Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1253 -0
trainer_state.json
ADDED
@@ -0,0 +1,1253 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 19.908045977011493,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1732,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.11494252873563218,
|
13 |
+
"grad_norm": 5.833221435546875,
|
14 |
+
"learning_rate": 2.2988505747126437e-05,
|
15 |
+
"loss": 1.1691,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.22988505747126436,
|
20 |
+
"grad_norm": 3.3890936374664307,
|
21 |
+
"learning_rate": 4.597701149425287e-05,
|
22 |
+
"loss": 0.4962,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.3448275862068966,
|
27 |
+
"grad_norm": 1.4635858535766602,
|
28 |
+
"learning_rate": 6.896551724137931e-05,
|
29 |
+
"loss": 0.2852,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.45977011494252873,
|
34 |
+
"grad_norm": 1.3263287544250488,
|
35 |
+
"learning_rate": 9.195402298850575e-05,
|
36 |
+
"loss": 0.2068,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.5747126436781609,
|
41 |
+
"grad_norm": 1.4933586120605469,
|
42 |
+
"learning_rate": 0.00011494252873563218,
|
43 |
+
"loss": 0.1671,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.6896551724137931,
|
48 |
+
"grad_norm": 3.461280584335327,
|
49 |
+
"learning_rate": 0.00013793103448275863,
|
50 |
+
"loss": 0.163,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.8045977011494253,
|
55 |
+
"grad_norm": 0.8741048574447632,
|
56 |
+
"learning_rate": 0.00016091954022988506,
|
57 |
+
"loss": 0.1625,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.9195402298850575,
|
62 |
+
"grad_norm": 0.9238091111183167,
|
63 |
+
"learning_rate": 0.0001839080459770115,
|
64 |
+
"loss": 0.1292,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 1.0344827586206897,
|
69 |
+
"grad_norm": 1.1964523792266846,
|
70 |
+
"learning_rate": 0.00019999835873288997,
|
71 |
+
"loss": 0.1137,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 1.1494252873563218,
|
76 |
+
"grad_norm": 0.6014072895050049,
|
77 |
+
"learning_rate": 0.0001999691821496584,
|
78 |
+
"loss": 0.1066,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 1.264367816091954,
|
83 |
+
"grad_norm": 0.6493586897850037,
|
84 |
+
"learning_rate": 0.00019990354521250803,
|
85 |
+
"loss": 0.1051,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 1.3793103448275863,
|
90 |
+
"grad_norm": 0.6839510202407837,
|
91 |
+
"learning_rate": 0.00019980147186027586,
|
92 |
+
"loss": 0.0926,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 1.4942528735632183,
|
97 |
+
"grad_norm": 0.5401099920272827,
|
98 |
+
"learning_rate": 0.00019966299932074023,
|
99 |
+
"loss": 0.0895,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 1.6091954022988506,
|
104 |
+
"grad_norm": 1.1563774347305298,
|
105 |
+
"learning_rate": 0.000199488178097043,
|
106 |
+
"loss": 0.0915,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.7241379310344827,
|
111 |
+
"grad_norm": 2.1474926471710205,
|
112 |
+
"learning_rate": 0.00019927707194927066,
|
113 |
+
"loss": 0.0853,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.839080459770115,
|
118 |
+
"grad_norm": 0.7267495393753052,
|
119 |
+
"learning_rate": 0.00019902975787119956,
|
120 |
+
"loss": 0.0873,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.9540229885057472,
|
125 |
+
"grad_norm": 0.5461205244064331,
|
126 |
+
"learning_rate": 0.00019874632606221545,
|
127 |
+
"loss": 0.0739,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 2.0689655172413794,
|
132 |
+
"grad_norm": 0.43381860852241516,
|
133 |
+
"learning_rate": 0.00019842687989441604,
|
134 |
+
"loss": 0.0682,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 2.1839080459770113,
|
139 |
+
"grad_norm": 0.619968593120575,
|
140 |
+
"learning_rate": 0.00019807153587490963,
|
141 |
+
"loss": 0.0725,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 2.2988505747126435,
|
146 |
+
"grad_norm": 0.5531813502311707,
|
147 |
+
"learning_rate": 0.00019768042360332325,
|
148 |
+
"loss": 0.0649,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 2.413793103448276,
|
153 |
+
"grad_norm": 0.4325454831123352,
|
154 |
+
"learning_rate": 0.00019725368572453539,
|
155 |
+
"loss": 0.0629,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 2.528735632183908,
|
160 |
+
"grad_norm": 1.132866382598877,
