Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1375 -0
trainer_state.json
ADDED
@@ -0,0 +1,1375 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 3.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 9822,
|
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 |
+
"learning_rate": 1.9898187741804116e-05,
|
14 |
+
"loss": 0.6683,
|
15 |
+
"step": 50
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.03,
|
19 |
+
"learning_rate": 1.9796375483608227e-05,
|
20 |
+
"loss": 0.5088,
|
21 |
+
"step": 100
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.05,
|
25 |
+
"learning_rate": 1.9694563225412342e-05,
|
26 |
+
"loss": 0.4559,
|
27 |
+
"step": 150
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.06,
|
31 |
+
"learning_rate": 1.9592750967216453e-05,
|
32 |
+
"loss": 0.4482,
|
33 |
+
"step": 200
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.08,
|
37 |
+
"learning_rate": 1.9490938709020567e-05,
|
38 |
+
"loss": 0.4275,
|
39 |
+
"step": 250
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.09,
|
43 |
+
"learning_rate": 1.938912645082468e-05,
|
44 |
+
"loss": 0.4156,
|
45 |
+
"step": 300
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.11,
|
49 |
+
"learning_rate": 1.9287314192628793e-05,
|
50 |
+
"loss": 0.4342,
|
51 |
+
"step": 350
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.12,
|
55 |
+
"learning_rate": 1.9185501934432908e-05,
|
56 |
+
"loss": 0.4266,
|
57 |
+
"step": 400
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.14,
|
61 |
+
"learning_rate": 1.9083689676237022e-05,
|
62 |
+
"loss": 0.4291,
|
63 |
+
"step": 450
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.15,
|
67 |
+
"learning_rate": 1.8981877418041133e-05,
|
68 |
+
"loss": 0.3773,
|
69 |
+
"step": 500
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.15,
|
73 |
+
"eval_accuracy": 0.8431264872780524,
|
74 |
+
"eval_loss": 0.3870041072368622,
|
75 |
+
"eval_runtime": 22.3103,
|
76 |
+
"eval_samples_per_second": 244.864,
|
77 |
+
"eval_steps_per_second": 15.329,
|
78 |
+
"step": 500
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 0.17,
|
82 |
+
"learning_rate": 1.8880065159845248e-05,
|
83 |
+
"loss": 0.399,
|
84 |
+
"step": 550
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 0.18,
|
88 |
+
"learning_rate": 1.8778252901649362e-05,
|
89 |
+
"loss": 0.3695,
|
90 |
+
"step": 600
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"epoch": 0.2,
|
94 |
+
"learning_rate": 1.8676440643453473e-05,
|
95 |
+
"loss": 0.3644,
|
96 |
+
"step": 650
|
97 |
+
},
|
98 |
+
{
|
99 |
+
"epoch": 0.21,
|
100 |
+
"learning_rate": 1.8574628385257588e-05,
|
101 |
+
"loss": 0.3644,
|
102 |
+
"step": 700
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"epoch": 0.23,
|
106 |
+
"learning_rate": 1.84728161270617e-05,
|
107 |
+
"loss": 0.361,
|
108 |
+
"step": 750
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.24,
|
112 |
+
"learning_rate": 1.8371003868865813e-05,
|
113 |
+
"loss": 0.353,
|
114 |
+
"step": 800
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.26,
|
118 |
+
"learning_rate": 1.8269191610669924e-05,
|
119 |
+
"loss": 0.3826,
|
120 |
+
"step": 850
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"epoch": 0.27,
|
124 |
+
"learning_rate": 1.816737935247404e-05,
|
125 |
+
"loss": 0.3904,
|
126 |
+
"step": 900
|
127 |
+
},
|
128 |
+
{
|
129 |
+
"epoch": 0.29,
|
130 |
+
"learning_rate": 1.8065567094278153e-05,
|
131 |
+
"loss": 0.3585,
|
132 |
+
"step": 950
|
133 |
+
},
|
134 |
+
{
|
135 |
+
"epoch": 0.31,
|
136 |
+
"learning_rate": 1.7963754836082265e-05,
|
137 |
+
"loss": 0.3547,
|
138 |
+
"step": 1000
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"epoch": 0.31,
|
142 |
+
"eval_accuracy": 0.8658246384770273,
|
143 |
+
"eval_loss": 0.3174993097782135,
|
144 |
+
"eval_runtime": 22.3389,
|
145 |
+
"eval_samples_per_second": 244.551,
|
146 |
+
"eval_steps_per_second": 15.31,
|
147 |
+
"step": 1000
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 0.32,
|
151 |
+
"learning_rate": 1.786194257788638e-05,
|
152 |
+
"loss": 0.3527,
|
153 |
+
"step": 1050
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 0.34,
|
157 |
+
"learning_rate": 1.776013031969049e-05,
|
158 |
+
"loss": 0.3777,
|
159 |
+
"step": 1100
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 0.35,
|
163 |
+
"learning_rate": 1.7658318061494605e-05,
|
164 |
+
"loss": 0.3722,
|
165 |
+
"step": 1150
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 0.37,
|
169 |
+
"learning_rate": 1.755650580329872e-05,
|
170 |
+
"loss": 0.3713,
|
171 |
+
"step": 1200
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.38,
|
175 |
+
"learning_rate": 1.7454693545102834e-05,
|
176 |
+
