End of training
Browse files- README.md +2 -2
- all_results.json +6 -6
- test_results.json +6 -6
- trainer_state.json +1109 -96
README.md
CHANGED
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
-
- Loss:
|
21 |
-
- Accuracy: 0.
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
17 |
|
18 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 1.8346
|
21 |
+
- Accuracy: 0.6275
|
22 |
|
23 |
## Model description
|
24 |
|
all_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy":
|
4 |
-
"eval_loss":
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second": 0.
|
7 |
-
"eval_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 29.01,
|
3 |
+
"eval_accuracy": 0.6274509803921569,
|
4 |
+
"eval_loss": 1.834560751914978,
|
5 |
+
"eval_runtime": 132.9753,
|
6 |
+
"eval_samples_per_second": 0.384,
|
7 |
+
"eval_steps_per_second": 0.196
|
8 |
}
|
test_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
-
"epoch":
|
3 |
-
"eval_accuracy":
|
4 |
-
"eval_loss":
|
5 |
-
"eval_runtime":
|
6 |
-
"eval_samples_per_second": 0.
|
7 |
-
"eval_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
+
"epoch": 29.01,
|
3 |
+
"eval_accuracy": 0.6274509803921569,
|
4 |
+
"eval_loss": 1.834560751914978,
|
5 |
+
"eval_runtime": 132.9753,
|
6 |
+
"eval_samples_per_second": 0.384,
|
7 |
+
"eval_steps_per_second": 0.196
|
8 |
}
|
trainer_state.json
CHANGED
@@ -1,136 +1,1149 @@
|
|
1 |
{
|
2 |
-
"best_metric":
|
3 |
-
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset-SBDtoy/checkpoint-
|
4 |
-
"epoch":
|
5 |
-
"
|
|
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
9 |
"log_history": [
|
10 |
{
|
11 |
-
"epoch": 0.
|
12 |
-
"
|
13 |
-
"
|
14 |
-
"
|
15 |
-
"
|
16 |
-
|
17 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
18 |
},
|
19 |
{
|
20 |
-
"epoch":
|
21 |
-
"
|
22 |
-
"
|
23 |
-
"
|
24 |
-
"
|
25 |
-
"eval_steps_per_second": 0.475,
|
26 |
-
"step": 4
|
27 |
},
|
28 |
{
|
29 |
-
"epoch":
|
30 |
-
"eval_accuracy":
|
31 |
-
"eval_loss":
|
32 |
-
"eval_runtime":
|
33 |
-
"eval_samples_per_second": 0.
|
34 |
-
"eval_steps_per_second": 0.
|
35 |
-
"step":
|
36 |
},
|
37 |
{
|
38 |
-
"epoch":
|
39 |
-
"
|
40 |
-
"
|
41 |
-
"
|
42 |
-
"
|
43 |
-
"eval_steps_per_second": 0.45,
|
44 |
-
"step": 8
|
45 |
},
|
46 |
{
|
47 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
"learning_rate": 2.777777777777778e-05,
|
49 |
-
"loss": 0.
|
50 |
-
"step":
|
51 |
},
|
52 |
{
|
53 |
-
"epoch":
|
54 |
-
"
|
55 |
-
"
|
56 |
-
"
|
57 |
-
"
|
58 |
-
"eval_steps_per_second": 0.41,
|
59 |
-
"step": 10
|
60 |
},
|
61 |
{
|
62 |
-
"epoch":
|
63 |
-
"eval_accuracy": 0.
|
64 |
-
"eval_loss":
|
65 |
-
"eval_runtime":
|
66 |
-
"eval_samples_per_second": 0.
|
67 |
-
"eval_steps_per_second": 0.
|
68 |
-
"step":
|
69 |
},
|
70 |
{
|
71 |
-
"epoch":
|
72 |
-
"
|
73 |
-
"
|
74 |
-
"
|
75 |
-
"
|
76 |
-
"eval_steps_per_second": 0.439,
|
77 |
-
"step": 14
|
78 |
},
|
79 |
{
|
80 |
-
"epoch":
|
81 |
-
"
|
82 |
-
"
|
83 |
-
"
|
84 |
-
"
|
85 |
-
"eval_steps_per_second": 0.485,
|
86 |
-
"step": 16
|
87 |
},
|
88 |
{
|
89 |
-
"epoch":
|
90 |
-
"
|
91 |
-
"
|
92 |
-
"
|
93 |
-
"
|
94 |
-
"eval_steps_per_second": 0.452,
|
95 |
-
"step": 18
|
96 |
},
|
97 |
{
|
98 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
99 |
"learning_rate": 0.0,
|
100 |
-
"loss": 0.
|
101 |
-
"step":
|
102 |
},
|
103 |
{
|
104 |
-
"epoch":
|
105 |
-
"eval_accuracy": 0.
|
106 |
-
"eval_loss":
|
107 |
-
"eval_runtime":
|
108 |
-
"eval_samples_per_second": 0.
|
109 |
-
"eval_steps_per_second": 0.
|
110 |
-
"step":
|
111 |
},
|
112 |
{
|
113 |
-
"epoch":
|
114 |
-
"step":
|
115 |
-
"total_flos": 2.
|
116 |
-
"train_loss": 0.
|
117 |
-
"train_runtime":
|
118 |
-
"train_samples_per_second": 0.
|
119 |
-
"train_steps_per_second": 0.
|
120 |
},
|
121 |
{
|
122 |
-
"epoch":
|
123 |
-
"eval_accuracy":
|
124 |
-
"eval_loss":
|
125 |
-
"eval_runtime":
|
126 |
-
"eval_samples_per_second": 0.
|
127 |
-
"eval_steps_per_second": 0.
