chansung commited on
Commit
6ff5c14
1 Parent(s): 03af989

Model save

Browse files
Files changed (4) hide show
  1. README.md +78 -0
  2. all_results.json +9 -0
  3. train_results.json +9 -0
  4. trainer_state.json +2005 -0
README.md ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: mistralai/Mistral-7B-v0.3
3
+ datasets:
4
+ - generator
5
+ library_name: peft
6
+ license: apache-2.0
7
+ tags:
8
+ - trl
9
+ - sft
10
+ - generated_from_trainer
11
+ model-index:
12
+ - name: mistral-7b_0_3-classification-gpt4o-100k
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
+ # mistral-7b_0_3-classification-gpt4o-100k
20
+
21
+ This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the generator dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 1.5428
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: 4
45
+ - seed: 42
46
+ - distributed_type: multi-GPU
47
+ - num_devices: 4
48
+ - gradient_accumulation_steps: 2
49
+ - total_train_batch_size: 64
50
+ - total_eval_batch_size: 16
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
+ | 0.8763 | 1.0 | 134 | 1.4227 |
61
+ | 0.7756 | 2.0 | 268 | 1.4397 |
62
+ | 0.7387 | 3.0 | 402 | 1.4401 |
63
+ | 0.7128 | 4.0 | 536 | 1.4540 |
64
+ | 0.6837 | 5.0 | 670 | 1.4658 |
65
+ | 0.6691 | 6.0 | 804 | 1.4930 |
66
+ | 0.6389 | 7.0 | 938 | 1.5040 |
67
+ | 0.6291 | 8.0 | 1072 | 1.5243 |
68
+ | 0.6197 | 9.0 | 1206 | 1.5375 |
69
+ | 0.6028 | 10.0 | 1340 | 1.5428 |
70
+
71
+
72
+ ### Framework versions
73
+
74
+ - PEFT 0.12.0
75
+ - Transformers 4.44.0
76
+ - Pytorch 2.4.0+cu121
77
+ - Datasets 2.20.0
78
+ - Tokenizers 0.19.1
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 3.755518175798624e+18,
4
+ "train_loss": 0.7269157813556159,
5
+ "train_runtime": 7393.4759,
6
+ "train_samples": 92634,
7
+ "train_samples_per_second": 11.586,
8
+ "train_steps_per_second": 0.181
9
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 3.755518175798624e+18,
4
+ "train_loss": 0.7269157813556159,
5
+ "train_runtime": 7393.4759,
6
+ "train_samples": 92634,
7
+ "train_samples_per_second": 11.586,
8
+ "train_steps_per_second": 0.181
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,2005 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 10.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1340,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.007462686567164179,
13
+ "grad_norm": 138.84262084960938,
14
+ "learning_rate": 1.4925373134328358e-06,
15
+ "loss": 2.7919,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.03731343283582089,
20
+ "grad_norm": 108.88359832763672,
21
+ "learning_rate": 7.4626865671641785e-06,
22
+ "loss": 2.7342,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.07462686567164178,
27
+ "grad_norm": 13.541873931884766,
28
+ "learning_rate": 1.4925373134328357e-05,
29
+ "loss": 2.3501,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.11194029850746269,
34
+ "grad_norm": 7.267947673797607,
35
+ "learning_rate": 2.238805970149254e-05,
36
+ "loss": 1.903,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.14925373134328357,
41
+ "grad_norm": 4.06025505065918,
42
+ "learning_rate": 2.9850746268656714e-05,
43
+ "loss": 1.5868,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.1865671641791045,
48
+ "grad_norm": 1.833858847618103,
49
+ "learning_rate": 3.73134328358209e-05,
50
+ "loss": 1.3748,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.22388059701492538,
55
+ "grad_norm": 1.4598793983459473,
56
+ "learning_rate": 4.477611940298508e-05,
57
+ "loss": 1.2384,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.26119402985074625,
62
+ "grad_norm": 1.4846118688583374,
63
+ "learning_rate": 5.223880597014925e-05,
64
+ "loss": 1.149,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.29850746268656714,
69
+ "grad_norm": 0.9875138998031616,
70
+ "learning_rate": 5.970149253731343e-05,
71
+ "loss": 1.0731,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.3358208955223881,
76
+ "grad_norm": 1.0995197296142578,
77
+ "learning_rate": 6.716417910447762e-05,
78
+ "loss": 1.0446,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.373134328358209,
83
+ "grad_norm": 0.6559099555015564,
84
+ "learning_rate": 7.46268656716418e-05,
85
+ "loss": 1.014,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.41044776119402987,
90
+ "grad_norm": 1.1354745626449585,
91
+ "learning_rate": 8.208955223880597e-05,
92
+ "loss": 0.994,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.44776119402985076,
97
+ "grad_norm": 0.49952608346939087,
98
+ "learning_rate": 8.955223880597016e-05,
99
+ "loss": 0.9708,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.48507462686567165,
104
+ "grad_norm": 0.8550089597702026,
105
+ "learning_rate": 9.701492537313434e-05,
106
+ "loss": 0.9634,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.5223880597014925,
111
+ "grad_norm": 0.5180774927139282,
112
+ "learning_rate": 0.0001044776119402985,
113
+ "loss": 0.951,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.5597014925373134,
118
+ "grad_norm": 0.6518343687057495,
119
+ "learning_rate": 0.00011194029850746269,
120
+ "loss": 0.9374,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.5970149253731343,
125
+ "grad_norm": 0.9855380654335022,
126
+ "learning_rate": 0.00011940298507462686,
127
+ "loss": 0.9345,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.6343283582089553,
132
+ "grad_norm": 1.1707885265350342,
133
+ "learning_rate": 0.00012686567164179105,
134
+ "loss": 0.9269,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.6716417910447762,
139
+ "grad_norm": 0.527793824672699,
140
+ "learning_rate": 0.00013432835820895525,
141
+ "loss": 0.9095,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.7089552238805971,
146
+ "grad_norm": 0.678547203540802,
147
+ "learning_rate": 0.00014179104477611942,
148
+ "loss": 0.9077,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.746268656716418,
153
+ "grad_norm": 1.1332663297653198,
154
+ "learning_rate": 0.0001492537313432836,
155
+ "loss": 0.8974,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.7835820895522388,
160
+ "grad_norm": 0.5298197865486145,
161
+ "learning_rate": 0.00015671641791044778,
162
+ "loss": 0.9071,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.8208955223880597,
167
+ "grad_norm": 0.5773002505302429,
168
+ "learning_rate": 0.00016417910447761195,
169
+ "loss": 0.8858,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.8582089552238806,
174
+ "grad_norm": 0.6888121962547302,
175
+ "learning_rate": 0.00017164179104477614,
176
+ "loss": 0.8891,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.8955223880597015,
181
+ "grad_norm": 0.550254762172699,
182
+ "learning_rate": 0.0001791044776119403,
183
