Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1176 -0
trainer_state.json
ADDED
@@ -0,0 +1,1176 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 5.975903614457831,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 162,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.03614457831325301,
|
13 |
+
"grad_norm": 2.8209025859832764,
|
14 |
+
"learning_rate": 5.882352941176471e-07,
|
15 |
+
"loss": 0.7616,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.07228915662650602,
|
20 |
+
"grad_norm": 2.7777888774871826,
|
21 |
+
"learning_rate": 1.1764705882352942e-06,
|
22 |
+
"loss": 0.7427,
|
23 |
+
"step": 2
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.10843373493975904,
|
27 |
+
"grad_norm": 2.813636064529419,
|
28 |
+
"learning_rate": 1.7647058823529414e-06,
|
29 |
+
"loss": 0.744,
|
30 |
+
"step": 3
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.14457831325301204,
|
34 |
+
"grad_norm": 2.675807476043701,
|
35 |
+
"learning_rate": 2.3529411764705885e-06,
|
36 |
+
"loss": 0.7432,
|
37 |
+
"step": 4
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.18072289156626506,
|
41 |
+
"grad_norm": 2.662332773208618,
|
42 |
+
"learning_rate": 2.9411764705882355e-06,
|
43 |
+
"loss": 0.7608,
|
44 |
+
"step": 5
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.21686746987951808,
|
48 |
+
"grad_norm": 2.7628424167633057,
|
49 |
+
"learning_rate": 3.529411764705883e-06,
|
50 |
+
"loss": 0.7493,
|
51 |
+
"step": 6
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.25301204819277107,
|
55 |
+
"grad_norm": 2.4826860427856445,
|
56 |
+
"learning_rate": 4.11764705882353e-06,
|
57 |
+
"loss": 0.7395,
|
58 |
+
"step": 7
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.2891566265060241,
|
62 |
+
"grad_norm": 2.5429294109344482,
|
63 |
+
"learning_rate": 4.705882352941177e-06,
|
64 |
+
"loss": 0.7363,
|
65 |
+
"step": 8
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.3253012048192771,
|
69 |
+
"grad_norm": 1.8845857381820679,
|
70 |
+
"learning_rate": 5.294117647058824e-06,
|
71 |
+
"loss": 0.7224,
|
72 |
+
"step": 9
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.3614457831325301,
|
76 |
+
"grad_norm": 1.3661439418792725,
|
77 |
+
"learning_rate": 5.882352941176471e-06,
|
78 |
+
"loss": 0.7079,
|
79 |
+
"step": 10
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.39759036144578314,
|
83 |
+
"grad_norm": 1.1115751266479492,
|
84 |
+
"learning_rate": 6.470588235294119e-06,
|
85 |
+
"loss": 0.6976,
|
86 |
+
"step": 11
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.43373493975903615,
|
90 |
+
"grad_norm": 0.9719029068946838,
|
91 |
+
"learning_rate": 7.058823529411766e-06,
|
92 |
+
"loss": 0.6877,
|
93 |
+
"step": 12
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.46987951807228917,
|
97 |
+
"grad_norm": 1.5337737798690796,
|
98 |
+
"learning_rate": 7.647058823529411e-06,
|
99 |
+
"loss": 0.6651,
|
100 |
+
"step": 13
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.5060240963855421,
|
104 |
+
"grad_norm": 1.6938939094543457,
|
105 |
+
"learning_rate": 8.23529411764706e-06,
|
106 |
+
"loss": 0.6765,
|
107 |
+
"step": 14
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.5421686746987951,
|
111 |
+
"grad_norm": 1.6760753393173218,
|
112 |
+
"learning_rate": 8.823529411764707e-06,
|
113 |
+
"loss": 0.6804,
|
114 |
+
"step": 15
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.5783132530120482,
|
118 |
+
"grad_norm": 1.392161250114441,
|
119 |
+
"learning_rate": 9.411764705882354e-06,
|
120 |
+
"loss": 0.6565,
|
121 |
+
"step": 16
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.6144578313253012,
|
125 |
+
"grad_norm": 1.2017204761505127,
|
126 |
+
"learning_rate": 1e-05,
|
127 |
+
"loss": 0.6346,
|
128 |
+
"step": 17
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.6506024096385542,
|
132 |
+
"grad_norm": 1.2758034467697144,
|
133 |
+
"learning_rate": 9.99882649009242e-06,
|
134 |
+
"loss": 0.669,
|
135 |
+
"step": 18
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.6867469879518072,
|
139 |
+
"grad_norm": 0.9672208428382874,
|
140 |
+
"learning_rate": 9.995306511219885e-06,
|
141 |
+
"loss": 0.6494,
|
142 |
+
"step": 19
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.7228915662650602,
|
146 |
+
"grad_norm": 0.7265082597732544,
|
147 |
+
"learning_rate": 9.989441715674422e-06,
|
148 |
+
"loss": 0.6397,
|
149 |
+
"step": 20
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.7590361445783133,
|
153 |
+
