MCPcopy Create free account

hub / github.com/Rivendile/Muri / functions

Functions770 in github.com/Rivendile/Muri

⨍ Functions 770◇ Types & classes 115
↓ 3 callersMethod__init__
(self, memory_size, keys=None)
cluster_exp/workloads/models/deep_rl/component/replay.py:21
↓ 3 callersMethod__init__
(self, state_dim, hidden_units=(64, 64), gate=F.relu, noisy_linear=False)
cluster_exp/workloads/models/deep_rl/network/network_bodies.py:51
↓ 3 callersFunction_define_helper
Registers 'flag_name' with 'default_value' and 'docstring'.
cluster_exp/flags.py:48
↓ 3 callersFunction_define_helper
Registers 'flag_name' with 'default_value' and 'docstring'.
simulator/flags.py:48
↓ 3 callersMethod_sample
(self)
cluster_exp/workloads/models/deep_rl/agent/BaseAgent.py:135
↓ 3 callersMethodactor
(self, phi)
cluster_exp/workloads/models/deep_rl/network/network_heads.py:166
↓ 3 callersMethodadd_job
Add job (job_dict) into job_list
simulator/jobs.py:184
↓ 3 callersMethodadd_job_end_event
(self, job)
cluster_exp/jobs.py:561
↓ 3 callersMethodadd_job_end_event
(self, job)
simulator/jobs.py:569
↓ 3 callersMethodadd_job_gpu_util
(self, job)
cluster_exp/switch.py:240
↓ 3 callersMethodadd_job_gpu_util
(self, job)
simulator/switch.py:222
↓ 3 callersMethodaugment
(self, u, v)
simulator/matching.py:335
↓ 3 callersMethodcheckpoint
Record cluster, and job information, including: time queue length blocking index gpu util cpu util
cluster_exp/log.py:165
↓ 3 callersFunctionclose_obj
(obj)
cluster_exp/workloads/models/deep_rl/utils/misc.py:50
↓ 3 callersMethodcritic
(self, phi, a)
cluster_exp/workloads/models/deep_rl/network/network_heads.py:169
↓ 3 callersFunctiondraw3_new
(title:str, name_lists:list, lists:list, ncol=3)
simulator/draw_fig9-10.py:61
↓ 3 callersFunctiondraw4_new
(title:str, name_lists:list, lists:list, ncol=2)
simulator/draw_fig9-10.py:89
↓ 3 callersMethoddump_all_logs
(self)
simulator/log.py:112
↓ 3 callersMethodfull
(self)
cluster_exp/workloads/models/deep_rl/component/replay.py:145
↓ 3 callersMethodget_pr
(self, b, xr)
simulator/matching.py:288
↓ 3 callersFunctionget_scale_factors_array
(jobs)
cluster_exp/run.py:1094
↓ 3 callersFunctionget_time_str
()
cluster_exp/workloads/models/deep_rl/utils/misc.py:38
↓ 3 callersMethodlog_prob
(self, action)
cluster_exp/workloads/models/deep_rl/utils/torch_utils.py:95
↓ 3 callersMethodms_yarn_placement
MS_YARN, all gpus should come from the same switch
cluster_exp/cluster.py:432
↓ 3 callersMethodms_yarn_placement
MS_YARN, all gpus should come from the same switch
simulator/cluster.py:430
↓ 3 callersMethodq
(self, obs, a)
cluster_exp/workloads/models/deep_rl/network/network_heads.py:287
↓ 3 callersFunctionrange_tensor
(end)
cluster_exp/workloads/models/deep_rl/utils/torch_utils.py:28
↓ 3 callersMethodrelease_gpus
release using gpus back to free list
cluster_exp/node.py:87
↓ 3 callersMethodrelease_gpus
release using gpus back to free list
simulator/node.py:85
↓ 3 callersMethodset_network
(self, net)
cluster_exp/workloads/models/deep_rl/agent/BaseAgent.py:179
↓ 3 callersMethodset_slack
(self, x)
cluster_exp/matching.py:236
↓ 3 callersMethodset_slack
(self, x)
simulator/matching.py:253
↓ 3 callersMethodtransform_noise
(self, x)
cluster_exp/workloads/models/deep_rl/network/network_utils.py:82
↓ 3 callersMethodupdate
(self, idx, p)
