Error while dp train: OSError: liquidData/type.raw not found.

[wjinjin]

while I use dp train input.json command, it shows OSError: liquidData/type.raw not found. But there is a type.raw in liquidData file. What could be wrong?

[njzjz]

Is liquidData in your current directory where you run dp train? Can you run ls liquidData/type.raw in that directory?

[wjinjin]

yes, I can. Now I reduce the data number in liquiddata, the error changes into "IndexError: list index out of range". Does this mean there are to much data in liquiddat?

[njzjz]

Something went wrong. Can you provide more information about your input.json and your data?
BTW, does the example work well in your machine?

[wjinjin]

Now I reduce the data number in liquiddata, it's working.
The input.json is shown as below.
"_comment": " model parameters",
"model": {
"descriptor" :{
"type": "se_a",
"sel": [30],
"rcut_smth": 2.50,
"rcut": 3.00,
"neuron": [50, 100],
"resnet_dt": false,
"axis_neuron": 16,
"seed": 1,
"_comment": " that's all"
},
"fitting_net" : {
"neuron": [240, 240, 240],
"resnet_dt": true,
"seed": 1,
"_comment": " that's all"
},
"_comment": " that's all"
},

"learning_rate" :{
    "type":         "exp",
    "start_lr":     0.001,
    "decay_steps":  15000,
    "decay_rate":   0.95,
    "_comment":     "that's all"
},

"loss" :{
    "start_pref_e": 0.02,
    "limit_pref_e": 1,
    "start_pref_f": 1000,
    "limit_pref_f": 1,

"start_pref_v": 0,
"limit_pref_v": 0,
"_comment": " that's all"
},

"_comment": " traing controls",
"training" : {
    "systems":      [
    "liqiudData"
    ],
    "set_prefix":   "set",
    "stop_batch":   3000000,
    "batch_size":   1,
    "seed":         1,
    "_comment": " display and restart",
    "_comment": " frequencies counted in batch",
    "disp_file":    "lcurve.out",
    "disp_freq":    100,
    "numb_test":    10,
    "save_freq":    1000,
    "save_ckpt":    "model.ckpt",
    "load_ckpt":    "model.ckpt",
    "disp_training":true,
    "time_training":true,

"profiling": false,
"profiling_file":"timeline.json",
"_comment": "that's all"
},

"_comment":         "that's all"

}

[jameswind]

What was the original number of data and what's the current number you use? I still feel it's a bit strange that this caused the problem.

…

On Mon, Dec 16, 2019 at 12:27 PM wjinjin ***@***.***> wrote: Now I reduce the data number in liquiddata, it's working. The input.json is shown as below. "_comment": " model parameters", "model": { "descriptor" :{ "type": "se_a", "sel": [30], "rcut_smth": 2.50, "rcut": 3.00, "neuron": [50, 100], "resnet_dt": false, "axis_neuron": 16, "seed": 1, "_comment": " that's all" }, "fitting_net" : { "neuron": [240, 240, 240], "resnet_dt": true, "seed": 1, "_comment": " that's all" }, "_comment": " that's all" }, "learning_rate" :{ "type": "exp", "start_lr": 0.001, "decay_steps": 15000, "decay_rate": 0.95, "_comment": "that's all" }, "loss" :{ "start_pref_e": 0.02, "limit_pref_e": 1, "start_pref_f": 1000, "limit_pref_f": 1, "start_pref_v": 0, "limit_pref_v": 0, "_comment": " that's all" }, "_comment": " traing controls", "training" : { "systems": [ "liqiudData" ], "set_prefix": "set", "stop_batch": 3000000, "batch_size": 1, "seed": 1, "_comment": " display and restart", "_comment": " frequencies counted in batch", "disp_file": "lcurve.out", "disp_freq": 100, "numb_test": 10, "save_freq": 1000, "save_ckpt": "model.ckpt", "load_ckpt": "model.ckpt", "disp_training":true, "time_training":true, "profiling": false, "profiling_file":"timeline.json", "_comment": "that's all" }, "_comment": "that's all" } — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <https://github.com/deepmodeling/deepmd-kit/issues/138?email_source=notifications&email_token=AEJ6DCYUO7565FWDKDBWJD3QY627TA5CNFSM4J2FF3AKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEG7OWEA#issuecomment-566160144>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEJ6DC6FRZIPTPVNGV5SJ7LQY627TANCNFSM4J2FF3AA> .