"loss": 0.3455,
|
177 |
+
"step": 1250
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.4,
|
181 |
+
"learning_rate": 1.7352881286906945e-05,
|
182 |
+
"loss": 0.3344,
|
183 |
+
"step": 1300
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 0.41,
|
187 |
+
"learning_rate": 1.725106902871106e-05,
|
188 |
+
"loss": 0.3504,
|
189 |
+
"step": 1350
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 0.43,
|
193 |
+
"learning_rate": 1.7149256770515174e-05,
|
194 |
+
"loss": 0.3794,
|
195 |
+
"step": 1400
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 0.44,
|
199 |
+
"learning_rate": 1.7047444512319285e-05,
|
200 |
+
"loss": 0.342,
|
201 |
+
"step": 1450
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 0.46,
|
205 |
+
"learning_rate": 1.69456322541234e-05,
|
206 |
+
"loss": 0.3385,
|
207 |
+
"step": 1500
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 0.46,
|
211 |
+
"eval_accuracy": 0.8738788211605345,
|
212 |
+
"eval_loss": 0.2986067831516266,
|
213 |
+
"eval_runtime": 22.3738,
|
214 |
+
"eval_samples_per_second": 244.169,
|
215 |
+
"eval_steps_per_second": 15.286,
|
216 |
+
"step": 1500
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"epoch": 0.47,
|
220 |
+
"learning_rate": 1.684381999592751e-05,
|
221 |
+
"loss": 0.352,
|
222 |
+
"step": 1550
|
223 |
+
},
|
224 |
+
{
|
225 |
+
"epoch": 0.49,
|
226 |
+
"learning_rate": 1.6742007737731625e-05,
|
227 |
+
"loss": 0.3375,
|
228 |
+
"step": 1600
|
229 |
+
},
|
230 |
+
{
|
231 |
+
"epoch": 0.5,
|
232 |
+
"learning_rate": 1.6640195479535736e-05,
|
233 |
+
"loss": 0.3277,
|
234 |
+
"step": 1650
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.52,
|
238 |
+
"learning_rate": 1.653838322133985e-05,
|
239 |
+
"loss": 0.3181,
|
240 |
+
"step": 1700
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.53,
|
244 |
+
"learning_rate": 1.6436570963143962e-05,
|
245 |
+
"loss": 0.3522,
|
246 |
+
"step": 1750
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 0.55,
|
250 |
+
"learning_rate": 1.6334758704948076e-05,
|
251 |
+
"loss": 0.346,
|
252 |
+
"step": 1800
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 0.57,
|
256 |
+
"learning_rate": 1.623294644675219e-05,
|
257 |
+
"loss": 0.3429,
|
258 |
+
"step": 1850
|
259 |
+
},
|
260 |
+
{
|
261 |
+
"epoch": 0.58,
|
262 |
+
"learning_rate": 1.6131134188556302e-05,
|
263 |
+
"loss": 0.3135,
|
264 |
+
"step": 1900
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 0.6,
|
268 |
+
"learning_rate": 1.6029321930360416e-05,
|
269 |
+
"loss": 0.3086,
|
270 |
+
"step": 1950
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 0.61,
|
274 |
+
"learning_rate": 1.592750967216453e-05,
|
275 |
+
"loss": 0.342,
|
276 |
+
"step": 2000
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.61,
|
280 |
+
"eval_accuracy": 0.8844956983342486,
|
281 |
+
"eval_loss": 0.27869123220443726,
|
282 |
+
"eval_runtime": 22.3248,
|
283 |
+
"eval_samples_per_second": 244.706,
|
284 |
+
"eval_steps_per_second": 15.319,
|
285 |
+
"step": 2000
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 0.63,
|
289 |
+
"learning_rate": 1.5825697413968642e-05,
|
290 |
+
"loss": 0.3349,
|
291 |
+
"step": 2050
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 0.64,
|
295 |
+
"learning_rate": 1.5723885155772757e-05,
|
296 |
+
"loss": 0.3191,
|
297 |
+
"step": 2100
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.66,
|
301 |
+
"learning_rate": 1.562207289757687e-05,
|
302 |
+
"loss": 0.3068,
|
303 |
+
"step": 2150
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.67,
|
307 |
+
"learning_rate": 1.5520260639380982e-05,
|
308 |
+
"loss": 0.307,
|
309 |
+
"step": 2200
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 0.69,
|
313 |
+
"learning_rate": 1.5418448381185097e-05,
|
314 |
+
"loss": 0.327,
|
315 |
+
"step": 2250
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 0.7,
|
319 |
+
"learning_rate": 1.531663612298921e-05,
|
320 |
+
"loss": 0.342,
|
321 |
+
"step": 2300
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 0.72,
|
325 |
+
"learning_rate": 1.5214823864793322e-05,
|
326 |
+
"loss": 0.3166,
|
327 |
+
"step": 2350
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 0.73,
|
331 |
+
"learning_rate": 1.5113011606597437e-05,
|
332 |
+
"loss": 0.3231,
|
333 |
+
"step": 2400
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 0.75,
|
337 |
+
"learning_rate": 1.5011199348401548e-05,
|
338 |
+
"loss": 0.3136,
|
339 |
+
"step": 2450
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.76,
|
343 |
+
"learning_rate": 1.4909387090205662e-05,
|
344 |
+
"loss": 0.3003,
|
345 |
+
"step": 2500
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.76,
|
349 |
+
"eval_accuracy": 0.8725974739154311,
