CN117764777A - Cooking method based on built-in AI large model and intelligent cooking device thereof - Google Patents
Cooking method based on built-in AI large model and intelligent cooking device thereof Download PDFInfo
- Publication number
- CN117764777A CN117764777A CN202311620532.0A CN202311620532A CN117764777A CN 117764777 A CN117764777 A CN 117764777A CN 202311620532 A CN202311620532 A CN 202311620532A CN 117764777 A CN117764777 A CN 117764777A
- Authority
- CN
- China
- Prior art keywords
- cooking
- model
- large model
- module
- intelligent
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000010411 cooking Methods 0.000 title claims abstract description 466
- 238000000034 method Methods 0.000 title claims abstract description 165
- 238000003860 storage Methods 0.000 claims abstract description 35
- 238000004891 communication Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims description 30
- 238000001514 detection method Methods 0.000 claims description 29
- 235000013305 food Nutrition 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 20
- 230000003993 interaction Effects 0.000 claims description 16
- 238000010025 steaming Methods 0.000 claims description 12
- 230000000694 effects Effects 0.000 claims description 10
- 238000009835 boiling Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 9
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 claims description 6
- 238000009826 distribution Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000013139 quantization Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000006872 improvement Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 239000003921 oil Substances 0.000 description 7
- 235000015277 pork Nutrition 0.000 description 6
- 235000021419 vinegar Nutrition 0.000 description 6
- 239000000052 vinegar Substances 0.000 description 6
- 239000000463 material Substances 0.000 description 5
- 235000009508 confectionery Nutrition 0.000 description 4
- 235000021186 dishes Nutrition 0.000 description 4
- 235000011389 fruit/vegetable juice Nutrition 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000013523 data management Methods 0.000 description 3
- 235000011194 food seasoning agent Nutrition 0.000 description 3
- 235000013311 vegetables Nutrition 0.000 description 3
- 235000015001 Cucumis melo var inodorus Nutrition 0.000 description 2
- 240000002495 Cucumis melo var. inodorus Species 0.000 description 2
- 239000004278 EU approved seasoning Substances 0.000 description 2
- 239000002199 base oil Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000010438 heat treatment Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 235000002595 Solanum tuberosum Nutrition 0.000 description 1
- 244000061456 Solanum tuberosum Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 235000013601 eggs Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 235000013372 meat Nutrition 0.000 description 1
- 238000005554 pickling Methods 0.000 description 1
- 235000012015 potatoes Nutrition 0.000 description 1
- 230000033764 rhythmic process Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 235000013580 sausages Nutrition 0.000 description 1
- 235000012046 side dish Nutrition 0.000 description 1
- 238000003756 stirring Methods 0.000 description 1
- 230000007306 turnover Effects 0.000 description 1
- 239000005418 vegetable material Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 238000005491 wire drawing Methods 0.000 description 1
Landscapes
- General Preparation And Processing Of Foods (AREA)
Abstract
The invention discloses a cooking method based on an embedded AI large model and an intelligent cooking device thereof, which are characterized in that 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, wherein the AI large model module is positioned in the storage module, the information collection module, the storage module, the AI large model module and the wireless communication module are electrically connected with 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 positioned in the storage module, and the AI large model module analyzes the cooking requirement information and generates a new cooking method.
Description
Technical Field
The invention relates to a cooking device and a cooking method, in particular to a cooking method based on a built-in AI large model and an intelligent cooking device thereof.
Background
Along with development of science and technology and acceleration of life rhythm, intelligent cooking devices are generated, and automatic cooking can be achieved by the intelligent cooking devices.
Regarding the combination of the AI large model and the intelligent cooking device, the latest prior art comprises an intelligent cooking machine cloud platform data management system and method based on the AI large model, with publication number of CN 116843510A and publication date of 2023, 10 and 03, and application number of 202310784823.7;
the application of the invention discloses an AI large model-based intelligent cooking machine cloud platform data management system and method, and relates to the technical field of data management systems, wherein the management method comprises the following steps: the intelligent stir-frying machine comprises a cloud platform, a processing end, an acquisition end, an AI large model, a management system, a cloud platform, a storage end, an intelligent stir-frying machine, a storage end and a warning system.
The effect of being applied to intelligent dish machine with AI large model among the prior art is that the processing end only analyzes the parameter data of storage through AI large model, and whether analysis this time dish in-process intelligent dish machine has the abnormality, when analysis intelligent dish machine has the abnormality, judge that the dish article taste exists the deviation according to analysis result, need change the quality and the taste of dish article are ensured to other intelligent dish machine stir-fry, and it is also not necessarily to change other intelligent dish machine stir-fry in fact, because the other intelligent dish machines of change also probably have the abnormality, the user needs to guarantee the quality and the taste of dish article through constantly changing intelligent dish machine like this.
The invention relates to a cooking method based on an internal AI large model and an intelligent cooking device thereof, which are characterized in that a massive cooking dataset is subjected to Fine tuning Fine-tuning training by using a third-party AI large model, a cooking AI large model with all relevant data of the cooking method is generated after training is completed, a user inputs cooking requirement information in a voice, video, text, picture or 3D model format, a new cooking method is generated by calling the cooking AI large model, and the intelligent cooking device executes the new cooking method generated by calling the cooking AI large model by the user to cook food meeting the requirements of the user.
