Photoresist model training method, device, equipment, medium and product
By selecting different data groups multiple times in the graph dataset and training and verification, and adjusting the parameters of the photoresist model, the problem of insufficient generalization ability of the traditional photoresist model is solved, and higher robustness and adaptability are achieved.
Patent Information
- Application Number
- CN202411558254.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-01
- Publication Date
- 2025-02-25
AI Technical Summary
Traditional photoresist models are trained on fixed data sets, which limit the generalization ability and applicability of the model, especially when facing different production conditions and material characteristics.
By selecting two-part graph data groups multiple times in the graph data set, a variety of selection results are formed. In each selection result, the data is divided into a training subset and a verification subset, the initial model is trained and its performance is evaluated, and the model parameters are adjusted based on multiple evaluation indicators to obtain the target photoresist model.
This method significantly improves the robustness and adaptability of the photoresist model, reduces the risk of model overfitting caused by data deviation, enhances the generalization ability and applicability of the model, and greatly improves its application value in complex manufacturing environments.
Smart Images
Figure CN119516305A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the field of semiconductor manufacturing, and in particular, relates to a method, device, equipment, medium and product for training a photoresist model. Background Art
[0002] With the continuous advancement of semiconductor manufacturing technology, the criticality of photolithography technology in chip manufacturing has become increasingly prominent. As a key component of the photolithography process, the quality and performance of photoresist directly affect the accuracy of chip circuit patterns and manufacturing quality.
[0003] Traditional photoresist models are usually trained and tested on fixed datasets, which may limit the generalization ability and applicability of the model, especially when faced with different production conditions and material properties. Summary of the invention
[0004] The embodiments of the present application provide a method, device, equipment, medium and product for training a photoresist model, which can improve the generalization ability and applicability of the photoresist model.
[0005] On the one hand, an embodiment of the present application provides a method for training a photoresist model, comprising:
[0006] Acquire a plurality of graphic data groups consisting of mask patterns and corresponding photoresist patterns to obtain graphic data sets;
[0007] Performing multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, so as to obtain multiple different selection results; the two parts of the graphic data set in each selection result are respectively a graphic data training subset and a graphic data verification subset;
[0008] For each of the selected results, after training an initial photoresist model based on the graphic data training subset, evaluating the trained initial photoresist model based on the graphic data verification subset to obtain an evaluation index;
[0009] Based on the multiple evaluation indicators, the parameters of the initial photoresist model are adjusted to obtain a target photoresist model.
[0010] On the other hand, the performing of multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, to obtain multiple different selection results, includes:
[0011] Dividing the graphic data set to obtain a plurality of mutually exclusive graphic data subsets;
[0012] Each time, one of the graphic data subsets is selected from each of the graphic data subsets, and the selected graphic data subset is determined as the graphic data verification subset;
[0013] The remaining subset of the graphic data is determined as the graphic data training subset.
[0014] On the other hand, the graphic data set is divided to obtain a plurality of mutually exclusive graphic data subsets, including:
[0015] The graphic data set is divided based on the proportion of each category of graphics in the graphic data set to obtain a plurality of mutually exclusive graphic data subsets.
[0016] On the other hand, for each of the selected results, after training the initial photoresist model based on the graphic data training subset, the trained initial photoresist model is evaluated based on the graphic data verification subset to obtain an evaluation index, including:
[0017] For each selection result, the following operations are performed:
[0018] Training the initial photoresist model based on the graphic data training subset to obtain a first photoresist model;
[0019] Inputting the mask pattern in the pattern data verification subset into the first photoresist model to obtain a simulation pattern;
[0020] Determining an evaluation index of the first photoresist model based on the simulation pattern and the corresponding photoresist pattern;
[0021] The adjusting the parameters of the initial photoresist model based on the plurality of evaluation indicators to obtain a target photoresist model includes:
[0022] Determine the target parameter value of the evaluation indicator obtained from each evaluation;
[0023] Based on the obtained target parameter value, the parameters of the initial photoresist model are adjusted.
