hugodk-sch commited on
Commit
3d27451
1 Parent(s): 95d79f7

Model save

Browse files
README.md ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ tags:
4
+ - trl
5
+ - sft
6
+ - generated_from_trainer
7
+ base_model: NorLLM-AI/NorMistral-7B
8
+ datasets:
9
+ - generator
10
+ model-index:
11
+ - name: norllm-ai-normistral-7b-sft-qlora
12
+ results: []
13
+ ---
14
+
15
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
16
+ should probably proofread and complete it, then remove this comment. -->
17
+
18
+ # norllm-ai-normistral-7b-sft-qlora
19
+
20
+ This model is a fine-tuned version of [NorLLM-AI/NorMistral-7B](https://huggingface.co/NorLLM-AI/NorMistral-7B) on the generator dataset.
21
+ It achieves the following results on the evaluation set:
22
+ - Loss: 1.4403
23
+
24
+ ## Model description
25
+
26
+ More information needed
27
+
28
+ ## Intended uses & limitations
29
+
30
+ More information needed
31
+
32
+ ## Training and evaluation data
33
+
34
+ More information needed
35
+
36
+ ## Training procedure
37
+
38
+ ### Training hyperparameters
39
+
40
+ The following hyperparameters were used during training:
41
+ - learning_rate: 0.0002
42
+ - train_batch_size: 4
43
+ - eval_batch_size: 8
44
+ - seed: 42
45
+ - distributed_type: multi-GPU
46
+ - gradient_accumulation_steps: 2
47
+ - total_train_batch_size: 8
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_ratio: 0.1
51
+ - num_epochs: 5
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:-----:|:----:|:---------------:|
57
+ | 1.7274 | 1.0 | 274 | 1.9432 |
58
+ | 1.1514 | 2.0 | 549 | 1.7111 |
59
+ | 0.645 | 3.0 | 823 | 1.5109 |
60
+ | 0.4291 | 4.0 | 1098 | 1.4415 |
61
+ | 0.3392 | 4.99 | 1370 | 1.4403 |
62
+
63
+
64
+ ### Framework versions
65
+
66
+ - PEFT 0.10.0
67
+ - Transformers 4.39.0.dev0
68
+ - Pytorch 2.1.2+cu121
69
+ - Datasets 2.14.6
70
+ - Tokenizers 0.15.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0fbca1c7381d2f9a014b80c2b3594fe6bbd9fbc9d4d931a448ff47475831c2b7
3
  size 83946192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc3e5e1f31fce526b019aaea0f5e020a78b105cc28acb9108258c5d58098e08e
3
  size 83946192
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 4.99,
3
+ "train_loss": 1.004000898782354,
4
+ "train_runtime": 10982.2688,
5
+ "train_samples": 4208,
6
+ "train_samples_per_second": 0.998,
7
+ "train_steps_per_second": 0.125
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 4.99,
3
+ "train_loss": 1.004000898782354,
4
+ "train_runtime": 10982.2688,
5
+ "train_samples": 4208,
6
+ "train_samples_per_second": 0.998,
7
+ "train_steps_per_second": 0.125
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1995 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 4.990892531876138,
5
+ "eval_steps": 500,
6
+ "global_step": 1370,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "grad_norm": 0.1083984375,
14
+ "learning_rate": 1.4598540145985402e-06,
15
+ "loss": 2.234,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.02,
20
+ "grad_norm": 0.09521484375,
21
+ "learning_rate": 7.2992700729927e-06,
22
+ "loss": 2.1882,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.04,
27
+ "grad_norm": 0.095703125,
28
+ "learning_rate": 1.45985401459854e-05,
29
+ "loss": 2.2055,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.05,
34
+ "grad_norm": 0.111328125,
35
+ "learning_rate": 2.1897810218978105e-05,
36
+ "loss": 2.176,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.07,
41
+ "grad_norm": 0.109375,
42
+ "learning_rate": 2.91970802919708e-05,
43
+ "loss": 2.1559,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.09,
48
+ "grad_norm": 0.11962890625,
49
+ "learning_rate": 3.649635036496351e-05,
50
+ "loss": 2.16,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.11,
55
+ "grad_norm": 0.1123046875,
56
+ "learning_rate": 4.379562043795621e-05,
57
+ "loss": 2.1681,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.13,
62
+ "grad_norm": 0.11572265625,
63
+ "learning_rate": 5.109489051094891e-05,
64
+ "loss": 2.0763,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.15,
69
+ "grad_norm": 0.1171875,
70
+ "learning_rate": 5.83941605839416e-05,
71
+ "loss": 2.114,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.16,
76
+ "grad_norm": 0.109375,
77
+ "learning_rate": 6.56934306569343e-05,
78
+ "loss": 2.0984,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.18,
83
+ "grad_norm": 0.111328125,
84
+ "learning_rate": 7.299270072992701e-05,
85
+ "loss": 2.0837,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.2,
90
+ "grad_norm": 0.1328125,
91
+ "learning_rate": 8.029197080291971e-05,
92
+ "loss": 2.0903,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.22,
97
+ "grad_norm": 0.1123046875,
98
+ "learning_rate": 8.759124087591242e-05,
99
+ "loss": 2.0727,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.24,
104
+ "grad_norm": 0.1201171875,
105
+ "learning_rate": 9.489051094890511e-05,
106
+ "loss": 2.0506,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.26,
111
+ "grad_norm": 0.1181640625,
112
+ "learning_rate": 0.00010218978102189782,
113
+ "loss": 2.0916,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.27,
118
+ "grad_norm": 0.1337890625,
119
+ "learning_rate": 0.00010948905109489052,
120
+ "loss": 2.0582,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.29,
125
+ "grad_norm": 0.1279296875,
126
+ "learning_rate": 0.0001167883211678832,
127
+ "loss": 2.0502,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.31,
132
+ "grad_norm": 0.12451171875,
133
+ "learning_rate": 0.0001240875912408759,
134
+ "loss": 2.0715,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.33,
139
+ "grad_norm": 0.1298828125,
140
+ "learning_rate": 0.0001313868613138686,
141
+ "loss": 1.9955,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.35,
146
+ "grad_norm": 0.146484375,
147
+ "learning_rate": 0.00013868613138686133,
148
+ "loss": 2.0099,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.36,
153
+ "grad_norm": 0.1376953125,
154
+ "learning_rate": 0.00014598540145985403,