"grad_norm": 0.605043351650238,
|
154 |
+
"learning_rate": 9.981234856414306e-06,
|
155 |
+
"loss": 0.645,
|
156 |
+
"step": 21
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.7951807228915663,
|
160 |
+
"grad_norm": 0.6995558142662048,
|
161 |
+
"learning_rate": 9.970689785771798e-06,
|
162 |
+
"loss": 0.6516,
|
163 |
+
"step": 22
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.8313253012048193,
|
167 |
+
"grad_norm": 0.7498184442520142,
|
168 |
+
"learning_rate": 9.957811453644848e-06,
|
169 |
+
"loss": 0.6296,
|
170 |
+
"step": 23
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.8674698795180723,
|
174 |
+
"grad_norm": 0.7003129124641418,
|
175 |
+
"learning_rate": 9.942605905173593e-06,
|
176 |
+
"loss": 0.6335,
|
177 |
+
"step": 24
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.9036144578313253,
|
181 |
+
"grad_norm": 0.721664309501648,
|
182 |
+
"learning_rate": 9.925080277902743e-06,
|
183 |
+
"loss": 0.641,
|
184 |
+
"step": 25
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.9397590361445783,
|
188 |
+
"grad_norm": 0.5983788371086121,
|
189 |
+
"learning_rate": 9.905242798431196e-06,
|
190 |
+
"loss": 0.6236,
|
191 |
+
"step": 26
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.9759036144578314,
|
195 |
+
"grad_norm": 0.5488296747207642,
|
196 |
+
"learning_rate": 9.883102778550434e-06,
|
197 |
+
"loss": 0.6434,
|
198 |
+
"step": 27
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 1.036144578313253,
|
202 |
+
"grad_norm": 1.0052872896194458,
|
203 |
+
"learning_rate": 9.858670610873528e-06,
|
204 |
+
"loss": 1.2338,
|
205 |
+
"step": 28
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 1.072289156626506,
|
209 |
+
"grad_norm": 0.4585392475128174,
|
210 |
+
"learning_rate": 9.831957763956814e-06,
|
211 |
+
"loss": 0.5655,
|
212 |
+
"step": 29
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.108433734939759,
|
216 |
+
"grad_norm": 0.5069596767425537,
|
217 |
+
"learning_rate": 9.802976776916493e-06,
|
218 |
+
"loss": 0.5669,
|
219 |
+
"step": 30
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.144578313253012,
|
223 |
+
"grad_norm": 0.5148389935493469,
|
224 |
+
"learning_rate": 9.771741253542742e-06,
|
225 |
+
"loss": 0.5765,
|
226 |
+
"step": 31
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 1.180722891566265,
|
230 |
+
"grad_norm": 0.5303863883018494,
|
231 |
+
"learning_rate": 9.738265855914014e-06,
|
232 |
+
"loss": 0.5701,
|
233 |
+
"step": 32
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 1.216867469879518,
|
237 |
+
"grad_norm": 0.5023403763771057,
|
238 |
+
"learning_rate": 9.70256629751462e-06,
|
239 |
+
"loss": 0.5825,
|
240 |
+
"step": 33
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 1.2530120481927711,
|
244 |
+
"grad_norm": 0.37979233264923096,
|
245 |
+
"learning_rate": 9.664659335858755e-06,
|
246 |
+
"loss": 0.5575,
|
247 |
+
"step": 34
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 1.2891566265060241,
|
251 |
+
"grad_norm": 0.46085426211357117,
|
252 |
+
"learning_rate": 9.624562764624445e-06,
|
253 |
+
"loss": 0.5377,
|
254 |
+
"step": 35
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 1.3253012048192772,
|
258 |
+
"grad_norm": 0.5160898566246033,
|
259 |
+
"learning_rate": 9.582295405301131e-06,
|
260 |
+
"loss": 0.5801,
|
261 |
+
"step": 36
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 1.3614457831325302,
|
265 |
+
"grad_norm": 0.4731687903404236,
|
266 |
+
"learning_rate": 9.537877098354787e-06,
|
267 |
+
"loss": 0.5462,
|
268 |
+
"step": 37
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 1.3975903614457832,
|
272 |
+
"grad_norm": 0.37070778012275696,
|
273 |
+
"learning_rate": 9.491328693914723e-06,
|
274 |
+
"loss": 0.5689,
|
275 |
+
"step": 38
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 1.4337349397590362,
|
279 |
+
"grad_norm": 0.3897731900215149,
|
280 |
+
"learning_rate": 9.442672041986456e-06,
|
281 |
+
"loss": 0.5429,
|
282 |
+
"step": 39
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 1.4698795180722892,
|
286 |
+
"grad_norm": 0.3995729684829712,
|
287 |
+
"learning_rate": 9.391929982195233e-06,
|
288 |
+
"loss": 0.5554,
|
289 |
+
"step": 40
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 1.5060240963855422,
|
293 |
+
"grad_norm": 0.4201774299144745,
|
294 |
+
"learning_rate": 9.339126333065008e-06,
|
295 |
+
"loss": 0.5513,
|
296 |
+
"step": 41
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.5421686746987953,
|
300 |
+
"grad_norm": 0.4044627249240875,