+ "loss": 0.8757,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.9328358208955224,
188
+ "grad_norm": 0.5638606548309326,
189
+ "learning_rate": 0.00018656716417910448,
190
+ "loss": 0.872,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.9701492537313433,
195
+ "grad_norm": 0.826892614364624,
196
+ "learning_rate": 0.00019402985074626867,
197
+ "loss": 0.8763,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 1.0,
202
+ "eval_loss": 1.4227064847946167,
203
+ "eval_runtime": 0.5192,
204
+ "eval_samples_per_second": 3.852,
205
+ "eval_steps_per_second": 1.926,
206
+ "step": 134
207
+ },
208
+ {
209
+ "epoch": 1.007462686567164,
210
+ "grad_norm": 0.6830571293830872,
211
+ "learning_rate": 0.00019999966070700876,
212
+ "loss": 0.8669,
213
+ "step": 135
214
+ },
215
+ {
216
+ "epoch": 1.044776119402985,
217
+ "grad_norm": 0.6296478509902954,
218
+ "learning_rate": 0.0001999877856940653,
219
+ "loss": 0.8547,
220
+ "step": 140
221
+ },
222
+ {
223
+ "epoch": 1.0820895522388059,
224
+ "grad_norm": 0.7137119770050049,
225
+ "learning_rate": 0.00019995894833388465,
226
+ "loss": 0.8498,
227
+ "step": 145
228
+ },
229
+ {
230
+ "epoch": 1.1194029850746268,
231
+ "grad_norm": 0.44721338152885437,
232
+ "learning_rate": 0.00019991315351855748,
233
+ "loss": 0.848,
234
+ "step": 150
235
+ },
236
+ {
237
+ "epoch": 1.1567164179104479,
238
+ "grad_norm": 0.4353409707546234,
239
+ "learning_rate": 0.0001998504090169083,
240
+ "loss": 0.8289,
241
+ "step": 155
242
+ },
243
+ {
244
+ "epoch": 1.1940298507462686,
245
+ "grad_norm": 0.4284411668777466,
246
+ "learning_rate": 0.0001997707254731775,
247
+ "loss": 0.8298,
248
+ "step": 160
249
+ },
250
+ {
251
+ "epoch": 1.2313432835820897,
252
+ "grad_norm": 0.49997657537460327,
253
+ "learning_rate": 0.00019967411640521552,
254
+ "loss": 0.8305,
255
+ "step": 165
256
+ },
257
+ {
258
+ "epoch": 1.2686567164179103,
259
+ "grad_norm": 0.5145930647850037,
260
+ "learning_rate": 0.00019956059820218982,
261
+ "loss": 0.8165,
262
+ "step": 170
263
+ },
264
+ {
265
+ "epoch": 1.3059701492537314,
266
+ "grad_norm": 0.9191280007362366,
267
+ "learning_rate": 0.0001994301901218043,
268
+ "loss": 0.8228,
269
+ "step": 175
270
+ },
271
+ {
272
+ "epoch": 1.3432835820895521,
273
+ "grad_norm": 0.4731077551841736,
274
+ "learning_rate": 0.00019928291428703262,
275
+ "loss": 0.8218,
276
+ "step": 180
277
+ },
278
+ {
279
+ "epoch": 1.3805970149253732,
280
+ "grad_norm": 0.4273313879966736,
281
+ "learning_rate": 0.0001991187956823649,
282
+ "loss": 0.8277,
283
+ "step": 185
284
+ },
285
+ {
286
+ "epoch": 1.417910447761194,
287
+ "grad_norm": 0.512370228767395,
288
+ "learning_rate": 0.00019893786214956945,
289
+ "loss": 0.8075,
290
+ "step": 190
291
+ },
292
+ {
293
+ "epoch": 1.455223880597015,
294
+ "grad_norm": 0.6519188284873962,
295
+ "learning_rate": 0.00019874014438296946,
296
+ "loss": 0.8178,
297
+ "step": 195
298
+ },
299
+ {
300
+ "epoch": 1.4925373134328357,
301
+ "grad_norm": 0.5379275679588318,
302
+ "learning_rate": 0.0001985256759242359,
303
+ "loss": 0.8134,
304
+ "step": 200
305
+ },
306
+ {
307
+ "epoch": 1.5298507462686568,
308
+ "grad_norm": 0.5376815795898438,
309
+ "learning_rate": 0.00019829449315669748,
310
+ "loss": 0.8005,
311
+ "step": 205
312
+ },
313
+ {
314
+ "epoch": 1.5671641791044775,
315
+ "grad_norm": 0.5267529487609863,
316
+ "learning_rate": 0.00019804663529916826,
317
+ "loss": 0.8021,
318
+ "step": 210
319
+ },
320
+ {
321
+ "epoch": 1.6044776119402986,
322
+ "grad_norm": 0.582000732421875,
323
+ "learning_rate": 0.00019778214439929452,
324
+ "loss": 0.8107,
325
+ "step": 215
326
+ },
327
+ {
328
+ "epoch": 1.6417910447761193,
329
+ "grad_norm": 0.4234386086463928,
330
+ "learning_rate": 0.0001975010653264216,
331
+ "loss": 0.7975,
332
+ "step": 220
333
+ },
334
+ {
335
+ "epoch": 1.6791044776119404,
336
+ "grad_norm": 0.4661960005760193,
337
+ "learning_rate": 0.00019720344576398203,
338
+ "loss": 0.7991,
339
+ "step": 225
340
+ },
341
+ {
342
+ "epoch": 1.716417910447761,
343
+ "grad_norm": 0.66214519739151,
344
+ "learning_rate": 0.00019688933620140637,
345
+ "loss": 0.7966,
346
+ "step": 230
347
+ },
348
+ {
349
+ "epoch": 1.7537313432835822,
350
+ "grad_norm": 0.4201449453830719,
351
+ "learning_rate": 0.000196558789925558,
352
+ "loss": 0.8005,
353
+ "step": 235
354
+ },
355
+ {
356
+ "epoch": 1.7910447761194028,
357
+ "grad_norm": 0.4008370637893677,
358
+ "learning_rate": 0.00019621186301169315,
359
+ "loss": 0.7921,
360
+ "step": 240
361
+ },
362
+ {
363
+ "epoch": 1.828358208955224,
364
+ "grad_norm": 0.4507235884666443,
365
+ "learning_rate": 0.00019584861431394822,
366
+ "loss": 0.7956,
367
+ "step": 245
368
+ },
369
+ {
370
+ "epoch": 1.8656716417910446,
371
+ "grad_norm": 0.46443891525268555,
372
+ "learning_rate": 0.00019546910545535558,
373
+ "loss": 0.7932,
374
+ "step": 250
375
+ },
376
+ {
377
+ "epoch": 1.9029850746268657,
378
+ "grad_norm": 0.5451995134353638,
379
+ "learning_rate": 0.0001950734008173893,
380
+ "loss": 0.7889,
381
+ "step": 255
382
+ },
383
+ {
384
+ "epoch": 1.9402985074626866,
385
+ "grad_norm": 0.507540762424469,
386
+ "learning_rate": 0.00019466156752904343,
387
+ "loss": 0.7794,
388
+ "step": 260
389
+ },
390
+ {
391
+ "epoch": 1.9776119402985075,
392
+ "grad_norm": 0.4746926724910736,
393
+ "learning_rate": 0.00019423367545544395,
394
+ "loss": 0.7756,
395
+ "step": 265
396
+ },
397
+ {
398
+ "epoch": 2.0,
399
+ "eval_loss": 1.439745306968689,
400
+ "eval_runtime": 0.5176,
401
+ "eval_samples_per_second": 3.864,
402
+ "eval_steps_per_second": 1.932,
403
+ "step": 268
404
+ },
405
+ {
406
+ "epoch": 2.014925373134328,
407
+ "grad_norm": 0.6268861889839172,
408
+ "learning_rate": 0.00019378979718599645,
409
+ "loss": 0.7882,
410
+ "step": 270
411
+ },
412
+ {
413
+ "epoch": 2.0522388059701493,
414
+ "grad_norm": 0.549802839756012,
415
+ "learning_rate": 0.0001933300080220719,
416
+ "loss": 0.7609,
417
+ "step": 275
418
+ },
419
+ {
420
+ "epoch": 2.08955223880597,
421
+ "grad_norm": 0.6178936958312988,
422
+ "learning_rate": 0.00019285438596423204,
423
+ "loss": 0.7603,
424
+ "step": 280
425
+ },
426
+ {
427
+ "epoch": 2.126865671641791,
428
+ "grad_norm": 0.41491350531578064,
429
+ "learning_rate": 0.00019236301169899713,
430
+ "loss": 0.7554,
431
+ "step": 285
432
+ },
433
+ {
434
+ "epoch": 2.1641791044776117,
435
+ "grad_norm": 0.5220763683319092,
436
+ "learning_rate": 0.000191855968585158,
437
+ "loss": 0.759,
438
+ "step": 290
439
+ },
440
+ {
441
+ "epoch": 2.201492537313433,
442
+ "grad_norm": 0.5099439024925232,
443
+ "learning_rate": 0.0001913333426396346,
444
+ "loss": 0.7614,
445
+ "step": 295
446
+ },
447
+ {
448
+ "epoch": 2.2388059701492535,
449
+ "grad_norm": 0.5702708959579468,
450
