thkim0305 commited on
Commit
aab31d5
·
verified ·
1 Parent(s): 9ef01e9

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round10.pth +3 -0
  2. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round12.pth +3 -0
  3. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round15.pth +3 -0
  4. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round17.pth +3 -0
  5. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round2.pth +3 -0
  6. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round20.pth +3 -0
  7. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round5.pth +3 -0
  8. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round7.pth +3 -0
  9. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_trainer_state.json +392 -0
  10. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round10.pth +3 -0
  11. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round12.pth +3 -0
  12. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round15.pth +3 -0
  13. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round17.pth +3 -0
  14. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round2.pth +3 -0
  15. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round20.pth +3 -0
  16. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round5.pth +3 -0
  17. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round7.pth +3 -0
  18. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_trainer_state.json +392 -0
  19. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round10.pth +3 -0
  20. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round12.pth +3 -0
  21. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round15.pth +3 -0
  22. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round17.pth +3 -0
  23. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round2.pth +3 -0
  24. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round20.pth +3 -0
  25. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round5.pth +3 -0
  26. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round7.pth +3 -0
  27. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_trainer_state.json +392 -0
  28. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round10.pth +3 -0
  29. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round12.pth +3 -0
  30. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round15.pth +3 -0
  31. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round17.pth +3 -0
  32. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round2.pth +3 -0
  33. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round20.pth +3 -0
  34. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round5.pth +3 -0
  35. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round7.pth +3 -0
  36. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_trainer_state.json +392 -0
  37. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round10.pth +3 -0
  38. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round12.pth +3 -0
  39. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round15.pth +3 -0
  40. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round17.pth +3 -0
  41. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round2.pth +3 -0
  42. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round20.pth +3 -0
  43. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round5.pth +3 -0
  44. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round7.pth +3 -0
  45. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_trainer_state.json +392 -0
  46. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round10.pth +3 -0
  47. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round12.pth +3 -0
  48. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round15.pth +3 -0
  49. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round17.pth +3 -0
  50. client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round2.pth +3 -0
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d0b98006a417849b6ac661229656393295a04d6f91d093dc59d2bdbb8ae617e
3
+ size 369838470
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round12.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8c03e1e43a63e7117790b6e2d671a5c5f4ccf8ae68353299ac809726946cc9ae
3
+ size 369838470
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round15.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec2fb7a6a5b938263cbd5ec10b400d715ab54d8b541a0ebe9fa494159c93c11e
3
+ size 369838470
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:445ebe780d6772c72be140743bb599ca60f379f3d29851b32505236a00e16327
3
+ size 369838470
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70a513b49a7974e643c1786ce1c86df0cfedbf34bfcb22a416161aeb36d235fc
3
+ size 369837282
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round20.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8788ca099c4405b8b862ba82f0ed3ad0055f2f2f309e7dddea265254122dfc88
3
+ size 369838470
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32d042e35bfb519e597c834a17e98803e3110f1ab1426918ccb732a2ef8ecf90
3
+ size 369837282
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_client_model_round7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4fd2af4f98756299ff860b2a7a1aff6992897784f0032707499d09ce69709d1
3
+ size 369837282
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/0_trainer_state.json ADDED
@@ -0,0 +1,392 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 2.673671245574951,
14
+ "learning_rate": 2e-05,
15
+ "loss": 1.0677,
16
+ "step": 2
17
+ },
18
+ {
19
+ "epoch": 0.04,
20
+ "grad_norm": 2.416391372680664,
21
+ "learning_rate": 2e-05,
22
+ "loss": 1.0969,
23
+ "step": 4
24
+ },
25
+ {
26
+ "epoch": 0.06,
27
+ "grad_norm": 4.597806453704834,
28
+ "learning_rate": 2e-05,
29
+ "loss": 0.9624,
30
+ "step": 6
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "grad_norm": 2.4499340057373047,
35
+ "learning_rate": 2e-05,
36
+ "loss": 1.9471,
37
+ "step": 8
38
+ },
39
+ {
40
+ "epoch": 0.1,
41
+ "grad_norm": 0.8575390577316284,
42
+ "learning_rate": 2e-05,
43
+ "loss": 0.3392,
44
+ "step": 10
45
+ },
46
+ {
47
+ "epoch": 0.12,
48
+ "grad_norm": 5.4067840576171875,
49
+ "learning_rate": 2e-05,
50
+ "loss": 0.5852,
51
+ "step": 12
52
+ },
53
+ {
54
+ "epoch": 0.14,
55
+ "grad_norm": 4.374941825866699,
56
+ "learning_rate": 2e-05,
57
+ "loss": 1.569,
58
+ "step": 14
59
+ },
60
+ {
61
+ "epoch": 0.16,
62
+ "grad_norm": 5.50150728225708,
63
+ "learning_rate": 2e-05,
64
+ "loss": 0.9978,
65