|
301 |
+
"learning_rate": 9.284285880837947e-06,
|
302 |
+
"loss": 0.5419,
|
303 |
+
"step": 42
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 1.5783132530120483,
|
307 |
+
"grad_norm": 0.38670021295547485,
|
308 |
+
"learning_rate": 9.22743436783966e-06,
|
309 |
+
"loss": 0.5504,
|
310 |
+
"step": 43
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 1.6144578313253013,
|
314 |
+
"grad_norm": 0.3694572150707245,
|
315 |
+
"learning_rate": 9.168598480395653e-06,
|
316 |
+
"loss": 0.5435,
|
317 |
+
"step": 44
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 1.6506024096385543,
|
321 |
+
"grad_norm": 0.37926754355430603,
|
322 |
+
"learning_rate": 9.107805836304658e-06,
|
323 |
+
"loss": 0.5437,
|
324 |
+
"step": 45
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 1.6867469879518073,
|
328 |
+
"grad_norm": 0.33715763688087463,
|
329 |
+
"learning_rate": 9.045084971874738e-06,
|
330 |
+
"loss": 0.5485,
|
331 |
+
"step": 46
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 1.7228915662650603,
|
335 |
+
"grad_norm": 0.373822420835495,
|
336 |
+
"learning_rate": 8.98046532852822e-06,
|
337 |
+
"loss": 0.5498,
|
338 |
+
"step": 47
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.7590361445783134,
|
342 |
+
"grad_norm": 0.37885019183158875,
|
343 |
+
"learning_rate": 8.91397723898178e-06,
|
344 |
+
"loss": 0.54,
|
345 |
+
"step": 48
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.7951807228915664,
|
349 |
+
"grad_norm": 0.3432829976081848,
|
350 |
+
"learning_rate": 8.845651913008145e-06,
|
351 |
+
"loss": 0.5558,
|
352 |
+
"step": 49
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.8313253012048194,
|
356 |
+
"grad_norm": 0.34714093804359436,
|
357 |
+
"learning_rate": 8.775521422786104e-06,
|
358 |
+
"loss": 0.5356,
|
359 |
+
"step": 50
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.8674698795180724,
|
363 |
+
"grad_norm": 0.34533223509788513,
|
364 |
+
"learning_rate": 8.703618687845697e-06,
|
365 |
+
"loss": 0.532,
|
366 |
+
"step": 51
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.9036144578313254,
|
370 |
+
"grad_norm": 0.33921247720718384,
|
371 |
+
"learning_rate": 8.629977459615655e-06,
|
372 |
+
"loss": 0.5414,
|
373 |
+
"step": 52
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.9397590361445785,
|
377 |
+
"grad_norm": 0.3631800413131714,
|
378 |
+
"learning_rate": 8.554632305580355e-06,
|
379 |
+
"loss": 0.5625,
|
380 |
+
"step": 53
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.9759036144578315,
|
384 |
+
"grad_norm": 0.3610064685344696,
|
385 |
+
"learning_rate": 8.477618593053693e-06,
|
386 |
+
"loss": 0.5483,
|
387 |
+
"step": 54
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 2.036144578313253,
|
391 |
+
"grad_norm": 0.6465901136398315,
|
392 |
+
"learning_rate": 8.39897247257754e-06,
|
393 |
+
"loss": 0.9698,
|
394 |
+
"step": 55
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 2.072289156626506,
|
398 |
+
"grad_norm": 0.4273433983325958,
|
399 |
+
"learning_rate": 8.318730860952523e-06,
|
400 |
+
"loss": 0.4964,
|
401 |
+
"step": 56
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 2.108433734939759,
|
405 |
+
"grad_norm": 0.3399488031864166,
|
406 |
+
"learning_rate": 8.23693142390914e-06,
|
407 |
+
"loss": 0.468,
|
408 |
+
"step": 57
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 2.144578313253012,
|
412 |
+
"grad_norm": 0.36461561918258667,
|
413 |
+
"learning_rate": 8.153612558427311e-06,
|
414 |
+
"loss": 0.4674,
|
415 |
+
"step": 58
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 2.180722891566265,
|
419 |
+
"grad_norm": 0.41741812229156494,
|
420 |
+
"learning_rate": 8.068813374712689e-06,
|
421 |
+
"loss": 0.46,
|
422 |
+
"step": 59
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 2.216867469879518,
|
426 |
+
"grad_norm": 0.539730966091156,
|
427 |
+
"learning_rate": 7.982573677838172e-06,
|
428 |
+
"loss": 0.4591,
|
429 |
+
"step": 60
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 2.253012048192771,
|
433 |
+
"grad_norm": 0.3364892303943634,
|
434 |
+
"learning_rate": 7.894933949059245e-06,
|
435 |
+
"loss": 0.4656,
|
436 |
+
"step": 61
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 2.289156626506024,
|
440 |
+
"grad_norm": 0.3342113196849823,
|
441 |
+
"learning_rate": 7.805935326811913e-06,
|
442 |
+
"loss": 0.4669,
|
443 |
+
"step": 62
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 2.325301204819277,
|
447 |
+
"grad_norm": 0.4477402865886688,
