Prikshit7766
commited on
Upload log_history.json with huggingface_hub
Browse files- log_history.json +847 -0
log_history.json
ADDED
@@ -0,0 +1,847 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"loss": 1.3916,
|
4 |
+
"grad_norm": 7.633892059326172,
|
5 |
+
"learning_rate": 1.9831190798376187e-05,
|
6 |
+
"epoch": 0.08457374830852503,
|
7 |
+
"step": 500
|
8 |
+
},
|
9 |
+
{
|
10 |
+
"loss": 1.2408,
|
11 |
+
"grad_norm": 4.5788655281066895,
|
12 |
+
"learning_rate": 1.9662043301759137e-05,
|
13 |
+
"epoch": 0.16914749661705006,
|
14 |
+
"step": 1000
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"loss": 1.2028,
|
18 |
+
"grad_norm": 5.243614673614502,
|
19 |
+
"learning_rate": 1.9492895805142083e-05,
|
20 |
+
"epoch": 0.25372124492557513,
|
21 |
+
"step": 1500
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"loss": 1.1534,
|
25 |
+
"grad_norm": 3.415882110595703,
|
26 |
+
"learning_rate": 1.9323748308525033e-05,
|
27 |
+
"epoch": 0.3382949932341001,
|
28 |
+
"step": 2000
|
29 |
+
},
|
30 |
+
{
|
31 |
+
"loss": 1.1093,
|
32 |
+
"grad_norm": 3.2337677478790283,
|
33 |
+
"learning_rate": 1.9154600811907986e-05,
|
34 |
+
"epoch": 0.42286874154262516,
|
35 |
+
"step": 2500
|
36 |
+
},
|
37 |
+
{
|
38 |
+
"loss": 1.0754,
|
39 |
+
"grad_norm": 3.4433956146240234,
|
40 |
+
"learning_rate": 1.8985453315290936e-05,
|
41 |
+
"epoch": 0.5074424898511503,
|
42 |
+
"step": 3000
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"loss": 1.0467,
|
46 |
+
"grad_norm": 4.641908168792725,
|
47 |
+
"learning_rate": 1.8816305818673886e-05,
|
48 |
+
"epoch": 0.5920162381596752,
|
49 |
+
"step": 3500
|
50 |
+
},
|
51 |
+
{
|
52 |
+
"loss": 1.0355,
|
53 |
+
"grad_norm": 3.7774765491485596,
|
54 |
+
"learning_rate": 1.8647158322056836e-05,
|
55 |
+
"epoch": 0.6765899864682002,
|
56 |
+
"step": 4000
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"loss": 1.0421,
|
60 |
+
"grad_norm": 3.130302906036377,
|
61 |
+
"learning_rate": 1.8478349120433018e-05,
|
62 |
+
"epoch": 0.7611637347767253,
|
63 |
+
"step": 4500
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"loss": 1.0158,
|
67 |
+
"grad_norm": 2.876955509185791,
|
68 |
+
"learning_rate": 1.8309201623815968e-05,
|
69 |
+
"epoch": 0.8457374830852503,
|
70 |
+
"step": 5000
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"loss": 1.0,
|
74 |
+
"grad_norm": 3.218794107437134,
|
75 |
+
"learning_rate": 1.8140054127198918e-05,
|
76 |
+
"epoch": 0.9303112313937754,
|
77 |
+
"step": 5500
|
78 |
+
},
|
79 |
+
{
|
80 |
+
"loss": 0.9665,
|
81 |
+
"grad_norm": 2.9098243713378906,
|
82 |
+
"learning_rate": 1.7971244925575103e-05,
|
83 |
+
"epoch": 1.0148849797023005,
|
84 |
+
"step": 6000
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"loss": 0.8815,
|
88 |
+
"grad_norm": 2.203686475753784,
|
89 |
+
"learning_rate": 1.7802097428958052e-05,
|
90 |
+
"epoch": 1.0994587280108254,
|
91 |
+
"step": 6500
|
92 |
+
},
|
93 |
+
{
|
94 |
+
"loss": 0.9049,
|
95 |
+
"grad_norm": 3.389420986175537,
|
96 |
+
"learning_rate": 1.7632949932341002e-05,
|
97 |
+
"epoch": 1.1840324763193504,
|
98 |
+
"step": 7000
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"loss": 0.9059,
|
102 |
+
"grad_norm": 2.778923749923706,
|
103 |