|
350 |
+
"eval_loss": 0.30752307176589966,
|
351 |
+
"eval_runtime": 22.2828,
|
352 |
+
"eval_samples_per_second": 245.166,
|
353 |
+
"eval_steps_per_second": 15.348,
|
354 |
+
"step": 2500
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 0.78,
|
358 |
+
"learning_rate": 1.4807574832009775e-05,
|
359 |
+
"loss": 0.3412,
|
360 |
+
"step": 2550
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 0.79,
|
364 |
+
"learning_rate": 1.4705762573813888e-05,
|
365 |
+
"loss": 0.3169,
|
366 |
+
"step": 2600
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 0.81,
|
370 |
+
"learning_rate": 1.4603950315618001e-05,
|
371 |
+
"loss": 0.3147,
|
372 |
+
"step": 2650
|
373 |
+
},
|
374 |
+
{
|
375 |
+
"epoch": 0.82,
|
376 |
+
"learning_rate": 1.4502138057422115e-05,
|
377 |
+
"loss": 0.3263,
|
378 |
+
"step": 2700
|
379 |
+
},
|
380 |
+
{
|
381 |
+
"epoch": 0.84,
|
382 |
+
"learning_rate": 1.440032579922623e-05,
|
383 |
+
"loss": 0.3003,
|
384 |
+
"step": 2750
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"epoch": 0.86,
|
388 |
+
"learning_rate": 1.4298513541030341e-05,
|
389 |
+
"loss": 0.3454,
|
390 |
+
"step": 2800
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 0.87,
|
394 |
+
"learning_rate": 1.4196701282834456e-05,
|
395 |
+
"loss": 0.2966,
|
396 |
+
"step": 2850
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 0.89,
|
400 |
+
"learning_rate": 1.4094889024638567e-05,
|
401 |
+
"loss": 0.3109,
|
402 |
+
"step": 2900
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 0.9,
|
406 |
+
"learning_rate": 1.3993076766442681e-05,
|
407 |
+
"loss": 0.2894,
|
408 |
+
"step": 2950
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 0.92,
|
412 |
+
"learning_rate": 1.3891264508246794e-05,
|
413 |
+
"loss": 0.3298,
|
414 |
+
"step": 3000
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 0.92,
|
418 |
+
"eval_accuracy": 0.8806516565989383,
|
419 |
+
"eval_loss": 0.27810677886009216,
|
420 |
+
"eval_runtime": 22.29,
|
421 |
+
"eval_samples_per_second": 245.087,
|
422 |
+
"eval_steps_per_second": 15.343,
|
423 |
+
"step": 3000
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.93,
|
427 |
+
"learning_rate": 1.3789452250050907e-05,
|
428 |
+
"loss": 0.3099,
|
429 |
+
"step": 3050
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.95,
|
433 |
+
"learning_rate": 1.368763999185502e-05,
|
434 |
+
"loss": 0.308,
|
435 |
+
"step": 3100
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 0.96,
|
439 |
+
"learning_rate": 1.3585827733659134e-05,
|
440 |
+
"loss": 0.3037,
|
441 |
+
"step": 3150
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 0.98,
|
445 |
+
"learning_rate": 1.3484015475463249e-05,
|
446 |
+
"loss": 0.2799,
|
447 |
+
"step": 3200
|
448 |
+
},
|
449 |
+
{
|
450 |
+
"epoch": 0.99,
|
451 |
+
"learning_rate": 1.338220321726736e-05,
|
452 |
+
"loss": 0.3148,
|
453 |
+
"step": 3250
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 1.01,
|
457 |
+
"learning_rate": 1.3280390959071474e-05,
|
458 |
+
"loss": 0.2428,
|
459 |
+
"step": 3300
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 1.02,
|
463 |
+
"learning_rate": 1.3178578700875585e-05,
|
464 |
+
"loss": 0.2694,
|
465 |
+
"step": 3350
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 1.04,
|
469 |
+
"learning_rate": 1.30767664426797e-05,
|
470 |
+
"loss": 0.2405,
|
471 |
+
"step": 3400
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.05,
|
475 |
+
"learning_rate": 1.2974954184483813e-05,
|
476 |
+
"loss": 0.237,
|
477 |
+
"step": 3450
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 1.07,
|
481 |
+
"learning_rate": 1.2873141926287927e-05,
|
482 |
+
"loss": 0.2475,
|
483 |
+
"step": 3500
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 1.07,
|
487 |
+
"eval_accuracy": 0.8941973274757459,
|
488 |
+
"eval_loss": 0.2694980800151825,
|
489 |
+
"eval_runtime": 22.3802,
|
490 |
+
"eval_samples_per_second": 244.099,
|
491 |
+
"eval_steps_per_second": 15.281,
|
492 |
+
"step": 3500
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.08,
|
496 |
+
"learning_rate": 1.2771329668092038e-05,
|
497 |
+
"loss": 0.2633,
|
498 |
+
"step": 3550
|
499 |
+
},
|
500 |
+
{
|
501 |
+
"epoch": 1.1,
|
502 |
+
"learning_rate": 1.2669517409896153e-05,
|
503 |
+
"loss": 0.2533,
|
504 |
+
"step": 3600
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 1.11,
|
508 |
+
"learning_rate": 1.2567705151700267e-05,
|
509 |
+
"loss": 0.2391,
|
510 |
+
"step": 3650
|
511 |
+
},
|
512 |
+
{
|
513 |
+
"epoch": 1.13,
|
514 |
+
"learning_rate": 1.2465892893504378e-05,
|
515 |
+
"loss": 0.2579,
|
516 |
+
"step": 3700
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 1.15,
|
520 |
+