|
448 |
+
"learning_rate": 7.715619587402165e-06,
|
449 |
+
"loss": 0.4774,
|
450 |
+
"step": 63
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 2.36144578313253,
|
454 |
+
"grad_norm": 0.4410865902900696,
|
455 |
+
"learning_rate": 7.624029125396004e-06,
|
456 |
+
"loss": 0.4386,
|
457 |
+
"step": 64
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 2.397590361445783,
|
461 |
+
"grad_norm": 0.3660714328289032,
|
462 |
+
"learning_rate": 7.53120693371927e-06,
|
463 |
+
"loss": 0.4412,
|
464 |
+
"step": 65
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 2.433734939759036,
|
468 |
+
"grad_norm": 0.3396703898906708,
|
469 |
+
"learning_rate": 7.437196583476597e-06,
|
470 |
+
"loss": 0.4442,
|
471 |
+
"step": 66
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 2.4698795180722892,
|
475 |
+
"grad_norm": 0.352255254983902,
|
476 |
+
"learning_rate": 7.342042203498952e-06,
|
477 |
+
"loss": 0.4664,
|
478 |
+
"step": 67
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 2.5060240963855422,
|
482 |
+
"grad_norm": 0.39523279666900635,
|
483 |
+
"learning_rate": 7.245788459629397e-06,
|
484 |
+
"loss": 0.4386,
|
485 |
+
"step": 68
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 2.5421686746987953,
|
489 |
+
"grad_norm": 0.42037248611450195,
|
490 |
+
"learning_rate": 7.148480533756759e-06,
|
491 |
+
"loss": 0.4406,
|
492 |
+
"step": 69
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 2.5783132530120483,
|
496 |
+
"grad_norm": 0.3557772636413574,
|
497 |
+
"learning_rate": 7.050164102607081e-06,
|
498 |
+
"loss": 0.4568,
|
499 |
+
"step": 70
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 2.6144578313253013,
|
503 |
+
"grad_norm": 0.3083244562149048,
|
504 |
+
"learning_rate": 6.950885316302773e-06,
|
505 |
+
"loss": 0.4193,
|
506 |
+
"step": 71
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 2.6506024096385543,
|
510 |
+
"grad_norm": 0.3360963463783264,
|
511 |
+
"learning_rate": 6.850690776699574e-06,
|
512 |
+
"loss": 0.4265,
|
513 |
+
"step": 72
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 2.6867469879518073,
|
517 |
+
"grad_norm": 0.3748219907283783,
|
518 |
+
"learning_rate": 6.749627515511443e-06,
|
519 |
+
"loss": 0.4544,
|
520 |
+
"step": 73
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 2.7228915662650603,
|
524 |
+
"grad_norm": 0.3778473436832428,
|
525 |
+
"learning_rate": 6.647742972233703e-06,
|
526 |
+
"loss": 0.435,
|
527 |
+
"step": 74
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 2.7590361445783134,
|
531 |
+
"grad_norm": 0.3551197946071625,
|
532 |
+
"learning_rate": 6.545084971874738e-06,
|
533 |
+
"loss": 0.4466,
|
534 |
+
"step": 75
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 2.7951807228915664,
|
538 |
+
"grad_norm": 0.35094714164733887,
|
539 |
+
"learning_rate": 6.441701702506755e-06,
|
540 |
+
"loss": 0.4189,
|
541 |
+
"step": 76
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 2.8313253012048194,
|
545 |
+
"grad_norm": 0.32956451177597046,
|
546 |
+
"learning_rate": 6.337641692646106e-06,
|
547 |
+
"loss": 0.419,
|
548 |
+
"step": 77
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 2.8674698795180724,
|
552 |
+
"grad_norm": 0.36045852303504944,
|
553 |
+
"learning_rate": 6.2329537884738115e-06,
|
554 |
+
"loss": 0.4361,
|
555 |
+
"step": 78
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 2.9036144578313254,
|
559 |
+
"grad_norm": 0.3273407816886902,
|
560 |
+
"learning_rate": 6.127687130906972e-06,
|
561 |
+
"loss": 0.4588,
|
562 |
+
"step": 79
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 2.9397590361445785,
|
566 |
+
"grad_norm": 0.3656028211116791,
|
567 |
+
"learning_rate": 6.021891132531825e-06,
|
568 |
+
"loss": 0.4462,
|
569 |
+
"step": 80
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 2.9759036144578315,
|
573 |
+
"grad_norm": 0.35299059748649597,
|
574 |
+
"learning_rate": 5.915615454409281e-06,
|
575 |
+
"loss": 0.4599,
|
576 |
+
"step": 81
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 3.036144578313253,
|
580 |
+
"grad_norm": 0.5190630555152893,
|
581 |
+
"learning_rate": 5.808909982763825e-06,
|
582 |
+
"loss": 0.8681,
|
583 |
+
"step": 82
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 3.072289156626506,
|
587 |
+
"grad_norm": 0.40097540616989136,
|
588 |
+
"learning_rate": 5.701824805566722e-06,
|
589 |
+
"loss": 0.387,
|
590 |
+
"step": 83
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 3.108433734939759,
|
594 |
+
"grad_norm": 0.37181419134140015,
