chansung commited on
Commit
0545094
1 Parent(s): 34b9900

Model save

Browse files
README.md ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: gemma
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - sft
7
+ - generated_from_trainer
8
+ base_model: google/gemma-2b
9
+ datasets:
10
+ - generator
11
+ model-index:
12
+ - name: gemma2b-summarize-claude3sonnet-32k
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # gemma2b-summarize-claude3sonnet-32k
20
+
21
+ This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the generator dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 2.5238
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 0.0002
43
+ - train_batch_size: 8
44
+ - eval_batch_size: 8
45
+ - seed: 42
46
+ - distributed_type: multi-GPU
47
+ - num_devices: 3
48
+ - gradient_accumulation_steps: 2
49
+ - total_train_batch_size: 48
50
+ - total_eval_batch_size: 24
51
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
+ - lr_scheduler_type: cosine
53
+ - lr_scheduler_warmup_ratio: 0.1
54
+ - num_epochs: 10
55
+
56
+ ### Training results
57
+
58
+ | Training Loss | Epoch | Step | Validation Loss |
59
+ |:-------------:|:------:|:----:|:---------------:|
60
+ | 1.2361 | 0.9951 | 101 | 2.5186 |
61
+ | 1.0968 | 2.0 | 203 | 2.4845 |
62
+ | 1.0436 | 2.9951 | 304 | 2.4796 |
63
+ | 1.0084 | 4.0 | 406 | 2.4944 |
64
+ | 0.9913 | 4.9951 | 507 | 2.5010 |
65
+ | 0.9588 | 6.0 | 609 | 2.5066 |
66
+ | 0.9459 | 6.9951 | 710 | 2.5164 |
67
+ | 0.943 | 8.0 | 812 | 2.5233 |
68
+ | 0.9169 | 8.9951 | 913 | 2.5240 |
69
+ | 0.925 | 9.9507 | 1010 | 2.5238 |
70
+
71
+
72
+ ### Framework versions
73
+
74
+ - PEFT 0.10.0
75
+ - Transformers 4.40.0
76
+ - Pytorch 2.2.2+cu121
77
+ - Datasets 2.19.1
78
+ - Tokenizers 0.19.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:89cf8225948b3f164c920775b0877a11ba0669213ca5c3924568d1acd73b285c
3
  size 19644912
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e2c2d17c5a93a5d7d4bb2cf165f327db8c660c1433f24a1514fe1f345583238
3
  size 19644912
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.950738916256158,
3
+ "total_flos": 5.932470720905871e+17,
4
+ "train_loss": 1.07745361446154,
5
+ "train_runtime": 3468.402,
6
+ "train_samples": 31747,
7
+ "train_samples_per_second": 13.998,
8
+ "train_steps_per_second": 0.291
9
+ }
runs/Jun05_14-32-00_user-HP-Z8-Fury-G5-Workstation-Desktop-PC/events.out.tfevents.1717565547.user-HP-Z8-Fury-G5-Workstation-Desktop-PC.936870.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:82374c32ccf95447dda8774786f5a657881bd77de761959eafa80db9ea918f08
3
- size 50217
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30034432e79e3a971d61eaa3835bed7a09263839e86de4b2aed22c24d8c52e2a
3
+ size 51264
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.950738916256158,
3
+ "total_flos": 5.932470720905871e+17,
4
+ "train_loss": 1.07745361446154,
5
+ "train_runtime": 3468.402,
6
+ "train_samples": 31747,
7
+ "train_samples_per_second": 13.998,
8
+ "train_steps_per_second": 0.291
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,1531 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.950738916256158,
5
+ "eval_steps": 500,
6
+ "global_step": 1010,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.009852216748768473,
13
+ "grad_norm": 1.765625,
14
+ "learning_rate": 1.9801980198019803e-06,
15
+ "loss": 2.8604,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.04926108374384237,
20
+ "grad_norm": 1.4921875,
21
+ "learning_rate": 9.900990099009901e-06,
22
+ "loss": 2.8185,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.09852216748768473,
27
+ "grad_norm": 2.1875,
28
+ "learning_rate": 1.9801980198019803e-05,
29
+ "loss": 2.8289,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.1477832512315271,
34
+ "grad_norm": 2.0625,
35
+ "learning_rate": 2.9702970297029702e-05,
36
+ "loss": 2.7485,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.19704433497536947,
41
+ "grad_norm": 3.3125,
42
+ "learning_rate": 3.9603960396039605e-05,
43
+ "loss": 2.6773,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.24630541871921183,
48
+ "grad_norm": 2.3125,
49
+ "learning_rate": 4.950495049504951e-05,
50
+ "loss": 2.5348,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.2955665024630542,
55
+ "grad_norm": 1.3359375,
56
+ "learning_rate": 5.9405940594059404e-05,
57
+ "loss": 2.375,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.3448275862068966,
62
+ "grad_norm": 1.5,
63
+ "learning_rate": 6.93069306930693e-05,
64
+ "loss": 2.225,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.39408866995073893,
69
+ "grad_norm": 1.1796875,
70
+ "learning_rate": 7.920792079207921e-05,
71
+ "loss": 2.0577,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.4433497536945813,
76
+ "grad_norm": 2.3125,
77
+ "learning_rate": 8.910891089108912e-05,
78
+ "loss": 1.9332,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.49261083743842365,
83
+ "grad_norm": 0.890625,
84
+ "learning_rate": 9.900990099009902e-05,
85
+ "loss": 1.8054,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.541871921182266,
