End of training
Browse files
all_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch": 1.
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second": 5.
|
7 |
-
"eval_steps_per_second": 1.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_accuracy": 0.8129032258064516,
|
4 |
+
"eval_loss": 0.5109545588493347,
|
5 |
+
"eval_runtime": 27.7194,
|
6 |
+
"eval_samples_per_second": 5.592,
|
7 |
+
"eval_steps_per_second": 1.407
|
8 |
}
|
runs/May22_12-56-09_940520f3922b/events.out.tfevents.1716382778.940520f3922b.34.3
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f0a0fbfa3a79d6dbf24fc976de7897320a990513c19b89ac9c024b5476f9d0b9
|
3 |
+
size 722
|
test_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch": 1.
|
3 |
-
"eval_accuracy": 0.
|
4 |
-
"eval_loss": 0.
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second": 5.
|
7 |
-
"eval_steps_per_second": 1.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 1.0,
|
3 |
+
"eval_accuracy": 0.8129032258064516,
|
4 |
+
"eval_loss": 0.5109545588493347,
|
5 |
+
"eval_runtime": 27.7194,
|
6 |
+
"eval_samples_per_second": 5.592,
|
7 |
+
"eval_steps_per_second": 1.407
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,170 +1,105 @@
|
|
1 |
{
|
2 |
-
"best_metric": 0.
|
3 |
-
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-
|
4 |
-
"epoch": 1.
|
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":
|
15 |
-
"loss":
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm":
|
21 |
-
"learning_rate": 4.
|
22 |
-
"loss":
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm":
|
28 |
-
"learning_rate":
|
29 |
-
"loss":
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
-
"epoch": 0.
|
34 |
-
"grad_norm":
|
35 |
-
"learning_rate":
|
36 |
-
"loss":
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
-
"epoch": 0.
|
41 |
-
"grad_norm":
|
42 |
-
"learning_rate":
|
43 |
-
"loss":
|
44 |
"step": 50
|
45 |
},
|
46 |
{
|
47 |
-
"epoch": 0.
|
48 |
-
"grad_norm":
|
49 |
-
"learning_rate":
|
50 |
-
"loss":
|
51 |
"step": 60
|
52 |
},
|
53 |
{
|
54 |
-
"epoch": 0.
|
55 |
-
"grad_norm":
|
56 |
-
"learning_rate":
|
57 |
-
"loss":
|
58 |
"step": 70
|
59 |
},
|
60 |
{
|
61 |
-
"epoch": 0
|
62 |
-
"eval_accuracy": 0.
|
63 |
-
"eval_loss":
|
64 |
-
"eval_runtime":
|
65 |
-
"eval_samples_per_second": 5.
|
66 |
-
"eval_steps_per_second": 1.
|
67 |
"step": 75
|
68 |
},
|
69 |
{
|
70 |
-
"epoch": 1.
|
71 |
-
"
|
72 |
-
"
|
73 |
-
"
|
74 |
-
"
|
75 |
-
|
76 |
-
|
77 |
-
"epoch": 1.1,
|
78 |
-
"grad_norm": 12.44357967376709,
|
79 |
-
"learning_rate": 2.2222222222222223e-05,
|
80 |
-
"loss": 1.2583,
|
81 |
-
"step": 90
|
82 |
-
},
|
83 |
-
{
|
84 |
-
"epoch": 1.17,
|
85 |
-
"grad_norm": 14.58968448638916,
|
86 |
-
"learning_rate": 1.8518518518518518e-05,
|
87 |
-
"loss": 1.1709,
|
88 |
-
"step": 100
|
89 |
-
},
|
90 |
-
{
|
91 |
-
"epoch": 1.23,
|
92 |
-
"grad_norm": 8.881256103515625,
|
93 |
-
"learning_rate": 1.4814814814814815e-05,
|
94 |
-
"loss": 0.9056,
|
95 |
-
"step": 110
|
96 |
-
},
|
97 |
-
{
|
98 |
-
"epoch": 1.3,
|
99 |
-
"grad_norm": 11.71822738647461,
|
100 |
-
"learning_rate": 1.1111111111111112e-05,
|
101 |
-
"loss": 0.9813,
|
102 |
-
"step": 120
|
103 |
-
},
|
104 |
-
{
|
105 |
-
"epoch": 1.37,
|
106 |
-
"grad_norm": 5.7398858070373535,
|
107 |
-
"learning_rate": 7.4074074074074075e-06,
|
108 |
-
"loss": 0.709,
|
109 |
-
"step": 130
|
110 |
-
},
|
111 |
-
{
|
112 |
-
"epoch": 1.43,
|
113 |
-
"grad_norm": 9.218167304992676,
|
114 |
-
"learning_rate": 3.7037037037037037e-06,
|
115 |
-
"loss": 0.6339,
|
116 |
-
"step": 140
|
117 |
-
},
|
118 |
-
{
|
119 |
-
"epoch": 1.5,
|
120 |
-
"grad_norm": 35.40492630004883,
|
121 |
-
"learning_rate": 0.0,
|
122 |
-
"loss": 0.895,
|
123 |
-
"step": 150
|
124 |
},
|
125 |
{
|
126 |
-
"epoch": 1.
|
127 |
-
"eval_accuracy": 0.
|
128 |
-
"eval_loss": 0.
|
129 |
-
"eval_runtime":
|
130 |
-
"eval_samples_per_second": 5.
|
131 |
-
"eval_steps_per_second": 1.
