Hugging Face's logo

language-and-voice-lab
/
bloom560ftrmh

Text Generation
Transformers
PyTorch
Icelandic
bloom
text-generation-inference
Inference Endpoints
Model card Files Community
stefanola commited on
Commit
e55ada6
•
1 Parent(s): d16224e

Upload 10 files

Browse files
Files changed (11) hide show
  1. .gitattributes +1 -0
  2. all_results.json +8 -0
  3. config.json +32 -0
  4. generation_config.json +8 -0
  5. pytorch_model.bin +3 -0
  6. special_tokens_map.json +7 -0
  7. tokenizer.json +3 -0
  8. tokenizer_config.json +11 -0
  9. train_results.json +8 -0
  10. trainer_state.json +2023 -0
  11. training_args.bin +3 -0
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.8373472213664984,
4
+ "train_runtime": 584567.9437,
5
+ "train_samples": 10663527,
6
+ "train_samples_per_second": 18.242,
7
+ "train_steps_per_second": 0.285
8
+ }
config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bigscience/bloom-560m",
3
+ "apply_residual_connection_post_layernorm": false,
4
+ "architectures": [
5
+ "BloomForCausalLM"
6
+ ],
7
+ "attention_dropout": 0.0,
8
+ "attention_softmax_in_fp32": true,
9
+ "bias_dropout_fusion": true,
10
+ "bos_token_id": 1,
11
+ "eos_token_id": 2,
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 1024,
14
+ "initializer_range": 0.02,
15
+ "layer_norm_epsilon": 1e-05,
16
+ "masked_softmax_fusion": true,
17
+ "model_type": "bloom",
18
+ "n_head": 16,
19
+ "n_inner": null,
20
+ "n_layer": 24,
21
+ "offset_alibi": 100,
22
+ "pad_token_id": 3,
23
+ "pretraining_tp": 1,
24
+ "skip_bias_add": true,
25
+ "skip_bias_add_qkv": false,
26
+ "slow_but_exact": false,
27
+ "torch_dtype": "float16",
28
+ "transformers_version": "4.29.0.dev0",
29
+ "unk_token_id": 0,
30
+ "use_cache": false,
31
+ "vocab_size": 250680
32
+ }
generation_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 3,
6
+ "transformers_version": "4.29.0.dev0",
7
+ "use_cache": false
8
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3360a8f7163a594a54ef180a6119f5a33435db045dab64a8e4b1c8720db57c20
3
+ size 1118059773
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "mask_token": "<unk>",
5
+ "pad_token": "<pad>",
6
+ "unk_token": "<unk>"
7
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17a208233d2ee8d8c83b23bc214df737c44806a1919f444e89b31e586cd956ba
3
+ size 14500471
tokenizer_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<s>",
4
+ "clean_up_tokenization_spaces": false,
5
+ "eos_token": "</s>",
6
+ "model_max_length": 1000000000000000019884624838656,
7
+ "pad_token": "<pad>",
8
+ "padding_side": "left",
9
+ "tokenizer_class": "BloomTokenizer",
10
+ "unk_token": "<unk>"
11
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.8373472213664984,
4
+ "train_runtime": 584567.9437,
5
+ "train_samples": 10663527,
6
+ "train_samples_per_second": 18.242,
7
+ "train_steps_per_second": 0.285
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9999962488962377,
5
+ "global_step": 166617,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 5e-05,
13
+ "loss": 3.0218,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 5e-05,
19
+ "loss": 2.376,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 5e-05,
25
+ "loss": 2.2374,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 5e-05,
31
+ "loss": 2.1414,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 5e-05,
37
+ "loss": 2.0934,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 5e-05,
43
+ "loss": 2.0597,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.02,
48
+ "learning_rate": 5e-05,
49
+ "loss": 2.0249,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.02,
54
+ "learning_rate": 5e-05,
55
+ "loss": 1.9947,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 5e-05,
61
+ "loss": 1.969,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.03,
66
+ "learning_rate": 5e-05,
67
+ "loss": 1.9525,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.03,
72
+ "learning_rate": 5e-05,
73
+ "loss": 1.9319,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.04,
78
+ "learning_rate": 5e-05,
79
+ "loss": 1.9208,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.04,
84
+ "learning_rate": 5e-05,
85
+ "loss": 1.9116,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.04,
90
+ "learning_rate": 5e-05,
91
+ "loss": 1.899,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.05,
96
+ "learning_rate": 5e-05,
97
+ "loss": 1.8899,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.05,
102
+ "learning_rate": 5e-05,
103
+ "loss": 1.8775,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.05,
108
+ "learning_rate": 5e-05,
109
+ "loss": 1.8701,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.05,
114
+ "learning_rate": 5e-05,
115
+ "loss": 1.8654,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.06,
120
+ "learning_rate": 5e-05,
121
+ "loss": 1.8557,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.06,
126
+ "learning_rate": 5e-05,
127
+ "loss": 1.8487,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.06,
132
+ "learning_rate": 5e-05,
133
+ "loss": 1.8401,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.07,
138
+ "learning_rate": 5e-05,