+ "learning_rate": 0.00019079522252288386,
451
+ "loss": 0.7485,
452
+ "step": 300
453
+ },
454
+ {
455
+ "epoch": 2.2761194029850746,
456
+ "grad_norm": 0.4422476887702942,
457
+ "learning_rate": 0.00019024169952385885,
458
+ "loss": 0.76,
459
+ "step": 305
460
+ },
461
+ {
462
+ "epoch": 2.3134328358208958,
463
+ "grad_norm": 0.44903138279914856,
464
+ "learning_rate": 0.00018967286754452214,
465
+ "loss": 0.7632,
466
+ "step": 310
467
+ },
468
+ {
469
+ "epoch": 2.3507462686567164,
470
+ "grad_norm": 0.49397289752960205,
471
+ "learning_rate": 0.00018908882308391598,
472
+ "loss": 0.7568,
473
+ "step": 315
474
+ },
475
+ {
476
+ "epoch": 2.388059701492537,
477
+ "grad_norm": 0.4179913103580475,
478
+ "learning_rate": 0.00018848966522179168,
479
+ "loss": 0.7596,
480
+ "step": 320
481
+ },
482
+ {
483
+ "epoch": 2.425373134328358,
484
+ "grad_norm": 0.4103606045246124,
485
+ "learning_rate": 0.0001878754956018014,
486
+ "loss": 0.7487,
487
+ "step": 325
488
+ },
489
+ {
490
+ "epoch": 2.4626865671641793,
491
+ "grad_norm": 0.41152262687683105,
492
+ "learning_rate": 0.00018724641841425478,
493
+ "loss": 0.7565,
494
+ "step": 330
495
+ },
496
+ {
497
+ "epoch": 2.5,
498
+ "grad_norm": 0.4100894629955292,
499
+ "learning_rate": 0.00018660254037844388,
500
+ "loss": 0.7538,
501
+ "step": 335
502
+ },
503
+ {
504
+ "epoch": 2.5373134328358207,
505
+ "grad_norm": 0.3992711901664734,
506
+ "learning_rate": 0.00018594397072453856,
507
+ "loss": 0.7524,
508
+ "step": 340
509
+ },
510
+ {
511
+ "epoch": 2.574626865671642,
512
+ "grad_norm": 0.42304080724716187,
513
+ "learning_rate": 0.0001852708211750564,
514
+ "loss": 0.7533,
515
+ "step": 345
516
+ },
517
+ {
518
+ "epoch": 2.611940298507463,
519
+ "grad_norm": 0.4376600384712219,
520
+ "learning_rate": 0.00018458320592590975,
521
+ "loss": 0.7635,
522
+ "step": 350
523
+ },
524
+ {
525
+ "epoch": 2.6492537313432836,
526
+ "grad_norm": 0.39463093876838684,
527
+ "learning_rate": 0.00018388124162703275,
528
+ "loss": 0.7438,
529
+ "step": 355
530
+ },
531
+ {
532
+ "epoch": 2.6865671641791042,
533
+ "grad_norm": 0.3841034770011902,
534
+ "learning_rate": 0.00018316504736259255,
535
+ "loss": 0.759,
536
+ "step": 360
537
+ },
538
+ {
539
+ "epoch": 2.7238805970149254,
540
+ "grad_norm": 0.43507230281829834,
541
+ "learning_rate": 0.00018243474463078735,
542
+ "loss": 0.7477,
543
+ "step": 365
544
+ },
545
+ {
546
+ "epoch": 2.7611940298507465,
547
+ "grad_norm": 0.41791486740112305,
548
+ "learning_rate": 0.00018169045732323492,
549
+ "loss": 0.7463,
550
+ "step": 370
551
+ },
552
+ {
553
+ "epoch": 2.798507462686567,
554
+ "grad_norm": 0.4145572781562805,
555
+ "learning_rate": 0.00018093231170395507,
556
+ "loss": 0.7438,
557
+ "step": 375
558
+ },
559
+ {
560
+ "epoch": 2.835820895522388,
561
+ "grad_norm": 0.39668190479278564,
562
+ "learning_rate": 0.00018016043638794974,
563
+ "loss": 0.7545,
564
+ "step": 380
565
+ },
566
+ {
567
+ "epoch": 2.873134328358209,
568
+ "grad_norm": 0.36246854066848755,
569
+ "learning_rate": 0.0001793749623193842,
570
+ "loss": 0.7483,
571
+ "step": 385
572
+ },
573
+ {
574
+ "epoch": 2.91044776119403,
575
+ "grad_norm": 0.4155827462673187,
576
+ "learning_rate": 0.00017857602274937308,
577
+ "loss": 0.7499,
578
+ "step": 390
579
+ },
580
+ {
581
+ "epoch": 2.9477611940298507,
582
+ "grad_norm": 0.3993385136127472,
583
+ "learning_rate": 0.00017776375321337521,
584
+ "loss": 0.7402,
585
+ "step": 395
586
+ },
587
+ {
588
+ "epoch": 2.9850746268656714,
589
+ "grad_norm": 0.40633976459503174,
590
+ "learning_rate": 0.00017693829150820068,
591
+ "loss": 0.7387,
592
+ "step": 400
593
+ },
594
+ {
595
+ "epoch": 3.0,
596
+ "eval_loss": 1.440061092376709,
597
+ "eval_runtime": 0.5179,
598
+ "eval_samples_per_second": 3.862,
599
+ "eval_steps_per_second": 1.931,
600
+ "step": 402
601
+ },
602
+ {
603
+ "epoch": 3.0223880597014925,
604
+ "grad_norm": 0.43004971742630005,
605
+ "learning_rate": 0.00017609977766863458,
606
+ "loss": 0.7429,
607
+ "step": 405
608
+ },
609
+ {
610
+ "epoch": 3.0597014925373136,
611
+ "grad_norm": 0.4013059735298157,
612
+ "learning_rate": 0.00017524835394368065,
613
+ "loss": 0.7087,
614
+ "step": 410
615
+ },
616
+ {
617
+ "epoch": 3.0970149253731343,
618
+ "grad_norm": 0.4613398313522339,
619
+ "learning_rate": 0.0001743841647724299,
620
+ "loss": 0.7234,
621
+ "step": 415
622
+ },
623
+ {
624
+ "epoch": 3.1343283582089554,
625
+ "grad_norm": 0.3847793936729431,
626
+ "learning_rate": 0.00017350735675955697,
627
+ "loss": 0.7214,
628
+ "step": 420
629
+ },
630
+ {
631
+ "epoch": 3.171641791044776,
632
+ "grad_norm": 0.4573831558227539,
633
+ "learning_rate": 0.00017261807865044993,
634
+ "loss": 0.7228,
635
+ "step": 425
636
+ },
637
+ {
638
+ "epoch": 3.208955223880597,
639
+ "grad_norm": 0.4393455684185028,
640
+ "learning_rate": 0.00017171648130597612,
641
+ "loss": 0.7202,
642
+ "step": 430
643
+ },
644
+ {
645
+ "epoch": 3.246268656716418,
646
+ "grad_norm": 0.45842474699020386,
647
+ "learning_rate": 0.00017080271767688963,
648
+ "loss": 0.7098,
649
+ "step": 435
650
+ },
651
+ {
652
+ "epoch": 3.283582089552239,
653
+ "grad_norm": 0.46594393253326416,
654
+ "learning_rate": 0.00016987694277788417,
655
+ "loss": 0.7159,
656
+ "step": 440
657
+ },
658
+ {
659
+ "epoch": 3.3208955223880596,
660
+ "grad_norm": 0.40664443373680115,
661
+ "learning_rate": 0.00016893931366129562,
662
+ "loss": 0.7128,
663
+ "step": 445
664
+ },
665
+ {
666
+ "epoch": 3.3582089552238807,
667
+ "grad_norm": 0.38857197761535645,
668
+ "learning_rate": 0.00016798998939045895,
669
+ "loss": 0.7223,
670
+ "step": 450
671
+ },
672
+ {
673
+ "epoch": 3.3955223880597014,
674
+ "grad_norm": 0.443276047706604,
675
+ "learning_rate": 0.0001670291310127242,
676
+ "loss": 0.7132,
677
+ "step": 455
678
+ },
679
+ {
680
+ "epoch": 3.4328358208955225,
681
+ "grad_norm": 0.3902607858181,
682
+ "learning_rate": 0.0001660569015321357,
683
+ "loss": 0.7139,
684
+ "step": 460
685
+ },
686
+ {
687
+ "epoch": 3.470149253731343,
688
+ "grad_norm": 0.3990344703197479,
689
+ "learning_rate": 0.00016507346588177935,
690
+ "loss": 0.7191,
691
+ "step": 465
692
+ },
693
+ {
694
+ "epoch": 3.5074626865671643,
695
+ "grad_norm": 0.4015812277793884,
696
+ "learning_rate": 0.00016407899089580262,
697
+ "loss": 0.7146,
698
+ "step": 470
699
+ },
700
+ {
701
+ "epoch": 3.544776119402985,
702
+ "grad_norm": 0.44737905263900757,
703
+ "learning_rate": 0.0001630736452811122,
704
+ "loss": 0.7182,
705
+ "step": 475
706
+ },
707
+ {
708
+ "epoch": 3.582089552238806,
709
+ "grad_norm": 0.4523073136806488,
710
+ "learning_rate": 0.0001620575995887538,
711
+ "loss": 0.7212,
712
+ "step": 480
713
+ },
714
+ {
715
+ "epoch": 3.6194029850746268,
716
