All Dunn!!!
Browse files- all_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +423 -45
all_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 2.0,
|
3 |
-
"total_flos":
|
4 |
-
"train_loss":
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 2.0,
|
3 |
+
"total_flos": 1.9906356553640313e+19,
|
4 |
+
"train_loss": 0.6271170827069549,
|
5 |
+
"train_runtime": 2733.4434,
|
6 |
+
"train_samples_per_second": 4.92,
|
7 |
+
"train_steps_per_second": 0.615
|
8 |
}
|
train_results.json
CHANGED
@@ -1,8 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 2.0,
|
3 |
-
"total_flos":
|
4 |
-
"train_loss":
|
5 |
-
"train_runtime":
|
6 |
-
"train_samples_per_second":
|
7 |
-
"train_steps_per_second": 0.
|
8 |
}
|
|
|
1 |
{
|
2 |
"epoch": 2.0,
|
3 |
+
"total_flos": 1.9906356553640313e+19,
|
4 |
+
"train_loss": 0.6271170827069549,
|
5 |
+
"train_runtime": 2733.4434,
|
6 |
+
"train_samples_per_second": 4.92,
|
7 |
+
"train_steps_per_second": 0.615
|
8 |
}
|
trainer_state.json
CHANGED
@@ -3,76 +3,454 @@
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 2.0,
|
5 |
"eval_steps": 500,
|
6 |
-
"global_step":
|
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.
|
13 |
-
"grad_norm":
|
14 |
-
"learning_rate": 4.
|
15 |
-
"loss":
|
16 |
"step": 1
|
17 |
},
|
18 |
{
|
19 |
-
"epoch": 0.
|
20 |
-
"grad_norm":
|
21 |
-
"learning_rate":
|
22 |
-
"loss": 2.
|
23 |
"step": 29
|
24 |
},
|
25 |
{
|
26 |
-
"epoch": 0.
|
27 |
-
"grad_norm":
|
28 |
-
"learning_rate":
|
29 |
-
"loss":
|
30 |
"step": 58
|
31 |
},
|
32 |
{
|
33 |
-
"epoch":
|
34 |
-
"
|
35 |
-
"
|
36 |
-
"
|
37 |
-
"eval_samples_per_second": 0.485,
|
38 |
-
"eval_steps_per_second": 0.064,
|
39 |
-
"step": 61
|
40 |
-
},
|
41 |
-
{
|
42 |
-
"epoch": 1.4262295081967213,
|
43 |
-
"grad_norm": 30.61115837097168,
|
44 |
-
"learning_rate": 1.4344262295081968e-05,
|
45 |
-
"loss": 1.1471,
|
46 |
"step": 87
|
47 |
},
|
48 |
{
|
49 |
-
"epoch":
|
50 |
-
"grad_norm":
|
51 |
-
"learning_rate":
|
52 |
-
"loss":
|
53 |
"step": 116
|
54 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
55 |
{
|
56 |
"epoch": 2.0,
|
57 |
-
"eval_cer": 0.
|
58 |
-
"eval_loss": 0.
|
59 |
-
"eval_runtime":
|
60 |
-
"eval_samples_per_second":
|
61 |
-
"eval_steps_per_second": 0.
|
62 |
-
"step":
|
63 |
},
|
64 |
{
|
65 |
"epoch": 2.0,
|
66 |
-
"step":
|
67 |
-
"total_flos":
|
68 |
-
"train_loss":
|
69 |
-
"train_runtime":
|
70 |
-
"train_samples_per_second":
|
71 |
-
"train_steps_per_second": 0.
