amauriciogonzalez
commited on
Commit
•
6187fdb
1
Parent(s):
27b99d3
End of training
Browse files- all_results.json +11 -11
- eval_results.json +6 -6
- runs/Sep06_22-24-21_2de8c0b5efc2/events.out.tfevents.1725661898.2de8c0b5efc2.36.6 +3 -0
- train_results.json +6 -6
- trainer_state.json +37 -104
all_results.json
CHANGED
@@ -1,13 +1,13 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 1.
|
8 |
-
"total_flos":
|
9 |
-
"train_loss":
|
10 |
-
"train_runtime":
|
11 |
-
"train_samples_per_second":
|
12 |
-
"train_steps_per_second": 0.
|
13 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_accuracy": 0.8634920634920635,
|
4 |
+
"eval_loss": 0.44639232754707336,
|
5 |
+
"eval_runtime": 28.206,
|
6 |
+
"eval_samples_per_second": 67.007,
|
7 |
+
"eval_steps_per_second": 1.064,
|
8 |
+
"total_flos": 1.0941516662243328e+18,
|
9 |
+
"train_loss": 1.2471325397491455,
|
10 |
+
"train_runtime": 397.8778,
|
11 |
+
"train_samples_per_second": 26.918,
|
12 |
+
"train_steps_per_second": 0.106
|
13 |
}
|
eval_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second":
|
7 |
-
"eval_steps_per_second": 1.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_accuracy": 0.8634920634920635,
|
4 |
+
"eval_loss": 0.44639232754707336,
|
5 |
+
"eval_runtime": 28.206,
|
6 |
+
"eval_samples_per_second": 67.007,
|
7 |
+
"eval_steps_per_second": 1.064
|
8 |
}
|
runs/Sep06_22-24-21_2de8c0b5efc2/events.out.tfevents.1725661898.2de8c0b5efc2.36.6
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e8983b84d875ab0888b68e5a5f73192c6c50d04509d3951ddac0472a9a61f203
|
3 |
+
size 405
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"total_flos":
|
4 |
-
"train_loss":
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"total_flos": 1.0941516662243328e+18,
|
4 |
+
"train_loss": 1.2471325397491455,
|
5 |
+
"train_runtime": 397.8778,
|
6 |
+
"train_samples_per_second": 26.918,
|
7 |
+
"train_steps_per_second": 0.106
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,131 +1,64 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "
|
4 |
-
"epoch":
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
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.
|
13 |
-
"grad_norm":
|
14 |
-
"learning_rate": 4.
|
15 |
-
"loss": 2.
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm":
|
21 |
-
"learning_rate":
|
22 |
-
"loss": 1.
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm":
|
28 |
-
"learning_rate":
|
29 |
-
"loss": 0.
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
-
"epoch": 0.
|
34 |
-
"
|
35 |
-
"
|
36 |
-
"
|
37 |
-
"eval_samples_per_second": 48.246,
|
38 |
-
"eval_steps_per_second": 0.766,
|
39 |
-
"step": 37
|
40 |
-
},
|
41 |
-
{
|
42 |
-
"epoch": 1.0596026490066226,
|
43 |
-
"grad_norm": 22.94610023498535,
|
44 |
-
"learning_rate": 3.5858585858585855e-05,
|
45 |
-
"loss": 0.863,
|
46 |
"step": 40
|
47 |
},
|
48 |
{
|
49 |
-
"epoch": 1.
|
50 |
-
"
|
51 |
-
"
|
52 |
-
"
|
53 |
-
"
|
54 |
-
|
55 |
-
|
56 |
-
"epoch": 1.589403973509934,
|
57 |
-
"grad_norm": 16.806970596313477,
|
58 |
-
"learning_rate": 2.575757575757576e-05,
|
59 |
-
"loss": 0.7287,
|
60 |
-
"step": 60
|
61 |
-
},
|
62 |
-
{
|
63 |
-
"epoch": 1.8543046357615895,
|
64 |
-
"grad_norm": 16.49362564086914,
|
65 |
-
"learning_rate": 2.070707070707071e-05,
|
66 |
-
"loss": 0.7052,
|
67 |
-
"step": 70
|
68 |
-
},
|
69 |
-
{
|
70 |
-
"epoch": 1.9867549668874172,
|
71 |
-
"eval_accuracy": 0.8468720821661998,
|
72 |
-
"eval_loss": 0.4805677831172943,
|
73 |
-
"eval_runtime": 15.9267,
|
74 |
-
"eval_samples_per_second": 67.246,
|
75 |
-
"eval_steps_per_second": 1.067,
|
76 |
-
"step": 75
|
77 |
-
},
|
78 |
-
{
|
79 |
-
