interpolation_logs / qqp_interpol@15000steps /19568992_interpolate_models.out
connectivity's picture
Upload qqp_interpol@15000steps/19568992_interpolate_models.out
51c1d3f
raw
history blame
25.3 kB
Sun May 15 14:08:15 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.57.02 Driver Version: 470.57.02 CUDA Version: 11.4 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 Quadro RTX 8000 On | 00000000:86:00.0 Off | 0 |
| N/A 26C P8 13W / 250W | 0MiB / 45556MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
Obtaining file:///scratch/rb5139/constellations/src
Preparing metadata (setup.py): started
Preparing metadata (setup.py): finished with status 'done'
Installing collected packages: constellations
Attempting uninstall: constellations
Found existing installation: constellations 0.0.0
Uninstalling constellations-0.0.0:
Successfully uninstalled constellations-0.0.0
Running setup.py develop for constellations
Successfully installed constellations-0.0.0
Creating .16MB9L5NLTNNO4QPJACL, for loading in git data
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Creating .7PIFW8RDK8XC03JY1VH1, for loading in git data
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Euclidean distance between 95 and 96: 18.774574279785156.
Interpolating between 95 and 96 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2561 0.8945 0.8448
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2484 0.9004 0.8563
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2476 0.9121 0.8760
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2538 0.9062 0.8696
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2608 0.9062 0.8696
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2575 0.9062 0.8696
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2499 0.9023 0.8619
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2452 0.9004 0.8595
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2443 0.9023 0.8603
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2482 0.9062 0.8644
Completed interpolation from 95 to 96
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Creating .C0VPFTNWXB2T78080HD5, for loading in git data
Encountered Error: connectivity/bert_ft_qqp-97 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-97.pt
Euclidean distance between 95 and 97: 19.503650665283203.
Interpolating between 95 and 97 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2330 0.8984 0.8547
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2306 0.9102 0.8743
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2359 0.9082 0.8719
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2448 0.9062 0.8703
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2491 0.8984 0.8610
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2460 0.8965 0.8579
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2417 0.9004 0.8633
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2396 0.8984 0.8571
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2410 0.9004 0.8579
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2482 0.9062 0.8644
Completed interpolation from 95 to 97
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Creating .ZTHIZRW7BI40A12MP24L, for loading in git data
Encountered Error: connectivity/bert_ft_qqp-98 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-98.pt
Euclidean distance between 95 and 98: 19.072668075561523.
Interpolating between 95 and 98 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2532 0.8965 0.8464
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2395 0.9004 0.8547
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2359 0.9102 0.8722
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2402 0.9043 0.8672
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2444 0.9023 0.8656
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2421 0.9062 0.8703
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2389 0.9043 0.8658
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2372 0.8984 0.8564
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2394 0.9043 0.8627
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2482 0.9062 0.8644
Completed interpolation from 95 to 98
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Encountered Error: connectivity/bert_ft_qqp-95 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-95.pt
Creating .A6IVVJA8UD7RHGWVF3IE, for loading in git data
Encountered Error: connectivity/bert_ft_qqp-99 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-99.pt
Euclidean distance between 95 and 99: 20.792173385620117.
Interpolating between 95 and 99 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2399 0.8926 0.8424
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2247 0.8926 0.8451
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2326 0.9023 0.8641
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2543 0.9004 0.8668
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2682 0.8965 0.8651
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2583 0.8984 0.8660
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2403 0.9023 0.8691
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2309 0.8965 0.8556
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2327 0.9023 0.8611
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2482 0.9062 0.8644
Completed interpolation from 95 to 99
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Encountered Error: connectivity/bert_ft_qqp-97 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-97.pt
Euclidean distance between 96 and 97: 19.26505470275879.
Interpolating between 96 and 97 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2330 0.8984 0.8547
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2295 0.9102 0.8729
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2337 0.9062 0.8689
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2426 0.9043 0.8665
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2475 0.9043 0.8672
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2446 0.9043 0.8665
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2391 0.9043 0.8658
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2380 0.9043 0.8650
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2442 0.9023 0.8596
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2561 0.8945 0.8448
Completed interpolation from 96 to 97
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Encountered Error: connectivity/bert_ft_qqp-98 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-98.pt
Euclidean distance between 96 and 98: 19.161643981933594.