90
+ "grad_norm": 0.82421875,
91
+ "learning_rate": 0.00010891089108910893,
92
+ "loss": 1.7066,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.5911330049261084,
97
+ "grad_norm": 0.55078125,
98
+ "learning_rate": 0.00011881188118811881,
99
+ "loss": 1.6055,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.6403940886699507,
104
+ "grad_norm": 0.44921875,
105
+ "learning_rate": 0.00012871287128712872,
106
+ "loss": 1.5102,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.6896551724137931,
111
+ "grad_norm": 0.8125,
112
+ "learning_rate": 0.0001386138613861386,
113
+ "loss": 1.4332,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.7389162561576355,
118
+ "grad_norm": 0.48046875,
119
+ "learning_rate": 0.0001485148514851485,
120
+ "loss": 1.3703,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.7881773399014779,
125
+ "grad_norm": 0.62109375,
126
+ "learning_rate": 0.00015841584158415842,
127
+ "loss": 1.3483,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.8374384236453202,
132
+ "grad_norm": 0.5625,
133
+ "learning_rate": 0.00016831683168316833,
134
+ "loss": 1.2989,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.8866995073891626,
139
+ "grad_norm": 0.55078125,
140
+ "learning_rate": 0.00017821782178217824,
141
+ "loss": 1.2627,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.9359605911330049,
146
+ "grad_norm": 0.71875,
147
+ "learning_rate": 0.00018811881188118812,
148
+ "loss": 1.2456,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.9852216748768473,
153
+ "grad_norm": 0.5,
154
+ "learning_rate": 0.00019801980198019803,
155
+ "loss": 1.2361,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.9950738916256158,
160
+ "eval_loss": 2.518617868423462,
161
+ "eval_runtime": 0.6245,
162
+ "eval_samples_per_second": 16.014,
163
+ "eval_steps_per_second": 1.601,
164
+ "step": 101
165
+ },
166
+ {
167
+ "epoch": 1.0344827586206897,
168
+ "grad_norm": 0.609375,
169
+ "learning_rate": 0.000199990444464082,
170
+ "loss": 1.2023,
171
+ "step": 105
172
+ },
173
+ {
174
+ "epoch": 1.083743842364532,
175
+ "grad_norm": 0.5546875,
176
+ "learning_rate": 0.00019995162822919883,
177
+ "loss": 1.1906,
178
+ "step": 110
179
+ },
180
+ {
181
+ "epoch": 1.1330049261083743,
182
+ "grad_norm": 0.4375,
183
+ "learning_rate": 0.00019988296565626987,
184
+ "loss": 1.1786,
185
+ "step": 115
186
+ },
187
+ {
188
+ "epoch": 1.1822660098522166,
189
+ "grad_norm": 0.62109375,
190
+ "learning_rate": 0.00019978447724847652,
191
+ "loss": 1.1654,
192
+ "step": 120
193
+ },
194
+ {
195
+ "epoch": 1.2315270935960592,
196
+ "grad_norm": 0.466796875,
197
+ "learning_rate": 0.0001996561924152278,
198
+ "loss": 1.1533,
199
+ "step": 125
200
+ },
201
+ {
202
+ "epoch": 1.2807881773399015,
203
+ "grad_norm": 0.466796875,
204
+ "learning_rate": 0.00019949814946337838,
205
+ "loss": 1.149,
206
+ "step": 130
207
+ },
208
+ {
209
+ "epoch": 1.3300492610837438,
210
+ "grad_norm": 0.5234375,
211
+ "learning_rate": 0.00019931039558578997,
212
+ "loss": 1.1363,
213
+ "step": 135
214
+ },
215
+ {
216
+ "epoch": 1.3793103448275863,
217
+ "grad_norm": 0.427734375,
218
+ "learning_rate": 0.00019909298684723904,
219
+ "loss": 1.1292,
220
+ "step": 140
221
+ },
222
+ {
223
+ "epoch": 1.4285714285714286,
224
+ "grad_norm": 0.7109375,
225
+ "learning_rate": 0.00019884598816767563,
226
+ "loss": 1.1319,
227
+ "step": 145
228
+ },
229
+ {
230
+ "epoch": 1.477832512315271,
231
+ "grad_norm": 0.384765625,
232
+ "learning_rate": 0.00019856947330283752,
233
+ "loss": 1.1248,
234
+ "step": 150
235
+ },
236
+ {
237
+ "epoch": 1.5270935960591134,
238
+ "grad_norm": 0.578125,
239
+ "learning_rate": 0.00019826352482222638,
240
+ "loss": 1.1241,
241
+ "step": 155
242
+ },
243
+ {
244
+ "epoch": 1.5763546798029555,
245
+ "grad_norm": 0.47265625,
246
+ "learning_rate": 0.00019792823408445174,
247
+ "loss": 1.1179,
248
+ "step": 160
249
+ },
250
+ {
251
+ "epoch": 1.625615763546798,
252
+ "grad_norm": 0.55859375,
253
+ "learning_rate": 0.00019756370120995066,
254
+ "loss": 1.1112,
255
+ "step": 165
256
+ },
257
+ {
258
+ "epoch": 1.6748768472906403,
259
+ "grad_norm": 0.462890625,
260
+ "learning_rate": 0.00019717003505109095,
261
+ "loss": 1.1021,
262
+ "step": 170
263
+ },
264
+ {
265
+ "epoch": 1.7241379310344827,
266
+ "grad_norm": 0.93359375,
267
+ "learning_rate": 0.0001967473531596671,
268
+ "loss": 1.1162,
269
+ "step": 175
270
+ },
271
+ {
272
+ "epoch": 1.7733990147783252,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 0.0001962957817517982,
275
+ "loss": 1.0989,
276
+ "step": 180
277
+ },
278
+ {
279
+ "epoch": 1.8226600985221675,
280
+ "grad_norm": 0.482421875,
281
+ "learning_rate": 0.000195815455670239,
282
+ "loss": 1.1004,
283
+ "step": 185
284
+ },
285
+ {
286
+ "epoch": 1.8719211822660098,
287
+ "grad_norm": 0.55859375,
288
+ "learning_rate": 0.00019530651834411474,
289
+ "loss": 1.0986,
290
+ "step": 190
291
+ },
292
+ {
293
+ "epoch": 1.9211822660098523,
294
+ "grad_norm": 0.40625,
295
+ "learning_rate": 0.0001947691217460921,
296
+ "loss": 1.0921,
297
+ "step": 195