+ "step": 16
66
+ },
67
+ {
68
+ "epoch": 0.18,
69
+ "grad_norm": 3.8701274394989014,
70
+ "learning_rate": 2e-05,
71
+ "loss": 0.781,
72
+ "step": 18
73
+ },
74
+ {
75
+ "epoch": 0.2,
76
+ "grad_norm": 0.7540118098258972,
77
+ "learning_rate": 2e-05,
78
+ "loss": 0.7018,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.22,
83
+ "grad_norm": 4.309854984283447,
84
+ "learning_rate": 2e-05,
85
+ "loss": 0.8156,
86
+ "step": 22
87
+ },
88
+ {
89
+ "epoch": 0.24,
90
+ "grad_norm": 1.7167375087738037,
91
+ "learning_rate": 2e-05,
92
+ "loss": 0.5814,
93
+ "step": 24
94
+ },
95
+ {
96
+ "epoch": 0.26,
97
+ "grad_norm": 2.9055559635162354,
98
+ "learning_rate": 2e-05,
99
+ "loss": 1.0377,
100
+ "step": 26
101
+ },
102
+ {
103
+ "epoch": 0.28,
104
+ "grad_norm": 2.938593864440918,
105
+ "learning_rate": 2e-05,
106
+ "loss": 0.6945,
107
+ "step": 28
108
+ },
109
+ {
110
+ "epoch": 0.3,
111
+ "grad_norm": 4.305431842803955,
112
+ "learning_rate": 2e-05,
113
+ "loss": 1.1069,
114
+ "step": 30
115
+ },
116
+ {
117
+ "epoch": 0.32,
118
+ "grad_norm": 2.340209484100342,
119
+ "learning_rate": 2e-05,
120
+ "loss": 1.5016,
121
+ "step": 32
122
+ },
123
+ {
124
+ "epoch": 0.34,
125
+ "grad_norm": 1.1784175634384155,
126
+ "learning_rate": 2e-05,
127
+ "loss": 0.4982,
128
+ "step": 34
129
+ },
130
+ {
131
+ "epoch": 0.36,
132
+ "grad_norm": 3.8605434894561768,
133
+ "learning_rate": 2e-05,
134
+ "loss": 1.5003,
135
+ "step": 36
136
+ },
137
+ {
138
+ "epoch": 0.38,
139
+ "grad_norm": 5.080264091491699,
140
+ "learning_rate": 2e-05,
141
+ "loss": 3.3415,
142
+ "step": 38
143
+ },
144
+ {
145
+ "epoch": 0.4,
146
+ "grad_norm": 2.7285072803497314,
147
+ "learning_rate": 2e-05,
148
+ "loss": 1.3715,
149
+ "step": 40
150
+ },
151
+ {
152
+ "epoch": 0.42,
153
+ "grad_norm": 3.899984359741211,
154
+ "learning_rate": 2e-05,
155
+ "loss": 1.2467,
156
+ "step": 42
157
+ },
158
+ {
159
+ "epoch": 0.44,
160
+ "grad_norm": 11.371481895446777,
161
+ "learning_rate": 2e-05,
162
+ "loss": 1.2012,
163
+ "step": 44
164
+ },
165
+ {
166
+ "epoch": 0.46,
167
+ "grad_norm": 3.17934513092041,
168
+ "learning_rate": 2e-05,
169
+ "loss": 1.1335,
170
+ "step": 46
171
+ },
172
+ {
173
+ "epoch": 0.48,
174
+ "grad_norm": 1.391940951347351,
175
+ "learning_rate": 2e-05,
176
+ "loss": 0.5662,
177
+ "step": 48
178
+ },
179
+ {
180
+ "epoch": 0.5,
181
+ "grad_norm": 2.64958119392395,
182
+ "learning_rate": 2e-05,
183
+ "loss": 1.1034,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.52,
188
+ "grad_norm": 2.085965156555176,
189
+ "learning_rate": 2e-05,
190
+ "loss": 0.8947,
191
+ "step": 52
192
+ },
193
+ {
194
+ "epoch": 0.54,
195
+ "grad_norm": 5.4762749671936035,
196
+ "learning_rate": 2e-05,
197
+ "loss": 2.1323,
198
+ "step": 54
199
+ },
200
+ {
201
+ "epoch": 0.56,
202
+ "grad_norm": 5.612151145935059,
203
+ "learning_rate": 2e-05,
204
+ "loss": 1.8613,
205
+ "step": 56
206
+ },
207
+ {
208
+ "epoch": 0.58,
209
+ "grad_norm": 12.62922477722168,
210
+ "learning_rate": 2e-05,
211
+ "loss": 1.5807,
212
+ "step": 58
213
+ },
214
+ {
215
+ "epoch": 0.6,
216
+ "grad_norm": 3.3207852840423584,
217
+ "learning_rate": 2e-05,
218
+ "loss": 0.71,
219
+ "step": 60
220
+ },
221
+ {
222
+ "epoch": 0.62,
223
+ "grad_norm": 5.958663463592529,
224
+ "learning_rate": 2e-05,
225
+ "loss": 1.88,
226
+ "step": 62
227
+ },
228
+ {
229
+ "epoch": 0.64,
230
+ "grad_norm": 3.557206630706787,
231
+ "learning_rate": 2e-05,
232
+ "loss": 1.6649,
233
+ "step": 64
234
+ },
235
+ {
236
+ "epoch": 0.66,
237
+ "grad_norm": 3.5797131061553955,
238
+ "learning_rate": 2e-05,
239
+ "loss": 0.91,
240
+ "step": 66
241
+ },
242
+ {
243
+ "epoch": 0.68,
244
+ "grad_norm": 3.187143087387085,
245
+ "learning_rate": 2e-05,
246
+ "loss": 0.6656,
247
+ "step": 68
248
+ },
249
+ {
250
+ "epoch": 0.7,
251
+ "grad_norm": 6.657142162322998,
252
+ "learning_rate": 2e-05,
253
+ "loss": 1.3212,
254
+ "step": 70
255
+ },
256
+ {
257
+ "epoch": 0.72,
258
+ "grad_norm": 2.342517614364624,
259
+ "learning_rate": 2e-05,
260
+ "loss": 0.8374,
261
+ "step": 72
262
+ },
263
+ {
264
+ "epoch": 0.74,
265
+ "grad_norm": 2.212033271789551,
266
+ "learning_rate": 2e-05,
267
+ "loss": 1.0031,
268
+ "step": 74
269
+ },
270
+ {
271
+ "epoch": 0.76,
272
+ "grad_norm": 1.835802674293518,
273
+ "learning_rate": 2e-05,
274
+ "loss": 0.957,
275
+ "step": 76
276
+ },
277
+ {
278
+ "epoch": 0.78,
279
+ "grad_norm": 3.243709087371826,
280
+ "learning_rate": 2e-05,
281
+ "loss": 1.4119,
282
+ "step": 78
283
+ },
284
+ {
285
+ "epoch": 0.8,
286
+ "grad_norm": 2.2901952266693115,
287
+ "learning_rate": 2e-05,
288
+ "loss": 1.581,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.82,
293
+ "grad_norm": 3.645848035812378,
294
+ "learning_rate": 2e-05,
295
+ "loss": 1.3598,
296
+ "step": 82
297
+ },
298
+ {
299
+ "epoch": 0.84,
300
+ "grad_norm": 1.8024070262908936,
301
+ "learning_rate": 2e-05,
302
+ "loss": 1.3748,
303
+ "step": 84
304
+ },
305
+ {
306
+ "epoch": 0.86,
307
+ "grad_norm": 1.429137945175171,
308
+ "learning_rate": 2e-05,
309
+ "loss": 0.8242,
310
+ "step": 86
311
+ },
312
+ {
313
+ "epoch": 0.88,
314
+ "grad_norm": 4.3826003074646,
315
+ "learning_rate": 2e-05,
316
+ "loss": 1.2812,
317
+ "step": 88
318
+ },
319
+ {
320
+ "epoch": 0.9,
321
+ "grad_norm": 1.7045739889144897,
322
+ "learning_rate": 2e-05,
323
+ "loss": 0.4062,
324
+ "step": 90
325
+ },
326
+ {
327
+ "epoch": 0.92,
328
+ "grad_norm": 2.06375789642334,
329
+ "learning_rate": 2e-05,
330
+ "loss": 0.526,
331
+ "step": 92
332
+ },
333
+ {
334
+ "epoch": 0.94,
335
+ "grad_norm": 2.061166286468506,
336
+ "learning_rate": 2e-05,
337
+ "loss": 0.9437,
338
+ "step": 94
339
+ },
340
+ {
341
+ "epoch": 0.96,
342
+ "grad_norm": 1.7989872694015503,
343
+ "learning_rate": 2e-05,
344
+ "loss": 0.7991,
345
+ "step": 96
346
+ },
347
+ {
348
+ "epoch": 0.98,
349
+ "grad_norm": 2.4970247745513916,
350
+ "learning_rate": 2e-05,
351
+ "loss": 1.1118,
352
+ "step": 98
353
+ },
354
+ {
355
+ "epoch": 1.0,
356
+ "grad_norm": 2.229430913925171,
357
+ "learning_rate": 2e-05,
358
+ "loss": 0.5631,
359
+ "step": 100
360
+ },
361
+ {
362
+ "epoch": 1.0,
363
+ "step": 100,
364
+ "total_flos": 5831214409711616.0,
365
+ "train_loss": 1.1269862699508666,
366
+ "train_runtime": 270.2671,
367
+ "train_samples_per_second": 1.48,
368
+ "train_steps_per_second": 0.37
369
+ }
370
+ ],
371
+ "logging_steps": 2,
372
+ "max_steps": 100,
373
+ "num_input_tokens_seen": 0,
374
+ "num_train_epochs": 1,
375
