abhinavp commited on
Commit
2fa20ab
1 Parent(s): a82e735

Training in progress, epoch 1, checkpoint

Browse files
checkpoint-152640/config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_remove_final_layer_norm": false,
3
+ "activation_function": "relu",
4
+ "architectures": [
5
+ "OPTForCausalLM"
6
+ ],
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 2,
9
+ "do_layer_norm_before": true,
10
+ "dropout": 0.1,
11
+ "enable_bias": true,
12
+ "eos_token_id": 2,
13
+ "ffn_dim": 768,
14
+ "hidden_size": 768,
15
+ "init_std": 0.02,
16
+ "layer_norm_elementwise_affine": true,
17
+ "layerdrop": 0.0,
18
+ "max_position_embeddings": 512,
19
+ "model_type": "opt",
20
+ "num_attention_heads": 8,
21
+ "num_hidden_layers": 8,
22
+ "pad_token_id": 1,
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.33.3",
25
+ "use_cache": true,
26
+ "vocab_size": 50002,
27
+ "word_embed_proj_dim": 768
28
+ }
checkpoint-152640/generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 1,
6
+ "transformers_version": "4.33.3"
7
+ }
checkpoint-152640/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58c2990bf3dc5379ff5150e2b5d7ccb185f06e8e5724b5f04e04d16a85c65a51
3
+ size 537476677
checkpoint-152640/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1c67128bdd21811fbf27a90bda221512320e17fee24e222b0ed312ba27a754a
3
+ size 268727709
checkpoint-152640/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c20d46edc8129ae36968126f0e50fbe328de0aade0122b51152db2033be203db
3
+ size 14575
checkpoint-152640/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:09eacd27f135edc084693a84552efff395e7611dc852885ea5300230e0db19cd
3
+ size 627
checkpoint-152640/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "<eos>",
3
+ "pad_token": "<pad>",
4
+ "unk_token": "<unk>"
5
+ }
checkpoint-152640/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-152640/tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "eos_token": "<eos>",
4
+ "max_length": null,
5
+ "model_max_length": 512,
6
+ "pad_to_multiple_of": null,
7
+ "pad_token": "<pad>",
8
+ "pad_token_type_id": 0,
9
+ "padding_side": "right",
10
+ "tokenizer_class": "PreTrainedTokenizerFast",
11
+ "unk_token": "<unk>"
12
+ }
checkpoint-152640/trainer_state.json ADDED
@@ -0,0 +1,1829 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 4.026838779449463,
3
+ "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-only/transformer/3/checkpoints/checkpoint-152640",
4
+ "epoch": 1.0250006060157382,
5
+ "eval_steps": 10,
6
+ "global_step": 152640,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 4.999998362119627e-05,
14
+ "loss": 10.9828,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.0,
19
+ "learning_rate": 4.999161405248948e-05,
20
+ "loss": 6.815,
21
+ "step": 512
22
+ },
23
+ {
24
+ "epoch": 0.0,
25
+ "learning_rate": 4.998322810497896e-05,
26
+ "loss": 6.1821,
27
+ "step": 1024
28
+ },
29
+ {
30
+ "epoch": 0.0,
31
+ "learning_rate": 4.997484215746844e-05,
32
+ "loss": 5.9729,
33
+ "step": 1536
34
+ },
35
+ {
36
+ "epoch": 0.0,
37
+ "learning_rate": 4.996645620995792e-05,
38
+ "loss": 5.8165,
39
+ "step": 2048
40
+ },
41
+ {
42
+ "epoch": 0.0,
43
+ "learning_rate": 4.99580702624474e-05,
44
+ "loss": 5.7082,
45
+ "step": 2560
46
+ },
47
+ {
48
+ "epoch": 0.0,
49
+ "learning_rate": 4.994968431493688e-05,
50
+ "loss": 5.6061,
51
+ "step": 3072
52
+ },
53
+ {
54
+ "epoch": 0.0,
55
+ "learning_rate": 4.994129836742636e-05,
56
+ "loss": 5.5386,
57
+ "step": 3584
58
+ },
59
+ {
60
+ "epoch": 0.0,
61
+ "learning_rate": 4.993291241991584e-05,
62
+ "loss": 5.4514,
63
+ "step": 4096
64
+ },
65
+ {
66
+ "epoch": 0.0,
67
+ "learning_rate": 4.992452647240532e-05,
68
+ "loss": 5.4102,
69
+ "step": 4608
70
+ },
71
+ {
72
+ "epoch": 0.0,
73
+ "learning_rate": 4.99161405248948e-05,
74
+ "loss": 5.3599,
75
+ "step": 5120
76
+ },
77
+ {
78
+ "epoch": 0.0,
79
+ "learning_rate": 4.990775457738428e-05,
80
+ "loss": 5.3198,
81
+ "step": 5632
82
+ },
83
+ {
84
+ "epoch": 0.0,
85
+ "learning_rate": 4.989938500867749e-05,
86
+ "loss": 5.2546,
87
+ "step": 6144
88
+ },
89
+ {
90
+ "epoch": 0.0,
91
+ "learning_rate": 4.989099906116697e-05,
92
+ "loss": 5.2159,
93
+ "step": 6656
94
+ },
95
+ {
96
+ "epoch": 0.0,
97
+ "learning_rate": 4.988261311365645e-05,
98
+ "loss": 5.183,
99
+ "step": 7168
100
+ },
101
+ {
102
+ "epoch": 0.0,
103
+ "learning_rate": 4.987422716614593e-05,
104
+ "loss": 5.1335,
105
+ "step": 7680
106
+ },
107
+ {
108
+ "epoch": 0.0,
109
+ "learning_rate": 4.986584121863541e-05,
110
+ "loss": 5.1172,
111
+ "step": 8192
112
+ },
113
+ {
114
+ "epoch": 0.0,
115
+ "learning_rate": 4.985745527112489e-05,
116
+ "loss": 5.0848,
117
+ "step": 8704
118
+ },
119
+ {
120
+ "epoch": 0.0,
121
+ "learning_rate": 4.984906932361437e-05,
122
+ "loss": 5.0452,
123
+ "step": 9216
124
+ },
125
+ {
126
+ "epoch": 0.0,
127
+ "learning_rate": 4.984068337610385e-05,
128
+ "loss": 5.023,
129
+ "step": 9728
130
+ },
131
+ {
132
+ "epoch": 0.0,
133
+ "learning_rate": 4.983229742859333e-05,
134
+ "loss": 5.0069,
135
+ "step": 10240
136
+ },
137
+ {
138
+ "epoch": 0.0,
139
+ "learning_rate": 4.9823927859886547e-05,
140
+ "loss": 4.9763,
141
+ "step": 10752
142
+ },
143
+ {
144
+ "epoch": 0.0,
145
+ "learning_rate": 4.9815541912376026e-05,