"learning_rate": 1.2364080635308493e-05,
|
521 |
+
"loss": 0.2388,
|
522 |
+
"step": 3750
|
523 |
+
},
|
524 |
+
{
|
525 |
+
"epoch": 1.16,
|
526 |
+
"learning_rate": 1.2262268377112606e-05,
|
527 |
+
"loss": 0.2487,
|
528 |
+
"step": 3800
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 1.18,
|
532 |
+
"learning_rate": 1.2160456118916718e-05,
|
533 |
+
"loss": 0.2673,
|
534 |
+
"step": 3850
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 1.19,
|
538 |
+
"learning_rate": 1.2058643860720831e-05,
|
539 |
+
"loss": 0.2532,
|
540 |
+
"step": 3900
|
541 |
+
},
|
542 |
+
{
|
543 |
+
"epoch": 1.21,
|
544 |
+
"learning_rate": 1.1956831602524946e-05,
|
545 |
+
"loss": 0.2341,
|
546 |
+
"step": 3950
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"epoch": 1.22,
|
550 |
+
"learning_rate": 1.1855019344329057e-05,
|
551 |
+
"loss": 0.2441,
|
552 |
+
"step": 4000
|
553 |
+
},
|
554 |
+
{
|
555 |
+
"epoch": 1.22,
|
556 |
+
"eval_accuracy": 0.8940142778693025,
|
557 |
+
"eval_loss": 0.261508047580719,
|
558 |
+
"eval_runtime": 22.3379,
|
559 |
+
"eval_samples_per_second": 244.561,
|
560 |
+
"eval_steps_per_second": 15.31,
|
561 |
+
"step": 4000
|
562 |
+
},
|
563 |
+
{
|
564 |
+
"epoch": 1.24,
|
565 |
+
"learning_rate": 1.1753207086133171e-05,
|
566 |
+
"loss": 0.262,
|
567 |
+
"step": 4050
|
568 |
+
},
|
569 |
+
{
|
570 |
+
"epoch": 1.25,
|
571 |
+
"learning_rate": 1.1651394827937284e-05,
|
572 |
+
"loss": 0.2542,
|
573 |
+
"step": 4100
|
574 |
+
},
|
575 |
+
{
|
576 |
+
"epoch": 1.27,
|
577 |
+
"learning_rate": 1.1549582569741397e-05,
|
578 |
+
"loss": 0.2583,
|
579 |
+
"step": 4150
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 1.28,
|
583 |
+
"learning_rate": 1.1447770311545512e-05,
|
584 |
+
"loss": 0.2694,
|
585 |
+
"step": 4200
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 1.3,
|
589 |
+
"learning_rate": 1.1345958053349624e-05,
|
590 |
+
"loss": 0.2462,
|
591 |
+
"step": 4250
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 1.31,
|
595 |
+
"learning_rate": 1.1244145795153739e-05,
|
596 |
+
"loss": 0.2443,
|
597 |
+
"step": 4300
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 1.33,
|
601 |
+
"learning_rate": 1.114233353695785e-05,
|
602 |
+
"loss": 0.2575,
|
603 |
+
"step": 4350
|
604 |
+
},
|
605 |
+
{
|
606 |
+
"epoch": 1.34,
|
607 |
+
"learning_rate": 1.1040521278761964e-05,
|
608 |
+
"loss": 0.2463,
|
609 |
+
"step": 4400
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 1.36,
|
613 |
+
"learning_rate": 1.0938709020566076e-05,
|
614 |
+
"loss": 0.233,
|
615 |
+
"step": 4450
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 1.37,
|
619 |
+
"learning_rate": 1.083689676237019e-05,
|
620 |
+
"loss": 0.249,
|
621 |
+
"step": 4500
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 1.37,
|
625 |
+
"eval_accuracy": 0.8958447739337361,
|
626 |
+
"eval_loss": 0.25481221079826355,
|
627 |
+
"eval_runtime": 22.2902,
|
628 |
+
"eval_samples_per_second": 245.085,
|
629 |
+
"eval_steps_per_second": 15.343,
|
630 |
+
"step": 4500
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"epoch": 1.39,
|
634 |
+
"learning_rate": 1.0735084504174303e-05,
|
635 |
+
"loss": 0.2645,
|
636 |
+
"step": 4550
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"epoch": 1.41,
|
640 |
+
"learning_rate": 1.0633272245978416e-05,
|
641 |
+
"loss": 0.2525,
|
642 |
+
"step": 4600
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"epoch": 1.42,
|
646 |
+
"learning_rate": 1.053145998778253e-05,
|
647 |
+
"loss": 0.2185,
|
648 |
+
"step": 4650
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"epoch": 1.44,
|
652 |
+
"learning_rate": 1.0429647729586643e-05,
|
653 |
+
"loss": 0.2528,
|
654 |
+
"step": 4700
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 1.45,
|
658 |
+
"learning_rate": 1.0327835471390758e-05,
|
659 |
+
"loss": 0.2261,
|
660 |
+
"step": 4750
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 1.47,
|
664 |
+
"learning_rate": 1.0226023213194869e-05,
|
665 |
+
"loss": 0.2415,
|
666 |
+
"step": 4800
|
667 |
+
},
|
668 |
+
{
|
669 |
+
"epoch": 1.48,
|
670 |
+
"learning_rate": 1.0124210954998983e-05,
|
671 |
+
"loss": 0.2569,
|
672 |
+
"step": 4850
|
673 |
+
},
|
674 |
+
{
|
675 |
+
"epoch": 1.5,
|
676 |
+
"learning_rate": 1.0022398696803094e-05,
|
677 |
+
"loss": 0.2308,
|
678 |
+
"step": 4900
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 1.51,
|
682 |
+
"learning_rate": 9.920586438607209e-06,
|
683 |
+
"loss": 0.2538,
|
684 |
+
"step": 4950
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"epoch": 1.53,
|
688 |
+
"learning_rate": 9.818774180411322e-06,
|
689 |
+
"loss": 0.2261,
|
690 |
+
"step": 5000
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 1.53,