298
+ },
299
+ {
300
+ "epoch": 1.9704433497536946,
301
+ "grad_norm": 0.59375,
302
+ "learning_rate": 0.0001942034263469989,
303
+ "loss": 1.0968,
304
+ "step": 200
305
+ },
306
+ {
307
+ "epoch": 2.0,
308
+ "eval_loss": 2.484497547149658,
309
+ "eval_runtime": 0.5386,
310
+ "eval_samples_per_second": 18.567,
311
+ "eval_steps_per_second": 1.857,
312
+ "step": 203
313
+ },
314
+ {
315
+ "epoch": 2.019704433497537,
316
+ "grad_norm": 0.75390625,
317
+ "learning_rate": 0.00019360960106790643,
318
+ "loss": 1.0792,
319
+ "step": 205
320
+ },
321
+ {
322
+ "epoch": 2.0689655172413794,
323
+ "grad_norm": 0.455078125,
324
+ "learning_rate": 0.00019298782322968815,
325
+ "loss": 1.0645,
326
+ "step": 210
327
+ },
328
+ {
329
+ "epoch": 2.1182266009852215,
330
+ "grad_norm": 0.6015625,
331
+ "learning_rate": 0.00019233827850007027,
332
+ "loss": 1.069,
333
+ "step": 215
334
+ },
335
+ {
336
+ "epoch": 2.167487684729064,
337
+ "grad_norm": 0.48046875,
338
+ "learning_rate": 0.00019166116083819002,
339
+ "loss": 1.0589,
340
+ "step": 220
341
+ },
342
+ {
343
+ "epoch": 2.2167487684729066,
344
+ "grad_norm": 0.53515625,
345
+ "learning_rate": 0.0001909566724366779,
346
+ "loss": 1.0646,
347
+ "step": 225
348
+ },
349
+ {
350
+ "epoch": 2.2660098522167487,
351
+ "grad_norm": 0.5546875,
352
+ "learning_rate": 0.00019022502366128135,
353
+ "loss": 1.0631,
354
+ "step": 230
355
+ },
356
+ {
357
+ "epoch": 2.315270935960591,
358
+ "grad_norm": 0.55078125,
359
+ "learning_rate": 0.00018946643298804793,
360
+ "loss": 1.0559,
361
+ "step": 235
362
+ },
363
+ {
364
+ "epoch": 2.3645320197044333,
365
+ "grad_norm": 0.609375,
366
+ "learning_rate": 0.00018868112693808665,
367
+ "loss": 1.0469,
368
+ "step": 240
369
+ },
370
+ {
371
+ "epoch": 2.413793103448276,
372
+ "grad_norm": 0.6953125,
373
+ "learning_rate": 0.00018786934000992688,
374
+ "loss": 1.0502,
375
+ "step": 245
376
+ },
377
+ {
378
+ "epoch": 2.4630541871921183,
379
+ "grad_norm": 0.53515625,
380
+ "learning_rate": 0.00018703131460949554,
381
+ "loss": 1.058,
382
+ "step": 250
383
+ },
384
+ {
385
+ "epoch": 2.512315270935961,
386
+ "grad_norm": 0.53515625,
387
+ "learning_rate": 0.0001861673009777325,
388
+ "loss": 1.0501,
389
+ "step": 255
390
+ },
391
+ {
392
+ "epoch": 2.561576354679803,
393
+ "grad_norm": 0.5546875,
394
+ "learning_rate": 0.00018527755711586678,
395
+ "loss": 1.0516,
396
+ "step": 260
397
+ },
398
+ {
399
+ "epoch": 2.6108374384236455,
400
+ "grad_norm": 0.6484375,
401
+ "learning_rate": 0.00018436234870837547,
402
+ "loss": 1.0503,
403
+ "step": 265
404
+ },
405
+ {
406
+ "epoch": 2.6600985221674875,
407
+ "grad_norm": 0.52734375,
408
+ "learning_rate": 0.00018342194904364813,
409
+ "loss": 1.0539,
410
+ "step": 270
411
+ },
412
+ {
413
+ "epoch": 2.70935960591133,
414
+ "grad_norm": 0.5078125,
415
+ "learning_rate": 0.00018245663893238075,
416
+ "loss": 1.0407,
417
+ "step": 275
418
+ },
419
+ {
420
+ "epoch": 2.7586206896551726,
421
+ "grad_norm": 0.466796875,
422
+ "learning_rate": 0.00018146670662372354,
423
+ "loss": 1.0412,
424
+ "step": 280
425
+ },
426
+ {
427
+ "epoch": 2.8078817733990147,
428
+ "grad_norm": 0.72265625,
429
+ "learning_rate": 0.0001804524477192075,
430
+ "loss": 1.0412,
431
+ "step": 285
432
+ },
433
+ {
434
+ "epoch": 2.857142857142857,
435
+ "grad_norm": 0.4921875,
436
+ "learning_rate": 0.00017941416508447536,
437
+ "loss": 1.0286,
438
+ "step": 290
439
+ },
440
+ {
441
+ "epoch": 2.9064039408866993,
442
+ "grad_norm": 0.63671875,
443
+ "learning_rate": 0.00017835216875884368,
444
+ "loss": 1.0476,
445
+ "step": 295
446
+ },
447
+ {
448
+ "epoch": 2.955665024630542,
449
+ "grad_norm": 0.5390625,
450
+ "learning_rate": 0.00017726677586272263,
451
+ "loss": 1.0436,
452
+ "step": 300
453
+ },
454
+ {
455
+ "epoch": 2.9950738916256157,
456
+ "eval_loss": 2.479555606842041,
457
+ "eval_runtime": 0.6669,
458
+ "eval_samples_per_second": 14.995,
459
+ "eval_steps_per_second": 1.5,
460
+ "step": 304
461
+ },
462
+ {
463
+ "epoch": 3.0049261083743843,
464
+ "grad_norm": 0.455078125,
465
+ "learning_rate": 0.0001761583105029213,
466
+ "loss": 1.0275,
467
+ "step": 305
468
+ },
469
+ {
470
+ "epoch": 3.0541871921182264,
471
+ "grad_norm": 0.6640625,
472
+ "learning_rate": 0.00017502710367586687,
473
+ "loss": 1.0062,
474
+ "step": 310
475
+ },
476
+ {
477
+ "epoch": 3.103448275862069,
478
+ "grad_norm": 0.6484375,
479
+ "learning_rate": 0.00017387349316876666,
480
+ "loss": 1.018,
481
+ "step": 315
482
+ },
483
+ {
484
+ "epoch": 3.1527093596059115,
485
+ "grad_norm": 0.68359375,
486
+ "learning_rate": 0.00017269782345874203,
487
+ "loss": 1.005,
488
+ "step": 320
489
+ },
490
+ {
491
+ "epoch": 3.2019704433497536,
492
+ "grad_norm": 0.66015625,
493
+ "learning_rate": 0.00017150044560996488,
494
+ "loss": 1.0104,
495
+ "step": 325
496
+ },
497
+ {
498
+ "epoch": 3.251231527093596,
499
+ "grad_norm": 0.609375,
500
+ "learning_rate": 0.00017028171716882714,
501
+ "loss": 1.0039,
502
+ "step": 330
503
+ },
504
+ {
505
+ "epoch": 3.3004926108374386,
506
+ "grad_norm": 0.494140625,