+
"learning_rate": 1.7463802435723952e-05,
|
104 |
+
"epoch": 1.2686062246278755,
|
105 |
+
"step": 7500
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"loss": 0.8858,
|
109 |
+
"grad_norm": 5.216675758361816,
|
110 |
+
"learning_rate": 1.7294993234100137e-05,
|
111 |
+
"epoch": 1.3531799729364005,
|
112 |
+
"step": 8000
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"loss": 0.8881,
|
116 |
+
"grad_norm": 3.9376771450042725,
|
117 |
+
"learning_rate": 1.7125845737483087e-05,
|
118 |
+
"epoch": 1.4377537212449256,
|
119 |
+
"step": 8500
|
120 |
+
},
|
121 |
+
{
|
122 |
+
"loss": 0.8885,
|
123 |
+
"grad_norm": 3.8835389614105225,
|
124 |
+
"learning_rate": 1.6956698240866037e-05,
|
125 |
+
"epoch": 1.5223274695534506,
|
126 |
+
"step": 9000
|
127 |
+
},
|
128 |
+
{
|
129 |
+
"loss": 0.8774,
|
130 |
+
"grad_norm": 3.470211982727051,
|
131 |
+
"learning_rate": 1.6787550744248987e-05,
|
132 |
+
"epoch": 1.6069012178619757,
|
133 |
+
"step": 9500
|
134 |
+
},
|
135 |
+
{
|
136 |
+
"loss": 0.872,
|
137 |
+
"grad_norm": 3.03437876701355,
|
138 |
+
"learning_rate": 1.6618403247631937e-05,
|
139 |
+
"epoch": 1.6914749661705006,
|
140 |
+
"step": 10000
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"loss": 0.8883,
|
144 |
+
"grad_norm": 3.5217363834381104,
|
145 |
+
"learning_rate": 1.6449255751014887e-05,
|
146 |
+
"epoch": 1.7760487144790256,
|
147 |
+
"step": 10500
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"loss": 0.8614,
|
151 |
+
"grad_norm": 3.965338706970215,
|
152 |
+
"learning_rate": 1.6280446549391072e-05,
|
153 |
+
"epoch": 1.8606224627875507,
|
154 |
+
"step": 11000
|
155 |
+
},
|
156 |
+
{
|
157 |
+
"loss": 0.8682,
|
158 |
+
"grad_norm": 2.316436767578125,
|
159 |
+
"learning_rate": 1.6111299052774022e-05,
|
160 |
+
"epoch": 1.9451962110960759,
|
161 |
+
"step": 11500
|
162 |
+
},
|
163 |
+
{
|
164 |
+
"loss": 0.8286,
|
165 |
+
"grad_norm": 3.9571282863616943,
|
166 |
+
"learning_rate": 1.5942151556156972e-05,
|
167 |
+
"epoch": 2.029769959404601,
|
168 |
+
"step": 12000
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"loss": 0.7808,
|
172 |
+
"grad_norm": 3.676339864730835,
|
173 |
+
"learning_rate": 1.577300405953992e-05,
|
174 |
+
"epoch": 2.1143437077131257,
|
175 |
+
"step": 12500
|
176 |
+
},
|
177 |
+
{
|
178 |
+
"loss": 0.7811,
|
179 |
+
"grad_norm": 2.8704166412353516,
|
180 |
+
"learning_rate": 1.560385656292287e-05,
|
181 |
+
"epoch": 2.198917456021651,
|
182 |
+
"step": 13000
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"loss": 0.8,
|
186 |
+
"grad_norm": 4.323342800140381,
|
187 |
+
"learning_rate": 1.5434709066305818e-05,
|
188 |
+
"epoch": 2.283491204330176,
|
189 |
+
"step": 13500
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"loss": 0.79,
|
193 |
+
"grad_norm": 2.5053164958953857,
|
194 |
+
"learning_rate": 1.5265561569688768e-05,
|
195 |
+
"epoch": 2.3680649526387008,
|
196 |
+
"step": 14000
|
197 |
+
},
|
198 |
+
{
|
199 |
+
"loss": 0.7849,
|
200 |
+
"grad_norm": 3.5477893352508545,
|
201 |
+
"learning_rate": 1.509641407307172e-05,
|
202 |
+
"epoch": 2.452638700947226,
|
203 |
+
"step": 14500
|
204 |
+
},
|
205 |
+
{
|
206 |
+
"loss": 0.7839,
|
207 |
+
"grad_norm": 3.699144124984741,
|
208 |