|
128 |
-
"step":
|
129 |
}
|
130 |
],
|
131 |
-
"
|
|
|
|
|
132 |
"num_train_epochs": 9223372036854775807,
|
133 |
-
"
|
|
|
|
|
134 |
"trial_name": null,
|
135 |
"trial_params": null
|
136 |
}
|
|
|
1 |
{
|
2 |
+
"best_metric": 0.6274509803921569,
|
3 |
+
"best_model_checkpoint": "/users/hailey/code/03_code/SBD_ai/videomae-base-finetuned-ucf101-subset-SBDtoy/checkpoint-492",
|
4 |
+
"epoch": 29.009166666666665,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1200,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.01,
|
13 |
+
"grad_norm": 8.483016014099121,
|
14 |
+
"learning_rate": 4.166666666666667e-06,
|
15 |
+
"loss": 0.8897,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.02,
|
20 |
+
"grad_norm": 10.62829303741455,
|
21 |
+
"learning_rate": 8.333333333333334e-06,
|
22 |
+
"loss": 0.7501,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.03,
|
27 |
+
"grad_norm": 10.224696159362793,
|
28 |
+
"learning_rate": 1.25e-05,
|
29 |
+
"loss": 0.6748,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.03,
|
34 |
+
"grad_norm": 5.6519365310668945,
|
35 |
+
"learning_rate": 1.6666666666666667e-05,
|
36 |
+
"loss": 0.7634,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.03,
|
41 |
+
"eval_accuracy": 0.5882352941176471,
|
42 |
+
"eval_loss": 0.6320443153381348,
|
43 |
+
"eval_runtime": 137.732,
|
44 |
+
"eval_samples_per_second": 0.37,
|
45 |
+
"eval_steps_per_second": 0.189,
|
46 |
+
"step": 41
|
47 |
+
},
|
48 |
+
{
|
49 |
+
"epoch": 1.01,
|
50 |
+
"grad_norm": 8.263680458068848,
|
51 |
+
"learning_rate": 2.0833333333333336e-05,
|
52 |
+
"loss": 0.7384,
|
53 |
+
"step": 50
|
54 |
+
},
|
55 |
+
{
|
56 |
+
"epoch": 1.02,
|
57 |
+
"grad_norm": 9.806657791137695,
|
58 |
+
"learning_rate": 2.5e-05,
|
59 |
+
"loss": 0.6558,
|
60 |
+
"step": 60
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"epoch": 1.02,
|
64 |
+
"grad_norm": 4.578342437744141,
|
65 |
+
"learning_rate": 2.916666666666667e-05,
|
66 |
+
"loss": 0.569,
|
67 |
+
"step": 70
|
68 |
+
},
|
69 |
+
{
|
70 |
+
"epoch": 1.03,
|
71 |
+
"grad_norm": 1.7302507162094116,
|
72 |
+
"learning_rate": 3.3333333333333335e-05,
|
73 |
+
"loss": 0.7445,
|
74 |
+
"step": 80
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 1.03,
|
78 |
+
"eval_accuracy": 0.5686274509803921,
|
79 |
+
"eval_loss": 1.4197746515274048,
|
80 |
+
"eval_runtime": 132.2118,
|
81 |
+
"eval_samples_per_second": 0.386,
|
82 |
+
"eval_steps_per_second": 0.197,
|
83 |
+
"step": 82
|
84 |
+
},
|
85 |
+
{
|
86 |
+
"epoch": 2.01,
|
87 |
+
"grad_norm": 29.237693786621094,
|
88 |
+
"learning_rate": 3.7500000000000003e-05,
|
89 |
+
"loss": 0.7204,
|
90 |
+
"step": 90
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 2.02,
|
94 |
+
"grad_norm": 3.5326147079467773,
|
95 |
+
"learning_rate": 4.166666666666667e-05,
|
96 |
+
"loss": 1.7924,
|
97 |
+
"step": 100
|
98 |
+
},
|
99 |
+
{
|
100 |
+
"epoch": 2.02,
|
101 |
+
"grad_norm": 10.499820709228516,
|
102 |
+
"learning_rate": 4.5833333333333334e-05,
|
103 |
+
"loss": 0.6213,
|
104 |
+
"step": 110
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 2.03,
|
108 |
+
"grad_norm": 6.973377227783203,
|
109 |
+
"learning_rate": 5e-05,
|
110 |
+
"loss": 0.4931,
|
111 |
+
"step": 120
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 2.03,
|
115 |
+
"eval_accuracy": 0.5686274509803921,
|
116 |
+
"eval_loss": 1.3688796758651733,
|
117 |
+
"eval_runtime": 132.0446,
|
118 |
+
"eval_samples_per_second": 0.386,
|
119 |
+
"eval_steps_per_second": 0.197,
|
120 |
+
"step": 123
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 3.01,
|
124 |
+
"grad_norm": 24.18391990661621,
|
125 |
+
"learning_rate": 4.9537037037037035e-05,
|
126 |
+
"loss": 0.6786,
|
127 |
+
"step": 130
|
128 |
+
},
|
129 |
+
{
|
130 |
+
"epoch": 3.01,
|
131 |
+
"grad_norm": 19.578279495239258,
|
132 |
+
"learning_rate": 4.9074074074074075e-05,
|
133 |
+
"loss": 0.597,
|
134 |
+
"step": 140
|
135 |
+
},
|
136 |
+
{
|
137 |
+
"epoch": 3.02,
|
138 |
+
"grad_norm": 0.4028966724872589,
|
139 |
+
"learning_rate": 4.8611111111111115e-05,
|
140 |
+
"loss": 0.7246,
|
141 |
+
"step": 150
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 3.03,
|
145 |
+
"grad_norm": 29.119096755981445,
|
146 |
+
"learning_rate": 4.814814814814815e-05,
|
147 |
+
"loss": 0.7206,
|
148 |
+
"step": 160
|
149 |
+
},
|
150 |
+
{
|
151 |
+
"epoch": 3.03,
|
152 |
+
"eval_accuracy": 0.5882352941176471,
|
153 |
+
"eval_loss": 1.328170657157898,
|
154 |
+
"eval_runtime": 134.48,
|
155 |
+
"eval_samples_per_second": 0.379,
|
156 |
+
"eval_steps_per_second": 0.193,
|
157 |
+
"step": 164
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 4.0,
|
161 |
+
"grad_norm": 4.413028717041016,
|
162 |
+
"learning_rate": 4.768518518518519e-05,
|
163 |
+
"loss": 0.6192,
|
164 |
+
"step": 170
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 4.01,