+ "save_steps": 500,
376
+ "stateful_callbacks": {
377
+ "TrainerControl": {
378
+ "args": {
379
+ "should_epoch_stop": false,
380
+ "should_evaluate": false,
381
+ "should_log": false,
382
+ "should_save": false,
383
+ "should_training_stop": false
384
+ },
385
+ "attributes": {}
386
+ }
387
+ },
388
+ "total_flos": 5831214409711616.0,
389
+ "train_batch_size": 1,
390
+ "trial_name": null,
391
+ "trial_params": null
392
+ }
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:138d506fddbf706bb6e41a017e4b9f2ab4b63a8cfb469edeff3bb360587136b6
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round12.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa7e2d9e865a75c8230a00824f1560f031639bca31cb98b4c27a3f112a69b7e6
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round15.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d499d9d346b71286e8ae97114529f323e96f244e321d64e0292ec29e83c81690
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d6dcbba22c491daf3769ba29f1bb162206fa57f69658476f22771aee70fffb5
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50b1af7ea78d2bda758b82278f8e726b8328ca406e0fb004f90add894a133c24
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round20.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64f07e9eba6c814a47dc54d354838e8d98dc5df9aee63bb04f35246b5006955c
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60bbc244045c277e610418ce7382f1e2b81d566983b94b08f40b50695a927e61
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_client_model_round7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b96ff1284fe3996e65338ab2297c9ccb067288c7952a64937d954566be48da4c
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/1_trainer_state.json ADDED
@@ -0,0 +1,392 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 4.277268409729004,
14
+ "learning_rate": 2e-05,
15
+ "loss": 1.6152,
16
+ "step": 2
17
+ },
18
+ {
19
+ "epoch": 0.04,
20
+ "grad_norm": 3.284519672393799,
21
+ "learning_rate": 2e-05,
22
+ "loss": 1.3525,
23
+ "step": 4
24
+ },
25
+ {
26
+ "epoch": 0.06,
27
+ "grad_norm": 2.114292621612549,
28
+ "learning_rate": 2e-05,
29
+ "loss": 1.2699,
30
+ "step": 6
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "grad_norm": 1.3881689310073853,
35
+ "learning_rate": 2e-05,
36
+ "loss": 1.2783,
37
+ "step": 8
38
+ },
39
+ {
40
+ "epoch": 0.1,
41
+ "grad_norm": 1.070853590965271,
42
+ "learning_rate": 2e-05,
43
+ "loss": 1.052,
44
+ "step": 10
45
+ },
46
+ {
47
+ "epoch": 0.12,
48
+ "grad_norm": 1.7331618070602417,
49
+ "learning_rate": 2e-05,
50
+ "loss": 0.9035,
51
+ "step": 12
52
+ },
53
+ {
54
+ "epoch": 0.14,
55
+ "grad_norm": 2.8790841102600098,
56
+ "learning_rate": 2e-05,
57
+ "loss": 1.3837,
58
+ "step": 14
59
+ },
60
+ {
61
+ "epoch": 0.16,
62
+ "grad_norm": 1.225778579711914,
63
+ "learning_rate": 2e-05,
64
+ "loss": 1.0574,
65
+ "step": 16
66
+ },
67
+ {
68
+ "epoch": 0.18,
69
+ "grad_norm": 1.4563435316085815,
70
+ "learning_rate": 2e-05,
71
+ "loss": 0.8274,
72
+ "step": 18
73
+ },
74
+ {
75
+ "epoch": 0.2,
76
+ "grad_norm": 1.5500695705413818,
77
+ "learning_rate": 2e-05,
78
+ "loss": 0.8812,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.22,
83
+ "grad_norm": 1.326027512550354,
84
+ "learning_rate": 2e-05,
85
+ "loss": 0.9961,
86
+ "step": 22
87
+ },
88
+ {
89
+ "epoch": 0.24,
90
+ "grad_norm": 1.095481514930725,
91
+ "learning_rate": 2e-05,
92
+ "loss": 1.0845,
93
+ "step": 24
94
+ },
95
+ {
96
+ "epoch": 0.26,
97
+ "grad_norm": 1.2606724500656128,
98
+ "learning_rate": 2e-05,
99
+ "loss": 0.6172,
100
+ "step": 26
101
+ },
102
+ {
103
+ "epoch": 0.28,
104
+ "grad_norm": 1.4941548109054565,
105
+ "learning_rate": 2e-05,
106
+ "loss": 0.7191,
107
+ "step": 28
108
+ },
109
+ {
110
+ "epoch": 0.3,
111
+ "grad_norm": 1.3917955160140991,
112
+ "learning_rate": 2e-05,
113
+ "loss": 1.2529,
114
+ "step": 30
115
+ },
116
+ {
117
+ "epoch": 0.32,
118
+ "grad_norm": 2.953693389892578,
119
+ "learning_rate": 2e-05,
120
+ "loss": 1.371,
121
+ "step": 32
122
+ },
123
+ {
124
+ "epoch": 0.34,
125
+ "grad_norm": 0.6573340892791748,
126
+ "learning_rate": 2e-05,
127
+ "loss": 1.0989,
128
+ "step": 34
129
+ },
130
+ {
131
+ "epoch": 0.36,
132
+ "grad_norm": 1.684754729270935,
133
+ "learning_rate": 2e-05,
134
+ "loss": 0.9304,
135
+ "step": 36
136
+ },
137
+ {
138
+ "epoch": 0.38,
139
+ "grad_norm": 0.9705935120582581,
140
+ "learning_rate": 2e-05,
141
+ "loss": 0.4247,
142
+ "step": 38
143
+ },
144
+ {
145
+ "epoch": 0.4,
146
+ "grad_norm": 1.424384355545044,
147
+ "learning_rate": 2e-05,
148
+ "loss": 1.1189,
149
+ "step": 40
150
+ },
151
+ {
152
+ "epoch": 0.42,
153
+ "grad_norm": 1.9079763889312744,
154
+ "learning_rate": 2e-05,
155
+ "loss": 1.0473,
156
+ "step": 42
157
+ },
158
+ {
159
+ "epoch": 0.44,
160
+ "grad_norm": 0.7125902771949768,
161
+ "learning_rate": 2e-05,
162
+ "loss": 1.035,
163
+ "step": 44
164
+ },
165
+ {
166
+ "epoch": 0.46,
167
+ "grad_norm": 1.202347993850708,
168
+ "learning_rate": 2e-05,
169
+ "loss": 0.7511,
170
+ "step": 46
171
+ },
172
+ {
173
+ "epoch": 0.48,
174
+ "grad_norm": 1.2924076318740845,
175
+ "learning_rate": 2e-05,
176
+ "loss": 0.6636,
177
+ "step": 48
178
+ },
179
+ {
180
+ "epoch": 0.5,
181
+ "grad_norm": 1.6152644157409668,
182
+ "learning_rate": 2e-05,
183
+ "loss": 1.1831,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.52,
188
+ "grad_norm": 2.278852939605713,
189
+ "learning_rate": 2e-05,
190
+ "loss": 1.4946,
191
+ "step": 52
192
+ },
193
+ {
194
+ "epoch": 0.54,
195
+ "grad_norm": 2.523404121398926,
196
+ "learning_rate": 2e-05,
197
+ "loss": 1.3058,
198
+ "step": 54
199
+ },
200
+ {
201
+ "epoch": 0.56,
202
+ "grad_norm": 0.9222102165222168,
203
+ "learning_rate": 2e-05,
204
+ "loss": 0.4836,
205
+ "step": 56
206
+ },
207
+ {
208
+ "epoch": 0.58,
209
+ "grad_norm": 1.4422670602798462,
210
+ "learning_rate": 2e-05,
211
+ "loss": 0.7916,
212
+ "step": 58
213
+ },
214
+ {
215
+ "epoch": 0.6,
216
+ "grad_norm": 1.1267635822296143,
217
+ "learning_rate": 2e-05,
218
+ "loss": 0.8997,
219
+ "step": 60
220
+ },
221
+ {
222
+ "epoch": 0.62,
223
+ "grad_norm": 1.4941160678863525,
224
+ "learning_rate": 2e-05,
225
+ "loss": 1.3463,
226
+ "step": 62
227
+ },
228
+ {
229
+ "epoch": 0.64,
230
+ "grad_norm": 1.4414238929748535,
231
+ "learning_rate": 2e-05,
232
+ "loss": 0.8191,
233
+ "step": 64
234
+ },
235
+ {
236
+ "epoch": 0.66,
237
+ "grad_norm": 2.048666000366211,
238
+ "learning_rate": 2e-05,
239
+ "loss": 1.3579,
240
+ "step": 66
241