146
+ "loss": 4.9588,
147
+ "step": 11264
148
+ },
149
+ {
150
+ "epoch": 0.0,
151
+ "learning_rate": 4.9807155964865506e-05,
152
+ "loss": 4.926,
153
+ "step": 11776
154
+ },
155
+ {
156
+ "epoch": 0.0,
157
+ "learning_rate": 4.9798770017354986e-05,
158
+ "loss": 4.9188,
159
+ "step": 12288
160
+ },
161
+ {
162
+ "epoch": 0.0,
163
+ "learning_rate": 4.9790400448648195e-05,
164
+ "loss": 4.8964,
165
+ "step": 12800
166
+ },
167
+ {
168
+ "epoch": 0.0,
169
+ "learning_rate": 4.9782014501137675e-05,
170
+ "loss": 4.8732,
171
+ "step": 13312
172
+ },
173
+ {
174
+ "epoch": 0.0,
175
+ "learning_rate": 4.9773628553627155e-05,
176
+ "loss": 4.8532,
177
+ "step": 13824
178
+ },
179
+ {
180
+ "epoch": 0.0,
181
+ "learning_rate": 4.9765242606116635e-05,
182
+ "loss": 4.828,
183
+ "step": 14336
184
+ },
185
+ {
186
+ "epoch": 0.0,
187
+ "learning_rate": 4.9756873037409844e-05,
188
+ "loss": 4.8199,
189
+ "step": 14848
190
+ },
191
+ {
192
+ "epoch": 0.01,
193
+ "learning_rate": 4.9748487089899324e-05,
194
+ "loss": 4.8034,
195
+ "step": 15360
196
+ },
197
+ {
198
+ "epoch": 0.01,
199
+ "learning_rate": 4.9740117521192533e-05,
200
+ "loss": 4.7967,
201
+ "step": 15872
202
+ },
203
+ {
204
+ "epoch": 0.01,
205
+ "learning_rate": 4.9731731573682013e-05,
206
+ "loss": 4.7807,
207
+ "step": 16384
208
+ },
209
+ {
210
+ "epoch": 0.01,
211
+ "learning_rate": 4.97233456261715e-05,
212
+ "loss": 4.7672,
213
+ "step": 16896
214
+ },
215
+ {
216
+ "epoch": 0.01,
217
+ "learning_rate": 4.971495967866098e-05,
218
+ "loss": 4.7584,
219
+ "step": 17408
220
+ },
221
+ {
222
+ "epoch": 0.01,
223
+ "learning_rate": 4.970657373115046e-05,
224
+ "loss": 4.7461,
225
+ "step": 17920
226
+ },
227
+ {
228
+ "epoch": 0.01,
229
+ "learning_rate": 4.969818778363994e-05,
230
+ "loss": 4.7364,
231
+ "step": 18432
232
+ },
233
+ {
234
+ "epoch": 0.01,
235
+ "learning_rate": 4.968981821493315e-05,
236
+ "loss": 4.7037,
237
+ "step": 18944
238
+ },
239
+ {
240
+ "epoch": 0.01,
241
+ "learning_rate": 4.968143226742263e-05,
242
+ "loss": 4.7038,
243
+ "step": 19456
244
+ },
245
+ {
246
+ "epoch": 0.01,
247
+ "learning_rate": 4.967304631991211e-05,
248
+ "loss": 4.6797,
249
+ "step": 19968
250
+ },
251
+ {
252
+ "epoch": 0.01,
253
+ "learning_rate": 4.966466037240159e-05,
254
+ "loss": 4.6841,
255
+ "step": 20480
256
+ },
257
+ {
258
+ "epoch": 0.01,
259
+ "learning_rate": 4.965627442489107e-05,
260
+ "loss": 4.6644,
261
+ "step": 20992
262
+ },
263
+ {
264
+ "epoch": 0.01,
265
+ "learning_rate": 4.964788847738054e-05,
266
+ "loss": 4.6666,
267
+ "step": 21504
268
+ },
269
+ {
270
+ "epoch": 0.01,
271
+ "learning_rate": 4.963950252987002e-05,
272
+ "loss": 4.6455,
273
+ "step": 22016
274
+ },
275
+ {
276
+ "epoch": 0.01,
277
+ "learning_rate": 4.96311165823595e-05,
278
+ "loss": 4.6428,
279
+ "step": 22528
280
+ },
281
+ {
282
+ "epoch": 0.01,
283
+ "learning_rate": 4.962274701365272e-05,
284
+ "loss": 4.6342,
285
+ "step": 23040
286
+ },
287
+ {
288
+ "epoch": 0.01,
289
+ "learning_rate": 4.96143610661422e-05,
290
+ "loss": 4.6276,
291
+ "step": 23552
292
+ },
293
+ {
294
+ "epoch": 0.01,
295
+ "learning_rate": 4.960597511863168e-05,
296
+ "loss": 4.6153,
297
+ "step": 24064
298
+ },
299
+ {
300
+ "epoch": 0.01,
301
+ "learning_rate": 4.959758917112116e-05,
302
+ "loss": 4.5907,
303
+ "step": 24576
304
+ },
305
+ {
306
+ "epoch": 0.01,
307
+ "learning_rate": 4.9589219602414374e-05,
308
+ "loss": 4.5971,
309
+ "step": 25088
310
+ },
311
+ {
312
+ "epoch": 0.01,
313
+ "learning_rate": 4.958083365490385e-05,
314
+ "loss": 4.5978,
315
+ "step": 25600
316
+ },
317
+ {
318
+ "epoch": 0.01,
319
+ "learning_rate": 4.957244770739333e-05,
320
+ "loss": 4.5916,
321
+ "step": 26112
322
+ },
323
+ {
324
+ "epoch": 0.01,
325
+ "learning_rate": 4.956406175988281e-05,
326
+ "loss": 4.5719,
327
+ "step": 26624
328
+ },
329
+ {
330
+ "epoch": 0.01,
331
+ "learning_rate": 4.9555692191176016e-05,
332
+ "loss": 4.5592,
333
+ "step": 27136
334
+ },
335
+ {
336
+ "epoch": 0.01,
337
+ "learning_rate": 4.954732262246923e-05,
338
+ "loss": 4.5614,
339
+ "step": 27648
340
+ },
341
+ {
342
+ "epoch": 0.01,
343
+ "learning_rate": 4.953893667495871e-05,
344
+ "loss": 4.5481,
345
+ "step": 28160
346
+ },
347
+ {
348
+ "epoch": 0.01,
349
+ "learning_rate": 4.953055072744819e-05,
350
+ "loss": 4.563,
351
+ "step": 28672
352
+ },
353
+ {
354
+ "epoch": 0.01,
355
+ "learning_rate": 4.952216477993767e-05,
356
+ "loss": 4.5249,
357
+ "step": 29184
358
+ },
359
+ {
360
+ "epoch": 0.01,
361
+ "learning_rate": 4.951377883242715e-05,
362
+ "loss": 4.5405,
363
+ "step": 29696
364
+ },
365
+ {
366
+ "epoch": 0.01,
367
+ "learning_rate": 4.950539288491663e-05,
368
+ "loss": 4.5303,
369
+ "step": 30208
370
+ },
371
+ {
372
+ "epoch": 0.01,
373
+ "learning_rate": 4.949700693740611e-05,
374
+ "loss": 4.5116,
375
+ "step": 30720
376
+ },
377
+ {
378
+ "epoch": 0.01,
379
+ "learning_rate": 4.948862098989559e-05,
380
+ "loss": 4.5213,
381
+ "step": 31232
382
+ },
383
+ {
384
+ "epoch": 0.01,
385
+ "learning_rate": 4.94802514211888e-05,
386
+ "loss": 4.5053,
387
+ "step": 31744
388
+ },
389
+ {
390
+ "epoch": 0.01,
391
+ "learning_rate": 4.947186547367828e-05,
392
+ "loss": 4.4964,
393
+ "step": 32256
394
+ },
395
+ {