|
161 |
+
"learning_rate": 0.00019679147787665126,
|
162 |
+
"loss": 0.0597,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 2.6436781609195403,
|
167 |
+
"grad_norm": 0.5158783793449402,
|
168 |
+
"learning_rate": 0.00019629396863423911,
|
169 |
+
"loss": 0.0658,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 2.7586206896551726,
|
174 |
+
"grad_norm": 0.5275442600250244,
|
175 |
+
"learning_rate": 0.0001957613394468484,
|
176 |
+
"loss": 0.0624,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 2.873563218390805,
|
181 |
+
"grad_norm": 0.26212960481643677,
|
182 |
+
"learning_rate": 0.0001951937845728321,
|
183 |
+
"loss": 0.0565,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 2.9885057471264367,
|
188 |
+
"grad_norm": 0.4064450263977051,
|
189 |
+
"learning_rate": 0.00019459151100849784,
|
190 |
+
"loss": 0.0586,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 3.103448275862069,
|
195 |
+
"grad_norm": 0.497156023979187,
|
196 |
+
"learning_rate": 0.0001939547384126128,
|
197 |
+
"loss": 0.0569,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 3.218390804597701,
|
202 |
+
"grad_norm": 0.23238833248615265,
|
203 |
+
"learning_rate": 0.00019328369902629083,
|
204 |
+
"loss": 0.048,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 3.3333333333333335,
|
209 |
+
"grad_norm": 0.33057811856269836,
|
210 |
+
"learning_rate": 0.00019257863758829035,
|
211 |
+
"loss": 0.0508,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 3.4482758620689653,
|
216 |
+
"grad_norm": 0.2923976182937622,
|
217 |
+
"learning_rate": 0.00019183981124575418,
|
218 |
+
"loss": 0.059,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 3.5632183908045976,
|
223 |
+
"grad_norm": 0.40444961190223694,
|
224 |
+
"learning_rate": 0.00019106748946042407,
|
225 |
+
"loss": 0.0589,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 3.67816091954023,
|
230 |
+
"grad_norm": 0.3618530333042145,
|
231 |
+
"learning_rate": 0.00019026195391036338,
|
232 |
+
"loss": 0.0493,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 3.793103448275862,
|
237 |
+
"grad_norm": 0.2655580937862396,
|
238 |
+
"learning_rate": 0.00018942349838722486,
|
239 |
+
"loss": 0.0502,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 3.9080459770114944,
|
244 |
+
"grad_norm": 0.30788642168045044,
|
245 |
+
"learning_rate": 0.0001885524286891002,
|
246 |
+
"loss": 0.0562,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 4.022988505747127,
|
251 |
+
"grad_norm": 0.3801023066043854,
|
252 |
+
"learning_rate": 0.00018764906250899027,
|
253 |
+
"loss": 0.0536,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 4.137931034482759,
|
258 |
+
"grad_norm": 0.34299996495246887,
|
259 |
+
"learning_rate": 0.00018671372931893773,
|
260 |
+
"loss": 0.0583,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 4.252873563218391,
|
265 |
+
"grad_norm": 0.5825142860412598,
|
266 |
+
"learning_rate": 0.0001857467702498633,
|
267 |
+
"loss": 0.0465,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 4.3678160919540225,
|
272 |
+
"grad_norm": 0.46258264780044556,
|
273 |
+
"learning_rate": 0.0001847485379671496,
|
274 |
+
"loss": 0.0469,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 4.482758620689655,
|
279 |
+
"grad_norm": 0.23550616204738617,
|
280 |
+
"learning_rate": 0.0001837193965420188,
|
281 |
+
"loss": 0.0477,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 4.597701149425287,
|
286 |
+
"grad_norm": 0.609255850315094,
|
287 |
+
"learning_rate": 0.00018265972131874987,
|
288 |
+
"loss": 0.0467,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 4.712643678160919,
|
293 |
+
"grad_norm": 0.3701482117176056,
|
294 |
+
"learning_rate": 0.00018156989877778461,
|
295 |
+
"loss": 0.0471,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 4.827586206896552,
|
300 |
+
"grad_norm": 0.4651474356651306,
|
301 |
+
"learning_rate": 0.00018045032639477194,
|
302 |
+
"loss": 0.0434,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 4.942528735632184,
|
307 |
+
"grad_norm": 0.34028705954551697,
|
308 |
+
"learning_rate": 0.00017930141249560233,
|
309 |
+
"loss": 0.0452,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 5.057471264367816,
|
314 |
+
"grad_norm": 0.2748933732509613,
|
315 |
+
"learning_rate": 0.00017812357610748488,
|
316 |
+