+ "grad_norm": 0.4242749512195587,
717
+ "learning_rate": 0.00016103102618497922,
718
+ "loss": 0.7168,
719
+ "step": 485
720
+ },
721
+ {
722
+ "epoch": 3.656716417910448,
723
+ "grad_norm": 0.4483940899372101,
724
+ "learning_rate": 0.0001599940992220053,
725
+ "loss": 0.7145,
726
+ "step": 490
727
+ },
728
+ {
729
+ "epoch": 3.6940298507462686,
730
+ "grad_norm": 0.4250345826148987,
731
+ "learning_rate": 0.00015894699460847016,
732
+ "loss": 0.6993,
733
+ "step": 495
734
+ },
735
+ {
736
+ "epoch": 3.7313432835820897,
737
+ "grad_norm": 0.38691166043281555,
738
+ "learning_rate": 0.00015788988997959114,
739
+ "loss": 0.7135,
740
+ "step": 500
741
+ },
742
+ {
743
+ "epoch": 3.7686567164179103,
744
+ "grad_norm": 0.4665898382663727,
745
+ "learning_rate": 0.00015682296466703024,
746
+ "loss": 0.7226,
747
+ "step": 505
748
+ },
749
+ {
750
+ "epoch": 3.8059701492537314,
751
+ "grad_norm": 0.501005232334137,
752
+ "learning_rate": 0.00015574639966847126,
753
+ "loss": 0.7189,
754
+ "step": 510
755
+ },
756
+ {
757
+ "epoch": 3.843283582089552,
758
+ "grad_norm": 0.3910924792289734,
759
+ "learning_rate": 0.0001546603776169149,
760
+ "loss": 0.7208,
761
+ "step": 515
762
+ },
763
+ {
764
+ "epoch": 3.8805970149253732,
765
+ "grad_norm": 0.4152643382549286,
766
+ "learning_rate": 0.00015356508274969594,
767
+ "loss": 0.7248,
768
+ "step": 520
769
+ },
770
+ {
771
+ "epoch": 3.917910447761194,
772
+ "grad_norm": 0.4302676022052765,
773
+ "learning_rate": 0.00015246070087722845,
774
+ "loss": 0.7169,
775
+ "step": 525
776
+ },
777
+ {
778
+ "epoch": 3.955223880597015,
779
+ "grad_norm": 0.39138561487197876,
780
+ "learning_rate": 0.0001513474193514842,
781
+ "loss": 0.7243,
782
+ "step": 530
783
+ },
784
+ {
785
+ "epoch": 3.9925373134328357,
786
+ "grad_norm": 0.4384176731109619,
787
+ "learning_rate": 0.0001502254270342095,
788
+ "loss": 0.7128,
789
+ "step": 535
790
+ },
791
+ {
792
+ "epoch": 4.0,
793
+ "eval_loss": 1.4540088176727295,
794
+ "eval_runtime": 0.518,
795
+ "eval_samples_per_second": 3.861,
796
+ "eval_steps_per_second": 1.93,
797
+ "step": 536
798
+ },
799
+ {
800
+ "epoch": 4.029850746268656,
801
+ "grad_norm": 0.4122828543186188,
802
+ "learning_rate": 0.00014909491426488578,
803
+ "loss": 0.6889,
804
+ "step": 540
805
+ },
806
+ {
807
+ "epoch": 4.067164179104478,
808
+ "grad_norm": 0.4475855827331543,
809
+ "learning_rate": 0.0001479560728284398,
810
+ "loss": 0.6847,
811
+ "step": 545
812
+ },
813
+ {
814
+ "epoch": 4.104477611940299,
815
+ "grad_norm": 0.47523927688598633,
816
+ "learning_rate": 0.0001468090959227082,
817
+ "loss": 0.6887,
818
+ "step": 550
819
+ },
820
+ {
821
+ "epoch": 4.141791044776119,
822
+ "grad_norm": 0.39732715487480164,
823
+ "learning_rate": 0.00014565417812566285,
824
+ "loss": 0.6877,
825
+ "step": 555
826
+ },
827
+ {
828
+ "epoch": 4.17910447761194,
829
+ "grad_norm": 0.4284641444683075,
830
+ "learning_rate": 0.00014449151536240166,
831
+ "loss": 0.6932,
832
+ "step": 560
833
+ },
834
+ {
835
+ "epoch": 4.2164179104477615,
836
+ "grad_norm": 0.43158912658691406,
837
+ "learning_rate": 0.0001433213048719111,
838
+ "loss": 0.6893,
839
+ "step": 565
840
+ },
841
+ {
842
+ "epoch": 4.253731343283582,
843
+ "grad_norm": 0.4248037338256836,
844
+ "learning_rate": 0.00014214374517360575,
845
+ "loss": 0.6846,
846
+ "step": 570
847
+ },
848
+ {
849
+ "epoch": 4.291044776119403,
850
+ "grad_norm": 0.4019497334957123,
851
+ "learning_rate": 0.00014095903603365066,
852
+ "loss": 0.6872,
853
+ "step": 575
854
+ },
855
+ {
856
+ "epoch": 4.3283582089552235,
857
+ "grad_norm": 0.42748814821243286,
858
+ "learning_rate": 0.00013976737843107202,
859
+ "loss": 0.6876,
860
+ "step": 580
861
+ },
862
+ {
863
+ "epoch": 4.365671641791045,
864
+ "grad_norm": 0.4337392747402191,
865
+ "learning_rate": 0.00013856897452366242,
866
+ "loss": 0.6981,
867
+ "step": 585
868
+ },
869
+ {
870
+ "epoch": 4.402985074626866,
871
+ "grad_norm": 0.4018155634403229,
872
+ "learning_rate": 0.00013736402761368598,
873
+ "loss": 0.6872,
874
+ "step": 590
875
+ },
876
+ {
877
+ "epoch": 4.440298507462686,
878
+ "grad_norm": 0.40716463327407837,
879
+ "learning_rate": 0.00013615274211338909,
880
+ "loss": 0.6874,
881
+ "step": 595
882
+ },
883
+ {
884
+ "epoch": 4.477611940298507,
885
+ "grad_norm": 0.4064732491970062,
886
+ "learning_rate": 0.0001349353235103232,
887
+ "loss": 0.69,
888
+ "step": 600
889
+ },
890
+ {
891
+ "epoch": 4.514925373134329,
892
+ "grad_norm": 0.38808923959732056,
893
+ "learning_rate": 0.00013371197833248507,
894
+ "loss": 0.6906,
895
+ "step": 605
896
+ },
897
+ {
898
+ "epoch": 4.552238805970149,
899
+ "grad_norm": 0.4051991105079651,
900
+ "learning_rate": 0.00013248291411328047,
901
+ "loss": 0.6898,
902
+ "step": 610
903
+ },
904
+ {
905
+ "epoch": 4.58955223880597,
906
+ "grad_norm": 0.4059875011444092,
907
+ "learning_rate": 0.00013124833935631726,
908
+ "loss": 0.688,
909
+ "step": 615
910
+ },
911
+ {
912
+ "epoch": 4.6268656716417915,
913
+ "grad_norm": 0.4350437521934509,
914
+ "learning_rate": 0.0001300084635000341,
915
+ "loss": 0.6868,
916
+ "step": 620
917
+ },
918
+ {
919
+ "epoch": 4.664179104477612,
920
+ "grad_norm": 0.43591079115867615,
921
+ "learning_rate": 0.0001287634968821703,
922
+ "loss": 0.6843,
923
+ "step": 625
924
+ },
925
+ {
926
+ "epoch": 4.701492537313433,
927
+ "grad_norm": 0.3807081878185272,
928
+ "learning_rate": 0.00012751365070408333,
929
+ "loss": 0.6792,
930
+ "step": 630
931
+ },
932
+ {
933
+ "epoch": 4.7388059701492535,
934
+ "grad_norm": 0.3958300054073334,
935
+ "learning_rate": 0.00012625913699491984,
936
+ "loss": 0.6896,
937
+ "step": 635
938
+ },
939
+ {
940
+ "epoch": 4.776119402985074,
941
+ "grad_norm": 0.4042377471923828,
942
+ "learning_rate": 0.00012500016857564585,
943
+ "loss": 0.6875,
944
+ "step": 640
945
+ },
946
+ {
947
+ "epoch": 4.813432835820896,
948
+ "grad_norm": 0.40110722184181213,
949
+ "learning_rate": 0.00012373695902294314,
950
+ "loss": 0.6923,
951
+ "step": 645
952
+ },
953
+ {
954
+ "epoch": 4.850746268656716,
955
+ "grad_norm": 0.4298909604549408,
956
+ "learning_rate": 0.0001224697226329772,
957
+ "loss": 0.6975,
958
+ "step": 650
959
+ },
960
+ {
961
+ "epoch": 4.888059701492537,
962
+ "grad_norm": 0.4291529655456543,
963
+ "learning_rate": 0.00012119867438504302,
964
+ "loss": 0.6908,
965
+ "step": 655
966
+ },
967
+ {
968
+ "epoch": 4.925373134328359,
969
+ "grad_norm": 0.4685961902141571,
970
+ "learning_rate": 0.00011992402990509515,
971
+ "loss": 0.6836,
972
+ "step": 660
973
+ },
974
+ {
975
+ "epoch": 4.962686567164179,
976
+ "grad_norm": 0.4041878879070282,
977
+ "learning_rate": 0.00011864600542916813,
978
+ "loss": 0.6921,
979
+ "step": 665
980
+ },
981
+ {
982