|
168 |
+
"grad_norm": 0.44842419028282166,
|
169 |
+
"learning_rate": 4.722222222222222e-05,
|
170 |
+
"loss": 0.3903,
|
171 |
+
"step": 180
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 4.02,
|
175 |
+
"grad_norm": 9.306065559387207,
|
176 |
+
"learning_rate": 4.675925925925926e-05,
|
177 |
+
"loss": 0.7218,
|
178 |
+
"step": 190
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 4.03,
|
182 |
+
"grad_norm": 52.77680587768555,
|
183 |
+
"learning_rate": 4.62962962962963e-05,
|
184 |
+
"loss": 0.735,
|
185 |
+
"step": 200
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 4.03,
|
189 |
+
"eval_accuracy": 0.5686274509803921,
|
190 |
+
"eval_loss": 2.0849499702453613,
|
191 |
+
"eval_runtime": 132.8084,
|
192 |
+
"eval_samples_per_second": 0.384,
|
193 |
+
"eval_steps_per_second": 0.196,
|
194 |
+
"step": 205
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 5.0,
|
198 |
+
"grad_norm": 20.23072624206543,
|
199 |
+
"learning_rate": 4.5833333333333334e-05,
|
200 |
+
"loss": 0.9199,
|
201 |
+
"step": 210
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 5.01,
|
205 |
+
"grad_norm": 3.353752613067627,
|
206 |
+
"learning_rate": 4.5370370370370374e-05,
|
207 |
+
"loss": 0.4171,
|
208 |
+
"step": 220
|
209 |
+
},
|
210 |
+
{
|
211 |
+
"epoch": 5.02,
|
212 |
+
"grad_norm": 0.07062011957168579,
|
213 |
+
"learning_rate": 4.490740740740741e-05,
|
214 |
+
"loss": 0.1716,
|
215 |
+
"step": 230
|
216 |
+
},
|
217 |
+
{
|
218 |
+
"epoch": 5.03,
|
219 |
+
"grad_norm": 17.275123596191406,
|
220 |
+
"learning_rate": 4.4444444444444447e-05,
|
221 |
+
"loss": 1.769,
|
222 |
+
"step": 240
|
223 |
+
},
|
224 |
+
{
|
225 |
+
"epoch": 5.03,
|
226 |
+
"eval_accuracy": 0.6078431372549019,
|
227 |
+
"eval_loss": 1.0332462787628174,
|
228 |
+
"eval_runtime": 141.0066,
|
229 |
+
"eval_samples_per_second": 0.362,
|
230 |
+
"eval_steps_per_second": 0.184,
|
231 |
+
"step": 246
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 6.0,
|
235 |
+
"grad_norm": 1.1296720504760742,
|
236 |
+
"learning_rate": 4.3981481481481486e-05,
|
237 |
+
"loss": 0.4991,
|
238 |
+
"step": 250
|
239 |
+
},
|
240 |
+
{
|
241 |
+
"epoch": 6.01,
|
242 |
+
"grad_norm": 0.18362000584602356,
|
243 |
+
"learning_rate": 4.351851851851852e-05,
|
244 |
+
"loss": 0.4565,
|
245 |
+
"step": 260
|
246 |
+
},
|
247 |
+
{
|
248 |
+
"epoch": 6.02,
|
249 |
+
"grad_norm": 9.544737815856934,
|
250 |
+
"learning_rate": 4.305555555555556e-05,
|
251 |
+
"loss": 1.5794,
|
252 |
+
"step": 270
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 6.03,
|
256 |
+
"grad_norm": 9.105454444885254,
|
257 |
+
"learning_rate": 4.259259259259259e-05,
|
258 |
+
"loss": 0.9494,
|
259 |
+
"step": 280
|
260 |
+
},
|
261 |
+
{
|
262 |
+
"epoch": 6.03,
|
263 |
+
"eval_accuracy": 0.6078431372549019,
|
264 |
+
"eval_loss": 2.130749464035034,
|
265 |
+
"eval_runtime": 147.5482,
|
266 |
+
"eval_samples_per_second": 0.346,
|
267 |
+
"eval_steps_per_second": 0.176,
|
268 |
+
"step": 287
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 7.0,
|
272 |
+
"grad_norm": 1.585872769355774,
|
273 |
+
"learning_rate": 4.212962962962963e-05,
|
274 |
+
"loss": 0.3491,
|
275 |
+
"step": 290
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 7.01,
|
279 |
+
"grad_norm": 6.12731409072876,
|
280 |
+
"learning_rate": 4.166666666666667e-05,
|
281 |
+
"loss": 0.5312,
|
282 |
+
"step": 300
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 7.02,
|
286 |
+
"grad_norm": 20.4097957611084,
|
287 |
+
"learning_rate": 4.1203703703703705e-05,
|
288 |
+
"loss": 0.6494,
|
289 |
+
"step": 310
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 7.03,
|
293 |
+
"grad_norm": 1.075332522392273,
|
294 |
+
"learning_rate": 4.074074074074074e-05,
|
295 |
+
"loss": 0.3857,
|
296 |
+
"step": 320
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 7.03,
|
300 |
+
"eval_accuracy": 0.6078431372549019,
|
301 |
+
"eval_loss": 2.598536252975464,
|
302 |
+
"eval_runtime": 141.9381,
|
303 |
+
"eval_samples_per_second": 0.359,
|
304 |
+
"eval_steps_per_second": 0.183,
|
305 |
+
"step": 328
|
306 |
+
},
|
307 |
+
{
|
308 |
+
"epoch": 8.0,
|
309 |
+
"grad_norm": 0.12138959765434265,
|
310 |
+
"learning_rate": 4.027777777777778e-05,
|
311 |
+
"loss": 0.8439,
|
312 |
+
"step": 330
|
313 |
+
},
|
314 |
+
{
|
315 |
+
"epoch": 8.01,
|
316 |
+
"grad_norm": 68.02975463867188,
|
317 |
+
"learning_rate": 3.981481481481482e-05,
|
318 |
+
"loss": 0.8624,
|
319 |
+
"step": 340
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"epoch": 8.02,
|
323 |
+
"grad_norm": 0.4764052927494049,
|
324 |
+
"learning_rate": 3.935185185185186e-05,
|
325 |
+
"loss": 0.3225,
|
326 |
+
"step": 350
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 8.03,
|
330 |
+
"grad_norm": 1.0622520446777344,
|
331 |
+
"learning_rate": 3.888888888888889e-05,