396
+ "epoch": 0.01,
397
+ "learning_rate": 4.946347952616776e-05,
398
+ "loss": 4.4904,
399
+ "step": 32768
400
+ },
401
+ {
402
+ "epoch": 0.01,
403
+ "learning_rate": 4.945509357865724e-05,
404
+ "loss": 4.5027,
405
+ "step": 33280
406
+ },
407
+ {
408
+ "epoch": 0.01,
409
+ "learning_rate": 4.944672400995045e-05,
410
+ "loss": 4.4811,
411
+ "step": 33792
412
+ },
413
+ {
414
+ "epoch": 0.01,
415
+ "learning_rate": 4.943833806243993e-05,
416
+ "loss": 4.4772,
417
+ "step": 34304
418
+ },
419
+ {
420
+ "epoch": 0.01,
421
+ "learning_rate": 4.942995211492941e-05,
422
+ "loss": 4.4696,
423
+ "step": 34816
424
+ },
425
+ {
426
+ "epoch": 0.01,
427
+ "learning_rate": 4.942156616741889e-05,
428
+ "loss": 4.4621,
429
+ "step": 35328
430
+ },
431
+ {
432
+ "epoch": 0.01,
433
+ "learning_rate": 4.9413196598712105e-05,
434
+ "loss": 4.4697,
435
+ "step": 35840
436
+ },
437
+ {
438
+ "epoch": 0.01,
439
+ "learning_rate": 4.940482703000532e-05,
440
+ "loss": 4.4767,
441
+ "step": 36352
442
+ },
443
+ {
444
+ "epoch": 0.01,
445
+ "learning_rate": 4.9396441082494794e-05,
446
+ "loss": 4.4497,
447
+ "step": 36864
448
+ },
449
+ {
450
+ "epoch": 0.01,
451
+ "learning_rate": 4.9388055134984274e-05,
452
+ "loss": 4.4607,
453
+ "step": 37376
454
+ },
455
+ {
456
+ "epoch": 0.01,
457
+ "learning_rate": 4.9379669187473754e-05,
458
+ "loss": 4.4656,
459
+ "step": 37888
460
+ },
461
+ {
462
+ "epoch": 0.01,
463
+ "learning_rate": 4.9371283239963234e-05,
464
+ "loss": 4.4472,
465
+ "step": 38400
466
+ },
467
+ {
468
+ "epoch": 0.01,
469
+ "learning_rate": 4.936291367125644e-05,
470
+ "loss": 4.4359,
471
+ "step": 38912
472
+ },
473
+ {
474
+ "epoch": 0.01,
475
+ "learning_rate": 4.935452772374592e-05,
476
+ "loss": 4.4374,
477
+ "step": 39424
478
+ },
479
+ {
480
+ "epoch": 0.01,
481
+ "learning_rate": 4.93461417762354e-05,
482
+ "loss": 4.4348,
483
+ "step": 39936
484
+ },
485
+ {
486
+ "epoch": 0.01,
487
+ "learning_rate": 4.933775582872488e-05,
488
+ "loss": 4.421,
489
+ "step": 40448
490
+ },
491
+ {
492
+ "epoch": 0.01,
493
+ "learning_rate": 4.932936988121436e-05,
494
+ "loss": 4.4193,
495
+ "step": 40960
496
+ },
497
+ {
498
+ "epoch": 0.01,
499
+ "learning_rate": 4.932098393370384e-05,
500
+ "loss": 4.4273,
501
+ "step": 41472
502
+ },
503
+ {
504
+ "epoch": 0.01,
505
+ "learning_rate": 4.931261436499706e-05,
506
+ "loss": 4.4282,
507
+ "step": 41984
508
+ },
509
+ {
510
+ "epoch": 0.01,
511
+ "learning_rate": 4.930422841748654e-05,
512
+ "loss": 4.4108,
513
+ "step": 42496
514
+ },
515
+ {
516
+ "epoch": 0.01,
517
+ "learning_rate": 4.929584246997602e-05,
518
+ "loss": 4.3884,
519
+ "step": 43008
520
+ },
521
+ {
522
+ "epoch": 0.01,
523
+ "learning_rate": 4.92874565224655e-05,
524
+ "loss": 4.4021,
525
+ "step": 43520
526
+ },
527
+ {
528
+ "epoch": 0.01,
529
+ "learning_rate": 4.927907057495498e-05,
530
+ "loss": 4.406,
531
+ "step": 44032
532
+ },
533
+ {
534
+ "epoch": 0.01,
535
+ "learning_rate": 4.927068462744446e-05,
536
+ "loss": 4.3941,
537
+ "step": 44544
538
+ },
539
+ {
540
+ "epoch": 0.01,
541
+ "learning_rate": 4.926229867993394e-05,
542
+ "loss": 4.3996,
543
+ "step": 45056
544
+ },
545
+ {
546
+ "epoch": 0.01,
547
+ "learning_rate": 4.925392911122715e-05,
548
+ "loss": 4.3806,
549
+ "step": 45568
550
+ },
551
+ {
552
+ "epoch": 0.02,
553
+ "learning_rate": 4.924554316371663e-05,
554
+ "loss": 4.3806,
555
+ "step": 46080
556
+ },
557
+ {
558
+ "epoch": 0.02,
559
+ "learning_rate": 4.923715721620611e-05,
560
+ "loss": 4.3754,
561
+ "step": 46592
562
+ },
563
+ {
564
+ "epoch": 0.02,
565
+ "learning_rate": 4.922878764749932e-05,
566
+ "loss": 4.3734,
567
+ "step": 47104
568
+ },
569
+ {
570
+ "epoch": 0.02,
571
+ "learning_rate": 4.92204016999888e-05,
572
+ "loss": 4.3795,
573
+ "step": 47616
574
+ },
575
+ {
576
+ "epoch": 0.02,
577
+ "learning_rate": 4.921201575247828e-05,
578
+ "loss": 4.3668,
579
+ "step": 48128
580
+ },
581
+ {
582
+ "epoch": 0.02,
583
+ "learning_rate": 4.920362980496776e-05,
584
+ "loss": 4.3828,
585
+ "step": 48640
586
+ },
587
+ {
588
+ "epoch": 0.02,
589
+ "learning_rate": 4.919524385745724e-05,
590
+ "loss": 4.3664,
591
+ "step": 49152
592
+ },
593
+ {
594
+ "epoch": 0.02,
595
+ "learning_rate": 4.918685790994672e-05,
596
+ "loss": 4.3492,
597
+ "step": 49664
598
+ },
599
+ {
600
+ "epoch": 0.02,
601
+ "learning_rate": 4.91784719624362e-05,
602
+ "loss": 4.3528,
603
+ "step": 50176
604
+ },
605
+ {
606
+ "epoch": 0.02,
607
+ "learning_rate": 4.9170086014925676e-05,
608
+ "loss": 4.3598,
609
+ "step": 50688
610
+ },
611
+ {
612
+ "epoch": 0.02,
613
+ "learning_rate": 4.916171644621889e-05,
614
+ "loss": 4.3504,
615
+ "step": 51200
616
+ },
617
+ {
618
+ "epoch": 0.02,
619
+ "learning_rate": 4.915333049870837e-05,
620
+ "loss": 4.3481,
621
+ "step": 51712
622
+ },
623
+ {
624
+ "epoch": 0.02,
625
+ "learning_rate": 4.914494455119785e-05,
626
+ "loss": 4.3471,
627
+ "step": 52224
628
+ },
629
+ {
630
+ "epoch": 0.02,
631
+ "learning_rate": 4.9136558603687325e-05,
632
+ "loss": 4.3356,
633
+ "step": 52736
634
+ },
635
+ {
636
+ "epoch": 0.02,
637
+ "learning_rate": 4.9128172656176805e-05,
638
+ "loss": 4.3382,
639
+ "step": 53248
640
+ },
641
+ {
642
+ "epoch": 0.02,
643
+ "learning_rate": 4.9119786708666285e-05,
