ur-whitelab / hoomd-tf

A plugin that allows the use of Tensorflow in Hoomd-Blue for GPU-accelerated ML+MD
https://hoomd-tf.readthedocs.io
MIT License
30 stars 8 forks source link

Adding Jupyter notebooks to CI #276

Closed hgandhi2411 closed 3 years ago

hgandhi2411 commented 3 years ago
whitead commented 3 years ago

@hgandhi2411 this is really great work, nice job Heta! Is there anyway though you can remove the notebook output from the log file? It makes the log files hard to read.

hgandhi2411 commented 3 years ago

@hgandhi2411 this is really great work, nice job Heta! Is there anyway though you can remove the notebook output from the log file? It makes the log files hard to read.

Thanks @whitead. I can look into clearing out the notebook output.

hgandhi2411 commented 3 years ago

@whitead In the 08. N-Body Neural Force Fields.ipynb notebook, the following error shows up occasionally and disappears on re-running the notebook. Do we need to worry about it?

########### Hoomd-Sim Code ################
hoomd.context.initialize('--mode=cpu')

tfcompute = htf.tfcompute(model)

# create a square lattice
system = hoomd.init.create_lattice(unitcell=hoomd.lattice.sq(a=1.2),
                                    n=[16,16])
nlist = hoomd.md.nlist.cell()
# NVT ensemble
hoomd.md.integrate.mode_standard(dt=0.001)
hoomd.md.integrate.nvt(group=hoomd.group.all(), kT=0.1, tau=0.5).randomize_velocities(seed=1)
tfcompute.attach(nlist, r_cut=5, save_output_period=5)
#run with our silly model
hoomd.run(1e3)
------------------

---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
<ipython-input-1-127d1819638d> in <module>
     13 tfcompute.attach(nlist, r_cut=5, save_output_period=5)
     14 #run with our silly model
---> 15 hoomd.run(1e3)

/usr/share/miniconda/lib/python3.7/site-packages/hoomd/__init__.py in run(tsteps, profile, limit_hours, limit_multiple, callback_period, callback, quiet)
    199     if not quiet:
    200         context.msg.notice(1, "** starting run **\n");
--> 201     context.current.system.run(int(tsteps), callback_period, callback, limit_hours, int(limit_multiple));
    202     if not quiet:
    203         context.msg.notice(1, "** run complete **\n");

RuntimeError: Error computing cell list
RuntimeError: Error computing cell list

Error: Process completed with exit code 1.
hgandhi2411 commented 3 years ago

@whitead Now only errors are logged when notebooks are run. Let me know if you'd like any other changes.

whitead commented 3 years ago

@hgandhi2411 that is bizarre, especially since it uses a random number seed. Can you try adjusting the check period nlist.cell(check_period=1)?

whitead commented 3 years ago

@hgandhi2411 looks like there are some build errors still

hgandhi2411 commented 3 years ago

@whitead Everything works now!