139
+ "loss": 1.8307,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.07,
144
+ "learning_rate": 5e-05,
145
+ "loss": 1.8303,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.07,
150
+ "learning_rate": 5e-05,
151
+ "loss": 1.8217,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.08,
156
+ "learning_rate": 5e-05,
157
+ "loss": 1.8192,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.08,
162
+ "learning_rate": 5e-05,
163
+ "loss": 1.8156,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.08,
168
+ "learning_rate": 5e-05,
169
+ "loss": 1.8077,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.08,
174
+ "learning_rate": 5e-05,
175
+ "loss": 1.802,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.09,
180
+ "learning_rate": 5e-05,
181
+ "loss": 1.7997,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.09,
186
+ "learning_rate": 5e-05,
187
+ "loss": 1.7962,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.09,
192
+ "learning_rate": 5e-05,
193
+ "loss": 1.7918,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.1,
198
+ "learning_rate": 5e-05,
199
+ "loss": 1.7898,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.1,
204
+ "learning_rate": 5e-05,
205
+ "loss": 1.7861,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.1,
210
+ "learning_rate": 5e-05,
211
+ "loss": 1.7817,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.11,
216
+ "learning_rate": 5e-05,
217
+ "loss": 1.7743,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.11,
222
+ "learning_rate": 5e-05,
223
+ "loss": 1.7747,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.11,
228
+ "learning_rate": 5e-05,
229
+ "loss": 1.7718,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.11,
234
+ "learning_rate": 5e-05,
235
+ "loss": 1.7672,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.12,
240
+ "learning_rate": 5e-05,
241
+ "loss": 1.7676,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.12,
246
+ "learning_rate": 5e-05,
247
+ "loss": 1.7618,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.12,
252
+ "learning_rate": 5e-05,
253
+ "loss": 1.7554,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.13,
258
+ "learning_rate": 5e-05,
259
+ "loss": 1.7548,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.13,
264
+ "learning_rate": 5e-05,
265
+ "loss": 1.7542,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.13,
270
+ "learning_rate": 5e-05,
271
+ "loss": 1.7487,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.14,
276
+ "learning_rate": 5e-05,
277
+ "loss": 1.7504,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.14,
282
+ "learning_rate": 5e-05,
283
+ "loss": 1.7464,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.14,
288
+ "learning_rate": 5e-05,
289
+ "loss": 1.7497,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.14,
294
+ "learning_rate": 5e-05,
295
+ "loss": 1.7451,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.15,
300
+ "learning_rate": 5e-05,
301
+ "loss": 1.745,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.15,
306
+ "learning_rate": 5e-05,
307
+ "loss": 1.7409,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.15,
312
+ "learning_rate": 5e-05,
313
+ "loss": 1.7416,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.16,
318
+ "learning_rate": 5e-05,
319
+ "loss": 1.7341,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.16,
324
+ "learning_rate": 5e-05,
325
+ "loss": 1.7313,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.16,
330
+ "learning_rate": 5e-05,
331
+ "loss": 1.7351,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.17,
336
+ "learning_rate": 5e-05,
337
+ "loss": 1.7269,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.17,
342
+ "learning_rate": 5e-05,
343
+ "loss": 1.7292,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.17,
348
+ "learning_rate": 5e-05,
349
+ "loss": 1.7279,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.17,
354
+ "learning_rate": 5e-05,
355
+ "loss": 1.7259,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.18,
360
+ "learning_rate": 5e-05,
361
+ "loss": 1.7254,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.18,
366
+ "learning_rate": 5e-05,
367
+ "loss": 1.7262,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.18,
372
+ "learning_rate": 5e-05,
373
+ "loss": 1.7213,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.19,
378
+ "learning_rate": 5e-05,
379
+ "loss": 1.7207,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.19,
384
+ "learning_rate": 5e-05,
385
+ "loss": 1.72,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.19,
390
+ "learning_rate": 5e-05,
391
+ "loss": 1.7178,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.2,
396
+ "learning_rate": 5e-05,
397
+ "loss": 1.71,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.2,
402
+ "learning_rate": 5e-05,
403
+ "loss": 1.7128,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.2,
408
+ "learning_rate": 5e-05,
409
+ "loss": 1.7101,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.2,
414
+ "learning_rate": 5e-05,
415
+ "loss": 1.7108,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.21,
420