|
332 |
+
"loss": 0.3639,
|
333 |
+
"step": 360
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 8.03,
|
337 |
+
"eval_accuracy": 0.6078431372549019,
|
338 |
+
"eval_loss": 2.2299933433532715,
|
339 |
+
"eval_runtime": 147.3983,
|
340 |
+
"eval_samples_per_second": 0.346,
|
341 |
+
"eval_steps_per_second": 0.176,
|
342 |
+
"step": 369
|
343 |
+
},
|
344 |
+
{
|
345 |
+
"epoch": 9.0,
|
346 |
+
"grad_norm": 24.868986129760742,
|
347 |
+
"learning_rate": 3.8425925925925924e-05,
|
348 |
+
"loss": 0.6411,
|
349 |
+
"step": 370
|
350 |
+
},
|
351 |
+
{
|
352 |
+
"epoch": 9.01,
|
353 |
+
"grad_norm": 0.08051346242427826,
|
354 |
+
"learning_rate": 3.7962962962962964e-05,
|
355 |
+
"loss": 0.2678,
|
356 |
+
"step": 380
|
357 |
+
},
|
358 |
+
{
|
359 |
+
"epoch": 9.02,
|
360 |
+
"grad_norm": 14.524049758911133,
|
361 |
+
"learning_rate": 3.7500000000000003e-05,
|
362 |
+
"loss": 1.1313,
|
363 |
+
"step": 390
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 9.03,
|
367 |
+
"grad_norm": 5.631989479064941,
|
368 |
+
"learning_rate": 3.7037037037037037e-05,
|
369 |
+
"loss": 0.7205,
|
370 |
+
"step": 400
|
371 |
},
|
372 |
{
|
373 |
+
"epoch": 9.03,
|
374 |
+
"grad_norm": 0.37318071722984314,
|
375 |
+
"learning_rate": 3.6574074074074076e-05,
|
376 |
+
"loss": 0.6456,
|
377 |
+
"step": 410
|
|
|
|
|
378 |
},
|
379 |
{
|
380 |
+
"epoch": 9.03,
|
381 |
+
"eval_accuracy": 0.6078431372549019,
|
382 |
+
"eval_loss": 1.8992350101470947,
|
383 |
+
"eval_runtime": 135.9679,
|
384 |
+
"eval_samples_per_second": 0.375,
|
385 |
+
"eval_steps_per_second": 0.191,
|
386 |
+
"step": 410
|
387 |
},
|
388 |
{
|
389 |
+
"epoch": 10.01,
|
390 |
+
"grad_norm": 0.06262947618961334,
|
391 |
+
"learning_rate": 3.611111111111111e-05,
|
392 |
+
"loss": 0.331,
|
393 |
+
"step": 420
|
|
|
|
|
394 |
},
|
395 |
{
|
396 |
+
"epoch": 10.02,
|
397 |
+
"grad_norm": 0.16807079315185547,
|
398 |
+
"learning_rate": 3.564814814814815e-05,
|
399 |
+
"loss": 0.4851,
|
400 |
+
"step": 430
|
401 |
+
},
|
402 |
+
{
|
403 |
+
"epoch": 10.03,
|
404 |
+
"grad_norm": 0.23209631443023682,
|
405 |
+
"learning_rate": 3.518518518518519e-05,
|
406 |
+
"loss": 0.6103,
|
407 |
+
"step": 440
|
408 |
+
},
|
409 |
+
{
|
410 |
+
"epoch": 10.03,
|
411 |
+
"grad_norm": 0.20488207042217255,
|
412 |
+
"learning_rate": 3.472222222222222e-05,
|
413 |
+
"loss": 0.9483,
|
414 |
+
"step": 450
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 10.03,
|
418 |
+
"eval_accuracy": 0.6078431372549019,
|
419 |
+
"eval_loss": 1.955637812614441,
|
420 |
+
"eval_runtime": 130.9716,
|
421 |
+
"eval_samples_per_second": 0.389,
|
422 |
+
"eval_steps_per_second": 0.199,
|
423 |
+
"step": 451
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 11.01,
|
427 |
+
"grad_norm": 42.45355987548828,
|
428 |
+
"learning_rate": 3.425925925925926e-05,
|
429 |
+
"loss": 0.3743,
|
430 |
+
"step": 460
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 11.02,
|
434 |
+
"grad_norm": 0.16700521111488342,
|
435 |
+
"learning_rate": 3.3796296296296295e-05,
|
436 |
+
"loss": 0.368,
|
437 |
+
"step": 470
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 11.02,
|
441 |
+
"grad_norm": 1.189526915550232,
|
442 |
+
"learning_rate": 3.3333333333333335e-05,
|
443 |
+
"loss": 0.8548,
|
444 |
+
"step": 480
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 11.03,
|
448 |
+
"grad_norm": 23.483234405517578,
|
449 |
+
"learning_rate": 3.2870370370370375e-05,
|
450 |
+
"loss": 0.4518,
|
451 |
+
"step": 490
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 11.03,
|
455 |
+
"eval_accuracy": 0.6274509803921569,
|
456 |
+
"eval_loss": 1.834560751914978,
|
457 |
+
"eval_runtime": 139.8695,
|
458 |
+
"eval_samples_per_second": 0.365,
|
459 |
+
"eval_steps_per_second": 0.186,
|
460 |
+
"step": 492
|
461 |
+
},
|
462 |
+
{
|
463 |
+
"epoch": 12.01,
|
464 |
+
"grad_norm": 14.332037925720215,
|
465 |
+
"learning_rate": 3.240740740740741e-05,
|
466 |
+
"loss": 0.3553,
|
467 |
+
"step": 500
|
468 |
+
},
|
469 |
+
{
|
470 |
+
"epoch": 12.02,
|
471 |
+
"grad_norm": 0.2652936279773712,
|
472 |
+
"learning_rate": 3.194444444444444e-05,
|
473 |
+
"loss": 0.3726,
|
474 |
+
"step": 510
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 12.02,
|
478 |
+
"grad_norm": 0.1738712340593338,
|
479 |
+
"learning_rate": 3.148148148148148e-05,
|
480 |
+
"loss": 0.4627,
|
481 |
+
"step": 520
|
482 |
+
},
|
483 |
+
{
|
484 |
+
"epoch": 12.03,
|
485 |
+
"grad_norm": 16.079055786132812,
|
486 |
+
"learning_rate": 3.101851851851852e-05,
|
487 |
+
"loss": 0.9109,
|
488 |
+
"step": 530
|
489 |
+
},
|
490 |
+
{
|
491 |
+
"epoch": 12.03,
|
492 |
+
"eval_accuracy": 0.6078431372549019,
|
493 |
+
"eval_loss": 1.993680477142334,
|
494 |
+
"eval_runtime": 138.9405,
|
495 |
+
"eval_samples_per_second": 0.367,
|
496 |
+
"eval_steps_per_second": 0.187,