[0024] On the other hand, the training of the initial photoresist model based on the graphic data training subset to obtain a first photoresist model includes:
[0025] Determine the weight corresponding to each category of graphics based on the proportion of each category of graphics in the graphic data set; the proportion of each category of graphics in the graphic data set is inversely correlated with the weight;
[0026] The initial photoresist model is trained according to the weights corresponding to each category of graphics to obtain the first photoresist model.
[0027] On the other hand, after adjusting the parameters of the initial photoresist model based on the plurality of evaluation indicators to obtain a target photoresist model, the method further includes:
[0028] Verifying the performance of the target photoresist model to obtain a verification result;
[0029] When the verification result meets the preset conditions, it is determined that the target photoresist model meets the actual requirements.
[0030] On the other hand, an embodiment of the present application provides a photoresist model training device, comprising:
[0031] An acquisition module, used for acquiring a plurality of graphic data groups consisting of mask patterns and corresponding photoresist patterns to obtain a graphic data set;
[0032] A selection module, used for performing multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, so as to obtain multiple different selection results; the two parts of the graphic data set in each selection result are a graphic data training subset and a graphic data verification subset;
[0033] An evaluation module, for each of the selected results, after training an initial photoresist model based on the graphic data training subset, and then evaluating the trained initial photoresist model based on the graphic data verification subset to obtain an evaluation index;
[0034] The adjustment module is used to adjust the parameters of the initial photoresist model based on the multiple evaluation indicators to obtain a target photoresist model.
[0035] On the other hand, an embodiment of the present application provides a photoresist model training device, including: a processor and a memory storing computer program instructions;
[0036] When the processor executes the computer program instructions, the photoresist model training method as described above is implemented.
[0037] On the other hand, an embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the training method of the photoresist model as described above is implemented.
[0038] On the other hand, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the photoresist model training method as described above.
[0039] A method for training a photoresist model provided in an embodiment of the present application selects two different groups of graphic data in a graphic data set multiple times to obtain a corresponding plurality of selection results. Then, for each selection result, the initial photoresist model is trained based on the graphic data training subset, and the trained initial photoresist model is evaluated based on the graphic data verification subset to obtain an evaluation index; finally, based on multiple evaluation indexes, the parameters of the initial photoresist model are adjusted to obtain a target photoresist model. The embodiment of the present application uses different parts of the graphic data set as training sets and verification sets in turn to comprehensively evaluate the performance of the model, thereby effectively enhancing the robustness and adaptability of the photoresist model. Compared with the traditional method of training with a fixed data set, this technology can significantly reduce the risk of model overfitting caused by data deviation with less training data, improve the generalization ability and applicability of the photoresist model, and thus improve the practical application value of the model in a complex manufacturing environment. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0041] Figure 1 A schematic diagram of a process of training a photoresist model provided by an embodiment of the present application is shown;
[0042] Figure 2 Schematic diagram of the fitting effect of training the photoresist model for a traditional fixed data set;
[0043] Figure 3 A schematic diagram of the fitting effect of the photoresist model trained using a multi-fold cross-validation method for this application;
[0044] Figure 4 A schematic diagram of the structure of a photoresist model training device provided in an embodiment of the present application is shown;
[0045] Figure 5 A schematic diagram of the hardware structure of a photoresist model training device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION
[0046] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.
[0047] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "include..." does not exclude the presence of other identical elements in the process, method, article or device including the element.
[0048] In order to solve the problems of the traditional solution, the embodiment of the present application provides a method, device, equipment, medium and product for training a photoresist model. The following first introduces the method for training a photoresist model provided by the embodiment of the present application. Figure 1 FIG. 1 is a flow chart of a method for training a photoresist model provided by an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:
[0049] S101: Acquire a plurality of graphic data groups consisting of mask patterns and corresponding photoresist patterns to obtain graphic data sets.