"epoch": 2.119205298013245,
|
80 |
-
"grad_norm": 22.60504722595215,
|
81 |
-
"learning_rate": 1.565656565656566e-05,
|
82 |
-
"loss": 0.6016,
|
83 |
-
"step": 80
|
84 |
-
},
|
85 |
-
{
|
86 |
-
"epoch": 2.384105960264901,
|
87 |
-
"grad_norm": 15.832114219665527,
|
88 |
-
"learning_rate": 1.0606060606060607e-05,
|
89 |
-
"loss": 0.5896,
|
90 |
-
"step": 90
|
91 |
-
},
|
92 |
-
{
|
93 |
-
"epoch": 2.6490066225165565,
|
94 |
-
"grad_norm": 12.651689529418945,
|
95 |
-
"learning_rate": 5.555555555555556e-06,
|
96 |
-
"loss": 0.5613,
|
97 |
-
"step": 100
|
98 |
-
},
|
99 |
-
{
|
100 |
-
"epoch": 2.9139072847682117,
|
101 |
-
"grad_norm": 17.284223556518555,
|
102 |
-
"learning_rate": 5.050505050505052e-07,
|
103 |
-
"loss": 0.5414,
|
104 |
-
"step": 110
|
105 |
-
},
|
106 |
-
{
|
107 |
-
"epoch": 2.9403973509933774,
|
108 |
-
"eval_accuracy": 0.8683473389355743,
|
109 |
-
"eval_loss": 0.4009244740009308,
|
110 |
-
"eval_runtime": 15.8918,
|
111 |
-
"eval_samples_per_second": 67.393,
|
112 |
-
"eval_steps_per_second": 1.07,
|
113 |
-
"step": 111
|
114 |
},
|
115 |
{
|
116 |
-
"epoch":
|
117 |
-
"step":
|
118 |
-
"total_flos":
|
119 |
-
"train_loss":
|
120 |
-
"train_runtime":
|
121 |
-
"train_samples_per_second":
|
122 |
-
"train_steps_per_second": 0.
|
123 |
}
|
124 |
],
|
125 |
"logging_steps": 10,
|
126 |
-
"max_steps":
|
127 |
"num_input_tokens_seen": 0,
|
128 |
-
"num_train_epochs":
|
129 |
"save_steps": 500,
|
130 |
"stateful_callbacks": {
|
131 |
"TrainerControl": {
|
@@ -139,7 +72,7 @@
|
|
139 |
"attributes": {}
|
140 |
}
|
141 |
},
|
142 |
-
"total_flos":
|
143 |
"train_batch_size": 64,
|
144 |
"trial_name": null,
|
145 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.8634920634920635,
|
3 |
+
"best_model_checkpoint": "dinov2-base-finetuned-har/checkpoint-42",
|
4 |
+
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 42,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.23809523809523808,
|
13 |
+
"grad_norm": 44.80964660644531,
|
14 |
+
"learning_rate": 4.324324324324325e-05,
|
15 |
+
"loss": 2.2183,
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.47619047619047616,
|
20 |
+
"grad_norm": 20.761051177978516,
|
21 |
+
"learning_rate": 2.9729729729729733e-05,
|
22 |
+
"loss": 1.1574,
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.7142857142857143,
|
27 |
+
"grad_norm": 28.136566162109375,
|
28 |
+
"learning_rate": 1.6216216216216218e-05,
|
29 |
+
"loss": 0.8864,
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.9523809523809523,
|
34 |
+
"grad_norm": 15.380955696105957,
|
35 |
+
"learning_rate": 2.702702702702703e-06,
|
36 |
+
"loss": 0.8084,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 1.0,
|
41 |
+
"eval_accuracy": 0.8634920634920635,
|
42 |
+
"eval_loss": 0.44639232754707336,
|
43 |
+
"eval_runtime": 28.1795,
|
44 |
+
"eval_samples_per_second": 67.07,
|
45 |
+
"eval_steps_per_second": 1.065,
|
46 |
+
"step": 42
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
47 |
},
|
48 |
{
|
49 |
+
"epoch": 1.0,
|
50 |
+
"step": 42,
|
51 |
+
"total_flos": 1.0941516662243328e+18,
|
52 |
+
"train_loss": 1.2471325397491455,
|
53 |
+
"train_runtime": 397.8778,
|
54 |
+
"train_samples_per_second": 26.918,
|
55 |
+
"train_steps_per_second": 0.106
|
56 |
}
|
57 |
],
|
58 |
"logging_steps": 10,
|
59 |
+
"max_steps": 42,
|
60 |
"num_input_tokens_seen": 0,
|
61 |
+
"num_train_epochs": 1,
|
62 |
"save_steps": 500,
|
63 |
"stateful_callbacks": {
|
64 |
"TrainerControl": {
|
|
|
72 |
"attributes": {}
|
73 |
}
|
74 |
},
|
75 |
+
"total_flos": 1.0941516662243328e+18,
|
76 |
"train_batch_size": 64,
|
77 |
"trial_name": null,
|
78 |
"trial_params": null
|