"loss": 0.0413,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 5.172413793103448,
|
321 |
+
"grad_norm": 0.4612014591693878,
|
322 |
+
"learning_rate": 0.00017691724680612118,
|
323 |
+
"loss": 0.0423,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 5.287356321839081,
|
328 |
+
"grad_norm": 0.37961891293525696,
|
329 |
+
"learning_rate": 0.00017568286455903258,
|
330 |
+
"loss": 0.0422,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 5.402298850574713,
|
335 |
+
"grad_norm": 0.3245999813079834,
|
336 |
+
"learning_rate": 0.00017442087956509665,
|
337 |
+
"loss": 0.0396,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 5.517241379310345,
|
342 |
+
"grad_norm": 0.5230941772460938,
|
343 |
+
"learning_rate": 0.00017313175209035268,
|
344 |
+
"loss": 0.0405,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 5.6321839080459775,
|
349 |
+
"grad_norm": 0.2870311737060547,
|
350 |
+
"learning_rate": 0.00017181595230013525,
|
351 |
+
"loss": 0.0343,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 5.747126436781609,
|
356 |
+
"grad_norm": 0.2876773774623871,
|
357 |
+
"learning_rate": 0.00017047396008759754,
|
358 |
+
"loss": 0.0436,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 5.862068965517241,
|
363 |
+
"grad_norm": 0.4095667898654938,
|
364 |
+
"learning_rate": 0.00016910626489868649,
|
365 |
+
"loss": 0.0408,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 5.977011494252873,
|
370 |
+
"grad_norm": 0.377605140209198,
|
371 |
+
"learning_rate": 0.00016771336555363418,
|
372 |
+
"loss": 0.0415,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 6.091954022988506,
|
377 |
+
"grad_norm": 0.28248798847198486,
|
378 |
+
"learning_rate": 0.00016629577006503009,
|
379 |
+
"loss": 0.0386,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 6.206896551724138,
|
384 |
+
"grad_norm": 0.36199840903282166,
|
385 |
+
"learning_rate": 0.0001648539954525409,
|
386 |
+
"loss": 0.0405,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 6.32183908045977,
|
391 |
+
"grad_norm": 0.2778664529323578,
|
392 |
+
"learning_rate": 0.00016338856755434503,
|
393 |
+
"loss": 0.0359,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 6.436781609195402,
|
398 |
+
"grad_norm": 0.23418012261390686,
|
399 |
+
"learning_rate": 0.00016190002083535122,
|
400 |
+
"loss": 0.0382,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 6.551724137931035,
|
405 |
+
"grad_norm": 0.3027312457561493,
|
406 |
+
"learning_rate": 0.00016038889819227045,
|
407 |
+
"loss": 0.0394,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 6.666666666666667,
|
412 |
+
"grad_norm": 0.2858007550239563,
|
413 |
+
"learning_rate": 0.00015885575075561326,
|
414 |
+
"loss": 0.042,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 6.781609195402299,
|
419 |
+
"grad_norm": 0.2762337923049927,
|
420 |
+
"learning_rate": 0.00015730113768868312,
|
421 |
+
"loss": 0.039,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 6.896551724137931,
|
426 |
+
"grad_norm": 0.40732237696647644,
|
427 |
+
"learning_rate": 0.0001557256259836412,
|
428 |
+
"loss": 0.0404,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 7.011494252873563,
|
433 |
+
"grad_norm": 0.36847805976867676,
|
434 |
+
"learning_rate": 0.00015412979025471488,
|
435 |
+
"loss": 0.0368,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 7.126436781609195,
|
440 |
+
"grad_norm": 0.2492237538099289,
|
441 |
+
"learning_rate": 0.00015251421252862707,
|
442 |
+
"loss": 0.0336,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 7.241379310344827,
|
447 |
+
"grad_norm": 0.2626156210899353,
|
448 |
+
"learning_rate": 0.00015087948203232156,
|
449 |
+
"loss": 0.0352,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 7.35632183908046,
|
454 |
+
"grad_norm": 0.6365396976470947,
|
455 |
+
"learning_rate": 0.00014922619497806277,
|
456 |
+
"loss": 0.0342,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 7.471264367816092,
|
461 |
+
"grad_norm": 0.3000635802745819,
|
462 |
+
"learning_rate": 0.00014755495434598745,
|
463 |
+
"loss": 0.037,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 7.586206896551724,
|
468 |
+
"grad_norm": 0.21869853138923645,
|
469 |
+
"learning_rate": 0.0001458663696641884,
|
470 |
+
"loss": 0.0365,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 7.7011494252873565,
|
475 |
+