|
595 |
+
"learning_rate": 5.594410189024533e-06,
|
596 |
+
"loss": 0.3973,
|
597 |
+
"step": 84
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 3.144578313253012,
|
601 |
+
"grad_norm": 0.36105361580848694,
|
602 |
+
"learning_rate": 5.4867165539839505e-06,
|
603 |
+
"loss": 0.3706,
|
604 |
+
"step": 85
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 3.180722891566265,
|
608 |
+
"grad_norm": 0.32242706418037415,
|
609 |
+
"learning_rate": 5.378794452264053e-06,
|
610 |
+
"loss": 0.3647,
|
611 |
+
"step": 86
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 3.216867469879518,
|
615 |
+
"grad_norm": 0.37879207730293274,
|
616 |
+
"learning_rate": 5.270694542927089e-06,
|
617 |
+
"loss": 0.3568,
|
618 |
+
"step": 87
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 3.253012048192771,
|
622 |
+
"grad_norm": 0.3197579085826874,
|
623 |
+
"learning_rate": 5.1624675684989035e-06,
|
624 |
+
"loss": 0.3512,
|
625 |
+
"step": 88
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 3.289156626506024,
|
629 |
+
"grad_norm": 0.3562887907028198,
|
630 |
+
"learning_rate": 5.054164331150199e-06,
|
631 |
+
"loss": 0.3561,
|
632 |
+
"step": 89
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 3.325301204819277,
|
636 |
+
"grad_norm": 0.343924343585968,
|
637 |
+
"learning_rate": 4.945835668849801e-06,
|
638 |
+
"loss": 0.358,
|
639 |
+
"step": 90
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 3.36144578313253,
|
643 |
+
"grad_norm": 0.35644856095314026,
|
644 |
+
"learning_rate": 4.837532431501098e-06,
|
645 |
+
"loss": 0.3673,
|
646 |
+
"step": 91
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 3.397590361445783,
|
650 |
+
"grad_norm": 0.4029666781425476,
|
651 |
+
"learning_rate": 4.729305457072913e-06,
|
652 |
+
"loss": 0.3534,
|
653 |
+
"step": 92
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 3.433734939759036,
|
657 |
+
"grad_norm": 0.3742657005786896,
|
658 |
+
"learning_rate": 4.621205547735949e-06,
|
659 |
+
"loss": 0.3623,
|
660 |
+
"step": 93
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 3.4698795180722892,
|
664 |
+
"grad_norm": 0.3467872142791748,
|
665 |
+
"learning_rate": 4.513283446016052e-06,
|
666 |
+
"loss": 0.3408,
|
667 |
+
"step": 94
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 3.5060240963855422,
|
671 |
+
"grad_norm": 0.36237695813179016,
|
672 |
+
"learning_rate": 4.4055898109754684e-06,
|
673 |
+
"loss": 0.3612,
|
674 |
+
"step": 95
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 3.5421686746987953,
|
678 |
+
"grad_norm": 0.3537854254245758,
|
679 |
+
"learning_rate": 4.298175194433279e-06,
|
680 |
+
"loss": 0.3696,
|
681 |
+
"step": 96
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 3.5783132530120483,
|
685 |
+
"grad_norm": 0.35789230465888977,
|
686 |
+
"learning_rate": 4.191090017236177e-06,
|
687 |
+
"loss": 0.3652,
|
688 |
+
"step": 97
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 3.6144578313253013,
|
692 |
+
"grad_norm": 0.34432247281074524,
|
693 |
+
"learning_rate": 4.0843845455907195e-06,
|
694 |
+
"loss": 0.3422,
|
695 |
+
"step": 98
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 3.6506024096385543,
|
699 |
+
"grad_norm": 0.36375707387924194,
|
700 |
+
"learning_rate": 3.9781088674681764e-06,
|
701 |
+
"loss": 0.3717,
|
702 |
+
"step": 99
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 3.6867469879518073,
|
706 |
+
"grad_norm": 0.34923499822616577,
|
707 |
+
"learning_rate": 3.87231286909303e-06,
|
708 |
+
"loss": 0.3436,
|
709 |
+
"step": 100
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 3.7228915662650603,
|
713 |
+
"grad_norm": 0.35821905732154846,
|
714 |
+
"learning_rate": 3.767046211526191e-06,
|
715 |
+
"loss": 0.3584,
|
716 |
+
"step": 101
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 3.7590361445783134,
|
720 |
+
"grad_norm": 0.3310084640979767,
|
721 |
+
"learning_rate": 3.662358307353897e-06,
|
722 |
+
"loss": 0.3751,
|
723 |
+
"step": 102
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 3.7951807228915664,
|
727 |
+
"grad_norm": 0.3326928913593292,
|
728 |
+
"learning_rate": 3.5582982974932467e-06,
|
729 |
+
"loss": 0.3578,
|
730 |
+
"step": 103
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 3.8313253012048194,
|
734 |
+
"grad_norm": 0.3595709502696991,
|
735 |
+
"learning_rate": 3.4549150281252635e-06,
|
736 |
+
"loss": 0.3504,
|
737 |
+
"step": 104
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 3.8674698795180724,
|