155
+ "loss": 2.0107,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.38,
160
+ "grad_norm": 0.1533203125,
161
+ "learning_rate": 0.00015328467153284672,
162
+ "loss": 1.9787,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.4,
167
+ "grad_norm": 0.1689453125,
168
+ "learning_rate": 0.00016058394160583942,
169
+ "loss": 1.992,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.42,
174
+ "grad_norm": 0.1748046875,
175
+ "learning_rate": 0.00016788321167883211,
176
+ "loss": 2.0616,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.44,
181
+ "grad_norm": 0.1748046875,
182
+ "learning_rate": 0.00017518248175182484,
183
+ "loss": 2.0284,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.46,
188
+ "grad_norm": 0.197265625,
189
+ "learning_rate": 0.00018248175182481753,
190
+ "loss": 1.9884,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.47,
195
+ "grad_norm": 0.2578125,
196
+ "learning_rate": 0.00018978102189781023,
197
+ "loss": 1.995,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 0.49,
202
+ "grad_norm": 0.2294921875,
203
+ "learning_rate": 0.00019708029197080293,
204
+ "loss": 2.0103,
205
+ "step": 135
206
+ },
207
+ {
208
+ "epoch": 0.51,
209
+ "grad_norm": 0.2021484375,
210
+ "learning_rate": 0.00019999707864731247,
211
+ "loss": 1.9678,
212
+ "step": 140
213
+ },
214
+ {
215
+ "epoch": 0.53,
216
+ "grad_norm": 0.255859375,
217
+ "learning_rate": 0.0001999792265545627,
218
+ "loss": 2.0033,
219
+ "step": 145
220
+ },
221
+ {
222
+ "epoch": 0.55,
223
+ "grad_norm": 0.25,
224
+ "learning_rate": 0.00019994514823658906,
225
+ "loss": 1.9277,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.56,
230
+ "grad_norm": 0.259765625,
231
+ "learning_rate": 0.00019989484922416502,
232
+ "loss": 2.0091,
233
+ "step": 155
234
+ },
235
+ {
236
+ "epoch": 0.58,
237
+ "grad_norm": 0.25,
238
+ "learning_rate": 0.0001998283376806175,
239
+ "loss": 1.9636,
240
+ "step": 160
241
+ },
242
+ {
243
+ "epoch": 0.6,
244
+ "grad_norm": 0.25390625,
245
+ "learning_rate": 0.00019974562440050177,
246
+ "loss": 1.9385,
247
+ "step": 165
248
+ },
249
+ {
250
+ "epoch": 0.62,
251
+ "grad_norm": 0.259765625,
252
+ "learning_rate": 0.00019964672280784954,
253
+ "loss": 1.9536,
254
+ "step": 170
255
+ },
256
+ {
257
+ "epoch": 0.64,
258
+ "grad_norm": 0.244140625,
259
+ "learning_rate": 0.00019953164895399042,
260
+ "loss": 1.9552,
261
+ "step": 175
262
+ },
263
+ {
264
+ "epoch": 0.66,
265
+ "grad_norm": 0.28125,
266
+ "learning_rate": 0.00019940042151494675,
267
+ "loss": 1.8901,
268
+ "step": 180
269
+ },
270
+ {
271
+ "epoch": 0.67,
272
+ "grad_norm": 0.283203125,
273
+ "learning_rate": 0.0001992530617884026,
274
+ "loss": 1.9541,
275
+ "step": 185
276
+ },
277
+ {
278
+ "epoch": 0.69,
279
+ "grad_norm": 0.26953125,
280
+ "learning_rate": 0.00019908959369024728,
281
+ "loss": 1.9515,
282
+ "step": 190
283
+ },
284
+ {
285
+ "epoch": 0.71,
286
+ "grad_norm": 0.279296875,
287
+ "learning_rate": 0.00019891004375069377,
288
+ "loss": 1.896,
289
+ "step": 195
290
+ },
291
+ {
292
+ "epoch": 0.73,
293
+ "grad_norm": 0.279296875,
294
+ "learning_rate": 0.0001987144411099731,
295
+ "loss": 1.9335,
296
+ "step": 200
297
+ },
298
+ {
299
+ "epoch": 0.75,
300
+ "grad_norm": 0.380859375,
301
+ "learning_rate": 0.00019850281751360497,
302
+ "loss": 1.8805,
303
+ "step": 205
304
+ },
305
+ {
306
+ "epoch": 0.77,
307
+ "grad_norm": 0.296875,
308
+ "learning_rate": 0.00019827520730724543,
309
+ "loss": 1.8663,
310
+ "step": 210
311
+ },
312
+ {
313
+ "epoch": 0.78,
314
+ "grad_norm": 0.3125,
315
+ "learning_rate": 0.00019803164743111302,
316
+ "loss": 1.8169,
317
+ "step": 215
318
+ },
319
+ {
320
+ "epoch": 0.8,
321
+ "grad_norm": 0.3359375,
322
+ "learning_rate": 0.0001977721774139933,
323
+ "loss": 1.861,
324
+ "step": 220
325
+ },
326
+ {
327
+ "epoch": 0.82,
328
+ "grad_norm": 0.296875,
329
+ "learning_rate": 0.00019749683936682352,
330
+ "loss": 1.8507,
331
+ "step": 225
332
+ },
333
+ {
334
+ "epoch": 0.84,
335
+ "grad_norm": 0.3046875,
336
+ "learning_rate": 0.00019720567797585817,
337
+ "loss": 1.8928,
338
+ "step": 230
339
+ },
340
+ {
341
+ "epoch": 0.86,
342
+ "grad_norm": 0.296875,
343
+ "learning_rate": 0.00019689874049541673,
344
+ "loss": 1.8767,
345
+ "step": 235
346
+ },
347
+ {
348
+ "epoch": 0.87,
349
+ "grad_norm": 0.318359375,
350
+ "learning_rate": 0.00019657607674021424,
351
+ "loss": 1.9053,
352
+ "step": 240
353
+ },
354
+ {
355
+ "epoch": 0.89,
356
+ "grad_norm": 0.32421875,
357
+ "learning_rate": 0.00019623773907727682,
358
+ "loss": 1.7781,
359
+ "step": 245
360
+ },
361
+ {
362
+ "epoch": 0.91,
363
+ "grad_norm": 0.302734375,
364
+ "learning_rate": 0.00019588378241744254,
365
+ "loss": 1.8099,
366
+ "step": 250
367
+ },
368
+ {
369
+ "epoch": 0.93,
370
+ "grad_norm": 0.302734375,
371
+ "learning_rate": 0.00019551426420644967,
372
+ "loss": 1.8546,
373
+ "step": 255
374
+ },
375
+ {
376
+ "epoch": 0.95,
377
+ "grad_norm": 0.369140625,
378
+ "learning_rate": 0.00019512924441561348,
379
+ "loss": 1.8158,
380
+ "step": 260
381
+ },
382
+ {
383
+ "epoch": 0.97,
384
+ "grad_norm": 0.318359375,
385
+ "learning_rate": 0.00019472878553209312,
386
+ "loss": 1.7982,
387
+ "step": 265
388
+ },
389
+ {
390
+ "epoch": 0.98,
391
+ "grad_norm": 0.39453125,
392
+ "learning_rate": 0.0001943129525487502,
393
+ "loss": 1.7274,
394
+ "step": 270
395
+ },
396
+ {
397
+ "epoch": 1.0,
398
+ "eval_loss": 1.9431506395339966,
399
+ "eval_runtime": 74.7123,
400
+ "eval_samples_per_second": 3.667,
401
+ "eval_steps_per_second": 0.468,
402
+ "step": 274
403
+ },
404
+ {
405
+ "epoch": 1.0,
406
+ "grad_norm": 0.55078125,
407
+ "learning_rate": 0.00019388181295360078,
408
+ "loss": 1.7478,
409
+ "step": 275
410
+ },
411
+ {
412
+ "epoch": 1.02,
413
+ "grad_norm": 0.439453125,
414
+ "learning_rate": 0.00019343543671886218,
415
+ "loss": 1.61,
416
+ "step": 280
417
+ },
418
+ {
419
+ "epoch": 1.04,
420
+ "grad_norm": 0.435546875,
421
+ "learning_rate": 0.00019297389628959702,