|
497 |
+
"step": 533
|
498 |
+
},
|
499 |
+
{
|
500 |
+
"epoch": 13.01,
|
501 |
+
"grad_norm": 0.2999970018863678,
|
502 |
+
"learning_rate": 3.055555555555556e-05,
|
503 |
+
"loss": 0.1598,
|
504 |
+
"step": 540
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 13.01,
|
508 |
+
"grad_norm": 0.08949922025203705,
|
509 |
+
"learning_rate": 3.0092592592592593e-05,
|
510 |
+
"loss": 0.0925,
|
511 |
+
"step": 550
|
512 |
+
},
|
513 |
+
{
|
514 |
+
"epoch": 13.02,
|
515 |
+
"grad_norm": 13.087860107421875,
|
516 |
+
"learning_rate": 2.962962962962963e-05,
|
517 |
+
"loss": 0.8925,
|
518 |
+
"step": 560
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 13.03,
|
522 |
+
"grad_norm": 12.626638412475586,
|
523 |
+
"learning_rate": 2.916666666666667e-05,
|
524 |
+
"loss": 0.5441,
|
525 |
+
"step": 570
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 13.03,
|
529 |
+
"eval_accuracy": 0.6078431372549019,
|
530 |
+
"eval_loss": 1.401322364807129,
|
531 |
+
"eval_runtime": 140.7378,
|
532 |
+
"eval_samples_per_second": 0.362,
|
533 |
+
"eval_steps_per_second": 0.185,
|
534 |
+
"step": 574
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 14.01,
|
538 |
+
"grad_norm": 0.09884772449731827,
|
539 |
+
"learning_rate": 2.8703703703703706e-05,
|
540 |
+
"loss": 0.2266,
|
541 |
+
"step": 580
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 14.01,
|
545 |
+
"grad_norm": 0.16622687876224518,
|
546 |
+
"learning_rate": 2.824074074074074e-05,
|
547 |
+
"loss": 0.0233,
|
548 |
+
"step": 590
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 14.02,
|
552 |
+
"grad_norm": 0.08121990412473679,
|
553 |
"learning_rate": 2.777777777777778e-05,
|
554 |
+
"loss": 0.4697,
|
555 |
+
"step": 600
|
556 |
},
|
557 |
{
|
558 |
+
"epoch": 14.03,
|
559 |
+
"grad_norm": 0.04321848973631859,
|
560 |
+
"learning_rate": 2.7314814814814816e-05,
|
561 |
+
"loss": 0.4035,
|
562 |
+
"step": 610
|
|
|
|
|
563 |
},
|
564 |
{
|
565 |
+
"epoch": 14.03,
|
566 |
+
"eval_accuracy": 0.6078431372549019,
|
567 |
+
"eval_loss": 1.9077707529067993,
|
568 |
+
"eval_runtime": 140.6526,
|
569 |
+
"eval_samples_per_second": 0.363,
|
570 |
+
"eval_steps_per_second": 0.185,
|
571 |
+
"step": 615
|
572 |
},
|
573 |
{
|
574 |
+
"epoch": 15.0,
|
575 |
+
"grad_norm": 0.08672894537448883,
|
576 |
+
"learning_rate": 2.6851851851851855e-05,
|
577 |
+
"loss": 0.7602,
|
578 |
+
"step": 620
|
|
|
|
|
579 |
},
|
580 |
{
|
581 |
+
"epoch": 15.01,
|
582 |
+
"grad_norm": 0.03848231956362724,
|
583 |
+
"learning_rate": 2.6388888888888892e-05,
|
584 |
+
"loss": 0.0041,
|
585 |
+
"step": 630
|
|
|
|
|
586 |
},
|
587 |
{
|
588 |
+
"epoch": 15.02,
|
589 |
+
"grad_norm": 0.09817267954349518,
|
590 |
+
"learning_rate": 2.5925925925925925e-05,
|
591 |
+
"loss": 0.7232,
|
592 |
+
"step": 640
|
|
|
|
|
593 |
},
|
594 |
{
|
595 |
+
"epoch": 15.03,
|
596 |
+
"grad_norm": 2.0446386337280273,
|
597 |
+
"learning_rate": 2.5462962962962965e-05,
|
598 |
+
"loss": 0.1713,
|
599 |
+
"step": 650
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 15.03,
|
603 |
+
"eval_accuracy": 0.5882352941176471,
|
604 |
+
"eval_loss": 2.080268383026123,
|
605 |
+
"eval_runtime": 140.5712,
|
606 |
+
"eval_samples_per_second": 0.363,
|
607 |
+
"eval_steps_per_second": 0.185,
|
608 |
+
"step": 656
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 16.0,
|
612 |
+
"grad_norm": 13.678278923034668,
|
613 |
+
"learning_rate": 2.5e-05,
|
614 |
+
"loss": 0.4517,
|
615 |
+
"step": 660
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 16.01,
|
619 |
+
"grad_norm": 6.980419635772705,
|
620 |
+
"learning_rate": 2.4537037037037038e-05,
|
621 |
+
"loss": 0.0059,
|
622 |
+
"step": 670
|
623 |
+
},
|
624 |
+
{
|
625 |
+
"epoch": 16.02,
|
626 |
+
"grad_norm": 0.3049313724040985,
|
627 |
+
"learning_rate": 2.4074074074074074e-05,
|
628 |
+
"loss": 0.0051,
|
629 |
+
"step": 680
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 16.03,
|
633 |
+
"grad_norm": 0.06310291588306427,
|
634 |
+
"learning_rate": 2.361111111111111e-05,
|
635 |
+
"loss": 0.0542,
|
636 |
+
"step": 690
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"epoch": 16.03,
|
640 |
+
"eval_accuracy": 0.6078431372549019,
|
641 |
+
"eval_loss": 2.543215036392212,
|
642 |
+
"eval_runtime": 139.4428,
|
643 |
+
"eval_samples_per_second": 0.366,
|
644 |
+
"eval_steps_per_second": 0.186,
|
645 |
+
"step": 697
|
646 |
+
},
|
647 |
+
{
|
648 |
+
"epoch": 17.0,
|
649 |
+
"grad_norm": 0.047225985676050186,
|
650 |
+
"learning_rate": 2.314814814814815e-05,
|
651 |
+
"loss": 0.4816,
|
652 |
+
"step": 700
|
653 |
+
},
|
654 |
+
{
|
655 |
+
"epoch": 17.01,
|
656 |
+
"grad_norm": 0.025472501292824745,
|
657 |
+
"learning_rate": 2.2685185185185187e-05,
|
658 |
+
"loss": 0.2573,
|
659 |
+
"step": 710
|
660 |
+
},
|
661 |
+
{
|
662 |
+
"epoch": 17.02,
|
663 |
+