507
+ "learning_rate": 0.0001690420020571747,
508
+ "loss": 1.0177,
509
+ "step": 335
510
+ },
511
+ {
512
+ "epoch": 3.3497536945812807,
513
+ "grad_norm": 0.76171875,
514
+ "learning_rate": 0.00016778167046363734,
515
+ "loss": 1.0066,
516
+ "step": 340
517
+ },
518
+ {
519
+ "epoch": 3.399014778325123,
520
+ "grad_norm": 0.5859375,
521
+ "learning_rate": 0.00016650109873308765,
522
+ "loss": 1.0187,
523
+ "step": 345
524
+ },
525
+ {
526
+ "epoch": 3.4482758620689653,
527
+ "grad_norm": 0.498046875,
528
+ "learning_rate": 0.00016520066925426144,
529
+ "loss": 1.0157,
530
+ "step": 350
531
+ },
532
+ {
533
+ "epoch": 3.497536945812808,
534
+ "grad_norm": 0.734375,
535
+ "learning_rate": 0.00016388077034557355,
536
+ "loss": 1.0104,
537
+ "step": 355
538
+ },
539
+ {
540
+ "epoch": 3.5467980295566504,
541
+ "grad_norm": 0.5546875,
542
+ "learning_rate": 0.00016254179613916278,
543
+ "loss": 1.0177,
544
+ "step": 360
545
+ },
546
+ {
547
+ "epoch": 3.596059113300493,
548
+ "grad_norm": 0.4765625,
549
+ "learning_rate": 0.0001611841464632011,
550
+ "loss": 1.0193,
551
+ "step": 365
552
+ },
553
+ {
554
+ "epoch": 3.645320197044335,
555
+ "grad_norm": 0.52734375,
556
+ "learning_rate": 0.0001598082267225018,
557
+ "loss": 1.0096,
558
+ "step": 370
559
+ },
560
+ {
561
+ "epoch": 3.6945812807881775,
562
+ "grad_norm": 0.6640625,
563
+ "learning_rate": 0.0001584144477774623,
564
+ "loss": 1.025,
565
+ "step": 375
566
+ },
567
+ {
568
+ "epoch": 3.7438423645320196,
569
+ "grad_norm": 0.53515625,
570
+ "learning_rate": 0.00015700322582137827,
571
+ "loss": 1.0125,
572
+ "step": 380
573
+ },
574
+ {
575
+ "epoch": 3.793103448275862,
576
+ "grad_norm": 0.50390625,
577
+ "learning_rate": 0.00015557498225616487,
578
+ "loss": 1.0022,
579
+ "step": 385
580
+ },
581
+ {
582
+ "epoch": 3.8423645320197046,
583
+ "grad_norm": 0.578125,
584
+ "learning_rate": 0.00015413014356652286,
585
+ "loss": 1.007,
586
+ "step": 390
587
+ },
588
+ {
589
+ "epoch": 3.8916256157635467,
590
+ "grad_norm": 0.5078125,
591
+ "learning_rate": 0.000152669141192587,
592
+ "loss": 1.0013,
593
+ "step": 395
594
+ },
595
+ {
596
+ "epoch": 3.9408866995073892,
597
+ "grad_norm": 0.50390625,
598
+ "learning_rate": 0.00015119241140109467,
599
+ "loss": 1.0009,
600
+ "step": 400
601
+ },
602
+ {
603
+ "epoch": 3.9901477832512313,
604
+ "grad_norm": 0.5625,
605
+ "learning_rate": 0.00014970039515511304,
606
+ "loss": 1.0084,
607
+ "step": 405
608
+ },
609
+ {
610
+ "epoch": 4.0,
611
+ "eval_loss": 2.494363307952881,
612
+ "eval_runtime": 0.5386,
613
+ "eval_samples_per_second": 18.567,
614
+ "eval_steps_per_second": 1.857,
615
+ "step": 406
616
+ },
617
+ {
618
+ "epoch": 4.039408866995074,
619
+ "grad_norm": 0.50390625,
620
+ "learning_rate": 0.00014819353798236427,
621
+ "loss": 0.9801,
622
+ "step": 410
623
+ },
624
+ {
625
+ "epoch": 4.088669950738916,
626
+ "grad_norm": 0.60546875,
627
+ "learning_rate": 0.0001466722898421873,
628
+ "loss": 0.9817,
629
+ "step": 415
630
+ },
631
+ {
632
+ "epoch": 4.137931034482759,
633
+ "grad_norm": 0.5390625,
634
+ "learning_rate": 0.00014513710499117647,
635
+ "loss": 0.988,
636
+ "step": 420
637
+ },
638
+ {
639
+ "epoch": 4.187192118226601,
640
+ "grad_norm": 0.51171875,
641
+ "learning_rate": 0.00014358844184753712,
642
+ "loss": 0.9782,
643
+ "step": 425
644
+ },
645
+ {
646
+ "epoch": 4.236453201970443,
647
+ "grad_norm": 0.546875,
648
+ "learning_rate": 0.00014202676285419812,
649
+ "loss": 0.9812,
650
+ "step": 430
651
+ },
652
+ {
653
+ "epoch": 4.285714285714286,
654
+ "grad_norm": 0.7421875,
655
+ "learning_rate": 0.0001404525343407228,
656
+ "loss": 0.9897,
657
+ "step": 435
658
+ },
659
+ {
660
+ "epoch": 4.334975369458128,
661
+ "grad_norm": 0.859375,
662
+ "learning_rate": 0.00013886622638405952,
663
+ "loss": 0.992,
664
+ "step": 440
665
+ },
666
+ {
667
+ "epoch": 4.384236453201971,
668
+ "grad_norm": 0.5390625,
669
+ "learning_rate": 0.00013726831266817278,
670
+ "loss": 0.9933,
671
+ "step": 445
672
+ },
673
+ {
674
+ "epoch": 4.433497536945813,
675
+ "grad_norm": 0.54296875,
676
+ "learning_rate": 0.0001356592703425976,
677
+ "loss": 0.9742,
678
+ "step": 450
679
+ },
680
+ {
681
+ "epoch": 4.482758620689655,
682
+ "grad_norm": 0.6953125,
683
+ "learning_rate": 0.00013403957987995882,
684
+ "loss": 0.9777,
685
+ "step": 455
686
+ },
687
+ {
688
+ "epoch": 4.532019704433497,
689
+ "grad_norm": 0.63671875,
690
+ "learning_rate": 0.00013240972493249847,
691
+ "loss": 0.9853,
692
+ "step": 460
693
+ },
694
+ {
695
+ "epoch": 4.58128078817734,
696
+ "grad_norm": 0.474609375,
697
+ "learning_rate": 0.00013077019218765305,
698
+ "loss": 0.9791,
699
+ "step": 465
700
+ },
701
+ {
702
+ "epoch": 4.630541871921182,
703
+ "grad_norm": 0.53125,
704
+ "learning_rate": 0.00012912147122272523,
705
+ "loss": 0.9857,
706
+ "step": 470
707
+ },
708
+ {
709
+ "epoch": 4.679802955665025,
710
+ "grad_norm": 0.48046875,
711
+ "learning_rate": 0.00012746405435869198,
712
+ "loss": 0.9854,
713
+ "step": 475
714
+ },
715
+ {
716
+ "epoch": 4.7290640394088665,
717