+ "epoch": 5.0,
983
+ "grad_norm": 0.3824089467525482,
984
+ "learning_rate": 0.00011736481776669306,
985
+ "loss": 0.6837,
986
+ "step": 670
987
+ },
988
+ {
989
+ "epoch": 5.0,
990
+ "eval_loss": 1.465812087059021,
991
+ "eval_runtime": 0.5174,
992
+ "eval_samples_per_second": 3.866,
993
+ "eval_steps_per_second": 1.933,
994
+ "step": 670
995
+ },
996
+ {
997
+ "epoch": 5.037313432835821,
998
+ "grad_norm": 0.4383637309074402,
999
+ "learning_rate": 0.0001160806842637173,
1000
+ "loss": 0.6631,
1001
+ "step": 675
1002
+ },
1003
+ {
1004
+ "epoch": 5.074626865671641,
1005
+ "grad_norm": 0.40598100423812866,
1006
+ "learning_rate": 0.000114793822766033,
1007
+ "loss": 0.6596,
1008
+ "step": 680
1009
+ },
1010
+ {
1011
+ "epoch": 5.111940298507463,
1012
+ "grad_norm": 0.413703590631485,
1013
+ "learning_rate": 0.00011350445158222074,
1014
+ "loss": 0.6631,
1015
+ "step": 685
1016
+ },
1017
+ {
1018
+ "epoch": 5.149253731343284,
1019
+ "grad_norm": 0.44605740904808044,
1020
+ "learning_rate": 0.00011221278944661473,
1021
+ "loss": 0.6463,
1022
+ "step": 690
1023
+ },
1024
+ {
1025
+ "epoch": 5.186567164179104,
1026
+ "grad_norm": 0.43800660967826843,
1027
+ "learning_rate": 0.00011091905548219597,
1028
+ "loss": 0.6662,
1029
+ "step": 695
1030
+ },
1031
+ {
1032
+ "epoch": 5.223880597014926,
1033
+ "grad_norm": 0.4596647620201111,
1034
+ "learning_rate": 0.00010962346916341903,
1035
+ "loss": 0.6638,
1036
+ "step": 700
1037
+ },
1038
+ {
1039
+ "epoch": 5.2611940298507465,
1040
+ "grad_norm": 0.4407581090927124,
1041
+ "learning_rate": 0.00010832625027897978,
1042
+ "loss": 0.6644,
1043
+ "step": 705
1044
+ },
1045
+ {
1046
+ "epoch": 5.298507462686567,
1047
+ "grad_norm": 0.41823288798332214,
1048
+ "learning_rate": 0.0001070276188945293,
1049
+ "loss": 0.6565,
1050
+ "step": 710
1051
+ },
1052
+ {
1053
+ "epoch": 5.335820895522388,
1054
+ "grad_norm": 0.42799797654151917,
1055
+ "learning_rate": 0.00010572779531534112,
1056
+ "loss": 0.6555,
1057
+ "step": 715
1058
+ },
1059
+ {
1060
+ "epoch": 5.373134328358209,
1061
+ "grad_norm": 0.3943883776664734,
1062
+ "learning_rate": 0.00010442700004893764,
1063
+ "loss": 0.6572,
1064
+ "step": 720
1065
+ },
1066
+ {
1067
+ "epoch": 5.41044776119403,
1068
+ "grad_norm": 0.40598776936531067,
1069
+ "learning_rate": 0.00010312545376768246,
1070
+ "loss": 0.6564,
1071
+ "step": 725
1072
+ },
1073
+ {
1074
+ "epoch": 5.447761194029851,
1075
+ "grad_norm": 0.42347079515457153,
1076
+ "learning_rate": 0.0001018233772713443,
1077
+ "loss": 0.6677,
1078
+ "step": 730
1079
+ },
1080
+ {
1081
+ "epoch": 5.485074626865671,
1082
+ "grad_norm": 0.4139921963214874,
1083
+ "learning_rate": 0.00010052099144964004,
1084
+ "loss": 0.6701,
1085
+ "step": 735
1086
+ },
1087
+ {
1088
+ "epoch": 5.522388059701493,
1089
+ "grad_norm": 0.4295801520347595,
1090
+ "learning_rate": 9.92185172447616e-05,
1091
+ "loss": 0.6605,
1092
+ "step": 740
1093
+ },
1094
+ {
1095
+ "epoch": 5.559701492537314,
1096
+ "grad_norm": 0.41350170969963074,
1097
+ "learning_rate": 9.791617561389475e-05,
1098
+ "loss": 0.6671,
1099
+ "step": 745
1100
+ },
1101
+ {
1102
+ "epoch": 5.597014925373134,
1103
+ "grad_norm": 0.4078296720981598,
1104
+ "learning_rate": 9.661418749173467e-05,
1105
+ "loss": 0.669,
1106
+ "step": 750
1107
+ },
1108
+ {
1109
+ "epoch": 5.634328358208955,
1110
+ "grad_norm": 0.4294317066669464,
1111
+ "learning_rate": 9.531277375300599e-05,
1112
+ "loss": 0.6653,
1113
+ "step": 755
1114
+ },
1115
+ {
1116
+ "epoch": 5.6716417910447765,
1117
+ "grad_norm": 0.4155528247356415,
1118
+ "learning_rate": 9.40121551749925e-05,
1119
+ "loss": 0.6587,
1120
+ "step": 760
1121
+ },
1122
+ {
1123
+ "epoch": 5.708955223880597,
1124
+ "grad_norm": 0.4128773808479309,
1125
+ "learning_rate": 9.271255240008356e-05,
1126
+ "loss": 0.6572,
1127
+ "step": 765
1128
+ },
1129
+ {
1130
+ "epoch": 5.746268656716418,
1131
+ "grad_norm": 0.4138401746749878,
1132
+ "learning_rate": 9.141418589834339e-05,
1133
+ "loss": 0.6619,
1134
+ "step": 770
1135
+ },
1136
+ {
1137
+ "epoch": 5.7835820895522385,
1138
+ "grad_norm": 0.40413615107536316,
1139
+ "learning_rate": 9.011727593010981e-05,
1140
+ "loss": 0.6745,
1141
+ "step": 775
1142
+ },
1143
+ {
1144
+ "epoch": 5.82089552238806,
1145
+ "grad_norm": 0.447427898645401,
1146
+ "learning_rate": 8.882204250862796e-05,
1147
+ "loss": 0.6625,
1148
+ "step": 780
1149
+ },
1150
+ {
1151
+ "epoch": 5.858208955223881,
1152
+ "grad_norm": 0.4221794009208679,
1153
+ "learning_rate": 8.752870536272673e-05,
1154
+ "loss": 0.6743,
1155
+ "step": 785
1156
+ },
1157
+ {
1158
+ "epoch": 5.895522388059701,
1159
+ "grad_norm": 0.4193105399608612,
1160
+ "learning_rate": 8.623748389954283e-05,
1161
+ "loss": 0.6666,
1162
+ "step": 790
1163
+ },
1164
+ {
1165
+ "epoch": 5.932835820895522,
1166
+ "grad_norm": 0.4287867248058319,
1167
+ "learning_rate": 8.494859716729962e-05,
1168
+ "loss": 0.6676,
1169
+ "step": 795
1170
+ },
1171
+ {
1172
+ "epoch": 5.970149253731344,
1173
+ "grad_norm": 0.40571707487106323,
1174
+ "learning_rate": 8.366226381814697e-05,
1175
+ "loss": 0.6691,
1176
+ "step": 800
1177
+ },
1178
+ {
1179
+ "epoch": 6.0,
1180
+ "eval_loss": 1.4930399656295776,
1181
+ "eval_runtime": 0.5186,
1182
+ "eval_samples_per_second": 3.856,
1183
+ "eval_steps_per_second": 1.928,
1184
+ "step": 804
1185
+ },
1186
+ {
1187
+ "epoch": 6.007462686567164,
1188
+ "grad_norm": 0.4177990257740021,
1189
+ "learning_rate": 8.237870207106821e-05,
1190
+ "loss": 0.6566,
1191
+ "step": 805
1192
+ },
1193
+ {
1194
+ "epoch": 6.044776119402985,
1195
+ "grad_norm": 0.4325125813484192,
1196
+ "learning_rate": 8.109812967486025e-05,
1197
+ "loss": 0.6271,
1198
+ "step": 810
1199
+ },
1200
+ {
1201
+ "epoch": 6.082089552238806,
1202
+ "grad_norm": 0.42587417364120483,
1203
+ "learning_rate": 7.982076387119421e-05,
1204
+ "loss": 0.6384,
1205
+ "step": 815
1206
+ },
1207
+ {
1208
+ "epoch": 6.119402985074627,
1209
+ "grad_norm": 0.4367959499359131,
1210
+ "learning_rate": 7.854682135776131e-05,
1211
+ "loss": 0.642,
1212
+ "step": 820
1213
+ },
1214
+ {
1215
+ "epoch": 6.156716417910448,
1216
+ "grad_norm": 0.43222272396087646,
1217
+ "learning_rate": 7.727651825151145e-05,
1218
+ "loss": 0.6314,
1219
+ "step": 825
1220
+ },
1221
+ {
1222
+ "epoch": 6.1940298507462686,
1223
+ "grad_norm": 0.44278955459594727,
1224
+ "learning_rate": 7.601007005199021e-05,
1225
+ "loss": 0.6389,
1226
+ "step": 830
1227
+ },
1228
+ {
1229
+ "epoch": 6.231343283582089,
1230
+ "grad_norm": 0.4457733929157257,
1231
+ "learning_rate": 7.474769160478079e-05,
1232
+ "loss": 0.6411,
1233
+ "step": 835
1234
+ },
1235
+ {
1236
+ "epoch": 6.268656716417911,
1237
+ "grad_norm": 0.4409390091896057,
1238