"grad_norm": 0.1224554181098938,
|
664 |
+
"learning_rate": 2.2222222222222223e-05,
|
665 |
+
"loss": 0.0025,
|
666 |
+
"step": 720
|
667 |
+
},
|
668 |
+
{
|
669 |
+
"epoch": 17.03,
|
670 |
+
"grad_norm": 0.01003193762153387,
|
671 |
+
"learning_rate": 2.175925925925926e-05,
|
672 |
+
"loss": 0.3084,
|
673 |
+
"step": 730
|
674 |
+
},
|
675 |
+
{
|
676 |
+
"epoch": 17.03,
|
677 |
+
"eval_accuracy": 0.6078431372549019,
|
678 |
+
"eval_loss": 2.57525372505188,
|
679 |
+
"eval_runtime": 138.9495,
|
680 |
+
"eval_samples_per_second": 0.367,
|
681 |
+
"eval_steps_per_second": 0.187,
|
682 |
+
"step": 738
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 18.0,
|
686 |
+
"grad_norm": 1.1518467664718628,
|
687 |
+
"learning_rate": 2.1296296296296296e-05,
|
688 |
+
"loss": 0.242,
|
689 |
+
"step": 740
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 18.01,
|
693 |
+
"grad_norm": 0.033695243299007416,
|
694 |
+
"learning_rate": 2.0833333333333336e-05,
|
695 |
+
"loss": 0.2333,
|
696 |
+
"step": 750
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 18.02,
|
700 |
+
"grad_norm": 0.06799116730690002,
|
701 |
+
"learning_rate": 2.037037037037037e-05,
|
702 |
+
"loss": 0.0012,
|
703 |
+
"step": 760
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 18.03,
|
707 |
+
"grad_norm": 0.00848899595439434,
|
708 |
+
"learning_rate": 1.990740740740741e-05,
|
709 |
+
"loss": 0.2476,
|
710 |
+
"step": 770
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 18.03,
|
714 |
+
"eval_accuracy": 0.6274509803921569,
|
715 |
+
"eval_loss": 2.4252748489379883,
|
716 |
+
"eval_runtime": 140.2598,
|
717 |
+
"eval_samples_per_second": 0.364,
|
718 |
+
"eval_steps_per_second": 0.185,
|
719 |
+
"step": 779
|
720 |
+
},
|
721 |
+
{
|
722 |
+
"epoch": 19.0,
|
723 |
+
"grad_norm": 0.015138844028115273,
|
724 |
+
"learning_rate": 1.9444444444444445e-05,
|
725 |
+
"loss": 0.3527,
|
726 |
+
"step": 780
|
727 |
+
},
|
728 |
+
{
|
729 |
+
"epoch": 19.01,
|
730 |
+
"grad_norm": 0.008504153229296207,
|
731 |
+
"learning_rate": 1.8981481481481482e-05,
|
732 |
+
"loss": 0.0087,
|
733 |
+
"step": 790
|
734 |
+
},
|
735 |
+
{
|
736 |
+
"epoch": 19.02,
|
737 |
+
"grad_norm": 140.87295532226562,
|
738 |
+
"learning_rate": 1.8518518518518518e-05,
|
739 |
+
"loss": 0.2821,
|
740 |
+
"step": 800
|
741 |
+
},
|
742 |
+
{
|
743 |
+
"epoch": 19.03,
|
744 |
+
"grad_norm": 0.027083896100521088,
|
745 |
+
"learning_rate": 1.8055555555555555e-05,
|
746 |
+
"loss": 0.404,
|
747 |
+
"step": 810
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"epoch": 19.03,
|
751 |
+
"grad_norm": 0.15064358711242676,
|
752 |
+
"learning_rate": 1.7592592592592595e-05,
|
753 |
+
"loss": 0.006,
|
754 |
+
"step": 820
|
755 |
+
},
|
756 |
+
{
|
757 |
+
"epoch": 19.03,
|
758 |
+
"eval_accuracy": 0.6078431372549019,
|
759 |
+
"eval_loss": 2.4320342540740967,
|
760 |
+
"eval_runtime": 136.4673,
|
761 |
+
"eval_samples_per_second": 0.374,
|
762 |
+
"eval_steps_per_second": 0.191,
|
763 |
+
"step": 820
|
764 |
+
},
|
765 |
+
{
|
766 |
+
"epoch": 20.01,
|
767 |
+
"grad_norm": 0.05090714618563652,
|
768 |
+
"learning_rate": 1.712962962962963e-05,
|
769 |
+
"loss": 0.1214,
|
770 |
+
"step": 830
|
771 |
+
},
|
772 |
+
{
|
773 |
+
"epoch": 20.02,
|
774 |
+
"grad_norm": 0.05721502751111984,
|
775 |
+
"learning_rate": 1.6666666666666667e-05,
|
776 |
+
"loss": 0.3519,
|
777 |
+
"step": 840
|
778 |
+
},
|
779 |
+
{
|
780 |
+
"epoch": 20.02,
|
781 |
+
"grad_norm": 0.025916827842593193,
|
782 |
+
"learning_rate": 1.6203703703703704e-05,
|
783 |
+
"loss": 0.0014,
|
784 |
+
"step": 850
|
785 |
+
},
|
786 |
+
{
|
787 |
+
"epoch": 20.03,
|
788 |
+
"grad_norm": 0.04987427219748497,
|
789 |
+
"learning_rate": 1.574074074074074e-05,
|
790 |
+
"loss": 0.0021,
|
791 |
+
"step": 860
|
792 |
+
},
|
793 |
+
{
|
794 |
+
"epoch": 20.03,
|
795 |
+
"eval_accuracy": 0.6078431372549019,
|
796 |
+
"eval_loss": 2.718170166015625,
|
797 |
+
"eval_runtime": 143.3116,
|
798 |
+
"eval_samples_per_second": 0.356,
|
799 |
+
"eval_steps_per_second": 0.181,
|
800 |
+
"step": 861
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 21.01,
|
804 |
+
"grad_norm": 0.009684134274721146,
|
805 |
+
"learning_rate": 1.527777777777778e-05,
|
806 |
+
"loss": 0.22,
|
807 |
+
"step": 870
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 21.02,
|
811 |
+
"grad_norm": 20.616426467895508,
|
812 |
+
"learning_rate": 1.4814814814814815e-05,
|
813 |
+
"loss": 0.2033,
|
814 |
+
"step": 880
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 21.02,
|
818 |
+
"grad_norm": 0.028017116710543633,
|
819 |
+
"learning_rate": 1.4351851851851853e-05,
|
820 |
+
"loss": 0.4944,
|
821 |
+
"step": 890
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 21.03,
|
825 |
+
"grad_norm": 0.04853179305791855,
|
826 |
+
"learning_rate": 1.388888888888889e-05,
|
827 |
+