|
694 |
+
"eval_accuracy": 0.8945634266886326,
|
695 |
+
"eval_loss": 0.2588444650173187,
|
696 |
+
"eval_runtime": 22.3415,
|
697 |
+
"eval_samples_per_second": 244.523,
|
698 |
+
"eval_steps_per_second": 15.308,
|
699 |
+
"step": 5000
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 1.54,
|
703 |
+
"learning_rate": 9.716961922215436e-06,
|
704 |
+
"loss": 0.2207,
|
705 |
+
"step": 5050
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"epoch": 1.56,
|
709 |
+
"learning_rate": 9.615149664019549e-06,
|
710 |
+
"loss": 0.2676,
|
711 |
+
"step": 5100
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 1.57,
|
715 |
+
"learning_rate": 9.513337405823662e-06,
|
716 |
+
"loss": 0.2554,
|
717 |
+
"step": 5150
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 1.59,
|
721 |
+
"learning_rate": 9.411525147627775e-06,
|
722 |
+
"loss": 0.2516,
|
723 |
+
"step": 5200
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 1.6,
|
727 |
+
"learning_rate": 9.309712889431889e-06,
|
728 |
+
"loss": 0.2331,
|
729 |
+
"step": 5250
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 1.62,
|
733 |
+
"learning_rate": 9.207900631236002e-06,
|
734 |
+
"loss": 0.2158,
|
735 |
+
"step": 5300
|
736 |
+
},
|
737 |
+
{
|
738 |
+
"epoch": 1.63,
|
739 |
+
"learning_rate": 9.106088373040115e-06,
|
740 |
+
"loss": 0.2452,
|
741 |
+
"step": 5350
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 1.65,
|
745 |
+
"learning_rate": 9.004276114844227e-06,
|
746 |
+
"loss": 0.2281,
|
747 |
+
"step": 5400
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"epoch": 1.66,
|
751 |
+
"learning_rate": 8.902463856648342e-06,
|
752 |
+
"loss": 0.2418,
|
753 |
+
"step": 5450
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 1.68,
|
757 |
+
"learning_rate": 8.800651598452455e-06,
|
758 |
+
"loss": 0.2348,
|
759 |
+
"step": 5500
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 1.68,
|
763 |
+
"eval_accuracy": 0.8982244188174996,
|
764 |
+
"eval_loss": 0.2586851119995117,
|
765 |
+
"eval_runtime": 22.3033,
|
766 |
+
"eval_samples_per_second": 244.941,
|
767 |
+
"eval_steps_per_second": 15.334,
|
768 |
+
"step": 5500
|
769 |
+
},
|
770 |
+
{
|
771 |
+
"epoch": 1.7,
|
772 |
+
"learning_rate": 8.698839340256568e-06,
|
773 |
+
"loss": 0.2388,
|
774 |
+
"step": 5550
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 1.71,
|
778 |
+
"learning_rate": 8.59702708206068e-06,
|
779 |
+
"loss": 0.2605,
|
780 |
+
"step": 5600
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 1.73,
|
784 |
+
"learning_rate": 8.495214823864793e-06,
|
785 |
+
"loss": 0.2578,
|
786 |
+
"step": 5650
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 1.74,
|
790 |
+
"learning_rate": 8.393402565668908e-06,
|
791 |
+
"loss": 0.2498,
|
792 |
+
"step": 5700
|
793 |
+
},
|
794 |
+
{
|
795 |
+
"epoch": 1.76,
|
796 |
+
"learning_rate": 8.29159030747302e-06,
|
797 |
+
"loss": 0.2372,
|
798 |
+
"step": 5750
|
799 |
+
},
|
800 |
+
{
|
801 |
+
"epoch": 1.77,
|
802 |
+
"learning_rate": 8.189778049277133e-06,
|
803 |
+
"loss": 0.2346,
|
804 |
+
"step": 5800
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 1.79,
|
808 |
+
"learning_rate": 8.087965791081248e-06,
|
809 |
+
"loss": 0.2337,
|
810 |
+
"step": 5850
|
811 |
+
},
|
812 |
+
{
|
813 |
+
"epoch": 1.8,
|
814 |
+
"learning_rate": 7.98615353288536e-06,
|
815 |
+
"loss": 0.2382,
|
816 |
+
"step": 5900
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"epoch": 1.82,
|
820 |
+
"learning_rate": 7.884341274689473e-06,
|
821 |
+
"loss": 0.2454,
|
822 |
+
"step": 5950
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 1.83,
|
826 |
+
"learning_rate": 7.782529016493586e-06,
|
827 |
+
"loss": 0.2626,
|
828 |
+
"step": 6000
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 1.83,
|
832 |
+
"eval_accuracy": 0.8982244188174996,
|
833 |
+
"eval_loss": 0.25814592838287354,
|
834 |
+
"eval_runtime": 22.3236,
|
835 |
+
"eval_samples_per_second": 244.718,
|
836 |
+
"eval_steps_per_second": 15.32,
|
837 |
+
"step": 6000
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 1.85,
|
841 |
+
"learning_rate": 7.680716758297699e-06,
|
842 |
+
"loss": 0.2354,
|
843 |
+
"step": 6050
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 1.86,
|
847 |
+
"learning_rate": 7.578904500101813e-06,
|
848 |
+
"loss": 0.2501,
|
849 |
+
"step": 6100
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 1.88,
|
853 |
+
"learning_rate": 7.4770922419059255e-06,
|
854 |
+
"loss": 0.2369,
|
855 |
+
"step": 6150
|
856 |
+
},
|
857 |
+
{
|
858 |
+
"epoch": 1.89,
|
859 |
+
"learning_rate": 7.37527998371004e-06,
|
860 |
+
"loss": 0.2259,
|
861 |
+
"step": 6200
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 1.91,