422
+ "loss": 1.6174,
423
+ "step": 285
424
+ },
425
+ {
426
+ "epoch": 1.06,
427
+ "grad_norm": 0.412109375,
428
+ "learning_rate": 0.00019249726657195532,
429
+ "loss": 1.6191,
430
+ "step": 290
431
+ },
432
+ {
433
+ "epoch": 1.07,
434
+ "grad_norm": 0.392578125,
435
+ "learning_rate": 0.00019200562492101798,
436
+ "loss": 1.6129,
437
+ "step": 295
438
+ },
439
+ {
440
+ "epoch": 1.09,
441
+ "grad_norm": 0.41015625,
442
+ "learning_rate": 0.00019149905112824197,
443
+ "loss": 1.6052,
444
+ "step": 300
445
+ },
446
+ {
447
+ "epoch": 1.11,
448
+ "grad_norm": 0.44921875,
449
+ "learning_rate": 0.00019097762740851061,
450
+ "loss": 1.4974,
451
+ "step": 305
452
+ },
453
+ {
454
+ "epoch": 1.13,
455
+ "grad_norm": 0.4375,
456
+ "learning_rate": 0.0001904414383867907,
457
+ "loss": 1.598,
458
+ "step": 310
459
+ },
460
+ {
461
+ "epoch": 1.15,
462
+ "grad_norm": 0.4609375,
463
+ "learning_rate": 0.00018989057108439777,
464
+ "loss": 1.5788,
465
+ "step": 315
466
+ },
467
+ {
468
+ "epoch": 1.17,
469
+ "grad_norm": 0.373046875,
470
+ "learning_rate": 0.0001893251149048732,
471
+ "loss": 1.6078,
472
+ "step": 320
473
+ },
474
+ {
475
+ "epoch": 1.18,
476
+ "grad_norm": 0.55078125,
477
+ "learning_rate": 0.0001887451616194743,
478
+ "loss": 1.527,
479
+ "step": 325
480
+ },
481
+ {
482
+ "epoch": 1.2,
483
+ "grad_norm": 0.482421875,
484
+ "learning_rate": 0.0001881508053522801,
485
+ "loss": 1.4852,
486
+ "step": 330
487
+ },
488
+ {
489
+ "epoch": 1.22,
490
+ "grad_norm": 0.458984375,
491
+ "learning_rate": 0.00018754214256491562,
492
+ "loss": 1.5538,
493
+ "step": 335
494
+ },
495
+ {
496
+ "epoch": 1.24,
497
+ "grad_norm": 0.439453125,
498
+ "learning_rate": 0.0001869192720408963,
499
+ "loss": 1.5196,
500
+ "step": 340
501
+ },
502
+ {
503
+ "epoch": 1.26,
504
+ "grad_norm": 0.49609375,
505
+ "learning_rate": 0.00018628229486959596,
506
+ "loss": 1.495,
507
+ "step": 345
508
+ },
509
+ {
510
+ "epoch": 1.28,
511
+ "grad_norm": 0.443359375,
512
+ "learning_rate": 0.00018563131442984044,
513
+ "loss": 1.465,
514
+ "step": 350
515
+ },
516
+ {
517
+ "epoch": 1.29,
518
+ "grad_norm": 0.447265625,
519
+ "learning_rate": 0.00018496643637312942,
520
+ "loss": 1.3923,
521
+ "step": 355
522
+ },
523
+ {
524
+ "epoch": 1.31,
525
+ "grad_norm": 0.455078125,
526
+ "learning_rate": 0.00018428776860648994,
527
+ "loss": 1.5173,
528
+ "step": 360
529
+ },
530
+ {
531
+ "epoch": 1.33,
532
+ "grad_norm": 0.494140625,
533
+ "learning_rate": 0.0001835954212749632,
534
+ "loss": 1.4434,
535
+ "step": 365
536
+ },
537
+ {
538
+ "epoch": 1.35,
539
+ "grad_norm": 0.4765625,
540
+ "learning_rate": 0.0001828895067437287,
541
+ "loss": 1.4619,
542
+ "step": 370
543
+ },
544
+ {
545
+ "epoch": 1.37,
546
+ "grad_norm": 0.55078125,
547
+ "learning_rate": 0.0001821701395798677,
548
+ "loss": 1.3675,
549
+ "step": 375
550
+ },
551
+ {
552
+ "epoch": 1.38,
553
+ "grad_norm": 0.58984375,
554
+ "learning_rate": 0.00018143743653376942,
555
+ "loss": 1.4109,
556
+ "step": 380
557
+ },
558
+ {
559
+ "epoch": 1.4,
560
+ "grad_norm": 0.55078125,
561
+ "learning_rate": 0.0001806915165201831,
562
+ "loss": 1.3592,
563
+ "step": 385
564
+ },
565
+ {
566
+ "epoch": 1.42,
567
+ "grad_norm": 0.466796875,
568
+ "learning_rate": 0.00017993250059891838,
569
+ "loss": 1.3634,
570
+ "step": 390
571
+ },
572
+ {
573
+ "epoch": 1.44,
574
+ "grad_norm": 0.55859375,
575
+ "learning_rate": 0.00017916051195519797,
576
+ "loss": 1.2936,
577
+ "step": 395
578
+ },
579
+ {
580
+ "epoch": 1.46,
581
+ "grad_norm": 0.48828125,
582
+ "learning_rate": 0.00017837567587966507,
583
+ "loss": 1.4582,
584
+ "step": 400
585
+ },
586
+ {
587
+ "epoch": 1.48,
588
+ "grad_norm": 0.5078125,
589
+ "learning_rate": 0.00017757811974804927,
590
+ "loss": 1.3851,
591
+ "step": 405
592
+ },
593
+ {
594
+ "epoch": 1.49,
595
+ "grad_norm": 0.5234375,
596
+ "learning_rate": 0.00017676797300049393,
597
+ "loss": 1.4325,
598
+ "step": 410
599
+ },
600
+ {
601
+ "epoch": 1.51,
602
+ "grad_norm": 0.55859375,
603
+ "learning_rate": 0.0001759453671205487,
604
+ "loss": 1.3731,
605
+ "step": 415
606
+ },
607
+ {
608
+ "epoch": 1.53,
609
+ "grad_norm": 0.44921875,
610
+ "learning_rate": 0.00017511043561382997,
611
+ "loss": 1.5819,
612
+ "step": 420
613
+ },
614
+ {
615
+ "epoch": 1.55,
616
+ "grad_norm": 0.4921875,
617
+ "learning_rate": 0.0001742633139863538,
618
+ "loss": 1.4186,
619
+ "step": 425
620
+ },
621
+ {
622
+ "epoch": 1.57,
623
+ "grad_norm": 0.5859375,
624
+ "learning_rate": 0.00017340413972254367,
625
+ "loss": 1.2765,
626
+ "step": 430
627
+ },
628
+ {
629
+ "epoch": 1.58,
630
+ "grad_norm": 0.498046875,
631
+ "learning_rate": 0.00017253305226291724,
632
+ "loss": 1.2749,
633
+ "step": 435
634
+ },
635
+ {
636
+ "epoch": 1.6,
637
+ "grad_norm": 0.60546875,
638
+ "learning_rate": 0.00017165019298145585,
639
+ "loss": 1.3995,
640
+ "step": 440
641
+ },
642
+ {
643
+ "epoch": 1.62,
644
+ "grad_norm": 0.55078125,
645
+ "learning_rate": 0.00017075570516266012,
646
+ "loss": 1.3697,
647
+ "step": 445
648
+ },
649
+ {
650
+ "epoch": 1.64,
651
+ "grad_norm": 0.58984375,
652
+ "learning_rate": 0.00016984973397829524,
653
+ "loss": 1.2694,
654
+ "step": 450
655
+ },
656
+ {
657
+ "epoch": 1.66,
658
+ "grad_norm": 0.5390625,
659
+ "learning_rate": 0.0001689324264638304,
660
+ "loss": 1.3461,
661
+ "step": 455
662
+ },
663
+ {
664
+ "epoch": 1.68,
665
+ "grad_norm": 0.5859375,
666
+ "learning_rate": 0.00016800393149457539,
667
+ "loss": 1.2316,
668
+ "step": 460
669
+ },
670
+ {
671
+ "epoch": 1.69,
672
+ "grad_norm": 0.6015625,
673
+ "learning_rate": 0.0001670643997615189,
674
+ "loss": 1.2794,
675
+ "step": 465
676
+ },
677
+ {
678
+ "epoch": 1.71,
679
+ "grad_norm": 0.53125,
680
+ "learning_rate": 0.0001661139837468717,
681
+ "loss": 1.185,
682
+ "step": 470
683
+ },
684
+ {
685
+ "epoch": 1.73,
686
+ "grad_norm": 0.58203125,
687
+ "learning_rate": 0.00016515283769931972,
688
+ "loss": 1.2398,
689
+ "step": 475
690
+ },
691
+ {
692
+ "epoch": 1.75,
693