+ },
242
+ {
243
+ "epoch": 0.68,
244
+ "grad_norm": 3.2488296031951904,
245
+ "learning_rate": 2e-05,
246
+ "loss": 1.543,
247
+ "step": 68
248
+ },
249
+ {
250
+ "epoch": 0.7,
251
+ "grad_norm": 0.8362247347831726,
252
+ "learning_rate": 2e-05,
253
+ "loss": 1.2804,
254
+ "step": 70
255
+ },
256
+ {
257
+ "epoch": 0.72,
258
+ "grad_norm": 1.2172328233718872,
259
+ "learning_rate": 2e-05,
260
+ "loss": 1.0283,
261
+ "step": 72
262
+ },
263
+ {
264
+ "epoch": 0.74,
265
+ "grad_norm": 0.9695539474487305,
266
+ "learning_rate": 2e-05,
267
+ "loss": 0.8484,
268
+ "step": 74
269
+ },
270
+ {
271
+ "epoch": 0.76,
272
+ "grad_norm": 1.3703746795654297,
273
+ "learning_rate": 2e-05,
274
+ "loss": 0.9785,
275
+ "step": 76
276
+ },
277
+ {
278
+ "epoch": 0.78,
279
+ "grad_norm": 1.8802088499069214,
280
+ "learning_rate": 2e-05,
281
+ "loss": 0.6836,
282
+ "step": 78
283
+ },
284
+ {
285
+ "epoch": 0.8,
286
+ "grad_norm": 4.380638122558594,
287
+ "learning_rate": 2e-05,
288
+ "loss": 1.9597,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.82,
293
+ "grad_norm": 0.8395757675170898,
294
+ "learning_rate": 2e-05,
295
+ "loss": 0.6358,
296
+ "step": 82
297
+ },
298
+ {
299
+ "epoch": 0.84,
300
+ "grad_norm": 0.6108465790748596,
301
+ "learning_rate": 2e-05,
302
+ "loss": 0.8909,
303
+ "step": 84
304
+ },
305
+ {
306
+ "epoch": 0.86,
307
+ "grad_norm": 2.491547107696533,
308
+ "learning_rate": 2e-05,
309
+ "loss": 1.2093,
310
+ "step": 86
311
+ },
312
+ {
313
+ "epoch": 0.88,
314
+ "grad_norm": 1.081382155418396,
315
+ "learning_rate": 2e-05,
316
+ "loss": 0.9521,
317
+ "step": 88
318
+ },
319
+ {
320
+ "epoch": 0.9,
321
+ "grad_norm": 1.6292481422424316,
322
+ "learning_rate": 2e-05,
323
+ "loss": 1.1539,
324
+ "step": 90
325
+ },
326
+ {
327
+ "epoch": 0.92,
328
+ "grad_norm": 1.1277912855148315,
329
+ "learning_rate": 2e-05,
330
+ "loss": 0.98,
331
+ "step": 92
332
+ },
333
+ {
334
+ "epoch": 0.94,
335
+ "grad_norm": 2.7916040420532227,
336
+ "learning_rate": 2e-05,
337
+ "loss": 1.1936,
338
+ "step": 94
339
+ },
340
+ {
341
+ "epoch": 0.96,
342
+ "grad_norm": 1.3079055547714233,
343
+ "learning_rate": 2e-05,
344
+ "loss": 0.8721,
345
+ "step": 96
346
+ },
347
+ {
348
+ "epoch": 0.98,
349
+ "grad_norm": 1.0078022480010986,
350
+ "learning_rate": 2e-05,
351
+ "loss": 1.0526,
352
+ "step": 98
353
+ },
354
+ {
355
+ "epoch": 1.0,
356
+ "grad_norm": 0.9928855299949646,
357
+ "learning_rate": 2e-05,
358
+ "loss": 1.6038,
359
+ "step": 100
360
+ },
361
+ {
362
+ "epoch": 1.0,
363
+ "step": 100,
364
+ "total_flos": 1.3968703347490816e+16,
365
+ "train_loss": 1.0655856895446778,
366
+ "train_runtime": 439.7119,
367
+ "train_samples_per_second": 0.91,
368
+ "train_steps_per_second": 0.227
369
+ }
370
+ ],
371
+ "logging_steps": 2,
372
+ "max_steps": 100,
373
+ "num_input_tokens_seen": 0,
374
+ "num_train_epochs": 1,
375
+ "save_steps": 500,
376
+ "stateful_callbacks": {
377
+ "TrainerControl": {
378
+ "args": {
379
+ "should_epoch_stop": false,
380
+ "should_evaluate": false,
381
+ "should_log": false,
382
+ "should_save": false,
383
+ "should_training_stop": false
384
+ },
385
+ "attributes": {}
386
+ }
387
+ },
388
+ "total_flos": 1.3968703347490816e+16,
389
+ "train_batch_size": 1,
390
+ "trial_name": null,
391
+ "trial_params": null
392
+ }
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2bb9e20de42cfbe3f44987cbfca074a54706356ca0fca81fa97321ae9b04a7b
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round12.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6596b9ce6ff3eecfa1234ebf66a7d4fcafa5f2c09808098b40b46c4ad5bc89a8
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round15.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:841a689deef9a345c9e4c2e419fb614cbd810f900209cb4e132ed3455e303174
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a78d0a09b76f428bd0d271bcc7e6795dde8eaa878da812b0b1afd207e013869
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5fa9ec27a4e03e9226fa5fe66c2cdc46139f66cfee70957af64edba25e1c307
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round20.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f30b6f3fafe01f95015c5c455f65d1a0efd925f561d914e4443b426b60f3c741
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbb91e3a36051921f83b02c5a58fcec1e6178c517b573b0fc11af16d63e6f159
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_client_model_round7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37613e1c7b8f959cf59b3642057bcde40283595caff935de957e93c0583b6b67
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/2_trainer_state.json ADDED
@@ -0,0 +1,392 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 4.621701717376709,
14
+ "learning_rate": 2e-05,
15
+ "loss": 0.9244,
16
+ "step": 2
17
+ },
18
+ {
19
+ "epoch": 0.04,
20
+ "grad_norm": 0.73643958568573,
21
+ "learning_rate": 2e-05,
22
+ "loss": 0.3462,
23
+ "step": 4
24
+ },
25
+ {
26
+ "epoch": 0.06,
27
+ "grad_norm": 1.9696252346038818,
28
+ "learning_rate": 2e-05,
29
+ "loss": 0.687,
30
+ "step": 6
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "grad_norm": 2.036741256713867,
35
+ "learning_rate": 2e-05,
36
+ "loss": 0.3694,
37
+ "step": 8
38
+ },
39
+ {
40
+ "epoch": 0.1,
41
+ "grad_norm": 3.216172218322754,
42
+ "learning_rate": 2e-05,
43
+ "loss": 0.9554,
44
+ "step": 10
45
+ },
46
+ {
47
+ "epoch": 0.12,
48
+ "grad_norm": 2.133591651916504,
49
+ "learning_rate": 2e-05,
50
+ "loss": 0.8185,
51
+ "step": 12
52
+ },
53
+ {
54
+ "epoch": 0.14,
55
+ "grad_norm": 6.794097423553467,
56
+ "learning_rate": 2e-05,
57
+ "loss": 1.6995,
58
+ "step": 14
59
+ },
60
+ {
61
+ "epoch": 0.16,
62
+ "grad_norm": 1.703186273574829,
63
+ "learning_rate": 2e-05,
64
+ "loss": 0.6469,
65
+ "step": 16
66
+ },
67
+ {
68
+ "epoch": 0.18,
69
+ "grad_norm": 2.852735996246338,
70
+ "learning_rate": 2e-05,
71
+ "loss": 1.0313,
72
+ "step": 18
73
+ },
74
+ {
75
+ "epoch": 0.2,
76
+ "grad_norm": 1.5393972396850586,
77
+ "learning_rate": 2e-05,
78
+ "loss": 0.7196,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.22,
83
+ "grad_norm": 0.9888882040977478,
84
+ "learning_rate": 2e-05,
85
+ "loss": 0.8174,
86
+ "step": 22
87
+ },
88
+ {
89
+ "epoch": 0.24,
90
+ "grad_norm": 1.0866042375564575,
91
+ "learning_rate": 2e-05,
92
+ "loss": 0.677,
93
+ "step": 24
94
+ },
95
+ {
96
+ "epoch": 0.26,
97
+ "grad_norm": 0.9214580655097961,
98
+ "learning_rate": 2e-05,
99
+ "loss": 0.5664,
100
+ "step": 26
101
+ },
102
+ {
103
+ "epoch": 0.28,
104
+ "grad_norm": 4.819929599761963,