+
"learning_rate": 1.4927604871447903e-05,
|
209 |
+
"epoch": 2.537212449255751,
|
210 |
+
"step": 15000
|
211 |
+
},
|
212 |
+
{
|
213 |
+
"loss": 0.799,
|
214 |
+
"grad_norm": 2.969682455062866,
|
215 |
+
"learning_rate": 1.4758795669824088e-05,
|
216 |
+
"epoch": 2.621786197564276,
|
217 |
+
"step": 15500
|
218 |
+
},
|
219 |
+
{
|
220 |
+
"loss": 0.7799,
|
221 |
+
"grad_norm": 4.236715316772461,
|
222 |
+
"learning_rate": 1.4589648173207038e-05,
|
223 |
+
"epoch": 2.706359945872801,
|
224 |
+
"step": 16000
|
225 |
+
},
|
226 |
+
{
|
227 |
+
"loss": 0.7756,
|
228 |
+
"grad_norm": 3.2528302669525146,
|
229 |
+
"learning_rate": 1.4420500676589988e-05,
|
230 |
+
"epoch": 2.790933694181326,
|
231 |
+
"step": 16500
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"loss": 0.7837,
|
235 |
+
"grad_norm": 4.050159454345703,
|
236 |
+
"learning_rate": 1.4251353179972938e-05,
|
237 |
+
"epoch": 2.8755074424898512,
|
238 |
+
"step": 17000
|
239 |
+
},
|
240 |
+
{
|
241 |
+
"loss": 0.7907,
|
242 |
+
"grad_norm": 3.53711199760437,
|
243 |
+
"learning_rate": 1.4082543978349121e-05,
|
244 |
+
"epoch": 2.960081190798376,
|
245 |
+
"step": 17500
|
246 |
+
},
|
247 |
+
{
|
248 |
+
"loss": 0.7533,
|
249 |
+
"grad_norm": 4.171680927276611,
|
250 |
+
"learning_rate": 1.3913396481732071e-05,
|
251 |
+
"epoch": 3.044654939106901,
|
252 |
+
"step": 18000
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"loss": 0.724,
|
256 |
+
"grad_norm": 3.026613712310791,
|
257 |
+
"learning_rate": 1.3744248985115021e-05,
|
258 |
+
"epoch": 3.1292286874154263,
|
259 |
+
"step": 18500
|
260 |
+
},
|
261 |
+
{
|
262 |
+
"loss": 0.7289,
|
263 |
+
"grad_norm": 3.7827978134155273,
|
264 |
+
"learning_rate": 1.357510148849797e-05,
|
265 |
+
"epoch": 3.2138024357239514,
|
266 |
+
"step": 19000
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"loss": 0.7078,
|
270 |
+
"grad_norm": 2.7544713020324707,
|
271 |
+
"learning_rate": 1.3405953991880922e-05,
|
272 |
+
"epoch": 3.2983761840324766,
|
273 |
+
"step": 19500
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"loss": 0.7154,
|
277 |
+
"grad_norm": 2.6823747158050537,
|
278 |
+
"learning_rate": 1.3236806495263872e-05,
|
279 |
+
"epoch": 3.3829499323410013,
|
280 |
+
"step": 20000
|
281 |
+
},
|
282 |
+
{
|
283 |
+
"loss": 0.7314,
|
284 |
+
"grad_norm": 3.3478825092315674,
|
285 |
+
"learning_rate": 1.3067658998646822e-05,
|
286 |
+
"epoch": 3.4675236806495264,
|
287 |
+
"step": 20500
|
288 |
+
},
|
289 |
+
{
|
290 |
+
"loss": 0.726,
|
291 |
+
"grad_norm": 3.5908212661743164,
|
292 |
+
"learning_rate": 1.289851150202977e-05,
|
293 |
+
"epoch": 3.5520974289580516,
|
294 |
+
"step": 21000
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"loss": 0.7285,
|
298 |
+
"grad_norm": 3.1159133911132812,
|
299 |
+
"learning_rate": 1.2729702300405956e-05,
|
300 |
+
"epoch": 3.6366711772665763,
|
301 |
+
"step": 21500
|
302 |
+
},
|
303 |
+
{
|
304 |
+
"loss": 0.7162,
|
305 |
+
"grad_norm": 3.3147387504577637,
|
306 |
+
"learning_rate": 1.2560554803788905e-05,
|
307 |
+
"epoch": 3.7212449255751014,
|
308 |
+
"step": 22000
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"loss": 0.7233,
|
312 |
+
"grad_norm": 3.7307050228118896,
|
313 |
+