+ "learning_rate": 5e-05,
421
+ "loss": 1.7105,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.21,
426
+ "learning_rate": 5e-05,
427
+ "loss": 1.7076,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.21,
432
+ "learning_rate": 5e-05,
433
+ "loss": 1.7055,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.22,
438
+ "learning_rate": 5e-05,
439
+ "loss": 1.7068,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.22,
444
+ "learning_rate": 5e-05,
445
+ "loss": 1.707,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.22,
450
+ "learning_rate": 5e-05,
451
+ "loss": 1.7009,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.23,
456
+ "learning_rate": 5e-05,
457
+ "loss": 1.6997,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.23,
462
+ "learning_rate": 5e-05,
463
+ "loss": 1.6973,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.23,
468
+ "learning_rate": 5e-05,
469
+ "loss": 1.6963,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.23,
474
+ "learning_rate": 5e-05,
475
+ "loss": 1.6965,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.24,
480
+ "learning_rate": 5e-05,
481
+ "loss": 1.6944,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.24,
486
+ "learning_rate": 5e-05,
487
+ "loss": 1.6976,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.24,
492
+ "learning_rate": 5e-05,
493
+ "loss": 1.6853,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.25,
498
+ "learning_rate": 5e-05,
499
+ "loss": 1.6877,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.25,
504
+ "learning_rate": 5e-05,
505
+ "loss": 1.6891,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.25,
510
+ "learning_rate": 5e-05,
511
+ "loss": 1.6898,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.26,
516
+ "learning_rate": 5e-05,
517
+ "loss": 1.6844,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.26,
522
+ "learning_rate": 5e-05,
523
+ "loss": 1.6861,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.26,
528
+ "learning_rate": 5e-05,
529
+ "loss": 1.6819,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.26,
534
+ "learning_rate": 5e-05,
535
+ "loss": 1.6905,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.27,
540
+ "learning_rate": 5e-05,
541
+ "loss": 1.6818,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.27,
546
+ "learning_rate": 5e-05,
547
+ "loss": 1.6855,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.27,
552
+ "learning_rate": 5e-05,
553
+ "loss": 1.6987,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.28,
558
+ "learning_rate": 5e-05,
559
+ "loss": 1.6898,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.28,
564
+ "learning_rate": 5e-05,
565
+ "loss": 1.6831,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.28,
570
+ "learning_rate": 5e-05,
571
+ "loss": 1.6782,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.29,
576
+ "learning_rate": 5e-05,
577
+ "loss": 1.6789,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.29,
582
+ "learning_rate": 5e-05,
583
+ "loss": 1.6766,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.29,
588
+ "learning_rate": 5e-05,
589
+ "loss": 1.6708,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.29,
594
+ "learning_rate": 5e-05,
595
+ "loss": 1.6719,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.3,
600
+ "learning_rate": 5e-05,
601
+ "loss": 1.6703,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.3,
606
+ "learning_rate": 5e-05,
607
+ "loss": 1.667,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.3,
612
+ "learning_rate": 5e-05,
613
+ "loss": 1.6622,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.31,
618
+ "learning_rate": 5e-05,
619
+ "loss": 1.6677,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.31,
624
+ "learning_rate": 5e-05,
625
+ "loss": 1.6671,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.31,
630
+ "learning_rate": 5e-05,
631
+ "loss": 1.665,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.32,
636
+ "learning_rate": 5e-05,
637
+ "loss": 1.6644,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.32,
642
+ "learning_rate": 5e-05,
643
+ "loss": 1.6605,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.32,
648
+ "learning_rate": 5e-05,
649
+ "loss": 1.6554,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.32,
654
+ "learning_rate": 5e-05,
655
+ "loss": 1.659,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.33,
660
+ "learning_rate": 5e-05,
661
+ "loss": 1.6573,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.33,
666
+ "learning_rate": 5e-05,
667
+ "loss": 1.6538,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.33,
672
+ "learning_rate": 5e-05,
673
+ "loss": 1.6549,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.34,
678
+ "learning_rate": 5e-05,
679
+ "loss": 1.6529,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.34,
684
+ "learning_rate": 5e-05,
685
+ "loss": 1.6565,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.34,
690
+ "learning_rate": 5e-05,
691
+ "loss": 1.6556,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.35,
696
+ "learning_rate": 5e-05,
697
+ "loss": 1.6513,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.35,