644
+ "loss": 4.3182,
645
+ "step": 53760
646
+ },
647
+ {
648
+ "epoch": 0.02,
649
+ "learning_rate": 4.9111400761155765e-05,
650
+ "loss": 4.324,
651
+ "step": 54272
652
+ },
653
+ {
654
+ "epoch": 0.02,
655
+ "learning_rate": 4.910303119244898e-05,
656
+ "loss": 4.3292,
657
+ "step": 54784
658
+ },
659
+ {
660
+ "epoch": 0.02,
661
+ "learning_rate": 4.909464524493846e-05,
662
+ "loss": 4.3239,
663
+ "step": 55296
664
+ },
665
+ {
666
+ "epoch": 0.02,
667
+ "learning_rate": 4.908625929742794e-05,
668
+ "loss": 4.3165,
669
+ "step": 55808
670
+ },
671
+ {
672
+ "epoch": 0.02,
673
+ "learning_rate": 4.907787334991742e-05,
674
+ "loss": 4.3255,
675
+ "step": 56320
676
+ },
677
+ {
678
+ "epoch": 0.02,
679
+ "learning_rate": 4.90694874024069e-05,
680
+ "loss": 4.3005,
681
+ "step": 56832
682
+ },
683
+ {
684
+ "epoch": 0.02,
685
+ "learning_rate": 4.906110145489638e-05,
686
+ "loss": 4.3115,
687
+ "step": 57344
688
+ },
689
+ {
690
+ "epoch": 0.02,
691
+ "learning_rate": 4.905273188618959e-05,
692
+ "loss": 4.3071,
693
+ "step": 57856
694
+ },
695
+ {
696
+ "epoch": 0.02,
697
+ "learning_rate": 4.904434593867907e-05,
698
+ "loss": 4.3115,
699
+ "step": 58368
700
+ },
701
+ {
702
+ "epoch": 0.02,
703
+ "learning_rate": 4.903595999116855e-05,
704
+ "loss": 4.3189,
705
+ "step": 58880
706
+ },
707
+ {
708
+ "epoch": 0.02,
709
+ "learning_rate": 4.902757404365803e-05,
710
+ "loss": 4.3064,
711
+ "step": 59392
712
+ },
713
+ {
714
+ "epoch": 0.02,
715
+ "learning_rate": 4.901920447495124e-05,
716
+ "loss": 4.2984,
717
+ "step": 59904
718
+ },
719
+ {
720
+ "epoch": 0.02,
721
+ "learning_rate": 4.901081852744072e-05,
722
+ "loss": 4.2937,
723
+ "step": 60416
724
+ },
725
+ {
726
+ "epoch": 0.02,
727
+ "learning_rate": 4.90024325799302e-05,
728
+ "loss": 4.2998,
729
+ "step": 60928
730
+ },
731
+ {
732
+ "epoch": 0.02,
733
+ "learning_rate": 4.8994046632419686e-05,
734
+ "loss": 4.2947,
735
+ "step": 61440
736
+ },
737
+ {
738
+ "epoch": 0.02,
739
+ "learning_rate": 4.8985660684909166e-05,
740
+ "loss": 4.2915,
741
+ "step": 61952
742
+ },
743
+ {
744
+ "epoch": 0.02,
745
+ "learning_rate": 4.8977274737398646e-05,
746
+ "loss": 4.2846,
747
+ "step": 62464
748
+ },
749
+ {
750
+ "epoch": 0.02,
751
+ "learning_rate": 4.8968888789888125e-05,
752
+ "loss": 4.2819,
753
+ "step": 62976
754
+ },
755
+ {
756
+ "epoch": 0.02,
757
+ "learning_rate": 4.8960519221181335e-05,
758
+ "loss": 4.2823,
759
+ "step": 63488
760
+ },
761
+ {
762
+ "epoch": 0.02,
763
+ "learning_rate": 4.8952133273670815e-05,
764
+ "loss": 4.2714,
765
+ "step": 64000
766
+ },
767
+ {
768
+ "epoch": 0.02,
769
+ "learning_rate": 4.8943747326160294e-05,
770
+ "loss": 4.277,
771
+ "step": 64512
772
+ },
773
+ {
774
+ "epoch": 0.02,
775
+ "learning_rate": 4.8935361378649774e-05,
776
+ "loss": 4.2747,
777
+ "step": 65024
778
+ },
779
+ {
780
+ "epoch": 0.02,
781
+ "learning_rate": 4.8926991809942984e-05,
782
+ "loss": 4.271,
783
+ "step": 65536
784
+ },
785
+ {
786
+ "epoch": 0.02,
787
+ "learning_rate": 4.8918605862432463e-05,
788
+ "loss": 4.2606,
789
+ "step": 66048
790
+ },
791
+ {
792
+ "epoch": 0.02,
793
+ "learning_rate": 4.8910219914921943e-05,
794
+ "loss": 4.279,
795
+ "step": 66560
796
+ },
797
+ {
798
+ "epoch": 0.02,
799
+ "learning_rate": 4.8901833967411423e-05,
800
+ "loss": 4.2791,
801
+ "step": 67072
802
+ },
803
+ {
804
+ "epoch": 0.02,
805
+ "learning_rate": 4.88934480199009e-05,
806
+ "loss": 4.2664,
807
+ "step": 67584
808
+ },
809
+ {
810
+ "epoch": 0.02,
811
+ "learning_rate": 4.888507845119412e-05,
812
+ "loss": 4.2571,
813
+ "step": 68096
814
+ },
815
+ {
816
+ "epoch": 0.02,
817
+ "learning_rate": 4.88766925036836e-05,
818
+ "loss": 4.2637,
819
+ "step": 68608
820
+ },
821
+ {
822
+ "epoch": 0.02,
823
+ "learning_rate": 4.886830655617308e-05,
824
+ "loss": 4.2635,
825
+ "step": 69120
826
+ },
827
+ {
828
+ "epoch": 0.02,
829
+ "learning_rate": 4.885992060866256e-05,
830
+ "loss": 4.2636,
831
+ "step": 69632
832
+ },
833
+ {
834
+ "epoch": 0.02,
835
+ "learning_rate": 4.885153466115204e-05,
836
+ "loss": 4.2589,
837
+ "step": 70144
838
+ },
839
+ {
840
+ "epoch": 0.02,
841
+ "learning_rate": 4.884314871364151e-05,
842
+ "loss": 4.2573,
843
+ "step": 70656
844
+ },
845
+ {
846
+ "epoch": 0.02,
847
+ "learning_rate": 4.883477914493473e-05,
848
+ "loss": 4.2537,
849
+ "step": 71168
850
+ },
851
+ {
852
+ "epoch": 0.02,
853
+ "learning_rate": 4.882639319742421e-05,
854
+ "loss": 4.2512,
855
+ "step": 71680
856
+ },
857
+ {
858
+ "epoch": 0.02,
859
+ "learning_rate": 4.881800724991369e-05,
860
+ "loss": 4.2461,
861
+ "step": 72192
862
+ },
863
+ {
864
+ "epoch": 0.02,
865
+ "learning_rate": 4.880962130240316e-05,
866
+ "loss": 4.2477,
867
+ "step": 72704
868
+ },
869
+ {
870
+ "epoch": 0.02,
871
+ "learning_rate": 4.880123535489264e-05,
872
+ "loss": 4.239,
873
+ "step": 73216
874
+ },
875
+ {
876
+ "epoch": 0.02,
877
+ "learning_rate": 4.879286578618586e-05,
878
+ "loss": 4.2339,
879
+ "step": 73728
880
+ },
881
+ {
882
+ "epoch": 0.02,
883
+ "learning_rate": 4.878447983867534e-05,
884
+ "loss": 4.2479,
885
+ "step": 74240
886
+ },
887
+ {
888
+ "epoch": 0.02,
889
+ "learning_rate": 4.877609389116482e-05,
890
+ "loss": 4.2322,
891
+ "step": 74752