Disclosure of Invention
In order to solve the above-mentioned prior art problems, the present invention provides a cooking method based on a built-in AI big model, which is characterized in that the cooking method is applied to an intelligent cooking device, the intelligent cooking device is provided with an information collecting module, a storage module, an AI big model module, a processor and a wireless communication module, the AI big model module is located in the storage module, the information collecting module, the storage module, the AI big model module and the wireless communication module are electrically connected with the processor, the information collecting module is used for receiving cooking requirement information input by a user through voice, video, text, pictures or a 3D model format, pushing the 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, the new cooking method includes cooking operation instructions of the intelligent cooking device station, cooking operation steps, cooking parameters or cooking operation instructions that need to be matched by the user, and the new cooking method includes: the method comprises the steps of constructing a cooking AI big model and calling a cooking built-in AI big model, wherein the construction of the cooking AI big model comprises the steps of collecting cooking data, preprocessing the cooking data, selecting an AI big model applicable to cooking, training the cooking AI big model, verifying a test cooking AI big model and deploying and maintaining the cooking AI big model in the AI big model, a user calls the cooking built-in AI big model through the AI big model to generate the new cooking method through cooking requirement information input through voice, video, text, pictures or 3D model formats, the calling of the cooking built-in AI big model comprises the steps of assembling query sentences, carrying out reasoning operation on the built-in AI big model and returning results of the built-in AI big model, and the new cooking method is sent to the processor through the storage module after the reasoning operation of the cooking built-in AI big model is completed.
As an improvement of the cooking method based on the built-in AI large model of the invention, the construction of the cooking AI large model comprises:
first, collecting cooking data: collecting all information about food cooking methods presented by voice, video, text, pictures or 3D models;
step two, preprocessing cooking data: processing all collected information related to food cooking methods to ensure the integrity and usability of the information, including uniformly converting the information in different formats into texts, and editing the text information according to a certain format so as to facilitate the subsequent training of the AI large model;
third, selecting an AI large model applicable to cooking: selecting a large model of a third party AI at home and abroad, and measuring the large model by using accuracy, response speed and diversity indexes;
fourth step, training a cooking AI large model: finishing a cooking data set through the second step, and then carrying out Fine tuning Fine-tuning training on the cooking data set by using a third-party AI large model, and generating a cooking AI large model with all relevant data of a cooking method after training is finished;
fifth step, verifying and testing the large cooking AI model: performing effect detection evaluation of the specific task on the cooking AI large model generated in the fourth step, if the evaluation effect does not pass, continuously repeating the steps of the first step, the second step, the third step and the fourth step, retraining until the effect evaluation passes, generating the cooking AI large model and storing the cooking AI large model in a cloud platform;
Sixth, deploying and maintaining a large cooking AI model: and deploying the newly generated AI large model to the built-in AI large model module of the intelligent cooking device, continuously maintaining and updating, and periodically updating data to ensure timeliness and accuracy of the data.
As an improvement of the cooking method based on the built-in AI large model of the present invention, the third step of selecting the AI large model applicable to cooking is selected as AI large model of Baichuan2-13B, and the AI large model parameters are as follows: hidden layer dimension: 5,120, the number of layers: 40, attention header number: 40, vocabulary size: 64,000, total parameters: 13,264,901,120 training data (keys): 1.4 trillion, position coding: ALiBi, maximum length: 4,096;
as an improvement of the cooking method based on the built-in AI large model, the training process of training the cooking AI large model in the fourth step is as follows: firstly, downloading model weights of baichuan13b from huggingface, then downloading a belle data set train_0.5M_CN to the local and placing under a dataset folder under an item directory, finally running a sft_lora.py script, then quantizing the Baichuan LLM by adopting nf4 of qlora and a double quantization mode, and finally carrying out instruction fine adjustment by adopting lora.
As an improvement of the cooking method based on the built-in AI large model, the calling of the built-in AI large model specifically comprises the following steps:
firstly, assembling a query statement;
secondly, carrying out reasoning operation by a built-in AI large model: the query sentence is transmitted to the AI big model, and the AI big model carries out reasoning operation, wherein the reasoning operation comprises the following processes:
1) Understanding input: the distributed semantic parsing firstly receives a text sequence and converts the text sequence into word vectors, and the process is based on distributed semantic assumptions, namely word senses are determined by the use of the word senses in the context;
2) Parameter association: context focus linkage, inputting these word vectors into the transform's Encoder to generate a context representation;
3) Generating an answer: generating probability modeling, initializing a Decoder part of a transducer by the model, inputting the output of the Encoder and the current output sequence into the Decoder together, generating probability distribution of the next word by the Decoder, and selecting the word with the maximum probability or other set probability distribution as output, wherein the word is added into the output sequence;
4) Selecting the most appropriate answer: the dynamic word string evolves, the steps are repeated, new words are added to the output sequence each time, and a complete output sequence is generated;
Thirdly, the built-in AI large model returns the result: after the built-in AI large model reasoning operation is completed, returning information related to the new cooking method and an operation instruction of the new cooking method, wherein the information comprises contents in text, pictures, audio, video or 3D model formats provided for users;
the invention provides an intelligent cooking device which is characterized by comprising an information collecting module, a storage module, an AI large model module, a processor and a wireless communication module, wherein the AI large model module is positioned in the storage module, the information collecting module, the storage module, the AI large model module and the wireless communication module are electrically connected with the processor, the information collecting module is used for collecting all information about food cooking presented by a user through voice, video, text, pictures or a 3D model format, the wireless communication module can be used for connecting the processor with a cloud platform and a user mobile terminal, the processor is used for executing a cooking method based on an embedded AI large model which is deployed in the AI large model module in claim 1, and the novel cooking method is sent to the intelligent cooking device through the storage module after the reasoning operation of the embedded AI large model for cooking is completed and prompts the user to cook.
As an improvement of the intelligent cooking device, an intelligent cooking device station for frying, roasting, boiling, air frying, stewing, baking, steaming or frying is arranged, and the intelligent cooking device receives the new cooking method and prompts a user to cook and carry out cooking working steps at the corresponding intelligent cooking device station.
