I am doing calaculations using Cr-Ni using ESPEI. I am using the TDB file from thermocalc and it have 120 ZPF files . It is taking more than 2 day sto complete the optimization and after optimization I am not getting proper phase diagram . How can I accelerate the process ?
One more question is that whether we can include activity data in ESPEI? Some phases are not appearing in the phase diagram after mcmc . Why ?
I will attach here my mcmc_settings.yaml.
system:
phase_models: ../phase_models.json
datasets: ../input-data
tags:
dft:
excluded_model_contributions: ['idmix', 'mag']
estimated-entropy:
excluded_model_contributions: ['idmix', 'mag']
output:
output_db: Cr-Ni-mcmc.tdb
verbosity: 2
tracefile: null
probfile: null
logfile: null
mcmc:
iterations: 2000 # Increased from 2
save_interval: 20 # Increased from 1
scheduler: 'dask'
cores: 4
input_db: Cr-Ni.TDB
chains_per_parameter: 4
prior:
name: normal
loc_relative: 1.0
scale_relative: 0.5 # Reduced for stability
I am doing calaculations using Cr-Ni using ESPEI. I am using the TDB file from thermocalc and it have 120 ZPF files . It is taking more than 2 day sto complete the optimization and after optimization I am not getting proper phase diagram . How can I accelerate the process ?
One more question is that whether we can include activity data in ESPEI? Some phases are not appearing in the phase diagram after mcmc . Why ?
I will attach here my mcmc_settings.yaml.
system:
phase_models: ../phase_models.json
datasets: ../input-data
tags:
dft:
excluded_model_contributions: ['idmix', 'mag']
estimated-entropy:
excluded_model_contributions: ['idmix', 'mag']
output:
output_db: Cr-Ni-mcmc.tdb
verbosity: 2
tracefile: null
probfile: null
logfile: null
mcmc:
iterations: 2000 # Increased from 2
save_interval: 20 # Increased from 1
scheduler: 'dask'
cores: 4
input_db: Cr-Ni.TDB
chains_per_parameter: 4
prior:
name: normal
loc_relative: 1.0
scale_relative: 0.5 # Reduced for stability