"learning_rate": 1.2391407307171854e-05,
|
314 |
+
"epoch": 3.8058186738836266,
|
315 |
+
"step": 22500
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"loss": 0.7159,
|
319 |
+
"grad_norm": 2.382382392883301,
|
320 |
+
"learning_rate": 1.2222259810554804e-05,
|
321 |
+
"epoch": 3.8903924221921518,
|
322 |
+
"step": 23000
|
323 |
+
},
|
324 |
+
{
|
325 |
+
"loss": 0.722,
|
326 |
+
"grad_norm": 4.039222717285156,
|
327 |
+
"learning_rate": 1.2053112313937754e-05,
|
328 |
+
"epoch": 3.9749661705006765,
|
329 |
+
"step": 23500
|
330 |
+
},
|
331 |
+
{
|
332 |
+
"loss": 0.6786,
|
333 |
+
"grad_norm": 3.7908201217651367,
|
334 |
+
"learning_rate": 1.188430311231394e-05,
|
335 |
+
"epoch": 4.059539918809202,
|
336 |
+
"step": 24000
|
337 |
+
},
|
338 |
+
{
|
339 |
+
"loss": 0.6735,
|
340 |
+
"grad_norm": 3.7970995903015137,
|
341 |
+
"learning_rate": 1.171515561569689e-05,
|
342 |
+
"epoch": 4.144113667117726,
|
343 |
+
"step": 24500
|
344 |
+
},
|
345 |
+
{
|
346 |
+
"loss": 0.6736,
|
347 |
+
"grad_norm": 4.007111549377441,
|
348 |
+
"learning_rate": 1.1546008119079838e-05,
|
349 |
+
"epoch": 4.2286874154262515,
|
350 |
+
"step": 25000
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"loss": 0.664,
|
354 |
+
"grad_norm": 3.4510090351104736,
|
355 |
+
"learning_rate": 1.1376860622462788e-05,
|
356 |
+
"epoch": 4.313261163734777,
|
357 |
+
"step": 25500
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"loss": 0.6724,
|
361 |
+
"grad_norm": 3.279106378555298,
|
362 |
+
"learning_rate": 1.1207713125845738e-05,
|
363 |
+
"epoch": 4.397834912043302,
|
364 |
+
"step": 26000
|
365 |
+
},
|
366 |
+
{
|
367 |
+
"loss": 0.6696,
|
368 |
+
"grad_norm": 2.7026331424713135,
|
369 |
+
"learning_rate": 1.1038565629228688e-05,
|
370 |
+
"epoch": 4.482408660351827,
|
371 |
+
"step": 26500
|
372 |
+
},
|
373 |
+
{
|
374 |
+
"loss": 0.6738,
|
375 |
+
"grad_norm": 3.247185230255127,
|
376 |
+
"learning_rate": 1.0869418132611638e-05,
|
377 |
+
"epoch": 4.566982408660352,
|
378 |
+
"step": 27000
|
379 |
+
},
|
380 |
+
{
|
381 |
+
"loss": 0.6908,
|
382 |
+
"grad_norm": 3.6047909259796143,
|
383 |
+
"learning_rate": 1.0700608930987821e-05,
|
384 |
+
"epoch": 4.651556156968876,
|
385 |
+
"step": 27500
|
386 |
+
},
|
387 |
+
{
|
388 |
+
"loss": 0.6687,
|
389 |
+
"grad_norm": 4.114670753479004,
|
390 |
+
"learning_rate": 1.0531461434370771e-05,
|
391 |
+
"epoch": 4.7361299052774015,
|
392 |
+
"step": 28000
|
393 |
+
},
|
394 |
+
{
|
395 |
+
"loss": 0.6632,
|
396 |
+
"grad_norm": 2.717122793197632,
|
397 |
+
"learning_rate": 1.0362313937753723e-05,
|
398 |
+
"epoch": 4.820703653585927,
|
399 |
+
"step": 28500
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"loss": 0.6842,
|
403 |
+
"grad_norm": 2.6075901985168457,
|
404 |
+
"learning_rate": 1.0193166441136673e-05,
|
405 |
+
"epoch": 4.905277401894452,
|
406 |
+
"step": 29000
|
407 |
+
},
|
408 |
+
{
|
409 |
+
"loss": 0.6767,
|
410 |
+
"grad_norm": 3.6151123046875,
|
411 |
+
"learning_rate": 1.0024018944519623e-05,
|
412 |
+
"epoch": 4.989851150202977,
|
413 |
+
"step": 29500
|
414 |
+
},
|
415 |
+
{
|
416 |
+
"loss": 0.6441,
|
417 |
+
"grad_norm": 3.7179670333862305,
|
418 |
+
"learning_rate": 9.855209742895806e-06,
|
419 |
+