|
72 |
}
|
73 |
],
|
74 |
"logging_steps": 29,
|
75 |
-
"max_steps":
|
76 |
"num_input_tokens_seen": 0,
|
77 |
"num_train_epochs": 2,
|
78 |
"save_steps": 500,
|
@@ -88,8 +466,8 @@
|
|
88 |
"attributes": {}
|
89 |
}
|
90 |
},
|
91 |
-
"total_flos":
|
92 |
-
"train_batch_size":
|
93 |
"trial_name": null,
|
94 |
"trial_params": null
|
95 |
}
|
|
|
3 |
"best_model_checkpoint": null,
|
4 |
"epoch": 2.0,
|
5 |
"eval_steps": 500,
|
6 |
+
"global_step": 1682,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.0011890606420927466,
|
13 |
+
"grad_norm": 128.86566162109375,
|
14 |
+
"learning_rate": 4.9970273483947685e-05,
|
15 |
+
"loss": 9.559,
|
16 |
"step": 1
|
17 |
},
|
18 |
{
|
19 |
+
"epoch": 0.034482758620689655,
|
20 |
+
"grad_norm": 23.844635009765625,
|
21 |
+
"learning_rate": 4.913793103448276e-05,
|
22 |
+
"loss": 2.0043,
|
23 |
"step": 29
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.06896551724137931,
|
27 |
+
"grad_norm": 43.76908493041992,
|
28 |
+
"learning_rate": 4.827586206896552e-05,
|
29 |
+
"loss": 1.6394,
|
30 |
"step": 58
|
31 |
},
|
32 |
{
|
33 |
+
"epoch": 0.10344827586206896,
|
34 |
+
"grad_norm": 112.72496032714844,
|
35 |
+
"learning_rate": 4.741379310344828e-05,
|
36 |
+
"loss": 1.1869,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
"step": 87
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.13793103448275862,
|
41 |
+
"grad_norm": 57.330135345458984,
|
42 |
+
"learning_rate": 4.655172413793104e-05,
|
43 |
+
"loss": 1.4015,
|
44 |
"step": 116
|
45 |
},
|
46 |
+
{
|
47 |
+
"epoch": 0.1724137931034483,
|
48 |
+
"grad_norm": 12.192914962768555,
|
49 |
+
"learning_rate": 4.5689655172413794e-05,
|
50 |
+
"loss": 1.1619,
|
51 |
+
"step": 145
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.20689655172413793,
|
55 |
+
"grad_norm": 49.972900390625,
|
56 |
+
"learning_rate": 4.482758620689655e-05,
|
57 |
+
"loss": 1.4928,
|
58 |
+
"step": 174
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.2413793103448276,
|
62 |
+
"grad_norm": 28.061901092529297,
|
63 |
+
"learning_rate": 4.396551724137931e-05,
|
64 |
+
"loss": 1.1286,
|
65 |
+
"step": 203
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.27586206896551724,
|
69 |
+
"grad_norm": 13.678406715393066,
|
70 |
+
"learning_rate": 4.3103448275862066e-05,
|
71 |
+
"loss": 0.9936,
|
72 |
+
"step": 232
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.3103448275862069,
|
76 |
+
"grad_norm": 56.66800308227539,
|
77 |
+
"learning_rate": 4.224137931034483e-05,
|
78 |
+
"loss": 1.1352,
|
79 |
+
"step": 261
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.3448275862068966,
|
83 |
+
"grad_norm": 18.401317596435547,
|
84 |
+
"learning_rate": 4.1379310344827587e-05,
|
85 |
+
"loss": 1.0754,
|
86 |
+
"step": 290
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.3793103448275862,
|
90 |
+
"grad_norm": 28.412200927734375,
|
91 |
+
"learning_rate": 4.0517241379310344e-05,
|
92 |
+
"loss": 1.0104,
|
93 |
+
"step": 319
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.41379310344827586,
|
97 |
+
"grad_norm": 62.137596130371094,
|
98 |
+
"learning_rate": 3.965517241379311e-05,
|
99 |
+
"loss": 0.9393,
|
100 |
+
"step": 348
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.4482758620689655,
|
104 |
+
"grad_norm": 44.91804504394531,
|
105 |
+
"learning_rate": 3.8793103448275865e-05,
|
106 |
+
"loss": 0.727,
|
107 |
+
"step": 377
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.4827586206896552,
|
111 |
+
"grad_norm": 15.308109283447266,
|
112 |
+
"learning_rate": 3.793103448275862e-05,