[0050] After the mask pattern is exposed by photolithography, a photoresist pattern is formed on the photoresist. In order to train the photoresist model, the original mask pattern and the actual corresponding photoresist pattern are needed as training data. In the obtained graphic data set, the mask pattern and the photoresist pattern are one-to-one corresponding, and a mask pattern and the corresponding photoresist pattern constitute a graphic data set.
[0051] S102: performing multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, so as to obtain multiple different selection results.
[0052] In each selection result, the two selected graphic data groups are used as the graphic data training subset and the graphic data verification subset respectively; and each time the selection is made, it is necessary to ensure that the graphic data groups in the graphic data training subset and the graphic data verification subset are not exactly the same. If the graphic data training subset and the graphic data verification subset are the same, then the model verification through the graphic data verification subset may not be of reference value. In addition, it is also necessary to ensure that the selection results of each selection are different to avoid repeated training caused by the same selection results.
[0053] The embodiment of the present application does not limit the specific selection method, and completely mutually exclusive graphic data training subsets and graphic data verification subsets can be selected. Since the data volume of the training set is generally larger than the data volume of the verification set, the graphic data set can be divided into multiple graphic data subsets in advance, and one of the graphic data subsets is selected each time as the graphic data verification subset, and the remaining graphic data subsets are all used as graphic data training subsets.
[0054] S103: For each selection result, after training the initial photoresist model based on the graphic data training subset, the trained initial photoresist model is evaluated based on the graphic data verification subset to obtain an evaluation index.
[0055] A variety of selection results are obtained above. For each selection result, the initial photoresist model is trained and evaluated in turn to obtain a plurality of corresponding evaluation indicators.
[0056] The initial photoresist model mentioned in this step needs to be pre-established based on the model parameter data. Therefore, it is necessary to collect and pre-process the model parameter data first. For example, various data related to photoresist in historical data can be collected, including but not limited to the chemical composition of the glue, environmental factors (such as temperature and humidity), and exposure parameters. Then pre-process the data to ensure data quality and consistency.
[0057] After completing the collection and preprocessing of the model parameter data, an initial photoresist model is established based on the collected model parameter data. The model is established using different methods such as physical models combined with machine learning models to predict the behavioral characteristics of photoresists under different process conditions.
[0058] S104: Based on multiple evaluation indicators, adjust the parameters of the initial photoresist model to obtain a target photoresist model.
[0059] Finally, based on the multiple evaluation indicators mentioned above, the photoresist model is optimized and the performance differences of the photoresist model on different training sets and validation sets are analyzed. By adjusting the model parameters, adding data features, and improving the model structure, the photoresist model is optimized to improve its performance consistency and accuracy on different training sets and validation sets, thereby obtaining the target photoresist model.
[0060] In addition, the performance of the obtained target photoresist model can also be evaluated and verified. The optimized photoresist model also needs to undergo a comprehensive performance evaluation, such as the verification of performance indicators such as accuracy, stability, and generalization ability, so as to ensure the reliability and effectiveness of the model in actual semiconductor manufacturing production.
[0061] This application provides a photoresist model accuracy optimization method based on multi-fold cross-validation, which aims to improve the accuracy and stability of the photoresist model through effective data analysis and model evaluation methods. This method can not only make full use of existing data resources, but also dynamically adjust during the model training process to adapt to different production environments and actual application needs.
[0062] A method for training a photoresist model provided in an embodiment of the present application selects two different groups of graphic data in a graphic data set multiple times to obtain a corresponding plurality of selection results. Then, for each selection result, the initial photoresist model is trained based on the graphic data training subset, and the trained initial photoresist model is evaluated based on the graphic data verification subset to obtain an evaluation index; finally, based on multiple evaluation indexes, the parameters of the initial photoresist model are adjusted to obtain a target photoresist model. The embodiment of the present application uses different parts of the graphic data set as training sets and verification sets in turn to comprehensively evaluate the performance of the model, thereby effectively enhancing the robustness and adaptability of the photoresist model. Compared with the traditional method of training with a fixed data set, this technology can significantly reduce the risk of model overfitting caused by data deviation with less training data, improve the generalization ability and applicability of the photoresist model, and thus improve the practical application value of the model in a complex manufacturing environment.