105
+ "learning_rate": 2e-05,
106
+ "loss": 0.9781,
107
+ "step": 28
108
+ },
109
+ {
110
+ "epoch": 0.3,
111
+ "grad_norm": 0.7382586598396301,
112
+ "learning_rate": 2e-05,
113
+ "loss": 0.6898,
114
+ "step": 30
115
+ },
116
+ {
117
+ "epoch": 0.32,
118
+ "grad_norm": 1.9690661430358887,
119
+ "learning_rate": 2e-05,
120
+ "loss": 0.5947,
121
+ "step": 32
122
+ },
123
+ {
124
+ "epoch": 0.34,
125
+ "grad_norm": 17.00811767578125,
126
+ "learning_rate": 2e-05,
127
+ "loss": 0.83,
128
+ "step": 34
129
+ },
130
+ {
131
+ "epoch": 0.36,
132
+ "grad_norm": 6.0308990478515625,
133
+ "learning_rate": 2e-05,
134
+ "loss": 1.1475,
135
+ "step": 36
136
+ },
137
+ {
138
+ "epoch": 0.38,
139
+ "grad_norm": 1.098817229270935,
140
+ "learning_rate": 2e-05,
141
+ "loss": 0.6007,
142
+ "step": 38
143
+ },
144
+ {
145
+ "epoch": 0.4,
146
+ "grad_norm": 1.9793791770935059,
147
+ "learning_rate": 2e-05,
148
+ "loss": 0.5569,
149
+ "step": 40
150
+ },
151
+ {
152
+ "epoch": 0.42,
153
+ "grad_norm": 6.214341163635254,
154
+ "learning_rate": 2e-05,
155
+ "loss": 0.9612,
156
+ "step": 42
157
+ },
158
+ {
159
+ "epoch": 0.44,
160
+ "grad_norm": 1.4605873823165894,
161
+ "learning_rate": 2e-05,
162
+ "loss": 0.356,
163
+ "step": 44
164
+ },
165
+ {
166
+ "epoch": 0.46,
167
+ "grad_norm": 1.5833739042282104,
168
+ "learning_rate": 2e-05,
169
+ "loss": 0.5266,
170
+ "step": 46
171
+ },
172
+ {
173
+ "epoch": 0.48,
174
+ "grad_norm": 0.6748036742210388,
175
+ "learning_rate": 2e-05,
176
+ "loss": 0.4992,
177
+ "step": 48
178
+ },
179
+ {
180
+ "epoch": 0.5,
181
+ "grad_norm": 11.380123138427734,
182
+ "learning_rate": 2e-05,
183
+ "loss": 1.5851,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.52,
188
+ "grad_norm": 4.65256929397583,
189
+ "learning_rate": 2e-05,
190
+ "loss": 0.9587,
191
+ "step": 52
192
+ },
193
+ {
194
+ "epoch": 0.54,
195
+ "grad_norm": 1.627728819847107,
196
+ "learning_rate": 2e-05,
197
+ "loss": 0.8431,
198
+ "step": 54
199
+ },
200
+ {
201
+ "epoch": 0.56,
202
+ "grad_norm": 0.694214940071106,
203
+ "learning_rate": 2e-05,
204
+ "loss": 0.5186,
205
+ "step": 56
206
+ },
207
+ {
208
+ "epoch": 0.58,
209
+ "grad_norm": 1.7453944683074951,
210
+ "learning_rate": 2e-05,
211
+ "loss": 0.672,
212
+ "step": 58
213
+ },
214
+ {
215
+ "epoch": 0.6,
216
+ "grad_norm": 1.9566174745559692,
217
+ "learning_rate": 2e-05,
218
+ "loss": 0.8228,
219
+ "step": 60
220
+ },
221
+ {
222
+ "epoch": 0.62,
223
+ "grad_norm": 4.4222798347473145,
224
+ "learning_rate": 2e-05,
225
+ "loss": 1.1676,
226
+ "step": 62
227
+ },
228
+ {
229
+ "epoch": 0.64,
230
+ "grad_norm": 1.216988444328308,
231
+ "learning_rate": 2e-05,
232
+ "loss": 0.5975,
233
+ "step": 64
234
+ },
235
+ {
236
+ "epoch": 0.66,
237
+ "grad_norm": 3.127988576889038,
238
+ "learning_rate": 2e-05,
239
+ "loss": 1.2231,
240
+ "step": 66
241
+ },
242
+ {
243
+ "epoch": 0.68,
244
+ "grad_norm": 2.3479421138763428,
245
+ "learning_rate": 2e-05,
246
+ "loss": 0.4818,
247
+ "step": 68
248
+ },
249
+ {
250
+ "epoch": 0.7,
251
+ "grad_norm": 7.464072227478027,
252
+ "learning_rate": 2e-05,
253
+ "loss": 1.0901,
254
+ "step": 70
255
+ },
256
+ {
257
+ "epoch": 0.72,
258
+ "grad_norm": 1.146981954574585,
259
+ "learning_rate": 2e-05,
260
+ "loss": 0.679,
261
+ "step": 72
262
+ },
263
+ {
264
+ "epoch": 0.74,
265
+ "grad_norm": 2.462766408920288,
266
+ "learning_rate": 2e-05,
267
+ "loss": 1.0542,
268
+ "step": 74
269
+ },
270
+ {
271
+ "epoch": 0.76,
272
+ "grad_norm": 2.9858667850494385,
273
+ "learning_rate": 2e-05,
274
+ "loss": 1.316,
275
+ "step": 76
276
+ },
277
+ {
278
+ "epoch": 0.78,
279
+ "grad_norm": 0.8474032282829285,
280
+ "learning_rate": 2e-05,
281
+ "loss": 0.1557,
282
+ "step": 78
283
+ },
284
+ {
285
+ "epoch": 0.8,
286
+ "grad_norm": 1.6668614149093628,
287
+ "learning_rate": 2e-05,
288
+ "loss": 1.1471,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.82,
293
+ "grad_norm": 4.365744590759277,
294
+ "learning_rate": 2e-05,
295
+ "loss": 0.7832,
296
+ "step": 82
297
+ },
298
+ {
299
+ "epoch": 0.84,
300
+ "grad_norm": 1.0024306774139404,
301
+ "learning_rate": 2e-05,
302
+ "loss": 0.3819,
303
+ "step": 84
304
+ },
305
+ {
306
+ "epoch": 0.86,
307
+ "grad_norm": 1.6158033609390259,
308
+ "learning_rate": 2e-05,
309
+ "loss": 0.5178,
310
+ "step": 86
311
+ },
312
+ {
313
+ "epoch": 0.88,
314
+ "grad_norm": 3.335845947265625,
315
+ "learning_rate": 2e-05,
316
+ "loss": 1.0591,
317
+ "step": 88
318
+ },
319
+ {
320
+ "epoch": 0.9,
321
+ "grad_norm": 1.2282785177230835,
322
+ "learning_rate": 2e-05,
323
+ "loss": 0.5881,
324
+ "step": 90
325
+ },
326
+ {
327
+ "epoch": 0.92,
328
+ "grad_norm": 1.0900323390960693,
329
+ "learning_rate": 2e-05,
330
+ "loss": 0.6929,
331
+ "step": 92
332
+ },
333
+ {
334
+ "epoch": 0.94,
335
+ "grad_norm": 4.871852874755859,
336
+ "learning_rate": 2e-05,
337
+ "loss": 1.1615,
338
+ "step": 94
339
+ },
340
+ {
341
+ "epoch": 0.96,
342
+ "grad_norm": 5.758192539215088,
343
+ "learning_rate": 2e-05,
344
+ "loss": 1.7691,
345
+ "step": 96
346
+ },
347
+ {
348
+ "epoch": 0.98,
349
+ "grad_norm": 1.9010215997695923,
350
+ "learning_rate": 2e-05,
351
+ "loss": 1.1982,
352
+ "step": 98
353
+ },
354
+ {
355
+ "epoch": 1.0,
356
+ "grad_norm": 3.8769309520721436,
357
+ "learning_rate": 2e-05,
358
+ "loss": 1.0723,
359
+ "step": 100
360
+ },
361
+ {
362
+ "epoch": 1.0,
363
+ "step": 100,
364
+ "total_flos": 1.5469620907474944e+16,
365
+ "train_loss": 0.8307249307632446,
366
+ "train_runtime": 649.8095,
367
+ "train_samples_per_second": 0.616,
368
+ "train_steps_per_second": 0.154
369
+ }
370
+ ],
371
+ "logging_steps": 2,
372
+ "max_steps": 100,
373
+ "num_input_tokens_seen": 0,
374
+ "num_train_epochs": 1,
375
+ "save_steps": 500,
376
+ "stateful_callbacks": {
377
+ "TrainerControl": {
378
+ "args": {
379
+ "should_epoch_stop": false,
380
+ "should_evaluate": false,
381
+ "should_log": false,
382
+ "should_save": false,
383
+ "should_training_stop": false
384
+ },
385
+ "attributes": {}
386
+ }
387
+ },
388
+ "total_flos": 1.5469620907474944e+16,
389
+ "train_batch_size": 1,
390
+ "trial_name": null,
391
+ "trial_params": null
392
+ }
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:46cd618ae9e9b7adba74cfc50f99d22a1965dd688f237a15a022718d7390dd69
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round12.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c04f52e110d6016d15a506ba9fbb133001810c759ef599175be6d6297a372830