+ "grad_norm": 0.43359375,
718
+ "learning_rate": 0.0001257984365131938,
719
+ "loss": 0.9836,
720
+ "step": 480
721
+ },
722
+ {
723
+ "epoch": 4.778325123152709,
724
+ "grad_norm": 0.478515625,
725
+ "learning_rate": 0.00012412511505274844,
726
+ "loss": 0.9939,
727
+ "step": 485
728
+ },
729
+ {
730
+ "epoch": 4.827586206896552,
731
+ "grad_norm": 0.494140625,
732
+ "learning_rate": 0.00012244458964423327,
733
+ "loss": 0.9685,
734
+ "step": 490
735
+ },
736
+ {
737
+ "epoch": 4.876847290640394,
738
+ "grad_norm": 0.55859375,
739
+ "learning_rate": 0.0001207573621056809,
740
+ "loss": 0.9806,
741
+ "step": 495
742
+ },
743
+ {
744
+ "epoch": 4.926108374384237,
745
+ "grad_norm": 0.494140625,
746
+ "learning_rate": 0.00011906393625643244,
747
+ "loss": 0.979,
748
+ "step": 500
749
+ },
750
+ {
751
+ "epoch": 4.975369458128079,
752
+ "grad_norm": 0.6484375,
753
+ "learning_rate": 0.00011736481776669306,
754
+ "loss": 0.9913,
755
+ "step": 505
756
+ },
757
+ {
758
+ "epoch": 4.995073891625616,
759
+ "eval_loss": 2.50097918510437,
760
+ "eval_runtime": 0.6794,
761
+ "eval_samples_per_second": 14.718,
762
+ "eval_steps_per_second": 1.472,
763
+ "step": 507
764
+ },
765
+ {
766
+ "epoch": 5.024630541871921,
767
+ "grad_norm": 0.57421875,
768
+ "learning_rate": 0.00011566051400653486,
769
+ "loss": 0.9714,
770
+ "step": 510
771
+ },
772
+ {
773
+ "epoch": 5.073891625615763,
774
+ "grad_norm": 0.6171875,
775
+ "learning_rate": 0.00011395153389439233,
776
+ "loss": 0.9602,
777
+ "step": 515
778
+ },
779
+ {
780
+ "epoch": 5.123152709359606,
781
+ "grad_norm": 0.5234375,
782
+ "learning_rate": 0.00011223838774509514,
783
+ "loss": 0.9657,
784
+ "step": 520
785
+ },
786
+ {
787
+ "epoch": 5.172413793103448,
788
+ "grad_norm": 0.490234375,
789
+ "learning_rate": 0.00011052158711748434,
790
+ "loss": 0.9526,
791
+ "step": 525
792
+ },
793
+ {
794
+ "epoch": 5.221674876847291,
795
+ "grad_norm": 0.5234375,
796
+ "learning_rate": 0.00010880164466165674,
797
+ "loss": 0.958,
798
+ "step": 530
799
+ },
800
+ {
801
+ "epoch": 5.2709359605911335,
802
+ "grad_norm": 0.498046875,
803
+ "learning_rate": 0.00010707907396588361,
804
+ "loss": 0.9666,
805
+ "step": 535
806
+ },
807
+ {
808
+ "epoch": 5.320197044334975,
809
+ "grad_norm": 0.52734375,
810
+ "learning_rate": 0.0001053543894032493,
811
+ "loss": 0.9625,
812
+ "step": 540
813
+ },
814
+ {
815
+ "epoch": 5.369458128078818,
816
+ "grad_norm": 0.515625,
817
+ "learning_rate": 0.00010362810597805526,
818
+ "loss": 0.9657,
819
+ "step": 545
820
+ },
821
+ {
822
+ "epoch": 5.41871921182266,
823
+ "grad_norm": 0.482421875,
824
+ "learning_rate": 0.00010190073917203589,
825
+ "loss": 0.9655,
826
+ "step": 550
827
+ },
828
+ {
829
+ "epoch": 5.467980295566503,
830
+ "grad_norm": 0.515625,
831
+ "learning_rate": 0.00010017280479043147,
832
+ "loss": 0.9665,
833
+ "step": 555
834
+ },
835
+ {
836
+ "epoch": 5.517241379310345,
837
+ "grad_norm": 0.498046875,
838
+ "learning_rate": 9.844481880796491e-05,
839
+ "loss": 0.9587,
840
+ "step": 560
841
+ },
842
+ {
843
+ "epoch": 5.566502463054187,
844
+ "grad_norm": 0.609375,
845
+ "learning_rate": 9.671729721476746e-05,
846
+ "loss": 0.9665,
847
+ "step": 565
848
+ },
849
+ {
850
+ "epoch": 5.615763546798029,
851
+ "grad_norm": 0.5390625,
852
+ "learning_rate": 9.499075586230013e-05,
853
+ "loss": 0.9554,
854
+ "step": 570
855
+ },
856
+ {
857
+ "epoch": 5.665024630541872,
858
+ "grad_norm": 0.625,
859
+ "learning_rate": 9.326571030931637e-05,
860
+ "loss": 0.9607,
861
+ "step": 575
862
+ },
863
+ {
864
+ "epoch": 5.714285714285714,
865
+ "grad_norm": 0.68359375,
866
+ "learning_rate": 9.154267566791223e-05,
867
+ "loss": 0.9669,
868
+ "step": 580
869
+ },
870
+ {
871
+ "epoch": 5.763546798029557,
872
+ "grad_norm": 0.61328125,
873
+ "learning_rate": 8.982216644970979e-05,
874
+ "loss": 0.9628,
875
+ "step": 585
876
+ },
877
+ {
878
+ "epoch": 5.812807881773399,
879
+ "grad_norm": 0.6171875,
880
+ "learning_rate": 8.810469641222001e-05,
881
+ "loss": 0.9511,
882
+ "step": 590
883
+ },
884
+ {
885
+ "epoch": 5.862068965517241,
886
+ "grad_norm": 0.51171875,
887
+ "learning_rate": 8.639077840543077e-05,
888
+ "loss": 0.9684,
889
+ "step": 595
890
+ },
891
+ {
892
+ "epoch": 5.911330049261084,
893
+ "grad_norm": 0.671875,
894
+ "learning_rate": 8.468092421866573e-05,
895
+ "loss": 0.9678,
896
+ "step": 600
897
+ },
898
+ {
899
+ "epoch": 5.960591133004926,
900
+ "grad_norm": 0.65625,
901
+ "learning_rate": 8.297564442776014e-05,
902
+ "loss": 0.9588,
903
+ "step": 605
904
+ },
905
+ {
906
+ "epoch": 6.0,
907
+ "eval_loss": 2.5066332817077637,
908
+ "eval_runtime": 0.5391,
909
+ "eval_samples_per_second": 18.548,
910
+ "eval_steps_per_second": 1.855,
911
+ "step": 609
912
+ },
913
+ {
914
+ "epoch": 6.009852216748769,
915
+ "grad_norm": 0.68359375,
916
+ "learning_rate": 8.127544824259889e-05,
917
+ "loss": 0.953,
918
+ "step": 610
919
+ },
920
+ {
921
+ "epoch": 6.059113300492611,
922
+ "grad_norm": 0.734375,
923
+ "learning_rate": 7.958084335506239e-05,
924
+ "loss": 0.9536,
925
+ "step": 615