|
113 |
+
"loss": 0.8675,
|
114 |
+
"step": 406
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.5172413793103449,
|
118 |
+
"grad_norm": 11.947402000427246,
|
119 |
+
"learning_rate": 3.7068965517241385e-05,
|
120 |
+
"loss": 0.7525,
|
121 |
+
"step": 435
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.5517241379310345,
|
125 |
+
"grad_norm": 22.51788902282715,
|
126 |
+
"learning_rate": 3.620689655172414e-05,
|
127 |
+
"loss": 0.7872,
|
128 |
+
"step": 464
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.5862068965517241,
|
132 |
+
"grad_norm": 39.137386322021484,
|
133 |
+
"learning_rate": 3.53448275862069e-05,
|
134 |
+
"loss": 0.7889,
|
135 |
+
"step": 493
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.6206896551724138,
|
139 |
+
"grad_norm": 38.08049774169922,
|
140 |
+
"learning_rate": 3.4482758620689657e-05,
|
141 |
+
"loss": 0.7347,
|
142 |
+
"step": 522
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.6551724137931034,
|
146 |
+
"grad_norm": 10.072871208190918,
|
147 |
+
"learning_rate": 3.3620689655172414e-05,
|
148 |
+
"loss": 0.7422,
|
149 |
+
"step": 551
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.6896551724137931,
|
153 |
+
"grad_norm": 24.6478328704834,
|
154 |
+
"learning_rate": 3.275862068965517e-05,
|
155 |
+
"loss": 0.7217,
|
156 |
+
"step": 580
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.7241379310344828,
|
160 |
+
"grad_norm": 8.815550804138184,
|
161 |
+
"learning_rate": 3.1896551724137935e-05,
|
162 |
+
"loss": 0.767,
|
163 |
+
"step": 609
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.7586206896551724,
|
167 |
+
"grad_norm": 7.418780326843262,
|
168 |
+
"learning_rate": 3.103448275862069e-05,
|
169 |
+
"loss": 0.7365,
|
170 |
+
"step": 638
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.7931034482758621,
|
174 |
+
"grad_norm": 16.163270950317383,
|
175 |
+
"learning_rate": 3.017241379310345e-05,
|
176 |
+
"loss": 0.6203,
|
177 |
+
"step": 667
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.8275862068965517,
|
181 |
+
"grad_norm": 47.155818939208984,
|
182 |
+
"learning_rate": 2.9310344827586206e-05,
|
183 |
+
"loss": 0.7505,
|
184 |
+
"step": 696
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.8620689655172413,
|
188 |
+
"grad_norm": 17.693836212158203,
|
189 |
+
"learning_rate": 2.844827586206897e-05,
|
190 |
+
"loss": 0.6014,
|
191 |
+
"step": 725
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.896551724137931,
|
195 |
+
"grad_norm": 15.081289291381836,
|
196 |
+
"learning_rate": 2.7586206896551727e-05,
|
197 |
+
"loss": 0.5907,
|
198 |
+
"step": 754
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.9310344827586207,
|
202 |
+
"grad_norm": 235.15663146972656,
|
203 |
+
"learning_rate": 2.672413793103448e-05,
|
204 |
+
"loss": 0.5196,
|
205 |
+
"step": 783
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.9655172413793104,
|
209 |
+
"grad_norm": 13.673110961914062,
|
210 |
+
"learning_rate": 2.5862068965517244e-05,
|
211 |
+
"loss": 0.5441,
|
212 |
+
"step": 812
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.0,
|
216 |
+
"grad_norm": 22.076805114746094,
|
217 |
+
"learning_rate": 2.5e-05,
|
218 |
+
"loss": 0.5455,
|
219 |
+
"step": 841
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.0,
|
223 |
+
"eval_cer": 0.020121099208197483,
|
224 |
+
"eval_loss": 0.46177592873573303,
|
225 |
+
"eval_runtime": 644.2587,
|
226 |
+
"eval_samples_per_second": 2.611,
|
227 |
+
"eval_steps_per_second": 0.328,
|
228 |
+
"step": 841
|
229 |
+
},
|
230 |