Interpolating between 96 and 98 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2532 0.8965 0.8464
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2464 0.8984 0.8523
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2456 0.9004 0.8555
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2521 0.9043 0.8635
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2606 0.9023 0.8611
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2616 0.9043 0.8635
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2544 0.9043 0.8635
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2474 0.9043 0.8627
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2483 0.8965 0.8499
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2561 0.8945 0.8448
Completed interpolation from 96 to 98
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Encountered Error: connectivity/bert_ft_qqp-96 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-96.pt
Encountered Error: connectivity/bert_ft_qqp-99 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-99.pt
Euclidean distance between 96 and 99: 20.5821475982666.
Interpolating between 96 and 99 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2399 0.8926 0.8424
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2254 0.8926 0.8451
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2359 0.9043 0.8679
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2682 0.8945 0.8608
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2982 0.8887 0.8564
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2908 0.8867 0.8550
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2603 0.8926 0.8586
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2385 0.9043 0.8686
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2380 0.9043 0.8650
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2561 0.8945 0.8448
Completed interpolation from 96 to 99
Encountered Error: connectivity/bert_ft_qqp-97 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-97.pt
Encountered Error: connectivity/bert_ft_qqp-97 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-97.pt
Encountered Error: connectivity/bert_ft_qqp-98 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-98.pt
Euclidean distance between 97 and 98: 19.476078033447266.
Interpolating between 97 and 98 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2532 0.8965 0.8464
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2392 0.8984 0.8531
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2354 0.9023 0.8596
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2404 0.9023 0.8634
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2495 0.9004 0.8625
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2523 0.9004 0.8618
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2444 0.8984 0.8587
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2343 0.9023 0.8634
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2296 0.9062 0.8674
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2330 0.8984 0.8547
Completed interpolation from 97 to 98
Encountered Error: connectivity/bert_ft_qqp-97 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-97.pt
Encountered Error: connectivity/bert_ft_qqp-97 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-97.pt
Encountered Error: connectivity/bert_ft_qqp-99 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-99.pt
Euclidean distance between 97 and 99: 19.843801498413086.
Interpolating between 97 and 99 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2399 0.8926 0.8424
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2275 0.8965 0.8499
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2277 0.9043 0.8643
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2368 0.8965 0.8548
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.2451 0.8945 0.8548
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.2459 0.9004 0.8633
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2372 0.9004 0.8625
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2284 0.9004 0.8618
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2260 0.9043 0.8658
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2330 0.8984 0.8547
Completed interpolation from 97 to 99
Encountered Error: connectivity/bert_ft_qqp-98 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-98.pt
Encountered Error: connectivity/bert_ft_qqp-98 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-98.pt
Encountered Error: connectivity/bert_ft_qqp-99 does not appear to have a file named flax_model.msgpack.
Trying to load model from connectivity/bert_ft_qqp-99.pt
Euclidean distance between 98 and 99: 20.736886978149414.
Interpolating between 98 and 99 on qqp
[0.0, 1.0]
point_num loss accuracy f1
----------- -------- ---------- --------
0 0.2399 0.8926 0.8424
[0.1111111111111111, 0.8888888888888888]
point_num loss accuracy f1
----------- -------- ---------- --------
1 0.2303 0.8926 0.8424
[0.2222222222222222, 0.7777777777777778]
point_num loss accuracy f1
----------- -------- ---------- --------
2 0.2456 0.8926 0.8433
[0.3333333333333333, 0.6666666666666667]
point_num loss accuracy f1
----------- -------- ---------- --------
3 0.2832 0.8965 0.8490
[0.4444444444444444, 0.5555555555555556]
point_num loss accuracy f1
----------- -------- ---------- --------
4 0.3186 0.8965 0.8507
[0.5555555555555556, 0.4444444444444444]
point_num loss accuracy f1
----------- -------- ---------- --------
5 0.3067 0.8945 0.8500
[0.6666666666666666, 0.33333333333333337]
point_num loss accuracy f1
----------- -------- ---------- --------
6 0.2635 0.8965 0.8532
[0.7777777777777777, 0.22222222222222232]
point_num loss accuracy f1
----------- -------- ---------- --------
7 0.2376 0.9004 0.8579
[0.8888888888888888, 0.11111111111111116]
point_num loss accuracy f1
----------- -------- ---------- --------
8 0.2363 0.8965 0.8499
[1.0, 0.0]
point_num loss accuracy f1
----------- -------- ---------- --------
9 0.2532 0.8965 0.8464
Completed interpolation from 98 to 99
Wrote the values to ../logs/NLI/qqp_berts/qqp_interpol@15000steps/interpol_qqp_validation19.pkl!