Hugging Face's logo

xinyanghuang
/
Basic-Visual-Language-Model

Visual Question Answering
Model card Files Community
xinyanghuang commited on
Commit
a6334e8
•
1 Parent(s): 2d966ee

Upload 22 files

Browse files
Files changed (22) hide show
  1. checkpoint-19000/checkpoint-19000/adapter_config.json +18 -0
  2. checkpoint-19000/checkpoint-19000/adapter_model.bin +3 -0
  3. checkpoint-19000/checkpoint-19000/optimizer.pt +3 -0
  4. checkpoint-19000/checkpoint-19000/other_params.bin +3 -0
  5. checkpoint-19000/checkpoint-19000/rng_state_0.pth +3 -0
  6. checkpoint-19000/checkpoint-19000/rng_state_1.pth +3 -0
  7. checkpoint-19000/checkpoint-19000/rng_state_2.pth +3 -0
  8. checkpoint-19000/checkpoint-19000/rng_state_3.pth +3 -0
  9. checkpoint-19000/checkpoint-19000/scheduler.pt +3 -0
  10. checkpoint-19000/checkpoint-19000/trainer_state.json +1363 -0
  11. checkpoint-19000/checkpoint-19000/training_args.bin +3 -0
  12. checkpoint-36000/checkpoint-36000/adapter_config.json +18 -0
  13. checkpoint-36000/checkpoint-36000/adapter_model.bin +3 -0
  14. checkpoint-36000/checkpoint-36000/optimizer.pt +3 -0
  15. checkpoint-36000/checkpoint-36000/other_params.bin +3 -0
  16. checkpoint-36000/checkpoint-36000/rng_state_0.pth +3 -0
  17. checkpoint-36000/checkpoint-36000/rng_state_1.pth +3 -0
  18. checkpoint-36000/checkpoint-36000/rng_state_2.pth +3 -0
  19. checkpoint-36000/checkpoint-36000/rng_state_3.pth +3 -0
  20. checkpoint-36000/checkpoint-36000/scheduler.pt +3 -0
  21. checkpoint-36000/checkpoint-36000/trainer_state.json +2553 -0
  22. checkpoint-36000/checkpoint-36000/training_args.bin +3 -0
checkpoint-19000/checkpoint-19000/adapter_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "Qwen/Qwen-7B-Chat",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": false,
6
+ "init_lora_weights": true,
7
+ "lora_alpha": 32,
8
+ "lora_dropout": 0.1,
9
+ "modules_to_save": null,
10
+ "peft_type": "LORA",
11
+ "r": 128,
12
+ "target_modules": [
13
+ "c_attn",
14
+ "w1",
15
+ "w2"
16
+ ],
17
+ "task_type": "CAUSAL_LM"
18
+ }
checkpoint-19000/checkpoint-19000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0b0d273c5dcf7922768bf73e3284a8727839c083f676ec3948217cfa3fe8ee2
3
+ size 763431181
checkpoint-19000/checkpoint-19000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33e4150b7096b70bbe83927e2b2c6611b610616b0224717012734ae09a2c0817
3
+ size 1612910405
checkpoint-19000/checkpoint-19000/other_params.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7db2025e832208f4b1af4214ed077c5fcd2203f580c4c9687f8a220a7b0ad821
3
+ size 43009725
checkpoint-19000/checkpoint-19000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63034a2b25e07efb3e6bc582fb25af5b22e3932230b097e442d3734a38133cdb
3
+ size 17655
checkpoint-19000/checkpoint-19000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90b16dd91648e087e809e6e6eb9299749c52f015f4e4b7d2cb4014e05b5fd2c2
3
+ size 17655
checkpoint-19000/checkpoint-19000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5bfcbc96ad86fa7a01a43d625182ae6ac6f00ff2a591ed7138cf0a4202ea7d2
3
+ size 17655
checkpoint-19000/checkpoint-19000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f7ae09b17de1b0db3b05dd511e46cfaf000cb40e7184b95904cff5bf82c5a28
3
+ size 17655
checkpoint-19000/checkpoint-19000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc008ccfd1f4cfe14b07dd7404644bd07421c254c6c21defe9b2c2b2e246be81
3
+ size 627
checkpoint-19000/checkpoint-19000/trainer_state.json ADDED
@@ -0,0 +1,1363 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0528649008090436,
5
+ "eval_steps": 500,
6
+ "global_step": 19000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.00554139421478444,
13
+ "grad_norm": 0.7744311690330505,
14
+ "learning_rate": 2.9916879086778235e-05,
15
+ "loss": 2.174,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.01108278842956888,
20
+ "grad_norm": 1.1890419721603394,
21
+ "learning_rate": 2.983375817355647e-05,
22
+ "loss": 1.7489,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.01662418264435332,
27
+ "grad_norm": 2.874730348587036,
28
+ "learning_rate": 2.97506372603347e-05,
29
+ "loss": 1.6593,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.02216557685913776,
34
+ "grad_norm": 0.4631105363368988,
35
+ "learning_rate": 2.9667516347112933e-05,
36
+ "loss": 1.5928,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.0277069710739222,
41
+ "grad_norm": 1.1168841123580933,
42
+ "learning_rate": 2.9584395433891167e-05,
43
+ "loss": 1.5494,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.03324836528870664,
48
+ "grad_norm": 0.8050689101219177,
49
+ "learning_rate": 2.95012745206694e-05,
50
+ "loss": 1.5041,
51
+ "step": 600
52
+ },
53
+ {
54
+ "epoch": 0.038789759503491075,
55
+ "grad_norm": 2.871495008468628,
56
+ "learning_rate": 2.9418153607447632e-05,
57
+ "loss": 1.4829,
58
+ "step": 700
59
+ },
60
+ {
61
+ "epoch": 0.04433115371827552,
62
+ "grad_norm": 0.6971802711486816,
63
+ "learning_rate": 2.9335032694225866e-05,
64
+ "loss": 1.4492,
65
+ "step": 800
66
+ },
67
+ {
68
+ "epoch": 0.04987254793305996,
69
+ "grad_norm": 1.1272006034851074,
70
+ "learning_rate": 2.92519117810041e-05,
71
+ "loss": 1.45,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 0.0554139421478444,
76
+ "grad_norm": 0.9110287427902222,
77
+ "learning_rate": 2.9168790867782334e-05,
78
+ "loss": 1.4553,
79
+ "step": 1000
80
+ },
81
+ {
82
+ "epoch": 0.060955336362628836,
83
+ "grad_norm": 2.7382473945617676,
84
+ "learning_rate": 2.9085669954560568e-05,
85
+ "loss": 1.4244,
86
+ "step": 1100
87
+ },
88
+ {
89
+ "epoch": 0.06649673057741327,
90
+ "grad_norm": 2.678687334060669,
91
+ "learning_rate": 2.9002549041338802e-05,
92
+ "loss": 1.4399,
93
+ "step": 1200
94
+ },
95
+ {
96
+ "epoch": 0.07203812479219772,
97
+ "grad_norm": 0.8868110775947571,
98
+ "learning_rate": 2.8919428128117036e-05,
99
+ "loss": 1.408,
100
+ "step": 1300
101
+ },
102
+ {
103
+ "epoch": 0.07757951900698215,
104
+ "grad_norm": 0.7350853085517883,
105
+ "learning_rate": 2.883630721489527e-05,
106
+ "loss": 1.4128,
107
+ "step": 1400
108
+ },
109
+ {
110
+ "epoch": 0.0831209132217666,
111
+ "grad_norm": 2.349001169204712,
112
+ "learning_rate": 2.8753186301673504e-05,
113
+ "loss": 1.4154,
114
+ "step": 1500
115
+ },
116
+ {
117
+ "epoch": 0.08866230743655104,
118
+ "grad_norm": 2.588608503341675,
119
+ "learning_rate": 2.8670065388451735e-05,
120
+ "loss": 1.4102,
121
+ "step": 1600
122
+ },
123
+ {
124
+ "epoch": 0.09420370165133547,
125
+ "grad_norm": 3.461966037750244,
126
+ "learning_rate": 2.858694447522997e-05,
127
+ "loss": 1.3899,
128
+ "step": 1700
129
+ },
130
+ {
131
+ "epoch": 0.09974509586611992,
132
+ "grad_norm": 0.9384596943855286,
133
+ "learning_rate": 2.8503823562008203e-05,
134
+ "loss": 1.3946,
135
+ "step": 1800
136
+ },
137
+ {
138
+ "epoch": 0.10528649008090435,
139
+ "grad_norm": 2.6963839530944824,
140
+ "learning_rate": 2.8420702648786437e-05,
141
+ "loss": 1.3625,
142
+ "step": 1900
143
+ },
144
+ {
145
+ "epoch": 0.1108278842956888,
146
+ "grad_norm": 1.2156896591186523,
147
+ "learning_rate": 2.8337581735564668e-05,
148
+ "loss": 1.3998,
149
+ "step": 2000
150
+ },
151
+ {
152
+ "epoch": 0.11636927851047324,
153
+ "grad_norm": 0.9613581299781799,
154
+ "learning_rate": 2.82544608223429e-05,
155
+ "loss": 1.3672,
156
+ "step": 2100
157
+ },
158
+ {
159
+ "epoch": 0.12191067272525767,
160
+ "grad_norm": 0.8994789719581604,
161
+ "learning_rate": 2.8171339909121136e-05,
162
+ "loss": 1.3587,
163
+ "step": 2200
164
+ },
165
+ {
166
+ "epoch": 0.1274520669400421,
167
+ "grad_norm": 1.7463105916976929,
168
+ "learning_rate": 2.808821899589937e-05,
169
+ "loss": 1.3688,
170
+ "step": 2300
171
+ },
172
+ {
173
+ "epoch": 0.13299346115482655,
174
+ "grad_norm": 0.7639077305793762,
175
+ "learning_rate": 2.8005098082677604e-05,
176
+ "loss": 1.3798,
177
+ "step": 2400
178
+ },
179
+ {
180
+ "epoch": 0.138534855369611,
181
+ "grad_norm": 1.313306450843811,
182
+ "learning_rate": 2.7921977169455834e-05,
183
+ "loss": 1.3713,
184
+ "step": 2500
185
+ },
186
+ {
187
+ "epoch": 0.14407624958439544,
188
+ "grad_norm": 1.796030044555664,
189
+ "learning_rate": 2.783885625623407e-05,
190
+ "loss": 1.3589,
191
+ "step": 2600
192
+ },
193
+ {
194
+ "epoch": 0.14961764379917988,
195
+ "grad_norm": 0.40334078669548035,
196
+ "learning_rate": 2.7755735343012302e-05,
197
+ "loss": 1.3527,
198
+ "step": 2700
199
+ },
200
+ {
201
+ "epoch": 0.1551590380139643,
202
+ "grad_norm": 0.7727888822555542,
203
+ "learning_rate": 2.7672614429790536e-05,
204
+ "loss": 1.338,
205
+ "step": 2800
206
+ },
207
+ {
208
+ "epoch": 0.16070043222874875,
209
+ "grad_norm": 0.8815018534660339,
210
+ "learning_rate": 2.7589493516568767e-05,
211
+ "loss": 1.3638,
212
+ "step": 2900
213
+ },
214
+ {
215
+ "epoch": 0.1662418264435332,
216
+ "grad_norm": 1.1675957441329956,
217
+ "learning_rate": 2.7506372603347e-05,
218
+ "loss": 1.3462,
219
+ "step": 3000
220
+ },
221
+ {
222
+ "epoch": 0.17178322065831764,
223
+ "grad_norm": 0.9449867606163025,
224
+ "learning_rate": 2.7423251690125235e-05,
225
+ "loss": 1.3286,
226
+ "step": 3100
227
+ },
228
+ {
229
+ "epoch": 0.17732461487310208,
230
+ "grad_norm": 0.4934576749801636,
231
+ "learning_rate": 2.734013077690347e-05,
232
+ "loss": 1.3326,
233
+ "step": 3200
234
+ },
235
+ {
236
+ "epoch": 0.1828660090878865,
237
+ "grad_norm": 1.0242582559585571,
238
+ "learning_rate": 2.7257009863681703e-05,
239
+ "loss": 1.3423,
240
+ "step": 3300
241
+ },
242
+ {
243
+ "epoch": 0.18840740330267094,
244
+ "grad_norm": 1.253219723701477,
245
+ "learning_rate": 2.7173888950459934e-05,
246
+ "loss": 1.3473,
247
+ "step": 3400
248
+ },
249
+ {
250
+ "epoch": 0.1939487975174554,
251
+ "grad_norm": 2.4441049098968506,
252
+ "learning_rate": 2.7090768037238168e-05,
253
+ "loss": 1.3321,
254
+ "step": 3500
255
+ },
256
+ {
257
+ "epoch": 0.19949019173223984,
258
+ "grad_norm": 1.119807481765747,
259
+ "learning_rate": 2.7007647124016405e-05,
260
+ "loss": 1.3282,
261
+ "step": 3600
262
+ },
263
+ {
264
+ "epoch": 0.20503158594702428,
265
+ "grad_norm": 0.5911545157432556,
266
+ "learning_rate": 2.692452621079464e-05,
267
+ "loss": 1.329,
268
+ "step": 3700
269
+ },
270
+ {
271
+ "epoch": 0.2105729801618087,
272
+ "grad_norm": 1.959672451019287,
273
+ "learning_rate": 2.684140529757287e-05,
274
+ "loss": 1.3322,
275
+ "step": 3800
276
+ },
277
+ {
278
+ "epoch": 0.21611437437659314,
279
+ "grad_norm": 0.9169633984565735,
280
+ "learning_rate": 2.6758284384351104e-05,
281
+ "loss": 1.3248,
282
+ "step": 3900
283
+ },
284
+ {
285
+ "epoch": 0.2216557685913776,
286
+ "grad_norm": 0.8207225203514099,
287
+ "learning_rate": 2.6675163471129338e-05,
288
+ "loss": 1.3303,
289
+ "step": 4000
290
+ },
291
+ {
292
+ "epoch": 0.22719716280616203,
293
+ "grad_norm": 1.0271358489990234,
294
+ "learning_rate": 2.6592042557907572e-05,
295
+ "loss": 1.3231,
296
+ "step": 4100
297
+ },
298
+ {
299
+ "epoch": 0.23273855702094648,
300
+ "grad_norm": 1.2483460903167725,
301
+ "learning_rate": 2.6508921644685806e-05,
302
+ "loss": 1.3234,
303
+ "step": 4200
304
+ },
305
+ {
306
+ "epoch": 0.2382799512357309,
307
+ "grad_norm": 0.8641893863677979,
308
+ "learning_rate": 2.6425800731464037e-05,
309
+ "loss": 1.2856,
310
+ "step": 4300
311
+ },
312
+ {
313
+ "epoch": 0.24382134545051534,
314
+ "grad_norm": 1.6275527477264404,
315
+ "learning_rate": 2.634267981824227e-05,
316
+ "loss": 1.3039,
317
+ "step": 4400
318
+ },
319
+ {
320
+ "epoch": 0.2493627396652998,
321
+ "grad_norm": 2.3479559421539307,
322
+ "learning_rate": 2.6259558905020505e-05,
323
+ "loss": 1.331,
324
+ "step": 4500
325
+ },
326
+ {
327
+ "epoch": 0.2549041338800842,
328
+ "grad_norm": 2.3728585243225098,
329
+ "learning_rate": 2.617643799179874e-05,
330
+ "loss": 1.3115,
331
+ "step": 4600
332
+ },
333
+ {
334
+ "epoch": 0.26044552809486865,
335
+ "grad_norm": 1.7304351329803467,
336
+ "learning_rate": 2.609331707857697e-05,
337
+ "loss": 1.32,
338
+ "step": 4700
339
+ },
340
+ {
341
+ "epoch": 0.2659869223096531,
342
+ "grad_norm": 3.1707870960235596,
343
+ "learning_rate": 2.6010196165355203e-05,
344
+ "loss": 1.3166,
345
+ "step": 4800
346
+ },
347
+ {
348
+ "epoch": 0.27152831652443754,
349
+ "grad_norm": 0.6431360840797424,
350
+ "learning_rate": 2.5927075252133437e-05,
351
+ "loss": 1.3117,
352
+ "step": 4900
353
+ },
354
+ {
355
+ "epoch": 0.277069710739222,
356
+ "grad_norm": 0.7438284754753113,
357
+ "learning_rate": 2.584395433891167e-05,
358
+ "loss": 1.3049,
359
+ "step": 5000
360
+ },
361
+ {
362
+ "epoch": 0.28261110495400643,
363
+ "grad_norm": 0.8397496342658997,
364
+ "learning_rate": 2.5760833425689905e-05,
365
+ "loss": 1.3013,
366
+ "step": 5100
367
+ },
368
+ {
369
+ "epoch": 0.2881524991687909,
370
+ "grad_norm": 1.3787566423416138,
371
+ "learning_rate": 2.5677712512468136e-05,
372
+ "loss": 1.32,
373
+ "step": 5200
374
+ },
375
+ {
376
+ "epoch": 0.2936938933835753,
377
+ "grad_norm": 0.6342838406562805,
378
+ "learning_rate": 2.559459159924637e-05,
379
+ "loss": 1.3265,
380
+ "step": 5300
381
+ },
382
+ {
383
+ "epoch": 0.29923528759835977,
384
+ "grad_norm": 1.0924850702285767,
385
+ "learning_rate": 2.5511470686024604e-05,
386
+ "loss": 1.3224,
387
+ "step": 5400
388
+ },
389
+ {
390
+ "epoch": 0.3047766818131442,
391
+ "grad_norm": 3.104914665222168,
392
+ "learning_rate": 2.5428349772802838e-05,
393
+ "loss": 1.2897,
394
+ "step": 5500
395
+ },
396
+ {
397