926
+ },
927
+ {
928
+ "epoch": 6.108374384236453,
929
+ "grad_norm": 0.50390625,
930
+ "learning_rate": 7.789233578742582e-05,
931
+ "loss": 0.9446,
932
+ "step": 620
933
+ },
934
+ {
935
+ "epoch": 6.157635467980295,
936
+ "grad_norm": 0.498046875,
937
+ "learning_rate": 7.6210429741257e-05,
938
+ "loss": 0.9353,
939
+ "step": 625
940
+ },
941
+ {
942
+ "epoch": 6.206896551724138,
943
+ "grad_norm": 0.5078125,
944
+ "learning_rate": 7.453562744685778e-05,
945
+ "loss": 0.9449,
946
+ "step": 630
947
+ },
948
+ {
949
+ "epoch": 6.25615763546798,
950
+ "grad_norm": 0.5625,
951
+ "learning_rate": 7.286842901329412e-05,
952
+ "loss": 0.9535,
953
+ "step": 635
954
+ },
955
+ {
956
+ "epoch": 6.305418719211823,
957
+ "grad_norm": 0.498046875,
958
+ "learning_rate": 7.12093322790597e-05,
959
+ "loss": 0.9376,
960
+ "step": 640
961
+ },
962
+ {
963
+ "epoch": 6.3546798029556655,
964
+ "grad_norm": 0.51953125,
965
+ "learning_rate": 6.955883266341741e-05,
966
+ "loss": 0.9461,
967
+ "step": 645
968
+ },
969
+ {
970
+ "epoch": 6.403940886699507,
971
+ "grad_norm": 0.5,
972
+ "learning_rate": 6.791742301846326e-05,
973
+ "loss": 0.9503,
974
+ "step": 650
975
+ },
976
+ {
977
+ "epoch": 6.45320197044335,
978
+ "grad_norm": 0.46875,
979
+ "learning_rate": 6.62855934819569e-05,
980
+ "loss": 0.9447,
981
+ "step": 655
982
+ },
983
+ {
984
+ "epoch": 6.502463054187192,
985
+ "grad_norm": 0.6328125,
986
+ "learning_rate": 6.466383133096267e-05,
987
+ "loss": 0.9453,
988
+ "step": 660
989
+ },
990
+ {
991
+ "epoch": 6.551724137931035,
992
+ "grad_norm": 0.53125,
993
+ "learning_rate": 6.305262083634488e-05,
994
+ "loss": 0.944,
995
+ "step": 665
996
+ },
997
+ {
998
+ "epoch": 6.600985221674877,
999
+ "grad_norm": 0.53125,
1000
+ "learning_rate": 6.145244311816063e-05,
1001
+ "loss": 0.9366,
1002
+ "step": 670
1003
+ },
1004
+ {
1005
+ "epoch": 6.650246305418719,
1006
+ "grad_norm": 0.5234375,
1007
+ "learning_rate": 5.986377600199371e-05,
1008
+ "loss": 0.9405,
1009
+ "step": 675
1010
+ },
1011
+ {
1012
+ "epoch": 6.699507389162561,
1013
+ "grad_norm": 0.482421875,
1014
+ "learning_rate": 5.828709387627218e-05,
1015
+ "loss": 0.9521,
1016
+ "step": 680
1017
+ },
1018
+ {
1019
+ "epoch": 6.748768472906404,
1020
+ "grad_norm": 0.4453125,
1021
+ "learning_rate": 5.6722867550612116e-05,
1022
+ "loss": 0.9509,
1023
+ "step": 685
1024
+ },
1025
+ {
1026
+ "epoch": 6.798029556650246,
1027
+ "grad_norm": 0.59765625,
1028
+ "learning_rate": 5.5171564115230254e-05,
1029
+ "loss": 0.9625,
1030
+ "step": 690
1031
+ },
1032
+ {
1033
+ "epoch": 6.847290640394089,
1034
+ "grad_norm": 0.447265625,
1035
+ "learning_rate": 5.363364680146725e-05,
1036
+ "loss": 0.9496,
1037
+ "step": 695
1038
+ },
1039
+ {
1040
+ "epoch": 6.896551724137931,
1041
+ "grad_norm": 0.439453125,
1042
+ "learning_rate": 5.210957484346314e-05,
1043
+ "loss": 0.9457,
1044
+ "step": 700
1045
+ },
1046
+ {
1047
+ "epoch": 6.945812807881773,
1048
+ "grad_norm": 0.482421875,
1049
+ "learning_rate": 5.059980334102637e-05,
1050
+ "loss": 0.9377,
1051
+ "step": 705
1052
+ },
1053
+ {
1054
+ "epoch": 6.995073891625616,
1055
+ "grad_norm": 0.453125,
1056
+ "learning_rate": 4.9104783123737566e-05,
1057
+ "loss": 0.9459,
1058
+ "step": 710
1059
+ },
1060
+ {
1061
+ "epoch": 6.995073891625616,
1062
+ "eval_loss": 2.516418933868408,
1063
+ "eval_runtime": 0.674,
1064
+ "eval_samples_per_second": 14.837,
1065
+ "eval_steps_per_second": 1.484,
1066
+ "step": 710
1067
+ },
1068
+ {
1069
+ "epoch": 7.044334975369458,
1070
+ "grad_norm": 0.48828125,
1071
+ "learning_rate": 4.762496061632814e-05,
1072
+ "loss": 0.9341,
1073
+ "step": 715
1074
+ },
1075
+ {
1076
+ "epoch": 7.093596059113301,
1077
+ "grad_norm": 0.51953125,
1078
+ "learning_rate": 4.6160777705374524e-05,
1079
+ "loss": 0.938,
1080
+ "step": 720
1081
+ },
1082
+ {
1083
+ "epoch": 7.142857142857143,
1084
+ "grad_norm": 0.50390625,
1085
+ "learning_rate": 4.471267160734731e-05,
1086
+ "loss": 0.9366,
1087
+ "step": 725
1088
+ },
1089
+ {
1090
+ "epoch": 7.192118226600985,
1091
+ "grad_norm": 0.53125,
1092
+ "learning_rate": 4.328107473805487e-05,
1093
+ "loss": 0.9403,
1094
+ "step": 730
1095
+ },
1096
+ {
1097
+ "epoch": 7.241379310344827,
1098
+ "grad_norm": 0.45703125,
1099
+ "learning_rate": 4.1866414583520877e-05,
1100
+ "loss": 0.9387,
1101
+ "step": 735
1102
+ },
1103
+ {
1104
+ "epoch": 7.29064039408867,
1105
+ "grad_norm": 0.71484375,
1106
+ "learning_rate": 4.046911357233343e-05,
1107
+ "loss": 0.9334,
1108
+ "step": 740
1109
+ },
1110
+ {
1111
+ "epoch": 7.3399014778325125,
1112
+ "grad_norm": 0.498046875,
1113
+ "learning_rate": 3.9089588949504655e-05,
1114
+ "loss": 0.93,
1115
+ "step": 745
1116
+ },
1117
+ {
1118
+ "epoch": 7.389162561576355,
1119
+ "grad_norm": 0.453125,
1120
+ "learning_rate": 3.772825265187802e-05,
1121
+ "loss": 0.9298,
1122
+ "step": 750
1123
+ },
1124
+ {
1125
+ "epoch": 7.4384236453201975,
1126
+ "grad_norm": 0.474609375,
1127
+ "learning_rate": 3.638551118512089e-05,
1128
+ "loss": 0.9486,
1129
+ "step": 755
1130
+ },
1131
+ {
1132