|
132 |
-
"step":
|
133 |
-
},
|
134 |
-
{
|
135 |
-
"epoch": 1.5,
|
136 |
-
"step": 150,
|
137 |
-
"total_flos": 7.47692094062592e+17,
|
138 |
-
"train_loss": 1.4997930717468262,
|
139 |
-
"train_runtime": 266.4694,
|
140 |
-
"train_samples_per_second": 2.252,
|
141 |
-
"train_steps_per_second": 0.563
|
142 |
-
},
|
143 |
-
{
|
144 |
-
"epoch": 1.5,
|
145 |
-
"eval_accuracy": 0.7548387096774194,
|
146 |
-
"eval_loss": 0.8304252624511719,
|
147 |
-
"eval_runtime": 29.2907,
|
148 |
-
"eval_samples_per_second": 5.292,
|
149 |
-
"eval_steps_per_second": 1.331,
|
150 |
-
"step": 150
|
151 |
},
|
152 |
{
|
153 |
-
"epoch": 1.
|
154 |
-
"eval_accuracy": 0.
|
155 |
-
"eval_loss": 0.
|
156 |
-
"eval_runtime":
|
157 |
-
"eval_samples_per_second": 5.
|
158 |
-
"eval_steps_per_second": 1.
|
159 |
-
"step":
|
160 |
}
|
161 |
],
|
162 |
"logging_steps": 10,
|
163 |
-
"max_steps":
|
164 |
"num_input_tokens_seen": 0,
|
165 |
"num_train_epochs": 9223372036854775807,
|
166 |
"save_steps": 500,
|
167 |
-
"total_flos":
|
168 |
"train_batch_size": 4,
|
169 |
"trial_name": null,
|
170 |
"trial_params": null
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.7142857142857143,
|
3 |
+
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-75",
|
4 |
+
"epoch": 1.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 75,
|
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,
|
13 |
+
"grad_norm": 10.564898490905762,
|
14 |
+
"learning_rate": 4.850746268656717e-05,
|
15 |
+
"loss": 0.8833,
|
16 |
"step": 10
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.27,
|
20 |
+
"grad_norm": 20.67559051513672,
|
21 |
+
"learning_rate": 4.104477611940299e-05,
|
22 |
+
"loss": 0.9306,
|
23 |
"step": 20
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.4,
|
27 |
+
"grad_norm": 16.63642120361328,
|
28 |
+
"learning_rate": 3.358208955223881e-05,
|
29 |
+
"loss": 0.7969,
|
30 |
"step": 30
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.53,
|
34 |
+
"grad_norm": 3.232578992843628,
|
35 |
+
"learning_rate": 2.6119402985074626e-05,
|
36 |
+
"loss": 0.8249,
|
37 |
"step": 40
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.67,
|
41 |
+
"grad_norm": 16.193866729736328,
|
42 |
+
"learning_rate": 1.865671641791045e-05,
|
43 |
+
"loss": 0.7562,
|
44 |
"step": 50
|
45 |
},
|
46 |
{
|
47 |
+
"epoch": 0.8,
|
48 |
+
"grad_norm": 2.896273374557495,
|
49 |
+
"learning_rate": 1.119402985074627e-05,
|
50 |
+
"loss": 0.59,
|
51 |
"step": 60
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 0.93,
|
55 |
+
"grad_norm": 5.816062927246094,
|
56 |
+
"learning_rate": 3.7313432835820893e-06,
|
57 |
+
"loss": 0.4871,
|
58 |
"step": 70
|
59 |
},
|
60 |
{
|
61 |
+
"epoch": 1.0,
|
62 |
+
"eval_accuracy": 0.7142857142857143,
|
63 |
+
"eval_loss": 0.7370076775550842,
|
64 |
+
"eval_runtime": 12.5539,
|
65 |
+
"eval_samples_per_second": 5.576,
|
66 |
+
"eval_steps_per_second": 1.434,
|
67 |
"step": 75
|
68 |
},
|
69 |
{
|
70 |
+
"epoch": 1.0,
|
71 |
+
"step": 75,
|
72 |
+
"total_flos": 3.73846047031296e+17,
|
73 |
+
"train_loss": 0.7330940500895182,
|
74 |
+
"train_runtime": 114.9093,
|
75 |
+
"train_samples_per_second": 2.611,
|
76 |
+
"train_steps_per_second": 0.653
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
},
|
78 |
{
|
79 |
+
"epoch": 1.0,
|
80 |
+
"eval_accuracy": 0.8129032258064516,
|
81 |
+
"eval_loss": 0.5109546184539795,
|
82 |
+
"eval_runtime": 28.4314,
|
83 |
+
"eval_samples_per_second": 5.452,
|
84 |
+
"eval_steps_per_second": 1.372,
|
85 |
+
"step": 75
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
86 |
},
|
87 |
{
|
88 |
+
"epoch": 1.0,
|
89 |
+
"eval_accuracy": 0.8129032258064516,
|
90 |
+
"eval_loss": 0.5109545588493347,
|
91 |
+
"eval_runtime": 27.7194,
|
92 |
+
"eval_samples_per_second": 5.592,
|
93 |
+
"eval_steps_per_second": 1.407,
|
94 |
+
"step": 75
|
95 |
}
|
96 |
],
|
97 |
"logging_steps": 10,
|
98 |
+
"max_steps": 75,
|
99 |
"num_input_tokens_seen": 0,
|
100 |
"num_train_epochs": 9223372036854775807,
|
101 |
"save_steps": 500,
|
102 |
+
"total_flos": 3.73846047031296e+17,
|
103 |
"train_batch_size": 4,
|
104 |
"trial_name": null,
|
105 |
"trial_params": null
|