|
865 |
+
"learning_rate": 7.273467725514153e-06,
|
866 |
+
"loss": 0.2394,
|
867 |
+
"step": 6250
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 1.92,
|
871 |
+
"learning_rate": 7.171655467318266e-06,
|
872 |
+
"loss": 0.2411,
|
873 |
+
"step": 6300
|
874 |
+
},
|
875 |
+
{
|
876 |
+
"epoch": 1.94,
|
877 |
+
"learning_rate": 7.069843209122379e-06,
|
878 |
+
"loss": 0.2479,
|
879 |
+
"step": 6350
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 1.95,
|
883 |
+
"learning_rate": 6.968030950926492e-06,
|
884 |
+
"loss": 0.2386,
|
885 |
+
"step": 6400
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 1.97,
|
889 |
+
"learning_rate": 6.866218692730605e-06,
|
890 |
+
"loss": 0.2358,
|
891 |
+
"step": 6450
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 1.99,
|
895 |
+
"learning_rate": 6.764406434534719e-06,
|
896 |
+
"loss": 0.2463,
|
897 |
+
"step": 6500
|
898 |
+
},
|
899 |
+
{
|
900 |
+
"epoch": 1.99,
|
901 |
+
"eval_accuracy": 0.8963939227530661,
|
902 |
+
"eval_loss": 0.25200945138931274,
|
903 |
+
"eval_runtime": 22.2934,
|
904 |
+
"eval_samples_per_second": 245.05,
|
905 |
+
"eval_steps_per_second": 15.341,
|
906 |
+
"step": 6500
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 2.0,
|
910 |
+
"learning_rate": 6.662594176338831e-06,
|
911 |
+
"loss": 0.2455,
|
912 |
+
"step": 6550
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 2.02,
|
916 |
+
"learning_rate": 6.560781918142944e-06,
|
917 |
+
"loss": 0.1761,
|
918 |
+
"step": 6600
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"epoch": 2.03,
|
922 |
+
"learning_rate": 6.458969659947059e-06,
|
923 |
+
"loss": 0.1751,
|
924 |
+
"step": 6650
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 2.05,
|
928 |
+
"learning_rate": 6.3571574017511715e-06,
|
929 |
+
"loss": 0.1881,
|
930 |
+
"step": 6700
|
931 |
+
},
|
932 |
+
{
|
933 |
+
"epoch": 2.06,
|
934 |
+
"learning_rate": 6.255345143555285e-06,
|
935 |
+
"loss": 0.1525,
|
936 |
+
"step": 6750
|
937 |
+
},
|
938 |
+
{
|
939 |
+
"epoch": 2.08,
|
940 |
+
"learning_rate": 6.153532885359398e-06,
|
941 |
+
"loss": 0.1923,
|
942 |
+
"step": 6800
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 2.09,
|
946 |
+
"learning_rate": 6.051720627163511e-06,
|
947 |
+
"loss": 0.1872,
|
948 |
+
"step": 6850
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 2.11,
|
952 |
+
"learning_rate": 5.949908368967624e-06,
|
953 |
+
"loss": 0.1793,
|
954 |
+
"step": 6900
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.12,
|
958 |
+
"learning_rate": 5.848096110771737e-06,
|
959 |
+
"loss": 0.2016,
|
960 |
+
"step": 6950
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 2.14,
|
964 |
+
"learning_rate": 5.74628385257585e-06,
|
965 |
+
"loss": 0.1768,
|
966 |
+
"step": 7000
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"epoch": 2.14,
|
970 |
+
"eval_accuracy": 0.8951125755079626,
|
971 |
+
"eval_loss": 0.2794574797153473,
|
972 |
+
"eval_runtime": 22.3376,
|
973 |
+
"eval_samples_per_second": 244.566,
|
974 |
+
"eval_steps_per_second": 15.311,
|
975 |
+
"step": 7000
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 2.15,
|
979 |
+
"learning_rate": 5.644471594379963e-06,
|
980 |
+
"loss": 0.1761,
|
981 |
+
"step": 7050
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 2.17,
|
985 |
+
"learning_rate": 5.542659336184077e-06,
|
986 |
+
"loss": 0.1582,
|
987 |
+
"step": 7100
|
988 |
+
},
|
989 |
+
{
|
990 |
+
"epoch": 2.18,
|
991 |
+
"learning_rate": 5.44084707798819e-06,
|
992 |
+
"loss": 0.152,
|
993 |
+
"step": 7150
|
994 |
+
},
|
995 |
+
{
|
996 |
+
"epoch": 2.2,
|
997 |
+
"learning_rate": 5.339034819792304e-06,
|
998 |
+
"loss": 0.1648,
|
999 |
+
"step": 7200
|
1000 |
+
},
|
1001 |
+
{
|
1002 |
+
"epoch": 2.21,
|
1003 |
+
"learning_rate": 5.237222561596417e-06,
|
1004 |
+
"loss": 0.1817,
|
1005 |
+
"step": 7250
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 2.23,
|
1009 |
+
"learning_rate": 5.1354103034005295e-06,
|
1010 |
+
"loss": 0.1848,
|
1011 |
+
"step": 7300
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 2.24,
|
1015 |
+
"learning_rate": 5.033598045204643e-06,
|
1016 |
+
"loss": 0.1756,
|
1017 |
+
"step": 7350
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 2.26,
|
1021 |
+
"learning_rate": 4.931785787008756e-06,
|
1022 |
+
"loss": 0.2088,
|
1023 |
+
"step": 7400
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 2.28,
|
1027 |
+
"learning_rate": 4.82997352881287e-06,
|
1028 |
+
"loss": 0.173,
|
1029 |
+
"step": 7450
|
1030 |
+
},
|
1031 |
+
{
|
1032 |
+
"epoch": 2.29,
|
1033 |
+
"learning_rate": 4.728161270616982e-06,
|
1034 |
+
"loss": 0.1768,
|
1035 |
+
"step": 7500
|
1036 |
+
},
|
1037 |
+
{
|
1038 |
+