892
+ },
893
+ {
894
+ "epoch": 0.02,
895
+ "learning_rate": 4.87677079436543e-05,
896
+ "loss": 4.2427,
897
+ "step": 75264
898
+ },
899
+ {
900
+ "epoch": 0.02,
901
+ "learning_rate": 4.875932199614378e-05,
902
+ "loss": 4.2285,
903
+ "step": 75776
904
+ },
905
+ {
906
+ "epoch": 0.02,
907
+ "learning_rate": 4.875093604863326e-05,
908
+ "loss": 4.2223,
909
+ "step": 76288
910
+ },
911
+ {
912
+ "epoch": 0.03,
913
+ "eval_loss": 4.196377277374268,
914
+ "eval_runtime": 306.8923,
915
+ "eval_samples_per_second": 1243.404,
916
+ "eval_steps_per_second": 38.857,
917
+ "step": 76320
918
+ },
919
+ {
920
+ "epoch": 1.0,
921
+ "learning_rate": 4.874255010112274e-05,
922
+ "loss": 4.2163,
923
+ "step": 76800
924
+ },
925
+ {
926
+ "epoch": 1.0,
927
+ "learning_rate": 4.873416415361222e-05,
928
+ "loss": 4.2165,
929
+ "step": 77312
930
+ },
931
+ {
932
+ "epoch": 1.0,
933
+ "learning_rate": 4.8725794584905426e-05,
934
+ "loss": 4.2265,
935
+ "step": 77824
936
+ },
937
+ {
938
+ "epoch": 1.0,
939
+ "learning_rate": 4.8717408637394906e-05,
940
+ "loss": 4.2193,
941
+ "step": 78336
942
+ },
943
+ {
944
+ "epoch": 1.0,
945
+ "learning_rate": 4.8709022689884386e-05,
946
+ "loss": 4.2205,
947
+ "step": 78848
948
+ },
949
+ {
950
+ "epoch": 1.0,
951
+ "learning_rate": 4.8700636742373866e-05,
952
+ "loss": 4.21,
953
+ "step": 79360
954
+ },
955
+ {
956
+ "epoch": 1.0,
957
+ "learning_rate": 4.8692267173667075e-05,
958
+ "loss": 4.2079,
959
+ "step": 79872
960
+ },
961
+ {
962
+ "epoch": 1.0,
963
+ "learning_rate": 4.8683881226156555e-05,
964
+ "loss": 4.1989,
965
+ "step": 80384
966
+ },
967
+ {
968
+ "epoch": 1.0,
969
+ "learning_rate": 4.867549527864604e-05,
970
+ "loss": 4.2145,
971
+ "step": 80896
972
+ },
973
+ {
974
+ "epoch": 1.0,
975
+ "learning_rate": 4.866710933113552e-05,
976
+ "loss": 4.2058,
977
+ "step": 81408
978
+ },
979
+ {
980
+ "epoch": 1.0,
981
+ "learning_rate": 4.8658723383625e-05,
982
+ "loss": 4.2166,
983
+ "step": 81920
984
+ },
985
+ {
986
+ "epoch": 1.0,
987
+ "learning_rate": 4.865033743611448e-05,
988
+ "loss": 4.1978,
989
+ "step": 82432
990
+ },
991
+ {
992
+ "epoch": 1.0,
993
+ "learning_rate": 4.864196786740769e-05,
994
+ "loss": 4.1944,
995
+ "step": 82944
996
+ },
997
+ {
998
+ "epoch": 1.0,
999
+ "learning_rate": 4.863358191989717e-05,
1000
+ "loss": 4.1892,
1001
+ "step": 83456
1002
+ },
1003
+ {
1004
+ "epoch": 1.0,
1005
+ "learning_rate": 4.862519597238665e-05,
1006
+ "loss": 4.1842,
1007
+ "step": 83968
1008
+ },
1009
+ {
1010
+ "epoch": 1.0,
1011
+ "learning_rate": 4.861681002487613e-05,
1012
+ "loss": 4.1853,
1013
+ "step": 84480
1014
+ },
1015
+ {
1016
+ "epoch": 1.0,
1017
+ "learning_rate": 4.860842407736561e-05,
1018
+ "loss": 4.1941,
1019
+ "step": 84992
1020
+ },
1021
+ {
1022
+ "epoch": 1.0,
1023
+ "learning_rate": 4.860003812985509e-05,
1024
+ "loss": 4.1816,
1025
+ "step": 85504
1026
+ },
1027
+ {
1028
+ "epoch": 1.0,
1029
+ "learning_rate": 4.859165218234457e-05,
1030
+ "loss": 4.1797,
1031
+ "step": 86016
1032
+ },
1033
+ {
1034
+ "epoch": 1.0,
1035
+ "learning_rate": 4.858328261363778e-05,
1036
+ "loss": 4.1999,
1037
+ "step": 86528
1038
+ },
1039
+ {
1040
+ "epoch": 1.0,
1041
+ "learning_rate": 4.8574913044930995e-05,
1042
+ "loss": 4.1844,
1043
+ "step": 87040
1044
+ },
1045
+ {
1046
+ "epoch": 1.0,
1047
+ "learning_rate": 4.8566527097420475e-05,
1048
+ "loss": 4.1862,
1049
+ "step": 87552
1050
+ },
1051
+ {
1052
+ "epoch": 1.0,
1053
+ "learning_rate": 4.8558141149909955e-05,
1054
+ "loss": 4.1753,
1055
+ "step": 88064
1056
+ },
1057
+ {
1058
+ "epoch": 1.0,
1059
+ "learning_rate": 4.8549755202399435e-05,
1060
+ "loss": 4.1852,
1061
+ "step": 88576
1062
+ },
1063
+ {
1064
+ "epoch": 1.0,
1065
+ "learning_rate": 4.8541369254888915e-05,
1066
+ "loss": 4.1707,
1067
+ "step": 89088
1068
+ },
1069
+ {
1070
+ "epoch": 1.0,
1071
+ "learning_rate": 4.8532983307378395e-05,
1072
+ "loss": 4.172,
1073
+ "step": 89600
1074
+ },
1075
+ {
1076
+ "epoch": 1.0,
1077
+ "learning_rate": 4.8524597359867875e-05,
1078
+ "loss": 4.1704,
1079
+ "step": 90112
1080
+ },
1081
+ {
1082
+ "epoch": 1.0,
1083
+ "learning_rate": 4.851621141235735e-05,
1084
+ "loss": 4.1637,
1085
+ "step": 90624
1086
+ },
1087
+ {
1088
+ "epoch": 1.0,
1089
+ "learning_rate": 4.8507841843650564e-05,
1090
+ "loss": 4.1582,
1091
+ "step": 91136
1092
+ },
1093
+ {
1094
+ "epoch": 1.01,
1095
+ "learning_rate": 4.8499455896140044e-05,
1096
+ "loss": 4.1592,
1097
+ "step": 91648
1098
+ },
1099
+ {
1100
+ "epoch": 1.01,
1101
+ "learning_rate": 4.849110270623698e-05,
1102
+ "loss": 4.1676,
1103
+ "step": 92160
1104
+ },
1105
+ {
1106
+ "epoch": 1.01,
1107
+ "learning_rate": 4.848271675872646e-05,
1108
+ "loss": 4.1664,
1109
+ "step": 92672
1110
+ },
1111
+ {
1112
+ "epoch": 1.01,
1113
+ "learning_rate": 4.847433081121595e-05,
1114
+ "loss": 4.158,
1115
+ "step": 93184
1116
+ },
1117
+ {
1118
+ "epoch": 1.01,
1119
+ "learning_rate": 4.846594486370543e-05,
1120
+ "loss": 4.166,
1121
+ "step": 93696
1122
+ },
1123
+ {
1124
+ "epoch": 1.01,
1125
+ "learning_rate": 4.845755891619491e-05,
1126
+ "loss": 4.1614,
1127
+ "step": 94208
1128
+ },
1129
+ {
1130
+ "epoch": 1.01,
1131
+ "learning_rate": 4.844917296868439e-05,
1132
+ "loss": 4.1609,
1133