cluster_exp/workloads/models/deep_rl/utils/sum_tree.py:54
↓ 3 callersMethodupdate_priorities
(self, info)
cluster_exp/workloads/models/deep_rl/component/replay.py:148
↓ 2 callersMethod__array__
(self, dtype=None)
cluster_exp/workloads/models/deep_rl/component/envs.py:103
↓ 2 callersMethod_parse_flags
(self, args=None)
cluster_exp/flags.py:21
↓ 2 callersMethod_parse_flags
(self, args=None)
simulator/flags.py:21
↓ 2 callersMethod_set_up
(self)
cluster_exp/workloads/models/deep_rl/agent/BaseAgent.py:168
↓ 2 callersMethodadd_job
Add job (job_dict) into job_list
cluster_exp/jobs.py:184
↓ 2 callersMethodadd_scalar
(self, tag, value, step=None, log_level=0)
cluster_exp/workloads/models/deep_rl/utils/logger.py:55
↓ 2 callersMethodalloc_cpus
If enough free cpus, allocate gpus Return: True, for success; False, for failure
cluster_exp/node.py:116
↓ 2 callersMethodalloc_cpus
If enough free cpus, allocate gpus Return: True, for success; False, for failure
simulator/node.py:114
↓ 2 callersMethodaugment
(self, u, v)
cluster_exp/matching.py:289
↓ 2 callersFunctioncal_shortest_expected_remaining
(job_data, a)
simulator/run_sim.py:217
↓ 2 callersMethodcalc_self_used_ratio
(self)
cluster_exp/matching.py:186
↓ 2 callersMethodcalc_self_used_ratio
(self)
simulator/matching.py:202
↓ 2 callersMethodclose
(self)
cluster_exp/workloads/models/deep_rl/agent/BaseAgent.py:22
↓ 2 callersMethodcompute_loss
(self, transitions)
cluster_exp/workloads/models/deep_rl/agent/DQN_agent.py:81
↓ 2 callersMethodconstruct_transition
(self, index)
cluster_exp/workloads/models/deep_rl/component/replay.py:112
↓ 2 callersMethodcreate_multi_nodes_placement
(self, job, switch_id, node_list)
cluster_exp/jobs.py:411
↓ 2 callersMethodcreate_multi_nodes_placement
(self, job, switch_id, node_list)
simulator/jobs.py:419
↓ 2 callersFunctiondraw1
(title:str, name_lists:list, lists:list, color:str, ncol=2)
simulator/draw_fig11-13.py:65
↓ 2 callersFunctiondraw3
(title:str, name_lists:list, lists:list, ncol=3)
simulator/draw_fig11-13.py:96
↓ 2 callersFunctiondraw4
(title:str, name_lists:list, lists:list, ncol=4)
simulator/draw_fig11-13.py:126
↓ 2 callersMethoddump_job_logs
(self)
cluster_exp/log.py:260
↓ 2 callersMethoddump_job_logs
(self)
simulator/log.py:331
↓ 2 callersFunctionepsilon_greedy
(epsilon, x)
cluster_exp/workloads/models/deep_rl/utils/torch_utils.py:51
↓ 2 callersFunctionget_base_constraints
(x, scale_factors_array)
cluster_exp/run.py:1111
↓ 2 callersMethodget_gpu_reservation
Cluster manager should decide (dynamically) the reserved gpus for each g_job object
cluster_exp/jobs.py:72
↓ 2 callersMethodget_gpu_reservation
Cluster manager should decide (dynamically) the reserved gpus for each g_job object
simulator/jobs.py:70
↓ 2 callersFunctionget_isolated_throughputs
(jobs)
cluster_exp/run.py:1100
↓ 2 callersMethodget_node_with_gid
(self, gid)
cluster_exp/cluster.py:865
↓ 2 callersMethodget_node_with_gid
(self, gid)
simulator/cluster.py:863
↓ 2 callersMethodget_pr
(self, b, xr)
cluster_exp/matching.py:260
↓ 2 callersFunctionget_scale_factors_array
(jobs)
simulator/run_sim.py:1799
↓ 2 callersMethodget_step
(self, tag)
cluster_exp/workloads/models/deep_rl/utils/logger.py:48
↓ 2 callersMethodlazy_init_writer
(self)
cluster_exp/workloads/models/deep_rl/utils/logger.py:39
↓ 2 callersMethodmerge
(self, config_dict=None)