+ "epoch": 7.487684729064039,
1133
+ "grad_norm": 0.447265625,
1134
+ "learning_rate": 3.506176550233863e-05,
1135
+ "loss": 0.9373,
1136
+ "step": 760
1137
+ },
1138
+ {
1139
+ "epoch": 7.536945812807882,
1140
+ "grad_norm": 0.423828125,
1141
+ "learning_rate": 3.3757410884346894e-05,
1142
+ "loss": 0.939,
1143
+ "step": 765
1144
+ },
1145
+ {
1146
+ "epoch": 7.586206896551724,
1147
+ "grad_norm": 0.42578125,
1148
+ "learning_rate": 3.2472836821637744e-05,
1149
+ "loss": 0.9325,
1150
+ "step": 770
1151
+ },
1152
+ {
1153
+ "epoch": 7.635467980295567,
1154
+ "grad_norm": 0.458984375,
1155
+ "learning_rate": 3.120842689807468e-05,
1156
+ "loss": 0.932,
1157
+ "step": 775
1158
+ },
1159
+ {
1160
+ "epoch": 7.684729064039409,
1161
+ "grad_norm": 0.51171875,
1162
+ "learning_rate": 2.996455867635155e-05,
1163
+ "loss": 0.9259,
1164
+ "step": 780
1165
+ },
1166
+ {
1167
+ "epoch": 7.733990147783251,
1168
+ "grad_norm": 0.44921875,
1169
+ "learning_rate": 2.874160358524931e-05,
1170
+ "loss": 0.9277,
1171
+ "step": 785
1172
+ },
1173
+ {
1174
+ "epoch": 7.783251231527093,
1175
+ "grad_norm": 0.458984375,
1176
+ "learning_rate": 2.753992680872457e-05,
1177
+ "loss": 0.9259,
1178
+ "step": 790
1179
+ },
1180
+ {
1181
+ "epoch": 7.832512315270936,
1182
+ "grad_norm": 0.447265625,
1183
+ "learning_rate": 2.6359887176862718e-05,
1184
+ "loss": 0.9431,
1185
+ "step": 795
1186
+ },
1187
+ {
1188
+ "epoch": 7.8817733990147785,
1189
+ "grad_norm": 0.46875,
1190
+ "learning_rate": 2.5201837058728505e-05,
1191
+ "loss": 0.9323,
1192
+ "step": 800
1193
+ },
1194
+ {
1195
+ "epoch": 7.931034482758621,
1196
+ "grad_norm": 0.4296875,
1197
+ "learning_rate": 2.4066122257145894e-05,
1198
+ "loss": 0.9294,
1199
+ "step": 805
1200
+ },
1201
+ {
1202
+ "epoch": 7.980295566502463,
1203
+ "grad_norm": 0.453125,
1204
+ "learning_rate": 2.295308190543859e-05,
1205
+ "loss": 0.943,
1206
+ "step": 810
1207
+ },
1208
+ {
1209
+ "epoch": 8.0,
1210
+ "eval_loss": 2.523322582244873,
1211
+ "eval_runtime": 0.5404,
1212
+ "eval_samples_per_second": 18.504,
1213
+ "eval_steps_per_second": 1.85,
1214
+ "step": 812
1215
+ },
1216
+ {
1217
+ "epoch": 8.029556650246306,
1218
+ "grad_norm": 0.419921875,
1219
+ "learning_rate": 2.1863048366162208e-05,
1220
+ "loss": 0.9428,
1221
+ "step": 815
1222
+ },
1223
+ {
1224
+ "epoch": 8.078817733990148,
1225
+ "grad_norm": 0.44140625,
1226
+ "learning_rate": 2.0796347131858186e-05,
1227
+ "loss": 0.931,
1228
+ "step": 820
1229
+ },
1230
+ {
1231
+ "epoch": 8.12807881773399,
1232
+ "grad_norm": 0.431640625,
1233
+ "learning_rate": 1.9753296727859195e-05,
1234
+ "loss": 0.9263,
1235
+ "step": 825
1236
+ },
1237
+ {
1238
+ "epoch": 8.177339901477833,
1239
+ "grad_norm": 0.455078125,
1240
+ "learning_rate": 1.8734208617174988e-05,
1241
+ "loss": 0.9426,
1242
+ "step": 830
1243
+ },
1244
+ {
1245
+ "epoch": 8.226600985221674,
1246
+ "grad_norm": 0.453125,
1247
+ "learning_rate": 1.773938710748706e-05,
1248
+ "loss": 0.9267,
1249
+ "step": 835
1250
+ },
1251
+ {
1252
+ "epoch": 8.275862068965518,
1253
+ "grad_norm": 0.427734375,
1254
+ "learning_rate": 1.676912926028007e-05,
1255
+ "loss": 0.9269,
1256
+ "step": 840
1257
+ },
1258
+ {
1259
+ "epoch": 8.32512315270936,
1260
+ "grad_norm": 0.427734375,
1261
+ "learning_rate": 1.5823724802136865e-05,
1262
+ "loss": 0.9294,
1263
+ "step": 845
1264
+ },
1265
+ {
1266
+ "epoch": 8.374384236453203,
1267
+ "grad_norm": 0.419921875,
1268
+ "learning_rate": 1.4903456038223939e-05,
1269
+ "loss": 0.9222,
1270
+ "step": 850
1271
+ },
1272
+ {
1273
+ "epoch": 8.423645320197044,
1274
+ "grad_norm": 0.412109375,
1275
+ "learning_rate": 1.4008597767992871e-05,
1276
+ "loss": 0.929,
1277
+ "step": 855
1278
+ },
1279
+ {
1280
+ "epoch": 8.472906403940886,
1281
+ "grad_norm": 0.42578125,
1282
+ "learning_rate": 1.3139417203123027e-05,
1283
+ "loss": 0.9285,
1284
+ "step": 860
1285
+ },
1286
+ {
1287
+ "epoch": 8.52216748768473,
1288
+ "grad_norm": 0.421875,
1289
+ "learning_rate": 1.2296173887730123e-05,
1290
+ "loss": 0.9346,
1291
+ "step": 865
1292
+ },
1293
+ {
1294
+ "epoch": 8.571428571428571,
1295
+ "grad_norm": 0.42578125,
1296
+ "learning_rate": 1.1479119620864276e-05,
1297
+ "loss": 0.9252,
1298
+ "step": 870
1299
+ },
1300
+ {
1301
+ "epoch": 8.620689655172415,
1302
+ "grad_norm": 0.4375,
1303
+ "learning_rate": 1.0688498381320855e-05,
1304
+ "loss": 0.929,
1305
+ "step": 875
1306
+ },
1307
+ {
1308
+ "epoch": 8.669950738916256,
1309
+ "grad_norm": 0.66796875,
1310
+ "learning_rate": 9.924546254786493e-06,
1311
+ "loss": 0.9381,
1312
+ "step": 880
1313
+ },
1314
+ {
1315
+ "epoch": 8.719211822660098,
1316
+ "grad_norm": 0.453125,
1317
+ "learning_rate": 9.187491363342093e-06,
1318
+ "loss": 0.9374,
1319
+ "step": 885
1320
+ },
1321
+ {
1322
+ "epoch": 8.768472906403941,
1323
+ "grad_norm": 0.458984375,
1324
+ "learning_rate": 8.47755379734373e-06,
1325
+ "loss": 0.9267,
1326
+ "step": 890
1327
+ },
1328
+ {
1329
+ "epoch": 8.817733990147783,
1330
+ "grad_norm": 0.447265625,
1331
+ "learning_rate": 7.794945549701993e-06,
1332