+ "learning_rate": 7.348959706505626e-05,
1239
+ "loss": 0.6381,
1240
+ "step": 840
1241
+ },
1242
+ {
1243
+ "epoch": 6.3059701492537314,
1244
+ "grad_norm": 0.453755259513855,
1245
+ "learning_rate": 7.223599986124994e-05,
1246
+ "loss": 0.6458,
1247
+ "step": 845
1248
+ },
1249
+ {
1250
+ "epoch": 6.343283582089552,
1251
+ "grad_norm": 0.420446515083313,
1252
+ "learning_rate": 7.09871126588481e-05,
1253
+ "loss": 0.6409,
1254
+ "step": 850
1255
+ },
1256
+ {
1257
+ "epoch": 6.380597014925373,
1258
+ "grad_norm": 0.4332159757614136,
1259
+ "learning_rate": 6.974314732431258e-05,
1260
+ "loss": 0.6396,
1261
+ "step": 855
1262
+ },
1263
+ {
1264
+ "epoch": 6.417910447761194,
1265
+ "grad_norm": 0.44185054302215576,
1266
+ "learning_rate": 6.850431488913895e-05,
1267
+ "loss": 0.6461,
1268
+ "step": 860
1269
+ },
1270
+ {
1271
+ "epoch": 6.455223880597015,
1272
+ "grad_norm": 0.4363011121749878,
1273
+ "learning_rate": 6.727082551405636e-05,
1274
+ "loss": 0.6435,
1275
+ "step": 865
1276
+ },
1277
+ {
1278
+ "epoch": 6.492537313432836,
1279
+ "grad_norm": 0.426645427942276,
1280
+ "learning_rate": 6.604288845337453e-05,
1281
+ "loss": 0.6443,
1282
+ "step": 870
1283
+ },
1284
+ {
1285
+ "epoch": 6.529850746268656,
1286
+ "grad_norm": 0.4306298792362213,
1287
+ "learning_rate": 6.482071201948556e-05,
1288
+ "loss": 0.6405,
1289
+ "step": 875
1290
+ },
1291
+ {
1292
+ "epoch": 6.567164179104478,
1293
+ "grad_norm": 0.42580053210258484,
1294
+ "learning_rate": 6.360450354752458e-05,
1295
+ "loss": 0.636,
1296
+ "step": 880
1297
+ },
1298
+ {
1299
+ "epoch": 6.604477611940299,
1300
+ "grad_norm": 0.4214308261871338,
1301
+ "learning_rate": 6.239446936019657e-05,
1302
+ "loss": 0.6411,
1303
+ "step": 885
1304
+ },
1305
+ {
1306
+ "epoch": 6.641791044776119,
1307
+ "grad_norm": 0.44287511706352234,
1308
+ "learning_rate": 6.119081473277501e-05,
1309
+ "loss": 0.6427,
1310
+ "step": 890
1311
+ },
1312
+ {
1313
+ "epoch": 6.67910447761194,
1314
+ "grad_norm": 0.4150986671447754,
1315
+ "learning_rate": 5.99937438582782e-05,
1316
+ "loss": 0.6472,
1317
+ "step": 895
1318
+ },
1319
+ {
1320
+ "epoch": 6.7164179104477615,
1321
+ "grad_norm": 0.446879118680954,
1322
+ "learning_rate": 5.880345981282876e-05,
1323
+ "loss": 0.6408,
1324
+ "step": 900
1325
+ },
1326
+ {
1327
+ "epoch": 6.753731343283582,
1328
+ "grad_norm": 0.4269532561302185,
1329
+ "learning_rate": 5.7620164521203356e-05,
1330
+ "loss": 0.6367,
1331
+ "step": 905
1332
+ },
1333
+ {
1334
+ "epoch": 6.791044776119403,
1335
+ "grad_norm": 0.4269901216030121,
1336
+ "learning_rate": 5.6444058722577165e-05,
1337
+ "loss": 0.6396,
1338
+ "step": 910
1339
+ },
1340
+ {
1341
+ "epoch": 6.8283582089552235,
1342
+ "grad_norm": 0.42097654938697815,
1343
+ "learning_rate": 5.5275341936469396e-05,
1344
+ "loss": 0.6473,
1345
+ "step": 915
1346
+ },
1347
+ {
1348
+ "epoch": 6.865671641791045,
1349
+ "grad_norm": 0.4283384680747986,
1350
+ "learning_rate": 5.4114212428896424e-05,
1351
+ "loss": 0.6496,
1352
+ "step": 920
1353
+ },
1354
+ {
1355
+ "epoch": 6.902985074626866,
1356
+ "grad_norm": 0.4407792389392853,
1357
+ "learning_rate": 5.296086717873685e-05,
1358
+ "loss": 0.6459,
1359
+ "step": 925
1360
+ },
1361
+ {
1362
+ "epoch": 6.940298507462686,
1363
+ "grad_norm": 0.43775680661201477,
1364
+ "learning_rate": 5.1815501844315105e-05,
1365
+ "loss": 0.6414,
1366
+ "step": 930
1367
+ },
1368
+ {
1369
+ "epoch": 6.977611940298507,
1370
+ "grad_norm": 0.4383552372455597,
1371
+ "learning_rate": 5.0678310730209275e-05,
1372
+ "loss": 0.6389,
1373
+ "step": 935
1374
+ },
1375
+ {
1376
+ "epoch": 7.0,
1377
+ "eval_loss": 1.5040441751480103,
1378
+ "eval_runtime": 0.5191,
1379
+ "eval_samples_per_second": 3.853,
1380
+ "eval_steps_per_second": 1.927,
1381
+ "step": 938
1382
+ },
1383
+ {
1384
+ "epoch": 7.014925373134329,
1385
+ "grad_norm": 0.4299301505088806,
1386
+ "learning_rate": 4.954948675428853e-05,
1387
+ "loss": 0.6411,
1388
+ "step": 940
1389
+ },
1390
+ {
1391
+ "epoch": 7.052238805970149,
1392
+ "grad_norm": 0.44037163257598877,
1393
+ "learning_rate": 4.84292214149855e-05,
1394
+ "loss": 0.6185,
1395
+ "step": 945
1396
+ },
1397
+ {
1398
+ "epoch": 7.08955223880597,
1399
+ "grad_norm": 0.4394110441207886,
1400
+ "learning_rate": 4.7317704758809946e-05,
1401
+ "loss": 0.6158,
1402
+ "step": 950
1403
+ },
1404
+ {
1405
+ "epoch": 7.126865671641791,
1406
+ "grad_norm": 0.441795289516449,
1407
+ "learning_rate": 4.6215125348108247e-05,
1408
+ "loss": 0.6202,
1409
+ "step": 955
1410
+ },
1411
+ {
1412
+ "epoch": 7.164179104477612,
1413
+ "grad_norm": 0.44138744473457336,
1414
+ "learning_rate": 4.512167022907494e-05,
1415
+ "loss": 0.6238,
1416
+ "step": 960
1417
+ },
1418
+ {
1419
+ "epoch": 7.201492537313433,
1420
+ "grad_norm": 0.453387588262558,
1421
+ "learning_rate": 4.40375249000216e-05,
1422
+ "loss": 0.6258,
1423
+ "step": 965
1424
+ },
1425
+ {
1426
+ "epoch": 7.2388059701492535,
1427
+ "grad_norm": 0.44053417444229126,
1428
+ "learning_rate": 4.296287327990797e-05,
1429
+ "loss": 0.6237,
1430
+ "step": 970
1431
+ },
1432
+ {
1433
+ "epoch": 7.276119402985074,
1434
+ "grad_norm": 0.44405415654182434,
1435
+ "learning_rate": 4.189789767714102e-05,
1436
+ "loss": 0.6214,
1437
+ "step": 975
1438
+ },
1439
+ {
1440
+ "epoch": 7.313432835820896,
1441
+ "grad_norm": 0.4466931223869324,
1442
+ "learning_rate": 4.084277875864776e-05,
1443
+ "loss": 0.6258,
1444
+ "step": 980
1445
+ },
1446
+ {
1447
+ "epoch": 7.350746268656716,
1448
+ "grad_norm": 0.45868679881095886,
1449
+ "learning_rate": 3.979769551922592e-05,
1450
+ "loss": 0.6271,
1451
+ "step": 985
1452
+ },
1453
+ {
1454
+ "epoch": 7.388059701492537,
1455
+ "grad_norm": 0.44392016530036926,
1456
+ "learning_rate": 3.876282525117847e-05,
1457
+ "loss": 0.6191,
1458
+ "step": 990
1459
+ },
1460
+ {
1461
+ "epoch": 7.425373134328359,
1462
+ "grad_norm": 0.4744507670402527,
1463
+ "learning_rate": 3.7738343514237106e-05,
1464
+ "loss": 0.6291,
1465
+ "step": 995
1466
+ },
1467
+ {
1468
+ "epoch": 7.462686567164179,
1469
+ "grad_norm": 0.43945175409317017,
1470
+ "learning_rate": 3.672442410577965e-05,
1471
+ "loss": 0.6305,
1472
+ "step": 1000
1473
+ },
1474
+ {
1475
+ "epoch": 7.5,
1476
+ "grad_norm": 0.4612315893173218,
1477
+ "learning_rate": 3.5721239031346066e-05,
1478
+ "loss": 0.6236,
1479
+ "step": 1005
1480
+ },
1481
+ {
1482
+ "epoch": 7.537313432835821,
1483
+ "grad_norm": 0.45296919345855713,
1484
+ "learning_rate": 3.472895847545905e-05,
1485
+ "loss": 0.6282,
1486
+ "step": 1010
1487
+ },
1488
+ {
1489
+ "epoch": 7.574626865671641,
1490
+ "grad_norm": 0.4535280466079712,
1491
+ "learning_rate": 3.374775077275293e-05,
1492
+ "loss": 0.6249,
1493
+ "step": 1015
1494
+ },
1495
+ {
1496