cluster_exp/workloads/models/deep_rl/utils/config.py:84
↓ 2 callersMethodon_found_edge
(self, e)
cluster_exp/matching.py:410
↓ 2 callersMethodon_found_edge
(self, e)
simulator/matching.py:468
↓ 2 callersMethodplot_mean
(self, data, x=None, **kwargs)
cluster_exp/workloads/models/deep_rl/utils/plot.py:108
↓ 2 callersFunctionprint_fn
(log)
cluster_exp/utils.py:40
↓ 2 callersFunctionprint_fn
(log)
simulator/util.py:21
↓ 2 callersMethodquery_stats
(self, job_id_list)
cluster_exp/scheduler.py:70
↓ 2 callersMethodreduce_loss
(self, loss)
cluster_exp/workloads/models/deep_rl/agent/DQN_agent.py:78
↓ 2 callersMethodrun
(self)
cluster_exp/task.py:46
↓ 2 callersMethodrun
(self, run_jobs_dict, cluster_gpu, ordering=1)
simulator/matching.py:702
↓ 2 callersMethodset_match
(self, u, v)
cluster_exp/matching.py:275
↓ 2 callersMethodset_spec
(self, num_switch=0, num_node_p_switch=0, num_gpu_p_node=0, num_cpu_p_node=0, mem_p_node=0)
cluster_exp/cluster.py:56
↓ 2 callersMethodset_spec
(self, num_switch=0, num_node_p_switch=0, num_gpu_p_node=0, num_cpu_p_node=0, mem_p_node=0)
simulator/cluster.py:56
↓ 2 callersMethodset_st
(self, x, b)
cluster_exp/matching.py:251
↓ 2 callersMethodset_st
(self, x, b)
simulator/matching.py:279
↓ 2 callersFunctionshortest_first_sim_jobs
new jobs are added to the end of the ending queue but in the queue, shortest (gpu) job first be served, until no resource place=True: pla
cluster_exp/run.py:226
↓ 2 callersFunctionshortest_first_sim_jobs
new jobs are added to the end of the ending queue but in the queue, shortest (gpu) job first be served, until no resource
simulator/run_sim.py:227
↓ 2 callersMethodstep
(self)
cluster_exp/workloads/models/deep_rl/agent/BaseAgent.py:171
↓ 2 callersMethodto_numpy
(self, v)
cluster_exp/workloads/models/deep_rl/utils/logger.py:43
↓ 2 callersMethodtotal
(self)
cluster_exp/workloads/models/deep_rl/utils/sum_tree.py:35
↓ 2 callersMethodtry_cross_node_alloc
used in MS_YARN placement try get gpus from multiple nodes [need_gpu / gpu_p_node] nodes, and one node with [need_gpu % g
cluster_exp/switch.py:47
↓ 2 callersMethodtry_cross_node_alloc
used in MS_YARN placement try get gpus from multiple nodes [need_gpu / gpu_p_node] nodes, and one node with [need_gpu % g
simulator/switch.py:47
↓ 2 callersMethodtry_single_node_alloc
used in MS_YARN placement try get gpus from a single node if can't find a node, give up, and return False
cluster_exp/switch.py:164
↓ 2 callersMethodtry_single_node_alloc
used in MS_YARN placement try get gpus from a single node if can't find a node, give up, and return False
simulator/switch.py:152
↓ 2 callersMethodupdate_slack
(self, u, x)
cluster_exp/matching.py:231
↓ 2 callersMethodupdate_slack
(self, u, x)
simulator/matching.py:248
↓ 1 callersMethodDone
(self, request, context)
cluster_exp/runtime/rpc/master_server.py:33
↓ 1 callersMethodExecute
(self, request, context)
cluster_exp/runtime/rpc/worker_server.py:20
↓ 1 callersMethodExitCommand
(self, request, context)
cluster_exp/runtime/rpc/worker_server.py:41
↓ 1 callersMethodGetUtil
(self, request, context)
cluster_exp/runtime/rpc/worker_server.py:50
↓ 1 callersMethodKill
(self, request, context)
cluster_exp/runtime/rpc/worker_server.py:31
↓ 1 callersMethodQueryStats
(self, request, context)
cluster_exp/runtime/rpc/trainer_server.py:20
← previousnext →101–200 of 770, ranked by callers