As an improvement of the intelligent cooking device, the station of the intelligent cooking device for frying, roasting, boiling, air frying, stewing, baking, steaming or frying is provided with a corresponding operation detection feedback system, and the operation detection feedback system is used for detecting whether the cooking operation performed by a user in the intelligent cooking device meets the requirement of the new cooking method.
As an improvement of the intelligent cooking device, a man-machine interaction system is arranged, and the man-machine interaction system is used for information interaction between the intelligent cooking device and a user, and comprises the steps that the user operates a mobile terminal of the user to confirm cooking parameters and starts cooking operation instructions.
As an improvement of the intelligent cooking device, the new cooking method is sent to the user mobile terminal through the wireless communication module, the user mobile terminal receives the new cooking method and prompts a user to cook and carry out cooking work steps at the corresponding intelligent cooking device station, and the user operates the mobile terminal to confirm cooking parameters and start cooking operation instructions.
The invention relates to a cooking method based on a built-in AI large model and an intelligent cooking device thereof, which have the beneficial effects that: according to the cooking method based on the built-in AI large model and the intelligent cooking device thereof, all information about food cooking methods from people in all countries, groups and age groups which are presented through voice, video, text, pictures or 3D models is collected, even the cooking method comprises a specific individual custom or favorite cooking method of various foods, the third-party AI large model is used for Fine tuning of the Fine-tuning training, the cooking AI large model with all relevant data of the massive cooking methods is generated after the training is completed, a user inputs cooking requirement information through voice, video, text, pictures or 3D model formats, the intelligent cooking device executes a new cooking method generated by the built-in AI large model of the user, and the intelligent cooking device cooks foods meeting different requirements of different users, even cooks foods meeting different requirements of the same user, not only can meet different requirements of different users but also can meet different requirements of different environments of the same user in different time, but also is built in the intelligent cooking device, and the intelligent cooking device is free from the influence of the use of the AI large model.
Drawings
Fig. 1 is a flowchart of a cooking built-in AI large model construction method and an intelligent cooking apparatus according to a preferred embodiment of the present invention.
Fig. 2 is a flowchart of a cooking built-in AI large model according to a preferred embodiment of the cooking method based on the built-in AI large model and the intelligent cooking apparatus thereof of the present invention.
Fig. 3 is a flowchart showing a cooking method based on a built-in AI large model and an intelligent cooking apparatus according to a preferred embodiment of the present invention.
Fig. 4 is a flowchart of one of the other embodiments of the cooking method based on the built-in AI large model and the intelligent cooking apparatus according to the present invention.
Fig. 5 is a flowchart of a cooking method based on a built-in AI large model and a smart cooking device according to a second embodiment of the present invention.
Fig. 6 is a flowchart illustrating a cooking method based on a built-in AI large model and a third embodiment of an intelligent cooking apparatus according to the present invention.
Fig. 7 is a flowchart illustrating a cooking method based on a built-in AI large model and an intelligent cooking apparatus according to a fourth embodiment of the present invention.
Fig. 8 is a flowchart of a cooking method based on a built-in AI large model and other embodiments of the intelligent cooking apparatus according to the present invention.
Detailed Description
The present invention will be further described with reference to fig. 1 to 8, and specific embodiments and other examples, where it should be noted that, on the premise of no conflict, the following technical features may be arbitrarily combined to form new embodiments.
In a preferred embodiment, referring to fig. 1-3, the present invention provides a cooking method based on a built-in AI large model, which is characterized in that the cooking method is applied to an intelligent cooking device, the intelligent cooking device is provided with an information collecting 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 in the storage module 4, and the information collecting module 6, the storage module 4, the AI large model module 3 and the wireless communication module 7 are electrically connected with the processor 5;
in a preferred embodiment, the new cooking method comprises the intelligent cooking device station, a cooking working step, cooking parameters or cooking operation instructions which need to be matched by a user;
referring to fig. 2, in a preferred embodiment, the present invention is an improvement of a cooking method based on a built-in AI large model, said constructing the cooking AI large model comprising:
201. collecting cooking data: collecting all information about food cooking methods presented by voice, video, text, pictures or 3D models;
202. Preprocessing cooking data: processing all collected information related to food cooking methods to ensure the integrity and usability of the information, including uniformly converting the information in different formats into texts, and editing the text information according to a certain format so as to facilitate the subsequent training of the AI large model;
203. selection of AI large models applicable to cooking: selecting a large model of a third party AI at home and abroad, and measuring the large model by using accuracy, response speed and diversity indexes;
204. training a cooking AI big model: finishing a cooking data set through the second step, and then carrying out Fine tuning Fine-tuning training on the cooking data set by using a third-party AI large model, and generating a cooking AI large model with all relevant data of a cooking method after training is finished;
205. verification test cooking AI big model: performing effect detection evaluation of the specific task on the cooking AI large model generated in the fourth step, if the evaluation effect does not pass, continuously repeating the steps of the first step, the second step, the third step and the fourth step, retraining until the effect evaluation passes, generating the cooking AI large model and storing the cooking AI large model in a cloud platform;
206. deployment and maintenance of a cooking AI large model: the newly generated AI large model is deployed to the built-in AI large model module 3 of the intelligent cooking apparatus, and continuous maintenance and updating are performed, and data are updated periodically to ensure timeliness and accuracy of the data.
In this embodiment, the selection 203 of the process of constructing the cooking AI large model based on the cooking method with built-in AI large model is applied to the cooking AI large model, specifically, the AI large model of Baichuan2-13B is selected, and parameters of the AI large model are as follows: hidden layer dimension: 5,120, the number of layers: 40, attention header number: 40, vocabulary size: 64,000, total parameters: 13,264,901,120 training data (keys): 1.4 trillion, position coding: ALiBi, maximum length: 4,096;
in this embodiment, the training process of the cooking AI big model based on 204 of the process of constructing the cooking AI big model of the cooking method with built-in AI big model is as follows: firstly, downloading model weights of baichuan13b from huggingface, then downloading a belle data set train_0.5M_CN to the local and placing under a dataset folder under an item directory, finally running a sft_lora.py script, then quantizing the Baichuan LLM by adopting nf4 of qlora and a double quantization mode, and finally carrying out instruction fine adjustment by adopting lora.