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round15.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:618bfb4299137f6816838703471a4558ad71c24556cf450a9ab76448eed19616
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cfddf6b95087538040e23b831fe2cdea391a3b9932497295a668445f2541ef2
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8f1551a48c919d71615d890291b42a1b8d58c4693e3c91c9a2188c43e01e6fe
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round20.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:410ad579ba65adead23eece9de069fbcd2412996aacf5dbf4c642c74abc44fde
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61198ed3d6c412f256665eff1100075341608ef6f37fb551adc793d7eb41b9ed
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_client_model_round7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80cde2f70536d4c4895bcd3d8e43249040d674a84b423e6dd3dc94d3f564a6b0
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/3_trainer_state.json ADDED
@@ -0,0 +1,392 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 1.0928549766540527,
14
+ "learning_rate": 2e-05,
15
+ "loss": 0.8454,
16
+ "step": 2
17
+ },
18
+ {
19
+ "epoch": 0.04,
20
+ "grad_norm": 1.0574003458023071,
21
+ "learning_rate": 2e-05,
22
+ "loss": 1.0656,
23
+ "step": 4
24
+ },
25
+ {
26
+ "epoch": 0.06,
27
+ "grad_norm": 1.322147011756897,
28
+ "learning_rate": 2e-05,
29
+ "loss": 0.8001,
30
+ "step": 6
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "grad_norm": 0.7461885213851929,
35
+ "learning_rate": 2e-05,
36
+ "loss": 0.775,
37
+ "step": 8
38
+ },
39
+ {
40
+ "epoch": 0.1,
41
+ "grad_norm": 1.1713924407958984,
42
+ "learning_rate": 2e-05,
43
+ "loss": 0.6756,
44
+ "step": 10
45
+ },
46
+ {
47
+ "epoch": 0.12,
48
+ "grad_norm": 0.5914626121520996,
49
+ "learning_rate": 2e-05,
50
+ "loss": 0.4239,
51
+ "step": 12
52
+ },
53
+ {
54
+ "epoch": 0.14,
55
+ "grad_norm": 0.7302200198173523,
56
+ "learning_rate": 2e-05,
57
+ "loss": 0.4585,
58
+ "step": 14
59
+ },
60
+ {
61
+ "epoch": 0.16,
62
+ "grad_norm": 0.8082689642906189,
63
+ "learning_rate": 2e-05,
64
+ "loss": 0.717,
65
+ "step": 16
66
+ },
67
+ {
68
+ "epoch": 0.18,
69
+ "grad_norm": 1.3627772331237793,
70
+ "learning_rate": 2e-05,
71
+ "loss": 1.1646,
72
+ "step": 18
73
+ },
74
+ {
75
+ "epoch": 0.2,
76
+ "grad_norm": 0.7314639687538147,
77
+ "learning_rate": 2e-05,
78
+ "loss": 0.5912,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.22,
83
+ "grad_norm": 0.7471842169761658,
84
+ "learning_rate": 2e-05,
85
+ "loss": 0.2537,
86
+ "step": 22
87
+ },
88
+ {
89
+ "epoch": 0.24,
90
+ "grad_norm": 0.23702213168144226,
91
+ "learning_rate": 2e-05,
92
+ "loss": 0.2199,
93
+ "step": 24
94
+ },
95
+ {
96
+ "epoch": 0.26,
97
+ "grad_norm": 0.276137113571167,
98
+ "learning_rate": 2e-05,
99
+ "loss": 0.3347,
100
+ "step": 26
101
+ },
102
+ {
103
+ "epoch": 0.28,
104
+ "grad_norm": 2.8980002403259277,
105
+ "learning_rate": 2e-05,
106
+ "loss": 1.5936,
107
+ "step": 28
108
+ },
109
+ {
110
+ "epoch": 0.3,
111
+ "grad_norm": 1.8844953775405884,
112
+ "learning_rate": 2e-05,
113
+ "loss": 1.5603,
114
+ "step": 30
115
+ },
116
+ {
117
+ "epoch": 0.32,
118
+ "grad_norm": 0.5475688576698303,
119
+ "learning_rate": 2e-05,
120
+ "loss": 0.679,
121
+ "step": 32
122
+ },
123
+ {
124
+ "epoch": 0.34,
125
+ "grad_norm": 2.1111373901367188,
126
+ "learning_rate": 2e-05,
127
+ "loss": 1.0442,
128
+ "step": 34
129
+ },
130
+ {
131
+ "epoch": 0.36,
132
+ "grad_norm": 1.6544932126998901,
133
+ "learning_rate": 2e-05,
134
+ "loss": 0.6904,
135
+ "step": 36
136
+ },
137
+ {
138
+ "epoch": 0.38,
139
+ "grad_norm": 1.5641965866088867,
140
+ "learning_rate": 2e-05,
141
+ "loss": 0.8929,
142
+ "step": 38
143
+ },
144
+ {
145
+ "epoch": 0.4,
146
+ "grad_norm": 0.9792295694351196,
147
+ "learning_rate": 2e-05,
148
+ "loss": 0.5441,
149
+ "step": 40
150
+ },
151
+ {
152
+ "epoch": 0.42,
153
+ "grad_norm": 1.304323434829712,
154
+ "learning_rate": 2e-05,
155
+ "loss": 0.6259,
156
+ "step": 42
157
+ },
158
+ {
159
+ "epoch": 0.44,
160
+ "grad_norm": 1.5232268571853638,
161
+ "learning_rate": 2e-05,
162
+ "loss": 0.7603,
163
+ "step": 44
164
+ },
165
+ {
166
+ "epoch": 0.46,
167
+ "grad_norm": 1.5708423852920532,
168
+ "learning_rate": 2e-05,
169
+ "loss": 0.5623,
170
+ "step": 46
171
+ },
172
+ {
173
+ "epoch": 0.48,
174
+ "grad_norm": 0.6643423438072205,
175
+ "learning_rate": 2e-05,
176
+ "loss": 0.2971,
177
+ "step": 48
178
+ },
179
+ {
180
+ "epoch": 0.5,
181
+ "grad_norm": 0.7532520294189453,
182
+ "learning_rate": 2e-05,
183
+ "loss": 1.0968,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.52,
188
+ "grad_norm": 0.07311046123504639,
189
+ "learning_rate": 2e-05,
190
+ "loss": 0.3392,
191
+ "step": 52
192
+ },
193
+ {
194
+ "epoch": 0.54,
195
+ "grad_norm": 0.947688639163971,
196
+ "learning_rate": 2e-05,
197
+ "loss": 1.0041,
198
+ "step": 54
199
+ },
200
+ {
201
+ "epoch": 0.56,
202
+ "grad_norm": 1.7061188220977783,
203
+ "learning_rate": 2e-05,
204
+ "loss": 1.2263,
205
+ "step": 56
206
+ },
207
+ {
208
+ "epoch": 0.58,
209
+ "grad_norm": 1.699214220046997,
210
+ "learning_rate": 2e-05,
211
+ "loss": 0.677,
212
+ "step": 58
213
+ },
214
+ {
215
+ "epoch": 0.6,
216
+ "grad_norm": 1.089853286743164,
217
+ "learning_rate": 2e-05,
218
+ "loss": 0.4363,
219
+ "step": 60
220
+ },
221
+ {
222
+ "epoch": 0.62,
223
+ "grad_norm": 1.9666467905044556,
224
+ "learning_rate": 2e-05,
225
+ "loss": 0.7766,
226
+ "step": 62
227
+ },
228
+ {
229
+ "epoch": 0.64,
230
+ "grad_norm": 0.9256648421287537,
231
+ "learning_rate": 2e-05,
232
+ "loss": 1.0702,
233
+ "step": 64
234
+ },
235
+ {
236
+ "epoch": 0.66,
237
+ "grad_norm": 1.0910379886627197,
238
+ "learning_rate": 2e-05,
239
+ "loss": 0.6567,
240
+ "step": 66
241
+ },
242
+ {
243
+ "epoch": 0.68,
244
+ "grad_norm": 0.2857760190963745,
245
+ "learning_rate": 2e-05,
246
+ "loss": 0.3937,
247
+ "step": 68
248
+ },
249
+ {
250
+ "epoch": 0.7,
251
+ "grad_norm": 0.7233390212059021,
252
+ "learning_rate": 2e-05,
253
+ "loss": 0.4897,
254
+ "step": 70
255
+ },
256
+ {
257
+ "epoch": 0.72,
258
+ "grad_norm": 1.5829262733459473,
259
+ "learning_rate": 2e-05,
260
+ "loss": 0.4418,
261
+ "step": 72
262
+ },
263
+ {
264
+ "epoch": 0.74,
265
+ "grad_norm": 2.957613706588745,
266
+ "learning_rate": 2e-05,
267
+ "loss": 0.4509,
268
+ "step": 74
269
+ },
270
+ {
271
+ "epoch": 0.76,
272
+ "grad_norm": 1.2405678033828735,
273
+ "learning_rate": 2e-05,
274
+ "loss": 0.9233,
275
+ "step": 76
276
+ },
