Merge pull request #10 from pxlxingliang/develop

Develop

Author: pxlxingliang

src/abacustest.py

@@ -14,7 +14,7 @@
 def AbacusTestArgs(parser):
     parser.description = "This script is used to run a testing"
     parser.add_argument('-p', '--param', type=str, default="job.json",help='the job setting file, default is job.json')
-    parser.add_argument('-s', '--save', type=str, default=None,help='the folder where the results will be put in, default: result/date_of_today (e.g. result/20230101)')
+    parser.add_argument('-s', '--save', type=str, default=None,help='the folder where the results will be put in, default: result')
     parser.add_argument('--override', type=int, default=1,help="when the save folder exists, if override it. 0: no, 1: yes. ")
     parser.add_argument('--outinfo', type=int, default=1,help='if output detail informations, 0: no, 1: yes')
     parser.add_argument('--debug',nargs='?',type=int, const=1, default=0,help="debug mode for dflow")

src/launching/comm_class.py

@@ -1,6 +1,4 @@
-from dp.launching.cli import to_runner,SubParser,run_sp_and_exit
-from dp.launching.typing.basic import BaseModel, Int, String, Float,List,Optional,Union,Dict
-from dp.launching.cli import to_runner, default_minimal_exception_handler
+from dp.launching.typing.basic import BaseModel, Int, String, Float, List, Optional, Union, Dict
 from dp.launching.typing import InputFilePath, OutputDirectory
 from dp.launching.typing import (
     BaseModel,
@@ -22,10 +20,12 @@
     BenchmarkLabels,
     BenchmarkTags
 )
-from enum import Enum
-from typing import Literal
+
 import re
 
+class OutputSet(BaseModel):
+    IO_output_path: OutputDirectory = Field(default="./output")
+
 class ConfigSet(BaseModel):
     #Bohrium config
     Config_lbg_username:   BohriumUsername
@@ -36,102 +36,66 @@ class ConfigSet(BaseModel):
     Config_config_host: DflowArgoAPIServer
     Config_s3_config_endpoint: DflowStorageEndpoint
     Config_config_k8s_api_server: DflowK8sAPIServer
-    Config_config_token: DflowAccessToken 
+    Config_config_token: DflowAccessToken
 
     Config_dflow_labels: BenchmarkLabels
 
-class AbacusMetricEnum(String,Enum):
-    AbacusMetric_metric1 = 'version'
-    AbacusMetric_metric2 = 'ncore'
-    AbacusMetric_metric3 = 'normal_end'
-    AbacusMetric_metric4 = 'INPUT:ks_solver'
-#    AbacusMetric_metric5 = 'kpt'
-    AbacusMetric_metric6 = 'nbands'
-    AbacusMetric_metric7 = 'converge'
-    AbacusMetric_metric8 = 'total_mag'
-    AbacusMetric_metric9 = 'absolute_mag'
-    AbacusMetric_metric10 = 'nkstot'
-    AbacusMetric_metric11 = 'ibzk'
-    AbacusMetric_metric12 = 'natom'
-    AbacusMetric_metric13 = 'nelec'
-    AbacusMetric_metric14 = 'energy'
-    AbacusMetric_metric15 = 'volume'
-#    AbacusMetric_metric16 = 'fft_grid'
-    AbacusMetric_metric17 = 'efermi'
-    AbacusMetric_metric18 = 'energy_per_atom'
-#    AbacusMetric_metric19 = 'stress'
-#    AbacusMetric_metric20 = 'force'
-    AbacusMetric_metric21 = 'band_gap'
-    AbacusMetric_metric22 = 'total_time'
-    AbacusMetric_metric23 = 'stress_time'
-    AbacusMetric_metric24 = 'force_time'
-    AbacusMetric_metric25 = 'scf_time'
-    AbacusMetric_metric26 = 'scf_time_each_step'
-    AbacusMetric_metric27 = 'step1_time'
-    AbacusMetric_metric28 = 'scf_steps'
-    AbacusMetric_metric29 = 'atom_mag'
-#    AbacusMetric_metric30 = 'drho'
-    AbacusMetric_metric31 = 'lattice_constant'
-    AbacusMetric_metric32 = 'cell'
-#    AbacusMetric_metric33 = 'coordinate'
-    AbacusMetric_metric34 = 'element_list'
-    AbacusMetric_metric35 = 'atomlabel_list'
-#    AbacusMetric_metric36 = 'delta_energy'
-#    AbacusMetric_metric37 = 'delta_energyPerAtom'
-    AbacusMetric_metric38 = 'relax_converge'
-    AbacusMetric_metric39 = 'relax_steps'
-
-
-class SuperMetricMethodEnum(String,Enum):
-    SuperMetricMethod_method1 = "iGM"
-    SuperMetricMethod_method2 = "GM"
-    SuperMetricMethod_method3 = "TrueRatio"
-    SuperMetricMethod_method4 = "MEAN"
-
-'''
-class RunSetUpload(BaseModel):
-    uploaddatahub_project: String =  Field(default="",description="Please set the project where dataset will be uploaded to. If not set, will not upload")
-    uploaddatahub_datasetname: String =  Field(default="",description="Please set the path in project where dataset will be uploaded to. If not set, will not upload")
-    uploaddatahub_properties: Dict[String,String] =  Field(default={},description="")
-    uploaddatahub_tags: List[String] =  Field(default=[],description="")
-    uploadtracking_test_name: String =  Field(default="",description="Please set the name of test. If not set, will not upload metrics to tracking")
-    uploadtracking_experiment_name: String =  Field(default="",description="Please set the name of experiment. If not set, will not upload metrics to tracking")
-    uploadtracking_tags: List[String] =  Field(default=[],description="")
-'''
+
+class NgroupSet(BaseModel):
+    ngroup: Int = Field(
+        default=0, description="Number of groups to run in parallel. If not set, all examples will be run in parallel.", ge=0)
+
+
+class RundftCommandSet(BaseModel):
+    rundft_command: String = Field(default="OMP_NUM_THREADS=1 mpirun -np 16 abacus > log",
+                                   description="Command to run each example. Please note that the program will first enter each folder before executing this command. \
+During runtime, there will be no interaction between different examples",)
+
+
+class PostdftCommandSet(BaseModel):
+    postdft_command: String = Field(default="",
+                                    description="If you need to execute some custom scripts or commands, please enter the bash command here. \
+Usually used to generate some custom metrics. At this step, the program will first collect the results of all examples in rundft, and then execute the command. \
+The working directory is the same level directory as the outer layer of all examples.",)
+
 