"grad_norm": 0.22284150123596191,
|
476 |
+
"learning_rate": 0.0001441610567864096,
|
477 |
+
"loss": 0.035,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 7.816091954022989,
|
482 |
+
"grad_norm": 0.267621248960495,
|
483 |
+
"learning_rate": 0.00014243963766743495,
|
484 |
+
"loss": 0.029,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 7.931034482758621,
|
489 |
+
"grad_norm": 0.2817297875881195,
|
490 |
+
"learning_rate": 0.00014070274013625096,
|
491 |
+
"loss": 0.0303,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 8.045977011494253,
|
496 |
+
"grad_norm": 0.3712492287158966,
|
497 |
+
"learning_rate": 0.00013895099766706784,
|
498 |
+
"loss": 0.0297,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 8.160919540229886,
|
503 |
+
"grad_norm": 0.4549995958805084,
|
504 |
+
"learning_rate": 0.00013718504914828135,
|
505 |
+
"loss": 0.033,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 8.275862068965518,
|
510 |
+
"grad_norm": 0.28695234656333923,
|
511 |
+
"learning_rate": 0.00013540553864945976,
|
512 |
+
"loss": 0.0306,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 8.39080459770115,
|
517 |
+
"grad_norm": 0.34577062726020813,
|
518 |
+
"learning_rate": 0.00013361311518644172,
|
519 |
+
"loss": 0.0325,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 8.505747126436782,
|
524 |
+
"grad_norm": 0.3214464783668518,
|
525 |
+
"learning_rate": 0.00013180843248462973,
|
526 |
+
"loss": 0.0337,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 8.620689655172415,
|
531 |
+
"grad_norm": 0.33310961723327637,
|
532 |
+
"learning_rate": 0.00012999214874056595,
|
533 |
+
"loss": 0.0344,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 8.735632183908045,
|
538 |
+
"grad_norm": 0.25606226921081543,
|
539 |
+
"learning_rate": 0.00012816492638187762,
|
540 |
+
"loss": 0.0396,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 8.850574712643677,
|
545 |
+
"grad_norm": 0.36330148577690125,
|
546 |
+
"learning_rate": 0.00012632743182567905,
|
547 |
+
"loss": 0.0348,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 8.96551724137931,
|
552 |
+
"grad_norm": 0.314394474029541,
|
553 |
+
"learning_rate": 0.00012448033523551865,
|
554 |
+
"loss": 0.0399,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 9.080459770114942,
|
559 |
+
"grad_norm": 0.28129351139068604,
|
560 |
+
"learning_rate": 0.00012262431027695964,
|
561 |
+
"loss": 0.0298,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 9.195402298850574,
|
566 |
+
"grad_norm": 0.256881982088089,
|
567 |
+
"learning_rate": 0.00012076003387188353,
|
568 |
+
"loss": 0.0292,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 9.310344827586206,
|
573 |
+
"grad_norm": 0.1919921189546585,
|
574 |
+
"learning_rate": 0.00011888818595160584,
|
575 |
+
"loss": 0.0269,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 9.425287356321839,
|
580 |
+
"grad_norm": 0.2719796895980835,
|
581 |
+
"learning_rate": 0.00011700944920889436,
|
582 |
+
"loss": 0.0265,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 9.540229885057471,
|
587 |
+
"grad_norm": 0.2269754707813263,
|
588 |
+
"learning_rate": 0.00011512450884898022,
|
589 |
+
"loss": 0.0316,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 9.655172413793103,
|
594 |
+
"grad_norm": 0.23504453897476196,
|
595 |
+
"learning_rate": 0.00011323405233965256,
|
596 |
+
"loss": 0.0273,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 9.770114942528735,
|
601 |
+
"grad_norm": 0.22570957243442535,
|
602 |
+
"learning_rate": 0.00011133876916052821,
|
603 |
+
"loss": 0.0304,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 9.885057471264368,
|
608 |
+
"grad_norm": 0.19824576377868652,
|
609 |
+
"learning_rate": 0.00010943935055158734,
|
610 |
+
"loss": 0.0283,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 10.0,
|
615 |
+
"grad_norm": 0.41852012276649475,
|
616 |
+
"learning_rate": 0.00010753648926106723,
|
617 |
+
"loss": 0.0319,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 10.114942528735632,
|
622 |
+
"grad_norm": 0.20548714697360992,
|
623 |
+
"learning_rate": 0.00010563087929280613,
|
624 |
+
"loss": 0.0285,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 10.229885057471265,
|
629 |
+
"grad_norm": 0.22767336666584015,
|
630 |
+
"learning_rate": 0.00010372321565312872,
|
631 |
+