"epoch": 5.074424898511502,
|
420 |
+
"step": 30000
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"loss": 0.6274,
|
424 |
+
"grad_norm": 3.9154396057128906,
|
425 |
+
"learning_rate": 9.686062246278756e-06,
|
426 |
+
"epoch": 5.158998646820027,
|
427 |
+
"step": 30500
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"loss": 0.6214,
|
431 |
+
"grad_norm": 2.7849080562591553,
|
432 |
+
"learning_rate": 9.516914749661706e-06,
|
433 |
+
"epoch": 5.243572395128552,
|
434 |
+
"step": 31000
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"loss": 0.6362,
|
438 |
+
"grad_norm": 3.1513593196868896,
|
439 |
+
"learning_rate": 9.347767253044656e-06,
|
440 |
+
"epoch": 5.328146143437078,
|
441 |
+
"step": 31500
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"loss": 0.6366,
|
445 |
+
"grad_norm": 3.0636239051818848,
|
446 |
+
"learning_rate": 9.178619756427606e-06,
|
447 |
+
"epoch": 5.412719891745602,
|
448 |
+
"step": 32000
|
449 |
+
},
|
450 |
+
{
|
451 |
+
"loss": 0.635,
|
452 |
+
"grad_norm": 3.542881727218628,
|
453 |
+
"learning_rate": 9.009810554803789e-06,
|
454 |
+
"epoch": 5.497293640054127,
|
455 |
+
"step": 32500
|
456 |
+
},
|
457 |
+
{
|
458 |
+
"loss": 0.629,
|
459 |
+
"grad_norm": 2.9938108921051025,
|
460 |
+
"learning_rate": 8.840663058186739e-06,
|
461 |
+
"epoch": 5.581867388362652,
|
462 |
+
"step": 33000
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"loss": 0.6424,
|
466 |
+
"grad_norm": 3.608818769454956,
|
467 |
+
"learning_rate": 8.67151556156969e-06,
|
468 |
+
"epoch": 5.666441136671177,
|
469 |
+
"step": 33500
|
470 |
+
},
|
471 |
+
{
|
472 |
+
"loss": 0.6354,
|
473 |
+
"grad_norm": 4.858671188354492,
|
474 |
+
"learning_rate": 8.50236806495264e-06,
|
475 |
+
"epoch": 5.7510148849797025,
|
476 |
+
"step": 34000
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"loss": 0.6365,
|
480 |
+
"grad_norm": 3.254009246826172,
|
481 |
+
"learning_rate": 8.333220568335589e-06,
|
482 |
+
"epoch": 5.835588633288228,
|
483 |
+
"step": 34500
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"loss": 0.6343,
|
487 |
+
"grad_norm": 2.389611005783081,
|
488 |
+
"learning_rate": 8.164073071718539e-06,
|
489 |
+
"epoch": 5.920162381596752,
|
490 |
+
"step": 35000
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"loss": 0.6305,
|
494 |
+
"grad_norm": 3.2198381423950195,
|
495 |
+
"learning_rate": 7.994925575101489e-06,
|
496 |
+
"epoch": 6.004736129905277,
|
497 |
+
"step": 35500
|
498 |
+
},
|
499 |
+
{
|
500 |
+
"loss": 0.597,
|
501 |
+
"grad_norm": 2.834723711013794,
|
502 |
+
"learning_rate": 7.82577807848444e-06,
|
503 |
+
"epoch": 6.089309878213802,
|
504 |
+
"step": 36000
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"loss": 0.5914,
|
508 |
+
"grad_norm": 2.7054672241210938,
|
509 |
+
"learning_rate": 7.656630581867388e-06,
|
510 |
+
"epoch": 6.173883626522327,
|
511 |
+
"step": 36500
|
512 |
+
},
|
513 |
+
{
|
514 |
+
"loss": 0.6084,
|
515 |
+
"grad_norm": 2.8164889812469482,
|
516 |
+
"learning_rate": 7.487483085250339e-06,
|
517 |
+
"epoch": 6.2584573748308525,
|
518 |
+
"step": 37000
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"loss": 0.5947,
|
522 |
+
"grad_norm": 3.42501163482666,
|
523 |
+
