such as text, pictures, audio, video, and 3D models. The returned content is in the Jason format and is as follows:

[0097] {

[0098] “recipeName”:recipeName / / cooking method name,

[0099] "recipeText":text / / cooking method information,

[0100] "recipePic":pic_url / / Url of the picture of cooking method information,

[0101] "recipeAudio": audio_url / / audio URL of cooking method information,

[0102] "recipeVideo": video_url / / Video URL of cooking method information,

[0103] "recipe3D":3D_url / / 3D model URL of cooking method information,

[0104] “recipeCommand”:recipeCommand / / cooking method operation instructions

[0105] }

[0106] The cooking method based on the AI ​​large model and the intelligent cooking device thereof for cooking sweet and sour spare ribs are specifically operated as follows:

[0107] 1. The request for cooking sweet and sour spare ribs is sent to the cooking AI model;

[0108] 2. Call the cooking AI big model through the Prompt query statement, and the cooking AI big model performs inference operations.

[0109] 3. After the inference operation, the cooking AI model returns the following results:

[0110] {

[0111] "recipeName":"Sweet and Sour Spare Ribs Cooking Method" / / Cooking method name,

[0112] "recipeText":"1. Pour the pre-cooked spareribs into the pan and fry them when the oil temperature reaches 60%. Keep frying over medium heat. 2. Fry for 3 minutes. When the spareribs are slightly browned, remove them and drain the oil. 3. Leave a little oil in the pan, add the sweet and sour sauce, and bring to a boil over high heat. 4. Add the fried spareribs into the sweet and sour sauce and stir-fry evenly." / / Cooking method information,

[0113] "recipePic": "None" / / None,

[0114] "recipeAudio": "https: / / www.ixigua.com / 7289152022341747234?logTag=43fb2a078d33289b226b" / / Audio URL for cooking method information,

[0115] "recipeVideo": "https: / / www.ixigua.com / 7289152022341747234?logTag=43fb2a078d33289b226b" / / Video URL of cooking method information,

[0116] "recipe3D":3D_url / / 3D model URL of cooking method information,

[0117] "recipeCommand": "Fry for 30 seconds, adjust the temperature to 100 degrees, stir-fry for 20 seconds, steam for 2 minutes, and then finish the operation" / / Cooking method operation instructions

[0118] }

[0119] The present invention provides a cooking method based on a built-in AI big model and an intelligent cooking device thereof, which have the following beneficial effects: the present invention provides a cooking method based on a built-in AI big model and an intelligent cooking device thereof, which collects all information on cooking methods of ingredients from people of all countries and regions, all races, and all ages around the world presented through voice, video, text, pictures or 3D models, even including the cooking methods of various foods that a specific individual has been accustomed to or fond of in the past, and uses a third-party AI big model for fine-tuning training. After the training is completed, a cooking AI big model with all relevant data of massive cooking methods is generated. The user inputs cooking requirement information through voice, video, text, pictures or 3D model format, and the intelligent cooking device executes the new cooking method generated by the built-in cooking AI big model by the user to cook food that meets the different needs of different users, and even cooks food that meets the different needs of the same user. In this way, not only the different needs of different users can be met, but also the different needs of the same user at different times and in different environments can be met. Moreover, the AI ​​big model is built into the intelligent cooking device, and the use of the AI ​​big model to cook food by the intelligent cooking device is not affected by the Internet.

[0120] The present invention has been described in detail above. The above description is only a preferred embodiment of the present invention and should not limit the scope of implementation of the present invention. That is, all equivalent changes and modifications made within the scope of this application should still fall within the scope of the present invention.

Claims

1. A cooking method based on a built-in AI large model, characterized in that: Applied to an intelligent cooking device, the intelligent cooking device is provided with an information collection module, a storage module, an AI large model module, a processor and a wireless communication module, the AI ​​large model module is located in the storage module, the information collection module, the storage module, the AI ​​large model module, the wireless communication module are electrically connected to the processor, the information collection module is used to receive cooking demand information input by a user in voice, video, text, picture or 3D model format, and push the cooking demand information to the AI ​​large model module located in the storage module, the AI ​​large model module analyzes the cooking demand information and generates a new cooking method, and sends the new cooking method to the processor through the storage module, the new cooking method includes the workstation of the intelligent cooking device, cooking work steps, cooking parameters or cooking methods that require user cooperation Operation instructions, the AI ​​big model module generates a new cooking method, including: building a cooking AI big model and calling the built-in cooking AI big model, wherein building the cooking AI big model includes collecting cooking data, preprocessing cooking data, selecting an AI big model that can be applied to cooking, training the cooking AI big model, verifying and testing the cooking AI big model, and deploying and maintaining the cooking AI big model in the AI ​​big model module, the cooking requirement information input by the user in voice, video, text, picture or 3D model format is called by the AI ​​big model module to generate the new cooking method through the built-in cooking AI big model, the calling of the built-in cooking AI big model includes assembling query statements, the built-in AI big model performing reasoning operations and the built-in AI big model returning results, and the new cooking method is sent to the processor through the storage module after the reasoning operation of the built-in cooking AI big model is completed.