"loss": 0.031,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 10.344827586206897,
|
636 |
+
"grad_norm": 0.20542040467262268,
|
637 |
+
"learning_rate": 0.00010181419409736647,
|
638 |
+
"loss": 0.0316,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 10.459770114942529,
|
643 |
+
"grad_norm": 0.3105849027633667,
|
644 |
+
"learning_rate": 9.990451087610448e-05,
|
645 |
+
"loss": 0.027,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 10.574712643678161,
|
650 |
+
"grad_norm": 0.31816890835762024,
|
651 |
+
"learning_rate": 9.799486248124775e-05,
|
652 |
+
"loss": 0.025,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 10.689655172413794,
|
657 |
+
"grad_norm": 0.3295416235923767,
|
658 |
+
"learning_rate": 9.608594539199957e-05,
|
659 |
+
"loss": 0.0247,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 10.804597701149426,
|
664 |
+
"grad_norm": 0.17071272432804108,
|
665 |
+
"learning_rate": 9.417845582084448e-05,
|
666 |
+
"loss": 0.0291,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 10.919540229885058,
|
671 |
+
"grad_norm": 0.189552441239357,
|
672 |
+
"learning_rate": 9.227308945962827e-05,
|
673 |
+
"loss": 0.0243,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 11.03448275862069,
|
678 |
+
"grad_norm": 0.3179641664028168,
|
679 |
+
"learning_rate": 9.037054122582839e-05,
|
680 |
+
"loss": 0.0308,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 11.149425287356323,
|
685 |
+
"grad_norm": 0.3051457703113556,
|
686 |
+
"learning_rate": 8.847150500910618e-05,
|
687 |
+
"loss": 0.0275,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 11.264367816091955,
|
692 |
+
"grad_norm": 0.29757606983184814,
|
693 |
+
"learning_rate": 8.657667341823448e-05,
|
694 |
+
"loss": 0.0264,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 11.379310344827585,
|
699 |
+
"grad_norm": 0.2845855951309204,
|
700 |
+
"learning_rate": 8.4686737528492e-05,
|
701 |
+
"loss": 0.0249,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 11.494252873563218,
|
706 |
+
"grad_norm": 0.239737406373024,
|
707 |
+
"learning_rate": 8.280238662961728e-05,
|
708 |
+
"loss": 0.027,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 11.60919540229885,
|
713 |
+
"grad_norm": 0.2692360281944275,
|
714 |
+
"learning_rate": 8.092430797441364e-05,
|
715 |
+
"loss": 0.0216,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 11.724137931034482,
|
720 |
+
"grad_norm": 0.18495500087738037,
|
721 |
+
"learning_rate": 7.905318652809728e-05,
|
722 |
+
"loss": 0.0255,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 11.839080459770114,
|
727 |
+
"grad_norm": 0.2230875939130783,
|
728 |
+
"learning_rate": 7.718970471847923e-05,
|
729 |
+
"loss": 0.0262,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 11.954022988505747,
|
734 |
+
"grad_norm": 0.14376775920391083,
|
735 |
+
"learning_rate": 7.53345421870735e-05,
|
736 |
+
"loss": 0.0209,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 12.068965517241379,
|
741 |
+
"grad_norm": 0.20623371005058289,
|
742 |
+
"learning_rate": 7.348837554122057e-05,
|
743 |
+
"loss": 0.0192,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 12.183908045977011,
|
748 |
+
"grad_norm": 0.27209600806236267,
|
749 |
+
"learning_rate": 7.165187810731823e-05,
|
750 |
+
"loss": 0.0208,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 12.298850574712644,
|
755 |
+
"grad_norm": 0.19447851181030273,
|
756 |
+
"learning_rate": 6.982571968524847e-05,
|
757 |
+
"loss": 0.0201,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 12.413793103448276,
|
762 |
+
"grad_norm": 0.18613241612911224,
|
763 |
+
"learning_rate": 6.801056630409098e-05,
|
764 |
+
"loss": 0.0248,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 12.528735632183908,
|
769 |
+
"grad_norm": 0.24156583845615387,
|
770 |
+
"learning_rate": 6.620707997921157e-05,
|
771 |
+
"loss": 0.0197,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 12.64367816091954,
|
776 |
+
"grad_norm": 0.16912145912647247,
|
777 |
+
"learning_rate": 6.441591847081476e-05,
|
778 |
+
"loss": 0.022,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 12.758620689655173,
|
783 |
+
"grad_norm": 0.14165754616260529,
|
784 |
+
"learning_rate": 6.263773504404801e-05,
|
785 |
+
"loss": 0.0199,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 12.873563218390805,
|
790 |
+