741 |
+
"grad_norm": 0.3628636598587036,
|
742 |
+
"learning_rate": 3.3522570277662986e-06,
|
743 |
+
"loss": 0.3371,
|
744 |
+
"step": 105
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 3.9036144578313254,
|
748 |
+
"grad_norm": 0.2951819896697998,
|
749 |
+
"learning_rate": 3.250372484488558e-06,
|
750 |
+
"loss": 0.3565,
|
751 |
+
"step": 106
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 3.9397590361445785,
|
755 |
+
"grad_norm": 0.3485693037509918,
|
756 |
+
"learning_rate": 3.149309223300428e-06,
|
757 |
+
"loss": 0.3649,
|
758 |
+
"step": 107
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 3.9759036144578315,
|
762 |
+
"grad_norm": 0.3397277891635895,
|
763 |
+
"learning_rate": 3.0491146836972273e-06,
|
764 |
+
"loss": 0.3413,
|
765 |
+
"step": 108
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 4.036144578313253,
|
769 |
+
"grad_norm": 0.5771328210830688,
|
770 |
+
"learning_rate": 2.9498358973929197e-06,
|
771 |
+
"loss": 0.6546,
|
772 |
+
"step": 109
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 4.072289156626506,
|
776 |
+
"grad_norm": 0.377215176820755,
|
777 |
+
"learning_rate": 2.8515194662432423e-06,
|
778 |
+
"loss": 0.32,
|
779 |
+
"step": 110
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 4.108433734939759,
|
783 |
+
"grad_norm": 0.39817994832992554,
|
784 |
+
"learning_rate": 2.7542115403706067e-06,
|
785 |
+
"loss": 0.3009,
|
786 |
+
"step": 111
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 4.144578313253012,
|
790 |
+
"grad_norm": 0.3676982820034027,
|
791 |
+
"learning_rate": 2.65795779650105e-06,
|
792 |
+
"loss": 0.2888,
|
793 |
+
"step": 112
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 4.180722891566265,
|
797 |
+
"grad_norm": 0.34246042370796204,
|
798 |
+
"learning_rate": 2.562803416523405e-06,
|
799 |
+
"loss": 0.3289,
|
800 |
+
"step": 113
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 4.216867469879518,
|
804 |
+
"grad_norm": 0.3002203702926636,
|
805 |
+
"learning_rate": 2.46879306628073e-06,
|
806 |
+
"loss": 0.2918,
|
807 |
+
"step": 114
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 4.253012048192771,
|
811 |
+
"grad_norm": 0.3316577672958374,
|
812 |
+
"learning_rate": 2.375970874603998e-06,
|
813 |
+
"loss": 0.2918,
|
814 |
+
"step": 115
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 4.289156626506024,
|
818 |
+
"grad_norm": 0.36060383915901184,
|
819 |
+
"learning_rate": 2.2843804125978356e-06,
|
820 |
+
"loss": 0.3083,
|
821 |
+
"step": 116
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 4.325301204819277,
|
825 |
+
"grad_norm": 0.3468538522720337,
|
826 |
+
"learning_rate": 2.1940646731880887e-06,
|
827 |
+
"loss": 0.2965,
|
828 |
+
"step": 117
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 4.36144578313253,
|
832 |
+
"grad_norm": 0.33555057644844055,
|
833 |
+
"learning_rate": 2.105066050940758e-06,
|
834 |
+
"loss": 0.2753,
|
835 |
+
"step": 118
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 4.397590361445783,
|
839 |
+
"grad_norm": 0.3662334680557251,
|
840 |
+
"learning_rate": 2.0174263221618307e-06,
|
841 |
+
"loss": 0.3069,
|
842 |
+
"step": 119
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 4.433734939759036,
|
846 |
+
"grad_norm": 0.3135122060775757,
|
847 |
+
"learning_rate": 1.931186625287313e-06,
|
848 |
+
"loss": 0.2963,
|
849 |
+
"step": 120
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 4.469879518072289,
|
853 |
+
"grad_norm": 0.3595134913921356,
|
854 |
+
"learning_rate": 1.8463874415726918e-06,
|
855 |
+
"loss": 0.2941,
|
856 |
+
"step": 121
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 4.506024096385542,
|
860 |
+
"grad_norm": 0.3410852551460266,
|
861 |
+
"learning_rate": 1.7630685760908623e-06,
|
862 |
+
"loss": 0.2933,
|
863 |
+
"step": 122
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 4.542168674698795,
|
867 |
+
"grad_norm": 0.3322881758213043,
|
868 |
+
"learning_rate": 1.6812691390474788e-06,
|
869 |
+
"loss": 0.312,
|
870 |
+
"step": 123
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 4.578313253012048,
|
874 |
+
"grad_norm": 0.3299103379249573,
|
875 |
+
"learning_rate": 1.6010275274224607e-06,
|
876 |
+
"loss": 0.2687,
|
877 |
+
"step": 124
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 4.614457831325301,
|
881 |
+
"grad_norm": 0.3352324366569519,
|
882 |
+
"learning_rate": 1.5223814069463077e-06,
|
883 |
+
"loss": 0.2856,
|
884 |
+
"step": 125
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 4.650602409638554,