+ "step": 94720
1134
+ },
1135
+ {
1136
+ "epoch": 1.01,
1137
+ "learning_rate": 4.844078702117387e-05,
1138
+ "loss": 4.1469,
1139
+ "step": 95232
1140
+ },
1141
+ {
1142
+ "epoch": 1.01,
1143
+ "learning_rate": 4.843240107366335e-05,
1144
+ "loss": 4.1482,
1145
+ "step": 95744
1146
+ },
1147
+ {
1148
+ "epoch": 1.01,
1149
+ "learning_rate": 4.842401512615282e-05,
1150
+ "loss": 4.1377,
1151
+ "step": 96256
1152
+ },
1153
+ {
1154
+ "epoch": 1.01,
1155
+ "learning_rate": 4.841564555744604e-05,
1156
+ "loss": 4.1503,
1157
+ "step": 96768
1158
+ },
1159
+ {
1160
+ "epoch": 1.01,
1161
+ "learning_rate": 4.840725960993552e-05,
1162
+ "loss": 4.1356,
1163
+ "step": 97280
1164
+ },
1165
+ {
1166
+ "epoch": 1.01,
1167
+ "learning_rate": 4.8398873662425e-05,
1168
+ "loss": 4.1489,
1169
+ "step": 97792
1170
+ },
1171
+ {
1172
+ "epoch": 1.01,
1173
+ "learning_rate": 4.839048771491447e-05,
1174
+ "loss": 4.1375,
1175
+ "step": 98304
1176
+ },
1177
+ {
1178
+ "epoch": 1.01,
1179
+ "learning_rate": 4.838210176740395e-05,
1180
+ "loss": 4.1372,
1181
+ "step": 98816
1182
+ },
1183
+ {
1184
+ "epoch": 1.01,
1185
+ "learning_rate": 4.8373732198697167e-05,
1186
+ "loss": 4.1404,
1187
+ "step": 99328
1188
+ },
1189
+ {
1190
+ "epoch": 1.01,
1191
+ "learning_rate": 4.8365346251186646e-05,
1192
+ "loss": 4.1359,
1193
+ "step": 99840
1194
+ },
1195
+ {
1196
+ "epoch": 1.01,
1197
+ "learning_rate": 4.8356960303676126e-05,
1198
+ "loss": 4.1338,
1199
+ "step": 100352
1200
+ },
1201
+ {
1202
+ "epoch": 1.01,
1203
+ "learning_rate": 4.8348574356165606e-05,
1204
+ "loss": 4.1203,
1205
+ "step": 100864
1206
+ },
1207
+ {
1208
+ "epoch": 1.01,
1209
+ "learning_rate": 4.834020478745882e-05,
1210
+ "loss": 4.1251,
1211
+ "step": 101376
1212
+ },
1213
+ {
1214
+ "epoch": 1.01,
1215
+ "learning_rate": 4.8331818839948295e-05,
1216
+ "loss": 4.1366,
1217
+ "step": 101888
1218
+ },
1219
+ {
1220
+ "epoch": 1.01,
1221
+ "learning_rate": 4.8323432892437775e-05,
1222
+ "loss": 4.1329,
1223
+ "step": 102400
1224
+ },
1225
+ {
1226
+ "epoch": 1.01,
1227
+ "learning_rate": 4.8315046944927255e-05,
1228
+ "loss": 4.124,
1229
+ "step": 102912
1230
+ },
1231
+ {
1232
+ "epoch": 1.01,
1233
+ "learning_rate": 4.8306660997416735e-05,
1234
+ "loss": 4.1146,
1235
+ "step": 103424
1236
+ },
1237
+ {
1238
+ "epoch": 1.01,
1239
+ "learning_rate": 4.8298291428709944e-05,
1240
+ "loss": 4.1228,
1241
+ "step": 103936
1242
+ },
1243
+ {
1244
+ "epoch": 1.01,
1245
+ "learning_rate": 4.8289905481199424e-05,
1246
+ "loss": 4.1109,
1247
+ "step": 104448
1248
+ },
1249
+ {
1250
+ "epoch": 1.01,
1251
+ "learning_rate": 4.8281519533688904e-05,
1252
+ "loss": 4.1329,
1253
+ "step": 104960
1254
+ },
1255
+ {
1256
+ "epoch": 1.01,
1257
+ "learning_rate": 4.8273133586178384e-05,
1258
+ "loss": 4.1051,
1259
+ "step": 105472
1260
+ },
1261
+ {
1262
+ "epoch": 1.01,
1263
+ "learning_rate": 4.826474763866787e-05,
1264
+ "loss": 4.1251,
1265
+ "step": 105984
1266
+ },
1267
+ {
1268
+ "epoch": 1.01,
1269
+ "learning_rate": 4.825637806996108e-05,
1270
+ "loss": 4.1189,
1271
+ "step": 106496
1272
+ },
1273
+ {
1274
+ "epoch": 1.01,
1275
+ "learning_rate": 4.824799212245056e-05,
1276
+ "loss": 4.1047,
1277
+ "step": 107008
1278
+ },
1279
+ {
1280
+ "epoch": 1.01,
1281
+ "learning_rate": 4.823960617494004e-05,
1282
+ "loss": 4.1166,
1283
+ "step": 107520
1284
+ },
1285
+ {
1286
+ "epoch": 1.01,
1287
+ "learning_rate": 4.823122022742952e-05,
1288
+ "loss": 4.11,
1289
+ "step": 108032
1290
+ },
1291
+ {
1292
+ "epoch": 1.01,
1293
+ "learning_rate": 4.8222834279919e-05,
1294
+ "loss": 4.1003,
1295
+ "step": 108544
1296
+ },
1297
+ {
1298
+ "epoch": 1.01,
1299
+ "learning_rate": 4.821446471121221e-05,
1300
+ "loss": 4.101,
1301
+ "step": 109056
1302
+ },
1303
+ {
1304
+ "epoch": 1.01,
1305
+ "learning_rate": 4.820607876370169e-05,
1306
+ "loss": 4.1148,
1307
+ "step": 109568
1308
+ },
1309
+ {
1310
+ "epoch": 1.01,
1311
+ "learning_rate": 4.819769281619117e-05,
1312
+ "loss": 4.1036,
1313
+ "step": 110080
1314
+ },
1315
+ {
1316
+ "epoch": 1.01,
1317
+ "learning_rate": 4.818930686868065e-05,
1318
+ "loss": 4.1005,
1319
+ "step": 110592
1320
+ },
1321
+ {
1322
+ "epoch": 1.01,
1323
+ "learning_rate": 4.818093729997386e-05,
1324
+ "loss": 4.0983,
1325
+ "step": 111104
1326
+ },
1327
+ {
1328
+ "epoch": 1.01,
1329
+ "learning_rate": 4.817255135246334e-05,
1330
+ "loss": 4.0941,
1331
+ "step": 111616
1332
+ },
1333
+ {
1334
+ "epoch": 1.01,
1335
+ "learning_rate": 4.816416540495282e-05,
1336
+ "loss": 4.1028,
1337
+ "step": 112128
1338
+ },
1339
+ {
1340
+ "epoch": 1.01,
1341
+ "learning_rate": 4.8155779457442305e-05,
1342
+ "loss": 4.1173,
1343
+ "step": 112640
1344
+ },
1345
+ {
1346
+ "epoch": 1.01,
1347
+ "learning_rate": 4.8147393509931785e-05,
1348
+ "loss": 4.0947,
1349
+ "step": 113152
1350
+ },
1351
+ {
1352
+ "epoch": 1.01,
1353
+ "learning_rate": 4.8139023941224994e-05,
1354
+ "loss": 4.1104,
1355
+ "step": 113664
1356
+ },
1357
+ {
1358
+ "epoch": 1.01,
1359
+ "learning_rate": 4.8130637993714474e-05,
1360
+ "loss": 4.1093,
1361
+ "step": 114176
1362
+ },
1363
+ {
1364
+ "epoch": 1.01,
1365
+ "learning_rate": 4.8122252046203954e-05,
1366
+ "loss": 4.0994,
1367
+ "step": 114688
1368
+ },
1369
+ {
1370