[0063] The above embodiment mentioned that it is necessary to select multiple times in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, so as to obtain multiple different selection results, and then train and verify the model for each selection result. In practical applications, there is no limitation on how to select the graphic data training subset and the graphic data verification subset. The embodiment of the present application provides a specific implementation scheme:
[0064] The graphic data set is divided to obtain a plurality of mutually exclusive graphic data subsets; each time, one graphic data subset in each graphic data subset is determined as a graphic data verification subset, and the remaining graphic data subsets are determined as graphic data training subsets.
[0065] The embodiment of the present application adopts the idea of multi-fold cross-validation, which is a technique for evaluating the performance of machine learning models. It divides the data set into multiple subsets (folds), and then selects one subset as the validation set and the remaining subsets as the training set, and trains and validates the model multiple times to reduce the variance of the evaluation index.
[0066] When dividing the graphic data set, there is no specific requirement for the number of the specific divided graphic data subsets, for example, it can be set to 5 or 10, and can also be adjusted according to the size of the data set and specific circumstances. In addition, there is no specific requirement for the specific method of dividing the graphic data set, which can be random division or division based on the proportion of each category of graphics.
[0067] The embodiment of the present application divides the graphic data set into multiple mutually exclusive graphic data subsets to avoid overlap between the graphic data subsets and affect the effect of subsequent model training. Each subset is used as a validation set in turn, and the remaining subsets are used as training sets. Training and validation are repeated to ensure that each data subset has the opportunity to be used as a validation set and a training set, so as to comprehensively evaluate the generalization ability and stability of the model under limited data.
[0068] In addition, the traditional multi-fold cross-validation algorithm has a certain degree of randomness in data partitioning, which may lead to significant differences in model performance due to different data partitioning, meaning that the same model may show inconsistent accuracy and performance indicators in different cross-validation iterations. To this end, stratified sampling and adding category weights can be used to reduce the impact of randomness.
[0069] Specifically, the embodiment of the present application adopts the idea of stratified sampling to obtain the proportion of each category of graphics in the graphic data set, and then divides the graphic data set based on the proportion of each category of graphics to obtain multiple mutually exclusive graphic data subsets. It is ensured that the distribution ratio of each category of graphics in the multiple graphic data subsets obtained by division is similar to that of the overall graphic data set.
[0070] When dividing the graphic data set, the embodiment of the present application keeps the proportion of graphic samples of each category in each subset close to the proportion in the entire data set, so as to ensure that each subset contains a sufficient number of samples of each category, thereby improving the effect of model training.
[0071] The specific content of the evaluation index mentioned in the above embodiment is not specifically limited. The embodiment of the present application provides an embodiment. For each selection result, after the initial photoresist model is trained based on the graphic data training subset, the trained initial photoresist model is evaluated based on the graphic data verification subset to obtain the evaluation index, including:
[0072] For each selection result, do the following:
[0073] An initial photoresist model is trained based on a graphic data training subset to obtain a first photoresist model; a mask pattern in a graphic data verification subset is input into the first photoresist model to obtain a simulation pattern; based on the simulation pattern and the corresponding photoresist pattern, an evaluation index of the first photoresist model is determined, and the evaluation index may specifically include accuracy and / or error.
[0074] That is, under each selection result, the accuracy and / or error of the first photoresist model is obtained. Therefore, the target parameter value of the accuracy and / or error obtained in each evaluation can be determined. The target parameter value may include the mean value and / or standard deviation, etc., and then the parameters of the initial photoresist model are adjusted based on the obtained target parameter value.