+ "grad_norm": 0.55859375,
694
+ "learning_rate": 0.0001641811176089899,
695
+ "loss": 1.2424,
696
+ "step": 480
697
+ },
698
+ {
699
+ "epoch": 1.77,
700
+ "grad_norm": 0.60546875,
701
+ "learning_rate": 0.00016319898118213365,
702
+ "loss": 1.24,
703
+ "step": 485
704
+ },
705
+ {
706
+ "epoch": 1.79,
707
+ "grad_norm": 0.58984375,
708
+ "learning_rate": 0.00016220658781553176,
709
+ "loss": 1.2045,
710
+ "step": 490
711
+ },
712
+ {
713
+ "epoch": 1.8,
714
+ "grad_norm": 0.60546875,
715
+ "learning_rate": 0.00016120409857062507,
716
+ "loss": 1.2237,
717
+ "step": 495
718
+ },
719
+ {
720
+ "epoch": 1.82,
721
+ "grad_norm": 0.61328125,
722
+ "learning_rate": 0.0001601916761473747,
723
+ "loss": 1.329,
724
+ "step": 500
725
+ },
726
+ {
727
+ "epoch": 1.84,
728
+ "grad_norm": 0.7265625,
729
+ "learning_rate": 0.00015916948485785647,
730
+ "loss": 1.0863,
731
+ "step": 505
732
+ },
733
+ {
734
+ "epoch": 1.86,
735
+ "grad_norm": 0.55078125,
736
+ "learning_rate": 0.0001581376905995939,
737
+ "loss": 1.2565,
738
+ "step": 510
739
+ },
740
+ {
741
+ "epoch": 1.88,
742
+ "grad_norm": 0.59765625,
743
+ "learning_rate": 0.0001570964608286336,
744
+ "loss": 1.127,
745
+ "step": 515
746
+ },
747
+ {
748
+ "epoch": 1.89,
749
+ "grad_norm": 0.625,
750
+ "learning_rate": 0.0001560459645323678,
751
+ "loss": 1.147,
752
+ "step": 520
753
+ },
754
+ {
755
+ "epoch": 1.91,
756
+ "grad_norm": 0.62890625,
757
+ "learning_rate": 0.0001549863722021084,
758
+ "loss": 1.1599,
759
+ "step": 525
760
+ },
761
+ {
762
+ "epoch": 1.93,
763
+ "grad_norm": 0.65234375,
764
+ "learning_rate": 0.00015391785580541698,
765
+ "loss": 1.1667,
766
+ "step": 530
767
+ },
768
+ {
769
+ "epoch": 1.95,
770
+ "grad_norm": 0.6015625,
771
+ "learning_rate": 0.00015284058875819504,
772
+ "loss": 1.2302,
773
+ "step": 535
774
+ },
775
+ {
776
+ "epoch": 1.97,
777
+ "grad_norm": 0.66796875,
778
+ "learning_rate": 0.00015175474589653942,
779
+ "loss": 1.1781,
780
+ "step": 540
781
+ },
782
+ {
783
+ "epoch": 1.99,
784
+ "grad_norm": 0.734375,
785
+ "learning_rate": 0.00015066050344836706,
786
+ "loss": 1.1514,
787
+ "step": 545
788
+ },
789
+ {
790
+ "epoch": 2.0,
791
+ "eval_loss": 1.7110929489135742,
792
+ "eval_runtime": 74.7203,
793
+ "eval_samples_per_second": 3.667,
794
+ "eval_steps_per_second": 0.468,
795
+ "step": 549
796
+ },
797
+ {
798
+ "epoch": 2.0,
799
+ "grad_norm": 0.53515625,
800
+ "learning_rate": 0.00014955803900481389,
801
+ "loss": 1.0657,
802
+ "step": 550
803
+ },
804
+ {
805
+ "epoch": 2.02,
806
+ "grad_norm": 0.8125,
807
+ "learning_rate": 0.0001484475314914125,
808
+ "loss": 0.9519,
809
+ "step": 555
810
+ },
811
+ {
812
+ "epoch": 2.04,
813
+ "grad_norm": 0.66796875,
814
+ "learning_rate": 0.00014732916113905335,
815
+ "loss": 0.8776,
816
+ "step": 560
817
+ },
818
+ {
819
+ "epoch": 2.06,
820
+ "grad_norm": 0.76171875,
821
+ "learning_rate": 0.0001462031094547339,
822
+ "loss": 0.9081,
823
+ "step": 565
824
+ },
825
+ {
826
+ "epoch": 2.08,
827
+ "grad_norm": 0.703125,
828
+ "learning_rate": 0.0001450695591921008,
829
+ "loss": 0.9279,
830
+ "step": 570
831
+ },
832
+ {
833
+ "epoch": 2.09,
834
+ "grad_norm": 0.73828125,
835
+ "learning_rate": 0.00014392869432178971,
836
+ "loss": 0.7395,
837
+ "step": 575
838
+ },
839
+ {
840
+ "epoch": 2.11,
841
+ "grad_norm": 0.671875,
842
+ "learning_rate": 0.00014278070000156768,
843
+ "loss": 0.9795,
844
+ "step": 580
845
+ },
846
+ {
847
+ "epoch": 2.13,
848
+ "grad_norm": 0.72265625,
849
+ "learning_rate": 0.00014162576254628267,
850
+ "loss": 0.8563,
851
+ "step": 585
852
+ },
853
+ {
854
+ "epoch": 2.15,
855
+ "grad_norm": 0.73046875,
856
+ "learning_rate": 0.00014046406939762545,
857
+ "loss": 0.8419,
858
+ "step": 590
859
+ },
860
+ {
861
+ "epoch": 2.17,
862
+ "grad_norm": 0.70703125,
863
+ "learning_rate": 0.00013929580909370862,
864
+ "loss": 0.82,
865
+ "step": 595
866
+ },
867
+ {
868
+ "epoch": 2.19,
869
+ "grad_norm": 0.67578125,
870
+ "learning_rate": 0.00013812117123846747,
871
+ "loss": 0.9267,
872
+ "step": 600
873
+ },
874
+ {
875
+ "epoch": 2.2,
876
+ "grad_norm": 0.69921875,
877
+ "learning_rate": 0.0001369403464708884,
878
+ "loss": 0.8248,
879
+ "step": 605
880
+ },
881
+ {
882
+ "epoch": 2.22,
883
+ "grad_norm": 0.73046875,
884
+ "learning_rate": 0.00013575352643406844,
885
+ "loss": 0.8331,
886
+ "step": 610
887
+ },
888
+ {
889
+ "epoch": 2.24,
890
+ "grad_norm": 0.69140625,
891
+ "learning_rate": 0.00013456090374411284,
892
+ "loss": 0.8908,
893
+ "step": 615
894
+ },
895
+ {
896
+ "epoch": 2.26,
897
+ "grad_norm": 0.734375,
898
+ "learning_rate": 0.00013336267195887398,
899
+ "loss": 0.8693,
900
+ "step": 620
901
+ },
902
+ {
903
+ "epoch": 2.28,
904
+ "grad_norm": 0.953125,
905
+ "learning_rate": 0.00013215902554653772,
906
+ "loss": 0.8886,
907
+ "step": 625
908
+ },
909
+ {
910
+ "epoch": 2.3,
911
+ "grad_norm": 0.8046875,
912
+ "learning_rate": 0.00013095015985406215,
913
+ "loss": 0.7254,
914
+ "step": 630
915
+ },
916
+ {
917
+ "epoch": 2.31,
918
+ "grad_norm": 0.7421875,
919
+ "learning_rate": 0.00012973627107547346,
920
+ "loss": 0.8929,
921
+ "step": 635
922
+ },
923
+ {
924
+ "epoch": 2.33,
925
+ "grad_norm": 0.8671875,
926
+ "learning_rate": 0.00012851755622002444,
927
+ "loss": 0.7416,
928
+ "step": 640
929
+ },
930
+ {
931
+ "epoch": 2.35,
932
+ "grad_norm": 0.6796875,
933
+ "learning_rate": 0.0001272942130802207,
934
+ "loss": 0.7426,
935
+ "step": 645
936
+ },
937
+ {
938
+ "epoch": 2.37,
939
+ "grad_norm": 0.73046875,
940
+ "learning_rate": 0.00012606644019971968,
941
+ "loss": 0.7973,
942
+ "step": 650
943
+ },
944
+ {
945
+ "epoch": 2.39,
946
+ "grad_norm": 0.7265625,
947
+ "learning_rate": 0.00012483443684110793,
948
+ "loss": 0.7846,
949
+ "step": 655
950
+ },
951
+ {
952
+ "epoch": 2.4,
953
+ "grad_norm": 0.7265625,
954
+ "learning_rate": 0.00012359840295356151,
955
+ "loss": 0.9146,
956
+ "step": 660
957
+ },
958
+ {
959
+ "epoch": 2.42,
960
+ "grad_norm": 0.734375,
961
+ "learning_rate": 0.00012235853914039515,