277
+ {
278
+ "epoch": 0.78,
279
+ "grad_norm": 2.5274651050567627,
280
+ "learning_rate": 2e-05,
281
+ "loss": 0.8624,
282
+ "step": 78
283
+ },
284
+ {
285
+ "epoch": 0.8,
286
+ "grad_norm": 0.8004423379898071,
287
+ "learning_rate": 2e-05,
288
+ "loss": 0.4535,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.82,
293
+ "grad_norm": 0.7731671333312988,
294
+ "learning_rate": 2e-05,
295
+ "loss": 1.2591,
296
+ "step": 82
297
+ },
298
+ {
299
+ "epoch": 0.84,
300
+ "grad_norm": 2.7027082443237305,
301
+ "learning_rate": 2e-05,
302
+ "loss": 0.8519,
303
+ "step": 84
304
+ },
305
+ {
306
+ "epoch": 0.86,
307
+ "grad_norm": 1.2182791233062744,
308
+ "learning_rate": 2e-05,
309
+ "loss": 0.7901,
310
+ "step": 86
311
+ },
312
+ {
313
+ "epoch": 0.88,
314
+ "grad_norm": 1.952547311782837,
315
+ "learning_rate": 2e-05,
316
+ "loss": 1.0298,
317
+ "step": 88
318
+ },
319
+ {
320
+ "epoch": 0.9,
321
+ "grad_norm": 1.7673152685165405,
322
+ "learning_rate": 2e-05,
323
+ "loss": 0.5578,
324
+ "step": 90
325
+ },
326
+ {
327
+ "epoch": 0.92,
328
+ "grad_norm": 1.2945863008499146,
329
+ "learning_rate": 2e-05,
330
+ "loss": 0.6542,
331
+ "step": 92
332
+ },
333
+ {
334
+ "epoch": 0.94,
335
+ "grad_norm": 0.5422699451446533,
336
+ "learning_rate": 2e-05,
337
+ "loss": 0.6201,
338
+ "step": 94
339
+ },
340
+ {
341
+ "epoch": 0.96,
342
+ "grad_norm": 2.9008870124816895,
343
+ "learning_rate": 2e-05,
344
+ "loss": 0.6676,
345
+ "step": 96
346
+ },
347
+ {
348
+ "epoch": 0.98,
349
+ "grad_norm": 1.797958254814148,
350
+ "learning_rate": 2e-05,
351
+ "loss": 0.2743,
352
+ "step": 98
353
+ },
354
+ {
355
+ "epoch": 1.0,
356
+ "grad_norm": 0.26072898507118225,
357
+ "learning_rate": 2e-05,
358
+ "loss": 0.2183,
359
+ "step": 100
360
+ },
361
+ {
362
+ "epoch": 1.0,
363
+ "step": 100,
364
+ "total_flos": 2.159816985490227e+16,
365
+ "train_loss": 0.7159382379055024,
366
+ "train_runtime": 586.8643,
367
+ "train_samples_per_second": 0.682,
368
+ "train_steps_per_second": 0.17
369
+ }
370
+ ],
371
+ "logging_steps": 2,
372
+ "max_steps": 100,
373
+ "num_input_tokens_seen": 0,
374
+ "num_train_epochs": 1,
375
+ "save_steps": 500,
376
+ "stateful_callbacks": {
377
+ "TrainerControl": {
378
+ "args": {
379
+ "should_epoch_stop": false,
380
+ "should_evaluate": false,
381
+ "should_log": false,
382
+ "should_save": false,
383
+ "should_training_stop": false
384
+ },
385
+ "attributes": {}
386
+ }
387
+ },
388
+ "total_flos": 2.159816985490227e+16,
389
+ "train_batch_size": 1,
390
+ "trial_name": null,
391
+ "trial_params": null
392
+ }
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5fe17de3fc2f0d3c561085b8ed6efb817f23394aaf068de54a85111a7d625ca
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round12.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef8ee0bd84ed0c1804be2ee3ea072f8b3d0ecfaa732c6122fbbb11f7cab5017f
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round15.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79f8612ff1fed88ad674dee1359864a60a835afd06d08de137d423b5d0949318
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:127a70279ec5147588c30976b641c703c036d7365103aba2649acce2cf1bcf0c
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bad9f8bd5d6c91e9ae72aeba6e75796c8908ba04a5ea1661e9d5b2678e974d63
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round20.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9c3ea4181eb77e92851c99a2d72ab30bdf4175e5a46f66c945d96e8ac4cfd07
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4983e69fd61d0aa3bf4342e349f514b68282af6a75674f037a7b9c0f48ea9cf
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_client_model_round7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d60ec58e7de2db364bf4c25335a323b659f4cd584734f660ac3a6b57510e7acf
3
+ size 794706058
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/4_trainer_state.json ADDED
@@ -0,0 +1,392 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 1.158302664756775,
14
+ "learning_rate": 2e-05,
15
+ "loss": 0.3006,
16
+ "step": 2
17
+ },
18
+ {
19
+ "epoch": 0.04,
20
+ "grad_norm": 0.49584025144577026,
21
+ "learning_rate": 2e-05,
22
+ "loss": 0.2759,
23
+ "step": 4
24
+ },
25
+ {
26
+ "epoch": 0.06,
27
+ "grad_norm": 3.015228033065796,
28
+ "learning_rate": 2e-05,
29
+ "loss": 1.5356,
30
+ "step": 6
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "grad_norm": 0.681362509727478,
35
+ "learning_rate": 2e-05,
36
+ "loss": 0.382,
37
+ "step": 8
38
+ },
39
+ {
40
+ "epoch": 0.1,
41
+ "grad_norm": 2.733452558517456,
42
+ "learning_rate": 2e-05,
43
+ "loss": 0.8807,
44
+ "step": 10
45
+ },
46
+ {
47
+ "epoch": 0.12,
48
+ "grad_norm": 1.2124817371368408,
49
+ "learning_rate": 2e-05,
50
+ "loss": 0.61,
51
+ "step": 12
52
+ },
53
+ {
54
+ "epoch": 0.14,
55
+ "grad_norm": 0.5142900347709656,
56
+ "learning_rate": 2e-05,
57
+ "loss": 0.3694,
58
+ "step": 14
59
+ },
60
+ {
61
+ "epoch": 0.16,
62
+ "grad_norm": 3.4124550819396973,
63
+ "learning_rate": 2e-05,
64
+ "loss": 1.1068,
65
+ "step": 16
66
+ },
67
+ {
68
+ "epoch": 0.18,
69
+ "grad_norm": 0.49465346336364746,
70
+ "learning_rate": 2e-05,
71
+ "loss": 1.0515,
72
+ "step": 18
73
+ },
74
+ {
75
+ "epoch": 0.2,
76
+ "grad_norm": 2.906670570373535,
77
+ "learning_rate": 2e-05,
78
+ "loss": 0.574,
79
+ "step": 20
80
+ },
81
+ {
82
+ "epoch": 0.22,
83
+ "grad_norm": 3.5659115314483643,
84
+ "learning_rate": 2e-05,
85
+ "loss": 0.9978,
86
+ "step": 22
87
+ },
88
+ {
89
+ "epoch": 0.24,
90
+ "grad_norm": 2.965306520462036,
91
+ "learning_rate": 2e-05,
92
+ "loss": 1.4742,
93
+ "step": 24
94
+ },
95
+ {
96
+ "epoch": 0.26,
97
+ "grad_norm": 4.656124591827393,
98
+ "learning_rate": 2e-05,
99
+ "loss": 2.2981,
100
+ "step": 26
101
+ },
102
+ {
103
+ "epoch": 0.28,
104
+ "grad_norm": 4.481357097625732,
105
+ "learning_rate": 2e-05,
106
+ "loss": 1.6502,
107
+ "step": 28
108
+ },
109
+ {
110
+ "epoch": 0.3,
111
+ "grad_norm": 3.6990504264831543,
112
+ "learning_rate": 2e-05,
113
+ "loss": 2.1989,
114
+ "step": 30
115
+ },
116
+ {
117
+ "epoch": 0.32,
118
+ "grad_norm": 2.218642234802246,
119
+ "learning_rate": 2e-05,
120
+ "loss": 0.6695,
121
+ "step": 32
122
+ },
123
+ {
124
+ "epoch": 0.34,
125
+ "grad_norm": 1.512957215309143,
126
+ "learning_rate": 2e-05,
127
+ "loss": 1.0031,
128
+ "step": 34
129
+ },
130
+ {
131
+ "epoch": 0.36,
132
+ "grad_norm": 1.6833781003952026,
133
+ "learning_rate": 2e-05,
134
+ "loss": 1.9793,
135
+ "step": 36
136
+ },
137
+ {
138
+ "epoch": 0.38,
139