[0075] The embodiment of the present application implements model training through a training subset of graphic data, and then implements evaluation based on a verification subset of graphic data to obtain the accuracy and / or error of the first photoresist model, thereby obtaining model performance. The average value calculated in this embodiment is used as an estimated value of model performance, and the standard deviation reflects the stability of the model performance evaluation. Adjusting and optimizing the model based on these two values can better improve model performance.
[0076] The embodiment of the present application adds graphic category weights to train the model to reduce the impact of randomness. Specifically, firstly, based on the proportion of each category of graphics in the graphic data set, the weight corresponding to each category of graphics is determined; the proportion of each category of graphics in the graphic data set is made to be inversely correlated with the weight; then, the initial photoresist model is trained according to the weight corresponding to each category of graphics to obtain the first photoresist model.
[0077] The embodiments of the present application increase the weight of graphic samples of a smaller category or reduce the weight of majority category samples, so that the model pays more attention to minority category samples, thereby improving the recognition ability of minority categories.
[0078] After adjusting the parameters of the initial photoresist model based on multiple evaluation indicators to obtain the target photoresist model, a comprehensive performance evaluation of the model can be performed. Specifically, the accuracy, stability, and generalization ability of the target photoresist model can be verified to obtain a verification result; then, if the verification result meets the preset conditions, it is determined that the target photoresist model meets the actual needs.
[0079] The embodiment of the present application conducts a comprehensive performance evaluation of the optimized model, including verification of indicators such as accuracy, stability, and generalization ability, to ensure the reliability and effectiveness of the model in actual semiconductor manufacturing production.
[0080] The following is a further explanation of this solution with reference to specific examples. It should be noted that the exemplary embodiments described here are only for explanation and do not limit the scope of use of this application. First, use multiple data sets on the same Graphic Data System (GDS) file to train the photoresist model separately, and then apply the trained model to another GDS file to check the fitting effect. Figure 2 Schematic diagram of the fitting effect of training the photoresist model for the traditional fixed data set; Figure 2As shown, the photoresist pattern is simulated by the photoresist model trained by the traditional scheme. It can be seen that the fitting effect of the photoresist model trained by the traditional scheme is poor and the model has the phenomenon of overfitting. Figure 3 This is a schematic diagram of the fitting effect of the photoresist model trained using a multi-fold cross-validation method for this application; Figure 3 As shown, the photoresist pattern is simulated by the photoresist model trained by the scheme proposed in the embodiment of the present application. It can be seen that the photoresist model of this scheme has a better fitting effect on new data.
[0081] In order to solve the above technical problems, the present application also provides a photoresist model training device. Figure 4 FIG. 1 is a schematic diagram showing the structure of a photoresist model training device provided in an embodiment of the present application. Figure 4 As shown, the device includes the following modules:
[0082] An acquisition module 401 is used to acquire a plurality of graphic data groups consisting of mask patterns and corresponding photoresist patterns to obtain a graphic data set;
[0083] The selection module 402 is used to perform multiple selections in the graphic data set, each time selecting two parts of graphic data sets that are not completely the same, so as to obtain multiple different selection results; the two parts of graphic data sets in each selection result are respectively a graphic data training subset and a graphic data verification subset;
[0084] An evaluation module 403 is used for, for each selected result, after training the initial photoresist model based on the graphic data training subset, evaluating the trained initial photoresist model based on the graphic data verification subset to obtain an evaluation index;
[0085] The adjustment module 404 is used to adjust the parameters of the initial photoresist model based on multiple evaluation indicators to obtain a target photoresist model.
[0086] In some embodiments, the selection module 402 is specifically configured to:
[0087] Divide the graphic data set to obtain multiple mutually exclusive graphic data subsets;
[0088] Each time, a graphic data subset is selected from each graphic data subset, and the selected graphic data subset is determined as a graphic data verification subset;
[0089] The remaining subset of the graphic data is determined as the graphic data training subset.
[0090] In some embodiments, the selection module 402 is specifically configured to:
[0091] Based on the proportion of each category of graphics in the graphic data set, the graphic data set is divided to obtain multiple mutually exclusive graphic data subsets.