In a preferred embodiment, when the user uses the APP of the mobile terminal 2 or the information collection module 6 of the intelligent cooking apparatus to send the cooking requirement information presented by the user 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 invokes the AI built-in large model to analyze the cooking requirement information and generate a new cooking method, and then sends the new cooking method to the processor 5, the new cooking method has related intelligent cooking apparatus stations, cooking working steps, cooking parameters or cooking operation instructions requiring user cooperation, the processor 5 determines the intelligent cooking apparatus stations according to the new cooking method, then performs related parameter settings on the intelligent cooking apparatus stations, and then controls the intelligent cooking apparatus stations to cook.
In a preferred embodiment, when the user needs to cook the "hot and sour shredded potatoes" dish, the user uses the APP of the user mobile terminal 2 to send out the cooking requirement information, and the cooking requirement information includes these contents: the spicy degree of dishes, such as micro, medium and thick; the salty degree of dishes, such as light, medium and salty; acidity of dishes, such as micro, medium, and dense; the dishes are large, medium and small; if the user wants to add some innovation, some side dishes such as sausage, egg and meat balls can be added, the user can freely match with the existing food materials on the APP interface, and then the requirement is submitted on the APP. The 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 analyzes the cooking requirement information and generates a new cooking method, and then the new cooking method is transmitted to the APP of the user mobile terminal 2 of the user via the processor 5 by the wireless communication module 7, and after the user confirms to cook on the APP, the processor 5 performs related operations according to the new cooking method, wherein the new cooking method includes the following contents: a cooking device station, a cooking working step and cooking parameters, wherein the cooking working step; each step needs to use a corresponding intelligent cooking device station; the cooking parameters comprise parameters such as working time setting, power setting, temperature setting, water quantity setting and the like; cooking operation instructions matched with users are needed.
The specific implementation is as follows:
first, a hot pot: the processor 5 sets the working parameters of the intelligent cooking apparatus station fry 11 to 3000W power, heats to 200 degrees and maintains according to the new cooking method. After the parameters are set, starting the station frying 11 of the intelligent cooking device to work;
secondly, the user cooperates with oil discharge: simultaneously, the processor 7 synchronously sends information to the APP of the mobile terminal 2 of the user according to the new cooking method, reminds the user of dismantling the oil package, puts quantitative oil into the pot after hearing the prompt tone, and simultaneously, the processor 5 detects whether the user operates according to the requirement through the operation detection feedback system 14;
thirdly, the user cooperates with dish placing: after the processor 5 detects that the pot temperature reaches 200 ℃, the processor 5 synchronously sends information to the APP of the user mobile terminal 2 according to a new cooking method, reminds the user of dismantling the package of the vegetable materials, and after hearing the prompt tone, pours quantitative vegetables into the pot, and meanwhile, the processor 5 detects whether the user operates according to the requirement through the operation detection feedback system 14;
fourth, stir-frying: the processor 5 adjusts and sets the working parameters of the station frying 11 of the intelligent cooking device to 3000W power according to a new cooking method, heats the station frying 11 to 200 ℃ and keeps and starts to stir-fry for 5 minutes, simultaneously sends information to the APP of the mobile terminal 2 of the user, reminds the user to remove the vinegar material package, and after hearing the prompt tone, pours quantitative vinegar into the pot, and meanwhile, the processor 5 detects whether the user operates according to the requirement through the operation detection feedback system 14;
Fourth, adding seasonings: the processor 5 adjusts and sets the working parameters of the station frying 11 of the intelligent cooking device to 2000W power according to a new cooking method, heats the station frying 11 to 200 ℃ and keeps and starts to stir-fry for 1 minute, simultaneously sends information to the APP of the mobile terminal 2 of the user, reminds the user to dismantle the seasoning package, and after hearing the prompt tone, pours quantitative seasonings into the pot, and meanwhile, the processor 5 detects whether the user operates according to the requirement through the operation detection feedback system 14;
fifthly, vegetable discharging: the processor 5 sends information to the APP of the user mobile terminal 2 according to the new cooking method, reminds the user to put the dish at the designated position of the station frying 11 of the intelligent cooking device, controls the wok of the station frying 11 of the intelligent cooking device to turn over to pour the dish out of the dish after detecting that the user operates according to the requirement through the operation detection feedback system 14, and reminds the user to take the dish;
sixth, cleaning: the processor 5 controls the intelligent cooking apparatus station frying 11 to perform self-cleaning operation according to the new cooking method, and then performs a standby state.
The above is just to illustrate the operation process of one dish, and in the specific implementation, the operation process can be more intelligent and personalized according to the user requirement, and the built-in AI large model can customize a plurality of cooking methods according to the user requirement to be selected by the user.
