software:singularity:caviness

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
software:singularity:caviness [2018-08-10 23:23] – [Inside your interactive session or your batch job you must first issue the command] anitasoftware:singularity:caviness [2021-05-06 18:45] (current) – [Troubleshooting] anita
Line 73: Line 73:
 Singularity tensorflow.simg:~> grep Cpus_allowed_list /proc/$$/status Singularity tensorflow.simg:~> grep Cpus_allowed_list /proc/$$/status
 Cpus_allowed_list: 0-3 Cpus_allowed_list: 0-3
-Singularity tensorflow.simg:~>exit+Singularity tensorflow.simg:~> exit
 exit exit
 [traine@r00n45 ~]$ [traine@r00n45 ~]$
Line 82: Line 82:
 === Locale error === === Locale error ===
  
-If you experience the error below after starting a particular container, then it is likely due your current locale not being supported in this particular container's shell.  +If you experience the error below after starting a particular shell container, then it is likely due to your current locale not being supported in this particular container's shell.  
  
 <code bash> <code bash>
 +[traine@r00n45 ~]$ vpkg_require singularity
 +Adding package `singularity/2.5.1` to your environment
 [traine@r00n45 ~]$ singularity shell tensorflow.simg [traine@r00n45 ~]$ singularity shell tensorflow.simg
 Singularity: Invoking an interactive shell within container... Singularity: Invoking an interactive shell within container...
Line 115: Line 117:
 </code> </code>
  
-Of course in this particular case, checking the setting of account ''traine'' it is set to ''en_US.UTF-8'' which is not supported in this shell.  To fix the situation, simply set the ''LANG'' locale to one that is supported before starting the container shell.+Of course in this particular case, the locale setting for account ''traine'' is defined as ''en_US.UTF-8'' which is not supported in this shell.  To fix the situation, simply set the ''LANG'' locale to one that is supported before starting the container shell.
  
 <code bash> <code bash>
 +[traine@r00n45 ~]$ vpkg_require singularity
 +Adding package `singularity/2.5.1` to your environment
 [tainne@r00n45 ~]$ LANG=C.UTF-8 singularity shell tensorflow.simg [tainne@r00n45 ~]$ LANG=C.UTF-8 singularity shell tensorflow.simg
 Singularity: Invoking an interactive shell within container... Singularity: Invoking an interactive shell within container...
Line 135: Line 139:
 your TensorFlow runs and graphs. https://github.com/tensorflow/tensorboard your TensorFlow runs and graphs. https://github.com/tensorflow/tensorboard
  
-optional arguments: +[[ lots of additional information displayed for optional arguments
-  -h, --help            show this help message and exit +
-  --logdir PATH         Directory where TensorBoard will look to find +
-                        TensorFlow event files that it can display. +
-                        TensorBoard will recursively walk the directory +
-                        structure rooted at logdir, looking for .*tfevents.* +
-                        files. You may also pass a comma separated list of log +
-                        directories, and TensorBoard will watch each +
-                        directory. You can also assign names to individual log +
-                        directories by putting a colon between the name and +
-                        the path, as in: `tensorboard +
-                        --logdir=name1:/path/to/logs/1,name2:/path/to/logs/2` +
-  --host ADDR           What host to listen to. Defaults to serving on all +
-                        interfaces. Other commonly used values are 127.0.0.1 +
-                        (localhost) and :: (for IPv6). +
-  --port PORT           Port to serve TensorBoard on (default: 6006) +
-  --purge_orphaned_data BOOL +
-                        Whether to purge data that may have been orphaned due +
-                        to TensorBoard restarts. Setting +
-                        --purge_orphaned_data=False can be used to debug data +
-                        disappearance. (default: True) +
-  --reload_interval SECONDS +
-                        How often the backend should load more data, in +
-                        seconds. Set to 0 to load just once at startup and a +
-                        negative number to never reload at all. (default: 5.0) +
-  --db URI              [experimental] sets SQL database URI +
-  --inspect             Prints digests of event files to command line. This is +
-                        useful when no data is shown on TensorBoard, or the +
-                        data shown looks weird. Example usage: `tensorboard +
-                        --inspect --logdir mylogdir --tag loss` See +
-                        tensorflow/python/summary/event_file_inspector.py for +
-                        more info. +
-  --tag TAG             tag to query for; used with --inspect +
-  --event_file PATH     The particular event file to query for. Only used if +
-                        --inspect is present and --logdir is not specified. +
-  --path_prefix PATH    An optional, relative prefix to the path, e.g. +
-                        "/path/to/tensorboard". resulting in the new base url +
-                        being located at localhost:6006/path/to/tensorboard +
-                        under default settings. A leading slash is required +
-                        when specifying the path_prefix, however trailing +
-                        slashes can be omitted. The path_prefix can be +
-                        leveraged for path based routing of an elb when the +
-                        website base_url is not available e.g. +
-                        "example.site.com/path/to/tensorboard/". +
-  --window_title TEXT   changes title of browser window +
-  --max_reload_threads COUNT +
-                        The max number of threads that TensorBoard can use to +
-                        reload runs. Not relevant for db mode. Each thread +
-                        reloads one run at a time. +
-  --samples_per_plugin SAMPLES_PER_PLUGIN +
-                        An optional comma separated list of +
-                        plugin_name=num_samples pairs to explicitly specify +
-                        how many samples to keep per tag for that plugin. For +
-                        unspecified plugins, TensorBoard randomly downsamples +
-                        logged summaries to reasonable values to prevent out- +
-                        of-memory errors for long running jobs. This flag +
-                        allows fine control over that downsampling. Note that +
-                        0 means keep all samples of that type. For instance +
-                        "scalars=500,images=0" keeps 500 scalars and all +
-                        images. Most users should not need to set this flag. +
- +
-profile plugin: +
-  --master_tpu_unsecure_channel ADDR +
-                        IP address of "master tpu", used for getting streaming +
-                        trace data through tpu profiler analysis grpc. The +
-                        grpc channel is not secured. +
- +
-debugger plugin: +
-  --debugger_data_server_grpc_port PORT +
-                        The port at which the non-interactive debugger data +
-                        server should receive debugging data via gRPC from one +
-                        or more debugger-enabled TensorFlow runtimes. No +
-                        debugger plugin or debugger data server will be +
-                        started if this flag is not provided. This flag +
-                        differs from the `--debugger_port` flag in that it +
-                        starts a non-interactive mode. It is for use with the +
-                        "health pills" feature of the Graph Dashboard. This +
-                        flag is mutually exclusive with `--debugger_port`. +
-  --debugger_port PORT  The port at which the interactive debugger data server +
-                        (to be started by the debugger plugin) should receive +
-                        debugging data via gRPC from one or more debugger- +
-                        enabled TensorFlow runtimes. No debugger plugin or +
-                        debugger data server will be started if this flag is +
-                        not provided. This flag differs from the +
-                        `--debugger_data_server_grpc_port` flag in that it +
-                        starts an interactive mode that allows user to pause +
-                        at selected nodes inside a TensorFlow Graph or between +
-                        Session.runs. It is for use with the interactive +
-                        Debugger Dashboard. This flag is mutually exclusive +
-                        with `--debugger_data_server_grpc_port`. +
-Singularity tensorflow.simg:~>+
  
 +Singularity tensorflow.simg:~> exit
 +exit
 +[traine@r00n45 ~]$
 </code> </code>
- 
  
 ==== Common Singularity commands ==== ==== Common Singularity commands ====
  • software/singularity/caviness.1533957830.txt.gz
  • Last modified: 2018-08-10 23:23
  • by anita