962
+ "loss": 0.8241,
963
+ "step": 665
964
+ },
965
+ {
966
+ "epoch": 2.44,
967
+ "grad_norm": 0.6640625,
968
+ "learning_rate": 0.0001211150466265049,
969
+ "loss": 0.811,
970
+ "step": 670
971
+ },
972
+ {
973
+ "epoch": 2.46,
974
+ "grad_norm": 0.66015625,
975
+ "learning_rate": 0.00011986812722571039,
976
+ "loss": 0.7336,
977
+ "step": 675
978
+ },
979
+ {
980
+ "epoch": 2.48,
981
+ "grad_norm": 0.6875,
982
+ "learning_rate": 0.00011861798330800125,
983
+ "loss": 0.727,
984
+ "step": 680
985
+ },
986
+ {
987
+ "epoch": 2.5,
988
+ "grad_norm": 0.7578125,
989
+ "learning_rate": 0.00011736481776669306,
990
+ "loss": 0.6652,
991
+ "step": 685
992
+ },
993
+ {
994
+ "epoch": 2.51,
995
+ "grad_norm": 0.625,
996
+ "learning_rate": 0.00011610883398549879,
997
+ "loss": 0.6868,
998
+ "step": 690
999
+ },
1000
+ {
1001
+ "epoch": 2.53,
1002
+ "grad_norm": 0.734375,
1003
+ "learning_rate": 0.00011485023580552039,
1004
+ "loss": 0.7602,
1005
+ "step": 695
1006
+ },
1007
+ {
1008
+ "epoch": 2.55,
1009
+ "grad_norm": 0.73046875,
1010
+ "learning_rate": 0.00011358922749216614,
1011
+ "loss": 0.7176,
1012
+ "step": 700
1013
+ },
1014
+ {
1015
+ "epoch": 2.57,
1016
+ "grad_norm": 0.73828125,
1017
+ "learning_rate": 0.00011232601370199936,
1018
+ "loss": 0.7167,
1019
+ "step": 705
1020
+ },
1021
+ {
1022
+ "epoch": 2.59,
1023
+ "grad_norm": 0.76953125,
1024
+ "learning_rate": 0.00011106079944952317,
1025
+ "loss": 0.7278,
1026
+ "step": 710
1027
+ },
1028
+ {
1029
+ "epoch": 2.6,
1030
+ "grad_norm": 0.75390625,
1031
+ "learning_rate": 0.00010979379007390782,
1032
+ "loss": 0.778,
1033
+ "step": 715
1034
+ },
1035
+ {
1036
+ "epoch": 2.62,
1037
+ "grad_norm": 0.73046875,
1038
+ "learning_rate": 0.00010852519120566475,
1039
+ "loss": 0.7099,
1040
+ "step": 720
1041
+ },
1042
+ {
1043
+ "epoch": 2.64,
1044
+ "grad_norm": 0.671875,
1045
+ "learning_rate": 0.00010725520873327361,
1046
+ "loss": 0.781,
1047
+ "step": 725
1048
+ },
1049
+ {
1050
+ "epoch": 2.66,
1051
+ "grad_norm": 0.71875,
1052
+ "learning_rate": 0.00010598404876976749,
1053
+ "loss": 0.6719,
1054
+ "step": 730
1055
+ },
1056
+ {
1057
+ "epoch": 2.68,
1058
+ "grad_norm": 0.7890625,
1059
+ "learning_rate": 0.00010471191761928146,
1060
+ "loss": 0.8674,
1061
+ "step": 735
1062
+ },
1063
+ {
1064
+ "epoch": 2.7,
1065
+ "grad_norm": 0.73828125,
1066
+ "learning_rate": 0.00010343902174357039,
1067
+ "loss": 0.6813,
1068
+ "step": 740
1069
+ },
1070
+ {
1071
+ "epoch": 2.71,
1072
+ "grad_norm": 0.703125,
1073
+ "learning_rate": 0.00010216556772850094,
1074
+ "loss": 0.795,
1075
+ "step": 745
1076
+ },
1077
+ {
1078
+ "epoch": 2.73,
1079
+ "grad_norm": 0.7109375,
1080
+ "learning_rate": 0.00010089176225052354,
1081
+ "loss": 0.6741,
1082
+ "step": 750
1083
+ },
1084
+ {
1085
+ "epoch": 2.75,
1086
+ "grad_norm": 0.71484375,
1087
+ "learning_rate": 9.96178120431296e-05,
1088
+ "loss": 0.6977,
1089
+ "step": 755
1090
+ },
1091
+ {
1092
+ "epoch": 2.77,
1093
+ "grad_norm": 0.734375,
1094
+ "learning_rate": 9.83439238632996e-05,
1095
+ "loss": 0.6683,
1096
+ "step": 760
1097
+ },
1098
+ {
1099
+ "epoch": 2.79,
1100
+ "grad_norm": 0.65625,
1101
+ "learning_rate": 9.707030445794715e-05,
1102
+ "loss": 0.6262,
1103
+ "step": 765
1104
+ },
1105
+ {
1106
+ "epoch": 2.81,
1107
+ "grad_norm": 0.78125,
1108
+ "learning_rate": 9.579716053036479e-05,
1109
+ "loss": 0.6627,
1110
+ "step": 770
1111
+ },
1112
+ {
1113
+ "epoch": 2.82,
1114
+ "grad_norm": 0.765625,
1115
+ "learning_rate": 9.452469870667713e-05,
1116
+ "loss": 0.7535,
1117
+ "step": 775
1118
+ },
1119
+ {
1120
+ "epoch": 2.84,
1121
+ "grad_norm": 0.6640625,
1122
+ "learning_rate": 9.325312550230581e-05,
1123
+ "loss": 0.7468,
1124
+ "step": 780
1125
+ },
1126
+ {
1127
+ "epoch": 2.86,
1128
+ "grad_norm": 0.59765625,
1129
+ "learning_rate": 9.198264728845332e-05,
1130
+ "loss": 0.6141,
1131
+ "step": 785
1132
+ },
1133
+ {
1134
+ "epoch": 2.88,
1135
+ "grad_norm": 0.8125,
1136
+ "learning_rate": 9.071347025860958e-05,
1137
+ "loss": 0.6955,
1138
+ "step": 790
1139
+ },
1140
+ {
1141
+ "epoch": 2.9,
1142
+ "grad_norm": 0.65234375,
1143
+ "learning_rate": 8.944580039508748e-05,
1144
+ "loss": 0.6839,
1145
+ "step": 795
1146
+ },
1147
+ {
1148
+ "epoch": 2.91,
1149
+ "grad_norm": 0.71875,
1150
+ "learning_rate": 8.81798434355931e-05,
1151
+ "loss": 0.6004,
1152
+ "step": 800
1153
+ },
1154
+ {
1155
+ "epoch": 2.93,
1156
+ "grad_norm": 0.69140625,
1157
+ "learning_rate": 8.69158048398349e-05,
1158
+ "loss": 0.7424,
1159
+ "step": 805
1160
+ },
1161
+ {
1162
+ "epoch": 2.95,
1163
+ "grad_norm": 0.703125,
1164
+ "learning_rate": 8.565388975617873e-05,
1165
+ "loss": 0.6301,
1166
+ "step": 810
1167
+ },
1168
+ {
1169
+ "epoch": 2.97,
1170
+ "grad_norm": 0.6875,
1171
+ "learning_rate": 8.439430298835304e-05,
1172
+ "loss": 0.74,
1173
+ "step": 815
1174
+ },
1175
+ {
1176
+ "epoch": 2.99,
1177
+ "grad_norm": 0.7578125,
1178
+ "learning_rate": 8.313724896220994e-05,
1179
+ "loss": 0.645,
1180
+ "step": 820
1181
+ },
1182
+ {
1183
+ "epoch": 3.0,
1184
+ "eval_loss": 1.5109344720840454,
1185
+ "eval_runtime": 74.6963,
1186
+ "eval_samples_per_second": 3.668,
1187
+ "eval_steps_per_second": 0.469,
1188
+ "step": 823
1189
+ },
1190
+ {
1191
+ "epoch": 3.01,
1192
+ "grad_norm": 0.59765625,
1193
+ "learning_rate": 8.188293169254775e-05,
1194
+ "loss": 0.5792,
1195
+ "step": 825
1196
+ },
1197
+ {
1198
+ "epoch": 3.02,
1199
+ "grad_norm": 0.8359375,
1200
+ "learning_rate": 8.063155475000037e-05,
1201
+ "loss": 0.6678,
1202
+ "step": 830
1203
+ },
1204
+ {
1205
+ "epoch": 3.04,
1206
+ "grad_norm": 0.71875,
1207
+ "learning_rate": 7.938332122799834e-05,
1208
+ "loss": 0.3506,
1209
+ "step": 835
1210
+ },
1211
+ {
1212
+ "epoch": 3.06,
1213
+ "grad_norm": 0.828125,
1214
+ "learning_rate": 7.813843370980787e-05,
1215
+ "loss": 0.4464,
1216
+ "step": 840
1217
+ },
1218
+ {
1219
+ "epoch": 3.08,
1220
+ "grad_norm": 0.6640625,
1221
+ "learning_rate": 7.68970942356521e-05,
1222
+ "loss": 0.4639,
1223