"grad_norm": 0.3424724340438843,
|
791 |
+
"learning_rate": 6.087317823074565e-05,
|
792 |
+
"loss": 0.0209,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 12.988505747126437,
|
797 |
+
"grad_norm": 0.2658204138278961,
|
798 |
+
"learning_rate": 5.912289159289883e-05,
|
799 |
+
"loss": 0.0242,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 13.10344827586207,
|
804 |
+
"grad_norm": 0.21321730315685272,
|
805 |
+
"learning_rate": 5.7387513487938386e-05,
|
806 |
+
"loss": 0.0216,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 13.218390804597702,
|
811 |
+
"grad_norm": 0.2854823172092438,
|
812 |
+
"learning_rate": 5.566767683591553e-05,
|
813 |
+
"loss": 0.0227,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 13.333333333333334,
|
818 |
+
"grad_norm": 0.28919658064842224,
|
819 |
+
"learning_rate": 5.396400888866601e-05,
|
820 |
+
"loss": 0.0195,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 13.448275862068966,
|
825 |
+
"grad_norm": 0.22510255873203278,
|
826 |
+
"learning_rate": 5.2277131001041125e-05,
|
827 |
+
"loss": 0.0241,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 13.563218390804598,
|
832 |
+
"grad_norm": 0.21545900404453278,
|
833 |
+
"learning_rate": 5.060765840429e-05,
|
834 |
+
"loss": 0.023,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 13.678160919540229,
|
839 |
+
"grad_norm": 0.20618782937526703,
|
840 |
+
"learning_rate": 4.8956199981674656e-05,
|
841 |
+
"loss": 0.0181,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 13.793103448275861,
|
846 |
+
"grad_norm": 0.22331970930099487,
|
847 |
+
"learning_rate": 4.7323358046400844e-05,
|
848 |
+
"loss": 0.0212,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 13.908045977011493,
|
853 |
+
"grad_norm": 0.14791706204414368,
|
854 |
+
"learning_rate": 4.570972812194457e-05,
|
855 |
+
"loss": 0.0195,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 14.022988505747126,
|
860 |
+
"grad_norm": 0.1526448130607605,
|
861 |
+
"learning_rate": 4.4115898724855455e-05,
|
862 |
+
"loss": 0.0188,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 14.137931034482758,
|
867 |
+
"grad_norm": 0.18956783413887024,
|
868 |
+
"learning_rate": 4.254245115011506e-05,
|
869 |
+
"loss": 0.0188,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 14.25287356321839,
|
874 |
+
"grad_norm": 0.1313301920890808,
|
875 |
+
"learning_rate": 4.098995925912972e-05,
|
876 |
+
"loss": 0.019,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 14.367816091954023,
|
881 |
+
"grad_norm": 0.13764789700508118,
|
882 |
+
"learning_rate": 3.945898927043372e-05,
|
883 |
+
"loss": 0.0175,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 14.482758620689655,
|
888 |
+
"grad_norm": 0.19556942582130432,
|
889 |
+
"learning_rate": 3.7950099553180804e-05,
|
890 |
+
"loss": 0.0196,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 14.597701149425287,
|
895 |
+
"grad_norm": 0.14027345180511475,
|
896 |
+
"learning_rate": 3.646384042349764e-05,
|
897 |
+
"loss": 0.0177,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 14.71264367816092,
|
902 |
+
"grad_norm": 0.2918284833431244,
|
903 |
+
"learning_rate": 3.500075394377511e-05,
|
904 |
+
"loss": 0.0204,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 14.827586206896552,
|
909 |
+
"grad_norm": 0.12948164343833923,
|
910 |
+
"learning_rate": 3.3561373724969224e-05,
|
911 |
+
"loss": 0.0188,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 14.942528735632184,
|
916 |
+
"grad_norm": 0.1773224174976349,
|
917 |
+
"learning_rate": 3.214622473198492e-05,
|
918 |
+
"loss": 0.0212,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 15.057471264367816,
|
923 |
+
"grad_norm": 0.29863160848617554,
|
924 |
+
"learning_rate": 3.075582309221289e-05,
|
925 |
+
"loss": 0.0157,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 15.172413793103448,
|
930 |
+
"grad_norm": 0.18764474987983704,
|
931 |
+
"learning_rate": 2.939067590728972e-05,
|
932 |
+
"loss": 0.0175,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 15.28735632183908,
|
937 |
+
"grad_norm": 0.16273990273475647,
|
938 |
+
"learning_rate": 2.8051281068149803e-05,
|
939 |
+
"loss": 0.0135,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 15.402298850574713,
|
944 |
+
"grad_norm": 0.25088945031166077,
|
945 |
+
"learning_rate": 2.673812707343669e-05,
|
946 |
+