+ "epoch": 0.3103180760279286,
398
+ "grad_norm": 1.219292163848877,
399
+ "learning_rate": 2.534522885958107e-05,
400
+ "loss": 1.3013,
401
+ "step": 5600
402
+ },
403
+ {
404
+ "epoch": 0.31585947024271305,
405
+ "grad_norm": 0.743425726890564,
406
+ "learning_rate": 2.5262107946359303e-05,
407
+ "loss": 1.3069,
408
+ "step": 5700
409
+ },
410
+ {
411
+ "epoch": 0.3214008644574975,
412
+ "grad_norm": 3.0548131465911865,
413
+ "learning_rate": 2.5178987033137537e-05,
414
+ "loss": 1.3006,
415
+ "step": 5800
416
+ },
417
+ {
418
+ "epoch": 0.32694225867228194,
419
+ "grad_norm": 2.6230878829956055,
420
+ "learning_rate": 2.5095866119915774e-05,
421
+ "loss": 1.2954,
422
+ "step": 5900
423
+ },
424
+ {
425
+ "epoch": 0.3324836528870664,
426
+ "grad_norm": 1.5894700288772583,
427
+ "learning_rate": 2.5012745206694005e-05,
428
+ "loss": 1.2918,
429
+ "step": 6000
430
+ },
431
+ {
432
+ "epoch": 0.33802504710185083,
433
+ "grad_norm": 2.298469066619873,
434
+ "learning_rate": 2.492962429347224e-05,
435
+ "loss": 1.2974,
436
+ "step": 6100
437
+ },
438
+ {
439
+ "epoch": 0.3435664413166353,
440
+ "grad_norm": 2.3570539951324463,
441
+ "learning_rate": 2.4846503380250473e-05,
442
+ "loss": 1.2943,
443
+ "step": 6200
444
+ },
445
+ {
446
+ "epoch": 0.3491078355314197,
447
+ "grad_norm": 0.7656165361404419,
448
+ "learning_rate": 2.4763382467028707e-05,
449
+ "loss": 1.2774,
450
+ "step": 6300
451
+ },
452
+ {
453
+ "epoch": 0.35464922974620416,
454
+ "grad_norm": 0.9744777679443359,
455
+ "learning_rate": 2.468026155380694e-05,
456
+ "loss": 1.3062,
457
+ "step": 6400
458
+ },
459
+ {
460
+ "epoch": 0.3601906239609886,
461
+ "grad_norm": 2.929172992706299,
462
+ "learning_rate": 2.459714064058517e-05,
463
+ "loss": 1.2871,
464
+ "step": 6500
465
+ },
466
+ {
467
+ "epoch": 0.365732018175773,
468
+ "grad_norm": 1.1587573289871216,
469
+ "learning_rate": 2.4514019727363405e-05,
470
+ "loss": 1.3037,
471
+ "step": 6600
472
+ },
473
+ {
474
+ "epoch": 0.37127341239055744,
475
+ "grad_norm": 1.055335521697998,
476
+ "learning_rate": 2.443089881414164e-05,
477
+ "loss": 1.2823,
478
+ "step": 6700
479
+ },
480
+ {
481
+ "epoch": 0.3768148066053419,
482
+ "grad_norm": 2.3376619815826416,
483
+ "learning_rate": 2.4347777900919873e-05,
484
+ "loss": 1.2854,
485
+ "step": 6800
486
+ },
487
+ {
488
+ "epoch": 0.38235620082012634,
489
+ "grad_norm": 0.94229656457901,
490
+ "learning_rate": 2.4264656987698104e-05,
491
+ "loss": 1.3007,
492
+ "step": 6900
493
+ },
494
+ {
495
+ "epoch": 0.3878975950349108,
496
+ "grad_norm": 1.2927848100662231,
497
+ "learning_rate": 2.4181536074476338e-05,
498
+ "loss": 1.278,
499
+ "step": 7000
500
+ },
501
+ {
502
+ "epoch": 0.3934389892496952,
503
+ "grad_norm": 0.6448485255241394,
504
+ "learning_rate": 2.4098415161254572e-05,
505
+ "loss": 1.275,
506
+ "step": 7100
507
+ },
508
+ {
509
+ "epoch": 0.39898038346447967,
510
+ "grad_norm": 1.4731674194335938,
511
+ "learning_rate": 2.4015294248032806e-05,
512
+ "loss": 1.2737,
513
+ "step": 7200
514
+ },
515
+ {
516
+ "epoch": 0.4045217776792641,
517
+ "grad_norm": 2.395692825317383,
518
+ "learning_rate": 2.393217333481104e-05,
519
+ "loss": 1.285,
520
+ "step": 7300
521
+ },
522
+ {
523
+ "epoch": 0.41006317189404856,
524
+ "grad_norm": 0.8327000737190247,
525
+ "learning_rate": 2.384905242158927e-05,
526
+ "loss": 1.2514,
527
+ "step": 7400
528
+ },
529
+ {
530
+ "epoch": 0.415604566108833,
531
+ "grad_norm": 0.5383670330047607,
532
+ "learning_rate": 2.3765931508367505e-05,
533
+ "loss": 1.294,
534
+ "step": 7500
535
+ },
536
+ {
537
+ "epoch": 0.4211459603236174,
538
+ "grad_norm": 0.7433611154556274,
539
+ "learning_rate": 2.368281059514574e-05,
540
+ "loss": 1.2966,
541
+ "step": 7600
542
+ },
543
+ {
544
+ "epoch": 0.42668735453840184,
545
+ "grad_norm": 1.4880763292312622,
546
+ "learning_rate": 2.3599689681923973e-05,
547
+ "loss": 1.3076,
548
+ "step": 7700
549
+ },
550
+ {
551
+ "epoch": 0.4322287487531863,
552
+ "grad_norm": 1.2045841217041016,
553
+ "learning_rate": 2.3516568768702204e-05,
554
+ "loss": 1.2598,
555
+ "step": 7800
556
+ },
557
+ {
558
+ "epoch": 0.43777014296797073,
559
+ "grad_norm": 2.087782621383667,
560
+ "learning_rate": 2.3433447855480438e-05,
561
+ "loss": 1.2832,
562
+ "step": 7900
563
+ },
564
+ {
565
+ "epoch": 0.4433115371827552,
566
+ "grad_norm": 0.9906471371650696,
567
+ "learning_rate": 2.335032694225867e-05,
568
+ "loss": 1.2723,
569
+ "step": 8000
570
+ },
571
+ {
572
+ "epoch": 0.4488529313975396,
573
+ "grad_norm": 2.195661783218384,
574
+ "learning_rate": 2.3267206029036906e-05,
575
+ "loss": 1.2707,
576
+ "step": 8100
577
+ },
578
+ {
579
+ "epoch": 0.45439432561232407,
580
+ "grad_norm": 2.5563039779663086,
581
+ "learning_rate": 2.318408511581514e-05,
582
+ "loss": 1.2905,
583
+ "step": 8200
584
+ },
585
+ {
586
+ "epoch": 0.4599357198271085,
587
+ "grad_norm": 0.7435129284858704,
588
+ "learning_rate": 2.3100964202593374e-05,
589
+ "loss": 1.2814,
590
+ "step": 8300
591
+ },
592
+ {
593
+ "epoch": 0.46547711404189296,
594
+ "grad_norm": 0.8438817262649536,
595
+ "learning_rate": 2.3017843289371608e-05,
596
+ "loss": 1.2745,
597
+ "step": 8400
598
+ },
599
+ {
600
+ "epoch": 0.4710185082566774,
601
+ "grad_norm": 1.848270058631897,
602
+ "learning_rate": 2.293472237614984e-05,
603
+ "loss": 1.2735,
604
+ "step": 8500
605
+ },
606
+ {
607
+ "epoch": 0.4765599024714618,
608
+ "grad_norm": 0.9101561307907104,
609
+ "learning_rate": 2.2851601462928076e-05,
610
+ "loss": 1.2576,
611
+ "step": 8600
612
+ },
613
+ {
614
+ "epoch": 0.48210129668624624,
615
+ "grad_norm": 1.2922989130020142,
616
+ "learning_rate": 2.2768480549706306e-05,
617
+ "loss": 1.2824,
618
+ "step": 8700
619
+ },
620
+ {
621
+ "epoch": 0.4876426909010307,
622
+ "grad_norm": 0.7475651502609253,
623
+ "learning_rate": 2.268535963648454e-05,
624
+ "loss": 1.2877,
625
+ "step": 8800
626
+ },
627
+ {
628
+ "epoch": 0.49318408511581513,
629
+ "grad_norm": 1.3247780799865723,
630
+ "learning_rate": 2.2602238723262774e-05,
631
+ "loss": 1.278,
632
+ "step": 8900
633
+ },
634
+ {
635
+ "epoch": 0.4987254793305996,
636
+ "grad_norm": 0.9011294841766357,
637
+ "learning_rate": 2.251911781004101e-05,
638
+ "loss": 1.2776,
639
+ "step": 9000
640
+ },
641
+ {
642
+ "epoch": 0.504266873545384,
643
+ "grad_norm": 3.1133978366851807,
644
+ "learning_rate": 2.2435996896819242e-05,
645
+ "loss": 1.2732,
646
+ "step": 9100
647
+ },
648
+ {
649
+ "epoch": 0.5098082677601684,
650
+ "grad_norm": 0.9149213433265686,
651
+ "learning_rate": 2.2352875983597473e-05,
652
+ "loss": 1.2806,
653
+ "step": 9200
654
+ },
655
+ {
656
+ "epoch": 0.5153496619749529,
657
+ "grad_norm": 0.8003007769584656,
658
+ "learning_rate": 2.2269755070375707e-05,
659
+ "loss": 1.2795,
660
+ "step": 9300
661
+ },
662
+ {
663
+ "epoch": 0.5208910561897373,
664
+ "grad_norm": 0.8392866253852844,
665
+ "learning_rate": 2.218663415715394e-05,
666
+ "loss": 1.2715,
667
+ "step": 9400
668
+ },
669
+ {
670
+ "epoch": 0.5264324504045218,
671
+ "grad_norm": 0.7663789391517639,
672
+ "learning_rate": 2.2103513243932175e-05,
673
+ "loss": 1.2524,
674
+ "step": 9500
675
+ },
676
+ {
677
+ "epoch": 0.5319738446193062,
678
+ "grad_norm": 2.0740444660186768,
679
+ "learning_rate": 2.2020392330710406e-05,
680
+ "loss": 1.2687,
681
+ "step": 9600
682
+ },
683
+ {
684
+ "epoch": 0.5375152388340907,
685
+ "grad_norm": 3.680366277694702,
686
+ "learning_rate": 2.193727141748864e-05,
687
+ "loss": 1.2903,
688
+ "step": 9700
689
+ },
690
+ {
691
+ "epoch": 0.5430566330488751,
692
+ "grad_norm": 0.9677708745002747,
693
+ "learning_rate": 2.1854150504266874e-05,
694
+ "loss": 1.2655,
695
+ "step": 9800
696
+ },
697
+ {
698
+ "epoch": 0.5485980272636596,
699
+ "grad_norm": 1.019551157951355,
700
+ "learning_rate": 2.1771029591045108e-05,
701
+ "loss": 1.2664,
702
+ "step": 9900
703
+ },
704
+ {
705
+ "epoch": 0.554139421478444,
706
+ "grad_norm": 3.291146993637085,
707
+ "learning_rate": 2.168790867782334e-05,
708
+ "loss": 1.2716,
709
+ "step": 10000
710
+ },
711
+ {
712
+ "epoch": 0.5596808156932284,
713
+ "grad_norm": 2.067105770111084,
714
+ "learning_rate": 2.1604787764601572e-05,
715
+ "loss": 1.2703,
716
+ "step": 10100
717
+ },
718
+ {
719
+ "epoch": 0.5652222099080129,
720
+ "grad_norm": 0.6159241199493408,
721
+ "learning_rate": 2.1521666851379806e-05,
722
+ "loss": 1.269,
723
+ "step": 10200
724
+ },
725
+ {
726
+ "epoch": 0.5707636041227973,
727
+ "grad_norm": 0.8273797631263733,
728
+ "learning_rate": 2.143854593815804e-05,
729
+ "loss": 1.2679,
730
+ "step": 10300
731
+ },
732
+ {
733
+ "epoch": 0.5763049983375818,
734
+ "grad_norm": 1.0053974390029907,
735
+ "learning_rate": 2.1355425024936274e-05,
736
+ "loss": 1.2615,
737
+ "step": 10400
738
+ },
739
+ {
740
+ "epoch": 0.5818463925523661,
741
+ "grad_norm": 0.7567972540855408,
742
+ "learning_rate": 2.1272304111714505e-05,
743
+ "loss": 1.2706,
744
+ "step": 10500
745
+ },
746
+ {
747
+ "epoch": 0.5873877867671506,
748
+ "grad_norm": 1.5227950811386108,
749
+ "learning_rate": 2.1189183198492743e-05,
750
+ "loss": 1.2853,
751
+ "step": 10600
752
+ },
753
+ {
754
+ "epoch": 0.592929180981935,
755
+ "grad_norm": 0.7776534557342529,
756
+ "learning_rate": 2.1106062285270977e-05,
757
+ "loss": 1.267,
758
+ "step": 10700
759
+ },
760
+ {
761
+ "epoch": 0.5984705751967195,
762
+ "grad_norm": 2.273895502090454,
763
+ "learning_rate": 2.102294137204921e-05,
764
+ "loss": 1.2587,
765
+ "step": 10800
766
+ },
767
+ {
768
+ "epoch": 0.6040119694115039,
769
+ "grad_norm": 2.182058334350586,
770
+ "learning_rate": 2.093982045882744e-05,
771
+ "loss": 1.2676,
772
+ "step": 10900
773
+ },
774
+ {
775
+ "epoch": 0.6095533636262884,
776
+ "grad_norm": 2.346633195877075,
777
+ "learning_rate": 2.0856699545605675e-05,
778
+ "loss": 1.2512,
779
+ "step": 11000
780
+ },
781
+ {
782
+ "epoch": 0.6150947578410728,
783
+ "grad_norm": 2.9466536045074463,
784
+ "learning_rate": 2.077357863238391e-05,
785
+ "loss": 1.2636,
786
+ "step": 11100
787
+ },
788
+ {
789
+ "epoch": 0.6206361520558572,
790
+ "grad_norm": 2.7352640628814697,
791
+ "learning_rate": 2.0690457719162143e-05,
792
+ "loss": 1.2677,
793
+ "step": 11200
794
+ },
795
+ {
796
+ "epoch": 0.6261775462706417,
797
+ "grad_norm": 0.4942292869091034,
798
+ "learning_rate": 2.0607336805940377e-05,
799
+ "loss": 1.2512,
800
+ "step": 11300
801
+ },
802
+ {
803
+ "epoch": 0.6317189404854261,
804
+ "grad_norm": 1.3134087324142456,
805
+ "learning_rate": 2.0524215892718608e-05,
806
+ "loss": 1.2563,
807
+ "step": 11400
808
+ },
809
+ {
810
+ "epoch": 0.6372603347002106,
811
+ "grad_norm": 1.0919684171676636,
812
+ "learning_rate": 2.0441094979496842e-05,
813
+ "loss": 1.2402,
814
+ "step": 11500
815
+ },
816
+ {
817
+ "epoch": 0.642801728914995,
818
+ "grad_norm": 1.3876844644546509,
819
+ "learning_rate": 2.0357974066275076e-05,
820
+ "loss": 1.2788,
821
+ "step": 11600
822
+ },
823
+ {
824
+ "epoch": 0.6483431231297795,
825
+ "grad_norm": 2.135866165161133,
826
+ "learning_rate": 2.027485315305331e-05,
827
+ "loss": 1.2556,
828
+ "step": 11700
829
+ },
830
+ {
831
+ "epoch": 0.6538845173445639,
832
+ "grad_norm": 3.400996685028076,
833
+ "learning_rate": 2.019173223983154e-05,
834
+ "loss": 1.2706,
835
+ "step": 11800
836
+ },
837
+ {
838
+ "epoch": 0.6594259115593484,
839
+ "grad_norm": 1.1860913038253784,
840
+ "learning_rate": 2.0108611326609775e-05,
841
+ "loss": 1.2538,
842
+ "step": 11900
843
+ },
844
+ {
845
+ "epoch": 0.6649673057741328,
846
+ "grad_norm": 1.044102430343628,
847
+ "learning_rate": 2.002549041338801e-05,
848
+ "loss": 1.2573,
849
+ "step": 12000
850
+ },
851
+ {
852
+ "epoch": 0.6705086999889172,
853
+ "grad_norm": 1.0775806903839111,
854
+ "learning_rate": 1.9942369500166243e-05,
855
+ "loss": 1.2591,
856
+ "step": 12100
857
+ },
858
+ {
859
+ "epoch": 0.6760500942037017,
860
+ "grad_norm": 0.7967768311500549,
861
+ "learning_rate": 1.9859248586944477e-05,
862
+ "loss": 1.2394,
863
+ "step": 12200
864
+ },
865
+ {
866
+ "epoch": 0.681591488418486,
867
+ "grad_norm": 2.224622964859009,
868
+ "learning_rate": 1.9776127673722707e-05,
869
+ "loss": 1.2458,
870
+ "step": 12300
871
+ },
872
+ {
873
+ "epoch": 0.6871328826332705,
874
+ "grad_norm": 2.769935131072998,
875
+ "learning_rate": 1.969300676050094e-05,
876
+ "loss": 1.2665,
877
+ "step": 12400
878
+ },
879
+ {
880
+ "epoch": 0.6926742768480549,
881
+ "grad_norm": 1.0014837980270386,
882
+ "learning_rate": 1.9609885847279175e-05,
883
+ "loss": 1.2384,
884
+ "step": 12500
885
+ },
886
+ {
887
+ "epoch": 0.6982156710628394,
888
+ "grad_norm": 0.8904783725738525,
889
+ "learning_rate": 1.952676493405741e-05,
890
+ "loss": 1.2607,
891
+ "step": 12600
892
+ },
893
+ {
894
+ "epoch": 0.7037570652776238,
895
+ "grad_norm": 0.9549648761749268,
896
+ "learning_rate": 1.944364402083564e-05,
897
+ "loss": 1.2411,
898
+ "step": 12700
899
+ },
900
+ {
901
+ "epoch": 0.7092984594924083,
902
+ "grad_norm": 0.44161495566368103,
903
+ "learning_rate": 1.9360523107613874e-05,
904
+ "loss": 1.2645,
905
+ "step": 12800
906
+ },
907
+ {
908
+ "epoch": 0.7148398537071927,
909
+ "grad_norm": 1.002912163734436,
910
+ "learning_rate": 1.9277402194392108e-05,
911
+ "loss": 1.263,
912
+ "step": 12900
913
+ },
914
+ {
915
+ "epoch": 0.7203812479219772,
916
+ "grad_norm": 0.9343509078025818,
917
+ "learning_rate": 1.9194281281170345e-05,
918
+ "loss": 1.2663,
919
+ "step": 13000
920
+ },
921
+ {
922