+ "loss": 0.9369,
1333
+ "step": 895
1334
+ },
1335
+ {
1336
+ "epoch": 8.866995073891626,
1337
+ "grad_norm": 0.43359375,
1338
+ "learning_rate": 7.1398704525792e-06,
1339
+ "loss": 0.9371,
1340
+ "step": 900
1341
+ },
1342
+ {
1343
+ "epoch": 8.916256157635468,
1344
+ "grad_norm": 0.4375,
1345
+ "learning_rate": 6.512524116523633e-06,
1346
+ "loss": 0.9348,
1347
+ "step": 905
1348
+ },
1349
+ {
1350
+ "epoch": 8.96551724137931,
1351
+ "grad_norm": 0.44140625,
1352
+ "learning_rate": 5.913093872058528e-06,
1353
+ "loss": 0.9169,
1354
+ "step": 910
1355
+ },
1356
+ {
1357
+ "epoch": 8.995073891625616,
1358
+ "eval_loss": 2.523988723754883,
1359
+ "eval_runtime": 0.6711,
1360
+ "eval_samples_per_second": 14.9,
1361
+ "eval_steps_per_second": 1.49,
1362
+ "step": 913
1363
+ },
1364
+ {
1365
+ "epoch": 9.014778325123153,
1366
+ "grad_norm": 0.427734375,
1367
+ "learning_rate": 5.341758713743828e-06,
1368
+ "loss": 0.9254,
1369
+ "step": 915
1370
+ },
1371
+ {
1372
+ "epoch": 9.064039408866995,
1373
+ "grad_norm": 0.427734375,
1374
+ "learning_rate": 4.798689246727006e-06,
1375
+ "loss": 0.929,
1376
+ "step": 920
1377
+ },
1378
+ {
1379
+ "epoch": 9.113300492610838,
1380
+ "grad_norm": 0.421875,
1381
+ "learning_rate": 4.2840476357989825e-06,
1382
+ "loss": 0.9346,
1383
+ "step": 925
1384
+ },
1385
+ {
1386
+ "epoch": 9.16256157635468,
1387
+ "grad_norm": 0.4140625,
1388
+ "learning_rate": 3.797987556970495e-06,
1389
+ "loss": 0.936,
1390
+ "step": 930
1391
+ },
1392
+ {
1393
+ "epoch": 9.211822660098521,
1394
+ "grad_norm": 0.419921875,
1395
+ "learning_rate": 3.3406541515832003e-06,
1396
+ "loss": 0.9283,
1397
+ "step": 935
1398
+ },
1399
+ {
1400
+ "epoch": 9.261083743842365,
1401
+ "grad_norm": 0.41015625,
1402
+ "learning_rate": 2.912183982969385e-06,
1403
+ "loss": 0.9255,
1404
+ "step": 940
1405
+ },
1406
+ {
1407
+ "epoch": 9.310344827586206,
1408
+ "grad_norm": 0.421875,
1409
+ "learning_rate": 2.5127049956730207e-06,
1410
+ "loss": 0.9252,
1411
+ "step": 945
1412
+ },
1413
+ {
1414
+ "epoch": 9.35960591133005,
1415
+ "grad_norm": 0.431640625,
1416
+ "learning_rate": 2.1423364772445887e-06,
1417
+ "loss": 0.9276,
1418
+ "step": 950
1419
+ },
1420
+ {
1421
+ "epoch": 9.408866995073891,
1422
+ "grad_norm": 0.423828125,
1423
+ "learning_rate": 1.8011890226208527e-06,
1424
+ "loss": 0.9251,
1425
+ "step": 955
1426
+ },
1427
+ {
1428
+ "epoch": 9.458128078817733,
1429
+ "grad_norm": 0.4140625,
1430
+ "learning_rate": 1.489364501100332e-06,
1431
+ "loss": 0.9291,
1432
+ "step": 960
1433
+ },
1434
+ {
1435
+ "epoch": 9.507389162561577,
1436
+ "grad_norm": 0.423828125,
1437
+ "learning_rate": 1.2069560259243328e-06,
1438
+ "loss": 0.9333,
1439
+ "step": 965
1440
+ },
1441
+ {
1442
+ "epoch": 9.556650246305418,
1443
+ "grad_norm": 0.4375,
1444
+ "learning_rate": 9.540479264726676e-07,
1445
+ "loss": 0.9332,
1446
+ "step": 970
1447
+ },
1448
+ {
1449
+ "epoch": 9.605911330049262,
1450
+ "grad_norm": 0.4140625,
1451
+ "learning_rate": 7.307157230821426e-07,
1452
+ "loss": 0.9297,
1453
+ "step": 975
1454
+ },
1455
+ {
1456
+ "epoch": 9.655172413793103,
1457
+ "grad_norm": 0.40234375,
1458
+ "learning_rate": 5.370261044956971e-07,
1459
+ "loss": 0.9222,
1460
+ "step": 980
1461
+ },
1462
+ {
1463
+ "epoch": 9.704433497536947,
1464
+ "grad_norm": 0.421875,
1465
+ "learning_rate": 3.73036907948543e-07,
1466
+ "loss": 0.9396,
1467
+ "step": 985
1468
+ },
1469
+ {
1470
+ "epoch": 9.753694581280788,
1471
+ "grad_norm": 0.416015625,
1472
+ "learning_rate": 2.3879710189753656e-07,
1473
+ "loss": 0.931,
1474
+ "step": 990
1475
+ },
1476
+ {
1477
+ "epoch": 9.80295566502463,
1478
+ "grad_norm": 0.43359375,
1479
+ "learning_rate": 1.3434677139885222e-07,
1480
+ "loss": 0.9309,
1481
+ "step": 995
1482
+ },
1483
+ {
1484
+ "epoch": 9.852216748768473,
1485
+ "grad_norm": 0.4140625,
1486
+ "learning_rate": 5.971710613821291e-08,
1487
+ "loss": 0.9223,
1488
+ "step": 1000
1489
+ },
1490
+ {
1491
+ "epoch": 9.901477832512315,
1492
+ "grad_norm": 0.470703125,
1493
+ "learning_rate": 1.4930391117451426e-08,
1494
+ "loss": 0.9251,
1495
+ "step": 1005
1496
+ },
1497
+ {
1498
+ "epoch": 9.950738916256158,
1499
+ "grad_norm": 0.427734375,
1500
+ "learning_rate": 0.0,
1501
+ "loss": 0.925,
1502
+ "step": 1010
1503
+ },
1504
+ {
1505
+ "epoch": 9.950738916256158,
1506
+ "eval_loss": 2.523752212524414,
1507
+ "eval_runtime": 0.5416,
1508
+ "eval_samples_per_second": 18.464,
1509
+ "eval_steps_per_second": 1.846,
1510
+ "step": 1010
1511
+ },
1512
+ {
1513
+ "epoch": 9.950738916256158,
1514
+ "step": 1010,
1515
+ "total_flos": 5.932470720905871e+17,
1516
+ "train_loss": 1.07745361446154,
1517
+ "train_runtime": 3468.402,
1518
+ "train_samples_per_second": 13.998,
1519
+ "train_steps_per_second": 0.291
1520
+ }
1521
+ ],
1522
+ "logging_steps": 5,
1523
+ "max_steps": 1010,
1524
+ "num_input_tokens_seen": 0,
1525
+ "num_train_epochs": 10,
1526
+ "save_steps": 100,
1527
+ "total_flos": 5.932470720905871e+17,
1528
+ "train_batch_size": 8,
1529
+ "trial_name": null,
1530
+ "trial_params": null
1531
+ }