"learning_rate": 7.3186738836265225e-06,
|
524 |
+
"epoch": 6.343031123139378,
|
525 |
+
"step": 37500
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"loss": 0.604,
|
529 |
+
"grad_norm": 3.881469249725342,
|
530 |
+
"learning_rate": 7.149526387009473e-06,
|
531 |
+
"epoch": 6.427604871447903,
|
532 |
+
"step": 38000
|
533 |
+
},
|
534 |
+
{
|
535 |
+
"loss": 0.5896,
|
536 |
+
"grad_norm": 3.2387328147888184,
|
537 |
+
"learning_rate": 6.980378890392423e-06,
|
538 |
+
"epoch": 6.512178619756428,
|
539 |
+
"step": 38500
|
540 |
+
},
|
541 |
+
{
|
542 |
+
"loss": 0.5985,
|
543 |
+
"grad_norm": 3.245598316192627,
|
544 |
+
"learning_rate": 6.811231393775373e-06,
|
545 |
+
"epoch": 6.596752368064953,
|
546 |
+
"step": 39000
|
547 |
+
},
|
548 |
+
{
|
549 |
+
"loss": 0.6207,
|
550 |
+
"grad_norm": 4.7686448097229,
|
551 |
+
"learning_rate": 6.642422192151556e-06,
|
552 |
+
"epoch": 6.681326116373477,
|
553 |
+
"step": 39500
|
554 |
+
},
|
555 |
+
{
|
556 |
+
"loss": 0.6164,
|
557 |
+
"grad_norm": 3.3545920848846436,
|
558 |
+
"learning_rate": 6.473274695534507e-06,
|
559 |
+
"epoch": 6.7658998646820026,
|
560 |
+
"step": 40000
|
561 |
+
},
|
562 |
+
{
|
563 |
+
"loss": 0.5994,
|
564 |
+
"grad_norm": 3.037534713745117,
|
565 |
+
"learning_rate": 6.304127198917457e-06,
|
566 |
+
"epoch": 6.850473612990528,
|
567 |
+
"step": 40500
|
568 |
+
},
|
569 |
+
{
|
570 |
+
"loss": 0.6249,
|
571 |
+
"grad_norm": 4.4626784324646,
|
572 |
+
"learning_rate": 6.134979702300406e-06,
|
573 |
+
"epoch": 6.935047361299053,
|
574 |
+
"step": 41000
|
575 |
+
},
|
576 |
+
{
|
577 |
+
"loss": 0.5985,
|
578 |
+
"grad_norm": 2.8611857891082764,
|
579 |
+
"learning_rate": 5.96617050067659e-06,
|
580 |
+
"epoch": 7.019621109607578,
|
581 |
+
"step": 41500
|
582 |
+
},
|
583 |
+
{
|
584 |
+
"loss": 0.5841,
|
585 |
+
"grad_norm": 3.028613805770874,
|
586 |
+
"learning_rate": 5.79702300405954e-06,
|
587 |
+
"epoch": 7.104194857916103,
|
588 |
+
"step": 42000
|
589 |
+
},
|
590 |
+
{
|
591 |
+
"loss": 0.586,
|
592 |
+
"grad_norm": 2.902698040008545,
|
593 |
+
"learning_rate": 5.627875507442491e-06,
|
594 |
+
"epoch": 7.188768606224627,
|
595 |
+
"step": 42500
|
596 |
+
},
|
597 |
+
{
|
598 |
+
"loss": 0.5869,
|
599 |
+
"grad_norm": 2.2707433700561523,
|
600 |
+
"learning_rate": 5.45872801082544e-06,
|
601 |
+
"epoch": 7.273342354533153,
|
602 |
+
"step": 43000
|
603 |
+
},
|
604 |
+
{
|
605 |
+
"loss": 0.5888,
|
606 |
+
"grad_norm": 3.4720981121063232,
|
607 |
+
"learning_rate": 5.289918809201624e-06,
|
608 |
+
"epoch": 7.357916102841678,
|
609 |
+
"step": 43500
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"loss": 0.576,
|
613 |
+
"grad_norm": 2.8364577293395996,
|
614 |
+
"learning_rate": 5.120771312584573e-06,
|
615 |
+
"epoch": 7.442489851150203,
|
616 |
+
"step": 44000
|
617 |
+
},
|
618 |
+
{
|
619 |
+
"loss": 0.5759,
|
620 |
+
"grad_norm": 3.178103446960449,
|
621 |
+
"learning_rate": 4.951623815967524e-06,
|
622 |
+
"epoch": 7.527063599458728,
|
623 |
+
"step": 44500
|
624 |
+
},
|
625 |
+
{
|
626 |
+
"loss": 0.5742,
|
627 |
+
"grad_norm": 3.5063467025756836,
|
628 |
+
"learning_rate": 4.782476319350474e-06,
|
629 |
+
"epoch": 7.611637347767253,
|
630 |
+
"step": 45000