+ "epoch": 0.7259226421367616,
923
+ "grad_norm": 0.7143359184265137,
924
+ "learning_rate": 1.9111160367948576e-05,
925
+ "loss": 1.2509,
926
+ "step": 13100
927
+ },
928
+ {
929
+ "epoch": 0.731464036351546,
930
+ "grad_norm": 3.670598030090332,
931
+ "learning_rate": 1.902803945472681e-05,
932
+ "loss": 1.2424,
933
+ "step": 13200
934
+ },
935
+ {
936
+ "epoch": 0.7370054305663305,
937
+ "grad_norm": 0.9394696354866028,
938
+ "learning_rate": 1.8944918541505044e-05,
939
+ "loss": 1.2302,
940
+ "step": 13300
941
+ },
942
+ {
943
+ "epoch": 0.7425468247811149,
944
+ "grad_norm": 2.5993099212646484,
945
+ "learning_rate": 1.8861797628283278e-05,
946
+ "loss": 1.2454,
947
+ "step": 13400
948
+ },
949
+ {
950
+ "epoch": 0.7480882189958994,
951
+ "grad_norm": 1.5783851146697998,
952
+ "learning_rate": 1.8778676715061512e-05,
953
+ "loss": 1.2613,
954
+ "step": 13500
955
+ },
956
+ {
957
+ "epoch": 0.7536296132106838,
958
+ "grad_norm": 0.6382520794868469,
959
+ "learning_rate": 1.8695555801839743e-05,
960
+ "loss": 1.2583,
961
+ "step": 13600
962
+ },
963
+ {
964
+ "epoch": 0.7591710074254683,
965
+ "grad_norm": 0.6521400809288025,
966
+ "learning_rate": 1.8612434888617977e-05,
967
+ "loss": 1.2253,
968
+ "step": 13700
969
+ },
970
+ {
971
+ "epoch": 0.7647124016402527,
972
+ "grad_norm": 2.8146674633026123,
973
+ "learning_rate": 1.852931397539621e-05,
974
+ "loss": 1.2577,
975
+ "step": 13800
976
+ },
977
+ {
978
+ "epoch": 0.7702537958550372,
979
+ "grad_norm": 2.7363874912261963,
980
+ "learning_rate": 1.8446193062174445e-05,
981
+ "loss": 1.2403,
982
+ "step": 13900
983
+ },
984
+ {
985
+ "epoch": 0.7757951900698216,
986
+ "grad_norm": 2.4337146282196045,
987
+ "learning_rate": 1.8363072148952676e-05,
988
+ "loss": 1.2522,
989
+ "step": 14000
990
+ },
991
+ {
992
+ "epoch": 0.7813365842846061,
993
+ "grad_norm": 0.9723544716835022,
994
+ "learning_rate": 1.827995123573091e-05,
995
+ "loss": 1.2507,
996
+ "step": 14100
997
+ },
998
+ {
999
+ "epoch": 0.7868779784993905,
1000
+ "grad_norm": 1.1492934226989746,
1001
+ "learning_rate": 1.8196830322509144e-05,
1002
+ "loss": 1.23,
1003
+ "step": 14200
1004
+ },
1005
+ {
1006
+ "epoch": 0.7924193727141748,
1007
+ "grad_norm": 0.8592559099197388,
1008
+ "learning_rate": 1.8113709409287378e-05,
1009
+ "loss": 1.2508,
1010
+ "step": 14300
1011
+ },
1012
+ {
1013
+ "epoch": 0.7979607669289593,
1014
+ "grad_norm": 1.035599708557129,
1015
+ "learning_rate": 1.803058849606561e-05,
1016
+ "loss": 1.2465,
1017
+ "step": 14400
1018
+ },
1019
+ {
1020
+ "epoch": 0.8035021611437437,
1021
+ "grad_norm": 0.6385583281517029,
1022
+ "learning_rate": 1.7947467582843842e-05,
1023
+ "loss": 1.2476,
1024
+ "step": 14500
1025
+ },
1026
+ {
1027
+ "epoch": 0.8090435553585282,
1028
+ "grad_norm": 1.695744276046753,
1029
+ "learning_rate": 1.7864346669622076e-05,
1030
+ "loss": 1.239,
1031
+ "step": 14600
1032
+ },
1033
+ {
1034
+ "epoch": 0.8145849495733126,
1035
+ "grad_norm": 2.5061330795288086,
1036
+ "learning_rate": 1.778122575640031e-05,
1037
+ "loss": 1.2527,
1038
+ "step": 14700
1039
+ },
1040
+ {
1041
+ "epoch": 0.8201263437880971,
1042
+ "grad_norm": 1.0300838947296143,
1043
+ "learning_rate": 1.7698104843178544e-05,
1044
+ "loss": 1.2406,
1045
+ "step": 14800
1046
+ },
1047
+ {
1048
+ "epoch": 0.8256677380028815,
1049
+ "grad_norm": 0.8938758969306946,
1050
+ "learning_rate": 1.7614983929956775e-05,
1051
+ "loss": 1.236,
1052
+ "step": 14900
1053
+ },
1054
+ {
1055
+ "epoch": 0.831209132217666,
1056
+ "grad_norm": 0.9264464974403381,
1057
+ "learning_rate": 1.753186301673501e-05,
1058
+ "loss": 1.2439,
1059
+ "step": 15000
1060
+ },
1061
+ {
1062
+ "epoch": 0.8367505264324504,
1063
+ "grad_norm": 1.1339832544326782,
1064
+ "learning_rate": 1.7448742103513243e-05,
1065
+ "loss": 1.2358,
1066
+ "step": 15100
1067
+ },
1068
+ {
1069
+ "epoch": 0.8422919206472348,
1070
+ "grad_norm": 1.239995002746582,
1071
+ "learning_rate": 1.7365621190291477e-05,
1072
+ "loss": 1.2505,
1073
+ "step": 15200
1074
+ },
1075
+ {
1076
+ "epoch": 0.8478333148620193,
1077
+ "grad_norm": 2.820875644683838,
1078
+ "learning_rate": 1.7282500277069714e-05,
1079
+ "loss": 1.2502,
1080
+ "step": 15300
1081
+ },
1082
+ {
1083
+ "epoch": 0.8533747090768037,
1084
+ "grad_norm": 0.5509181022644043,
1085
+ "learning_rate": 1.7199379363847945e-05,
1086
+ "loss": 1.2285,
1087
+ "step": 15400
1088
+ },
1089
+ {
1090
+ "epoch": 0.8589161032915882,
1091
+ "grad_norm": 2.2472527027130127,
1092
+ "learning_rate": 1.711625845062618e-05,
1093
+ "loss": 1.2443,
1094
+ "step": 15500
1095
+ },
1096
+ {
1097
+ "epoch": 0.8644574975063726,
1098
+ "grad_norm": 0.6996986269950867,
1099
+ "learning_rate": 1.7033137537404413e-05,
1100
+ "loss": 1.2415,
1101
+ "step": 15600
1102
+ },
1103
+ {
1104
+ "epoch": 0.8699988917211571,
1105
+ "grad_norm": 1.9218405485153198,
1106
+ "learning_rate": 1.6950016624182647e-05,
1107
+ "loss": 1.2414,
1108
+ "step": 15700
1109
+ },
1110
+ {
1111
+ "epoch": 0.8755402859359415,
1112
+ "grad_norm": 1.069169521331787,
1113
+ "learning_rate": 1.6866895710960878e-05,
1114
+ "loss": 1.2365,
1115
+ "step": 15800
1116
+ },
1117
+ {
1118
+ "epoch": 0.881081680150726,
1119
+ "grad_norm": 2.6105518341064453,
1120
+ "learning_rate": 1.6783774797739112e-05,
1121
+ "loss": 1.2423,
1122
+ "step": 15900
1123
+ },
1124
+ {
1125
+ "epoch": 0.8866230743655104,
1126
+ "grad_norm": 3.099567413330078,
1127
+ "learning_rate": 1.6700653884517346e-05,
1128
+ "loss": 1.2474,
1129
+ "step": 16000
1130
+ },
1131
+ {
1132
+ "epoch": 0.8921644685802949,
1133
+ "grad_norm": 0.6464875340461731,
1134
+ "learning_rate": 1.661753297129558e-05,
1135
+ "loss": 1.2378,
1136
+ "step": 16100
1137
+ },
1138
+ {
1139
+ "epoch": 0.8977058627950792,
1140
+ "grad_norm": 1.130702018737793,
1141
+ "learning_rate": 1.6534412058073814e-05,
1142
+ "loss": 1.2355,
1143
+ "step": 16200
1144
+ },
1145
+ {
1146
+ "epoch": 0.9032472570098636,
1147
+ "grad_norm": 2.6416752338409424,
1148
+ "learning_rate": 1.6451291144852044e-05,
1149
+ "loss": 1.2399,
1150
+ "step": 16300
1151
+ },
1152
+ {
1153
+ "epoch": 0.9087886512246481,
1154
+ "grad_norm": 0.5979084372520447,
1155
+ "learning_rate": 1.636817023163028e-05,
1156
+ "loss": 1.2297,
1157
+ "step": 16400
1158
+ },
1159
+ {
1160
+ "epoch": 0.9143300454394325,
1161
+ "grad_norm": 2.5381157398223877,
1162
+ "learning_rate": 1.6285049318408512e-05,
1163
+ "loss": 1.2315,
1164
+ "step": 16500
1165
+ },
1166
+ {
1167
+ "epoch": 0.919871439654217,
1168
+ "grad_norm": 1.3977634906768799,
1169
+ "learning_rate": 1.6201928405186746e-05,
1170
+ "loss": 1.2441,
1171
+ "step": 16600
1172
+ },
1173
+ {
1174
+ "epoch": 0.9254128338690014,
1175
+ "grad_norm": 0.5126473307609558,
1176
+ "learning_rate": 1.6118807491964977e-05,
1177
+ "loss": 1.2377,
1178
+ "step": 16700
1179
+ },
1180
+ {
1181
+ "epoch": 0.9309542280837859,
1182
+ "grad_norm": 1.1525436639785767,
1183
+ "learning_rate": 1.603568657874321e-05,
1184
+ "loss": 1.2535,
1185
+ "step": 16800
1186
+ },
1187
+ {
1188
+ "epoch": 0.9364956222985703,
1189
+ "grad_norm": 1.0181013345718384,
1190
+ "learning_rate": 1.5952565665521445e-05,
1191
+ "loss": 1.2335,
1192
+ "step": 16900
1193
+ },
1194
+ {
1195
+ "epoch": 0.9420370165133548,
1196
+ "grad_norm": 1.0378520488739014,
1197
+ "learning_rate": 1.586944475229968e-05,
1198
+ "loss": 1.2429,
1199
+ "step": 17000
1200
+ },
1201
+ {
1202
+ "epoch": 0.9475784107281392,
1203
+ "grad_norm": 0.7528340220451355,
1204
+ "learning_rate": 1.578632383907791e-05,
1205
+ "loss": 1.2432,
1206
+ "step": 17100
1207
+ },
1208
+ {
1209
+ "epoch": 0.9531198049429236,
1210
+ "grad_norm": 0.6110236048698425,
1211
+ "learning_rate": 1.5703202925856144e-05,
1212
+ "loss": 1.251,
1213
+ "step": 17200
1214
+ },
1215
+ {
1216
+ "epoch": 0.9586611991577081,
1217
+ "grad_norm": 1.2787925004959106,
1218
+ "learning_rate": 1.5620082012634378e-05,
1219
+ "loss": 1.2243,
1220
+ "step": 17300
1221
+ },
1222
+ {
1223
+ "epoch": 0.9642025933724925,
1224
+ "grad_norm": 0.8089519739151001,
1225
+ "learning_rate": 1.5536961099412612e-05,
1226
+ "loss": 1.2375,
1227
+ "step": 17400
1228
+ },
1229
+ {
1230
+ "epoch": 0.969743987587277,
1231
+ "grad_norm": 0.702037513256073,
1232
+ "learning_rate": 1.5453840186190846e-05,
1233
+ "loss": 1.2231,
1234
+ "step": 17500
1235
+ },
1236
+ {
1237
+ "epoch": 0.9752853818020614,
1238
+ "grad_norm": 1.3774892091751099,
1239
+ "learning_rate": 1.5370719272969077e-05,
1240
+ "loss": 1.2516,
1241
+ "step": 17600
1242
+ },
1243
+ {
1244
+ "epoch": 0.9808267760168459,
1245
+ "grad_norm": 0.8553925156593323,
1246
+ "learning_rate": 1.5287598359747314e-05,
1247
+ "loss": 1.2329,
1248
+ "step": 17700
1249
+ },
1250
+ {
1251
+ "epoch": 0.9863681702316303,
1252
+ "grad_norm": 1.9921386241912842,
1253
+ "learning_rate": 1.5204477446525546e-05,
1254
+ "loss": 1.2474,
1255
+ "step": 17800
1256
+ },
1257
+ {
1258
+ "epoch": 0.9919095644464148,
1259
+ "grad_norm": 0.7573133707046509,
1260
+ "learning_rate": 1.512135653330378e-05,
1261
+ "loss": 1.2578,
1262
+ "step": 17900
1263
+ },
1264
+ {
1265
+ "epoch": 0.9974509586611991,
1266
+ "grad_norm": 2.220541000366211,
1267
+ "learning_rate": 1.5038235620082011e-05,
1268
+ "loss": 1.2222,
1269
+ "step": 18000
1270
+ },
1271
+ {
1272
+ "epoch": 1.0029923528759837,
1273
+ "grad_norm": 3.79048228263855,
1274
+ "learning_rate": 1.4955114706860245e-05,
1275
+ "loss": 1.2138,
1276
+ "step": 18100
1277
+ },
1278
+ {
1279
+ "epoch": 1.008533747090768,
1280
+ "grad_norm": 2.1642391681671143,
1281
+ "learning_rate": 1.4871993793638479e-05,
1282
+ "loss": 1.2038,
1283
+ "step": 18200
1284
+ },
1285
+ {
1286
+ "epoch": 1.0140751413055524,
1287
+ "grad_norm": 0.719645082950592,
1288
+ "learning_rate": 1.4788872880416715e-05,
1289
+ "loss": 1.213,
1290
+ "step": 18300
1291
+ },
1292
+ {
1293
+ "epoch": 1.0196165355203368,
1294
+ "grad_norm": 0.9910865426063538,
1295
+ "learning_rate": 1.4705751967194947e-05,
1296
+ "loss": 1.2227,
1297
+ "step": 18400
1298
+ },
1299
+ {
1300
+ "epoch": 1.0251579297351214,
1301
+ "grad_norm": 0.8456553220748901,
1302
+ "learning_rate": 1.4622631053973181e-05,
1303
+ "loss": 1.2076,
1304
+ "step": 18500
1305
+ },
1306
+ {
1307
+ "epoch": 1.0306993239499058,
1308
+ "grad_norm": 0.7198830842971802,
1309
+ "learning_rate": 1.4539510140751413e-05,
1310
+ "loss": 1.2184,
1311
+ "step": 18600
1312
+ },
1313
+ {
1314
+ "epoch": 1.0362407181646902,
1315
+ "grad_norm": 2.5999066829681396,
1316
+ "learning_rate": 1.4456389227529647e-05,
1317
+ "loss": 1.2161,
1318
+ "step": 18700
1319
+ },
1320
+ {
1321
+ "epoch": 1.0417821123794746,
1322
+ "grad_norm": 1.0215507745742798,
1323
+ "learning_rate": 1.437326831430788e-05,
1324
+ "loss": 1.2148,
1325
+ "step": 18800
1326
+ },
1327
+ {
1328
+ "epoch": 1.0473235065942592,
1329
+ "grad_norm": 2.206827402114868,
1330
+ "learning_rate": 1.4290147401086114e-05,
1331
+ "loss": 1.2189,
1332
+ "step": 18900
1333
+ },
1334
+ {
1335
+ "epoch": 1.0528649008090436,
1336
+ "grad_norm": 1.1960684061050415,
1337
+ "learning_rate": 1.4207026487864346e-05,
1338
+ "loss": 1.2073,
1339
+ "step": 19000
1340
+ }
1341
+ ],
1342
+ "logging_steps": 100,
1343
+ "max_steps": 36092,
1344
+ "num_input_tokens_seen": 0,
1345
+ "num_train_epochs": 2,
1346
+ "save_steps": 1000,
1347
+ "stateful_callbacks": {
1348
+ "TrainerControl": {
1349
+ "args": {
1350
+ "should_epoch_stop": false,
1351
+ "should_evaluate": false,
1352
+ "should_log": false,
1353
+ "should_save": true,
1354
+ "should_training_stop": false
1355
+ },
1356
+ "attributes": {}
1357
+ }
1358
+ },
1359
+ "total_flos": 0.0,
1360
+ "train_batch_size": 4,
1361
+ "trial_name": null,
1362
+ "trial_params": null
1363
+ }
checkpoint-19000/checkpoint-19000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa82d77ba90da2851bd1daf5e058cad113e2cbd97d871fd0f05dc83a889bec90
3
+ size 4667
checkpoint-36000/checkpoint-36000/adapter_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model_name_or_path": "Qwen/Qwen-7B-Chat",
3
+ "bias": "none",
4
+ "fan_in_fan_out": false,
5
+ "inference_mode": false,
6
+ "init_lora_weights": true,
7
+ "lora_alpha": 32,
8
+ "lora_dropout": 0.1,
9
+ "modules_to_save": null,
10
+ "peft_type": "LORA",
11
+ "r": 128,
12
+ "target_modules": [
13
+ "c_attn",
14
+ "w1",
15
+ "w2"
16
+ ],
17
+ "task_type": "CAUSAL_LM"
18
+ }
checkpoint-36000/checkpoint-36000/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bab67b05b624de1c56365e82c79ab3c8ace9dd28801d3efb8f1e9ce0a38b335c
3
+ size 763431181
checkpoint-36000/checkpoint-36000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d9b052419c6c187cca03d7302c882b84891e199a1d271c70adc715f3e8c4297
3
+ size 1612910405
checkpoint-36000/checkpoint-36000/other_params.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:620d634331b2311e75a44f796a1820ae95b7367cc38d474ee2b08a2009b53cea
3
+ size 43009725
checkpoint-36000/checkpoint-36000/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19be55a974d66e618a62c99aad12bf414b1a2db7bdb018b9bf690e99a0ac6997
3
+ size 17655
checkpoint-36000/checkpoint-36000/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4314274739eb66598c0d8f4f3d81210a48191823471c29311a96056657dfa2d
3
+ size 17655
checkpoint-36000/checkpoint-36000/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:957461f4a4fb9e87f19013d5b573b6adcd2b19853d69d912677c2710338c4dda
3
+ size 17655
checkpoint-36000/checkpoint-36000/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60bb1bd9e59ce510ae83cee467449ce535999d236027cdb9b0ddf1404921c0e1