702
+ "learning_rate": 5e-05,
703
+ "loss": 1.6533,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.35,
708
+ "learning_rate": 5e-05,
709
+ "loss": 1.6517,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.35,
714
+ "learning_rate": 5e-05,
715
+ "loss": 1.6497,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.36,
720
+ "learning_rate": 5e-05,
721
+ "loss": 1.6454,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.36,
726
+ "learning_rate": 5e-05,
727
+ "loss": 1.6492,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.36,
732
+ "learning_rate": 5e-05,
733
+ "loss": 1.6468,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.37,
738
+ "learning_rate": 5e-05,
739
+ "loss": 1.643,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.37,
744
+ "learning_rate": 5e-05,
745
+ "loss": 1.6443,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.37,
750
+ "learning_rate": 5e-05,
751
+ "loss": 1.6452,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.38,
756
+ "learning_rate": 5e-05,
757
+ "loss": 1.6479,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.38,
762
+ "learning_rate": 5e-05,
763
+ "loss": 1.6467,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.38,
768
+ "learning_rate": 5e-05,
769
+ "loss": 1.6442,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.38,
774
+ "learning_rate": 5e-05,
775
+ "loss": 1.6393,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.39,
780
+ "learning_rate": 5e-05,
781
+ "loss": 1.639,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.39,
786
+ "learning_rate": 5e-05,
787
+ "loss": 1.639,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.39,
792
+ "learning_rate": 5e-05,
793
+ "loss": 1.6413,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.4,
798
+ "learning_rate": 5e-05,
799
+ "loss": 1.6402,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.4,
804
+ "learning_rate": 5e-05,
805
+ "loss": 1.6377,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.4,
810
+ "learning_rate": 5e-05,
811
+ "loss": 1.6341,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.41,
816
+ "learning_rate": 5e-05,
817
+ "loss": 1.6347,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.41,
822
+ "learning_rate": 5e-05,
823
+ "loss": 1.6432,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.41,
828
+ "learning_rate": 5e-05,
829
+ "loss": 1.6381,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.41,
834
+ "learning_rate": 5e-05,
835
+ "loss": 1.6327,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.42,
840
+ "learning_rate": 5e-05,
841
+ "loss": 1.6392,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.42,
846
+ "learning_rate": 5e-05,
847
+ "loss": 1.6367,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.42,
852
+ "learning_rate": 5e-05,
853
+ "loss": 1.6341,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.43,
858
+ "learning_rate": 5e-05,
859
+ "loss": 1.6363,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.43,
864
+ "learning_rate": 5e-05,
865
+ "loss": 1.6303,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.43,
870
+ "learning_rate": 5e-05,
871
+ "loss": 1.6307,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.44,
876
+ "learning_rate": 5e-05,
877
+ "loss": 1.633,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.44,
882
+ "learning_rate": 5e-05,
883
+ "loss": 1.6341,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.44,
888
+ "learning_rate": 5e-05,
889
+ "loss": 1.6289,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.44,
894
+ "learning_rate": 5e-05,
895
+ "loss": 1.6357,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.45,
900
+ "learning_rate": 5e-05,
901
+ "loss": 1.6292,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 0.45,
906
+ "learning_rate": 5e-05,
907
+ "loss": 1.629,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 0.45,
912
+ "learning_rate": 5e-05,
913
+ "loss": 1.6286,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 0.46,
918
+ "learning_rate": 5e-05,
919
+ "loss": 1.6318,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 0.46,
924
+ "learning_rate": 5e-05,
925
+ "loss": 1.6278,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 0.46,
930
+ "learning_rate": 5e-05,
931
+ "loss": 1.6252,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 0.47,
936
+ "learning_rate": 5e-05,
937
+ "loss": 1.6302,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 0.47,
942
+ "learning_rate": 5e-05,
943
+ "loss": 1.6286,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 0.47,
948
+ "learning_rate": 5e-05,
949
+ "loss": 1.629,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 0.47,
954
+ "learning_rate": 5e-05,
955
+ "loss": 1.6246,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 0.48,
960
+ "learning_rate": 5e-05,
961
+ "loss": 1.6275,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 0.48,
966
+ "learning_rate": 5e-05,
967
+ "loss": 1.6251,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 0.48,
972
+ "learning_rate": 5e-05,
973
+ "loss": 1.6248,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 0.49,
978
+ "learning_rate": 5e-05,
979