[0092] In some embodiments, the evaluation module 403 is specifically configured to perform the following operations for each selection result:
[0093] Training an initial photoresist model based on the training subset of the graphic data to obtain a first photoresist model;
[0094] Inputting the mask pattern in the pattern data verification subset into the first photoresist model to obtain a simulation pattern;
[0095] Determining an evaluation index of a first photoresist model based on a simulation pattern and a corresponding photoresist pattern;
[0096] The adjustment module 404 is specifically used for:
[0097] Determine the target parameter value of the evaluation indicator obtained from each evaluation;
[0098] Based on the obtained target parameter values, the parameters of the initial photoresist model are adjusted.
[0099] In some embodiments, the evaluation module 403 is specifically configured to:
[0100] Based on the proportion of each category of graphics in the graphic data set, the weight corresponding to each category of graphics is determined; the proportion of each category of graphics in the graphic data set is inversely correlated with the weight;
[0101] The initial photoresist model is trained according to the weights corresponding to each category of graphics to obtain a first photoresist model.
[0102] In some embodiments, the photoresist model training device further includes:
[0103] A verification module, for adjusting the parameters of the initial photoresist model based on multiple evaluation indicators to obtain a target photoresist model, and then verifying the performance of the target photoresist model to obtain a verification result;
[0104] The determination module is used to determine whether the target photoresist model meets the actual requirements when the verification result meets the preset conditions.
[0105] The device provided in the embodiment of the present application is the same as the method in the above embodiment, so both have the same embodiments and beneficial effects, which will not be repeated here.
[0106] Figure 5 FIG. 1 is a schematic diagram showing the hardware structure of a photoresist model training device provided in an embodiment of the present application. Figure 5As shown, the photoresist model training device may include a processor 501 and a memory 502 storing computer program instructions.
[0107] Specifically, the processor 501 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.
[0108] The memory 502 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 502 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 502 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 502 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 502 is a non-volatile solid-state memory.
[0109] The memory 502 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Therefore, generally, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of the present disclosure.
[0110] The processor 501 implements any one of the photoresist model training methods in the above embodiments by reading and executing computer program instructions stored in the memory 502 .
[0111] In one example, the photoresist model training device may further include a communication interface 503 and a bus 504. The processor 501, the memory 502, and the communication interface 503 are connected via the bus 504 and communicate with each other.
[0112] The communication interface 503 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.
[0113] Bus 504 includes hardware, software or both, and the parts of the training equipment of the photoresist model are coupled to each other. For example, but not limitation, the bus may include Accelerate Graphical Port (AGP) or other graphics bus, Enhanced Industry Standard Architecture (EISA) bus, Front Side Bus (FSB), Hyper Transport (HT) interconnection, Industry Standard Architecture (ISA) bus, InfiniBand Interconnection, Low Pin Count (LPC) bus, Memory Bus, Micro Channel Architecture (MCA) bus, Peripheral Component Interconnect (PCI) bus, PCI-Express (PCI-X) bus, Serial Advanced Technology Attachment (SATA) bus, Video Electronics Standards Association Local Bus (VLB) bus or other suitable bus or two or more of these combinations. Where appropriate, bus 504 may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.
[0114] In addition, in combination with the photoresist model training method in the above embodiment, the present application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the photoresist model training methods in the above embodiment is implemented.
[0115] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed, implements any one of the photoresist model training methods in the above embodiments.
[0116] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.
[0117] The functional blocks shown in the block diagram above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an ASIC, appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (Erasable ROM, EROM), floppy disks, compact disc read-only memory (Compact Disc Read-Only Memory, CD-ROM), optical discs, hard disks, optical fiber media, radio frequency (Radio Frequency, RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.
[0118] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.
[0119] The above reference is to the flowchart and / or block diagram of a training method, device, equipment, medium and product of a photoresist model according to an embodiment of the present disclosure. Various aspects of the present disclosure are described. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the functions / actions specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.
[0120] The above contents are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application.