+
{
|
231 |
+
"epoch": 1.0344827586206897,
|
232 |
+
"grad_norm": 6.016767501831055,
|
233 |
+
"learning_rate": 2.413793103448276e-05,
|
234 |
+
"loss": 0.416,
|
235 |
+
"step": 870
|
236 |
+
},
|
237 |
+
{
|
238 |
+
"epoch": 1.0689655172413792,
|
239 |
+
"grad_norm": 5.592987060546875,
|
240 |
+
"learning_rate": 2.327586206896552e-05,
|
241 |
+
"loss": 0.4161,
|
242 |
+
"step": 899
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 1.103448275862069,
|
246 |
+
"grad_norm": 4.175529479980469,
|
247 |
+
"learning_rate": 2.2413793103448276e-05,
|
248 |
+
"loss": 0.4516,
|
249 |
+
"step": 928
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 1.1379310344827587,
|
253 |
+
"grad_norm": 7.126400470733643,
|
254 |
+
"learning_rate": 2.1551724137931033e-05,
|
255 |
+
"loss": 0.4583,
|
256 |
+
"step": 957
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"epoch": 1.1724137931034484,
|
260 |
+
"grad_norm": 5.696765899658203,
|
261 |
+
"learning_rate": 2.0689655172413793e-05,
|
262 |
+
"loss": 0.3918,
|
263 |
+
"step": 986
|
264 |
+
},
|
265 |
+
{
|
266 |
+
"epoch": 1.206896551724138,
|
267 |
+
"grad_norm": 11.261072158813477,
|
268 |
+
"learning_rate": 1.9827586206896554e-05,
|
269 |
+
"loss": 0.4423,
|
270 |
+
"step": 1015
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 1.2413793103448276,
|
274 |
+
"grad_norm": 3.23542857170105,
|
275 |
+
"learning_rate": 1.896551724137931e-05,
|
276 |
+
"loss": 0.3769,
|
277 |
+
"step": 1044
|
278 |
+
},
|
279 |
+
{
|
280 |
+
"epoch": 1.2758620689655173,
|
281 |
+
"grad_norm": 4.922264099121094,
|
282 |
+
"learning_rate": 1.810344827586207e-05,
|
283 |
+
"loss": 0.4311,
|
284 |
+
"step": 1073
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 1.3103448275862069,
|
288 |
+
"grad_norm": 3.692586898803711,
|
289 |
+
"learning_rate": 1.7241379310344828e-05,
|
290 |
+
"loss": 0.3667,
|
291 |
+
"step": 1102
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 1.3448275862068966,
|
295 |
+
"grad_norm": 2.88181471824646,
|
296 |
+
"learning_rate": 1.6379310344827585e-05,
|
297 |
+
"loss": 0.3167,
|
298 |
+
"step": 1131
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 1.3793103448275863,
|
302 |
+
"grad_norm": 3.277984142303467,
|
303 |
+
"learning_rate": 1.5517241379310346e-05,
|
304 |
+
"loss": 0.3331,
|
305 |
+
"step": 1160
|
306 |
+
},
|
307 |
+
{
|
308 |
+
"epoch": 1.4137931034482758,
|
309 |
+
"grad_norm": 7.566446304321289,
|
310 |
+
"learning_rate": 1.4655172413793103e-05,
|
311 |
+
"loss": 0.3046,
|
312 |
+
"step": 1189
|
313 |
+
},
|
314 |
+
{
|
315 |
+
"epoch": 1.4482758620689655,
|
316 |
+
"grad_norm": 17.953258514404297,
|
317 |
+
"learning_rate": 1.3793103448275863e-05,
|
318 |
+
"loss": 0.3332,
|
319 |
+
"step": 1218
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"epoch": 1.4827586206896552,
|
323 |
+
"grad_norm": 11.560026168823242,
|
324 |
+
"learning_rate": 1.2931034482758622e-05,
|
325 |
+
"loss": 0.3299,
|
326 |
+
"step": 1247
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 1.5172413793103448,
|
330 |
+
"grad_norm": 5.917276859283447,
|
331 |
+
"learning_rate": 1.206896551724138e-05,
|
332 |
+
"loss": 0.2961,
|
333 |
+
"step": 1276
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 1.5517241379310345,
|
337 |
+
"grad_norm": 3.665133476257324,
|
338 |
+
"learning_rate": 1.1206896551724138e-05,
|
339 |
+
"loss": 0.3142,
|
340 |
+
"step": 1305
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 1.5862068965517242,
|
344 |
+
"grad_norm": 2.3258779048919678,
|
345 |
+
"learning_rate": 1.0344827586206897e-05,