+ "loss": 1.6224,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 0.49,
984
+ "learning_rate": 5e-05,
985
+ "loss": 1.6265,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 0.49,
990
+ "learning_rate": 5e-05,
991
+ "loss": 1.6228,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 0.5,
996
+ "learning_rate": 5e-05,
997
+ "loss": 1.6258,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 0.5,
1002
+ "learning_rate": 5e-05,
1003
+ "loss": 1.6215,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 0.5,
1008
+ "learning_rate": 5e-05,
1009
+ "loss": 1.6231,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 0.5,
1014
+ "learning_rate": 5e-05,
1015
+ "loss": 1.6136,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 0.51,
1020
+ "learning_rate": 5e-05,
1021
+ "loss": 1.6216,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 0.51,
1026
+ "learning_rate": 5e-05,
1027
+ "loss": 1.6201,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 0.51,
1032
+ "learning_rate": 5e-05,
1033
+ "loss": 1.6197,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 0.52,
1038
+ "learning_rate": 5e-05,
1039
+ "loss": 1.6164,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 0.52,
1044
+ "learning_rate": 5e-05,
1045
+ "loss": 1.6236,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 0.52,
1050
+ "learning_rate": 5e-05,
1051
+ "loss": 1.6169,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 0.53,
1056
+ "learning_rate": 5e-05,
1057
+ "loss": 1.6159,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 0.53,
1062
+ "learning_rate": 5e-05,
1063
+ "loss": 1.616,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 0.53,
1068
+ "learning_rate": 5e-05,
1069
+ "loss": 1.6125,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 0.53,
1074
+ "learning_rate": 5e-05,
1075
+ "loss": 1.6174,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 0.54,
1080
+ "learning_rate": 5e-05,
1081
+ "loss": 1.6174,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 0.54,
1086
+ "learning_rate": 5e-05,
1087
+ "loss": 1.6125,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 0.54,
1092
+ "learning_rate": 5e-05,
1093
+ "loss": 1.6177,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 0.55,
1098
+ "learning_rate": 5e-05,
1099
+ "loss": 1.6111,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 0.55,
1104
+ "learning_rate": 5e-05,
1105
+ "loss": 1.6166,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 0.55,
1110
+ "learning_rate": 5e-05,
1111
+ "loss": 1.6126,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 0.56,
1116
+ "learning_rate": 5e-05,
1117
+ "loss": 1.6142,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 0.56,
1122
+ "learning_rate": 5e-05,
1123
+ "loss": 1.6111,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 0.56,
1128
+ "learning_rate": 5e-05,
1129
+ "loss": 1.6117,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 0.56,
1134
+ "learning_rate": 5e-05,
1135
+ "loss": 1.6123,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 0.57,
1140
+ "learning_rate": 5e-05,
1141
+ "loss": 1.6146,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 0.57,
1146
+ "learning_rate": 5e-05,
1147
+ "loss": 1.6097,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 0.57,
1152
+ "learning_rate": 5e-05,
1153
+ "loss": 1.6109,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 0.58,
1158
+ "learning_rate": 5e-05,
1159
+ "loss": 1.612,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 0.58,
1164
+ "learning_rate": 5e-05,
1165
+ "loss": 1.6137,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 0.58,
1170
+ "learning_rate": 5e-05,
1171
+ "loss": 1.6088,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 0.59,
1176
+ "learning_rate": 5e-05,
1177
+ "loss": 1.6127,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 0.59,
1182
+ "learning_rate": 5e-05,
1183
+ "loss": 1.608,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 0.59,
1188
+ "learning_rate": 5e-05,
1189
+ "loss": 1.6069,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 0.59,
1194
+ "learning_rate": 5e-05,
1195
+ "loss": 1.6115,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 0.6,
1200
+ "learning_rate": 5e-05,
1201
+ "loss": 1.608,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 0.6,
1206
+ "learning_rate": 5e-05,
1207
+ "loss": 1.6049,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 0.6,
1212
+ "learning_rate": 5e-05,
1213
+ "loss": 1.6104,
1214
+ "step": 100500
1215
+ },
1216
+ {
1217
+ "epoch": 0.61,
1218
+ "learning_rate": 5e-05,
1219
+ "loss": 1.607,
1220
+ "step": 101000
1221
+ },
1222
+ {
1223
+ "epoch": 0.61,
1224
+ "learning_rate": 5e-05,
1225
+ "loss": 1.6036,
1226
+ "step": 101500
1227
+ },
1228
+ {
1229
+ "epoch": 0.61,
1230
+ "learning_rate": 5e-05,
1231
+ "loss": 1.6053,
1232
+ "step": 102000
1233
+ },
1234
+ {
1235
+ "epoch": 0.62,
1236
+ "learning_rate": 5e-05,
1237
+ "loss": 1.6072,
1238
+ "step": 102500
1239
+ },
1240
+ {
1241
+ "epoch": 0.62,
1242
+ "learning_rate": 5e-05,
1243
+ "loss": 1.6054,
1244
+ "step": 103000
1245
+ },
1246
+ {
1247
+ "epoch": 0.62,
1248
+ "learning_rate": 5e-05,
1249
+ "loss": 1.6039,
1250
+ "step": 103500