"loss": 0.008,
|
828 |
+
"step": 900
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 21.03,
|
832 |
+
"eval_accuracy": 0.5882352941176471,
|
833 |
+
"eval_loss": 2.771120548248291,
|
834 |
+
"eval_runtime": 141.504,
|
835 |
+
"eval_samples_per_second": 0.36,
|
836 |
+
"eval_steps_per_second": 0.184,
|
837 |
+
"step": 902
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 22.01,
|
841 |
+
"grad_norm": 0.010195921175181866,
|
842 |
+
"learning_rate": 1.3425925925925928e-05,
|
843 |
+
"loss": 0.2491,
|
844 |
+
"step": 910
|
845 |
+
},
|
846 |
+
{
|
847 |
+
"epoch": 22.02,
|
848 |
+
"grad_norm": 0.058783795684576035,
|
849 |
+
"learning_rate": 1.2962962962962962e-05,
|
850 |
+
"loss": 0.1002,
|
851 |
+
"step": 920
|
852 |
+
},
|
853 |
+
{
|
854 |
+
"epoch": 22.02,
|
855 |
+
"grad_norm": 0.009438421577215195,
|
856 |
+
"learning_rate": 1.25e-05,
|
857 |
+
"loss": 0.2846,
|
858 |
+
"step": 930
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"epoch": 22.03,
|
862 |
+
"grad_norm": 0.017627988010644913,
|
863 |
+
"learning_rate": 1.2037037037037037e-05,
|
864 |
+
"loss": 0.0011,
|
865 |
+
"step": 940
|
866 |
+
},
|
867 |
+
{
|
868 |
+
"epoch": 22.03,
|
869 |
+
"eval_accuracy": 0.5882352941176471,
|
870 |
+
"eval_loss": 2.808924674987793,
|
871 |
+
"eval_runtime": 140.1443,
|
872 |
+
"eval_samples_per_second": 0.364,
|
873 |
+
"eval_steps_per_second": 0.186,
|
874 |
+
"step": 943
|
875 |
+
},
|
876 |
+
{
|
877 |
+
"epoch": 23.01,
|
878 |
+
"grad_norm": 0.009988226927816868,
|
879 |
+
"learning_rate": 1.1574074074074075e-05,
|
880 |
+
"loss": 0.0157,
|
881 |
+
"step": 950
|
882 |
+
},
|
883 |
+
{
|
884 |
+
"epoch": 23.01,
|
885 |
+
"grad_norm": 0.02875751443207264,
|
886 |
+
"learning_rate": 1.1111111111111112e-05,
|
887 |
+
"loss": 0.0036,
|
888 |
+
"step": 960
|
889 |
+
},
|
890 |
+
{
|
891 |
+
"epoch": 23.02,
|
892 |
+
"grad_norm": 0.5501790046691895,
|
893 |
+
"learning_rate": 1.0648148148148148e-05,
|
894 |
+
"loss": 0.0029,
|
895 |
+
"step": 970
|
896 |
+
},
|
897 |
+
{
|
898 |
+
"epoch": 23.03,
|
899 |
+
"grad_norm": 0.013142755255103111,
|
900 |
+
"learning_rate": 1.0185185185185185e-05,
|
901 |
+
"loss": 0.7546,
|
902 |
+
"step": 980
|
903 |
+
},
|
904 |
+
{
|
905 |
+
"epoch": 23.03,
|
906 |
+
"eval_accuracy": 0.6078431372549019,
|
907 |
+
"eval_loss": 2.8142335414886475,
|
908 |
+
"eval_runtime": 147.8203,
|
909 |
+
"eval_samples_per_second": 0.345,
|
910 |
+
"eval_steps_per_second": 0.176,
|
911 |
+
"step": 984
|
912 |
+
},
|
913 |
+
{
|
914 |
+
"epoch": 24.0,
|
915 |
+
"grad_norm": 1.282440423965454,
|
916 |
+
"learning_rate": 9.722222222222223e-06,
|
917 |
+
"loss": 0.0821,
|
918 |
+
"step": 990
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"epoch": 24.01,
|
922 |
+
"grad_norm": 0.011100966483354568,
|
923 |
+
"learning_rate": 9.259259259259259e-06,
|
924 |
+
"loss": 0.0014,
|
925 |
+
"step": 1000
|
926 |
+
},
|
927 |
+
{
|
928 |
+
"epoch": 24.02,
|
929 |
+
"grad_norm": 0.020793117582798004,
|
930 |
+
"learning_rate": 8.796296296296297e-06,
|
931 |
+
"loss": 0.1571,
|
932 |
+
"step": 1010
|
933 |
+
},
|
934 |
+
{
|
935 |
+
"epoch": 24.03,
|
936 |
+
"grad_norm": 0.02288602478802204,
|
937 |
+
"learning_rate": 8.333333333333334e-06,
|
938 |
+
"loss": 0.0006,
|
939 |
+
"step": 1020
|
940 |
+
},
|
941 |
+
{
|
942 |
+
"epoch": 24.03,
|
943 |
+
"eval_accuracy": 0.6078431372549019,
|
944 |
+
"eval_loss": 2.8796093463897705,
|
945 |
+
"eval_runtime": 142.281,
|
946 |
+
"eval_samples_per_second": 0.358,
|
947 |
+
"eval_steps_per_second": 0.183,
|
948 |
+
"step": 1025
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 25.0,
|
952 |
+
"grad_norm": 0.031101850792765617,
|
953 |
+
"learning_rate": 7.87037037037037e-06,
|
954 |
+
"loss": 0.0051,
|
955 |
+
"step": 1030
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 25.01,
|
959 |
+
"grad_norm": 0.017316602170467377,
|
960 |
+
"learning_rate": 7.4074074074074075e-06,
|
961 |
+
"loss": 0.0006,
|
962 |
+
"step": 1040
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 25.02,
|
966 |
+
"grad_norm": 0.015025320462882519,
|
967 |
+
"learning_rate": 6.944444444444445e-06,
|
968 |
+
"loss": 0.2444,
|
969 |
+
"step": 1050
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 25.03,
|
973 |
+
"grad_norm": 0.014637992717325687,
|
974 |
+
"learning_rate": 6.481481481481481e-06,
|
975 |
+
"loss": 0.0008,
|
976 |
+
"step": 1060
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 25.03,
|
980 |
+
"eval_accuracy": 0.6078431372549019,
|
981 |
+
"eval_loss": 2.8485913276672363,
|
982 |
+
"eval_runtime": 140.4219,
|
983 |
+
"eval_samples_per_second": 0.363,
|
984 |
+
"eval_steps_per_second": 0.185,
|
985 |
+
"step": 1066
|
986 |
+
},
|
987 |
+
{
|
988 |
+
"epoch": 26.0,
|
989 |
+
"grad_norm": 0.1402996927499771,
|
990 |
+
"learning_rate": 6.0185185185185185e-06,
|
991 |