"epoch": 2.29,
|
1039 |
+
"eval_accuracy": 0.8941973274757459,
|
1040 |
+
"eval_loss": 0.30689775943756104,
|
1041 |
+
"eval_runtime": 22.3211,
|
1042 |
+
"eval_samples_per_second": 244.746,
|
1043 |
+
"eval_steps_per_second": 15.322,
|
1044 |
+
"step": 7500
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 2.31,
|
1048 |
+
"learning_rate": 4.626349012421096e-06,
|
1049 |
+
"loss": 0.1592,
|
1050 |
+
"step": 7550
|
1051 |
+
},
|
1052 |
+
{
|
1053 |
+
"epoch": 2.32,
|
1054 |
+
"learning_rate": 4.524536754225209e-06,
|
1055 |
+
"loss": 0.1636,
|
1056 |
+
"step": 7600
|
1057 |
+
},
|
1058 |
+
{
|
1059 |
+
"epoch": 2.34,
|
1060 |
+
"learning_rate": 4.4227244960293225e-06,
|
1061 |
+
"loss": 0.209,
|
1062 |
+
"step": 7650
|
1063 |
+
},
|
1064 |
+
{
|
1065 |
+
"epoch": 2.35,
|
1066 |
+
"learning_rate": 4.320912237833435e-06,
|
1067 |
+
"loss": 0.1723,
|
1068 |
+
"step": 7700
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"epoch": 2.37,
|
1072 |
+
"learning_rate": 4.219099979637549e-06,
|
1073 |
+
"loss": 0.2041,
|
1074 |
+
"step": 7750
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 2.38,
|
1078 |
+
"learning_rate": 4.117287721441662e-06,
|
1079 |
+
"loss": 0.1642,
|
1080 |
+
"step": 7800
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 2.4,
|
1084 |
+
"learning_rate": 4.015475463245775e-06,
|
1085 |
+
"loss": 0.1762,
|
1086 |
+
"step": 7850
|
1087 |
+
},
|
1088 |
+
{
|
1089 |
+
"epoch": 2.41,
|
1090 |
+
"learning_rate": 3.913663205049888e-06,
|
1091 |
+
"loss": 0.1913,
|
1092 |
+
"step": 7900
|
1093 |
+
},
|
1094 |
+
{
|
1095 |
+
"epoch": 2.43,
|
1096 |
+
"learning_rate": 3.8118509468540015e-06,
|
1097 |
+
"loss": 0.1839,
|
1098 |
+
"step": 7950
|
1099 |
+
},
|
1100 |
+
{
|
1101 |
+
"epoch": 2.44,
|
1102 |
+
"learning_rate": 3.7100386886581147e-06,
|
1103 |
+
"loss": 0.1752,
|
1104 |
+
"step": 8000
|
1105 |
+
},
|
1106 |
+
{
|
1107 |
+
"epoch": 2.44,
|
1108 |
+
"eval_accuracy": 0.8971261211788395,
|
1109 |
+
"eval_loss": 0.278341680765152,
|
1110 |
+
"eval_runtime": 22.3616,
|
1111 |
+
"eval_samples_per_second": 244.303,
|
1112 |
+
"eval_steps_per_second": 15.294,
|
1113 |
+
"step": 8000
|
1114 |
+
},
|
1115 |
+
{
|
1116 |
+
"epoch": 2.46,
|
1117 |
+
"learning_rate": 3.6082264304622276e-06,
|
1118 |
+
"loss": 0.1954,
|
1119 |
+
"step": 8050
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 2.47,
|
1123 |
+
"learning_rate": 3.506414172266341e-06,
|
1124 |
+
"loss": 0.1862,
|
1125 |
+
"step": 8100
|
1126 |
+
},
|
1127 |
+
{
|
1128 |
+
"epoch": 2.49,
|
1129 |
+
"learning_rate": 3.4046019140704544e-06,
|
1130 |
+
"loss": 0.16,
|
1131 |
+
"step": 8150
|
1132 |
+
},
|
1133 |
+
{
|
1134 |
+
"epoch": 2.5,
|
1135 |
+
"learning_rate": 3.3027896558745677e-06,
|
1136 |
+
"loss": 0.1632,
|
1137 |
+
"step": 8200
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 2.52,
|
1141 |
+
"learning_rate": 3.2009773976786805e-06,
|
1142 |
+
"loss": 0.166,
|
1143 |
+
"step": 8250
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 2.54,
|
1147 |
+
"learning_rate": 3.0991651394827937e-06,
|
1148 |
+
"loss": 0.1771,
|
1149 |
+
"step": 8300
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 2.55,
|
1153 |
+
"learning_rate": 2.9973528812869074e-06,
|
1154 |
+
"loss": 0.1928,
|
1155 |
+
"step": 8350
|
1156 |
+
},
|
1157 |
+
{
|
1158 |
+
"epoch": 2.57,
|
1159 |
+
"learning_rate": 2.8955406230910206e-06,
|
1160 |
+
"loss": 0.1874,
|
1161 |
+
"step": 8400
|
1162 |
+
},
|
1163 |
+
{
|
1164 |
+
"epoch": 2.58,
|
1165 |
+
"learning_rate": 2.7937283648951334e-06,
|
1166 |
+
"loss": 0.1724,
|
1167 |
+
"step": 8450
|
1168 |
+
},
|
1169 |
+
{
|
1170 |
+
"epoch": 2.6,
|
1171 |
+
"learning_rate": 2.6919161066992467e-06,
|
1172 |
+
"loss": 0.1687,
|
1173 |
+
"step": 8500
|
1174 |
+
},
|
1175 |
+
{
|
1176 |
+
"epoch": 2.6,
|
1177 |
+
"eval_accuracy": 0.899505766062603,
|
1178 |
+
"eval_loss": 0.28996849060058594,
|
1179 |
+
"eval_runtime": 22.3143,
|
1180 |
+
"eval_samples_per_second": 244.821,
|
1181 |
+
"eval_steps_per_second": 15.327,
|
1182 |
+
"step": 8500
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"epoch": 2.61,
|
1186 |
+
"learning_rate": 2.5901038485033603e-06,
|
1187 |
+
"loss": 0.1714,
|
1188 |
+
"step": 8550
|
1189 |
+
},
|
1190 |
+
{
|
1191 |
+
"epoch": 2.63,
|
1192 |
+
"learning_rate": 2.4882915903074735e-06,
|
1193 |
+
"loss": 0.1672,
|
1194 |
+
"step": 8600
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"epoch": 2.64,
|
1198 |
+
"learning_rate": 2.3864793321115864e-06,
|
1199 |
+
"loss": 0.1946,
|
1200 |
+
"step": 8650
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 2.66,
|
1204 |
+
"learning_rate": 2.2846670739156996e-06,
|
1205 |
+
"loss": 0.1746,
|
1206 |
+