1251
+ },
1252
+ {
1253
+ "epoch": 0.62,
1254
+ "learning_rate": 5e-05,
1255
+ "loss": 1.606,
1256
+ "step": 104000
1257
+ },
1258
+ {
1259
+ "epoch": 0.63,
1260
+ "learning_rate": 5e-05,
1261
+ "loss": 1.6046,
1262
+ "step": 104500
1263
+ },
1264
+ {
1265
+ "epoch": 0.63,
1266
+ "learning_rate": 5e-05,
1267
+ "loss": 1.6053,
1268
+ "step": 105000
1269
+ },
1270
+ {
1271
+ "epoch": 0.63,
1272
+ "learning_rate": 5e-05,
1273
+ "loss": 1.6047,
1274
+ "step": 105500
1275
+ },
1276
+ {
1277
+ "epoch": 0.64,
1278
+ "learning_rate": 5e-05,
1279
+ "loss": 1.6021,
1280
+ "step": 106000
1281
+ },
1282
+ {
1283
+ "epoch": 0.64,
1284
+ "learning_rate": 5e-05,
1285
+ "loss": 1.6033,
1286
+ "step": 106500
1287
+ },
1288
+ {
1289
+ "epoch": 0.64,
1290
+ "learning_rate": 5e-05,
1291
+ "loss": 1.602,
1292
+ "step": 107000
1293
+ },
1294
+ {
1295
+ "epoch": 0.65,
1296
+ "learning_rate": 5e-05,
1297
+ "loss": 1.6026,
1298
+ "step": 107500
1299
+ },
1300
+ {
1301
+ "epoch": 0.65,
1302
+ "learning_rate": 5e-05,
1303
+ "loss": 1.6046,
1304
+ "step": 108000
1305
+ },
1306
+ {
1307
+ "epoch": 0.65,
1308
+ "learning_rate": 5e-05,
1309
+ "loss": 1.6088,
1310
+ "step": 108500
1311
+ },
1312
+ {
1313
+ "epoch": 0.65,
1314
+ "learning_rate": 5e-05,
1315
+ "loss": 1.6035,
1316
+ "step": 109000
1317
+ },
1318
+ {
1319
+ "epoch": 0.66,
1320
+ "learning_rate": 5e-05,
1321
+ "loss": 1.6004,
1322
+ "step": 109500
1323
+ },
1324
+ {
1325
+ "epoch": 0.66,
1326
+ "learning_rate": 5e-05,
1327
+ "loss": 1.6015,
1328
+ "step": 110000
1329
+ },
1330
+ {
1331
+ "epoch": 0.66,
1332
+ "learning_rate": 5e-05,
1333
+ "loss": 1.6001,
1334
+ "step": 110500
1335
+ },
1336
+ {
1337
+ "epoch": 0.67,
1338
+ "learning_rate": 5e-05,
1339
+ "loss": 1.6006,
1340
+ "step": 111000
1341
+ },
1342
+ {
1343
+ "epoch": 0.67,
1344
+ "learning_rate": 5e-05,
1345
+ "loss": 1.6007,
1346
+ "step": 111500
1347
+ },
1348
+ {
1349
+ "epoch": 0.67,
1350
+ "learning_rate": 5e-05,
1351
+ "loss": 1.6023,
1352
+ "step": 112000
1353
+ },
1354
+ {
1355
+ "epoch": 0.68,
1356
+ "learning_rate": 5e-05,
1357
+ "loss": 1.5983,
1358
+ "step": 112500
1359
+ },
1360
+ {
1361
+ "epoch": 0.68,
1362
+ "learning_rate": 5e-05,
1363
+ "loss": 1.5976,
1364
+ "step": 113000
1365
+ },
1366
+ {
1367
+ "epoch": 0.68,
1368
+ "learning_rate": 5e-05,
1369
+ "loss": 1.6027,
1370
+ "step": 113500
1371
+ },
1372
+ {
1373
+ "epoch": 0.68,
1374
+ "learning_rate": 5e-05,
1375
+ "loss": 1.6017,
1376
+ "step": 114000
1377
+ },
1378
+ {
1379
+ "epoch": 0.69,
1380
+ "learning_rate": 5e-05,
1381
+ "loss": 1.5954,
1382
+ "step": 114500
1383
+ },
1384
+ {
1385
+ "epoch": 0.69,
1386
+ "learning_rate": 5e-05,
1387
+ "loss": 1.6038,
1388
+ "step": 115000
1389
+ },
1390
+ {
1391
+ "epoch": 0.69,
1392
+ "learning_rate": 5e-05,
1393
+ "loss": 1.5949,
1394
+ "step": 115500
1395
+ },
1396
+ {
1397
+ "epoch": 0.7,
1398
+ "learning_rate": 5e-05,
1399
+ "loss": 1.5934,
1400
+ "step": 116000
1401
+ },
1402
+ {
1403
+ "epoch": 0.7,
1404
+ "learning_rate": 5e-05,
1405
+ "loss": 1.5981,
1406
+ "step": 116500
1407
+ },
1408
+ {
1409
+ "epoch": 0.7,
1410
+ "learning_rate": 5e-05,
1411
+ "loss": 1.5918,
1412
+ "step": 117000
1413
+ },
1414
+ {
1415
+ "epoch": 0.71,
1416
+ "learning_rate": 5e-05,
1417
+ "loss": 1.5938,
1418
+ "step": 117500
1419
+ },
1420
+ {
1421
+ "epoch": 0.71,
1422
+ "learning_rate": 5e-05,
1423
+ "loss": 1.5979,
1424
+ "step": 118000
1425
+ },
1426
+ {
1427
+ "epoch": 0.71,
1428
+ "learning_rate": 5e-05,
1429
+ "loss": 1.5936,
1430
+ "step": 118500
1431
+ },
1432
+ {
1433
+ "epoch": 0.71,
1434
+ "learning_rate": 5e-05,
1435
+ "loss": 1.5953,
1436
+ "step": 119000
1437
+ },
1438
+ {
1439
+ "epoch": 0.72,
1440
+ "learning_rate": 5e-05,
1441
+ "loss": 1.5914,
1442
+ "step": 119500
1443
+ },
1444
+ {
1445
+ "epoch": 0.72,
1446
+ "learning_rate": 5e-05,
1447
+ "loss": 1.5962,
1448
+ "step": 120000
1449
+ },
1450
+ {
1451
+ "epoch": 0.72,
1452
+ "learning_rate": 5e-05,
1453
+ "loss": 1.5931,
1454
+ "step": 120500
1455
+ },
1456
+ {
1457
+ "epoch": 0.73,
1458
+ "learning_rate": 5e-05,
1459
+ "loss": 1.5891,
1460
+ "step": 121000
1461
+ },
1462
+ {
1463
+ "epoch": 0.73,
1464
+ "learning_rate": 5e-05,
1465
+ "loss": 1.5919,
1466
+ "step": 121500
1467
+ },
1468
+ {
1469
+ "epoch": 0.73,
1470
+ "learning_rate": 5e-05,
1471
+ "loss": 1.5947,
1472
+ "step": 122000
1473
+ },
1474
+ {
1475
+ "epoch": 0.74,
1476
+ "learning_rate": 5e-05,
1477
+ "loss": 1.5943,
1478
+ "step": 122500
1479
+ },
1480
+ {
1481
+ "epoch": 0.74,
1482
+ "learning_rate": 5e-05,
1483
+ "loss": 1.5897,
1484
+ "step": 123000
1485
+ },
1486
+ {
1487
+ "epoch": 0.74,
1488
+ "learning_rate": 5e-05,
1489
+ "loss": 1.587,
1490
+ "step": 123500
1491
+ },
1492
+ {
1493
+ "epoch": 0.74,
1494
+ "learning_rate": 5e-05,
1495
+ "loss": 1.595,
1496
+ "step": 124000
1497
+ },
1498
+ {
1499
+ "epoch": 0.75,
1500
+ "learning_rate": 5e-05,
1501
+ "loss": 1.5921,
1502
+ "step": 124500
1503
+ },
1504
+ {
1505
+ "epoch": 0.75,
1506
+ "learning_rate": 5e-05,
1507
+ "loss": 1.5899,
1508
+ "step": 125000
1509
+ },
1510
+ {
1511
+ "epoch": 0.75,
1512
+ "learning_rate": 5e-05,