+
"loss": 0.1973,
|
992 |
+
"step": 1070
|
993 |
+
},
|
994 |
+
{
|
995 |
+
"epoch": 26.01,
|
996 |
+
"grad_norm": 0.006944969296455383,
|
997 |
+
"learning_rate": 5.555555555555556e-06,
|
998 |
+
"loss": 0.0005,
|
999 |
+
"step": 1080
|
1000 |
+
},
|
1001 |
+
{
|
1002 |
+
"epoch": 26.02,
|
1003 |
+
"grad_norm": 18.494768142700195,
|
1004 |
+
"learning_rate": 5.092592592592592e-06,
|
1005 |
+
"loss": 0.1727,
|
1006 |
+
"step": 1090
|
1007 |
+
},
|
1008 |
+
{
|
1009 |
+
"epoch": 26.03,
|
1010 |
+
"grad_norm": 0.010701355524361134,
|
1011 |
+
"learning_rate": 4.6296296296296296e-06,
|
1012 |
+
"loss": 0.1745,
|
1013 |
+
"step": 1100
|
1014 |
+
},
|
1015 |
+
{
|
1016 |
+
"epoch": 26.03,
|
1017 |
+
"eval_accuracy": 0.6078431372549019,
|
1018 |
+
"eval_loss": 2.8475284576416016,
|
1019 |
+
"eval_runtime": 143.4719,
|
1020 |
+
"eval_samples_per_second": 0.355,
|
1021 |
+
"eval_steps_per_second": 0.181,
|
1022 |
+
"step": 1107
|
1023 |
+
},
|
1024 |
+
{
|
1025 |
+
"epoch": 27.0,
|
1026 |
+
"grad_norm": 0.012416614219546318,
|
1027 |
+
"learning_rate": 4.166666666666667e-06,
|
1028 |
+
"loss": 0.0074,
|
1029 |
+
"step": 1110
|
1030 |
+
},
|
1031 |
+
{
|
1032 |
+
"epoch": 27.01,
|
1033 |
+
"grad_norm": 0.06672808527946472,
|
1034 |
+
"learning_rate": 3.7037037037037037e-06,
|
1035 |
+
"loss": 0.0006,
|
1036 |
+
"step": 1120
|
1037 |
+
},
|
1038 |
+
{
|
1039 |
+
"epoch": 27.02,
|
1040 |
+
"grad_norm": 0.008200524374842644,
|
1041 |
+
"learning_rate": 3.2407407407407406e-06,
|
1042 |
+
"loss": 0.0006,
|
1043 |
+
"step": 1130
|
1044 |
+
},
|
1045 |
+
{
|
1046 |
+
"epoch": 27.03,
|
1047 |
+
"grad_norm": 14.067534446716309,
|
1048 |
+
"learning_rate": 2.777777777777778e-06,
|
1049 |
+
"loss": 0.5421,
|
1050 |
+
"step": 1140
|
1051 |
+
},
|
1052 |
+
{
|
1053 |
+
"epoch": 27.03,
|
1054 |
+
"eval_accuracy": 0.6078431372549019,
|
1055 |
+
"eval_loss": 2.8462321758270264,
|
1056 |
+
"eval_runtime": 137.4454,
|
1057 |
+
"eval_samples_per_second": 0.371,
|
1058 |
+
"eval_steps_per_second": 0.189,
|
1059 |
+
"step": 1148
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 28.0,
|
1063 |
+
"grad_norm": 0.009883771650493145,
|
1064 |
+
"learning_rate": 2.3148148148148148e-06,
|
1065 |
+
"loss": 0.0889,
|
1066 |
+
"step": 1150
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 28.01,
|
1070 |
+
"grad_norm": 0.007996085099875927,
|
1071 |
+
"learning_rate": 1.8518518518518519e-06,
|
1072 |
+
"loss": 0.1251,
|
1073 |
+
"step": 1160
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 28.02,
|
1077 |
+
"grad_norm": 0.008312738500535488,
|
1078 |
+
"learning_rate": 1.388888888888889e-06,
|
1079 |
+
"loss": 0.339,
|
1080 |
+
"step": 1170
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 28.03,
|
1084 |
+
"grad_norm": 0.07012984901666641,
|
1085 |
+
"learning_rate": 9.259259259259259e-07,
|
1086 |
+
"loss": 0.1233,
|
1087 |
+
"step": 1180
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 28.03,
|
1091 |
+
"eval_accuracy": 0.6078431372549019,
|
1092 |
+
"eval_loss": 2.8324010372161865,
|
1093 |
+
"eval_runtime": 132.7093,
|
1094 |
+
"eval_samples_per_second": 0.384,
|
1095 |
+
"eval_steps_per_second": 0.196,
|
1096 |
+
"step": 1189
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 29.0,
|
1100 |
+
"grad_norm": 0.029733408242464066,
|
1101 |
+
"learning_rate": 4.6296296296296297e-07,
|
1102 |
+
"loss": 0.0054,
|
1103 |
+
"step": 1190
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"epoch": 29.01,
|
1107 |
+
"grad_norm": 0.033483583480119705,
|
1108 |
"learning_rate": 0.0,
|
1109 |
+
"loss": 0.1298,
|
1110 |
+
"step": 1200
|
1111 |
},
|
1112 |
{
|
1113 |
+
"epoch": 29.01,
|
1114 |
+
"eval_accuracy": 0.6078431372549019,
|
1115 |
+
"eval_loss": 2.833205461502075,
|
1116 |
+
"eval_runtime": 130.982,
|
1117 |
+
"eval_samples_per_second": 0.389,
|
1118 |
+
"eval_steps_per_second": 0.199,
|
1119 |
+
"step": 1200
|
1120 |
},
|
1121 |
{
|
1122 |
+
"epoch": 29.01,
|
1123 |
+
"step": 1200,
|
1124 |
+
"total_flos": 2.954419141637505e+18,
|
1125 |
+
"train_loss": 0.40525790957190716,
|
1126 |
+
"train_runtime": 10653.7383,
|
1127 |
+
"train_samples_per_second": 0.225,
|
1128 |
+
"train_steps_per_second": 0.113
|
1129 |
},
|
1130 |
{
|
1131 |
+
"epoch": 29.01,
|
1132 |
+
"eval_accuracy": 0.6274509803921569,
|
1133 |
+
"eval_loss": 1.834560751914978,
|
1134 |
+
"eval_runtime": 132.9753,
|
1135 |
+
"eval_samples_per_second": 0.384,
|
1136 |
+
"eval_steps_per_second": 0.196,
|
1137 |
+
"step": 1200
|
1138 |
}
|
1139 |
],
|
1140 |
+
"logging_steps": 10,
|
1141 |
+
"max_steps": 1200,
|
1142 |
+
"num_input_tokens_seen": 0,
|
1143 |
"num_train_epochs": 9223372036854775807,
|
1144 |
+
"save_steps": 500,
|
1145 |
+
"total_flos": 2.954419141637505e+18,
|
1146 |
+
"train_batch_size": 2,
|
1147 |
"trial_name": null,
|
1148 |
"trial_params": null
|
1149 |
}
|