+ "step": 845
1224
+ },
1225
+ {
1226
+ "epoch": 3.1,
1227
+ "grad_norm": 0.66015625,
1228
+ "learning_rate": 7.565950426992096e-05,
1229
+ "loss": 0.4815,
1230
+ "step": 850
1231
+ },
1232
+ {
1233
+ "epoch": 3.11,
1234
+ "grad_norm": 0.6953125,
1235
+ "learning_rate": 7.442586466847441e-05,
1236
+ "loss": 0.4089,
1237
+ "step": 855
1238
+ },
1239
+ {
1240
+ "epoch": 3.13,
1241
+ "grad_norm": 0.67578125,
1242
+ "learning_rate": 7.319637564604412e-05,
1243
+ "loss": 0.4647,
1244
+ "step": 860
1245
+ },
1246
+ {
1247
+ "epoch": 3.15,
1248
+ "grad_norm": 0.734375,
1249
+ "learning_rate": 7.197123674373942e-05,
1250
+ "loss": 0.4837,
1251
+ "step": 865
1252
+ },
1253
+ {
1254
+ "epoch": 3.17,
1255
+ "grad_norm": 0.71875,
1256
+ "learning_rate": 7.075064679666294e-05,
1257
+ "loss": 0.5208,
1258
+ "step": 870
1259
+ },
1260
+ {
1261
+ "epoch": 3.19,
1262
+ "grad_norm": 0.703125,
1263
+ "learning_rate": 6.953480390164e-05,
1264
+ "loss": 0.4568,
1265
+ "step": 875
1266
+ },
1267
+ {
1268
+ "epoch": 3.21,
1269
+ "grad_norm": 0.7890625,
1270
+ "learning_rate": 6.832390538506868e-05,
1271
+ "loss": 0.5692,
1272
+ "step": 880
1273
+ },
1274
+ {
1275
+ "epoch": 3.22,
1276
+ "grad_norm": 0.72265625,
1277
+ "learning_rate": 6.711814777089427e-05,
1278
+ "loss": 0.4652,
1279
+ "step": 885
1280
+ },
1281
+ {
1282
+ "epoch": 3.24,
1283
+ "grad_norm": 0.74609375,
1284
+ "learning_rate": 6.591772674871434e-05,
1285
+ "loss": 0.5665,
1286
+ "step": 890
1287
+ },
1288
+ {
1289
+ "epoch": 3.26,
1290
+ "grad_norm": 0.69921875,
1291
+ "learning_rate": 6.472283714201919e-05,
1292
+ "loss": 0.5147,
1293
+ "step": 895
1294
+ },
1295
+ {
1296
+ "epoch": 3.28,
1297
+ "grad_norm": 0.70703125,
1298
+ "learning_rate": 6.353367287657262e-05,
1299
+ "loss": 0.5033,
1300
+ "step": 900
1301
+ },
1302
+ {
1303
+ "epoch": 3.3,
1304
+ "grad_norm": 0.75,
1305
+ "learning_rate": 6.235042694893862e-05,
1306
+ "loss": 0.5159,
1307
+ "step": 905
1308
+ },
1309
+ {
1310
+ "epoch": 3.32,
1311
+ "grad_norm": 0.59765625,
1312
+ "learning_rate": 6.117329139515894e-05,
1313
+ "loss": 0.4865,
1314
+ "step": 910
1315
+ },
1316
+ {
1317
+ "epoch": 3.33,
1318
+ "grad_norm": 0.6953125,
1319
+ "learning_rate": 6.000245725958618e-05,
1320
+ "loss": 0.4963,
1321
+ "step": 915
1322
+ },
1323
+ {
1324
+ "epoch": 3.35,
1325
+ "grad_norm": 0.69140625,
1326
+ "learning_rate": 5.883811456387821e-05,
1327
+ "loss": 0.4766,
1328
+ "step": 920
1329
+ },
1330
+ {
1331
+ "epoch": 3.37,
1332
+ "grad_norm": 0.671875,
1333
+ "learning_rate": 5.7680452276158457e-05,
1334
+ "loss": 0.4681,
1335
+ "step": 925
1336
+ },
1337
+ {
1338
+ "epoch": 3.39,
1339
+ "grad_norm": 0.73046875,
1340
+ "learning_rate": 5.6529658280346906e-05,
1341
+ "loss": 0.4764,
1342
+ "step": 930
1343
+ },
1344
+ {
1345
+ "epoch": 3.41,
1346
+ "grad_norm": 0.62109375,
1347
+ "learning_rate": 5.5385919345667715e-05,
1348
+ "loss": 0.4547,
1349
+ "step": 935
1350
+ },
1351
+ {
1352
+ "epoch": 3.42,
1353
+ "grad_norm": 0.6953125,
1354
+ "learning_rate": 5.424942109633692e-05,
1355
+ "loss": 0.5382,
1356
+ "step": 940
1357
+ },
1358
+ {
1359
+ "epoch": 3.44,
1360
+ "grad_norm": 0.85546875,
1361
+ "learning_rate": 5.3120347981436444e-05,
1362
+ "loss": 0.4404,
1363
+ "step": 945
1364
+ },
1365
+ {
1366
+ "epoch": 3.46,
1367
+ "grad_norm": 0.703125,
1368
+ "learning_rate": 5.199888324497907e-05,
1369
+ "loss": 0.5514,
1370
+ "step": 950
1371
+ },
1372
+ {
1373
+ "epoch": 3.48,
1374
+ "grad_norm": 0.60546875,
1375
+ "learning_rate": 5.088520889616828e-05,
1376
+ "loss": 0.5287,
1377
+ "step": 955
1378
+ },
1379
+ {
1380
+ "epoch": 3.5,
1381
+ "grad_norm": 0.72265625,
1382
+ "learning_rate": 4.9779505679859096e-05,
1383
+ "loss": 0.4759,
1384
+ "step": 960
1385
+ },
1386
+ {
1387
+ "epoch": 3.52,
1388
+ "grad_norm": 0.73828125,
1389
+ "learning_rate": 4.8681953047223914e-05,
1390
+ "loss": 0.5279,
1391
+ "step": 965
1392
+ },
1393
+ {
1394
+ "epoch": 3.53,
1395
+ "grad_norm": 0.71484375,
1396
+ "learning_rate": 4.759272912662822e-05,
1397
+ "loss": 0.4184,
1398
+ "step": 970
1399
+ },
1400
+ {
1401
+ "epoch": 3.55,
1402
+ "grad_norm": 0.57421875,
1403
+ "learning_rate": 4.651201069472125e-05,
1404
+ "loss": 0.4301,
1405
+ "step": 975
1406
+ },
1407
+ {
1408
+ "epoch": 3.57,
1409
+ "grad_norm": 0.734375,
1410
+ "learning_rate": 4.543997314774553e-05,
1411
+ "loss": 0.4781,
1412
+ "step": 980
1413
+ },
1414
+ {
1415
+ "epoch": 3.59,
1416
+ "grad_norm": 0.64453125,
1417
+ "learning_rate": 4.437679047307116e-05,
1418
+ "loss": 0.5563,
1419
+ "step": 985
1420
+ },
1421
+ {
1422
+ "epoch": 3.61,
1423
+ "grad_norm": 0.6640625,
1424
+ "learning_rate": 4.332263522095815e-05,
1425
+ "loss": 0.4216,
1426
+ "step": 990
1427
+ },
1428
+ {
1429
+ "epoch": 3.62,
1430
+ "grad_norm": 0.72265625,
1431
+ "learning_rate": 4.227767847655205e-05,
1432
+ "loss": 0.5113,
1433
+ "step": 995
1434
+ },
1435
+ {
1436
+ "epoch": 3.64,
1437
+ "grad_norm": 0.6015625,
1438
+ "learning_rate": 4.124208983211799e-05,
1439
+ "loss": 0.5323,
1440
+ "step": 1000
1441
+ },
1442
+ {
1443
+ "epoch": 3.66,
1444
+ "grad_norm": 0.71875,
1445
+ "learning_rate": 4.021603735951608e-05,
1446
+ "loss": 0.5079,
1447
+ "step": 1005
1448
+ },
1449
+ {
1450
+ "epoch": 3.68,
1451
+ "grad_norm": 0.703125,
1452
+ "learning_rate": 3.919968758292425e-05,
1453
+ "loss": 0.4655,
1454
+ "step": 1010
1455
+ },
1456
+ {
1457
+ "epoch": 3.7,
1458
+ "grad_norm": 0.71875,
1459
+ "learning_rate": 3.8193205451811995e-05,
1460
+ "loss": 0.5073,
1461
+ "step": 1015
1462
+ },
1463
+ {
1464
+ "epoch": 3.72,
1465
+ "grad_norm": 0.609375,
1466
+ "learning_rate": 3.719675431416972e-05,
1467
+ "loss": 0.4882,
1468
+ "step": 1020
1469
+ },
1470
+ {
1471
+ "epoch": 3.73,
1472
+ "grad_norm": 0.640625,
1473
+ "learning_rate": 3.62104958899982e-05,
1474
+ "loss": 0.4177,
1475
+ "step": 1025
1476
+ },
1477
+ {
1478
+ "epoch": 3.75,
1479
+ "grad_norm": 0.671875,
1480
+ "learning_rate": 3.5234590245061684e-05,
1481
+ "loss": 0.5276,
1482
+ "step": 1030
1483
+ },
1484