1513
+ "loss": 1.5918,
1514
+ "step": 125500
1515
+ },
1516
+ {
1517
+ "epoch": 0.76,
1518
+ "learning_rate": 5e-05,
1519
+ "loss": 1.5938,
1520
+ "step": 126000
1521
+ },
1522
+ {
1523
+ "epoch": 0.76,
1524
+ "learning_rate": 5e-05,
1525
+ "loss": 1.5911,
1526
+ "step": 126500
1527
+ },
1528
+ {
1529
+ "epoch": 0.76,
1530
+ "learning_rate": 5e-05,
1531
+ "loss": 1.5926,
1532
+ "step": 127000
1533
+ },
1534
+ {
1535
+ "epoch": 0.77,
1536
+ "learning_rate": 5e-05,
1537
+ "loss": 1.5932,
1538
+ "step": 127500
1539
+ },
1540
+ {
1541
+ "epoch": 0.77,
1542
+ "learning_rate": 5e-05,
1543
+ "loss": 1.5895,
1544
+ "step": 128000
1545
+ },
1546
+ {
1547
+ "epoch": 0.77,
1548
+ "learning_rate": 5e-05,
1549
+ "loss": 1.5891,
1550
+ "step": 128500
1551
+ },
1552
+ {
1553
+ "epoch": 0.77,
1554
+ "learning_rate": 5e-05,
1555
+ "loss": 1.5884,
1556
+ "step": 129000
1557
+ },
1558
+ {
1559
+ "epoch": 0.78,
1560
+ "learning_rate": 5e-05,
1561
+ "loss": 1.5875,
1562
+ "step": 129500
1563
+ },
1564
+ {
1565
+ "epoch": 0.78,
1566
+ "learning_rate": 5e-05,
1567
+ "loss": 1.5901,
1568
+ "step": 130000
1569
+ },
1570
+ {
1571
+ "epoch": 0.78,
1572
+ "learning_rate": 5e-05,
1573
+ "loss": 1.59,
1574
+ "step": 130500
1575
+ },
1576
+ {
1577
+ "epoch": 0.79,
1578
+ "learning_rate": 5e-05,
1579
+ "loss": 1.5892,
1580
+ "step": 131000
1581
+ },
1582
+ {
1583
+ "epoch": 0.79,
1584
+ "learning_rate": 5e-05,
1585
+ "loss": 1.5896,
1586
+ "step": 131500
1587
+ },
1588
+ {
1589
+ "epoch": 0.79,
1590
+ "learning_rate": 5e-05,
1591
+ "loss": 1.589,
1592
+ "step": 132000
1593
+ },
1594
+ {
1595
+ "epoch": 0.8,
1596
+ "learning_rate": 5e-05,
1597
+ "loss": 1.5864,
1598
+ "step": 132500
1599
+ },
1600
+ {
1601
+ "epoch": 0.8,
1602
+ "learning_rate": 5e-05,
1603
+ "loss": 1.5866,
1604
+ "step": 133000
1605
+ },
1606
+ {
1607
+ "epoch": 0.8,
1608
+ "learning_rate": 5e-05,
1609
+ "loss": 1.5919,
1610
+ "step": 133500
1611
+ },
1612
+ {
1613
+ "epoch": 0.8,
1614
+ "learning_rate": 5e-05,
1615
+ "loss": 1.5839,
1616
+ "step": 134000
1617
+ },
1618
+ {
1619
+ "epoch": 0.81,
1620
+ "learning_rate": 5e-05,
1621
+ "loss": 1.5828,
1622
+ "step": 134500
1623
+ },
1624
+ {
1625
+ "epoch": 0.81,
1626
+ "learning_rate": 5e-05,
1627
+ "loss": 1.5888,
1628
+ "step": 135000
1629
+ },
1630
+ {
1631
+ "epoch": 0.81,
1632
+ "learning_rate": 5e-05,
1633
+ "loss": 1.5881,
1634
+ "step": 135500
1635
+ },
1636
+ {
1637
+ "epoch": 0.82,
1638
+ "learning_rate": 5e-05,
1639
+ "loss": 1.5854,
1640
+ "step": 136000
1641
+ },
1642
+ {
1643
+ "epoch": 0.82,
1644
+ "learning_rate": 5e-05,
1645
+ "loss": 1.5887,
1646
+ "step": 136500
1647
+ },
1648
+ {
1649
+ "epoch": 0.82,
1650
+ "learning_rate": 5e-05,
1651
+ "loss": 1.5847,
1652
+ "step": 137000
1653
+ },
1654
+ {
1655
+ "epoch": 0.83,
1656
+ "learning_rate": 5e-05,
1657
+ "loss": 1.5833,
1658
+ "step": 137500
1659
+ },
1660
+ {
1661
+ "epoch": 0.83,
1662
+ "learning_rate": 5e-05,
1663
+ "loss": 1.586,
1664
+ "step": 138000
1665
+ },
1666
+ {
1667
+ "epoch": 0.83,
1668
+ "learning_rate": 5e-05,
1669
+ "loss": 1.5838,
1670
+ "step": 138500
1671
+ },
1672
+ {
1673
+ "epoch": 0.83,
1674
+ "learning_rate": 5e-05,
1675
+ "loss": 1.5843,
1676
+ "step": 139000
1677
+ },
1678
+ {
1679
+ "epoch": 0.84,
1680
+ "learning_rate": 5e-05,
1681
+ "loss": 1.5837,
1682
+ "step": 139500
1683
+ },
1684
+ {
1685
+ "epoch": 0.84,
1686
+ "learning_rate": 5e-05,
1687
+ "loss": 1.5864,
1688
+ "step": 140000
1689
+ },
1690
+ {
1691
+ "epoch": 0.84,
1692
+ "learning_rate": 5e-05,
1693
+ "loss": 1.5867,
1694
+ "step": 140500
1695
+ },
1696
+ {
1697
+ "epoch": 0.85,
1698
+ "learning_rate": 5e-05,
1699
+ "loss": 1.5821,
1700
+ "step": 141000
1701
+ },
1702
+ {
1703
+ "epoch": 0.85,
1704
+ "learning_rate": 5e-05,
1705
+ "loss": 1.5867,
1706
+ "step": 141500
1707
+ },
1708
+ {
1709
+ "epoch": 0.85,
1710
+ "learning_rate": 5e-05,
1711
+ "loss": 1.5846,
1712
+ "step": 142000
1713
+ },
1714
+ {
1715
+ "epoch": 0.86,
1716
+ "learning_rate": 5e-05,
1717
+ "loss": 1.5863,
1718
+ "step": 142500
1719
+ },
1720
+ {
1721
+ "epoch": 0.86,
1722
+ "learning_rate": 5e-05,
1723
+ "loss": 1.5865,
1724
+ "step": 143000
1725
+ },
1726
+ {
1727
+ "epoch": 0.86,
1728
+ "learning_rate": 5e-05,
1729
+ "loss": 1.5867,
1730
+ "step": 143500
1731
+ },
1732
+ {
1733
+ "epoch": 0.86,
1734
+ "learning_rate": 5e-05,
1735
+ "loss": 1.5827,
1736
+ "step": 144000
1737
+ },
1738
+ {
1739
+ "epoch": 0.87,
1740
+ "learning_rate": 5e-05,
1741
+ "loss": 1.5816,
1742
+ "step": 144500
1743
+ },
1744
+ {
1745
+ "epoch": 0.87,
1746
+ "learning_rate": 5e-05,
1747
+ "loss": 1.5798,
1748
+ "step": 145000
1749
+ },
1750
+ {
1751
+ "epoch": 0.87,
1752
+ "learning_rate": 5e-05,
1753
+ "loss": 1.579,
1754
+ "step": 145500
1755
+ },
1756
+ {
1757
+ "epoch": 0.88,
1758
+ "learning_rate": 5e-05,
1759
+ "loss": 1.5825,
1760
+ "step": 146000
1761
+ },
1762
+ {
1763
+ "epoch": 0.88,
1764
+ "learning_rate": 5e-05,
1765
+ "loss": 1.5811,
1766
+ "step": 146500
1767
+ },
1768
+ {
1769
+ "epoch": 0.88,
1770
+ "learning_rate": 5e-05,
1771
+ "loss": 1.5786,
1772
+ "step": 147000
1773
+ },
1774
+ {
1775