 class TrackingSet(BaseModel):
-    Tracking_metrics: Boolean = Field(default=False,description="If tracking, will display historical metrics values based on test and experience")
-    Tracking_token: String = Field(default=None,description="If want to track metrics, please enter your token to access AIM")
+    Tracking_metrics: Boolean = Field(
+        default=False, description="If tracking, will display historical metrics values based on test and experience")
+    Tracking_token: String = Field(
+        default=None, description="If want to track metrics, please enter your token to access AIM")
     Tracking_tags: BenchmarkTags
     #tags = [f"benchmark-application-{application.name}", f"benchmark-version-{job.version}", f"benchmark-schedule-{job.properties.get('source_name', 'none')}",
     #        f"benchmark-job-{job.name}"]
 
     @classmethod
-    def parse_obj(cls,opts):
-        if opts.tracking_metrics and opts.tracking_token != None and opts.tracking_token.strip() != "":
-            default_tags = opts.default_tags 
-            schedule = "_".join(re.split("-",default_tags[2],2)[-1].strip().split())
-            application_name = re.split("-",default_tags[0],2)[-1]
-            job_name = re.split("-",default_tags[3],2)[-1]
-                
+    def parse_obj(cls, opts):
+        if opts.Tracking_metrics and opts.Tracking_token != None and opts.Tracking_token.strip() != "":
+            default_tags = opts.Tracking_tags
+            schedule = "_".join(
+                re.split("-", default_tags[2], 2)[-1].strip().split())
+            application_name = re.split("-", default_tags[0], 2)[-1]
+            job_name = re.split("-", default_tags[3], 2)[-1]
+
             return {
                 "name": schedule + "." + job_name,
                 "experiment": application_name + "/benchmark",
                 "tags": default_tags,
-                "token": opts.tracking.AIM_ACCESS_TOKEN
+                "token": opts.Tracking_token
             }
         else:
             return None
-    
+
 
 class myLog:
     def __init__(self):
         self.logs = ""
-    
-    def iprint(self,mess,*args):
+
+    def iprint(self, mess, *args):
         allmess = " ".join([str(mess)]+[str(i) for i in args])
         print(allmess)
         self.logs += allmess + "\n"
-    
-    def write(self,filename):
-        with open(filename,'w') as f1: f1.write(self.logs)
+
+    def write(self, filename):
+        with open(filename, 'w') as f1:
+            f1.write(self.logs)