+ {
1485
+ "epoch": 3.77,
1486
+ "grad_norm": 0.69921875,
1487
+ "learning_rate": 3.426919576491024e-05,
1488
+ "loss": 0.357,
1489
+ "step": 1035
1490
+ },
1491
+ {
1492
+ "epoch": 3.79,
1493
+ "grad_norm": 0.671875,
1494
+ "learning_rate": 3.3314469129174364e-05,
1495
+ "loss": 0.4146,
1496
+ "step": 1040
1497
+ },
1498
+ {
1499
+ "epoch": 3.81,
1500
+ "grad_norm": 0.64453125,
1501
+ "learning_rate": 3.237056528613624e-05,
1502
+ "loss": 0.3781,
1503
+ "step": 1045
1504
+ },
1505
+ {
1506
+ "epoch": 3.83,
1507
+ "grad_norm": 0.6328125,
1508
+ "learning_rate": 3.1437637427582564e-05,
1509
+ "loss": 0.4652,
1510
+ "step": 1050
1511
+ },
1512
+ {
1513
+ "epoch": 3.84,
1514
+ "grad_norm": 0.67578125,
1515
+ "learning_rate": 3.0515836963942056e-05,
1516
+ "loss": 0.5685,
1517
+ "step": 1055
1518
+ },
1519
+ {
1520
+ "epoch": 3.86,
1521
+ "grad_norm": 0.703125,
1522
+ "learning_rate": 2.960531349971204e-05,
1523
+ "loss": 0.5038,
1524
+ "step": 1060
1525
+ },
1526
+ {
1527
+ "epoch": 3.88,
1528
+ "grad_norm": 0.62109375,
1529
+ "learning_rate": 2.8706214809178355e-05,
1530
+ "loss": 0.4237,
1531
+ "step": 1065
1532
+ },
1533
+ {
1534
+ "epoch": 3.9,
1535
+ "grad_norm": 0.765625,
1536
+ "learning_rate": 2.7818686812432136e-05,
1537
+ "loss": 0.6172,
1538
+ "step": 1070
1539
+ },
1540
+ {
1541
+ "epoch": 3.92,
1542
+ "grad_norm": 0.72265625,
1543
+ "learning_rate": 2.6942873551687587e-05,
1544
+ "loss": 0.5244,
1545
+ "step": 1075
1546
+ },
1547
+ {
1548
+ "epoch": 3.93,
1549
+ "grad_norm": 0.62890625,
1550
+ "learning_rate": 2.6078917167904527e-05,
1551
+ "loss": 0.428,
1552
+ "step": 1080
1553
+ },
1554
+ {
1555
+ "epoch": 3.95,
1556
+ "grad_norm": 0.640625,
1557
+ "learning_rate": 2.5226957877719436e-05,
1558
+ "loss": 0.3934,
1559
+ "step": 1085
1560
+ },
1561
+ {
1562
+ "epoch": 3.97,
1563
+ "grad_norm": 0.64453125,
1564
+ "learning_rate": 2.4387133950688912e-05,
1565
+ "loss": 0.5376,
1566
+ "step": 1090
1567
+ },
1568
+ {
1569
+ "epoch": 3.99,
1570
+ "grad_norm": 0.62109375,
1571
+ "learning_rate": 2.3559581686849065e-05,
1572
+ "loss": 0.4291,
1573
+ "step": 1095
1574
+ },
1575
+ {
1576
+ "epoch": 4.0,
1577
+ "eval_loss": 1.4415178298950195,
1578
+ "eval_runtime": 74.7485,
1579
+ "eval_samples_per_second": 3.666,
1580
+ "eval_steps_per_second": 0.468,
1581
+ "step": 1098
1582
+ },
1583
+ {
1584
+ "epoch": 4.01,
1585
+ "grad_norm": 0.60546875,
1586
+ "learning_rate": 2.2744435394594497e-05,
1587
+ "loss": 0.4588,
1588
+ "step": 1100
1589
+ },
1590
+ {
1591
+ "epoch": 4.03,
1592
+ "grad_norm": 0.484375,
1593
+ "learning_rate": 2.1941827368880796e-05,
1594
+ "loss": 0.3858,
1595
+ "step": 1105
1596
+ },
1597
+ {
1598
+ "epoch": 4.04,
1599
+ "grad_norm": 0.6171875,
1600
+ "learning_rate": 2.115188786975336e-05,
1601
+ "loss": 0.4656,
1602
+ "step": 1110
1603
+ },
1604
+ {
1605
+ "epoch": 4.06,
1606
+ "grad_norm": 0.50390625,
1607
+ "learning_rate": 2.037474510120676e-05,
1608
+ "loss": 0.3973,
1609
+ "step": 1115
1610
+ },
1611
+ {
1612
+ "epoch": 4.08,
1613
+ "grad_norm": 0.6015625,
1614
+ "learning_rate": 1.9610525190377983e-05,
1615
+ "loss": 0.4184,
1616
+ "step": 1120
1617
+ },
1618
+ {
1619
+ "epoch": 4.1,
1620
+ "grad_norm": 0.61328125,
1621
+ "learning_rate": 1.885935216707617e-05,
1622
+ "loss": 0.3977,
1623
+ "step": 1125
1624
+ },
1625
+ {
1626
+ "epoch": 4.12,
1627
+ "grad_norm": 0.70703125,
1628
+ "learning_rate": 1.8121347943653332e-05,
1629
+ "loss": 0.4274,
1630
+ "step": 1130
1631
+ },
1632
+ {
1633
+ "epoch": 4.13,
1634
+ "grad_norm": 0.5859375,
1635
+ "learning_rate": 1.7396632295218317e-05,
1636
+ "loss": 0.326,
1637
+ "step": 1135
1638
+ },
1639
+ {
1640
+ "epoch": 4.15,
1641
+ "grad_norm": 0.625,
1642
+ "learning_rate": 1.668532284019789e-05,
1643
+ "loss": 0.4174,
1644
+ "step": 1140
1645
+ },
1646
+ {
1647
+ "epoch": 4.17,
1648
+ "grad_norm": 0.67578125,
1649
+ "learning_rate": 1.5987535021247667e-05,
1650
+ "loss": 0.3984,
1651
+ "step": 1145
1652
+ },
1653
+ {
1654
+ "epoch": 4.19,
1655
+ "grad_norm": 0.60546875,
1656
+ "learning_rate": 1.530338208651627e-05,
1657
+ "loss": 0.5323,
1658
+ "step": 1150
1659
+ },
1660
+ {
1661
+ "epoch": 4.21,
1662
+ "grad_norm": 0.67578125,
1663
+ "learning_rate": 1.463297507126562e-05,
1664
+ "loss": 0.398,
1665
+ "step": 1155
1666
+ },
1667
+ {
1668
+ "epoch": 4.23,
1669
+ "grad_norm": 0.703125,
1670
+ "learning_rate": 1.3976422779850384e-05,
1671
+ "loss": 0.3888,
1672
+ "step": 1160
1673
+ },
1674
+ {
1675
+ "epoch": 4.24,
1676
+ "grad_norm": 0.56640625,
1677
+ "learning_rate": 1.3333831768059301e-05,
1678
+ "loss": 0.3988,
1679
+ "step": 1165
1680
+ },
1681
+ {
1682
+ "epoch": 4.26,
1683
+ "grad_norm": 0.625,
1684
+ "learning_rate": 1.270530632582191e-05,
1685
+ "loss": 0.3928,
1686
+ "step": 1170
1687
+ },
1688
+ {
1689
+ "epoch": 4.28,
1690
+ "grad_norm": 0.61328125,
1691
+ "learning_rate": 1.2090948460282414e-05,
1692
+ "loss": 0.4208,
1693
+ "step": 1175
1694
+ },
1695
+ {
1696
+ "epoch": 4.3,
1697
+ "grad_norm": 0.58984375,
1698
+ "learning_rate": 1.149085787924451e-05,
1699
+ "loss": 0.4415,
1700
+ "step": 1180
1701
+ },
1702
+ {
1703
+ "epoch": 4.32,
1704
+ "grad_norm": 0.7109375,
1705
+ "learning_rate": 1.0905131974989158e-05,
1706
+ "loss": 0.3831,
1707
+ "step": 1185
1708
+ },
1709
+ {
1710
+ "epoch": 4.34,
1711
+ "grad_norm": 0.6640625,
1712
+ "learning_rate": 1.0333865808468202e-05,
1713
+ "loss": 0.4559,
1714
+ "step": 1190
1715
+ },
1716
+ {
1717
+ "epoch": 4.35,
1718
+ "grad_norm": 0.5625,
1719
+ "learning_rate": 9.777152093876396e-06,
1720
+ "loss": 0.3927,
1721
+ "step": 1195
1722
+ },
1723
+ {
1724
+ "epoch": 4.37,
1725
+ "grad_norm": 0.6171875,
1726
+ "learning_rate": 9.235081183604233e-06,
1727
+ "loss": 0.4036,
1728
+ "step": 1200
1729
+ },
1730
+ {
1731
+ "epoch": 4.39,
1732
+ "grad_norm": 0.5234375,
1733
+ "learning_rate": 8.70774105357407e-06,
1734
+ "loss": 0.3153,
1735
+ "step": 1205
1736
+ },
1737
+ {
1738
+ "epoch": 4.41,
1739
+ "grad_norm": 0.51953125,
1740
+ "learning_rate": 8.195217288962175e-06,