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"loss": 0.5912,
|
634 |
+
"grad_norm": 2.37205171585083,
|
635 |
+
"learning_rate": 4.613328822733424e-06,
|
636 |
+
"epoch": 7.696211096075778,
|
637 |
+
"step": 45500
|
638 |
+
},
|
639 |
+
{
|
640 |
+
"loss": 0.575,
|
641 |
+
"grad_norm": 3.2511661052703857,
|
642 |
+
"learning_rate": 4.444181326116374e-06,
|
643 |
+
"epoch": 7.7807848443843035,
|
644 |
+
"step": 46000
|
645 |
+
},
|
646 |
+
{
|
647 |
+
"loss": 0.5724,
|
648 |
+
"grad_norm": 3.2974693775177,
|
649 |
+
"learning_rate": 4.275033829499324e-06,
|
650 |
+
"epoch": 7.865358592692828,
|
651 |
+
"step": 46500
|
652 |
+
},
|
653 |
+
{
|
654 |
+
"loss": 0.5745,
|
655 |
+
"grad_norm": 3.180819511413574,
|
656 |
+
"learning_rate": 4.105886332882274e-06,
|
657 |
+
"epoch": 7.949932341001353,
|
658 |
+
"step": 47000
|
659 |
+
},
|
660 |
+
{
|
661 |
+
"loss": 0.5697,
|
662 |
+
"grad_norm": 2.4791033267974854,
|
663 |
+
"learning_rate": 3.937077131258458e-06,
|
664 |
+
"epoch": 8.034506089309879,
|
665 |
+
"step": 47500
|
666 |
+
},
|
667 |
+
{
|
668 |
+
"loss": 0.556,
|
669 |
+
"grad_norm": 3.5898892879486084,
|
670 |
+
"learning_rate": 3.7679296346414073e-06,
|
671 |
+
"epoch": 8.119079837618404,
|
672 |
+
"step": 48000
|
673 |
+
},
|
674 |
+
{
|
675 |
+
"loss": 0.5726,
|
676 |
+
"grad_norm": 2.7320892810821533,
|
677 |
+
"learning_rate": 3.5987821380243577e-06,
|
678 |
+
"epoch": 8.203653585926928,
|
679 |
+
"step": 48500
|
680 |
+
},
|
681 |
+
{
|
682 |
+
"loss": 0.5603,
|
683 |
+
"grad_norm": 3.3177103996276855,
|
684 |
+
"learning_rate": 3.429634641407307e-06,
|
685 |
+
"epoch": 8.288227334235453,
|
686 |
+
"step": 49000
|
687 |
+
},
|
688 |
+
{
|
689 |
+
"loss": 0.558,
|
690 |
+
"grad_norm": 2.1732170581817627,
|
691 |
+
"learning_rate": 3.2604871447902575e-06,
|
692 |
+
"epoch": 8.372801082543978,
|
693 |
+
"step": 49500
|
694 |
+
},
|
695 |
+
{
|
696 |
+
"loss": 0.5631,
|
697 |
+
"grad_norm": 3.1351735591888428,
|
698 |
+
"learning_rate": 3.091339648173207e-06,
|
699 |
+
"epoch": 8.457374830852503,
|
700 |
+
"step": 50000
|
701 |
+
},
|
702 |
+
{
|
703 |
+
"loss": 0.5486,
|
704 |
+
"grad_norm": 2.601547956466675,
|
705 |
+
"learning_rate": 2.9221921515561573e-06,
|
706 |
+
"epoch": 8.541948579161028,
|
707 |
+
"step": 50500
|
708 |
+
},
|
709 |
+
{
|
710 |
+
"loss": 0.5706,
|
711 |
+
"grad_norm": 4.160942554473877,
|
712 |
+
"learning_rate": 2.753382949932341e-06,
|
713 |
+
"epoch": 8.626522327469553,
|
714 |
+
"step": 51000
|
715 |
+
},
|
716 |
+
{
|
717 |
+
"loss": 0.5683,
|
718 |
+
"grad_norm": 3.348295211791992,
|
719 |
+
"learning_rate": 2.5842354533152914e-06,
|
720 |
+
"epoch": 8.711096075778078,
|
721 |
+
"step": 51500
|
722 |
+
},
|
723 |
+
{
|
724 |
+
"loss": 0.564,
|
725 |
+
"grad_norm": 3.148343563079834,
|
726 |
+
"learning_rate": 2.415087956698241e-06,
|
727 |
+
"epoch": 8.795669824086604,
|
728 |
+
"step": 52000
|
729 |
+
},
|
730 |
+
{
|
731 |
+
"loss": 0.5725,
|
732 |
+
"grad_norm": 3.285578489303589,
|
733 |
+
"learning_rate": 2.246278755074425e-06,
|
734 |
+
"epoch": 8.880243572395129,
|
735 |
+
"step": 52500
|
736 |
+
},
|
737 |
+
{
|
738 |
+
"loss": 0.567,
|
739 |
+
"grad_norm": 3.201730251312256,
|
740 |
+
"learning_rate": 2.077131258457375e-06,