+ "grad_norm": 1.107330083847046,
140
+ "learning_rate": 2e-05,
141
+ "loss": 0.3538,
142
+ "step": 38
143
+ },
144
+ {
145
+ "epoch": 0.4,
146
+ "grad_norm": 3.9457249641418457,
147
+ "learning_rate": 2e-05,
148
+ "loss": 1.4415,
149
+ "step": 40
150
+ },
151
+ {
152
+ "epoch": 0.42,
153
+ "grad_norm": 0.051738329231739044,
154
+ "learning_rate": 2e-05,
155
+ "loss": 0.4044,
156
+ "step": 42
157
+ },
158
+ {
159
+ "epoch": 0.44,
160
+ "grad_norm": 2.370415687561035,
161
+ "learning_rate": 2e-05,
162
+ "loss": 1.8381,
163
+ "step": 44
164
+ },
165
+ {
166
+ "epoch": 0.46,
167
+ "grad_norm": 1.5636987686157227,
168
+ "learning_rate": 2e-05,
169
+ "loss": 1.1786,
170
+ "step": 46
171
+ },
172
+ {
173
+ "epoch": 0.48,
174
+ "grad_norm": 1.8076434135437012,
175
+ "learning_rate": 2e-05,
176
+ "loss": 0.9974,
177
+ "step": 48
178
+ },
179
+ {
180
+ "epoch": 0.5,
181
+ "grad_norm": 1.4153037071228027,
182
+ "learning_rate": 2e-05,
183
+ "loss": 0.7157,
184
+ "step": 50
185
+ },
186
+ {
187
+ "epoch": 0.52,
188
+ "grad_norm": 1.8657974004745483,
189
+ "learning_rate": 2e-05,
190
+ "loss": 2.0883,
191
+ "step": 52
192
+ },
193
+ {
194
+ "epoch": 0.54,
195
+ "grad_norm": 0.06698138266801834,
196
+ "learning_rate": 2e-05,
197
+ "loss": 0.7553,
198
+ "step": 54
199
+ },
200
+ {
201
+ "epoch": 0.56,
202
+ "grad_norm": 1.0162607431411743,
203
+ "learning_rate": 2e-05,
204
+ "loss": 0.3858,
205
+ "step": 56
206
+ },
207
+ {
208
+ "epoch": 0.58,
209
+ "grad_norm": 1.3238625526428223,
210
+ "learning_rate": 2e-05,
211
+ "loss": 0.9815,
212
+ "step": 58
213
+ },
214
+ {
215
+ "epoch": 0.6,
216
+ "grad_norm": 1.3448885679244995,
217
+ "learning_rate": 2e-05,
218
+ "loss": 0.7539,
219
+ "step": 60
220
+ },
221
+ {
222
+ "epoch": 0.62,
223
+ "grad_norm": 1.4804365634918213,
224
+ "learning_rate": 2e-05,
225
+ "loss": 0.8375,
226
+ "step": 62
227
+ },
228
+ {
229
+ "epoch": 0.64,
230
+ "grad_norm": 2.679854393005371,
231
+ "learning_rate": 2e-05,
232
+ "loss": 1.5526,
233
+ "step": 64
234
+ },
235
+ {
236
+ "epoch": 0.66,
237
+ "grad_norm": 1.697770595550537,
238
+ "learning_rate": 2e-05,
239
+ "loss": 1.6394,
240
+ "step": 66
241
+ },
242
+ {
243
+ "epoch": 0.68,
244
+ "grad_norm": 0.08173129707574844,
245
+ "learning_rate": 2e-05,
246
+ "loss": 0.3275,
247
+ "step": 68
248
+ },
249
+ {
250
+ "epoch": 0.7,
251
+ "grad_norm": 1.562013864517212,
252
+ "learning_rate": 2e-05,
253
+ "loss": 1.2811,
254
+ "step": 70
255
+ },
256
+ {
257
+ "epoch": 0.72,
258
+ "grad_norm": 1.6425532102584839,
259
+ "learning_rate": 2e-05,
260
+ "loss": 0.8023,
261
+ "step": 72
262
+ },
263
+ {
264
+ "epoch": 0.74,
265
+ "grad_norm": 1.0354465246200562,
266
+ "learning_rate": 2e-05,
267
+ "loss": 1.1036,
268
+ "step": 74
269
+ },
270
+ {
271
+ "epoch": 0.76,
272
+ "grad_norm": 2.0127949714660645,
273
+ "learning_rate": 2e-05,
274
+ "loss": 1.0241,
275
+ "step": 76
276
+ },
277
+ {
278
+ "epoch": 0.78,
279
+ "grad_norm": 1.767486572265625,
280
+ "learning_rate": 2e-05,
281
+ "loss": 1.6514,
282
+ "step": 78
283
+ },
284
+ {
285
+ "epoch": 0.8,
286
+ "grad_norm": 3.19219708442688,
287
+ "learning_rate": 2e-05,
288
+ "loss": 1.3129,
289
+ "step": 80
290
+ },
291
+ {
292
+ "epoch": 0.82,
293
+ "grad_norm": 3.8570995330810547,
294
+ "learning_rate": 2e-05,
295
+ "loss": 2.3893,
296
+ "step": 82
297
+ },
298
+ {
299
+ "epoch": 0.84,
300
+ "grad_norm": 4.0963134765625,
301
+ "learning_rate": 2e-05,
302
+ "loss": 1.9847,
303
+ "step": 84
304
+ },
305
+ {
306
+ "epoch": 0.86,
307
+ "grad_norm": 1.6539266109466553,
308
+ "learning_rate": 2e-05,
309
+ "loss": 0.8877,
310
+ "step": 86
311
+ },
312
+ {
313
+ "epoch": 0.88,
314
+ "grad_norm": 1.8997998237609863,
315
+ "learning_rate": 2e-05,
316
+ "loss": 1.6619,
317
+ "step": 88
318
+ },
319
+ {
320
+ "epoch": 0.9,
321
+ "grad_norm": 1.5609619617462158,
322
+ "learning_rate": 2e-05,
323
+ "loss": 0.679,
324
+ "step": 90
325
+ },
326
+ {
327
+ "epoch": 0.92,
328
+ "grad_norm": 2.3895843029022217,
329
+ "learning_rate": 2e-05,
330
+ "loss": 1.4562,
331
+ "step": 92
332
+ },
333
+ {
334
+ "epoch": 0.94,
335
+ "grad_norm": 1.3494317531585693,
336
+ "learning_rate": 2e-05,
337
+ "loss": 0.6015,
338
+ "step": 94
339
+ },
340
+ {
341
+ "epoch": 0.96,
342
+ "grad_norm": 1.4236085414886475,
343
+ "learning_rate": 2e-05,
344
+ "loss": 1.3893,
345
+ "step": 96
346
+ },
347
+ {
348
+ "epoch": 0.98,
349
+ "grad_norm": 2.681549310684204,
350
+ "learning_rate": 2e-05,
351
+ "loss": 0.8135,
352
+ "step": 98
353
+ },
354
+ {
355
+ "epoch": 1.0,
356
+ "grad_norm": 1.336864948272705,
357
+ "learning_rate": 2e-05,
358
+ "loss": 1.0993,
359
+ "step": 100
360
+ },
361
+ {
362
+ "epoch": 1.0,
363
+ "step": 100,
364
+ "total_flos": 2.034972181397504e+16,
365
+ "train_loss": 1.1149367547035218,
366
+ "train_runtime": 565.793,
367
+ "train_samples_per_second": 0.707,
368
+ "train_steps_per_second": 0.177
369
+ }
370
+ ],
371
+ "logging_steps": 2,
372
+ "max_steps": 100,
373
+ "num_input_tokens_seen": 0,
374
+ "num_train_epochs": 1,
375
+ "save_steps": 500,
376
+ "stateful_callbacks": {
377
+ "TrainerControl": {
378
+ "args": {
379
+ "should_epoch_stop": false,
380
+ "should_evaluate": false,
381
+ "should_log": false,
382
+ "should_save": false,
383
+ "should_training_stop": false
384
+ },
385
+ "attributes": {}
386
+ }
387
+ },
388
+ "total_flos": 2.034972181397504e+16,
389
+ "train_batch_size": 1,
390
+ "trial_name": null,
391
+ "trial_params": null
392
+ }
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round10.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02b13e3ee14861a2302990aeadbe1e082ca2221d7208ed3c4a4d638fdde9d873
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round12.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2fddc721552f128395f74bf1358fb109ae8b6d949fb1bcc021c8d8f2eb50f53
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round15.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11e65c6af34bd3f9c4ff5d1eb16a8269f4b73eab559d293daa1245e243d42ea7
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round17.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d1bf5876493176d2a8ccb81e7af432ce52bf82a3399c59c18498832c1837e3f
3
+ size 794708086
client_states_feddualMultipqfullfreeze_homoAgg_moe_CCCC_pca_T05_bs4_saveoptim_lr2e-5_5e-5_sc132_4tasks_5rounds_fixitr100_T0125_decay099/5_client_model_round2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69d316fc4776caecfcd04561104c536ff710f255c7d78a2aa1e9d94a30f5c317
3
+ size 794706058