3
+ size 17655
checkpoint-36000/checkpoint-36000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b38b346d7bfcee081460f457625768a8763c99ab08a88a110bf23fd27b40e0e
3
+ size 627
checkpoint-36000/checkpoint-36000/trainer_state.json ADDED
@@ -0,0 +1,2553 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.9949019173223983,
5
+ "eval_steps": 500,
6
+ "global_step": 36000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.00554139421478444,
13
+ "grad_norm": 0.7744311690330505,
14
+ "learning_rate": 2.9916879086778235e-05,
15
+ "loss": 2.174,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.01108278842956888,
20
+ "grad_norm": 1.1890419721603394,
21
+ "learning_rate": 2.983375817355647e-05,
22
+ "loss": 1.7489,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.01662418264435332,
27
+ "grad_norm": 2.874730348587036,
28
+ "learning_rate": 2.97506372603347e-05,
29
+ "loss": 1.6593,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.02216557685913776,
34
+ "grad_norm": 0.4631105363368988,
35
+ "learning_rate": 2.9667516347112933e-05,
36
+ "loss": 1.5928,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.0277069710739222,
41
+ "grad_norm": 1.1168841123580933,
42
+ "learning_rate": 2.9584395433891167e-05,
43
+ "loss": 1.5494,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.03324836528870664,
48
+ "grad_norm": 0.8050689101219177,
49
+ "learning_rate": 2.95012745206694e-05,
50
+ "loss": 1.5041,
51
+ "step": 600
52
+ },
53
+ {
54
+ "epoch": 0.038789759503491075,
55
+ "grad_norm": 2.871495008468628,
56
+ "learning_rate": 2.9418153607447632e-05,
57
+ "loss": 1.4829,
58
+ "step": 700
59
+ },
60
+ {
61
+ "epoch": 0.04433115371827552,
62
+ "grad_norm": 0.6971802711486816,
63
+ "learning_rate": 2.9335032694225866e-05,
64
+ "loss": 1.4492,
65
+ "step": 800
66
+ },
67
+ {
68
+ "epoch": 0.04987254793305996,
69
+ "grad_norm": 1.1272006034851074,
70
+ "learning_rate": 2.92519117810041e-05,
71
+ "loss": 1.45,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 0.0554139421478444,
76
+ "grad_norm": 0.9110287427902222,
77
+ "learning_rate": 2.9168790867782334e-05,
78
+ "loss": 1.4553,
79
+ "step": 1000
80
+ },
81
+ {
82
+ "epoch": 0.060955336362628836,
83
+ "grad_norm": 2.7382473945617676,
84
+ "learning_rate": 2.9085669954560568e-05,
85
+ "loss": 1.4244,
86
+ "step": 1100
87
+ },
88
+ {
89
+ "epoch": 0.06649673057741327,
90
+ "grad_norm": 2.678687334060669,
91
+ "learning_rate": 2.9002549041338802e-05,
92
+ "loss": 1.4399,
93
+ "step": 1200
94
+ },
95
+ {
96
+ "epoch": 0.07203812479219772,
97
+ "grad_norm": 0.8868110775947571,
98
+ "learning_rate": 2.8919428128117036e-05,
99
+ "loss": 1.408,
100
+ "step": 1300
101
+ },
102
+ {
103
+ "epoch": 0.07757951900698215,
104
+ "grad_norm": 0.7350853085517883,
105
+ "learning_rate": 2.883630721489527e-05,
106
+ "loss": 1.4128,
107
+ "step": 1400
108
+ },
109
+ {
110
+ "epoch": 0.0831209132217666,
111
+ "grad_norm": 2.349001169204712,
112
+ "learning_rate": 2.8753186301673504e-05,
113
+ "loss": 1.4154,
114
+ "step": 1500
115
+ },
116
+ {
117
+ "epoch": 0.08866230743655104,
118
+ "grad_norm": 2.588608503341675,
119
+ "learning_rate": 2.8670065388451735e-05,
120
+ "loss": 1.4102,
121
+ "step": 1600
122
+ },
123
+ {
124
+ "epoch": 0.09420370165133547,
125
+ "grad_norm": 3.461966037750244,
126
+ "learning_rate": 2.858694447522997e-05,
127
+ "loss": 1.3899,
128
+ "step": 1700
129
+ },
130
+ {
131
+ "epoch": 0.09974509586611992,
132
+ "grad_norm": 0.9384596943855286,
133
+ "learning_rate": 2.8503823562008203e-05,
134
+ "loss": 1.3946,
135
+ "step": 1800
136
+ },
137
+ {
138
+ "epoch": 0.10528649008090435,
139
+ "grad_norm": 2.6963839530944824,
140
+ "learning_rate": 2.8420702648786437e-05,
141
+ "loss": 1.3625,
142
+ "step": 1900
143
+ },
144
+ {
145
+ "epoch": 0.1108278842956888,
146
+ "grad_norm": 1.2156896591186523,
147
+ "learning_rate": 2.8337581735564668e-05,
148
+ "loss": 1.3998,
149
+ "step": 2000
150
+ },
151
+ {
152
+ "epoch": 0.11636927851047324,
153
+ "grad_norm": 0.9613581299781799,
154
+ "learning_rate": 2.82544608223429e-05,
155
+ "loss": 1.3672,
156
+ "step": 2100
157
+ },
158
+ {
159
+ "epoch": 0.12191067272525767,
160
+ "grad_norm": 0.8994789719581604,
161
+ "learning_rate": 2.8171339909121136e-05,
162
+ "loss": 1.3587,
163
+ "step": 2200
164
+ },
165
+ {
166
+ "epoch": 0.1274520669400421,
167
+ "grad_norm": 1.7463105916976929,
168
+ "learning_rate": 2.808821899589937e-05,
169
+ "loss": 1.3688,
170
+ "step": 2300
171
+ },
172
+ {
173
+ "epoch": 0.13299346115482655,
174
+ "grad_norm": 0.7639077305793762,
175
+ "learning_rate": 2.8005098082677604e-05,
176
+ "loss": 1.3798,
177
+ "step": 2400
178
+ },
179
+ {
180
+ "epoch": 0.138534855369611,
181
+ "grad_norm": 1.313306450843811,
182
+ "learning_rate": 2.7921977169455834e-05,
183
+ "loss": 1.3713,
184
+ "step": 2500
185
+ },
186
+ {
187
+ "epoch": 0.14407624958439544,
188
+ "grad_norm": 1.796030044555664,
189
+ "learning_rate": 2.783885625623407e-05,
190
+ "loss": 1.3589,
191
+ "step": 2600
192
+ },
193
+ {
194
+ "epoch": 0.14961764379917988,
195
+ "grad_norm": 0.40334078669548035,
196
+ "learning_rate": 2.7755735343012302e-05,
197
+ "loss": 1.3527,
198
+ "step": 2700
199
+ },
200
+ {
201
+ "epoch": 0.1551590380139643,
202
+ "grad_norm": 0.7727888822555542,
203
+ "learning_rate": 2.7672614429790536e-05,
204
+ "loss": 1.338,
205
+ "step": 2800
206
+ },
207
+ {
208
+ "epoch": 0.16070043222874875,
209
+ "grad_norm": 0.8815018534660339,
210
+ "learning_rate": 2.7589493516568767e-05,
211
+ "loss": 1.3638,
212
+ "step": 2900
213
+ },
214
+ {
215
+ "epoch": 0.1662418264435332,
216
+ "grad_norm": 1.1675957441329956,
217
+ "learning_rate": 2.7506372603347e-05,
218
+ "loss": 1.3462,
219
+ "step": 3000
220
+ },
221
+ {
222
+ "epoch": 0.17178322065831764,
223
+ "grad_norm": 0.9449867606163025,
224
+ "learning_rate": 2.7423251690125235e-05,
225
+ "loss": 1.3286,
226
+ "step": 3100
227
+ },
228
+ {
229
+ "epoch": 0.17732461487310208,
230
+ "grad_norm": 0.4934576749801636,
231
+ "learning_rate": 2.734013077690347e-05,
232
+ "loss": 1.3326,
233
+ "step": 3200
234
+ },
235
+ {
236
+ "epoch": 0.1828660090878865,
237
+ "grad_norm": 1.0242582559585571,
238
+ "learning_rate": 2.7257009863681703e-05,
239
+ "loss": 1.3423,
240
+ "step": 3300
241
+ },
242
+ {
243
+ "epoch": 0.18840740330267094,
244
+ "grad_norm": 1.253219723701477,
245
+ "learning_rate": 2.7173888950459934e-05,
246
+ "loss": 1.3473,
247
+ "step": 3400
248
+ },
249
+ {
250
+ "epoch": 0.1939487975174554,
251
+ "grad_norm": 2.4441049098968506,
252
+ "learning_rate": 2.7090768037238168e-05,
253
+ "loss": 1.3321,
254
+ "step": 3500
255
+ },
256
+ {
257
+ "epoch": 0.19949019173223984,
258
+ "grad_norm": 1.119807481765747,
259
+ "learning_rate": 2.7007647124016405e-05,
260
+ "loss": 1.3282,
261
+ "step": 3600
262
+ },
263
+ {
264
+ "epoch": 0.20503158594702428,
265
+ "grad_norm": 0.5911545157432556,
266
+ "learning_rate": 2.692452621079464e-05,
267
+ "loss": 1.329,
268
+ "step": 3700
269
+ },
270
+ {
271
+ "epoch": 0.2105729801618087,
272
+ "grad_norm": 1.959672451019287,
273
+ "learning_rate": 2.684140529757287e-05,
274
+ "loss": 1.3322,
275
+ "step": 3800
276
+ },
277
+ {
278
+ "epoch": 0.21611437437659314,
279
+ "grad_norm": 0.9169633984565735,
280
+ "learning_rate": 2.6758284384351104e-05,
281
+ "loss": 1.3248,
282
+ "step": 3900
283
+ },
284
+ {
285
+ "epoch": 0.2216557685913776,
286
+ "grad_norm": 0.8207225203514099,
287
+ "learning_rate": 2.6675163471129338e-05,
288
+ "loss": 1.3303,
289
+ "step": 4000
290
+ },
291
+ {
292
+ "epoch": 0.22719716280616203,
293
+ "grad_norm": 1.0271358489990234,
294
+ "learning_rate": 2.6592042557907572e-05,
295
+ "loss": 1.3231,
296
+ "step": 4100
297
+ },
298
+ {
299
+ "epoch": 0.23273855702094648,
300
+ "grad_norm": 1.2483460903167725,
301
+ "learning_rate": 2.6508921644685806e-05,
302
+ "loss": 1.3234,
303
+ "step": 4200
304
+ },
305
+ {
306
+ "epoch": 0.2382799512357309,
307
+ "grad_norm": 0.8641893863677979,
308
+ "learning_rate": 2.6425800731464037e-05,
309
+ "loss": 1.2856,
310
+ "step": 4300
311
+ },
312
+ {
313
+ "epoch": 0.24382134545051534,
314
+ "grad_norm": 1.6275527477264404,
315
+ "learning_rate": 2.634267981824227e-05,
316
+ "loss": 1.3039,
317
+ "step": 4400
318
+ },
319
+ {
320
+ "epoch": 0.2493627396652998,
321
+ "grad_norm": 2.3479559421539307,
322
+ "learning_rate": 2.6259558905020505e-05,
323
+ "loss": 1.331,
324
+ "step": 4500
325
+ },
326
+ {
327
+ "epoch": 0.2549041338800842,
328
+ "grad_norm": 2.3728585243225098,
329
+ "learning_rate": 2.617643799179874e-05,
330
+ "loss": 1.3115,
331
+ "step": 4600
332
+ },
333
+ {
334
+ "epoch": 0.26044552809486865,
335
+ "grad_norm": 1.7304351329803467,
336
+ "learning_rate": 2.609331707857697e-05,
337
+ "loss": 1.32,
338
+ "step": 4700
339
+ },
340
+ {
341
+ "epoch": 0.2659869223096531,
342
+ "grad_norm": 3.1707870960235596,
343
+ "learning_rate": 2.6010196165355203e-05,
344
+ "loss": 1.3166,
345
+ "step": 4800
346
+ },
347
+ {
348
+ "epoch": 0.27152831652443754,
349
+ "grad_norm": 0.6431360840797424,
350
+ "learning_rate": 2.5927075252133437e-05,
351
+ "loss": 1.3117,
352
+ "step": 4900
353
+ },
354
+ {
355
+ "epoch": 0.277069710739222,
356
+ "grad_norm": 0.7438284754753113,
357
+ "learning_rate": 2.584395433891167e-05,
358
+ "loss": 1.3049,
359
+ "step": 5000
360
+ },
361
+ {
362
+ "epoch": 0.28261110495400643,
363
+ "grad_norm": 0.8397496342658997,
364
+ "learning_rate": 2.5760833425689905e-05,
365
+ "loss": 1.3013,
366
+ "step": 5100
367
+ },
368
+ {
369
+ "epoch": 0.2881524991687909,
370
+ "grad_norm": 1.3787566423416138,
371
+ "learning_rate": 2.5677712512468136e-05,
372
+ "loss": 1.32,
373
+ "step": 5200
374
+ },
375
+ {
376
+ "epoch": 0.2936938933835753,
377
+ "grad_norm": 0.6342838406562805,
378
+ "learning_rate": 2.559459159924637e-05,
379
+ "loss": 1.3265,
380
+ "step": 5300
381
+ },
382
+ {
383
+ "epoch": 0.29923528759835977,
384
+ "grad_norm": 1.0924850702285767,
385
+ "learning_rate": 2.5511470686024604e-05,
386
+ "loss": 1.3224,
387
+ "step": 5400
388
+ },
389
+ {
390
+ "epoch": 0.3047766818131442,
391
+ "grad_norm": 3.104914665222168,
392
+ "learning_rate": 2.5428349772802838e-05,
393
+ "loss": 1.2897,
394
+ "step": 5500
395
+ },
396
+ {
397
+ "epoch": 0.3103180760279286,
398
+ "grad_norm": 1.219292163848877,
399
+ "learning_rate": 2.534522885958107e-05,
400
+ "loss": 1.3013,
401
+ "step": 5600
402
+ },
403
+ {
404
+ "epoch": 0.31585947024271305,
405
+ "grad_norm": 0.743425726890564,
406
+ "learning_rate": 2.5262107946359303e-05,
407
+ "loss": 1.3069,
408
+ "step": 5700
409
+ },
410
+ {
411
+ "epoch": 0.3214008644574975,
412
+ "grad_norm": 3.0548131465911865,
413
+ "learning_rate": 2.5178987033137537e-05,
414
+ "loss": 1.3006,
415
+ "step": 5800
416
+ },
417
+ {
418
+ "epoch": 0.32694225867228194,
419
+ "grad_norm": 2.6230878829956055,
420
+ "learning_rate": 2.5095866119915774e-05,
421
+ "loss": 1.2954,
422
+ "step": 5900
423
+ },
424
+ {
425
+ "epoch": 0.3324836528870664,
426
+ "grad_norm": 1.5894700288772583,
427
+ "learning_rate": 2.5012745206694005e-05,
428
+ "loss": 1.2918,
429
+ "step": 6000
430
+ },
431
+ {
432
+ "epoch": 0.33802504710185083,
433
+ "grad_norm": 2.298469066619873,
434
+ "learning_rate": 2.492962429347224e-05,
435
+ "loss": 1.2974,
436
+ "step": 6100
437
+ },
438
+ {
439
+ "epoch": 0.3435664413166353,
440
+ "grad_norm": 2.3570539951324463,
441
+ "learning_rate": 2.4846503380250473e-05,
442
+ "loss": 1.2943,
443
+ "step": 6200
444
+ },
445
+ {
446
+ "epoch": 0.3491078355314197,
447
+ "grad_norm": 0.7656165361404419,
448
+ "learning_rate": 2.4763382467028707e-05,
449
+ "loss": 1.2774,
450
+ "step": 6300
451
+ },
452
+ {
453
+ "epoch": 0.35464922974620416,
454
+ "grad_norm": 0.9744777679443359,
455
+ "learning_rate": 2.468026155380694e-05,
456
+ "loss": 1.3062,
457
+ "step": 6400
458
+ },
459
+ {
460
+ "epoch": 0.3601906239609886,
461
+ "grad_norm": 2.929172992706299,
462
+ "learning_rate": 2.459714064058517e-05,
463
+ "loss": 1.2871,
464
+ "step": 6500
465
+ },
466
+ {
467
+ "epoch": 0.365732018175773,
468
+ "grad_norm": 1.1587573289871216,
469
+ "learning_rate": 2.4514019727363405e-05,
470
+ "loss": 1.3037,
471
+ "step": 6600
472
+ },
473
+ {
474
+ "epoch": 0.37127341239055744,
475
+ "grad_norm": 1.055335521697998,
476
+ "learning_rate": 2.443089881414164e-05,
477
+ "loss": 1.2823,
478
+ "step": 6700
479
+ },
480
+ {
481
+ "epoch": 0.3768148066053419,
482
+ "grad_norm": 2.3376619815826416,
483
+ "learning_rate": 2.4347777900919873e-05,
484
+ "loss": 1.2854,
485
+ "step": 6800
486
+ },
487
+ {
488
+ "epoch": 0.38235620082012634,
489
+ "grad_norm": 0.94229656457901,
490
+ "learning_rate": 2.4264656987698104e-05,
491
+ "loss": 1.3007,
492
+ "step": 6900
493
+ },
494
+ {
495
+ "epoch": 0.3878975950349108,
496
+ "grad_norm": 1.2927848100662231,
497
+ "learning_rate": 2.4181536074476338e-05,
498
+ "loss": 1.278,
499
+ "step": 7000
500
+ },
501
+ {
502
+ "epoch": 0.3934389892496952,
503
+ "grad_norm": 0.6448485255241394,
504
+ "learning_rate": 2.4098415161254572e-05,
505
+ "loss": 1.275,
506
+ "step": 7100
507
+ },
508
+ {
509
+ "epoch": 0.39898038346447967,
510
+ "grad_norm": 1.4731674194335938,
511
+ "learning_rate": 2.4015294248032806e-05,
512
+ "loss": 1.2737,
513
+ "step": 7200
514
+ },
515
+ {
516
+ "epoch": 0.4045217776792641,
517
+ "grad_norm": 2.395692825317383,
518
+ "learning_rate": 2.393217333481104e-05,
519
+ "loss": 1.285,