Claims
1. A method for training a photoresist model, characterized in that: include: Acquire a plurality of graphic data groups consisting of mask patterns and corresponding photoresist patterns to obtain graphic data sets; Performing multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, so as to obtain multiple different selection results; The two parts of the graphic data groups in each of the selected results are respectively a graphic data training subset and a graphic data verification subset; For each of the selected results, after training an initial photoresist model based on the graphic data training subset, evaluating the trained initial photoresist model based on the graphic data verification subset to obtain an evaluation index; Based on the multiple evaluation indicators, the parameters of the initial photoresist model are adjusted to obtain a target photoresist model.
2. The method for training a photoresist model according to claim 1, characterized in that: The step of performing multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, to obtain multiple different selection results, includes: Dividing the graphic data set to obtain a plurality of mutually exclusive graphic data subsets; Each time, one of the graphic data subsets is selected from each of the graphic data subsets, and the selected graphic data subset is determined as the graphic data verification subset; The remaining subset of the graphic data is determined as the graphic data training subset.
3. The method for training a photoresist model according to claim 2, characterized in that: The step of dividing the graphic data set to obtain a plurality of mutually exclusive graphic data subsets includes: The graphic data set is divided based on the proportion of each category of graphics in the graphic data set to obtain a plurality of mutually exclusive graphic data subsets.
4. The method for training a photoresist model according to any one of claims 1 to 3, characterized in that: For each of the selected results, after training the initial photoresist model based on the graphic data training subset, the trained initial photoresist model is evaluated based on the graphic data verification subset to obtain an evaluation index, including: For each selection result, the following operations are performed: Training the initial photoresist model based on the graphic data training subset to obtain a first photoresist model; Inputting the mask pattern in the pattern data verification subset into the first photoresist model to obtain a simulation pattern; Determining the evaluation index of the first photoresist model based on the simulation pattern and the corresponding photoresist pattern; The adjusting the parameters of the initial photoresist model based on the plurality of evaluation indicators to obtain a target photoresist model includes: Determining target parameter values based on the evaluation indicators obtained from each evaluation; Based on the obtained target parameter value, the parameters of the initial photoresist model are adjusted.
5. The photoresist model training method according to claim 4, characterized in that: The step of training the initial photoresist model based on the graphic data training subset to obtain a first photoresist model comprises: Determine the weight corresponding to each category of graphics based on the proportion of each category of graphics in the graphic data set; the proportion of each category of graphics in the graphic data set is inversely correlated with the weight; The initial photoresist model is trained according to the weights corresponding to each category of graphics to obtain the first photoresist model.
6. The method for training a photoresist model according to claim 1, characterized in that: After adjusting the parameters of the initial photoresist model based on the plurality of evaluation indicators to obtain a target photoresist model, the method further includes: Verifying the performance of the target photoresist model to obtain a verification result; When the verification result meets the preset conditions, it is determined that the target photoresist model meets the actual requirements.
7. A photoresist model training device, characterized in that: include: An acquisition module, used for acquiring a plurality of graphic data groups consisting of mask patterns and corresponding photoresist patterns to obtain a graphic data set; A selection module, used for performing multiple selections in the graphic data set, each time selecting two parts of the graphic data set that are not completely the same, so as to obtain multiple different selection results; The two parts of the graphic data groups in each of the selected results are respectively a graphic data training subset and a graphic data verification subset; An evaluation module, for each of the selected results, after training an initial photoresist model based on the graphic data training subset, and then evaluating the trained initial photoresist model based on the graphic data verification subset to obtain an evaluation index; The adjustment module is used to adjust the parameters of the initial photoresist model based on the multiple evaluation indicators to obtain a target photoresist model.
8. A photoresist model training device, characterized in that: include: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the photoresist model training method as described in any one of claims 1 to 7 is implemented.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the photoresist model training method according to any one of claims 1 to 7 is implemented.
10. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the photoresist model training method as described in any one of claims 1 to 7.