|
741 |
+
"epoch": 8.964817320703654,
|
742 |
+
"step": 53000
|
743 |
+
},
|
744 |
+
{
|
745 |
+
"loss": 0.553,
|
746 |
+
"grad_norm": 3.166001796722412,
|
747 |
+
"learning_rate": 1.907983761840325e-06,
|
748 |
+
"epoch": 9.049391069012179,
|
749 |
+
"step": 53500
|
750 |
+
},
|
751 |
+
{
|
752 |
+
"loss": 0.5614,
|
753 |
+
"grad_norm": 3.105032444000244,
|
754 |
+
"learning_rate": 1.7388362652232748e-06,
|
755 |
+
"epoch": 9.133964817320704,
|
756 |
+
"step": 54000
|
757 |
+
},
|
758 |
+
{
|
759 |
+
"loss": 0.5553,
|
760 |
+
"grad_norm": 4.7028937339782715,
|
761 |
+
"learning_rate": 1.5696887686062248e-06,
|
762 |
+
"epoch": 9.21853856562923,
|
763 |
+
"step": 54500
|
764 |
+
},
|
765 |
+
{
|
766 |
+
"loss": 0.552,
|
767 |
+
"grad_norm": 3.5488646030426025,
|
768 |
+
"learning_rate": 1.4005412719891747e-06,
|
769 |
+
"epoch": 9.303112313937755,
|
770 |
+
"step": 55000
|
771 |
+
},
|
772 |
+
{
|
773 |
+
"loss": 0.5483,
|
774 |
+
"grad_norm": 2.8257858753204346,
|
775 |
+
"learning_rate": 1.2313937753721246e-06,
|
776 |
+
"epoch": 9.387686062246278,
|
777 |
+
"step": 55500
|
778 |
+
},
|
779 |
+
{
|
780 |
+
"loss": 0.5476,
|
781 |
+
"grad_norm": 4.221645355224609,
|
782 |
+
"learning_rate": 1.0622462787550745e-06,
|
783 |
+
"epoch": 9.472259810554803,
|
784 |
+
"step": 56000
|
785 |
+
},
|
786 |
+
{
|
787 |
+
"loss": 0.5494,
|
788 |
+
"grad_norm": 3.1773674488067627,
|
789 |
+
"learning_rate": 8.930987821380243e-07,
|
790 |
+
"epoch": 9.556833558863328,
|
791 |
+
"step": 56500
|
792 |
+
},
|
793 |
+
{
|
794 |
+
"loss": 0.5477,
|
795 |
+
"grad_norm": 2.9824230670928955,
|
796 |
+
"learning_rate": 7.242895805142085e-07,
|
797 |
+
"epoch": 9.641407307171853,
|
798 |
+
"step": 57000
|
799 |
+
},
|
800 |
+
{
|
801 |
+
"loss": 0.5579,
|
802 |
+
"grad_norm": 2.51481294631958,
|
803 |
+
"learning_rate": 5.551420838971583e-07,
|
804 |
+
"epoch": 9.725981055480379,
|
805 |
+
"step": 57500
|
806 |
+
},
|
807 |
+
{
|
808 |
+
"loss": 0.5392,
|
809 |
+
"grad_norm": 3.0290215015411377,
|
810 |
+
"learning_rate": 3.8599458728010834e-07,
|
811 |
+
"epoch": 9.810554803788904,
|
812 |
+
"step": 58000
|
813 |
+
},
|
814 |
+
{
|
815 |
+
"loss": 0.5458,
|
816 |
+
"grad_norm": 2.9967031478881836,
|
817 |
+
"learning_rate": 2.168470906630582e-07,
|
818 |
+
"epoch": 9.895128552097429,
|
819 |
+
"step": 58500
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"loss": 0.5434,
|
823 |
+
"grad_norm": 2.6248016357421875,
|
824 |
+
"learning_rate": 4.769959404600812e-08,
|
825 |
+
"epoch": 9.979702300405954,
|
826 |
+
"step": 59000
|
827 |
+
},
|
828 |
+
{
|
829 |
+
"train_runtime": 8128.9827,
|
830 |
+
"train_samples_per_second": 232.692,
|
831 |
+
"train_steps_per_second": 7.273,
|
832 |
+
"total_flos": 4.123593985189478e+16,
|
833 |
+
"train_loss": 0.7104815463735867,
|
834 |
+
"epoch": 10.0,
|
835 |
+
"step": 59120
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"eval_loss": 0.7965446710586548,
|
839 |
+
"eval_model_preparation_time": 0.002,
|
840 |
+
"eval_bleu": 54.95663610516514,
|
841 |
+
"eval_runtime": 6012.2223,
|
842 |
+
"eval_samples_per_second": 3.496,
|
843 |
+
"eval_steps_per_second": 0.055,
|
844 |
+
"epoch": 10.0,
|
845 |
+
"step": 59120
|
846 |
+
}
|
847 |
+
]
|