520
+ "step": 7300
521
+ },
522
+ {
523
+ "epoch": 0.41006317189404856,
524
+ "grad_norm": 0.8327000737190247,
525
+ "learning_rate": 2.384905242158927e-05,
526
+ "loss": 1.2514,
527
+ "step": 7400
528
+ },
529
+ {
530
+ "epoch": 0.415604566108833,
531
+ "grad_norm": 0.5383670330047607,
532
+ "learning_rate": 2.3765931508367505e-05,
533
+ "loss": 1.294,
534
+ "step": 7500
535
+ },
536
+ {
537
+ "epoch": 0.4211459603236174,
538
+ "grad_norm": 0.7433611154556274,
539
+ "learning_rate": 2.368281059514574e-05,
540
+ "loss": 1.2966,
541
+ "step": 7600
542
+ },
543
+ {
544
+ "epoch": 0.42668735453840184,
545
+ "grad_norm": 1.4880763292312622,
546
+ "learning_rate": 2.3599689681923973e-05,
547
+ "loss": 1.3076,
548
+ "step": 7700
549
+ },
550
+ {
551
+ "epoch": 0.4322287487531863,
552
+ "grad_norm": 1.2045841217041016,
553
+ "learning_rate": 2.3516568768702204e-05,
554
+ "loss": 1.2598,
555
+ "step": 7800
556
+ },
557
+ {
558
+ "epoch": 0.43777014296797073,
559
+ "grad_norm": 2.087782621383667,
560
+ "learning_rate": 2.3433447855480438e-05,
561
+ "loss": 1.2832,
562
+ "step": 7900
563
+ },
564
+ {
565
+ "epoch": 0.4433115371827552,
566
+ "grad_norm": 0.9906471371650696,
567
+ "learning_rate": 2.335032694225867e-05,
568
+ "loss": 1.2723,
569
+ "step": 8000
570
+ },
571
+ {
572
+ "epoch": 0.4488529313975396,
573
+ "grad_norm": 2.195661783218384,
574
+ "learning_rate": 2.3267206029036906e-05,
575
+ "loss": 1.2707,
576
+ "step": 8100
577
+ },
578
+ {
579
+ "epoch": 0.45439432561232407,
580
+ "grad_norm": 2.5563039779663086,
581
+ "learning_rate": 2.318408511581514e-05,
582
+ "loss": 1.2905,
583
+ "step": 8200
584
+ },
585
+ {
586
+ "epoch": 0.4599357198271085,
587
+ "grad_norm": 0.7435129284858704,
588
+ "learning_rate": 2.3100964202593374e-05,
589
+ "loss": 1.2814,
590
+ "step": 8300
591
+ },
592
+ {
593
+ "epoch": 0.46547711404189296,
594
+ "grad_norm": 0.8438817262649536,
595
+ "learning_rate": 2.3017843289371608e-05,
596
+ "loss": 1.2745,
597
+ "step": 8400
598
+ },
599
+ {
600
+ "epoch": 0.4710185082566774,
601
+ "grad_norm": 1.848270058631897,
602
+ "learning_rate": 2.293472237614984e-05,
603
+ "loss": 1.2735,
604
+ "step": 8500
605
+ },
606
+ {
607
+ "epoch": 0.4765599024714618,
608
+ "grad_norm": 0.9101561307907104,
609
+ "learning_rate": 2.2851601462928076e-05,
610
+ "loss": 1.2576,
611
+ "step": 8600
612
+ },
613
+ {
614
+ "epoch": 0.48210129668624624,
615
+ "grad_norm": 1.2922989130020142,
616
+ "learning_rate": 2.2768480549706306e-05,
617
+ "loss": 1.2824,
618
+ "step": 8700
619
+ },
620
+ {
621
+ "epoch": 0.4876426909010307,
622
+ "grad_norm": 0.7475651502609253,
623
+ "learning_rate": 2.268535963648454e-05,
624
+ "loss": 1.2877,
625
+ "step": 8800
626
+ },
627
+ {
628
+ "epoch": 0.49318408511581513,
629
+ "grad_norm": 1.3247780799865723,
630
+ "learning_rate": 2.2602238723262774e-05,
631
+ "loss": 1.278,
632
+ "step": 8900
633
+ },
634
+ {
635
+ "epoch": 0.4987254793305996,
636
+ "grad_norm": 0.9011294841766357,
637
+ "learning_rate": 2.251911781004101e-05,
638
+ "loss": 1.2776,
639
+ "step": 9000
640
+ },
641
+ {
642
+ "epoch": 0.504266873545384,
643
+ "grad_norm": 3.1133978366851807,
644
+ "learning_rate": 2.2435996896819242e-05,
645
+ "loss": 1.2732,
646
+ "step": 9100
647
+ },
648
+ {
649
+ "epoch": 0.5098082677601684,
650
+ "grad_norm": 0.9149213433265686,
651
+ "learning_rate": 2.2352875983597473e-05,
652
+ "loss": 1.2806,
653
+ "step": 9200
654
+ },
655
+ {
656
+ "epoch": 0.5153496619749529,
657
+ "grad_norm": 0.8003007769584656,
658
+ "learning_rate": 2.2269755070375707e-05,
659
+ "loss": 1.2795,
660
+ "step": 9300
661
+ },
662
+ {
663
+ "epoch": 0.5208910561897373,
664
+ "grad_norm": 0.8392866253852844,
665
+ "learning_rate": 2.218663415715394e-05,
666
+ "loss": 1.2715,
667
+ "step": 9400
668
+ },
669
+ {
670
+ "epoch": 0.5264324504045218,
671
+ "grad_norm": 0.7663789391517639,
672
+ "learning_rate": 2.2103513243932175e-05,
673
+ "loss": 1.2524,
674
+ "step": 9500
675
+ },
676
+ {
677
+ "epoch": 0.5319738446193062,
678
+ "grad_norm": 2.0740444660186768,
679
+ "learning_rate": 2.2020392330710406e-05,
680
+ "loss": 1.2687,
681
+ "step": 9600
682
+ },
683
+ {
684
+ "epoch": 0.5375152388340907,
685
+ "grad_norm": 3.680366277694702,
686
+ "learning_rate": 2.193727141748864e-05,
687
+ "loss": 1.2903,
688
+ "step": 9700
689
+ },
690
+ {
691
+ "epoch": 0.5430566330488751,
692
+ "grad_norm": 0.9677708745002747,
693
+ "learning_rate": 2.1854150504266874e-05,
694
+ "loss": 1.2655,
695
+ "step": 9800
696
+ },
697
+ {
698
+ "epoch": 0.5485980272636596,
699
+ "grad_norm": 1.019551157951355,
700
+ "learning_rate": 2.1771029591045108e-05,
701
+ "loss": 1.2664,
702
+ "step": 9900
703
+ },
704
+ {
705
+ "epoch": 0.554139421478444,
706
+ "grad_norm": 3.291146993637085,
707
+ "learning_rate": 2.168790867782334e-05,
708
+ "loss": 1.2716,
709
+ "step": 10000
710
+ },
711
+ {
712
+ "epoch": 0.5596808156932284,
713
+ "grad_norm": 2.067105770111084,
714
+ "learning_rate": 2.1604787764601572e-05,
715
+ "loss": 1.2703,
716
+ "step": 10100
717
+ },
718
+ {
719
+ "epoch": 0.5652222099080129,
720
+ "grad_norm": 0.6159241199493408,
721
+ "learning_rate": 2.1521666851379806e-05,
722
+ "loss": 1.269,
723
+ "step": 10200
724
+ },
725
+ {
726
+ "epoch": 0.5707636041227973,
727
+ "grad_norm": 0.8273797631263733,
728
+ "learning_rate": 2.143854593815804e-05,
729
+ "loss": 1.2679,
730
+ "step": 10300
731
+ },
732
+ {
733
+ "epoch": 0.5763049983375818,
734
+ "grad_norm": 1.0053974390029907,
735
+ "learning_rate": 2.1355425024936274e-05,
736
+ "loss": 1.2615,
737
+ "step": 10400
738
+ },
739
+ {
740
+ "epoch": 0.5818463925523661,
741
+ "grad_norm": 0.7567972540855408,
742
+ "learning_rate": 2.1272304111714505e-05,
743
+ "loss": 1.2706,
744
+ "step": 10500
745
+ },
746
+ {
747
+ "epoch": 0.5873877867671506,
748
+ "grad_norm": 1.5227950811386108,
749
+ "learning_rate": 2.1189183198492743e-05,
750
+ "loss": 1.2853,
751
+ "step": 10600
752
+ },
753
+ {
754
+ "epoch": 0.592929180981935,
755
+ "grad_norm": 0.7776534557342529,
756
+ "learning_rate": 2.1106062285270977e-05,
757
+ "loss": 1.267,
758
+ "step": 10700
759
+ },
760
+ {
761
+ "epoch": 0.5984705751967195,
762
+ "grad_norm": 2.273895502090454,
763
+ "learning_rate": 2.102294137204921e-05,
764
+ "loss": 1.2587,
765
+ "step": 10800
766
+ },
767
+ {
768
+ "epoch": 0.6040119694115039,
769
+ "grad_norm": 2.182058334350586,
770
+ "learning_rate": 2.093982045882744e-05,
771
+ "loss": 1.2676,
772
+ "step": 10900
773
+ },
774
+ {
775
+ "epoch": 0.6095533636262884,
776
+ "grad_norm": 2.346633195877075,
777
+ "learning_rate": 2.0856699545605675e-05,
778
+ "loss": 1.2512,
779
+ "step": 11000
780
+ },
781
+ {
782
+ "epoch": 0.6150947578410728,
783
+ "grad_norm": 2.9466536045074463,
784
+ "learning_rate": 2.077357863238391e-05,
785
+ "loss": 1.2636,
786
+ "step": 11100
787
+ },
788
+ {
789
+ "epoch": 0.6206361520558572,
790
+ "grad_norm": 2.7352640628814697,
791
+ "learning_rate": 2.0690457719162143e-05,
792
+ "loss": 1.2677,
793
+ "step": 11200
794
+ },
795
+ {
796
+ "epoch": 0.6261775462706417,
797
+ "grad_norm": 0.4942292869091034,
798
+ "learning_rate": 2.0607336805940377e-05,
799
+ "loss": 1.2512,
800
+ "step": 11300
801
+ },
802
+ {
803
+ "epoch": 0.6317189404854261,
804
+ "grad_norm": 1.3134087324142456,
805
+ "learning_rate": 2.0524215892718608e-05,
806
+ "loss": 1.2563,
807
+ "step": 11400
808
+ },
809
+ {
810
+ "epoch": 0.6372603347002106,
811
+ "grad_norm": 1.0919684171676636,
812
+ "learning_rate": 2.0441094979496842e-05,
813
+ "loss": 1.2402,
814
+ "step": 11500
815
+ },
816
+ {
817
+ "epoch": 0.642801728914995,
818
+ "grad_norm": 1.3876844644546509,
819
+ "learning_rate": 2.0357974066275076e-05,
820
+ "loss": 1.2788,
821
+ "step": 11600
822
+ },
823
+ {
824
+ "epoch": 0.6483431231297795,
825
+ "grad_norm": 2.135866165161133,
826
+ "learning_rate": 2.027485315305331e-05,
827
+ "loss": 1.2556,
828
+ "step": 11700
829
+ },
830
+ {
831
+ "epoch": 0.6538845173445639,
832
+ "grad_norm": 3.400996685028076,
833
+ "learning_rate": 2.019173223983154e-05,
834
+ "loss": 1.2706,
835
+ "step": 11800
836
+ },
837
+ {
838
+ "epoch": 0.6594259115593484,
839
+ "grad_norm": 1.1860913038253784,
840
+ "learning_rate": 2.0108611326609775e-05,
841
+ "loss": 1.2538,
842
+ "step": 11900
843
+ },
844
+ {
845
+ "epoch": 0.6649673057741328,
846
+ "grad_norm": 1.044102430343628,
847
+ "learning_rate": 2.002549041338801e-05,
848
+ "loss": 1.2573,
849
+ "step": 12000
850
+ },
851
+ {
852
+ "epoch": 0.6705086999889172,
853
+ "grad_norm": 1.0775806903839111,
854
+ "learning_rate": 1.9942369500166243e-05,
855
+ "loss": 1.2591,
856
+ "step": 12100
857
+ },
858
+ {
859
+ "epoch": 0.6760500942037017,
860
+ "grad_norm": 0.7967768311500549,
861
+ "learning_rate": 1.9859248586944477e-05,
862
+ "loss": 1.2394,
863
+ "step": 12200
864
+ },
865
+ {
866
+ "epoch": 0.681591488418486,
867
+ "grad_norm": 2.224622964859009,
868
+ "learning_rate": 1.9776127673722707e-05,
869
+ "loss": 1.2458,
870
+ "step": 12300
871
+ },
872
+ {
873
+ "epoch": 0.6871328826332705,
874
+ "grad_norm": 2.769935131072998,
875
+ "learning_rate": 1.969300676050094e-05,
876
+ "loss": 1.2665,
877
+ "step": 12400
878
+ },
879
+ {
880
+ "epoch": 0.6926742768480549,
881
+ "grad_norm": 1.0014837980270386,
882
+ "learning_rate": 1.9609885847279175e-05,
883
+ "loss": 1.2384,
884
+ "step": 12500
885
+ },
886
+ {
887
+ "epoch": 0.6982156710628394,
888
+ "grad_norm": 0.8904783725738525,
889
+ "learning_rate": 1.952676493405741e-05,
890
+ "loss": 1.2607,
891
+ "step": 12600
892
+ },
893
+ {
894
+ "epoch": 0.7037570652776238,
895
+ "grad_norm": 0.9549648761749268,
896
+ "learning_rate": 1.944364402083564e-05,
897
+ "loss": 1.2411,
898
+ "step": 12700
899
+ },
900
+ {
901
+ "epoch": 0.7092984594924083,
902
+ "grad_norm": 0.44161495566368103,
903
+ "learning_rate": 1.9360523107613874e-05,
904
+ "loss": 1.2645,
905
+ "step": 12800
906
+ },
907
+ {
908
+ "epoch": 0.7148398537071927,
909
+ "grad_norm": 1.002912163734436,
910
+ "learning_rate": 1.9277402194392108e-05,
911
+ "loss": 1.263,
912
+ "step": 12900
913
+ },
914
+ {
915
+ "epoch": 0.7203812479219772,
916
+ "grad_norm": 0.9343509078025818,
917
+ "learning_rate": 1.9194281281170345e-05,
918
+ "loss": 1.2663,
919
+ "step": 13000
920
+ },
921
+ {
922
+ "epoch": 0.7259226421367616,
923
+ "grad_norm": 0.7143359184265137,
924
+ "learning_rate": 1.9111160367948576e-05,
925
+ "loss": 1.2509,
926
+ "step": 13100
927
+ },
928
+ {
929
+ "epoch": 0.731464036351546,
930
+ "grad_norm": 3.670598030090332,
931
+ "learning_rate": 1.902803945472681e-05,
932
+ "loss": 1.2424,
933
+ "step": 13200
934
+ },
935
+ {
936
+ "epoch": 0.7370054305663305,
937
+ "grad_norm": 0.9394696354866028,
938
+ "learning_rate": 1.8944918541505044e-05,
939
+ "loss": 1.2302,
940
+ "step": 13300
941
+ },
942
+ {
943
+ "epoch": 0.7425468247811149,
944
+ "grad_norm": 2.5993099212646484,
945
+ "learning_rate": 1.8861797628283278e-05,
946
+ "loss": 1.2454,
947
+ "step": 13400
948
+ },
949
+ {
950
+ "epoch": 0.7480882189958994,
951
+ "grad_norm": 1.5783851146697998,
952
+ "learning_rate": 1.8778676715061512e-05,
953
+ "loss": 1.2613,
954
+ "step": 13500
955
+ },
956
+ {
957
+ "epoch": 0.7536296132106838,
958
+ "grad_norm": 0.6382520794868469,
959
+ "learning_rate": 1.8695555801839743e-05,
960
+ "loss": 1.2583,
961
+ "step": 13600
962
+ },
963
+ {
964
+ "epoch": 0.7591710074254683,
965
+ "grad_norm": 0.6521400809288025,
966
+ "learning_rate": 1.8612434888617977e-05,
967
+ "loss": 1.2253,
968
+ "step": 13700
969
+ },
970
+ {
971
+ "epoch": 0.7647124016402527,
972
+ "grad_norm": 2.8146674633026123,
973
+ "learning_rate": 1.852931397539621e-05,
974
+ "loss": 1.2577,
975
+ "step": 13800
976
+ },
977
+ {
978
+ "epoch": 0.7702537958550372,
979
+ "grad_norm": 2.7363874912261963,
980
+ "learning_rate": 1.8446193062174445e-05,
981
+ "loss": 1.2403,
982
+ "step": 13900
983
+ },
984
+ {
985
+ "epoch": 0.7757951900698216,
986
+ "grad_norm": 2.4337146282196045,
987
+ "learning_rate": 1.8363072148952676e-05,
988
+ "loss": 1.2522,
989
+ "step": 14000
990
+ },
991
+ {
992
+ "epoch": 0.7813365842846061,
993
+ "grad_norm": 0.9723544716835022,
994
+ "learning_rate": 1.827995123573091e-05,
995
+ "loss": 1.2507,
996
+ "step": 14100
997
+ },
998
+ {
999
+ "epoch": 0.7868779784993905,
1000
+ "grad_norm": 1.1492934226989746,
1001
+ "learning_rate": 1.8196830322509144e-05,
1002
+ "loss": 1.23,
1003
+ "step": 14200
1004
+ },
1005
+ {
1006
+ "epoch": 0.7924193727141748,
1007
+ "grad_norm": 0.8592559099197388,
1008
+ "learning_rate": 1.8113709409287378e-05,
1009
+ "loss": 1.2508,
1010
+ "step": 14300
1011
+ },
1012
+ {
1013
+ "epoch": 0.7979607669289593,
1014
+ "grad_norm": 1.035599708557129,
1015
+ "learning_rate": 1.803058849606561e-05,
1016
+ "loss": 1.2465,
1017
+ "step": 14400
1018
+ },
1019
+ {
1020
+ "epoch": 0.8035021611437437,
1021
+ "grad_norm": 0.6385583281517029,
1022
+ "learning_rate": 1.7947467582843842e-05,
1023
+ "loss": 1.2476,
1024
+ "step": 14500
1025
+ },
1026
+ {
1027
+ "epoch": 0.8090435553585282,
1028
+ "grad_norm": 1.695744276046753,
1029
+ "learning_rate": 1.7864346669622076e-05,
1030
+ "loss": 1.239,
1031
+ "step": 14600
1032
+ },
1033
+ {
1034
+ "epoch": 0.8145849495733126,
1035
+ "grad_norm": 2.5061330795288086,
1036
+ "learning_rate": 1.778122575640031e-05,
1037
+ "loss": 1.2527,
1038
+ "step": 14700
1039
+ },
1040
+ {
1041
+ "epoch": 0.8201263437880971,
1042
+ "grad_norm": 1.0300838947296143,