2. The cooking method based on the built-in AI large model according to claim 1, characterized in that: The construction of the cooking AI big model includes: Step 1: Collect cooking data: Collect all information about cooking methods of ingredients presented through voice, video, text, pictures or 3D models; Step 2: Preprocess cooking data: process all collected information about cooking methods to ensure the integrity and availability of the information, including converting information in different formats into text and editing the text information in a certain format to facilitate subsequent training of the AI ​​model. Step 3: Select AI big models that can be applied to cooking: Select domestic and foreign third-party AI big models, and measure them by accuracy, response speed, and diversity indicators; Step 4: Train the cooking AI big model: After the cooking data set is sorted out in step 2, the cooking data set is fine-tuned using a third-party AI big model. After the training is completed, a cooking AI big model with all relevant data of the cooking method is generated; Step 5: Verify and test the cooking AI big model: Perform a specific task effect test and evaluation on the cooking AI big model generated in step 4. If the evaluation effect fails, continue to repeat the steps of step 1, step 2, step 3, and step 4, retrain until the effect evaluation passes, generate the cooking AI big model and store it on the cloud platform; Step 6. Deploy and maintain the cooking AI big model: deploy the newly generated AI big model to the AI ​​big model module of the intelligent cooking device, and perform continuous maintenance and updates, and update data regularly to ensure the timeliness and accuracy of the data.

3. The cooking method based on the built-in AI large model according to claim 2, characterized in that: The third step selects an AI big model that can be applied to cooking. The selected AI big model is Baichuan2-13B. The parameters of the AI ​​big model are as follows: hidden layer dimension: 5,120, number of layers: 40, number of attention heads: 40, vocabulary size: 64,000, total number of parameters: 13,264,901,120, training data (tokens): 1.4 trillion, Positional encoding: ALiBi, Maximum length: 4,096.

4. The cooking method based on the built-in AI large model according to claim 3 is characterized in that: The training process of the fourth step of training the cooking AI large model is as follows: first, download the model weights of baichuan13b from huggingface, then download the belle dataset train_0.5M_CN to the local and put it in the dataset folder under the project directory, and finally run the sft_lora.py script. Next, quantize Baichuan LLM using qlora's nf4 and double quantization methods, and finally, use lora to fine-tune the instructions.

5. The cooking method based on the built-in AI large model according to claim 1, characterized in that: The calling of the built-in AI big model for cooking specifically includes the following steps: The first step is to assemble the query statement; Step 2: Use the built-in AI big model for reasoning operations: Pass the query statement to the AI ​​big model, which then performs reasoning operations. This includes the following steps: 1) Understanding input: Distributed semantic parsing first receives a text sequence and converts it into a word vector. This process is based on the distributed semantics assumption that the meaning of a word is determined by its use in context. 2) Parameter association: context focus linkage, input these word vectors into the Transformer Encoder to generate context representation; 3) Generate answers: Generative probability modeling. The model initializes the decoder part of the Transformer and inputs the encoder output and the current output sequence into the decoder. The decoder generates the probability distribution of the next word and selects the word with the largest probability or other set probability distribution as the output. This word will be added to the output sequence. 4) Select the most appropriate answer: Dynamic word string evolution, repeat the above steps, adding new words to the output sequence each time until a complete output sequence is generated; Step 3. The built-in AI big model returns the result: After the built-in AI big model inference operation is completed, it returns the information related to the new cooking method and the operation instructions of the new cooking method, including providing the user with content in text, picture, audio, video or 3D model format.

6. An intelligent cooking device, characterized in that: An information collection module, a storage module, an AI big model module, a processor and a wireless communication module are provided. The AI ​​big model module is located inside the storage module. The information collection module, the storage module, the AI ​​big model module and the wireless communication module are electrically connected to the processor. The information collection module is used to collect all information about cooking ingredients presented by the user in voice, video, text, picture or 3D model format. The wireless communication module can be used to connect the processor with a cloud platform and a user mobile terminal. The processor is used to execute the cooking method based on the built-in AI big model deployed in the AI ​​big model module in claim 1. After the new cooking method is inferred by the cooking built-in AI big model, it is sent to the intelligent cooking device through the storage module and prompts the user to cook.

7. The intelligent cooking device according to claim 6, characterized in that: There are intelligent cooking device stations for frying, baking, boiling, air frying, braising, stewing, baking, steaming or stir-frying. The intelligent cooking device receives the new cooking method and prompts the user to cook and perform cooking work steps at the corresponding intelligent cooking device station.

8. The intelligent cooking device according to claim 7, characterized in that: The frying, baking, boiling, air frying, braising, stewing, baking, steaming or stir-frying intelligent cooking device workstations are provided with corresponding operation detection feedback systems, and the operation detection feedback systems are used to detect whether the cooking operations performed by the user on the intelligent cooking device meet the requirements of the new cooking method.

9. The intelligent cooking device according to claim 6, characterized in that: A human-computer interaction system is provided, and the human-computer interaction system is used for information interaction between the intelligent cooking device and the user, including the user operating the user mobile terminal to confirm cooking parameters and start cooking operation instructions.

10. The intelligent cooking device according to claim 6, characterized in that: The new cooking method is sent to the user's mobile terminal through the wireless communication module. The user's mobile terminal receives the new cooking method and prompts the user to cook and perform cooking steps at the corresponding intelligent cooking device workstation, and the user operates the mobile terminal to confirm cooking parameters and start cooking operation instructions.

Images