"loss": 0.0242,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 15.517241379310345,
|
951 |
+
"grad_norm": 0.25521960854530334,
|
952 |
+
"learning_rate": 2.545169285133965e-05,
|
953 |
+
"loss": 0.0188,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 15.632183908045977,
|
958 |
+
"grad_norm": 0.15780223906040192,
|
959 |
+
"learning_rate": 2.4192447584921195e-05,
|
960 |
+
"loss": 0.0194,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 15.74712643678161,
|
965 |
+
"grad_norm": 0.13658417761325836,
|
966 |
+
"learning_rate": 2.296085054099828e-05,
|
967 |
+
"loss": 0.0179,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 15.862068965517242,
|
972 |
+
"grad_norm": 0.14593394100666046,
|
973 |
+
"learning_rate": 2.175735090264058e-05,
|
974 |
+
"loss": 0.016,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 15.977011494252874,
|
979 |
+
"grad_norm": 0.20093883574008942,
|
980 |
+
"learning_rate": 2.0582387605346088e-05,
|
981 |
+
"loss": 0.0157,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 16.091954022988507,
|
986 |
+
"grad_norm": 0.22261527180671692,
|
987 |
+
"learning_rate": 1.943638917695453e-05,
|
988 |
+
"loss": 0.0175,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 16.20689655172414,
|
993 |
+
"grad_norm": 0.17486433684825897,
|
994 |
+
"learning_rate": 1.831977358135625e-05,
|
995 |
+
"loss": 0.0166,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 16.32183908045977,
|
1000 |
+
"grad_norm": 0.2138216346502304,
|
1001 |
+
"learning_rate": 1.723294806605428e-05,
|
1002 |
+
"loss": 0.0146,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 16.436781609195403,
|
1007 |
+
"grad_norm": 0.20112960040569305,
|
1008 |
+
"learning_rate": 1.6176309013634517e-05,
|
1009 |
+
"loss": 0.0159,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 16.551724137931036,
|
1014 |
+
"grad_norm": 0.15377485752105713,
|
1015 |
+
"learning_rate": 1.5150241797198883e-05,
|
1016 |
+
"loss": 0.016,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 16.666666666666668,
|
1021 |
+
"grad_norm": 0.23132722079753876,
|
1022 |
+
"learning_rate": 1.415512063981339e-05,
|
1023 |
+
"loss": 0.0134,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 16.7816091954023,
|
1028 |
+
"grad_norm": 0.15262600779533386,
|
1029 |
+
"learning_rate": 1.3191308478023212e-05,
|
1030 |
+
"loss": 0.017,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 16.896551724137932,
|
1035 |
+
"grad_norm": 0.0991855040192604,
|
1036 |
+
"learning_rate": 1.2259156829483842e-05,
|
1037 |
+
"loss": 0.0167,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 17.011494252873565,
|
1042 |
+
"grad_norm": 0.12278055399656296,
|
1043 |
+
"learning_rate": 1.1359005664756994e-05,
|
1044 |
+
"loss": 0.0146,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 17.126436781609197,
|
1049 |
+
"grad_norm": 0.17124158143997192,
|
1050 |
+
"learning_rate": 1.0491183283317997e-05,
|
1051 |
+
"loss": 0.0173,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 17.24137931034483,
|
1056 |
+
"grad_norm": 0.1300356686115265,
|
1057 |
+
"learning_rate": 9.656006193819633e-06,
|
1058 |
+
"loss": 0.0143,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 17.35632183908046,
|
1063 |
+
"grad_norm": 0.17946338653564453,
|
1064 |
+
"learning_rate": 8.853778998656537e-06,
|
1065 |
+
"loss": 0.0154,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 17.47126436781609,
|
1070 |
+
"grad_norm": 0.28736400604248047,
|
1071 |
+
"learning_rate": 8.084794282871689e-06,
|
1072 |
+
"loss": 0.0166,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 17.586206896551722,
|
1077 |
+
"grad_norm": 0.13112574815750122,
|
1078 |
+
"learning_rate": 7.3493325074461165e-06,
|
1079 |
+
"loss": 0.0132,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 17.701149425287355,
|
1084 |
+
"grad_norm": 0.12864838540554047,
|
1085 |
+
"learning_rate": 6.647661907010183e-06,
|
1086 |
+
"loss": 0.0171,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 17.816091954022987,
|
1091 |
+
"grad_norm": 0.16958807408809662,
|
1092 |
+
"learning_rate": 5.980038392014309e-06,
|
1093 |
+
"loss": 0.0161,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 17.93103448275862,
|
1098 |
+
"grad_norm": 0.36121729016304016,
|
1099 |
+
"learning_rate": 5.3467054553941405e-06,
|
1100 |
+