1043
+ "learning_rate": 1.7698104843178544e-05,
1044
+ "loss": 1.2406,
1045
+ "step": 14800
1046
+ },
1047
+ {
1048
+ "epoch": 0.8256677380028815,
1049
+ "grad_norm": 0.8938758969306946,
1050
+ "learning_rate": 1.7614983929956775e-05,
1051
+ "loss": 1.236,
1052
+ "step": 14900
1053
+ },
1054
+ {
1055
+ "epoch": 0.831209132217666,
1056
+ "grad_norm": 0.9264464974403381,
1057
+ "learning_rate": 1.753186301673501e-05,
1058
+ "loss": 1.2439,
1059
+ "step": 15000
1060
+ },
1061
+ {
1062
+ "epoch": 0.8367505264324504,
1063
+ "grad_norm": 1.1339832544326782,
1064
+ "learning_rate": 1.7448742103513243e-05,
1065
+ "loss": 1.2358,
1066
+ "step": 15100
1067
+ },
1068
+ {
1069
+ "epoch": 0.8422919206472348,
1070
+ "grad_norm": 1.239995002746582,
1071
+ "learning_rate": 1.7365621190291477e-05,
1072
+ "loss": 1.2505,
1073
+ "step": 15200
1074
+ },
1075
+ {
1076
+ "epoch": 0.8478333148620193,
1077
+ "grad_norm": 2.820875644683838,
1078
+ "learning_rate": 1.7282500277069714e-05,
1079
+ "loss": 1.2502,
1080
+ "step": 15300
1081
+ },
1082
+ {
1083
+ "epoch": 0.8533747090768037,
1084
+ "grad_norm": 0.5509181022644043,
1085
+ "learning_rate": 1.7199379363847945e-05,
1086
+ "loss": 1.2285,
1087
+ "step": 15400
1088
+ },
1089
+ {
1090
+ "epoch": 0.8589161032915882,
1091
+ "grad_norm": 2.2472527027130127,
1092
+ "learning_rate": 1.711625845062618e-05,
1093
+ "loss": 1.2443,
1094
+ "step": 15500
1095
+ },
1096
+ {
1097
+ "epoch": 0.8644574975063726,
1098
+ "grad_norm": 0.6996986269950867,
1099
+ "learning_rate": 1.7033137537404413e-05,
1100
+ "loss": 1.2415,
1101
+ "step": 15600
1102
+ },
1103
+ {
1104
+ "epoch": 0.8699988917211571,
1105
+ "grad_norm": 1.9218405485153198,
1106
+ "learning_rate": 1.6950016624182647e-05,
1107
+ "loss": 1.2414,
1108
+ "step": 15700
1109
+ },
1110
+ {
1111
+ "epoch": 0.8755402859359415,
1112
+ "grad_norm": 1.069169521331787,
1113
+ "learning_rate": 1.6866895710960878e-05,
1114
+ "loss": 1.2365,
1115
+ "step": 15800
1116
+ },
1117
+ {
1118
+ "epoch": 0.881081680150726,
1119
+ "grad_norm": 2.6105518341064453,
1120
+ "learning_rate": 1.6783774797739112e-05,
1121
+ "loss": 1.2423,
1122
+ "step": 15900
1123
+ },
1124
+ {
1125
+ "epoch": 0.8866230743655104,
1126
+ "grad_norm": 3.099567413330078,
1127
+ "learning_rate": 1.6700653884517346e-05,
1128
+ "loss": 1.2474,
1129
+ "step": 16000
1130
+ },
1131
+ {
1132
+ "epoch": 0.8921644685802949,
1133
+ "grad_norm": 0.6464875340461731,
1134
+ "learning_rate": 1.661753297129558e-05,
1135
+ "loss": 1.2378,
1136
+ "step": 16100
1137
+ },
1138
+ {
1139
+ "epoch": 0.8977058627950792,
1140
+ "grad_norm": 1.130702018737793,
1141
+ "learning_rate": 1.6534412058073814e-05,
1142
+ "loss": 1.2355,
1143
+ "step": 16200
1144
+ },
1145
+ {
1146
+ "epoch": 0.9032472570098636,
1147
+ "grad_norm": 2.6416752338409424,
1148
+ "learning_rate": 1.6451291144852044e-05,
1149
+ "loss": 1.2399,
1150
+ "step": 16300
1151
+ },
1152
+ {
1153
+ "epoch": 0.9087886512246481,
1154
+ "grad_norm": 0.5979084372520447,
1155
+ "learning_rate": 1.636817023163028e-05,
1156
+ "loss": 1.2297,
1157
+ "step": 16400
1158
+ },
1159
+ {
1160
+ "epoch": 0.9143300454394325,
1161
+ "grad_norm": 2.5381157398223877,
1162
+ "learning_rate": 1.6285049318408512e-05,
1163
+ "loss": 1.2315,
1164
+ "step": 16500
1165
+ },
1166
+ {
1167
+ "epoch": 0.919871439654217,
1168
+ "grad_norm": 1.3977634906768799,
1169
+ "learning_rate": 1.6201928405186746e-05,
1170
+ "loss": 1.2441,
1171
+ "step": 16600
1172
+ },
1173
+ {
1174
+ "epoch": 0.9254128338690014,
1175
+ "grad_norm": 0.5126473307609558,
1176
+ "learning_rate": 1.6118807491964977e-05,
1177
+ "loss": 1.2377,
1178
+ "step": 16700
1179
+ },
1180
+ {
1181
+ "epoch": 0.9309542280837859,
1182
+ "grad_norm": 1.1525436639785767,
1183
+ "learning_rate": 1.603568657874321e-05,
1184
+ "loss": 1.2535,
1185
+ "step": 16800
1186
+ },
1187
+ {
1188
+ "epoch": 0.9364956222985703,
1189
+ "grad_norm": 1.0181013345718384,
1190
+ "learning_rate": 1.5952565665521445e-05,
1191
+ "loss": 1.2335,
1192
+ "step": 16900
1193
+ },
1194
+ {
1195
+ "epoch": 0.9420370165133548,
1196
+ "grad_norm": 1.0378520488739014,
1197
+ "learning_rate": 1.586944475229968e-05,
1198
+ "loss": 1.2429,
1199
+ "step": 17000
1200
+ },
1201
+ {
1202
+ "epoch": 0.9475784107281392,
1203
+ "grad_norm": 0.7528340220451355,
1204
+ "learning_rate": 1.578632383907791e-05,
1205
+ "loss": 1.2432,
1206
+ "step": 17100
1207
+ },
1208
+ {
1209
+ "epoch": 0.9531198049429236,
1210
+ "grad_norm": 0.6110236048698425,
1211
+ "learning_rate": 1.5703202925856144e-05,
1212
+ "loss": 1.251,
1213
+ "step": 17200
1214
+ },
1215
+ {
1216
+ "epoch": 0.9586611991577081,
1217
+ "grad_norm": 1.2787925004959106,
1218
+ "learning_rate": 1.5620082012634378e-05,
1219
+ "loss": 1.2243,
1220
+ "step": 17300
1221
+ },
1222
+ {
1223
+ "epoch": 0.9642025933724925,
1224
+ "grad_norm": 0.8089519739151001,
1225
+ "learning_rate": 1.5536961099412612e-05,
1226
+ "loss": 1.2375,
1227
+ "step": 17400
1228
+ },
1229
+ {
1230
+ "epoch": 0.969743987587277,
1231
+ "grad_norm": 0.702037513256073,
1232
+ "learning_rate": 1.5453840186190846e-05,
1233
+ "loss": 1.2231,
1234
+ "step": 17500
1235
+ },
1236
+ {
1237
+ "epoch": 0.9752853818020614,
1238
+ "grad_norm": 1.3774892091751099,
1239
+ "learning_rate": 1.5370719272969077e-05,
1240
+ "loss": 1.2516,
1241
+ "step": 17600
1242
+ },
1243
+ {
1244
+ "epoch": 0.9808267760168459,
1245
+ "grad_norm": 0.8553925156593323,
1246
+ "learning_rate": 1.5287598359747314e-05,
1247
+ "loss": 1.2329,
1248
+ "step": 17700
1249
+ },
1250
+ {
1251
+ "epoch": 0.9863681702316303,
1252
+ "grad_norm": 1.9921386241912842,
1253
+ "learning_rate": 1.5204477446525546e-05,
1254
+ "loss": 1.2474,
1255
+ "step": 17800
1256
+ },
1257
+ {
1258
+ "epoch": 0.9919095644464148,
1259
+ "grad_norm": 0.7573133707046509,
1260
+ "learning_rate": 1.512135653330378e-05,
1261
+ "loss": 1.2578,
1262
+ "step": 17900
1263
+ },
1264
+ {
1265
+ "epoch": 0.9974509586611991,
1266
+ "grad_norm": 2.220541000366211,
1267
+ "learning_rate": 1.5038235620082011e-05,
1268
+ "loss": 1.2222,
1269
+ "step": 18000
1270
+ },
1271
+ {
1272
+ "epoch": 1.0029923528759837,
1273
+ "grad_norm": 3.79048228263855,
1274
+ "learning_rate": 1.4955114706860245e-05,
1275
+ "loss": 1.2138,
1276
+ "step": 18100
1277
+ },
1278
+ {
1279
+ "epoch": 1.008533747090768,
1280
+ "grad_norm": 2.1642391681671143,
1281
+ "learning_rate": 1.4871993793638479e-05,
1282
+ "loss": 1.2038,
1283
+ "step": 18200
1284
+ },
1285
+ {
1286
+ "epoch": 1.0140751413055524,
1287
+ "grad_norm": 0.719645082950592,
1288
+ "learning_rate": 1.4788872880416715e-05,
1289
+ "loss": 1.213,
1290
+ "step": 18300
1291
+ },
1292
+ {
1293
+ "epoch": 1.0196165355203368,
1294
+ "grad_norm": 0.9910865426063538,
1295
+ "learning_rate": 1.4705751967194947e-05,
1296
+ "loss": 1.2227,
1297
+ "step": 18400
1298
+ },
1299
+ {
1300
+ "epoch": 1.0251579297351214,
1301
+ "grad_norm": 0.8456553220748901,
1302
+ "learning_rate": 1.4622631053973181e-05,
1303
+ "loss": 1.2076,
1304
+ "step": 18500
1305
+ },
1306
+ {
1307
+ "epoch": 1.0306993239499058,
1308
+ "grad_norm": 0.7198830842971802,
1309
+ "learning_rate": 1.4539510140751413e-05,
1310
+ "loss": 1.2184,
1311
+ "step": 18600
1312
+ },
1313
+ {
1314
+ "epoch": 1.0362407181646902,
1315
+ "grad_norm": 2.5999066829681396,
1316
+ "learning_rate": 1.4456389227529647e-05,
1317
+ "loss": 1.2161,
1318
+ "step": 18700
1319
+ },
1320
+ {
1321
+ "epoch": 1.0417821123794746,
1322
+ "grad_norm": 1.0215507745742798,
1323
+ "learning_rate": 1.437326831430788e-05,
1324
+ "loss": 1.2148,
1325
+ "step": 18800
1326
+ },
1327
+ {
1328
+ "epoch": 1.0473235065942592,
1329
+ "grad_norm": 2.206827402114868,
1330
+ "learning_rate": 1.4290147401086114e-05,
1331
+ "loss": 1.2189,
1332
+ "step": 18900
1333
+ },
1334
+ {
1335
+ "epoch": 1.0528649008090436,
1336
+ "grad_norm": 1.1960684061050415,
1337
+ "learning_rate": 1.4207026487864346e-05,
1338
+ "loss": 1.2073,
1339
+ "step": 19000
1340
+ },
1341
+ {
1342
+ "epoch": 1.058406295023828,
1343
+ "grad_norm": 0.9036281704902649,
1344
+ "learning_rate": 1.412390557464258e-05,
1345
+ "loss": 1.19,
1346
+ "step": 19100
1347
+ },
1348
+ {
1349
+ "epoch": 1.0639476892386124,
1350
+ "grad_norm": 1.1484371423721313,
1351
+ "learning_rate": 1.4040784661420814e-05,
1352
+ "loss": 1.2059,
1353
+ "step": 19200
1354
+ },
1355
+ {
1356
+ "epoch": 1.0694890834533968,
1357
+ "grad_norm": 2.325908660888672,
1358
+ "learning_rate": 1.3957663748199046e-05,
1359
+ "loss": 1.1951,
1360
+ "step": 19300
1361
+ },
1362
+ {
1363
+ "epoch": 1.0750304776681814,
1364
+ "grad_norm": 0.8740373253822327,
1365
+ "learning_rate": 1.3874542834977282e-05,
1366
+ "loss": 1.2084,
1367
+ "step": 19400
1368
+ },
1369
+ {
1370
+ "epoch": 1.0805718718829658,
1371
+ "grad_norm": 2.4061365127563477,
1372
+ "learning_rate": 1.3791421921755514e-05,
1373
+ "loss": 1.2071,
1374
+ "step": 19500
1375
+ },
1376
+ {
1377
+ "epoch": 1.0861132660977502,
1378
+ "grad_norm": 1.086378812789917,
1379
+ "learning_rate": 1.3708301008533748e-05,
1380
+ "loss": 1.2235,
1381
+ "step": 19600
1382
+ },
1383
+ {
1384
+ "epoch": 1.0916546603125346,
1385
+ "grad_norm": 1.6212489604949951,
1386
+ "learning_rate": 1.362518009531198e-05,
1387
+ "loss": 1.1863,
1388
+ "step": 19700
1389
+ },
1390
+ {
1391
+ "epoch": 1.0971960545273192,
1392
+ "grad_norm": 3.0122616291046143,
1393
+ "learning_rate": 1.3542059182090215e-05,
1394
+ "loss": 1.1916,
1395
+ "step": 19800
1396
+ },
1397
+ {
1398
+ "epoch": 1.1027374487421036,
1399
+ "grad_norm": 0.6291629076004028,
1400
+ "learning_rate": 1.3458938268868447e-05,
1401
+ "loss": 1.2269,
1402
+ "step": 19900
1403
+ },
1404
+ {
1405
+ "epoch": 1.108278842956888,
1406
+ "grad_norm": 0.7421205043792725,
1407
+ "learning_rate": 1.3375817355646681e-05,
1408
+ "loss": 1.211,
1409
+ "step": 20000
1410
+ },
1411
+ {
1412
+ "epoch": 1.1138202371716723,
1413
+ "grad_norm": 1.2598263025283813,
1414
+ "learning_rate": 1.3292696442424913e-05,
1415
+ "loss": 1.2169,
1416
+ "step": 20100
1417
+ },
1418
+ {
1419
+ "epoch": 1.119361631386457,
1420
+ "grad_norm": 1.1422287225723267,
1421
+ "learning_rate": 1.3209575529203147e-05,
1422
+ "loss": 1.2289,
1423
+ "step": 20200
1424
+ },
1425
+ {
1426
+ "epoch": 1.1249030256012413,
1427
+ "grad_norm": 1.332324504852295,
1428
+ "learning_rate": 1.3126454615981382e-05,
1429
+ "loss": 1.2183,
1430
+ "step": 20300
1431
+ },
1432
+ {
1433
+ "epoch": 1.1304444198160257,
1434
+ "grad_norm": 2.4671413898468018,
1435
+ "learning_rate": 1.3043333702759614e-05,
1436
+ "loss": 1.2022,
1437
+ "step": 20400
1438
+ },
1439
+ {
1440
+ "epoch": 1.1359858140308101,
1441
+ "grad_norm": 1.7303521633148193,
1442
+ "learning_rate": 1.2960212789537848e-05,
1443
+ "loss": 1.2208,
1444
+ "step": 20500
1445
+ },
1446
+ {
1447
+ "epoch": 1.1415272082455945,
1448
+ "grad_norm": 1.3321056365966797,
1449
+ "learning_rate": 1.2877091876316082e-05,
1450
+ "loss": 1.186,
1451
+ "step": 20600
1452
+ },
1453
+ {
1454
+ "epoch": 1.1470686024603791,
1455
+ "grad_norm": 1.39114511013031,
1456
+ "learning_rate": 1.2793970963094316e-05,
1457
+ "loss": 1.2123,
1458
+ "step": 20700
1459
+ },
1460
+ {
1461
+ "epoch": 1.1526099966751635,
1462
+ "grad_norm": 1.1872011423110962,
1463
+ "learning_rate": 1.2710850049872548e-05,
1464
+ "loss": 1.2079,
1465
+ "step": 20800
1466
+ },
1467
+ {
1468
+ "epoch": 1.158151390889948,
1469
+ "grad_norm": 1.1009360551834106,
1470
+ "learning_rate": 1.2627729136650782e-05,
1471
+ "loss": 1.2134,
1472
+ "step": 20900
1473
+ },
1474
+ {
1475
+ "epoch": 1.1636927851047323,
1476
+ "grad_norm": 1.094801664352417,
1477
+ "learning_rate": 1.2544608223429015e-05,
1478
+ "loss": 1.2076,
1479
+ "step": 21000
1480
+ },
1481
+ {
1482
+ "epoch": 1.169234179319517,
1483
+ "grad_norm": 1.2555512189865112,
1484
+ "learning_rate": 1.2461487310207249e-05,
1485
+ "loss": 1.2218,
1486
+ "step": 21100
1487
+ },
1488
+ {
1489
+ "epoch": 1.1747755735343013,
1490
+ "grad_norm": 2.6357486248016357,
1491
+ "learning_rate": 1.2378366396985483e-05,
1492
+ "loss": 1.1952,
1493
+ "step": 21200
1494
+ },
1495
+ {
1496
+ "epoch": 1.1803169677490857,
1497
+ "grad_norm": 0.9188030958175659,
1498
+ "learning_rate": 1.2295245483763715e-05,
1499
+ "loss": 1.2052,
1500
+ "step": 21300
1501
+ },
1502
+ {
1503
+ "epoch": 1.18585836196387,
1504
+ "grad_norm": 0.6432448029518127,
1505
+ "learning_rate": 1.2212124570541949e-05,
1506
+ "loss": 1.2057,
1507
+ "step": 21400
1508
+ },
1509
+ {
1510
+ "epoch": 1.1913997561786545,
1511
+ "grad_norm": 0.6348960995674133,
1512
+ "learning_rate": 1.2129003657320181e-05,
1513
+ "loss": 1.2205,
1514
+ "step": 21500
1515
+ },
1516
+ {
1517
+ "epoch": 1.196941150393439,
1518
+ "grad_norm": 0.695725679397583,
1519
+ "learning_rate": 1.2045882744098415e-05,
1520
+ "loss": 1.2064,
1521
+ "step": 21600
1522
+ },
1523
+ {
1524
+ "epoch": 1.2024825446082235,
1525
+ "grad_norm": 1.1121282577514648,
1526
+ "learning_rate": 1.1962761830876648e-05,
1527
+ "loss": 1.2268,
1528
+ "step": 21700
1529
+ },
1530
+ {
1531
+ "epoch": 1.2080239388230078,
1532
+ "grad_norm": 1.0938053131103516,
1533
+ "learning_rate": 1.1879640917654883e-05,
1534
+ "loss": 1.2121,
1535
+ "step": 21800
1536
+ },
1537
+ {
1538
+ "epoch": 1.2135653330377922,
1539
+ "grad_norm": 3.3523197174072266,
1540
+ "learning_rate": 1.1796520004433116e-05,
1541
+ "loss": 1.2001,
1542
+ "step": 21900
1543
+ },
1544
+ {
1545
+ "epoch": 1.2191067272525768,
1546
+ "grad_norm": 2.4698383808135986,