|
888 |
+
"grad_norm": 0.3284769356250763,
|
889 |
+
"learning_rate": 1.4453676944196477e-06,
|
890 |
+
"loss": 0.2624,
|
891 |
+
"step": 126
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 4.686746987951807,
|
895 |
+
"grad_norm": 0.2953889071941376,
|
896 |
+
"learning_rate": 1.370022540384347e-06,
|
897 |
+
"loss": 0.295,
|
898 |
+
"step": 127
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 4.72289156626506,
|
902 |
+
"grad_norm": 0.30767425894737244,
|
903 |
+
"learning_rate": 1.296381312154305e-06,
|
904 |
+
"loss": 0.2872,
|
905 |
+
"step": 128
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 4.759036144578313,
|
909 |
+
"grad_norm": 0.3192180395126343,
|
910 |
+
"learning_rate": 1.2244785772138972e-06,
|
911 |
+
"loss": 0.289,
|
912 |
+
"step": 129
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 4.795180722891566,
|
916 |
+
"grad_norm": 0.3239184319972992,
|
917 |
+
"learning_rate": 1.1543480869918555e-06,
|
918 |
+
"loss": 0.2843,
|
919 |
+
"step": 130
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 4.831325301204819,
|
923 |
+
"grad_norm": 0.34262704849243164,
|
924 |
+
"learning_rate": 1.0860227610182222e-06,
|
925 |
+
"loss": 0.2841,
|
926 |
+
"step": 131
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 4.867469879518072,
|
930 |
+
"grad_norm": 0.3087993562221527,
|
931 |
+
"learning_rate": 1.0195346714717813e-06,
|
932 |
+
"loss": 0.2949,
|
933 |
+
"step": 132
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 4.903614457831325,
|
937 |
+
"grad_norm": 0.3130486011505127,
|
938 |
+
"learning_rate": 9.549150281252633e-07,
|
939 |
+
"loss": 0.2935,
|
940 |
+
"step": 133
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 4.9397590361445785,
|
944 |
+
"grad_norm": 0.34729984402656555,
|
945 |
+
"learning_rate": 8.921941636953435e-07,
|
946 |
+
"loss": 0.2852,
|
947 |
+
"step": 134
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 4.975903614457831,
|
951 |
+
"grad_norm": 0.2952175736427307,
|
952 |
+
"learning_rate": 8.314015196043501e-07,
|
953 |
+
"loss": 0.3026,
|
954 |
+
"step": 135
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 5.036144578313253,
|
958 |
+
"grad_norm": 0.33559635281562805,
|
959 |
+
"learning_rate": 7.725656321603414e-07,
|
960 |
+
"loss": 0.5577,
|
961 |
+
"step": 136
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 5.072289156626506,
|
965 |
+
"grad_norm": 0.30693137645721436,
|
966 |
+
"learning_rate": 7.157141191620548e-07,
|
967 |
+
"loss": 0.2559,
|
968 |
+
"step": 137
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 5.108433734939759,
|
972 |
+
"grad_norm": 0.3629642724990845,
|
973 |
+
"learning_rate": 6.60873666934993e-07,
|
974 |
+
"loss": 0.2612,
|
975 |
+
"step": 138
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 5.144578313253012,
|
979 |
+
"grad_norm": 0.36669614911079407,
|
980 |
+
"learning_rate": 6.080700178047688e-07,
|
981 |
+
"loss": 0.2997,
|
982 |
+
"step": 139
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 5.180722891566265,
|
986 |
+
"grad_norm": 0.314728707075119,
|
987 |
+
"learning_rate": 5.573279580135438e-07,
|
988 |
+
"loss": 0.2893,
|
989 |
+
"step": 140
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 5.216867469879518,
|
993 |
+
"grad_norm": 0.29658058285713196,
|
994 |
+
"learning_rate": 5.086713060852788e-07,
|
995 |
+
"loss": 0.2676,
|
996 |
+
"step": 141
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 5.253012048192771,
|
1000 |
+
"grad_norm": 0.33411288261413574,
|
1001 |
+
"learning_rate": 4.6212290164521554e-07,
|
1002 |
+
"loss": 0.2433,
|
1003 |
+
"step": 142
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 5.289156626506024,
|
1007 |
+
"grad_norm": 0.3197992742061615,
|
1008 |
+
"learning_rate": 4.1770459469887003e-07,
|
1009 |
+
"loss": 0.2898,
|
1010 |
+
"step": 143
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 5.325301204819277,
|
1014 |
+
"grad_norm": 0.30660751461982727,
|
1015 |
+
"learning_rate": 3.754372353755559e-07,
|
1016 |
+
"loss": 0.2554,
|
1017 |
+
"step": 144
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 5.36144578313253,
|
1021 |
+
"grad_norm": 0.2993546724319458,
|
1022 |
+
"learning_rate": 3.35340664141246e-07,
|
1023 |
+
"loss": 0.2363,
|
1024 |
+
"step": 145
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 5.397590361445783,
|
1028 |
+
"grad_norm": 0.3170936107635498,
|
1029 |
+
"learning_rate": 2.974337024853802e-07,
|
1030 |
+
"loss": 0.2471,
|
1031 |
+
"step": 146