1741
+ "loss": 0.2749,
1742
+ "step": 1210
1743
+ },
1744
+ {
1745
+ "epoch": 4.43,
1746
+ "grad_norm": 0.71484375,
1747
+ "learning_rate": 7.697593070308317e-06,
1748
+ "loss": 0.4395,
1749
+ "step": 1215
1750
+ },
1751
+ {
1752
+ "epoch": 4.44,
1753
+ "grad_norm": 0.6796875,
1754
+ "learning_rate": 7.214949160016115e-06,
1755
+ "loss": 0.4126,
1756
+ "step": 1220
1757
+ },
1758
+ {
1759
+ "epoch": 4.46,
1760
+ "grad_norm": 0.67578125,
1761
+ "learning_rate": 6.747363889245606e-06,
1762
+ "loss": 0.4111,
1763
+ "step": 1225
1764
+ },
1765
+ {
1766
+ "epoch": 4.48,
1767
+ "grad_norm": 0.6328125,
1768
+ "learning_rate": 6.294913145200321e-06,
1769
+ "loss": 0.3986,
1770
+ "step": 1230
1771
+ },
1772
+ {
1773
+ "epoch": 4.5,
1774
+ "grad_norm": 0.625,
1775
+ "learning_rate": 5.857670358811096e-06,
1776
+ "loss": 0.4057,
1777
+ "step": 1235
1778
+ },
1779
+ {
1780
+ "epoch": 4.52,
1781
+ "grad_norm": 0.71875,
1782
+ "learning_rate": 5.435706492818693e-06,
1783
+ "loss": 0.478,
1784
+ "step": 1240
1785
+ },
1786
+ {
1787
+ "epoch": 4.54,
1788
+ "grad_norm": 0.640625,
1789
+ "learning_rate": 5.0290900302566335e-06,
1790
+ "loss": 0.449,
1791
+ "step": 1245
1792
+ },
1793
+ {
1794
+ "epoch": 4.55,
1795
+ "grad_norm": 0.70703125,
1796
+ "learning_rate": 4.63788696333678e-06,
1797
+ "loss": 0.4115,
1798
+ "step": 1250
1799
+ },
1800
+ {
1801
+ "epoch": 4.57,
1802
+ "grad_norm": 0.52734375,
1803
+ "learning_rate": 4.262160782739067e-06,
1804
+ "loss": 0.3495,
1805
+ "step": 1255
1806
+ },
1807
+ {
1808
+ "epoch": 4.59,
1809
+ "grad_norm": 0.640625,
1810
+ "learning_rate": 3.901972467307191e-06,
1811
+ "loss": 0.3864,
1812
+ "step": 1260
1813
+ },
1814
+ {
1815
+ "epoch": 4.61,
1816
+ "grad_norm": 0.69921875,
1817
+ "learning_rate": 3.5573804741519833e-06,
1818
+ "loss": 0.4566,
1819
+ "step": 1265
1820
+ },
1821
+ {
1822
+ "epoch": 4.63,
1823
+ "grad_norm": 0.51171875,
1824
+ "learning_rate": 3.2284407291641174e-06,
1825
+ "loss": 0.4479,
1826
+ "step": 1270
1827
+ },
1828
+ {
1829
+ "epoch": 4.64,
1830
+ "grad_norm": 0.66015625,
1831
+ "learning_rate": 2.9152066179374603e-06,
1832
+ "loss": 0.4232,
1833
+ "step": 1275
1834
+ },
1835
+ {
1836
+ "epoch": 4.66,
1837
+ "grad_norm": 0.6328125,
1838
+ "learning_rate": 2.6177289771049274e-06,
1839
+ "loss": 0.3441,
1840
+ "step": 1280
1841
+ },
1842
+ {
1843
+ "epoch": 4.68,
1844
+ "grad_norm": 0.68359375,
1845
+ "learning_rate": 2.33605608608779e-06,
1846
+ "loss": 0.3774,
1847
+ "step": 1285
1848
+ },
1849
+ {
1850
+ "epoch": 4.7,
1851
+ "grad_norm": 0.70703125,
1852
+ "learning_rate": 2.0702336592602122e-06,
1853
+ "loss": 0.4859,
1854
+ "step": 1290
1855
+ },
1856
+ {
1857
+ "epoch": 4.72,
1858
+ "grad_norm": 0.625,
1859
+ "learning_rate": 1.8203048385299181e-06,
1860
+ "loss": 0.4231,
1861
+ "step": 1295
1862
+ },
1863
+ {
1864
+ "epoch": 4.74,
1865
+ "grad_norm": 0.671875,
1866
+ "learning_rate": 1.5863101863364483e-06,
1867
+ "loss": 0.441,
1868
+ "step": 1300
1869
+ },
1870
+ {
1871
+ "epoch": 4.75,
1872
+ "grad_norm": 0.734375,
1873
+ "learning_rate": 1.3682876790681254e-06,
1874
+ "loss": 0.3982,
1875
+ "step": 1305
1876
+ },
1877
+ {
1878
+ "epoch": 4.77,
1879
+ "grad_norm": 0.71875,
1880
+ "learning_rate": 1.1662727008984964e-06,
1881
+ "loss": 0.4582,
1882
+ "step": 1310
1883
+ },
1884
+ {
1885
+ "epoch": 4.79,
1886
+ "grad_norm": 0.5625,
1887
+ "learning_rate": 9.802980380437588e-07,
1888
+ "loss": 0.3836,
1889
+ "step": 1315
1890
+ },
1891
+ {
1892
+ "epoch": 4.81,
1893
+ "grad_norm": 0.72265625,
1894
+ "learning_rate": 8.103938734415839e-07,
1895
+ "loss": 0.4008,
1896
+ "step": 1320
1897
+ },
1898
+ {
1899
+ "epoch": 4.83,
1900
+ "grad_norm": 0.5859375,
1901
+ "learning_rate": 6.565877818526245e-07,
1902
+ "loss": 0.353,
1903
+ "step": 1325
1904
+ },
1905
+ {
1906
+ "epoch": 4.85,
1907
+ "grad_norm": 0.5859375,
1908
+ "learning_rate": 5.189047253851942e-07,
1909
+ "loss": 0.4758,
1910
+ "step": 1330
1911
+ },
1912
+ {
1913
+ "epoch": 4.86,
1914
+ "grad_norm": 0.609375,
1915
+ "learning_rate": 3.973670494440307e-07,
1916
+ "loss": 0.4848,
1917
+ "step": 1335
1918
+ },
1919
+ {
1920
+ "epoch": 4.88,
1921
+ "grad_norm": 0.5703125,
1922
+ "learning_rate": 2.919944791037632e-07,
1923
+ "loss": 0.3638,
1924
+ "step": 1340
1925
+ },
1926
+ {
1927
+ "epoch": 4.9,
1928
+ "grad_norm": 0.6796875,
1929
+ "learning_rate": 2.0280411590754e-07,
1930
+ "loss": 0.4863,
1931
+ "step": 1345
1932
+ },
1933
+ {
1934
+ "epoch": 4.92,
1935
+ "grad_norm": 0.6484375,
1936
+ "learning_rate": 1.2981043509159297e-07,
1937
+ "loss": 0.4502,
1938
+ "step": 1350
1939
+ },
1940
+ {
1941
+ "epoch": 4.94,
1942
+ "grad_norm": 0.7578125,
1943
+ "learning_rate": 7.302528323589464e-08,
1944
+ "loss": 0.4023,
1945
+ "step": 1355
1946
+ },
1947
+ {
1948
+ "epoch": 4.95,
1949
+ "grad_norm": 0.65234375,
1950
+ "learning_rate": 3.2457876341585033e-08,
1951
+ "loss": 0.4602,
1952
+ "step": 1360
1953
+ },
1954
+ {
1955
+ "epoch": 4.97,
1956
+ "grad_norm": 0.61328125,
1957
+ "learning_rate": 8.114798335157048e-09,
1958
+ "loss": 0.4061,
1959
+ "step": 1365
1960
+ },
1961
+ {
1962
+ "epoch": 4.99,
1963
+ "grad_norm": 0.421875,
1964
+ "learning_rate": 0.0,
1965
+ "loss": 0.3392,
1966
+ "step": 1370
1967
+ },
1968
+ {
1969
+ "epoch": 4.99,
1970
+ "eval_loss": 1.4403423070907593,
1971
+ "eval_runtime": 74.7328,
1972
+ "eval_samples_per_second": 3.666,
1973
+ "eval_steps_per_second": 0.468,
1974
+ "step": 1370
1975
+ },
1976
+ {
1977
+ "epoch": 4.99,
1978
+ "step": 1370,
1979
+ "total_flos": 9.82065973089534e+17,
1980
+ "train_loss": 1.004000898782354,
1981
+ "train_runtime": 10982.2688,
1982
+ "train_samples_per_second": 0.998,
1983
+ "train_steps_per_second": 0.125
1984
+ }
1985
+ ],
1986
+ "logging_steps": 5,
1987
+ "max_steps": 1370,
1988
+ "num_input_tokens_seen": 0,
1989
+ "num_train_epochs": 5,
1990
+ "save_steps": 100,
1991
+ "total_flos": 9.82065973089534e+17,
1992
+ "train_batch_size": 4,
1993
+ "trial_name": null,
1994
+ "trial_params": null
1995
+ }