"step": 8700
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 2.67,
|
1210 |
+
"learning_rate": 2.182854815719813e-06,
|
1211 |
+
"loss": 0.1698,
|
1212 |
+
"step": 8750
|
1213 |
+
},
|
1214 |
+
{
|
1215 |
+
"epoch": 2.69,
|
1216 |
+
"learning_rate": 2.081042557523926e-06,
|
1217 |
+
"loss": 0.185,
|
1218 |
+
"step": 8800
|
1219 |
+
},
|
1220 |
+
{
|
1221 |
+
"epoch": 2.7,
|
1222 |
+
"learning_rate": 1.9792302993280393e-06,
|
1223 |
+
"loss": 0.1877,
|
1224 |
+
"step": 8850
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 2.72,
|
1228 |
+
"learning_rate": 1.8774180411321525e-06,
|
1229 |
+
"loss": 0.1751,
|
1230 |
+
"step": 8900
|
1231 |
+
},
|
1232 |
+
{
|
1233 |
+
"epoch": 2.73,
|
1234 |
+
"learning_rate": 1.7756057829362658e-06,
|
1235 |
+
"loss": 0.1513,
|
1236 |
+
"step": 8950
|
1237 |
+
},
|
1238 |
+
{
|
1239 |
+
"epoch": 2.75,
|
1240 |
+
"learning_rate": 1.6737935247403788e-06,
|
1241 |
+
"loss": 0.163,
|
1242 |
+
"step": 9000
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 2.75,
|
1246 |
+
"eval_accuracy": 0.8969430715723962,
|
1247 |
+
"eval_loss": 0.2827822268009186,
|
1248 |
+
"eval_runtime": 22.4022,
|
1249 |
+
"eval_samples_per_second": 243.86,
|
1250 |
+
"eval_steps_per_second": 15.266,
|
1251 |
+
"step": 9000
|
1252 |
+
},
|
1253 |
+
{
|
1254 |
+
"epoch": 2.76,
|
1255 |
+
"learning_rate": 1.571981266544492e-06,
|
1256 |
+
"loss": 0.1762,
|
1257 |
+
"step": 9050
|
1258 |
+
},
|
1259 |
+
{
|
1260 |
+
"epoch": 2.78,
|
1261 |
+
"learning_rate": 1.4701690083486052e-06,
|
1262 |
+
"loss": 0.175,
|
1263 |
+
"step": 9100
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 2.79,
|
1267 |
+
"learning_rate": 1.3683567501527185e-06,
|
1268 |
+
"loss": 0.1695,
|
1269 |
+
"step": 9150
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 2.81,
|
1273 |
+
"learning_rate": 1.2665444919568317e-06,
|
1274 |
+
"loss": 0.1681,
|
1275 |
+
"step": 9200
|
1276 |
+
},
|
1277 |
+
{
|
1278 |
+
"epoch": 2.83,
|
1279 |
+
"learning_rate": 1.164732233760945e-06,
|
1280 |
+
"loss": 0.1725,
|
1281 |
+
"step": 9250
|
1282 |
+
},
|
1283 |
+
{
|
1284 |
+
"epoch": 2.84,
|
1285 |
+
"learning_rate": 1.0629199755650582e-06,
|
1286 |
+
"loss": 0.1956,
|
1287 |
+
"step": 9300
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 2.86,
|
1291 |
+
"learning_rate": 9.611077173691712e-07,
|
1292 |
+
"loss": 0.186,
|
1293 |
+
"step": 9350
|
1294 |
+
},
|
1295 |
+
{
|
1296 |
+
"epoch": 2.87,
|
1297 |
+
"learning_rate": 8.592954591732845e-07,
|
1298 |
+
"loss": 0.1692,
|
1299 |
+
"step": 9400
|
1300 |
+
},
|
1301 |
+
{
|
1302 |
+
"epoch": 2.89,
|
1303 |
+
"learning_rate": 7.574832009773978e-07,
|
1304 |
+
"loss": 0.1738,
|
1305 |
+
"step": 9450
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 2.9,
|
1309 |
+
"learning_rate": 6.55670942781511e-07,
|
1310 |
+
"loss": 0.1547,
|
1311 |
+
"step": 9500
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 2.9,
|
1315 |
+
"eval_accuracy": 0.8980413692110562,
|
1316 |
+
"eval_loss": 0.28731775283813477,
|
1317 |
+
"eval_runtime": 22.3397,
|
1318 |
+
"eval_samples_per_second": 244.542,
|
1319 |
+
"eval_steps_per_second": 15.309,
|
1320 |
+
"step": 9500
|
1321 |
+
},
|
1322 |
+
{
|
1323 |
+
"epoch": 2.92,
|
1324 |
+
"learning_rate": 5.538586845856241e-07,
|
1325 |
+
"loss": 0.1529,
|
1326 |
+
"step": 9550
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 2.93,
|
1330 |
+
"learning_rate": 4.5204642638973736e-07,
|
1331 |
+
"loss": 0.1577,
|
1332 |
+
"step": 9600
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 2.95,
|
1336 |
+
"learning_rate": 3.502341681938506e-07,
|
1337 |
+
"loss": 0.1574,
|
1338 |
+
"step": 9650
|
1339 |
+
},
|
1340 |
+
{
|
1341 |
+
"epoch": 2.96,
|
1342 |
+
"learning_rate": 2.484219099979638e-07,
|
1343 |
+
"loss": 0.1777,
|
1344 |
+
"step": 9700
|
1345 |
+
},
|
1346 |
+
{
|
1347 |
+
"epoch": 2.98,
|
1348 |
+
"learning_rate": 1.4660965180207698e-07,
|
1349 |
+
"loss": 0.1511,
|
1350 |
+
"step": 9750
|
1351 |
+
},
|
1352 |
+
{
|
1353 |
+
"epoch": 2.99,
|
1354 |
+
"learning_rate": 4.4797393606190183e-08,
|
1355 |
+
"loss": 0.1514,
|
1356 |
+
"step": 9800
|
1357 |
+
},
|
1358 |
+
{
|
1359 |
+
"epoch": 3.0,
|
1360 |
+
"step": 9822,
|
1361 |
+
"total_flos": 3.4902311012018772e+16,
|
1362 |
+
"train_loss": 0.25818555016693384,
|
1363 |
+
"train_runtime": 4997.6653,
|
1364 |
+
"train_samples_per_second": 62.875,
|
1365 |
+
"train_steps_per_second": 1.965
|
1366 |
+
}
|
1367 |
+
],
|
1368 |
+
"logging_steps": 50,
|
1369 |
+
"max_steps": 9822,
|
1370 |
+
"num_train_epochs": 3,
|
1371 |
+
"save_steps": 500,
|
1372 |
+
"total_flos": 3.4902311012018772e+16,
|
1373 |
+
"trial_name": null,
|
1374 |
+
"trial_params": null
|
1375 |
+
}
|