+ "epoch": 0.89,
1776
+ "learning_rate": 5e-05,
1777
+ "loss": 1.5813,
1778
+ "step": 147500
1779
+ },
1780
+ {
1781
+ "epoch": 0.89,
1782
+ "learning_rate": 5e-05,
1783
+ "loss": 1.5829,
1784
+ "step": 148000
1785
+ },
1786
+ {
1787
+ "epoch": 0.89,
1788
+ "learning_rate": 5e-05,
1789
+ "loss": 1.58,
1790
+ "step": 148500
1791
+ },
1792
+ {
1793
+ "epoch": 0.89,
1794
+ "learning_rate": 5e-05,
1795
+ "loss": 1.5797,
1796
+ "step": 149000
1797
+ },
1798
+ {
1799
+ "epoch": 0.9,
1800
+ "learning_rate": 5e-05,
1801
+ "loss": 1.579,
1802
+ "step": 149500
1803
+ },
1804
+ {
1805
+ "epoch": 0.9,
1806
+ "learning_rate": 5e-05,
1807
+ "loss": 1.5804,
1808
+ "step": 150000
1809
+ },
1810
+ {
1811
+ "epoch": 0.9,
1812
+ "learning_rate": 5e-05,
1813
+ "loss": 1.5787,
1814
+ "step": 150500
1815
+ },
1816
+ {
1817
+ "epoch": 0.91,
1818
+ "learning_rate": 5e-05,
1819
+ "loss": 1.5814,
1820
+ "step": 151000
1821
+ },
1822
+ {
1823
+ "epoch": 0.91,
1824
+ "learning_rate": 5e-05,
1825
+ "loss": 1.5796,
1826
+ "step": 151500
1827
+ },
1828
+ {
1829
+ "epoch": 0.91,
1830
+ "learning_rate": 5e-05,
1831
+ "loss": 1.5742,
1832
+ "step": 152000
1833
+ },
1834
+ {
1835
+ "epoch": 0.92,
1836
+ "learning_rate": 5e-05,
1837
+ "loss": 1.5762,
1838
+ "step": 152500
1839
+ },
1840
+ {
1841
+ "epoch": 0.92,
1842
+ "learning_rate": 5e-05,
1843
+ "loss": 1.5822,
1844
+ "step": 153000
1845
+ },
1846
+ {
1847
+ "epoch": 0.92,
1848
+ "learning_rate": 5e-05,
1849
+ "loss": 1.578,
1850
+ "step": 153500
1851
+ },
1852
+ {
1853
+ "epoch": 0.92,
1854
+ "learning_rate": 5e-05,
1855
+ "loss": 1.5816,
1856
+ "step": 154000
1857
+ },
1858
+ {
1859
+ "epoch": 0.93,
1860
+ "learning_rate": 5e-05,
1861
+ "loss": 1.5772,
1862
+ "step": 154500
1863
+ },
1864
+ {
1865
+ "epoch": 0.93,
1866
+ "learning_rate": 5e-05,
1867
+ "loss": 1.5786,
1868
+ "step": 155000
1869
+ },
1870
+ {
1871
+ "epoch": 0.93,
1872
+ "learning_rate": 5e-05,
1873
+ "loss": 1.5784,
1874
+ "step": 155500
1875
+ },
1876
+ {
1877
+ "epoch": 0.94,
1878
+ "learning_rate": 5e-05,
1879
+ "loss": 1.5784,
1880
+ "step": 156000
1881
+ },
1882
+ {
1883
+ "epoch": 0.94,
1884
+ "learning_rate": 5e-05,
1885
+ "loss": 1.5786,
1886
+ "step": 156500
1887
+ },
1888
+ {
1889
+ "epoch": 0.94,
1890
+ "learning_rate": 5e-05,
1891
+ "loss": 1.5769,
1892
+ "step": 157000
1893
+ },
1894
+ {
1895
+ "epoch": 0.95,
1896
+ "learning_rate": 5e-05,
1897
+ "loss": 1.577,
1898
+ "step": 157500
1899
+ },
1900
+ {
1901
+ "epoch": 0.95,
1902
+ "learning_rate": 5e-05,
1903
+ "loss": 1.5767,
1904
+ "step": 158000
1905
+ },
1906
+ {
1907
+ "epoch": 0.95,
1908
+ "learning_rate": 5e-05,
1909
+ "loss": 1.5769,
1910
+ "step": 158500
1911
+ },
1912
+ {
1913
+ "epoch": 0.95,
1914
+ "learning_rate": 5e-05,
1915
+ "loss": 1.576,
1916
+ "step": 159000
1917
+ },
1918
+ {
1919
+ "epoch": 0.96,
1920
+ "learning_rate": 5e-05,
1921
+ "loss": 1.5736,
1922
+ "step": 159500
1923
+ },
1924
+ {
1925
+ "epoch": 0.96,
1926
+ "learning_rate": 5e-05,
1927
+ "loss": 1.5746,
1928
+ "step": 160000
1929
+ },
1930
+ {
1931
+ "epoch": 0.96,
1932
+ "learning_rate": 5e-05,
1933
+ "loss": 1.5692,
1934
+ "step": 160500
1935
+ },
1936
+ {
1937
+ "epoch": 0.97,
1938
+ "learning_rate": 5e-05,
1939
+ "loss": 1.5742,
1940
+ "step": 161000
1941
+ },
1942
+ {
1943
+ "epoch": 0.97,
1944
+ "learning_rate": 5e-05,
1945
+ "loss": 1.5767,
1946
+ "step": 161500
1947
+ },
1948
+ {
1949
+ "epoch": 0.97,
1950
+ "learning_rate": 5e-05,
1951
+ "loss": 1.5728,
1952
+ "step": 162000
1953
+ },
1954
+ {
1955
+ "epoch": 0.98,
1956
+ "learning_rate": 5e-05,
1957
+ "loss": 1.5704,
1958
+ "step": 162500
1959
+ },
1960
+ {
1961
+ "epoch": 0.98,
1962
+ "learning_rate": 5e-05,
1963
+ "loss": 1.5698,
1964
+ "step": 163000
1965
+ },
1966
+ {
1967
+ "epoch": 0.98,
1968
+ "learning_rate": 5e-05,
1969
+ "loss": 1.5765,
1970
+ "step": 163500
1971
+ },
1972
+ {
1973
+ "epoch": 0.98,
1974
+ "learning_rate": 5e-05,
1975
+ "loss": 1.5712,
1976
+ "step": 164000
1977
+ },
1978
+ {
1979
+ "epoch": 0.99,
1980
+ "learning_rate": 5e-05,
1981
+ "loss": 1.5747,
1982
+ "step": 164500
1983
+ },
1984
+ {
1985
+ "epoch": 0.99,
1986
+ "learning_rate": 5e-05,
1987
+ "loss": 1.5784,
1988
+ "step": 165000
1989
+ },
1990
+ {
1991
+ "epoch": 0.99,
1992
+ "learning_rate": 5e-05,
1993
+ "loss": 1.575,
1994
+ "step": 165500
1995
+ },
1996
+ {
1997
+ "epoch": 1.0,
1998
+ "learning_rate": 5e-05,
1999
+ "loss": 1.5729,
2000
+ "step": 166000
2001
+ },
2002
+ {
2003
+ "epoch": 1.0,
2004
+ "learning_rate": 5e-05,
2005
+ "loss": 1.5715,
2006
+ "step": 166500
2007
+ },
2008
+ {
2009
+ "epoch": 1.0,
2010
+ "step": 166617,
2011
+ "total_flos": 9.903255447295492e+18,
2012
+ "train_loss": 0.8373472213664984,
2013
+ "train_runtime": 584567.9437,
2014
+ "train_samples_per_second": 18.242,
2015
+ "train_steps_per_second": 0.285
2016
+ }
2017
+ ],
2018
+ "max_steps": 166617,
2019
+ "num_train_epochs": 1,
2020
+ "total_flos": 9.903255447295492e+18,
2021
+ "trial_name": null,
2022
+ "trial_params": null
2023
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0bd0859c20cb0c686f75f5def82bed79e637fc08caf445aca43184aa66870b58
3
+ size 4987