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 5.433734939759036,
|
1035 |
+
"grad_norm": 0.3209405541419983,
|
1036 |
+
"learning_rate": 2.617341440859883e-07,
|
1037 |
+
"loss": 0.2399,
|
1038 |
+
"step": 147
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 5.469879518072289,
|
1042 |
+
"grad_norm": 0.29624640941619873,
|
1043 |
+
"learning_rate": 2.2825874645725942e-07,
|
1044 |
+
"loss": 0.2651,
|
1045 |
+
"step": 148
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 5.506024096385542,
|
1049 |
+
"grad_norm": 0.30279141664505005,
|
1050 |
+
"learning_rate": 1.9702322308350675e-07,
|
1051 |
+
"loss": 0.2525,
|
1052 |
+
"step": 149
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 5.542168674698795,
|
1056 |
+
"grad_norm": 0.3102453351020813,
|
1057 |
+
"learning_rate": 1.6804223604318825e-07,
|
1058 |
+
"loss": 0.2871,
|
1059 |
+
"step": 150
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 5.578313253012048,
|
1063 |
+
"grad_norm": 0.27690696716308594,
|
1064 |
+
"learning_rate": 1.413293891264722e-07,
|
1065 |
+
"loss": 0.242,
|
1066 |
+
"step": 151
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 5.614457831325301,
|
1070 |
+
"grad_norm": 0.2901451885700226,
|
1071 |
+
"learning_rate": 1.1689722144956672e-07,
|
1072 |
+
"loss": 0.2704,
|
1073 |
+
"step": 152
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 5.650602409638554,
|
1077 |
+
"grad_norm": 0.3027832508087158,
|
1078 |
+
"learning_rate": 9.475720156880419e-08,
|
1079 |
+
"loss": 0.2681,
|
1080 |
+
"step": 153
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 5.686746987951807,
|
1084 |
+
"grad_norm": 0.3067132830619812,
|
1085 |
+
"learning_rate": 7.491972209725807e-08,
|
1086 |
+
"loss": 0.2789,
|
1087 |
+
"step": 154
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 5.72289156626506,
|
1091 |
+
"grad_norm": 0.2636028826236725,
|
1092 |
+
"learning_rate": 5.739409482640956e-08,
|
1093 |
+
"loss": 0.239,
|
1094 |
+
"step": 155
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 5.759036144578313,
|
1098 |
+
"grad_norm": 0.2763376533985138,
|
1099 |
+
"learning_rate": 4.2188546355153016e-08,
|
1100 |
+
"loss": 0.2624,
|
1101 |
+
"step": 156
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 5.795180722891566,
|
1105 |
+
"grad_norm": 0.2844434082508087,
|
1106 |
+
"learning_rate": 2.9310214228202016e-08,
|
1107 |
+
"loss": 0.2616,
|
1108 |
+
"step": 157
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 5.831325301204819,
|
1112 |
+
"grad_norm": 0.30939143896102905,
|
1113 |
+
"learning_rate": 1.8765143585693924e-08,
|
1114 |
+
"loss": 0.2651,
|
1115 |
+
"step": 158
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 5.867469879518072,
|
1119 |
+
"grad_norm": 0.2921662926673889,
|
1120 |
+
"learning_rate": 1.0558284325578038e-08,
|
1121 |
+
"loss": 0.2655,
|
1122 |
+
"step": 159
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 5.903614457831325,
|
1126 |
+
"grad_norm": 0.2594563364982605,
|
1127 |
+
"learning_rate": 4.69348878011644e-09,
|
1128 |
+
"loss": 0.2887,
|
1129 |
+
"step": 160
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 5.9397590361445785,
|
1133 |
+
"grad_norm": 0.2833521068096161,
|
1134 |
+
"learning_rate": 1.173509907579362e-09,
|
1135 |
+
"loss": 0.278,
|
1136 |
+
"step": 161
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 5.975903614457831,
|
1140 |
+
"grad_norm": 0.2993119955062866,
|
1141 |
+
"learning_rate": 0.0,
|
1142 |
+
"loss": 0.2823,
|
1143 |
+
"step": 162
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 5.975903614457831,
|
1147 |
+
"step": 162,
|
1148 |
+
"total_flos": 212201928261632.0,
|
1149 |
+
"train_loss": 0.4494376038143664,
|
1150 |
+
"train_runtime": 18091.5652,
|
1151 |
+
"train_samples_per_second": 0.439,
|
1152 |
+
"train_steps_per_second": 0.009
|
1153 |
+
}
|
1154 |
+
],
|
1155 |
+
"logging_steps": 1,
|
1156 |
+
"max_steps": 162,
|
1157 |
+
"num_input_tokens_seen": 0,
|
1158 |
+
"num_train_epochs": 6,
|
1159 |
+
"save_steps": 28,
|
1160 |
+
"stateful_callbacks": {
|
1161 |
+
"TrainerControl": {
|
1162 |
+
"args": {
|
1163 |
+
"should_epoch_stop": false,
|
1164 |
+
"should_evaluate": false,
|
1165 |
+
"should_log": false,
|
1166 |
+
"should_save": true,
|
1167 |
+
"should_training_stop": true
|
1168 |
+
},
|
1169 |
+
"attributes": {}
|
1170 |
+
}
|
1171 |
+
},
|
1172 |
+
"total_flos": 212201928261632.0,
|
1173 |
+
"train_batch_size": 1,
|
1174 |
+
"trial_name": null,
|
1175 |
+
"trial_params": null
|
1176 |
+
}
|