+ "epoch": 7.611940298507463,
1497
+ "grad_norm": 0.4502476155757904,
1498
+ "learning_rate": 3.27777823794168e-05,
1499
+ "loss": 0.6267,
1500
+ "step": 1020
1501
+ },
1502
+ {
1503
+ "epoch": 7.649253731343284,
1504
+ "grad_norm": 0.4433761239051819,
1505
+ "learning_rate": 3.1819217844956214e-05,
1506
+ "loss": 0.6344,
1507
+ "step": 1025
1508
+ },
1509
+ {
1510
+ "epoch": 7.686567164179104,
1511
+ "grad_norm": 0.4449211657047272,
1512
+ "learning_rate": 3.0872219784278354e-05,
1513
+ "loss": 0.6241,
1514
+ "step": 1030
1515
+ },
1516
+ {
1517
+ "epoch": 7.723880597014926,
1518
+ "grad_norm": 0.47005486488342285,
1519
+ "learning_rate": 2.9936948850105152e-05,
1520
+ "loss": 0.6166,
1521
+ "step": 1035
1522
+ },
1523
+ {
1524
+ "epoch": 7.7611940298507465,
1525
+ "grad_norm": 0.4615201950073242,
1526
+ "learning_rate": 2.901356370571967e-05,
1527
+ "loss": 0.6269,
1528
+ "step": 1040
1529
+ },
1530
+ {
1531
+ "epoch": 7.798507462686567,
1532
+ "grad_norm": 0.44223588705062866,
1533
+ "learning_rate": 2.8102220998049843e-05,
1534
+ "loss": 0.6299,
1535
+ "step": 1045
1536
+ },
1537
+ {
1538
+ "epoch": 7.835820895522388,
1539
+ "grad_norm": 0.4457671642303467,
1540
+ "learning_rate": 2.7203075331094017e-05,
1541
+ "loss": 0.627,
1542
+ "step": 1050
1543
+ },
1544
+ {
1545
+ "epoch": 7.8731343283582085,
1546
+ "grad_norm": 0.4446623623371124,
1547
+ "learning_rate": 2.6316279239693465e-05,
1548
+ "loss": 0.6195,
1549
+ "step": 1055
1550
+ },
1551
+ {
1552
+ "epoch": 7.91044776119403,
1553
+ "grad_norm": 0.4469725489616394,
1554
+ "learning_rate": 2.54419831636557e-05,
1555
+ "loss": 0.612,
1556
+ "step": 1060
1557
+ },
1558
+ {
1559
+ "epoch": 7.947761194029851,
1560
+ "grad_norm": 0.4488806426525116,
1561
+ "learning_rate": 2.458033542223316e-05,
1562
+ "loss": 0.6139,
1563
+ "step": 1065
1564
+ },
1565
+ {
1566
+ "epoch": 7.985074626865671,
1567
+ "grad_norm": 0.44639790058135986,
1568
+ "learning_rate": 2.3731482188961818e-05,
1569
+ "loss": 0.6291,
1570
+ "step": 1070
1571
+ },
1572
+ {
1573
+ "epoch": 8.0,
1574
+ "eval_loss": 1.5243113040924072,
1575
+ "eval_runtime": 0.5203,
1576
+ "eval_samples_per_second": 3.844,
1577
+ "eval_steps_per_second": 1.922,
1578
+ "step": 1072
1579
+ },
1580
+ {
1581
+ "epoch": 8.022388059701493,
1582
+ "grad_norm": 0.43840205669403076,
1583
+ "learning_rate": 2.28955674668636e-05,
1584
+ "loss": 0.6144,
1585
+ "step": 1075
1586
+ },
1587
+ {
1588
+ "epoch": 8.059701492537313,
1589
+ "grad_norm": 0.44431746006011963,
1590
+ "learning_rate": 2.2072733064017103e-05,
1591
+ "loss": 0.6062,
1592
+ "step": 1080
1593
+ },
1594
+ {
1595
+ "epoch": 8.097014925373134,
1596
+ "grad_norm": 0.44257909059524536,
1597
+ "learning_rate": 2.1263118569500795e-05,
1598
+ "loss": 0.6067,
1599
+ "step": 1085
1600
+ },
1601
+ {
1602
+ "epoch": 8.134328358208956,
1603
+ "grad_norm": 0.4406292140483856,
1604
+ "learning_rate": 2.0466861329712473e-05,
1605
+ "loss": 0.6169,
1606
+ "step": 1090
1607
+ },
1608
+ {
1609
+ "epoch": 8.171641791044776,
1610
+ "grad_norm": 0.45199882984161377,
1611
+ "learning_rate": 1.968409642506913e-05,
1612
+ "loss": 0.6133,
1613
+ "step": 1095
1614
+ },
1615
+ {
1616
+ "epoch": 8.208955223880597,
1617
+ "grad_norm": 0.4403633773326874,
1618
+ "learning_rate": 1.89149566470915e-05,
1619
+ "loss": 0.6059,
1620
+ "step": 1100
1621
+ },
1622
+ {
1623
+ "epoch": 8.246268656716419,
1624
+ "grad_norm": 0.45532822608947754,
1625
+ "learning_rate": 1.8159572475876718e-05,
1626
+ "loss": 0.6042,
1627
+ "step": 1105
1628
+ },
1629
+ {
1630
+ "epoch": 8.283582089552239,
1631
+ "grad_norm": 0.4559955596923828,
1632
+ "learning_rate": 1.741807205796314e-05,
1633
+ "loss": 0.6189,
1634
+ "step": 1110
1635
+ },
1636
+ {
1637
+ "epoch": 8.32089552238806,
1638
+ "grad_norm": 0.45486971735954285,
1639
+ "learning_rate": 1.6690581184590858e-05,
1640
+ "loss": 0.6116,
1641
+ "step": 1115
1642
+ },
1643
+ {
1644
+ "epoch": 8.35820895522388,
1645
+ "grad_norm": 0.44671645760536194,
1646
+ "learning_rate": 1.5977223270362196e-05,
1647
+ "loss": 0.6149,
1648
+ "step": 1120
1649
+ },
1650
+ {
1651
+ "epoch": 8.395522388059701,
1652
+ "grad_norm": 0.45056024193763733,
1653
+ "learning_rate": 1.5278119332305053e-05,
1654
+ "loss": 0.614,
1655
+ "step": 1125
1656
+ },
1657
+ {
1658
+ "epoch": 8.432835820895523,
1659
+ "grad_norm": 0.4756515920162201,
1660
+ "learning_rate": 1.459338796934293e-05,
1661
+ "loss": 0.6163,
1662
+ "step": 1130
1663
+ },
1664
+ {
1665
+ "epoch": 8.470149253731343,
1666
+ "grad_norm": 0.4487175941467285,
1667
+ "learning_rate": 1.3923145342175482e-05,
1668
+ "loss": 0.6098,
1669
+ "step": 1135
1670
+ },
1671
+ {
1672
+ "epoch": 8.507462686567164,
1673
+ "grad_norm": 0.46132373809814453,
1674
+ "learning_rate": 1.3267505153572501e-05,
1675
+ "loss": 0.6099,
1676
+ "step": 1140
1677
+ },
1678
+ {
1679
+ "epoch": 8.544776119402986,
1680
+ "grad_norm": 0.44755882024765015,
1681
+ "learning_rate": 1.2626578629084784e-05,
1682
+ "loss": 0.6169,
1683
+ "step": 1145
1684
+ },
1685
+ {
1686
+ "epoch": 8.582089552238806,
1687
+ "grad_norm": 0.46651872992515564,
1688
+ "learning_rate": 1.2000474498175552e-05,
1689
+ "loss": 0.6099,
1690
+ "step": 1150
1691
+ },
1692
+ {
1693
+ "epoch": 8.619402985074627,
1694
+ "grad_norm": 0.47415512800216675,
1695
+ "learning_rate": 1.138929897577493e-05,
1696
+ "loss": 0.6104,
1697
+ "step": 1155
1698
+ },
1699
+ {
1700
+ "epoch": 8.656716417910447,
1701
+ "grad_norm": 0.4649696350097656,
1702
+ "learning_rate": 1.0793155744261351e-05,
1703
+ "loss": 0.6094,
1704
+ "step": 1160
1705
+ },
1706
+ {
1707
+ "epoch": 8.694029850746269,
1708
+ "grad_norm": 0.44566088914871216,
1709
+ "learning_rate": 1.0212145935872375e-05,
1710
+ "loss": 0.6223,
1711
+ "step": 1165
1712
+ },
1713
+ {
1714
+ "epoch": 8.73134328358209,
1715
+ "grad_norm": 0.4482034742832184,
1716
+ "learning_rate": 9.646368115548232e-06,
1717
+ "loss": 0.6064,
1718
+ "step": 1170
1719
+ },
1720
+ {
1721
+ "epoch": 8.76865671641791,
1722
+ "grad_norm": 0.4628354012966156,
1723
+ "learning_rate": 9.095918264210779e-06,
1724
+ "loss": 0.6139,
1725
+ "step": 1175
1726
+ },
1727
+ {
1728
+ "epoch": 8.805970149253731,
1729
+ "grad_norm": 0.4561688303947449,
1730
+ "learning_rate": 8.56088976248095e-06,
1731
+ "loss": 0.6141,
1732
+ "step": 1180
1733
+ },
1734
+ {
1735
+ "epoch": 8.843283582089553,
1736
+ "grad_norm": 0.4501047730445862,
1737
+ "learning_rate": 8.041373374837302e-06,
1738
+ "loss": 0.6184,
1739
+ "step": 1185
1740
+ },
1741
+ {
1742
+ "epoch": 8.880597014925373,
1743
+ "grad_norm": 0.4514229893684387,
1744
+ "learning_rate": 7.53745723421827e-06,
1745
+ "loss": 0.6069,
1746
+ "step": 1190
1747
+ },
1748
+ {
1749
+ "epoch": 8.917910447761194,
1750