|
346 |
+
"loss": 0.3005,
|
347 |
+
"step": 1334
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 1.6206896551724137,
|
351 |
+
"grad_norm": 2.856088638305664,
|
352 |
+
"learning_rate": 9.482758620689655e-06,
|
353 |
+
"loss": 0.2652,
|
354 |
+
"step": 1363
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 1.6551724137931034,
|
358 |
+
"grad_norm": 8.568778991699219,
|
359 |
+
"learning_rate": 8.620689655172414e-06,
|
360 |
+
"loss": 0.2652,
|
361 |
+
"step": 1392
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 1.6896551724137931,
|
365 |
+
"grad_norm": 4.4803667068481445,
|
366 |
+
"learning_rate": 7.758620689655173e-06,
|
367 |
+
"loss": 0.2541,
|
368 |
+
"step": 1421
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 1.7241379310344827,
|
372 |
+
"grad_norm": 13.121492385864258,
|
373 |
+
"learning_rate": 6.896551724137932e-06,
|
374 |
+
"loss": 0.2754,
|
375 |
+
"step": 1450
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 1.7586206896551724,
|
379 |
+
"grad_norm": 2.48468279838562,
|
380 |
+
"learning_rate": 6.03448275862069e-06,
|
381 |
+
"loss": 0.2379,
|
382 |
+
"step": 1479
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 1.793103448275862,
|
386 |
+
"grad_norm": 1.497287631034851,
|
387 |
+
"learning_rate": 5.172413793103448e-06,
|
388 |
+
"loss": 0.2273,
|
389 |
+
"step": 1508
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 1.8275862068965516,
|
393 |
+
"grad_norm": 2.972078800201416,
|
394 |
+
"learning_rate": 4.310344827586207e-06,
|
395 |
+
"loss": 0.2254,
|
396 |
+
"step": 1537
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 1.8620689655172413,
|
400 |
+
"grad_norm": 12.911340713500977,
|
401 |
+
"learning_rate": 3.448275862068966e-06,
|
402 |
+
"loss": 0.2448,
|
403 |
+
"step": 1566
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 1.896551724137931,
|
407 |
+
"grad_norm": 1.3689017295837402,
|
408 |
+
"learning_rate": 2.586206896551724e-06,
|
409 |
+
"loss": 0.2089,
|
410 |
+
"step": 1595
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 1.9310344827586206,
|
414 |
+
"grad_norm": 4.04969596862793,
|
415 |
+
"learning_rate": 1.724137931034483e-06,
|
416 |
+
"loss": 0.2174,
|
417 |
+
"step": 1624
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 1.9655172413793105,
|
421 |
+
"grad_norm": 2.9180474281311035,
|
422 |
+
"learning_rate": 8.620689655172415e-07,
|
423 |
+
"loss": 0.2381,
|
424 |
+
"step": 1653
|
425 |
+
},
|
426 |
+
{
|
427 |
+
"epoch": 2.0,
|
428 |
+
"grad_norm": 4.1190409660339355,
|
429 |
+
"learning_rate": 0.0,
|
430 |
+
"loss": 0.2068,
|
431 |
+
"step": 1682
|
432 |
+
},
|
433 |
{
|
434 |
"epoch": 2.0,
|
435 |
+
"eval_cer": 0.0032914143766495886,
|
436 |
+
"eval_loss": 0.23910197615623474,
|
437 |
+
"eval_runtime": 636.093,
|
438 |
+
"eval_samples_per_second": 2.644,
|
439 |
+
"eval_steps_per_second": 0.332,
|
440 |
+
"step": 1682
|
441 |
},
|
442 |
{
|
443 |
"epoch": 2.0,
|
444 |
+
"step": 1682,
|
445 |
+
"total_flos": 1.9906356553640313e+19,
|
446 |
+
"train_loss": 0.6271170827069549,
|
447 |
+
"train_runtime": 2733.4434,
|
448 |
+
"train_samples_per_second": 4.92,
|
449 |
+
"train_steps_per_second": 0.615
|
450 |
}
|
451 |
],
|
452 |
"logging_steps": 29,
|
453 |
+
"max_steps": 1682,
|
454 |
"num_input_tokens_seen": 0,
|
455 |
"num_train_epochs": 2,
|
456 |
"save_steps": 500,
|
|
|
466 |
"attributes": {}
|
467 |
}
|
468 |
},
|
469 |
+
"total_flos": 1.9906356553640313e+19,
|
470 |
+
"train_batch_size": 8,
|
471 |
"trial_name": null,
|
472 |
"trial_params": null
|
473 |
}
|