1547
+ "learning_rate": 1.171339909121135e-05,
1548
+ "loss": 1.2044,
1549
+ "step": 22000
1550
+ },
1551
+ {
1552
+ "epoch": 1.2246481214673612,
1553
+ "grad_norm": 0.9351718425750732,
1554
+ "learning_rate": 1.1630278177989582e-05,
1555
+ "loss": 1.1884,
1556
+ "step": 22100
1557
+ },
1558
+ {
1559
+ "epoch": 1.2301895156821456,
1560
+ "grad_norm": 0.9378387331962585,
1561
+ "learning_rate": 1.1547157264767816e-05,
1562
+ "loss": 1.2056,
1563
+ "step": 22200
1564
+ },
1565
+ {
1566
+ "epoch": 1.23573090989693,
1567
+ "grad_norm": 2.597338914871216,
1568
+ "learning_rate": 1.146403635154605e-05,
1569
+ "loss": 1.2116,
1570
+ "step": 22300
1571
+ },
1572
+ {
1573
+ "epoch": 1.2412723041117144,
1574
+ "grad_norm": 0.8649385571479797,
1575
+ "learning_rate": 1.1380915438324282e-05,
1576
+ "loss": 1.213,
1577
+ "step": 22400
1578
+ },
1579
+ {
1580
+ "epoch": 1.246813698326499,
1581
+ "grad_norm": 3.0743801593780518,
1582
+ "learning_rate": 1.1297794525102516e-05,
1583
+ "loss": 1.1996,
1584
+ "step": 22500
1585
+ },
1586
+ {
1587
+ "epoch": 1.2523550925412834,
1588
+ "grad_norm": 1.3909724950790405,
1589
+ "learning_rate": 1.1214673611880749e-05,
1590
+ "loss": 1.198,
1591
+ "step": 22600
1592
+ },
1593
+ {
1594
+ "epoch": 1.2578964867560678,
1595
+ "grad_norm": 2.2874038219451904,
1596
+ "learning_rate": 1.1131552698658983e-05,
1597
+ "loss": 1.1817,
1598
+ "step": 22700
1599
+ },
1600
+ {
1601
+ "epoch": 1.2634378809708522,
1602
+ "grad_norm": 1.2329871654510498,
1603
+ "learning_rate": 1.1048431785437215e-05,
1604
+ "loss": 1.1886,
1605
+ "step": 22800
1606
+ },
1607
+ {
1608
+ "epoch": 1.2689792751856368,
1609
+ "grad_norm": 1.1210861206054688,
1610
+ "learning_rate": 1.0965310872215449e-05,
1611
+ "loss": 1.2003,
1612
+ "step": 22900
1613
+ },
1614
+ {
1615
+ "epoch": 1.2745206694004212,
1616
+ "grad_norm": 0.7688116431236267,
1617
+ "learning_rate": 1.0882189958993683e-05,
1618
+ "loss": 1.2227,
1619
+ "step": 23000
1620
+ },
1621
+ {
1622
+ "epoch": 1.2800620636152056,
1623
+ "grad_norm": 1.4471969604492188,
1624
+ "learning_rate": 1.0799069045771917e-05,
1625
+ "loss": 1.2255,
1626
+ "step": 23100
1627
+ },
1628
+ {
1629
+ "epoch": 1.28560345782999,
1630
+ "grad_norm": 0.9862343668937683,
1631
+ "learning_rate": 1.0715948132550151e-05,
1632
+ "loss": 1.2196,
1633
+ "step": 23200
1634
+ },
1635
+ {
1636
+ "epoch": 1.2911448520447744,
1637
+ "grad_norm": 0.8841775059700012,
1638
+ "learning_rate": 1.0632827219328383e-05,
1639
+ "loss": 1.2149,
1640
+ "step": 23300
1641
+ },
1642
+ {
1643
+ "epoch": 1.296686246259559,
1644
+ "grad_norm": 2.7124688625335693,
1645
+ "learning_rate": 1.0549706306106618e-05,
1646
+ "loss": 1.1915,
1647
+ "step": 23400
1648
+ },
1649
+ {
1650
+ "epoch": 1.3022276404743434,
1651
+ "grad_norm": 2.0986595153808594,
1652
+ "learning_rate": 1.046658539288485e-05,
1653
+ "loss": 1.2212,
1654
+ "step": 23500
1655
+ },
1656
+ {
1657
+ "epoch": 1.3077690346891278,
1658
+ "grad_norm": 0.6581986546516418,
1659
+ "learning_rate": 1.0383464479663084e-05,
1660
+ "loss": 1.2075,
1661
+ "step": 23600
1662
+ },
1663
+ {
1664
+ "epoch": 1.3133104289039121,
1665
+ "grad_norm": 1.219954252243042,
1666
+ "learning_rate": 1.0300343566441316e-05,
1667
+ "loss": 1.2003,
1668
+ "step": 23700
1669
+ },
1670
+ {
1671
+ "epoch": 1.3188518231186968,
1672
+ "grad_norm": 1.0948656797409058,
1673
+ "learning_rate": 1.021722265321955e-05,
1674
+ "loss": 1.2031,
1675
+ "step": 23800
1676
+ },
1677
+ {
1678
+ "epoch": 1.3243932173334811,
1679
+ "grad_norm": 1.120632290840149,
1680
+ "learning_rate": 1.0134101739997783e-05,
1681
+ "loss": 1.2069,
1682
+ "step": 23900
1683
+ },
1684
+ {
1685
+ "epoch": 1.3299346115482655,
1686
+ "grad_norm": 1.3407847881317139,
1687
+ "learning_rate": 1.0050980826776017e-05,
1688
+ "loss": 1.2022,
1689
+ "step": 24000
1690
+ },
1691
+ {
1692
+ "epoch": 1.33547600576305,
1693
+ "grad_norm": 2.170480728149414,
1694
+ "learning_rate": 9.96785991355425e-06,
1695
+ "loss": 1.1866,
1696
+ "step": 24100
1697
+ },
1698
+ {
1699
+ "epoch": 1.3410173999778343,
1700
+ "grad_norm": 0.999652087688446,
1701
+ "learning_rate": 9.884739000332485e-06,
1702
+ "loss": 1.2043,
1703
+ "step": 24200
1704
+ },
1705
+ {
1706
+ "epoch": 1.346558794192619,
1707
+ "grad_norm": 1.229710578918457,
1708
+ "learning_rate": 9.801618087110719e-06,
1709
+ "loss": 1.2099,
1710
+ "step": 24300
1711
+ },
1712
+ {
1713
+ "epoch": 1.3521001884074033,
1714
+ "grad_norm": 0.7365024089813232,
1715
+ "learning_rate": 9.718497173888951e-06,
1716
+ "loss": 1.2158,
1717
+ "step": 24400
1718
+ },
1719
+ {
1720
+ "epoch": 1.3576415826221877,
1721
+ "grad_norm": 2.9519498348236084,
1722
+ "learning_rate": 9.635376260667185e-06,
1723
+ "loss": 1.1904,
1724
+ "step": 24500
1725
+ },
1726
+ {
1727
+ "epoch": 1.3631829768369723,
1728
+ "grad_norm": 1.277997612953186,
1729
+ "learning_rate": 9.552255347445417e-06,
1730
+ "loss": 1.2056,
1731
+ "step": 24600
1732
+ },
1733
+ {
1734
+ "epoch": 1.3687243710517567,
1735
+ "grad_norm": 1.2643914222717285,
1736
+ "learning_rate": 9.469134434223651e-06,
1737
+ "loss": 1.1899,
1738
+ "step": 24700
1739
+ },
1740
+ {
1741
+ "epoch": 1.374265765266541,
1742
+ "grad_norm": 1.0967351198196411,
1743
+ "learning_rate": 9.386013521001884e-06,
1744
+ "loss": 1.189,
1745
+ "step": 24800
1746
+ },
1747
+ {
1748
+ "epoch": 1.3798071594813255,
1749
+ "grad_norm": 2.875250816345215,
1750
+ "learning_rate": 9.302892607780118e-06,
1751
+ "loss": 1.1995,
1752
+ "step": 24900
1753
+ },
1754
+ {
1755
+ "epoch": 1.3853485536961099,
1756
+ "grad_norm": 2.1933434009552,
1757
+ "learning_rate": 9.21977169455835e-06,
1758
+ "loss": 1.2042,
1759
+ "step": 25000
1760
+ },
1761
+ {
1762
+ "epoch": 1.3908899479108943,
1763
+ "grad_norm": 1.2143278121948242,
1764
+ "learning_rate": 9.136650781336584e-06,
1765
+ "loss": 1.203,
1766
+ "step": 25100
1767
+ },
1768
+ {
1769
+ "epoch": 1.3964313421256789,
1770
+ "grad_norm": 0.855842649936676,
1771
+ "learning_rate": 9.053529868114816e-06,
1772
+ "loss": 1.1996,
1773
+ "step": 25200
1774
+ },
1775
+ {
1776
+ "epoch": 1.4019727363404633,
1777
+ "grad_norm": 1.3569194078445435,
1778
+ "learning_rate": 8.970408954893052e-06,
1779
+ "loss": 1.2199,
1780
+ "step": 25300
1781
+ },
1782
+ {
1783
+ "epoch": 1.4075141305552477,
1784
+ "grad_norm": 2.6152353286743164,
1785
+ "learning_rate": 8.887288041671286e-06,
1786
+ "loss": 1.2051,
1787
+ "step": 25400
1788
+ },
1789
+ {
1790
+ "epoch": 1.4130555247700323,
1791
+ "grad_norm": 2.4184043407440186,
1792
+ "learning_rate": 8.804167128449518e-06,
1793
+ "loss": 1.2098,
1794
+ "step": 25500
1795
+ },
1796
+ {
1797
+ "epoch": 1.4185969189848167,
1798
+ "grad_norm": 1.34404718875885,
1799
+ "learning_rate": 8.721046215227752e-06,
1800
+ "loss": 1.2057,
1801
+ "step": 25600
1802
+ },
1803
+ {
1804
+ "epoch": 1.424138313199601,
1805
+ "grad_norm": 0.8461852669715881,
1806
+ "learning_rate": 8.637925302005985e-06,
1807
+ "loss": 1.1872,
1808
+ "step": 25700
1809
+ },
1810
+ {
1811
+ "epoch": 1.4296797074143854,
1812
+ "grad_norm": 3.817380666732788,
1813
+ "learning_rate": 8.554804388784219e-06,
1814
+ "loss": 1.2055,
1815
+ "step": 25800
1816
+ },
1817
+ {
1818
+ "epoch": 1.4352211016291698,
1819
+ "grad_norm": 2.2559361457824707,
1820
+ "learning_rate": 8.471683475562451e-06,
1821
+ "loss": 1.208,
1822
+ "step": 25900
1823
+ },
1824
+ {
1825
+ "epoch": 1.4407624958439542,
1826
+ "grad_norm": 1.232676386833191,
1827
+ "learning_rate": 8.388562562340685e-06,
1828
+ "loss": 1.197,
1829
+ "step": 26000
1830
+ },
1831
+ {
1832
+ "epoch": 1.4463038900587388,
1833
+ "grad_norm": 0.9286311268806458,
1834
+ "learning_rate": 8.305441649118917e-06,
1835
+ "loss": 1.2056,
1836
+ "step": 26100
1837
+ },
1838
+ {
1839
+ "epoch": 1.4518452842735232,
1840
+ "grad_norm": 1.5397043228149414,
1841
+ "learning_rate": 8.222320735897151e-06,
1842
+ "loss": 1.1884,
1843
+ "step": 26200
1844
+ },
1845
+ {
1846
+ "epoch": 1.4573866784883076,
1847
+ "grad_norm": 0.9644666314125061,
1848
+ "learning_rate": 8.139199822675385e-06,
1849
+ "loss": 1.1912,
1850
+ "step": 26300
1851
+ },
1852
+ {
1853
+ "epoch": 1.4629280727030922,
1854
+ "grad_norm": 1.5598095655441284,
1855
+ "learning_rate": 8.056078909453618e-06,
1856
+ "loss": 1.1926,
1857
+ "step": 26400
1858
+ },
1859
+ {
1860
+ "epoch": 1.4684694669178766,
1861
+ "grad_norm": 0.8418022394180298,
1862
+ "learning_rate": 7.972957996231853e-06,
1863
+ "loss": 1.2007,
1864
+ "step": 26500
1865
+ },
1866
+ {
1867
+ "epoch": 1.474010861132661,
1868
+ "grad_norm": 1.2986925840377808,
1869
+ "learning_rate": 7.889837083010086e-06,
1870
+ "loss": 1.1946,
1871
+ "step": 26600
1872
+ },
1873
+ {
1874
+ "epoch": 1.4795522553474454,
1875
+ "grad_norm": 1.0484559535980225,
1876
+ "learning_rate": 7.80671616978832e-06,
1877
+ "loss": 1.1863,
1878
+ "step": 26700
1879
+ },
1880
+ {
1881
+ "epoch": 1.4850936495622298,
1882
+ "grad_norm": 2.89473557472229,
1883
+ "learning_rate": 7.723595256566552e-06,
1884
+ "loss": 1.172,
1885
+ "step": 26800
1886
+ },
1887
+ {
1888
+ "epoch": 1.4906350437770142,
1889
+ "grad_norm": 1.2736523151397705,
1890
+ "learning_rate": 7.640474343344786e-06,
1891
+ "loss": 1.1993,
1892
+ "step": 26900
1893
+ },
1894
+ {
1895
+ "epoch": 1.4961764379917988,
1896
+ "grad_norm": 0.6851234436035156,
1897
+ "learning_rate": 7.5573534301230185e-06,
1898
+ "loss": 1.2056,
1899
+ "step": 27000
1900
+ },
1901
+ {
1902
+ "epoch": 1.5017178322065832,
1903
+ "grad_norm": 1.546315312385559,
1904
+ "learning_rate": 7.4742325169012525e-06,
1905
+ "loss": 1.1737,
1906
+ "step": 27100
1907
+ },
1908
+ {
1909
+ "epoch": 1.5072592264213678,
1910
+ "grad_norm": 0.6313077807426453,
1911
+ "learning_rate": 7.391111603679486e-06,
1912
+ "loss": 1.1984,
1913
+ "step": 27200
1914
+ },
1915
+ {
1916
+ "epoch": 1.5128006206361522,
1917
+ "grad_norm": 1.1231598854064941,
1918
+ "learning_rate": 7.307990690457719e-06,
1919
+ "loss": 1.1926,
1920
+ "step": 27300
1921
+ },
1922
+ {
1923
+ "epoch": 1.5183420148509366,
1924
+ "grad_norm": 0.8349178433418274,
1925
+ "learning_rate": 7.224869777235953e-06,
1926
+ "loss": 1.2113,
1927
+ "step": 27400
1928
+ },
1929
+ {
1930
+ "epoch": 1.523883409065721,
1931
+ "grad_norm": 0.9475273489952087,
1932
+ "learning_rate": 7.141748864014186e-06,
1933
+ "loss": 1.2028,
1934
+ "step": 27500
1935
+ },
1936
+ {
1937
+ "epoch": 1.5294248032805053,
1938
+ "grad_norm": 2.9551515579223633,
1939
+ "learning_rate": 7.058627950792419e-06,
1940
+ "loss": 1.1975,
1941
+ "step": 27600
1942
+ },
1943
+ {
1944
+ "epoch": 1.5349661974952897,
1945
+ "grad_norm": 0.7484708428382874,
1946
+ "learning_rate": 6.975507037570652e-06,
1947
+ "loss": 1.1814,
1948
+ "step": 27700
1949
+ },
1950
+ {
1951
+ "epoch": 1.5405075917100741,
1952
+ "grad_norm": 0.8442228436470032,
1953
+ "learning_rate": 6.8923861243488864e-06,
1954
+ "loss": 1.19,
1955
+ "step": 27800
1956
+ },
1957
+ {
1958
+ "epoch": 1.5460489859248587,
1959
+ "grad_norm": 3.3256871700286865,
1960
+ "learning_rate": 6.80926521112712e-06,
1961
+ "loss": 1.2252,
1962
+ "step": 27900
1963
+ },
1964
+ {
1965
+ "epoch": 1.5515903801396431,
1966
+ "grad_norm": 0.900709331035614,
1967
+ "learning_rate": 6.726144297905354e-06,
1968
+ "loss": 1.1801,
1969
+ "step": 28000
1970
+ },
1971
+ {
1972
+ "epoch": 1.5571317743544277,
1973
+ "grad_norm": 1.3425995111465454,
1974
+ "learning_rate": 6.643023384683587e-06,
1975
+ "loss": 1.1907,
1976
+ "step": 28100
1977
+ },
1978
+ {
1979
+ "epoch": 1.5626731685692121,
1980
+ "grad_norm": 0.5740347504615784,
1981
+ "learning_rate": 6.55990247146182e-06,
1982
+ "loss": 1.194,
1983
+ "step": 28200
1984
+ },
1985
+ {
1986
+ "epoch": 1.5682145627839965,
1987
+ "grad_norm": 1.318077564239502,
1988
+ "learning_rate": 6.476781558240053e-06,
1989
+ "loss": 1.2083,
1990
+ "step": 28300
1991
+ },
1992
+ {
1993
+ "epoch": 1.573755956998781,
1994
+ "grad_norm": 1.2379876375198364,
1995
+ "learning_rate": 6.393660645018286e-06,
1996
+ "loss": 1.1931,
1997
+ "step": 28400
1998
+ },
1999
+ {
2000
+ "epoch": 1.5792973512135653,
2001
+ "grad_norm": 3.2320945262908936,
2002
+ "learning_rate": 6.31053973179652e-06,
2003
+ "loss": 1.2153,
2004
+ "step": 28500
2005
+ },
2006
+ {
2007
+ "epoch": 1.5848387454283497,
2008
+ "grad_norm": 1.2113683223724365,
2009
+ "learning_rate": 6.2274188185747535e-06,
2010
+ "loss": 1.2092,
2011
+ "step": 28600
2012
+ },
2013
+ {
2014
+ "epoch": 1.590380139643134,
2015
+ "grad_norm": 1.2382327318191528,
2016
+ "learning_rate": 6.144297905352987e-06,
2017
+ "loss": 1.1976,
2018
+ "step": 28700
2019
+ },
2020
+ {
2021
+ "epoch": 1.5959215338579187,
2022
+ "grad_norm": 2.7310492992401123,
2023
+ "learning_rate": 6.061176992131221e-06,
2024
+ "loss": 1.1967,
2025
+ "step": 28800
2026
+ },
2027
+ {
2028
+ "epoch": 1.601462928072703,
2029
+ "grad_norm": 1.2863633632659912,
2030
+ "learning_rate": 5.978056078909454e-06,
2031
+ "loss": 1.2075,
2032
+ "step": 28900
2033
+ },
2034
+ {
2035
+ "epoch": 1.6070043222874877,
2036
+ "grad_norm": 1.1544476747512817,
2037
+ "learning_rate": 5.894935165687687e-06,
2038
+ "loss": 1.1807,
2039
+ "step": 29000
2040
+ },
2041
+ {
2042
+ "epoch": 1.612545716502272,
2043
+ "grad_norm": 2.8650896549224854,
2044
+ "learning_rate": 5.811814252465921e-06,
2045
+ "loss": 1.2076,
2046
+ "step": 29100
2047
+ },
2048
+ {
2049
+ "epoch": 1.6180871107170565,
2050
+ "grad_norm": 0.744802713394165,
2051
+ "learning_rate": 5.728693339244154e-06,