"loss": 0.0158,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 18.04597701149425,
|
1105 |
+
"grad_norm": 0.2107989490032196,
|
1106 |
+
"learning_rate": 4.7478940837649924e-06,
|
1107 |
+
"loss": 0.0147,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 18.160919540229884,
|
1112 |
+
"grad_norm": 0.15654149651527405,
|
1113 |
+
"learning_rate": 4.183822673177229e-06,
|
1114 |
+
"loss": 0.0164,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 18.275862068965516,
|
1119 |
+
"grad_norm": 0.1438828557729721,
|
1120 |
+
"learning_rate": 3.6546969494637986e-06,
|
1121 |
+
"loss": 0.0131,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 18.39080459770115,
|
1126 |
+
"grad_norm": 0.2543192207813263,
|
1127 |
+
"learning_rate": 3.16070989320868e-06,
|
1128 |
+
"loss": 0.0157,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 18.50574712643678,
|
1133 |
+
"grad_norm": 0.13453112542629242,
|
1134 |
+
"learning_rate": 2.702041669363875e-06,
|
1135 |
+
"loss": 0.017,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 18.620689655172413,
|
1140 |
+
"grad_norm": 0.16369780898094177,
|
1141 |
+
"learning_rate": 2.2788595615403474e-06,
|
1142 |
+
"loss": 0.0157,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 18.735632183908045,
|
1147 |
+
"grad_norm": 0.14639818668365479,
|
1148 |
+
"learning_rate": 1.8913179109969482e-06,
|
1149 |
+
"loss": 0.0122,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 18.850574712643677,
|
1154 |
+
"grad_norm": 0.23813354969024658,
|
1155 |
+
"learning_rate": 1.5395580603498328e-06,
|
1156 |
+
"loss": 0.0157,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 18.96551724137931,
|
1161 |
+
"grad_norm": 0.15577834844589233,
|
1162 |
+
"learning_rate": 1.2237083020224526e-06,
|
1163 |
+
"loss": 0.0144,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 19.080459770114942,
|
1168 |
+
"grad_norm": 0.06880059838294983,
|
1169 |
+
"learning_rate": 9.438838314553056e-07,
|
1170 |
+
"loss": 0.0109,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 19.195402298850574,
|
1175 |
+
"grad_norm": 0.19819270074367523,
|
1176 |
+
"learning_rate": 7.001867050923095e-07,
|
1177 |
+
"loss": 0.0134,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 19.310344827586206,
|
1182 |
+
"grad_norm": 0.10673543065786362,
|
1183 |
+
"learning_rate": 4.92705803159188e-07,
|
1184 |
+
"loss": 0.0155,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 19.42528735632184,
|
1189 |
+
"grad_norm": 0.16529639065265656,
|
1190 |
+
"learning_rate": 3.2151679724748975e-07,
|
1191 |
+
"loss": 0.0175,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 19.54022988505747,
|
1196 |
+
"grad_norm": 0.1206677109003067,
|
1197 |
+
"learning_rate": 1.8668212271585327e-07,
|
1198 |
+
"loss": 0.0188,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 19.655172413793103,
|
1203 |
+
"grad_norm": 0.10180158913135529,
|
1204 |
+
"learning_rate": 8.825095591891152e-08,
|
1205 |
+
"loss": 0.0158,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 19.770114942528735,
|
1210 |
+
"grad_norm": 0.14544282853603363,
|
1211 |
+
"learning_rate": 2.625919627188278e-08,
|
1212 |
+
"loss": 0.015,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 19.885057471264368,
|
1217 |
+
"grad_norm": 0.14054107666015625,
|
1218 |
+
"learning_rate": 7.294531574553176e-10,
|
1219 |
+
"loss": 0.0139,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 19.908045977011493,
|
1224 |
+
"step": 1732,
|
1225 |
+
"total_flos": 2.431760592612004e+17,
|
1226 |
+
"train_loss": 0.04785273781403314,
|
1227 |
+
"train_runtime": 1980.9382,
|
1228 |
+
"train_samples_per_second": 55.957,
|
1229 |
+
"train_steps_per_second": 0.874
|
1230 |
+
}
|
1231 |
+
],
|
1232 |
+
"logging_steps": 10,
|
1233 |
+
"max_steps": 1732,
|
1234 |
+
"num_input_tokens_seen": 0,
|
1235 |
+
"num_train_epochs": 20,
|
1236 |
+
"save_steps": 10000,
|
1237 |
+
"stateful_callbacks": {
|
1238 |
+
"TrainerControl": {
|
1239 |
+
"args": {
|
1240 |
+
"should_epoch_stop": false,
|
1241 |
+
"should_evaluate": false,
|
1242 |
+
"should_log": false,
|
1243 |
+
"should_save": true,
|
1244 |
+
"should_training_stop": true
|
1245 |
+
},
|
1246 |
+
"attributes": {}
|
1247 |
+
}
|
1248 |
+
},
|
1249 |
+
"total_flos": 2.431760592612004e+17,
|
1250 |
+
"train_batch_size": 64,
|
1251 |
+
"trial_name": null,
|
1252 |
+
"trial_params": null
|
1253 |
+
}
|