In this embodiment, the calling cooking built-in AI large model of the cooking method based on the built-in AI large model of the present invention specifically includes the following steps:
301. assembling a query statement;
302. and (3) carrying out reasoning operation on the built-in AI large model: the query sentence is transmitted to the AI big model module 3, and the AI big model module 3 performs reasoning operation, wherein the following four processes are included:
302-1, understanding input: the distributed semantic parsing firstly receives a text sequence and converts the text sequence into word vectors, and the process is based on distributed semantic assumptions, namely word senses are determined by the use of the word senses in the context;
302-2, parameter association: context focus linkage, inputting these word vectors into the transform's Encoder to generate a context representation;
302-3, generating an answer: generating probability modeling, initializing a Decoder part of a transducer by the model, inputting the output of the Encoder and the current output sequence into the Decoder together, generating probability distribution of the next word by the Decoder, and selecting the word with the maximum probability or other set probability distribution as output, wherein the word is added into the output sequence;
302-4, selecting the most appropriate answer: the dynamic word string evolves, the steps are repeated, new words are added to the output sequence each time, and a complete output sequence is generated;
303. And (5) returning a result by the built-in AI large model: after the built-in AI large model reasoning operation is completed, returning information related to the new cooking method and an operation instruction of the new cooking method, wherein the information comprises contents in text, pictures, audio, video or 3D model formats provided for users;
the invention provides an intelligent cooking apparatus, which is characterized in that the intelligent cooking apparatus is provided with an information collecting module 6, a storage module 4, an AI large model module 3, a processor 5 and a wireless communication module 7, wherein the AI large model module 3 is positioned in the storage module 4, the information collecting module 6, the storage module 4, the AI large model module 3 and the wireless communication module 7 are electrically connected with the processor 5, the information collecting module 6 is used for collecting all food cooking related information presented by a user through voice, video, text, pictures or a 3D model format, the processor 5 is connected with a cloud platform 16 and a user mobile terminal 2 through the wireless communication module 7, the processor 5 is used for executing the cooking method based on the built-in AI large model and deployed in the AI large model module 3 in claim 1, the new cooking method is characterized in that after the built-in AI large model reasoning operation is finished, the new cooking method is sent to an intelligent cooking device through the storage module 4 and prompts a user to cook, the information collection module 6 is provided with information input ports of voice, video, text, pictures and 3D models, a user can input cooking requirement information through the information input ports, or the information collection module 6 is connected with the user mobile terminal 2 through the wireless communication module 7 to receive the cooking requirement information, the large AI model is deployed in the AI large AI 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 of each intelligent cooking device station of the intelligent cooking device and control each intelligent cooking device station to execute related instructions, when the cooking requirement information of the user is input through the information collection module 6 and then is transmitted to the AI large model module 3 through the processor 5, the AI large model module 3 analyzes and generates a cooking method, and the cooking method comprises specific cooking data information, the processor 5 can set parameters of related intelligent cooking device stations according to the cooking data information and control the relevant intelligent cooking device stations to execute related instructions, and during the putting and using of the intelligent cooking device, the cloud platform 16 or the site can regularly update and maintain the built-in cooking AI large model of the intelligent cooking device.
In this embodiment, the intelligent cooking apparatus of the present invention is provided with intelligent cooking apparatus stations for frying 8, baking 9, cooking 10, air frying 12, steaming 13 and frying 11, the intelligent cooking apparatus stations for frying 8, baking 9, cooking 10, air frying 12, steaming 13 and frying 11 are provided with a set of common operation detection feedback system 14, the operation detection feedback system 14 is used for detecting whether the cooking operation performed by the user in the intelligent cooking apparatus meets the requirements of the cooking method based on the built-in AI large model, the operation detection feedback system 14 comprises detection devices such as a camera, infrared detection, radar detection, magnetic detection, weight detection and the like, and is used for detecting whether the cooking operation performed by the user in the intelligent cooking apparatus meets the requirements of the cooking method based on the built-in AI large model, for example, in the intelligent cooking apparatus stations for frying, the intelligent cooking apparatus requires the user to put the food to be cooked on the stations, but when the user is not put or misplaced, the operation detection feedback system 14 is used for detecting that the food to be fed back to the processor 5 of the intelligent cooking apparatus through the camera, the information is fed back to the intelligent cooking apparatus, the processor 5 is controlled by the processor 5, the intelligent cooking apparatus is also used for carrying out the combination of the information processing by the camera detection system after the detection and the detection device is used for detecting that the operation by the intelligent cooking apparatus is stopped by the intelligent cooking apparatus, and the user is only used for carrying out the detection system;
In a first embodiment, referring to fig. 4, the intelligent cooking apparatus stations of the intelligent cooking apparatus of the present invention, namely, the fry 8, bake 9, cook 10, air fry 12 and steam 13, are provided with a common operation detection feedback system 14, otherwise identical to the preferred embodiment of the present invention;
in a second embodiment, the intelligent cooking appliance stations of the intelligent cooking appliance of the present invention, including the bake 9, cook 10, air fry 12 and steam 13 of fig. 5, are provided with a common operation detection feedback system 14, otherwise identical to the preferred embodiment of the present invention;
in a third embodiment, the intelligent cooking appliance station of the intelligent cooking appliance of the present invention, shown in fig. 6, for the cooking appliance 10 and the frying appliance 11, is provided with a common operation detection feedback system 14, otherwise identical to the preferred embodiment of the present invention;
in a fourth embodiment, referring to fig. 7, a set of operation detection feedback systems 14 are provided at the intelligent cooking apparatus station of the fry 11 of the intelligent cooking apparatus according to the present invention, and the other embodiments are the same as the preferred embodiments of the present invention;
in a fifth embodiment, referring to fig. 8, the intelligent cooking apparatus of the present invention is not provided with the user mobile terminal 2, and the other embodiments are the same as the preferred embodiments of the present invention;
in other embodiments, the intelligent cooking appliance is one or a combination of a cooker, air fryer, electromagnetic oven, microwave oven, steam oven, electric cooker, electric pressure cooker, electric stewpan, integrated stove, or frying pan, and the one or more intelligent cooking appliances include one or more cooking functions of frying, air frying, roasting, frying, stewing, steaming, boiling, or baking to perform cooking food material work.