2052
+ "loss": 1.2009,
2053
+ "step": 29200
2054
+ },
2055
+ {
2056
+ "epoch": 1.6236285049318409,
2057
+ "grad_norm": 0.8797061443328857,
2058
+ "learning_rate": 5.6455724260223874e-06,
2059
+ "loss": 1.2058,
2060
+ "step": 29300
2061
+ },
2062
+ {
2063
+ "epoch": 1.6291698991466252,
2064
+ "grad_norm": 2.258319854736328,
2065
+ "learning_rate": 5.562451512800621e-06,
2066
+ "loss": 1.1872,
2067
+ "step": 29400
2068
+ },
2069
+ {
2070
+ "epoch": 1.6347112933614096,
2071
+ "grad_norm": 3.3380839824676514,
2072
+ "learning_rate": 5.479330599578854e-06,
2073
+ "loss": 1.1878,
2074
+ "step": 29500
2075
+ },
2076
+ {
2077
+ "epoch": 1.640252687576194,
2078
+ "grad_norm": 0.7560855150222778,
2079
+ "learning_rate": 5.396209686357087e-06,
2080
+ "loss": 1.1878,
2081
+ "step": 29600
2082
+ },
2083
+ {
2084
+ "epoch": 1.6457940817909786,
2085
+ "grad_norm": 1.3115606307983398,
2086
+ "learning_rate": 5.313088773135321e-06,
2087
+ "loss": 1.2017,
2088
+ "step": 29700
2089
+ },
2090
+ {
2091
+ "epoch": 1.651335476005763,
2092
+ "grad_norm": 1.267406940460205,
2093
+ "learning_rate": 5.229967859913555e-06,
2094
+ "loss": 1.2,
2095
+ "step": 29800
2096
+ },
2097
+ {
2098
+ "epoch": 1.6568768702205476,
2099
+ "grad_norm": 3.0515575408935547,
2100
+ "learning_rate": 5.146846946691788e-06,
2101
+ "loss": 1.2027,
2102
+ "step": 29900
2103
+ },
2104
+ {
2105
+ "epoch": 1.662418264435332,
2106
+ "grad_norm": 3.467629909515381,
2107
+ "learning_rate": 5.063726033470021e-06,
2108
+ "loss": 1.1921,
2109
+ "step": 30000
2110
+ },
2111
+ {
2112
+ "epoch": 1.6679596586501164,
2113
+ "grad_norm": 2.8003575801849365,
2114
+ "learning_rate": 4.9806051202482545e-06,
2115
+ "loss": 1.1704,
2116
+ "step": 30100
2117
+ },
2118
+ {
2119
+ "epoch": 1.6735010528649008,
2120
+ "grad_norm": 3.9951558113098145,
2121
+ "learning_rate": 4.897484207026488e-06,
2122
+ "loss": 1.2098,
2123
+ "step": 30200
2124
+ },
2125
+ {
2126
+ "epoch": 1.6790424470796852,
2127
+ "grad_norm": 1.0794897079467773,
2128
+ "learning_rate": 4.814363293804722e-06,
2129
+ "loss": 1.2046,
2130
+ "step": 30300
2131
+ },
2132
+ {
2133
+ "epoch": 1.6845838412944696,
2134
+ "grad_norm": 2.093555450439453,
2135
+ "learning_rate": 4.731242380582955e-06,
2136
+ "loss": 1.196,
2137
+ "step": 30400
2138
+ },
2139
+ {
2140
+ "epoch": 1.690125235509254,
2141
+ "grad_norm": 1.4370906352996826,
2142
+ "learning_rate": 4.648121467361188e-06,
2143
+ "loss": 1.2083,
2144
+ "step": 30500
2145
+ },
2146
+ {
2147
+ "epoch": 1.6956666297240386,
2148
+ "grad_norm": 3.102911949157715,
2149
+ "learning_rate": 4.565000554139421e-06,
2150
+ "loss": 1.2052,
2151
+ "step": 30600
2152
+ },
2153
+ {
2154
+ "epoch": 1.701208023938823,
2155
+ "grad_norm": 2.760039806365967,
2156
+ "learning_rate": 4.481879640917654e-06,
2157
+ "loss": 1.1986,
2158
+ "step": 30700
2159
+ },
2160
+ {
2161
+ "epoch": 1.7067494181536076,
2162
+ "grad_norm": 1.47969651222229,
2163
+ "learning_rate": 4.398758727695888e-06,
2164
+ "loss": 1.1799,
2165
+ "step": 30800
2166
+ },
2167
+ {
2168
+ "epoch": 1.712290812368392,
2169
+ "grad_norm": 0.8171949982643127,
2170
+ "learning_rate": 4.3156378144741224e-06,
2171
+ "loss": 1.1991,
2172
+ "step": 30900
2173
+ },
2174
+ {
2175
+ "epoch": 1.7178322065831764,
2176
+ "grad_norm": 2.6968746185302734,
2177
+ "learning_rate": 4.232516901252356e-06,
2178
+ "loss": 1.2016,
2179
+ "step": 31000
2180
+ },
2181
+ {
2182
+ "epoch": 1.7233736007979608,
2183
+ "grad_norm": 1.193534016609192,
2184
+ "learning_rate": 4.149395988030589e-06,
2185
+ "loss": 1.1719,
2186
+ "step": 31100
2187
+ },
2188
+ {
2189
+ "epoch": 1.7289149950127451,
2190
+ "grad_norm": 3.1099929809570312,
2191
+ "learning_rate": 4.066275074808822e-06,
2192
+ "loss": 1.1943,
2193
+ "step": 31200
2194
+ },
2195
+ {
2196
+ "epoch": 1.7344563892275295,
2197
+ "grad_norm": 0.7732232213020325,
2198
+ "learning_rate": 3.983154161587055e-06,
2199
+ "loss": 1.1989,
2200
+ "step": 31300
2201
+ },
2202
+ {
2203
+ "epoch": 1.739997783442314,
2204
+ "grad_norm": 2.8280205726623535,
2205
+ "learning_rate": 3.900033248365288e-06,
2206
+ "loss": 1.2068,
2207
+ "step": 31400
2208
+ },
2209
+ {
2210
+ "epoch": 1.7455391776570985,
2211
+ "grad_norm": 1.3707168102264404,
2212
+ "learning_rate": 3.816912335143522e-06,
2213
+ "loss": 1.2044,
2214
+ "step": 31500
2215
+ },
2216
+ {
2217
+ "epoch": 1.751080571871883,
2218
+ "grad_norm": 0.6743613481521606,
2219
+ "learning_rate": 3.733791421921756e-06,
2220
+ "loss": 1.1776,
2221
+ "step": 31600
2222
+ },
2223
+ {
2224
+ "epoch": 1.7566219660866675,
2225
+ "grad_norm": 0.9411177635192871,
2226
+ "learning_rate": 3.650670508699989e-06,
2227
+ "loss": 1.1659,
2228
+ "step": 31700
2229
+ },
2230
+ {
2231
+ "epoch": 1.762163360301452,
2232
+ "grad_norm": 3.406217575073242,
2233
+ "learning_rate": 3.5675495954782223e-06,
2234
+ "loss": 1.1898,
2235
+ "step": 31800
2236
+ },
2237
+ {
2238
+ "epoch": 1.7677047545162363,
2239
+ "grad_norm": 1.51300847530365,
2240
+ "learning_rate": 3.484428682256456e-06,
2241
+ "loss": 1.1827,
2242
+ "step": 31900
2243
+ },
2244
+ {
2245
+ "epoch": 1.7732461487310207,
2246
+ "grad_norm": 1.2516944408416748,
2247
+ "learning_rate": 3.401307769034689e-06,
2248
+ "loss": 1.2037,
2249
+ "step": 32000
2250
+ },
2251
+ {
2252
+ "epoch": 1.778787542945805,
2253
+ "grad_norm": 1.4199496507644653,
2254
+ "learning_rate": 3.3181868558129226e-06,
2255
+ "loss": 1.2044,
2256
+ "step": 32100
2257
+ },
2258
+ {
2259
+ "epoch": 1.7843289371605895,
2260
+ "grad_norm": 3.167757511138916,
2261
+ "learning_rate": 3.2350659425911562e-06,
2262
+ "loss": 1.2057,
2263
+ "step": 32200
2264
+ },
2265
+ {
2266
+ "epoch": 1.7898703313753739,
2267
+ "grad_norm": 1.2943912744522095,
2268
+ "learning_rate": 3.1519450293693894e-06,
2269
+ "loss": 1.1913,
2270
+ "step": 32300
2271
+ },
2272
+ {
2273
+ "epoch": 1.7954117255901585,
2274
+ "grad_norm": 0.8653062582015991,
2275
+ "learning_rate": 3.0688241161476226e-06,
2276
+ "loss": 1.1809,
2277
+ "step": 32400
2278
+ },
2279
+ {
2280
+ "epoch": 1.8009531198049429,
2281
+ "grad_norm": 0.7564247846603394,
2282
+ "learning_rate": 2.985703202925856e-06,
2283
+ "loss": 1.1961,
2284
+ "step": 32500
2285
+ },
2286
+ {
2287
+ "epoch": 1.8064945140197275,
2288
+ "grad_norm": 2.7736010551452637,
2289
+ "learning_rate": 2.9025822897040898e-06,
2290
+ "loss": 1.2027,
2291
+ "step": 32600
2292
+ },
2293
+ {
2294
+ "epoch": 1.8120359082345119,
2295
+ "grad_norm": 1.2347112894058228,
2296
+ "learning_rate": 2.819461376482323e-06,
2297
+ "loss": 1.1861,
2298
+ "step": 32700
2299
+ },
2300
+ {
2301
+ "epoch": 1.8175773024492963,
2302
+ "grad_norm": 3.789325475692749,
2303
+ "learning_rate": 2.7363404632605565e-06,
2304
+ "loss": 1.1915,
2305
+ "step": 32800
2306
+ },
2307
+ {
2308
+ "epoch": 1.8231186966640807,
2309
+ "grad_norm": 1.139970302581787,
2310
+ "learning_rate": 2.6532195500387897e-06,
2311
+ "loss": 1.1971,
2312
+ "step": 32900
2313
+ },
2314
+ {
2315
+ "epoch": 1.828660090878865,
2316
+ "grad_norm": 3.750394344329834,
2317
+ "learning_rate": 2.570098636817023e-06,
2318
+ "loss": 1.1647,
2319
+ "step": 33000
2320
+ },
2321
+ {
2322
+ "epoch": 1.8342014850936494,
2323
+ "grad_norm": 1.769946813583374,
2324
+ "learning_rate": 2.486977723595257e-06,
2325
+ "loss": 1.1855,
2326
+ "step": 33100
2327
+ },
2328
+ {
2329
+ "epoch": 1.839742879308434,
2330
+ "grad_norm": 1.2825562953948975,
2331
+ "learning_rate": 2.40385681037349e-06,
2332
+ "loss": 1.1785,
2333
+ "step": 33200
2334
+ },
2335
+ {
2336
+ "epoch": 1.8452842735232184,
2337
+ "grad_norm": 2.801413059234619,
2338
+ "learning_rate": 2.3207358971517232e-06,
2339
+ "loss": 1.1722,
2340
+ "step": 33300
2341
+ },
2342
+ {
2343
+ "epoch": 1.8508256677380028,
2344
+ "grad_norm": 1.2618361711502075,
2345
+ "learning_rate": 2.237614983929957e-06,
2346
+ "loss": 1.1898,
2347
+ "step": 33400
2348
+ },
2349
+ {
2350
+ "epoch": 1.8563670619527874,
2351
+ "grad_norm": 1.056923508644104,
2352
+ "learning_rate": 2.15449407070819e-06,
2353
+ "loss": 1.192,
2354
+ "step": 33500
2355
+ },
2356
+ {
2357
+ "epoch": 1.8619084561675718,
2358
+ "grad_norm": 1.3828442096710205,
2359
+ "learning_rate": 2.0713731574864236e-06,
2360
+ "loss": 1.1757,
2361
+ "step": 33600
2362
+ },
2363
+ {
2364
+ "epoch": 1.8674498503823562,
2365
+ "grad_norm": 1.426632046699524,
2366
+ "learning_rate": 1.988252244264657e-06,
2367
+ "loss": 1.201,
2368
+ "step": 33700
2369
+ },
2370
+ {
2371
+ "epoch": 1.8729912445971406,
2372
+ "grad_norm": 2.116184711456299,
2373
+ "learning_rate": 1.9051313310428904e-06,
2374
+ "loss": 1.1962,
2375
+ "step": 33800
2376
+ },
2377
+ {
2378
+ "epoch": 1.878532638811925,
2379
+ "grad_norm": 1.293491244316101,
2380
+ "learning_rate": 1.822010417821124e-06,
2381
+ "loss": 1.1815,
2382
+ "step": 33900
2383
+ },
2384
+ {
2385
+ "epoch": 1.8840740330267094,
2386
+ "grad_norm": 3.251264810562134,
2387
+ "learning_rate": 1.7388895045993574e-06,
2388
+ "loss": 1.1709,
2389
+ "step": 34000
2390
+ },
2391
+ {
2392
+ "epoch": 1.889615427241494,
2393
+ "grad_norm": 1.7608253955841064,
2394
+ "learning_rate": 1.6557685913775905e-06,
2395
+ "loss": 1.1867,
2396
+ "step": 34100
2397
+ },
2398
+ {
2399
+ "epoch": 1.8951568214562784,
2400
+ "grad_norm": 0.7705633640289307,
2401
+ "learning_rate": 1.5726476781558241e-06,
2402
+ "loss": 1.1918,
2403
+ "step": 34200
2404
+ },
2405
+ {
2406
+ "epoch": 1.9006982156710628,
2407
+ "grad_norm": 2.4283251762390137,
2408
+ "learning_rate": 1.4895267649340575e-06,
2409
+ "loss": 1.2103,
2410
+ "step": 34300
2411
+ },
2412
+ {
2413
+ "epoch": 1.9062396098858474,
2414
+ "grad_norm": 1.4676234722137451,
2415
+ "learning_rate": 1.4064058517122907e-06,
2416
+ "loss": 1.2032,
2417
+ "step": 34400
2418
+ },
2419
+ {
2420
+ "epoch": 1.9117810041006318,
2421
+ "grad_norm": 0.6845753788948059,
2422
+ "learning_rate": 1.3232849384905243e-06,
2423
+ "loss": 1.1716,
2424
+ "step": 34500
2425
+ },
2426
+ {
2427
+ "epoch": 1.9173223983154162,
2428
+ "grad_norm": 2.020423650741577,
2429
+ "learning_rate": 1.2401640252687577e-06,
2430
+ "loss": 1.1925,
2431
+ "step": 34600
2432
+ },
2433
+ {
2434
+ "epoch": 1.9228637925302006,
2435
+ "grad_norm": 2.275815010070801,
2436
+ "learning_rate": 1.157043112046991e-06,
2437
+ "loss": 1.1704,
2438
+ "step": 34700
2439
+ },
2440
+ {
2441
+ "epoch": 1.928405186744985,
2442
+ "grad_norm": 1.063697338104248,
2443
+ "learning_rate": 1.0739221988252244e-06,
2444
+ "loss": 1.1893,
2445
+ "step": 34800
2446
+ },
2447
+ {
2448
+ "epoch": 1.9339465809597693,
2449
+ "grad_norm": 1.1643952131271362,
2450
+ "learning_rate": 9.908012856034578e-07,
2451
+ "loss": 1.1788,
2452
+ "step": 34900
2453
+ },
2454
+ {
2455
+ "epoch": 1.939487975174554,
2456
+ "grad_norm": 1.824187994003296,
2457
+ "learning_rate": 9.076803723816913e-07,
2458
+ "loss": 1.2061,
2459
+ "step": 35000
2460
+ },
2461
+ {
2462
+ "epoch": 1.9450293693893383,
2463
+ "grad_norm": 1.3975834846496582,
2464
+ "learning_rate": 8.245594591599247e-07,
2465
+ "loss": 1.1763,
2466
+ "step": 35100
2467
+ },
2468
+ {
2469
+ "epoch": 1.950570763604123,
2470
+ "grad_norm": 2.2744956016540527,
2471
+ "learning_rate": 7.41438545938158e-07,
2472
+ "loss": 1.1912,
2473
+ "step": 35200
2474
+ },
2475
+ {
2476
+ "epoch": 1.9561121578189073,
2477
+ "grad_norm": 0.8214097619056702,
2478
+ "learning_rate": 6.583176327163915e-07,
2479
+ "loss": 1.1899,
2480
+ "step": 35300
2481
+ },
2482
+ {
2483
+ "epoch": 1.9616535520336917,
2484
+ "grad_norm": 2.8134753704071045,
2485
+ "learning_rate": 5.751967194946249e-07,
2486
+ "loss": 1.2095,
2487
+ "step": 35400
2488
+ },
2489
+ {
2490
+ "epoch": 1.9671949462484761,
2491
+ "grad_norm": 1.693952202796936,
2492
+ "learning_rate": 4.920758062728582e-07,
2493
+ "loss": 1.1836,
2494
+ "step": 35500
2495
+ },
2496
+ {
2497
+ "epoch": 1.9727363404632605,
2498
+ "grad_norm": 3.5731701850891113,
2499
+ "learning_rate": 4.089548930510917e-07,
2500
+ "loss": 1.1897,
2501
+ "step": 35600
2502
+ },
2503
+ {
2504
+ "epoch": 1.978277734678045,
2505
+ "grad_norm": 0.9789274334907532,
2506
+ "learning_rate": 3.2583397982932507e-07,
2507
+ "loss": 1.1918,
2508
+ "step": 35700
2509
+ },
2510
+ {
2511
+ "epoch": 1.9838191288928293,
2512
+ "grad_norm": 1.1598788499832153,
2513
+ "learning_rate": 2.4271306660755845e-07,
2514
+ "loss": 1.1898,
2515
+ "step": 35800
2516
+ },
2517
+ {
2518
+ "epoch": 1.989360523107614,
2519
+ "grad_norm": 1.2938345670700073,
2520
+ "learning_rate": 1.595921533857919e-07,
2521
+ "loss": 1.16,
2522
+ "step": 35900
2523
+ },
2524
+ {
2525
+ "epoch": 1.9949019173223983,
2526
+ "grad_norm": 0.6811911463737488,
2527
+ "learning_rate": 7.647124016402526e-08,
2528
+ "loss": 1.2095,
2529
+ "step": 36000
2530
+ }
2531
+ ],
2532
+ "logging_steps": 100,
2533
+ "max_steps": 36092,
2534
+ "num_input_tokens_seen": 0,
2535
+ "num_train_epochs": 2,
2536
+ "save_steps": 1000,
2537
+ "stateful_callbacks": {
2538
+ "TrainerControl": {
2539
+ "args": {
2540
+ "should_epoch_stop": false,
2541
+ "should_evaluate": false,
2542
+ "should_log": false,
2543
+ "should_save": true,
2544
+ "should_training_stop": false
2545
+ },
2546
+ "attributes": {}
2547
+ }
2548
+ },
2549
+ "total_flos": 0.0,
2550
+ "train_batch_size": 4,
2551
+ "trial_name": null,
2552
+ "trial_params": null
2553
+ }
checkpoint-36000/checkpoint-36000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4e5af0143e69f43eb36d258cc9e086252cf7c687e6d2897008ce099643325fb
3
+ size 4667