+ "epoch": 1.01,
1371
+ "learning_rate": 4.8113866098693434e-05,
1372
+ "loss": 4.0894,
1373
+ "step": 115200
1374
+ },
1375
+ {
1376
+ "epoch": 1.01,
1377
+ "learning_rate": 4.810549652998664e-05,
1378
+ "loss": 4.1022,
1379
+ "step": 115712
1380
+ },
1381
+ {
1382
+ "epoch": 1.01,
1383
+ "learning_rate": 4.809711058247612e-05,
1384
+ "loss": 4.0984,
1385
+ "step": 116224
1386
+ },
1387
+ {
1388
+ "epoch": 1.01,
1389
+ "learning_rate": 4.80887246349656e-05,
1390
+ "loss": 4.0893,
1391
+ "step": 116736
1392
+ },
1393
+ {
1394
+ "epoch": 1.01,
1395
+ "learning_rate": 4.808033868745508e-05,
1396
+ "loss": 4.0905,
1397
+ "step": 117248
1398
+ },
1399
+ {
1400
+ "epoch": 1.01,
1401
+ "learning_rate": 4.807196911874829e-05,
1402
+ "loss": 4.0945,
1403
+ "step": 117760
1404
+ },
1405
+ {
1406
+ "epoch": 1.01,
1407
+ "learning_rate": 4.806358317123777e-05,
1408
+ "loss": 4.1003,
1409
+ "step": 118272
1410
+ },
1411
+ {
1412
+ "epoch": 1.01,
1413
+ "learning_rate": 4.805519722372726e-05,
1414
+ "loss": 4.0885,
1415
+ "step": 118784
1416
+ },
1417
+ {
1418
+ "epoch": 1.01,
1419
+ "learning_rate": 4.804681127621674e-05,
1420
+ "loss": 4.0734,
1421
+ "step": 119296
1422
+ },
1423
+ {
1424
+ "epoch": 1.01,
1425
+ "learning_rate": 4.803844170750995e-05,
1426
+ "loss": 4.0813,
1427
+ "step": 119808
1428
+ },
1429
+ {
1430
+ "epoch": 1.01,
1431
+ "learning_rate": 4.803005575999943e-05,
1432
+ "loss": 4.0892,
1433
+ "step": 120320
1434
+ },
1435
+ {
1436
+ "epoch": 1.01,
1437
+ "learning_rate": 4.802166981248891e-05,
1438
+ "loss": 4.0851,
1439
+ "step": 120832
1440
+ },
1441
+ {
1442
+ "epoch": 1.01,
1443
+ "learning_rate": 4.801328386497839e-05,
1444
+ "loss": 4.0895,
1445
+ "step": 121344
1446
+ },
1447
+ {
1448
+ "epoch": 1.01,
1449
+ "learning_rate": 4.8004914296271596e-05,
1450
+ "loss": 4.0757,
1451
+ "step": 121856
1452
+ },
1453
+ {
1454
+ "epoch": 1.02,
1455
+ "learning_rate": 4.7996528348761076e-05,
1456
+ "loss": 4.0717,
1457
+ "step": 122368
1458
+ },
1459
+ {
1460
+ "epoch": 1.02,
1461
+ "learning_rate": 4.7988142401250556e-05,
1462
+ "loss": 4.0725,
1463
+ "step": 122880
1464
+ },
1465
+ {
1466
+ "epoch": 1.02,
1467
+ "learning_rate": 4.7979756453740036e-05,
1468
+ "loss": 4.0767,
1469
+ "step": 123392
1470
+ },
1471
+ {
1472
+ "epoch": 1.02,
1473
+ "learning_rate": 4.7971386885033245e-05,
1474
+ "loss": 4.0792,
1475
+ "step": 123904
1476
+ },
1477
+ {
1478
+ "epoch": 1.02,
1479
+ "learning_rate": 4.7963000937522725e-05,
1480
+ "loss": 4.0746,
1481
+ "step": 124416
1482
+ },
1483
+ {
1484
+ "epoch": 1.02,
1485
+ "learning_rate": 4.795463136881594e-05,
1486
+ "loss": 4.0931,
1487
+ "step": 124928
1488
+ },
1489
+ {
1490
+ "epoch": 1.02,
1491
+ "learning_rate": 4.794624542130542e-05,
1492
+ "loss": 4.0785,
1493
+ "step": 125440
1494
+ },
1495
+ {
1496
+ "epoch": 1.02,
1497
+ "learning_rate": 4.79378594737949e-05,
1498
+ "loss": 4.0617,
1499
+ "step": 125952
1500
+ },
1501
+ {
1502
+ "epoch": 1.02,
1503
+ "learning_rate": 4.792947352628438e-05,
1504
+ "loss": 4.0713,
1505
+ "step": 126464
1506
+ },
1507
+ {
1508
+ "epoch": 1.02,
1509
+ "learning_rate": 4.792108757877386e-05,
1510
+ "loss": 4.0731,
1511
+ "step": 126976
1512
+ },
1513
+ {
1514
+ "epoch": 1.02,
1515
+ "learning_rate": 4.791270163126334e-05,
1516
+ "loss": 4.0726,
1517
+ "step": 127488
1518
+ },
1519
+ {
1520
+ "epoch": 1.02,
1521
+ "learning_rate": 4.790431568375282e-05,
1522
+ "loss": 4.0717,
1523
+ "step": 128000
1524
+ },
1525
+ {
1526
+ "epoch": 1.02,
1527
+ "learning_rate": 4.78959297362423e-05,
1528
+ "loss": 4.0686,
1529
+ "step": 128512
1530
+ },
1531
+ {
1532
+ "epoch": 1.02,
1533
+ "learning_rate": 4.788756016753551e-05,
1534
+ "loss": 4.0607,
1535
+ "step": 129024
1536
+ },
1537
+ {
1538
+ "epoch": 1.02,
1539
+ "learning_rate": 4.787917422002499e-05,
1540
+ "loss": 4.0708,
1541
+ "step": 129536
1542
+ },
1543
+ {
1544
+ "epoch": 1.02,
1545
+ "learning_rate": 4.787078827251447e-05,
1546
+ "loss": 4.0472,
1547
+ "step": 130048
1548
+ },
1549
+ {
1550
+ "epoch": 1.02,
1551
+ "learning_rate": 4.786240232500395e-05,
1552
+ "loss": 4.0587,
1553
+ "step": 130560
1554
+ },
1555
+ {
1556
+ "epoch": 1.02,
1557
+ "learning_rate": 4.7854032756297166e-05,
1558
+ "loss": 4.0586,
1559
+ "step": 131072
1560
+ },
1561
+ {
1562
+ "epoch": 1.02,
1563
+ "learning_rate": 4.7845646808786646e-05,
1564
+ "loss": 4.0626,
1565
+ "step": 131584
1566
+ },
1567
+ {
1568
+ "epoch": 1.02,
1569
+ "learning_rate": 4.7837260861276126e-05,
1570
+ "loss": 4.0525,
1571
+ "step": 132096
1572
+ },
1573
+ {
1574
+ "epoch": 1.02,
1575
+ "learning_rate": 4.7828874913765606e-05,
1576
+ "loss": 4.0614,
1577
+ "step": 132608
1578
+ },
1579
+ {
1580
+ "epoch": 1.02,
1581
+ "learning_rate": 4.782048896625508e-05,
1582
+ "loss": 4.0475,
1583
+ "step": 133120
1584
+ },
1585
+ {
1586
+ "epoch": 1.02,
1587
+ "learning_rate": 4.781210301874456e-05,
1588
+ "loss": 4.0531,
1589
+ "step": 133632
1590
+ },
1591
+ {
1592
+ "epoch": 1.02,
1593
+ "learning_rate": 4.7803733450037775e-05,
1594
+ "loss": 4.0525,
1595
+ "step": 134144
1596
+ },
1597
+ {
1598
+ "epoch": 1.02,
1599
+ "learning_rate": 4.779534750252725e-05,
1600
+ "loss": 4.0539,
1601
+ "step": 134656
1602
+ },
1603
+ {
1604
+ "epoch": 1.02,
1605