In other embodiments, when the intelligent cooking device prompts a user to cook different food materials, the intelligent cooking device is placed at different stations on the intelligent cooking device to perform operation cooking, for example, a stir-frying station is different from a steaming, air frying and frying station in the cooking position in the intelligent cooking device, and the user needs to be prompted to cook at the corresponding intelligent cooking device station; the cooking working steps comprise the steps that the intelligent cooking device prompts a user to do cooking steps when cooking different food materials; the cooking parameters comprise the cooking time, cooking temperature, cooking power and the like of the food; the cooking operation instructions to be matched by the user comprise operations of confirming cooking parameters or confirming that the intelligent cooking device station and the cooking working steps are completed in the mobile terminal.
In other embodiments, the intelligent cooking apparatus is a variety of integrated bodies with a plurality of cooking functions of frying, air frying, baking, frying, stewing, steaming, boiling or baking, for example, the intelligent cooking apparatus is a cooking apparatus with integrated functions of air frying, baking, frying, steaming, or the like, or a cooking apparatus with integrated functions of frying, air frying, stewing, boiling, and the like.
In other embodiments, the intelligent cooking apparatus is provided with a plurality of cooking functions including one or more cooking functions of stir-frying, quick-frying, cooking, frying, pasting, burning, stewing, steaming, quick-boiling, braising, quick-frying, stirring, pickling, baking, marinating, freezing, wire drawing, honeydew, fumigating, rolling, sliding or baking, such as a cooking appliance with functions of honeydew, baking, frying, steaming, or a cooking appliance with functions of stir-frying, air frying, braising, stewing, boiling, and the like.
In this embodiment, the intelligent cooking apparatus of the present invention is provided with a man-machine interaction system 15, the man-machine interaction system 15 is used for information interaction between the intelligent cooking apparatus and the user 1, including the user operating the user mobile terminal 2 to confirm cooking parameters and start cooking operation instructions, the man-machine interaction system 15 includes a touch screen, a sound input/output unit, an image input unit and a camera device, the man-machine interaction system 15 is electrically connected with the processor 5 for information interaction between the intelligent cooking apparatus and the user 1, the user 1 can input cooking requirement information in a voice, video, text, picture or 3D model manner through the man-machine interaction system 15, the man-machine interaction system 15 sends the cooking requirement information to the processor 5, the processor 5 transmits the cooking requirement information to the AI large model module 3, the AI large model module 3 analyzes the cooking requirement information and generates a new cooking method, and then sends the new cooking method to the processor 5, the processor 5 sends the information to the user through the man-machine interaction system 15 in one or a combination of the voice, the user 1 can input cooking requirement information in a voice, a text, a picture or a text or a human-machine interaction system 1 is set to confirm the cooking process by the intelligent cooking apparatus according to the user gesture, and the cooking process is set up in the relevant to the intelligent cooking apparatus.
The cooking method based on the built-in AI large model and the related specific implementation steps of the intelligent cooking device for cooking are as follows:
the specific implementation process of preprocessing cooking data of the process 202 for constructing the cooking AI large model based on the cooking method with the built-in AI large model is as follows: all the acquired information is processed, so that the integrity and usability of the information are ensured. The method comprises the steps of uniformly converting information in different formats into texts, and editing the text information according to a certain format so as to facilitate subsequent training of the large AI model.
The data format is as follows:
instrucing: task instructions cannot be null.
input: the task input may be null. If not, instruction, input is spliced together as input of task when training data is processed inside the project
output: task output, cannot be empty
Examples of data are as follows:
{
"construction": "sugar and vinegar spareribs prefabricated vegetable cooking method",
"input":"",
1, pouring the prefabricated spareribs into a pot for frying, frying under the condition of heating oil temperature of 6, and keeping medium-fire slow frying. 2. The pork rib is slightly yellowish brown in color after 3 minutes of frying process, so that the pork rib can be filled with the oil. 3. A small amount of base oil is left in the pot, sugar and vinegar juice is added, and the pot is boiled with strong fire. 4. Putting the fried spareribs into sweet and sour juice, and stir-frying uniformly "
}
The specific implementation process of the built-in AI large model return result in step 303 of the flow of calling the built-in AI large model based on the AI large model cooking method is as follows: and after the built-in AI large model reasoning operation is completed, returning cooking method related information and cooking method operation instructions, wherein the cooking method operation instructions comprise contents in various formats of texts, pictures, audios, videos and 3D models. Returning to the Jason format, the contents are as follows:
{
"recobiname": recobiname// cooking method name,
"recontext": text// cooking method information,
"recoppic": pic _ Url// picture of cooking method information Url,
"recoperaudio": audio_url// audio of cooking method information,
"recop video": video_url// video of cooking method information Url,
"record 3D":3D_url// 3D model Url of cooking method information,
"recoeCommand": recoeCommand// cooking method operation instruction
}
The invention discloses a cooking method based on an AI large model and an intelligent cooking device for cooking sweet and sour pork ribs, wherein the cooking method comprises the following steps of:
1. transmitting the request information of cooking the sweet and sour pork ribs to a cooking AI large model;
2. and calling the cooking AI large model through the Prompt query statement, and carrying out reasoning operation on the cooking AI large model.