+ "grad_norm": 0.4529721438884735,
1751
+ "learning_rate": 7.0492268270709875e-06,
1752
+ "loss": 0.6136,
1753
+ "step": 1195
1754
+ },
1755
+ {
1756
+ "epoch": 8.955223880597014,
1757
+ "grad_norm": 0.4562200605869293,
1758
+ "learning_rate": 6.576764978849004e-06,
1759
+ "loss": 0.6053,
1760
+ "step": 1200
1761
+ },
1762
+ {
1763
+ "epoch": 8.992537313432836,
1764
+ "grad_norm": 0.455243319272995,
1765
+ "learning_rate": 6.1201518399613635e-06,
1766
+ "loss": 0.6197,
1767
+ "step": 1205
1768
+ },
1769
+ {
1770
+ "epoch": 9.0,
1771
+ "eval_loss": 1.5375313758850098,
1772
+ "eval_runtime": 0.5174,
1773
+ "eval_samples_per_second": 3.865,
1774
+ "eval_steps_per_second": 1.933,
1775
+ "step": 1206
1776
+ },
1777
+ {
1778
+ "epoch": 9.029850746268657,
1779
+ "grad_norm": 0.44451919198036194,
1780
+ "learning_rate": 5.679464872175666e-06,
1781
+ "loss": 0.6126,
1782
+ "step": 1210
1783
+ },
1784
+ {
1785
+ "epoch": 9.067164179104477,
1786
+ "grad_norm": 0.4510916471481323,
1787
+ "learning_rate": 5.254778835477015e-06,
1788
+ "loss": 0.619,
1789
+ "step": 1215
1790
+ },
1791
+ {
1792
+ "epoch": 9.104477611940299,
1793
+ "grad_norm": 0.4504685699939728,
1794
+ "learning_rate": 4.846165775385459e-06,
1795
+ "loss": 0.6059,
1796
+ "step": 1220
1797
+ },
1798
+ {
1799
+ "epoch": 9.14179104477612,
1800
+ "grad_norm": 0.4434657692909241,
1801
+ "learning_rate": 4.4536950107339605e-06,
1802
+ "loss": 0.601,
1803
+ "step": 1225
1804
+ },
1805
+ {
1806
+ "epoch": 9.17910447761194,
1807
+ "grad_norm": 0.46363726258277893,
1808
+ "learning_rate": 4.077433121908747e-06,
1809
+ "loss": 0.6165,
1810
+ "step": 1230
1811
+ },
1812
+ {
1813
+ "epoch": 9.216417910447761,
1814
+ "grad_norm": 0.45644015073776245,
1815
+ "learning_rate": 3.717443939554388e-06,
1816
+ "loss": 0.6102,
1817
+ "step": 1235
1818
+ },
1819
+ {
1820
+ "epoch": 9.253731343283581,
1821
+ "grad_norm": 0.45389410853385925,
1822
+ "learning_rate": 3.3737885337452814e-06,
1823
+ "loss": 0.6029,
1824
+ "step": 1240
1825
+ },
1826
+ {
1827
+ "epoch": 9.291044776119403,
1828
+ "grad_norm": 0.4506882131099701,
1829
+ "learning_rate": 3.0465252036255165e-06,
1830
+ "loss": 0.6091,
1831
+ "step": 1245
1832
+ },
1833
+ {
1834
+ "epoch": 9.328358208955224,
1835
+ "grad_norm": 0.44706588983535767,
1836
+ "learning_rate": 2.735709467518699e-06,
1837
+ "loss": 0.6114,
1838
+ "step": 1250
1839
+ },
1840
+ {
1841
+ "epoch": 9.365671641791044,
1842
+ "grad_norm": 0.45065855979919434,
1843
+ "learning_rate": 2.4413940535096156e-06,
1844
+ "loss": 0.603,
1845
+ "step": 1255
1846
+ },
1847
+ {
1848
+ "epoch": 9.402985074626866,
1849
+ "grad_norm": 0.4460483491420746,
1850
+ "learning_rate": 2.1636288904992585e-06,
1851
+ "loss": 0.6019,
1852
+ "step": 1260
1853
+ },
1854
+ {
1855
+ "epoch": 9.440298507462687,
1856
+ "grad_norm": 0.4578142762184143,
1857
+ "learning_rate": 1.902461099734587e-06,
1858
+ "loss": 0.6065,
1859
+ "step": 1265
1860
+ },
1861
+ {
1862
+ "epoch": 9.477611940298507,
1863
+ "grad_norm": 0.46080294251441956,
1864
+ "learning_rate": 1.6579349868147687e-06,
1865
+ "loss": 0.5979,
1866
+ "step": 1270
1867
+ },
1868
+ {
1869
+ "epoch": 9.514925373134329,
1870
+ "grad_norm": 0.451742559671402,
1871
+ "learning_rate": 1.430092034174979e-06,
1872
+ "loss": 0.6133,
1873
+ "step": 1275
1874
+ },
1875
+ {
1876
+ "epoch": 9.552238805970148,
1877
+ "grad_norm": 0.4507807493209839,
1878
+ "learning_rate": 1.2189708940490652e-06,
1879
+ "loss": 0.604,
1880
+ "step": 1280
1881
+ },
1882
+ {
1883
+ "epoch": 9.58955223880597,
1884
+ "grad_norm": 0.45165741443634033,
1885
+ "learning_rate": 1.0246073819125345e-06,
1886
+ "loss": 0.604,
1887
+ "step": 1285
1888
+ },
1889
+ {
1890
+ "epoch": 9.626865671641792,
1891
+ "grad_norm": 0.45481327176094055,
1892
+ "learning_rate": 8.470344704066046e-07,
1893
+ "loss": 0.6077,
1894
+ "step": 1290
1895
+ },
1896
+ {
1897
+ "epoch": 9.664179104477611,
1898
+ "grad_norm": 0.4481595456600189,
1899
+ "learning_rate": 6.862822837445881e-07,
1900
+ "loss": 0.6021,
1901
+ "step": 1295
1902
+ },
1903
+ {
1904
+ "epoch": 9.701492537313433,
1905
+ "grad_norm": 0.4543148875236511,
1906
+ "learning_rate": 5.42378092601481e-07,
1907
+ "loss": 0.5979,
1908
+ "step": 1300
1909
+ },
1910
+ {
1911
+ "epoch": 9.738805970149254,
1912
+ "grad_norm": 0.4538787305355072,
1913
+ "learning_rate": 4.153463094877186e-07,
1914
+ "loss": 0.6013,
1915
+ "step": 1305
1916
+ },
1917
+ {
1918
+ "epoch": 9.776119402985074,
1919
+ "grad_norm": 0.4568292796611786,
1920
+ "learning_rate": 3.0520848460765527e-07,
1921
+ "loss": 0.6035,
1922
+ "step": 1310
1923
+ },
1924
+ {
1925
+ "epoch": 9.813432835820896,
1926
+ "grad_norm": 0.4435996115207672,
1927
+ "learning_rate": 2.119833022037221e-07,
1928
+ "loss": 0.6022,
1929
+ "step": 1315
1930
+ },
1931
+ {
1932
+ "epoch": 9.850746268656717,
1933
+ "grad_norm": 0.4439810514450073,
1934
+ "learning_rate": 1.3568657738678435e-07,
1935
+ "loss": 0.6025,
1936
+ "step": 1320
1937
+ },
1938
+ {
1939
+ "epoch": 9.888059701492537,
1940
+ "grad_norm": 0.4461255371570587,
1941
+ "learning_rate": 7.633125345317682e-08,
1942
+ "loss": 0.5931,
1943
+ "step": 1325
1944
+ },
1945
+ {
1946
+ "epoch": 9.925373134328359,
1947
+ "grad_norm": 0.4546414315700531,
1948
+ "learning_rate": 3.392739968894887e-08,
1949
+ "loss": 0.6182,
1950
+ "step": 1330
1951
+ },
1952
+ {
1953
+ "epoch": 9.962686567164178,
1954
+ "grad_norm": 0.45576563477516174,
1955
+ "learning_rate": 8.48220966164215e-09,
1956
+ "loss": 0.6126,
1957
+ "step": 1335
1958
+ },
1959
+ {
1960
+ "epoch": 10.0,
1961
+ "grad_norm": 0.4497128427028656,
1962
+ "learning_rate": 0.0,
1963
+ "loss": 0.6028,
1964
+ "step": 1340
1965
+ },
1966
+ {
1967
+ "epoch": 10.0,
1968
+ "eval_loss": 1.542757511138916,
1969
+ "eval_runtime": 0.5408,
1970
+ "eval_samples_per_second": 3.698,
1971
+ "eval_steps_per_second": 1.849,
1972
+ "step": 1340
1973
+ },
1974
+ {
1975
+ "epoch": 10.0,
1976
+ "step": 1340,
1977
+ "total_flos": 3.755518175798624e+18,
1978
+ "train_loss": 0.7269157813556159,
1979
+ "train_runtime": 7393.4759,
1980
+ "train_samples_per_second": 11.586,
1981
+ "train_steps_per_second": 0.181
1982
+ }
1983
+ ],
1984
+ "logging_steps": 5,
1985
+ "max_steps": 1340,
1986
+ "num_input_tokens_seen": 0,
1987
+ "num_train_epochs": 10,
1988
+ "save_steps": 100,
1989
+ "stateful_callbacks": {
1990
+ "TrainerControl": {
1991
+ "args": {
1992
+ "should_epoch_stop": false,
1993
+ "should_evaluate": false,
1994
+ "should_log": false,
1995
+ "should_save": true,
1996
+ "should_training_stop": true
1997
+ },
1998
+ "attributes": {}
1999
+ }
2000
+ },
2001
+ "total_flos": 3.755518175798624e+18,
2002
+ "train_batch_size": 8,
2003
+ "trial_name": null,
2004
+ "trial_params": null
2005
+ }