+ "learning_rate": 4.778696155501673e-05,
1606
+ "loss": 4.0702,
1607
+ "step": 135168
1608
+ },
1609
+ {
1610
+ "epoch": 1.02,
1611
+ "learning_rate": 4.777857560750621e-05,
1612
+ "loss": 4.0595,
1613
+ "step": 135680
1614
+ },
1615
+ {
1616
+ "epoch": 1.02,
1617
+ "learning_rate": 4.7770206038799423e-05,
1618
+ "loss": 4.0532,
1619
+ "step": 136192
1620
+ },
1621
+ {
1622
+ "epoch": 1.02,
1623
+ "learning_rate": 4.7761820091288903e-05,
1624
+ "loss": 4.0492,
1625
+ "step": 136704
1626
+ },
1627
+ {
1628
+ "epoch": 1.02,
1629
+ "learning_rate": 4.7753434143778383e-05,
1630
+ "loss": 4.0514,
1631
+ "step": 137216
1632
+ },
1633
+ {
1634
+ "epoch": 1.02,
1635
+ "learning_rate": 4.774504819626786e-05,
1636
+ "loss": 4.0508,
1637
+ "step": 137728
1638
+ },
1639
+ {
1640
+ "epoch": 1.02,
1641
+ "learning_rate": 4.773666224875734e-05,
1642
+ "loss": 4.0468,
1643
+ "step": 138240
1644
+ },
1645
+ {
1646
+ "epoch": 1.02,
1647
+ "learning_rate": 4.772827630124682e-05,
1648
+ "loss": 4.0447,
1649
+ "step": 138752
1650
+ },
1651
+ {
1652
+ "epoch": 1.02,
1653
+ "learning_rate": 4.77198903537363e-05,
1654
+ "loss": 4.043,
1655
+ "step": 139264
1656
+ },
1657
+ {
1658
+ "epoch": 1.02,
1659
+ "learning_rate": 4.771150440622578e-05,
1660
+ "loss": 4.0457,
1661
+ "step": 139776
1662
+ },
1663
+ {
1664
+ "epoch": 1.02,
1665
+ "learning_rate": 4.770313483751899e-05,
1666
+ "loss": 4.0343,
1667
+ "step": 140288
1668
+ },
1669
+ {
1670
+ "epoch": 1.02,
1671
+ "learning_rate": 4.769474889000847e-05,
1672
+ "loss": 4.0423,
1673
+ "step": 140800
1674
+ },
1675
+ {
1676
+ "epoch": 1.02,
1677
+ "learning_rate": 4.768636294249795e-05,
1678
+ "loss": 4.0405,
1679
+ "step": 141312
1680
+ },
1681
+ {
1682
+ "epoch": 1.02,
1683
+ "learning_rate": 4.767799337379116e-05,
1684
+ "loss": 4.0413,
1685
+ "step": 141824
1686
+ },
1687
+ {
1688
+ "epoch": 1.02,
1689
+ "learning_rate": 4.766960742628064e-05,
1690
+ "loss": 4.0288,
1691
+ "step": 142336
1692
+ },
1693
+ {
1694
+ "epoch": 1.02,
1695
+ "learning_rate": 4.766122147877012e-05,
1696
+ "loss": 4.0497,
1697
+ "step": 142848
1698
+ },
1699
+ {
1700
+ "epoch": 1.02,
1701
+ "learning_rate": 4.76528355312596e-05,
1702
+ "loss": 4.0547,
1703
+ "step": 143360
1704
+ },
1705
+ {
1706
+ "epoch": 1.02,
1707
+ "learning_rate": 4.764444958374909e-05,
1708
+ "loss": 4.0394,
1709
+ "step": 143872
1710
+ },
1711
+ {
1712
+ "epoch": 1.02,
1713
+ "learning_rate": 4.763606363623857e-05,
1714
+ "loss": 4.0319,
1715
+ "step": 144384
1716
+ },
1717
+ {
1718
+ "epoch": 1.02,
1719
+ "learning_rate": 4.762767768872805e-05,
1720
+ "loss": 4.0387,
1721
+ "step": 144896
1722
+ },
1723
+ {
1724
+ "epoch": 1.02,
1725
+ "learning_rate": 4.761929174121753e-05,
1726
+ "loss": 4.0416,
1727
+ "step": 145408
1728
+ },
1729
+ {
1730
+ "epoch": 1.02,
1731
+ "learning_rate": 4.761092217251074e-05,
1732
+ "loss": 4.0439,
1733
+ "step": 145920
1734
+ },
1735
+ {
1736
+ "epoch": 1.02,
1737
+ "learning_rate": 4.760253622500022e-05,
1738
+ "loss": 4.0386,
1739
+ "step": 146432
1740
+ },
1741
+ {
1742
+ "epoch": 1.02,
1743
+ "learning_rate": 4.75941502774897e-05,
1744
+ "loss": 4.0385,
1745
+ "step": 146944
1746
+ },
1747
+ {
1748
+ "epoch": 1.02,
1749
+ "learning_rate": 4.758576432997918e-05,
1750
+ "loss": 4.0334,
1751
+ "step": 147456
1752
+ },
1753
+ {
1754
+ "epoch": 1.02,
1755
+ "learning_rate": 4.7577394761272386e-05,
1756
+ "loss": 4.0388,
1757
+ "step": 147968
1758
+ },
1759
+ {
1760
+ "epoch": 1.02,
1761
+ "learning_rate": 4.7569008813761866e-05,
1762
+ "loss": 4.0283,
1763
+ "step": 148480
1764
+ },
1765
+ {
1766
+ "epoch": 1.02,
1767
+ "learning_rate": 4.7560622866251346e-05,
1768
+ "loss": 4.0369,
1769
+ "step": 148992
1770
+ },
1771
+ {
1772
+ "epoch": 1.02,
1773
+ "learning_rate": 4.7552236918740826e-05,
1774
+ "loss": 4.0321,
1775
+ "step": 149504
1776
+ },
1777
+ {
1778
+ "epoch": 1.02,
1779
+ "learning_rate": 4.754388372883777e-05,
1780
+ "loss": 4.0208,
1781
+ "step": 150016
1782
+ },
1783
+ {
1784
+ "epoch": 1.02,
1785
+ "learning_rate": 4.753549778132725e-05,
1786
+ "loss": 4.0379,
1787
+ "step": 150528
1788
+ },
1789
+ {
1790
+ "epoch": 1.02,
1791
+ "learning_rate": 4.752711183381673e-05,
1792
+ "loss": 4.0271,
1793
+ "step": 151040
1794
+ },
1795
+ {
1796
+ "epoch": 1.02,
1797
+ "learning_rate": 4.751872588630621e-05,
1798
+ "loss": 4.0329,
1799
+ "step": 151552
1800
+ },
1801
+ {
1802
+ "epoch": 1.02,
1803
+ "learning_rate": 4.751033993879569e-05,
1804
+ "loss": 4.0203,
1805
+ "step": 152064
1806
+ },
1807
+ {
1808
+ "epoch": 1.02,
1809
+ "learning_rate": 4.750195399128517e-05,
1810
+ "loss": 4.0204,
1811
+ "step": 152576
1812
+ },
1813
+ {
1814
+ "epoch": 1.03,
1815
+ "eval_loss": 4.026838779449463,
1816
+ "eval_runtime": 307.1642,
1817
+ "eval_samples_per_second": 1242.303,
1818
+ "eval_steps_per_second": 38.823,
1819
+ "step": 152640
1820
+ }
1821
+ ],
1822
+ "logging_steps": 512,
1823
+ "max_steps": 3052726,
1824
+ "num_train_epochs": 9223372036854775807,
1825
+ "save_steps": 10,
1826
+ "total_flos": 1.0644055367242752e+17,
1827
+ "trial_name": null,
1828
+ "trial_params": null
1829
+ }
checkpoint-152640/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf06229907a00c361bcc45ea31a88489cc0a27e0af480a8dc3b171af1c1e740a
3
+ size 4219