3. After the cooking AI big model is subjected to reasoning operation, the following results are returned:
{
"recoername": sugar and vinegar spareribs cooking method "// cooking method name,
1, pouring the prefabricated spareribs into a pot for frying, frying under the condition of heating oil temperature of 6, and keeping medium-fire slow frying. 2. The pork rib is slightly yellowish brown in color after 3 minutes of frying process, so that the pork rib can be filled with the oil. 3. A small amount of base oil is left in the pot, sugar and vinegar juice is added, and the pot is boiled with strong fire. 4. Putting the fried spareribs into sweet and sour juice, and stir-frying uniformly "
Information on the method of cooking/the method of cooking,
"recoppic": "None"// None,
“recipeAudio”:
"https:// www.ixigua.com/7289152022341747234 logtag=43fb 2a078d33289b226b"// audio Url of cooking method information,
“recipeVideo”:
"https:// www.ixigua.com/7289152022341747234 logtag=43fb 2a078d33289b226b"// video of cooking method information Url,
"record 3D":3D_url// 3D model Url of cooking method information,
"recoecommand" fry, 30 seconds, temperature set to 100 degrees fry, 20 seconds, steam, 2 minutes, end operation "// cooking method operation instruction
}
The invention relates to a cooking method based on a built-in AI large model and an intelligent cooking device thereof, which have the beneficial effects that: according to the cooking method based on the built-in AI large model and the intelligent cooking device thereof, all information about food cooking methods from people in all countries, groups and age groups which are presented through voice, video, text, pictures or 3D models is collected, even the cooking method comprises a specific individual custom or favorite cooking method of various foods, the third-party AI large model is used for Fine tuning of the Fine-tuning training, the cooking AI large model with all relevant data of the massive cooking methods is generated after the training is completed, a user inputs cooking requirement information through voice, video, text, pictures or 3D model formats, the intelligent cooking device executes a new cooking method generated by the built-in AI large model of the user, and the intelligent cooking device cooks foods meeting different requirements of different users, even cooks foods meeting different requirements of the same user, not only can meet different requirements of different users but also can meet different requirements of different environments of the same user in different time, but also is built in the intelligent cooking device, and the intelligent cooking device is free from the influence of the use of the AI large model.
The foregoing detailed description of the invention has been presented for purposes of illustration and description, but is not intended to limit the scope of the invention, i.e., the invention is not limited to the details shown and described.
Claims (10)
1. The utility model provides a cooking method based on built-in AI large model, which is characterized in that is applied to intelligent cooking device, intelligent cooking device is equipped with information collection module, storage module, AI large model module, treater and wireless communication module, AI large model module is located the storage module the inside, information collection module, storage module, AI large model module, wireless communication module with treater electric connection, information collection module is used for receiving the cooking demand information that the user input through pronunciation, video, text, picture or 3D model form to this cooking demand information is pushed to the AI large model module that is located the storage module, AI large model module carries out analysis and generates new cooking method to the cooking demand information, and will new cooking method is passed through the storage module and is sent to the treater, new cooking method includes intelligent cooking device station, culinary art working procedure, culinary art parameter or the culinary art operation instruction that needs the user to cooperate, the new cooking method that the AI large model produced includes: the method comprises the steps of constructing a cooking AI big model and calling a cooking built-in AI big model, wherein the construction of the cooking AI big model comprises the steps of collecting cooking data, preprocessing the cooking data, selecting an AI big model applicable to cooking, training the cooking AI big model, verifying a test cooking AI big model and deploying and maintaining the cooking AI big model in the AI big model, a user calls the cooking built-in AI big model through the AI big model to generate the new cooking method through cooking requirement information input through voice, video, text, pictures or 3D model formats, the calling of the cooking built-in AI big model comprises the steps of assembling query sentences, carrying out reasoning operation on the built-in AI big model and returning results of the built-in AI big model, and the new cooking method is sent to the processor through the storage module after the reasoning operation of the cooking built-in AI big model is completed.
2. The cooking method based on the built-in AI large model according to claim 1, wherein: the constructing the cooking AI large model includes:
first, collecting cooking data: collecting all information about food cooking methods presented by voice, video, text, pictures or 3D models;
step two, preprocessing cooking data: processing all collected information related to food cooking methods to ensure the integrity and usability of the information, including uniformly converting the information in different formats into texts, and editing the text information according to a certain format so as to facilitate the subsequent training of the AI large model;
third, selecting an AI large model applicable to cooking: selecting a large model of a third party AI at home and abroad, and measuring the large model by using accuracy, response speed and diversity indexes;
fourth step, training a cooking AI large model: finishing a cooking data set through the second step, and then carrying out Fine tuning Fine-tuning training on the cooking data set by using a third-party AI large model, and generating a cooking AI large model with all relevant data of a cooking method after training is finished;
fifth step, verifying and testing the large cooking AI model: performing effect detection evaluation of the specific task on the cooking AI large model generated in the fourth step, if the evaluation effect does not pass, continuously repeating the steps of the first step, the second step, the third step and the fourth step, retraining until the effect evaluation passes, generating the cooking AI large model and storing the cooking AI large model in a cloud platform;
Sixth, deploying and maintaining a large cooking AI model: and deploying the newly generated AI large model to the built-in AI large model module of the intelligent cooking device, continuously maintaining and updating, and periodically updating data to ensure 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 of selection can be applied to a cooking AI large model, wherein the AI large model of Baichuan2-13B is selected, and parameters of the AI large model are as follows: hidden layer dimension: 5,120, the number of layers: 40, attention header number: 40, vocabulary size: 64,000, total parameters: 13,264,901,120 training data (keys): 1.4 trillion, position coding: ALiBi, maximum length: 4,096.
4. The AI-large model-based cooking method according to claim 3, wherein: the training process of the fourth step of training the cooking AI large model is as follows: firstly, downloading model weights of baichuan13b from huggingface, then downloading a belle data set train_0.5M_CN to the local and placing under a dataset folder under an item directory, finally running a sft_lora.py script, then quantizing the Baichuan LLM by adopting nf4 of qlora and a double quantization mode, and finally carrying out instruction fine adjustment by adopting lora.
5. The cooking method based on the built-in AI large model according to claim 1, wherein: the method for calling the built-in AI large model for cooking specifically comprises the following steps:
firstly, assembling a query statement;
secondly, carrying out reasoning operation by a built-in AI large model: the query sentence is transmitted to the AI big model, and the AI big model carries out reasoning operation, wherein the reasoning operation comprises the following processes:
1) Understanding input: the distributed semantic parsing firstly receives a text sequence and converts the text sequence into word vectors, and the process is based on distributed semantic assumptions, namely word senses are determined by the use of the word senses in the context;
2) Parameter association: context focus linkage, inputting these word vectors into the transform's Encoder to generate a context representation;
3) Generating an answer: generating probability modeling, initializing a Decoder part of a transducer by the model, inputting the output of the Encoder and the current output sequence into the Decoder together, generating probability distribution of the next word by the Decoder, and selecting the word with the maximum probability or other set probability distribution as output, wherein the word is added into the output sequence;
4) Selecting the most appropriate answer: the dynamic word string evolves, the steps are repeated, new words are added to the output sequence each time, and a complete output sequence is generated;
Thirdly, the built-in AI large model returns the result: and after the built-in AI large model reasoning operation is completed, returning information related to the new cooking method and operation instructions of the new cooking method, wherein the information comprises contents provided for text, pictures, audio, video or 3D model formats of a user.
6. The intelligent cooking device is characterized by comprising an information collecting module, a storage module, an AI large model module, a processor and a wireless communication module, wherein the AI large model module is positioned in the storage module, the information collecting module, the storage module, the AI large model module and the wireless communication module are electrically connected with the processor, the information collecting module is used for collecting all food cooking-related information presented by a user through voice, video, text, pictures or a 3D model format, the wireless communication module can be used for connecting the processor with a cloud platform and a user mobile terminal, the processor is used for executing the cooking method based on the built-in AI large model module, which is deployed in the claim 1, and the new cooking method is sent to the intelligent cooking device through the storage module after the built-in AI large model reasoning operation of cooking is completed and prompts the user to cook.
7. The intelligent cooking appliance of claim 6, wherein: the intelligent cooking device comprises intelligent cooking device stations for frying, roasting, boiling, air frying, stewing, baking, steaming or frying, and the intelligent cooking device receives the new cooking method and prompts a user to cook and carry out cooking work steps at the corresponding intelligent cooking device stations.
8. The intelligent cooking appliance of claim 7, wherein: the intelligent cooking device stations for frying, roasting, boiling, air frying, stewing, baking, steaming or frying are provided with corresponding operation detection feedback systems, and the operation detection feedback systems are used for detecting whether the cooking operation performed by a user in the intelligent cooking device meets the requirements of the new cooking method.
9. The intelligent cooking appliance of claim 6, wherein: the intelligent cooking device is provided with a man-machine interaction system, wherein the man-machine interaction system is used for information interaction between the intelligent cooking device and a user, and comprises the steps that the user operates a user mobile terminal to confirm cooking parameters and start cooking operation instructions.
10. The intelligent cooking appliance of claim 6, wherein the new cooking method is transmitted to a user mobile terminal via the wireless communication module, the user mobile terminal receives the new cooking method and prompts the user to perform cooking and cooking work steps at the corresponding intelligent cooking appliance station, and the user operates the mobile terminal to confirm cooking parameters and initiate cooking operation instructions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311620532.0A CN117764777A (en) | 2023-11-29 | 2023-11-29 | Cooking method based on built-in AI large model and intelligent cooking device thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311620532.0A CN117764777A (en) | 2023-11-29 | 2023-11-29 | Cooking method based on built-in AI large model and intelligent cooking device thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117764777A true CN117764777A (en) | 2024-03-26 |
Family
ID=90324737
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311620532.0A Pending CN117764777A (en) | 2023-11-29 | 2023-11-29 | Cooking method based on built-in AI large model and intelligent cooking device thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117764777A (en) |
-
2023
- 2023-11-29 CN CN202311620532.0A patent/CN117764777A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104595949B (en) | Realize the domestic intelligent cooking system that cuisines are shared | |
| US9316400B2 (en) | Appliance control method, speech-based appliance control system, and cooking appliance | |
| CN105358913B (en) | Heating device | |
| CN105981999A (en) | Automatic cooking method and system, intelligent cooking equipment and intelligent oven | |
| CN111596563B (en) | Intelligent smoke kitchen system and cooking guiding method thereof | |
| CN106662334A (en) | Method for data communication with a domestic appliance by a mobile computer device, mobile computer device and domestic appliance | |
| CN105444222A (en) | Cooking control method and system of microwave oven, cloud server and microwave oven | |
| CN106264065A (en) | A kind of intelligent kitchen cooking system and the method for intelligence auxiliary cooking | |
| CN108447543A (en) | Menu method for pushing based on cooking equipment and device | |
| US20230168651A1 (en) | Generating method, processing method and executing method of data and device | |
| CN107664959A (en) | Intelligent cooking system and its menu generation, cooking methods | |
| CN110953609A (en) | Cooking control method, storage medium, cooking control device and cooking system | |
| CN106773859A (en) | A kind of intelligent cooking control method | |
| CN109918522A (en) | A kind of vegetable cooking learning method and device based on cooking platform | |
| CN111345665B (en) | Cooking method and cooking device | |
| CN110045638A (en) | A kind of recommended method of cooking information, device and storage medium | |
| CN106453545A (en) | Recipe showing and interacting method and system, and intelligent device | |
| CN206378716U (en) | Kitchen house keeper robot | |
| CN112383455B (en) | Data generation method and execution method and equipment | |
| CN110858279A (en) | Food material identification method and device | |
| CN110604459A (en) | Totally-enclosed oil-smoke-free intelligent cooking robot and control system thereof | |
| CN110084372A (en) | Intelligent menu generation method and intelligent cooking method based on adaptive learning | |
| CN113662446A (en) | Internet of things-based cooking assistance method and device, intelligent terminal and storage medium | |
| CN117812770A (en) | Intelligent induction cooker cooking method, system, electronic equipment and medium | |
| CN110522289A (en) | A kind of control method of intelligent electric cooker, device and intelligent electric cooker |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |