Ogamon commited on
Commit
c954183
·
verified ·
1 Parent(s): 583be7a

second commit

Browse files
all_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
- "epoch": 4.951768488745981,
3
- "num_input_tokens_seen": 5776160,
4
- "total_flos": 2.6009787341943603e+17,
5
- "train_loss": 1.2260490821160424,
6
- "train_runtime": 5290.2175,
7
- "train_samples_per_second": 18.789,
8
- "train_steps_per_second": 0.073
9
  }
 
1
  {
2
+ "predict_bleu-4": 86.3840886328125,
3
+ "predict_rouge-1": 94.296875,
4
+ "predict_rouge-2": 0.0,
5
+ "predict_rouge-l": 94.296875,
6
+ "predict_runtime": 17.1267,
7
+ "predict_samples_per_second": 149.124,
8
+ "predict_steps_per_second": 9.342
9
  }
generated_predictions.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
llamaboard_config.yaml CHANGED
@@ -1,5 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
1
  top.booster: auto
2
- top.checkpoint_path: null
3
  top.finetuning_type: full
4
  top.model_name: LLaMA3-8B-Chat
5
  top.quantization_bit: none
@@ -7,59 +18,3 @@ top.quantization_method: bitsandbytes
7
  top.rope_scaling: none
8
  top.template: llama3
9
  top.visual_inputs: false
10
- train.additional_target: ''
11
- train.badam_mode: layer
12
- train.badam_switch_interval: 50
13
- train.badam_switch_mode: ascending
14
- train.badam_update_ratio: 0.05
15
- train.batch_size: 4
16
- train.compute_type: bf16
17
- train.create_new_adapter: false
18
- train.cutoff_len: 1024
19
- train.dataset:
20
- - truth_train
21
- train.dataset_dir: data
22
- train.ds_offload: false
23
- train.ds_stage: '2'
24
- train.freeze_extra_modules: ''
25
- train.freeze_trainable_layers: 2
26
- train.freeze_trainable_modules: all
27
- train.galore_rank: 16
28
- train.galore_scale: 0.25
29
- train.galore_target: all
30
- train.galore_update_interval: 200
31
- train.gradient_accumulation_steps: 8
32
- train.learning_rate: 5e-6
33
- train.logging_steps: 1
34
- train.lora_alpha: 16
35
- train.lora_dropout: 0
36
- train.lora_rank: 8
37
- train.lora_target: ''
38
- train.loraplus_lr_ratio: 0
39
- train.lr_scheduler_type: cosine
40
- train.max_grad_norm: '1.0'
41
- train.max_samples: '100000'
42
- train.neat_packing: false
43
- train.neftune_alpha: 0
44
- train.num_train_epochs: '5.0'
45
- train.optim: adamw_torch
46
- train.packing: false
47
- train.ppo_score_norm: false
48
- train.ppo_whiten_rewards: false
49
- train.pref_beta: 0.1
50
- train.pref_ftx: 0
51
- train.pref_loss: sigmoid
52
- train.report_to: false
53
- train.resize_vocab: false
54
- train.reward_model: null
55
- train.save_steps: 1000
56
- train.shift_attn: false
57
- train.training_stage: Supervised Fine-Tuning
58
- train.use_badam: false
59
- train.use_dora: false
60
- train.use_galore: false
61
- train.use_llama_pro: false
62
- train.use_pissa: false
63
- train.use_rslora: false
64
- train.val_size: 0
65
- train.warmup_steps: 600
 
1
+ eval.batch_size: 2
2
+ eval.cutoff_len: 1024
3
+ eval.dataset:
4
+ - truth_dev
5
+ eval.dataset_dir: data
6
+ eval.max_new_tokens: 512
7
+ eval.max_samples: '100000'
8
+ eval.output_dir: eval_2024-07-11-10-49-45
9
+ eval.predict: true
10
+ eval.temperature: 0.95
11
+ eval.top_p: 0.7
12
  top.booster: auto
13
+ top.checkpoint_path: train_2024-07-11-10-49-45_inst_llama3
14
  top.finetuning_type: full
15
  top.model_name: LLaMA3-8B-Chat
16
  top.quantization_bit: none
 
18
  top.rope_scaling: none
19
  top.template: llama3
20
  top.visual_inputs: false
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
predict_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "predict_bleu-4": 86.3840886328125,
3
+ "predict_rouge-1": 94.296875,
4
+ "predict_rouge-2": 0.0,
5
+ "predict_rouge-l": 94.296875,
6
+ "predict_runtime": 17.1267,
7
+ "predict_samples_per_second": 149.124,
8
+ "predict_steps_per_second": 9.342
9
+ }
running_log.txt CHANGED
@@ -1,95 +1,79 @@
1
- [INFO|parser.py:325] 2024-07-11 11:02:30,806 >> Process rank: 0, device: cuda:0, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
2
 
3
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 6, device: cuda:6, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
4
 
5
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 2, device: cuda:2, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
6
 
7
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 7, device: cuda:7, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
8
 
9
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 4, device: cuda:4, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
10
 
11
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 5, device: cuda:5, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
12
 
13
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 3, device: cuda:3, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
14
 
15
- 07/11/2024 11:02:31 - INFO - llamafactory.hparams.parser - Process rank: 1, device: cuda:1, n_gpu: 1, distributed training: True, compute dtype: torch.bfloat16
16
 
17
- [INFO|tokenization_utils_base.py:2161] 2024-07-11 11:02:33,758 >> loading file tokenizer.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-Instruct/snapshots/e1945c40cd546c78e41f1151f4db032b271faeaa/tokenizer.json
18
 
19
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
20
 
21
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
22
 
23
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
24
 
25
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
26
 
27
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
28
 
29
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
30
 
31
- [INFO|tokenization_utils_base.py:2161] 2024-07-11 11:02:33,758 >> loading file added_tokens.json from cache at None
32
 
33
- [INFO|tokenization_utils_base.py:2161] 2024-07-11 11:02:33,758 >> loading file special_tokens_map.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-Instruct/snapshots/e1945c40cd546c78e41f1151f4db032b271faeaa/special_tokens_map.json
34
 
35
- [INFO|tokenization_utils_base.py:2161] 2024-07-11 11:02:33,758 >> loading file tokenizer_config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-Instruct/snapshots/e1945c40cd546c78e41f1151f4db032b271faeaa/tokenizer_config.json
36
 
37
- [WARNING|logging.py:313] 2024-07-11 11:02:34,060 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
38
 
39
- [INFO|template.py:270] 2024-07-11 11:02:34,061 >> Replace eos token: <|eot_id|>
40
 
41
- [INFO|template.py:372] 2024-07-11 11:02:34,061 >> Add pad token: <|eot_id|>
42
 
43
- [INFO|loader.py:50] 2024-07-11 11:02:34,061 >> Loading dataset train_output.json...
44
 
45
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
46
 
47
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
48
 
49
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
50
 
51
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
52
 
53
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
54
 
55
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
56
 
57
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
58
 
59
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
60
 
61
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
62
 
63
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
64
 
65
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
66
 
67
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
68
 
69
- 07/11/2024 11:02:34 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
70
 
71
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
72
 
73
- 07/11/2024 11:02:34 - INFO - llamafactory.data.template - Add pad token: <|eot_id|>
74
 
75
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
76
-
77
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
78
-
79
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
80
-
81
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
82
-
83
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
84
-
85
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
86
-
87
- 07/11/2024 11:02:35 - INFO - llamafactory.data.loader - Loading dataset train_output.json...
88
-
89
- [INFO|configuration_utils.py:733] 2024-07-11 11:02:39,782 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-Instruct/snapshots/e1945c40cd546c78e41f1151f4db032b271faeaa/config.json
90
-
91
- [INFO|configuration_utils.py:800] 2024-07-11 11:02:39,785 >> Model config LlamaConfig {
92
- "_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
93
  "architectures": [
94
  "LlamaForCausalLM"
95
  ],
@@ -114,40 +98,46 @@
114
  "tie_word_embeddings": false,
115
  "torch_dtype": "bfloat16",
116
  "transformers_version": "4.42.3",
117
- "use_cache": true,
118
  "vocab_size": 128256
119
  }
120
 
121
 
122
- [INFO|modeling_utils.py:3556] 2024-07-11 11:02:40,864 >> loading weights file model.safetensors from cache at /root/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-Instruct/snapshots/e1945c40cd546c78e41f1151f4db032b271faeaa/model.safetensors.index.json
 
 
123
 
124
- [INFO|modeling_utils.py:1531] 2024-07-11 11:05:05,483 >> Instantiating LlamaForCausalLM model under default dtype torch.bfloat16.
125
 
126
- [INFO|configuration_utils.py:1000] 2024-07-11 11:05:05,489 >> Generate config GenerationConfig {
127
  "bos_token_id": 128000,
128
  "eos_token_id": 128009
129
  }
130
 
131
 
132
- 07/11/2024 11:05:10 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
 
 
 
 
133
 
134
- 07/11/2024 11:05:10 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
135
 
136
- 07/11/2024 11:05:10 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
137
 
138
- 07/11/2024 11:05:10 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
139
 
140
- 07/11/2024 11:05:10 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
141
 
142
- [INFO|modeling_utils.py:4364] 2024-07-11 11:05:10,887 >> All model checkpoint weights were used when initializing LlamaForCausalLM.
143
 
144
 
145
- [INFO|modeling_utils.py:4372] 2024-07-11 11:05:10,887 >> All the weights of LlamaForCausalLM were initialized from the model checkpoint at meta-llama/Meta-Llama-3-8B-Instruct.
146
  If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
147
 
148
- [INFO|configuration_utils.py:955] 2024-07-11 11:05:11,059 >> loading configuration file generation_config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-Instruct/snapshots/e1945c40cd546c78e41f1151f4db032b271faeaa/generation_config.json
149
 
150
- [INFO|configuration_utils.py:1000] 2024-07-11 11:05:11,059 >> Generate config GenerationConfig {
151
  "bos_token_id": 128000,
152
  "do_sample": true,
153
  "eos_token_id": [
@@ -160,898 +150,60 @@ If your task is similar to the task the model of the checkpoint was trained on,
160
  }
161
 
162
 
163
- [INFO|checkpointing.py:103] 2024-07-11 11:05:11,066 >> Gradient checkpointing enabled.
164
-
165
- [INFO|attention.py:80] 2024-07-11 11:05:11,066 >> Using torch SDPA for faster training and inference.
166
-
167
- [INFO|adapter.py:302] 2024-07-11 11:05:11,066 >> Upcasting trainable params to float32.
168
-
169
- [INFO|adapter.py:48] 2024-07-11 11:05:11,066 >> Fine-tuning method: Full
170
-
171
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
172
-
173
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
174
-
175
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
176
-
177
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
178
-
179
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
180
-
181
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
182
-
183
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
184
-
185
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
186
-
187
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
188
-
189
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
190
-
191
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
192
-
193
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
194
-
195
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
196
-
197
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
198
-
199
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
200
-
201
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
202
-
203
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
204
-
205
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
206
-
207
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
208
-
209
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
210
-
211
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled.
212
-
213
- 07/11/2024 11:05:11 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
214
-
215
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32.
216
-
217
- 07/11/2024 11:05:11 - INFO - llamafactory.model.adapter - Fine-tuning method: Full
218
-
219
- [INFO|loader.py:196] 2024-07-11 11:05:11,161 >> trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
220
-
221
- 07/11/2024 11:05:11 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
222
-
223
- 07/11/2024 11:05:11 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
224
-
225
- 07/11/2024 11:05:11 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
226
-
227
- 07/11/2024 11:05:11 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
228
-
229
- 07/11/2024 11:05:11 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
230
-
231
- 07/11/2024 11:05:11 - INFO - llamafactory.model.loader - trainable params: 8,030,261,248 || all params: 8,030,261,248 || trainable%: 100.0000
232
-
233
- [INFO|trainer.py:642] 2024-07-11 11:05:11,167 >> Using auto half precision backend
234
-
235
- [INFO|trainer.py:2128] 2024-07-11 11:05:33,940 >> ***** Running training *****
236
-
237
- [INFO|trainer.py:2129] 2024-07-11 11:05:33,940 >> Num examples = 19,880
238
-
239
- [INFO|trainer.py:2130] 2024-07-11 11:05:33,940 >> Num Epochs = 5
240
-
241
- [INFO|trainer.py:2131] 2024-07-11 11:05:33,940 >> Instantaneous batch size per device = 4
242
-
243
- [INFO|trainer.py:2134] 2024-07-11 11:05:33,940 >> Total train batch size (w. parallel, distributed & accumulation) = 256
244
-
245
- [INFO|trainer.py:2135] 2024-07-11 11:05:33,940 >> Gradient Accumulation steps = 8
246
-
247
- [INFO|trainer.py:2136] 2024-07-11 11:05:33,940 >> Total optimization steps = 385
248
-
249
- [INFO|trainer.py:2137] 2024-07-11 11:05:33,941 >> Number of trainable parameters = 8,030,261,248
250
-
251
- [INFO|callbacks.py:310] 2024-07-11 11:05:52,173 >> {'loss': 13.9619, 'learning_rate': 8.3333e-09, 'epoch': 0.01, 'throughput': 830.24}
252
-
253
- [INFO|callbacks.py:310] 2024-07-11 11:06:05,761 >> {'loss': 13.7129, 'learning_rate': 1.6667e-08, 'epoch': 0.03, 'throughput': 949.38}
254
-
255
- [INFO|callbacks.py:310] 2024-07-11 11:06:19,304 >> {'loss': 13.8474, 'learning_rate': 2.5000e-08, 'epoch': 0.04, 'throughput': 1000.31}
256
-
257
- [INFO|callbacks.py:310] 2024-07-11 11:06:32,865 >> {'loss': 13.8844, 'learning_rate': 3.3333e-08, 'epoch': 0.05, 'throughput': 1017.74}
258
-
259
- [INFO|callbacks.py:310] 2024-07-11 11:06:46,377 >> {'loss': 14.1380, 'learning_rate': 4.1667e-08, 'epoch': 0.06, 'throughput': 1037.30}
260
-
261
- [INFO|callbacks.py:310] 2024-07-11 11:06:59,966 >> {'loss': 13.9077, 'learning_rate': 5.0000e-08, 'epoch': 0.08, 'throughput': 1055.71}
262
-
263
- [INFO|callbacks.py:310] 2024-07-11 11:07:13,513 >> {'loss': 13.8396, 'learning_rate': 5.8333e-08, 'epoch': 0.09, 'throughput': 1060.56}
264
-
265
- [INFO|callbacks.py:310] 2024-07-11 11:07:27,086 >> {'loss': 13.9828, 'learning_rate': 6.6667e-08, 'epoch': 0.10, 'throughput': 1059.47}
266
-
267
- [INFO|callbacks.py:310] 2024-07-11 11:07:40,694 >> {'loss': 14.0361, 'learning_rate': 7.5000e-08, 'epoch': 0.12, 'throughput': 1065.89}
268
-
269
- [INFO|callbacks.py:310] 2024-07-11 11:07:54,240 >> {'loss': 13.9392, 'learning_rate': 8.3333e-08, 'epoch': 0.13, 'throughput': 1069.27}
270
-
271
- [INFO|callbacks.py:310] 2024-07-11 11:08:07,884 >> {'loss': 14.0256, 'learning_rate': 9.1667e-08, 'epoch': 0.14, 'throughput': 1073.24}
272
-
273
- [INFO|callbacks.py:310] 2024-07-11 11:08:21,425 >> {'loss': 13.6693, 'learning_rate': 1.0000e-07, 'epoch': 0.15, 'throughput': 1070.54}
274
-
275
- [INFO|callbacks.py:310] 2024-07-11 11:08:34,965 >> {'loss': 13.9031, 'learning_rate': 1.0833e-07, 'epoch': 0.17, 'throughput': 1067.89}
276
-
277
- [INFO|callbacks.py:310] 2024-07-11 11:08:48,495 >> {'loss': 13.8575, 'learning_rate': 1.1667e-07, 'epoch': 0.18, 'throughput': 1073.07}
278
-
279
- [INFO|callbacks.py:310] 2024-07-11 11:09:02,114 >> {'loss': 13.8366, 'learning_rate': 1.2500e-07, 'epoch': 0.19, 'throughput': 1073.42}
280
-
281
- [INFO|callbacks.py:310] 2024-07-11 11:09:15,701 >> {'loss': 13.8705, 'learning_rate': 1.3333e-07, 'epoch': 0.21, 'throughput': 1076.20}
282
-
283
- [INFO|callbacks.py:310] 2024-07-11 11:09:29,305 >> {'loss': 13.1816, 'learning_rate': 1.4167e-07, 'epoch': 0.22, 'throughput': 1076.26}
284
-
285
- [INFO|callbacks.py:310] 2024-07-11 11:09:42,873 >> {'loss': 13.2292, 'learning_rate': 1.5000e-07, 'epoch': 0.23, 'throughput': 1078.02}
286
-
287
- [INFO|callbacks.py:310] 2024-07-11 11:09:56,380 >> {'loss': 13.4366, 'learning_rate': 1.5833e-07, 'epoch': 0.24, 'throughput': 1076.68}
288
-
289
- [INFO|callbacks.py:310] 2024-07-11 11:10:09,892 >> {'loss': 12.9904, 'learning_rate': 1.6667e-07, 'epoch': 0.26, 'throughput': 1076.72}
290
-
291
- [INFO|callbacks.py:310] 2024-07-11 11:10:23,491 >> {'loss': 12.8780, 'learning_rate': 1.7500e-07, 'epoch': 0.27, 'throughput': 1078.97}
292
-
293
- [INFO|callbacks.py:310] 2024-07-11 11:10:37,113 >> {'loss': 12.7794, 'learning_rate': 1.8333e-07, 'epoch': 0.28, 'throughput': 1081.79}
294
-
295
- [INFO|callbacks.py:310] 2024-07-11 11:10:50,680 >> {'loss': 11.3144, 'learning_rate': 1.9167e-07, 'epoch': 0.30, 'throughput': 1082.54}
296
-
297
- [INFO|callbacks.py:310] 2024-07-11 11:11:04,260 >> {'loss': 10.8531, 'learning_rate': 2.0000e-07, 'epoch': 0.31, 'throughput': 1086.37}
298
-
299
- [INFO|callbacks.py:310] 2024-07-11 11:11:17,809 >> {'loss': 10.7149, 'learning_rate': 2.0833e-07, 'epoch': 0.32, 'throughput': 1087.02}
300
-
301
- [INFO|callbacks.py:310] 2024-07-11 11:11:31,356 >> {'loss': 10.5802, 'learning_rate': 2.1667e-07, 'epoch': 0.33, 'throughput': 1089.07}
302
-
303
- [INFO|callbacks.py:310] 2024-07-11 11:11:44,856 >> {'loss': 10.3671, 'learning_rate': 2.2500e-07, 'epoch': 0.35, 'throughput': 1090.07}
304
-
305
- [INFO|callbacks.py:310] 2024-07-11 11:11:58,442 >> {'loss': 10.1751, 'learning_rate': 2.3333e-07, 'epoch': 0.36, 'throughput': 1090.50}
306
-
307
- [INFO|callbacks.py:310] 2024-07-11 11:12:11,969 >> {'loss': 9.7707, 'learning_rate': 2.4167e-07, 'epoch': 0.37, 'throughput': 1089.46}
308
-
309
- [INFO|callbacks.py:310] 2024-07-11 11:12:25,588 >> {'loss': 9.6489, 'learning_rate': 2.5000e-07, 'epoch': 0.39, 'throughput': 1090.10}
310
-
311
- [INFO|callbacks.py:310] 2024-07-11 11:12:39,159 >> {'loss': 8.6805, 'learning_rate': 2.5833e-07, 'epoch': 0.40, 'throughput': 1090.83}
312
-
313
- [INFO|callbacks.py:310] 2024-07-11 11:12:52,757 >> {'loss': 5.9207, 'learning_rate': 2.6667e-07, 'epoch': 0.41, 'throughput': 1091.16}
314
-
315
- [INFO|callbacks.py:310] 2024-07-11 11:13:06,329 >> {'loss': 5.7661, 'learning_rate': 2.7500e-07, 'epoch': 0.42, 'throughput': 1093.01}
316
-
317
- [INFO|callbacks.py:310] 2024-07-11 11:13:19,934 >> {'loss': 5.6168, 'learning_rate': 2.8333e-07, 'epoch': 0.44, 'throughput': 1094.61}
318
-
319
- [INFO|callbacks.py:310] 2024-07-11 11:13:33,491 >> {'loss': 5.3367, 'learning_rate': 2.9167e-07, 'epoch': 0.45, 'throughput': 1095.23}
320
-
321
- [INFO|callbacks.py:310] 2024-07-11 11:13:47,054 >> {'loss': 4.9751, 'learning_rate': 3.0000e-07, 'epoch': 0.46, 'throughput': 1094.37}
322
-
323
- [INFO|callbacks.py:310] 2024-07-11 11:14:00,636 >> {'loss': 4.7041, 'learning_rate': 3.0833e-07, 'epoch': 0.48, 'throughput': 1093.52}
324
-
325
- [INFO|callbacks.py:310] 2024-07-11 11:14:14,270 >> {'loss': 4.4631, 'learning_rate': 3.1667e-07, 'epoch': 0.49, 'throughput': 1093.46}
326
-
327
- [INFO|callbacks.py:310] 2024-07-11 11:14:27,823 >> {'loss': 4.1912, 'learning_rate': 3.2500e-07, 'epoch': 0.50, 'throughput': 1093.70}
328
-
329
- [INFO|callbacks.py:310] 2024-07-11 11:14:41,381 >> {'loss': 3.9146, 'learning_rate': 3.3333e-07, 'epoch': 0.51, 'throughput': 1093.91}
330
-
331
- [INFO|callbacks.py:310] 2024-07-11 11:14:55,032 >> {'loss': 3.0530, 'learning_rate': 3.4167e-07, 'epoch': 0.53, 'throughput': 1094.67}
332
-
333
- [INFO|callbacks.py:310] 2024-07-11 11:15:08,599 >> {'loss': 1.5544, 'learning_rate': 3.5000e-07, 'epoch': 0.54, 'throughput': 1095.05}
334
-
335
- [INFO|callbacks.py:310] 2024-07-11 11:15:22,200 >> {'loss': 1.0549, 'learning_rate': 3.5833e-07, 'epoch': 0.55, 'throughput': 1094.70}
336
-
337
- [INFO|callbacks.py:310] 2024-07-11 11:15:35,757 >> {'loss': 0.7110, 'learning_rate': 3.6667e-07, 'epoch': 0.57, 'throughput': 1095.09}
338
-
339
- [INFO|callbacks.py:310] 2024-07-11 11:15:49,369 >> {'loss': 0.5127, 'learning_rate': 3.7500e-07, 'epoch': 0.58, 'throughput': 1095.93}
340
-
341
- [INFO|callbacks.py:310] 2024-07-11 11:16:02,947 >> {'loss': 0.4143, 'learning_rate': 3.8333e-07, 'epoch': 0.59, 'throughput': 1095.72}
342
-
343
- [INFO|callbacks.py:310] 2024-07-11 11:16:16,516 >> {'loss': 0.4014, 'learning_rate': 3.9167e-07, 'epoch': 0.60, 'throughput': 1096.29}
344
-
345
- [INFO|callbacks.py:310] 2024-07-11 11:16:30,096 >> {'loss': 0.3840, 'learning_rate': 4.0000e-07, 'epoch': 0.62, 'throughput': 1096.57}
346
-
347
- [INFO|callbacks.py:310] 2024-07-11 11:16:43,682 >> {'loss': 0.3017, 'learning_rate': 4.0833e-07, 'epoch': 0.63, 'throughput': 1098.22}
348
-
349
- [INFO|callbacks.py:310] 2024-07-11 11:16:57,245 >> {'loss': 0.3031, 'learning_rate': 4.1667e-07, 'epoch': 0.64, 'throughput': 1099.04}
350
-
351
- [INFO|callbacks.py:310] 2024-07-11 11:17:10,757 >> {'loss': 0.2818, 'learning_rate': 4.2500e-07, 'epoch': 0.66, 'throughput': 1098.90}
352
-
353
- [INFO|callbacks.py:310] 2024-07-11 11:17:24,345 >> {'loss': 0.2881, 'learning_rate': 4.3333e-07, 'epoch': 0.67, 'throughput': 1098.82}
354
-
355
- [INFO|callbacks.py:310] 2024-07-11 11:17:37,932 >> {'loss': 0.2943, 'learning_rate': 4.4167e-07, 'epoch': 0.68, 'throughput': 1099.73}
356
-
357
- [INFO|callbacks.py:310] 2024-07-11 11:17:51,440 >> {'loss': 0.2781, 'learning_rate': 4.5000e-07, 'epoch': 0.69, 'throughput': 1098.98}
358
-
359
- [INFO|callbacks.py:310] 2024-07-11 11:18:04,974 >> {'loss': 0.2724, 'learning_rate': 4.5833e-07, 'epoch': 0.71, 'throughput': 1099.88}
360
-
361
- [INFO|callbacks.py:310] 2024-07-11 11:18:18,537 >> {'loss': 0.2520, 'learning_rate': 4.6667e-07, 'epoch': 0.72, 'throughput': 1099.29}
362
-
363
- [INFO|callbacks.py:310] 2024-07-11 11:18:32,042 >> {'loss': 0.2482, 'learning_rate': 4.7500e-07, 'epoch': 0.73, 'throughput': 1099.87}
364
-
365
- [INFO|callbacks.py:310] 2024-07-11 11:18:45,666 >> {'loss': 0.2212, 'learning_rate': 4.8333e-07, 'epoch': 0.75, 'throughput': 1100.79}
366
-
367
- [INFO|callbacks.py:310] 2024-07-11 11:18:59,238 >> {'loss': 0.2408, 'learning_rate': 4.9167e-07, 'epoch': 0.76, 'throughput': 1100.40}
368
-
369
- [INFO|callbacks.py:310] 2024-07-11 11:19:12,837 >> {'loss': 0.2133, 'learning_rate': 5.0000e-07, 'epoch': 0.77, 'throughput': 1100.76}
370
-
371
- [INFO|callbacks.py:310] 2024-07-11 11:19:26,415 >> {'loss': 0.2319, 'learning_rate': 5.0833e-07, 'epoch': 0.78, 'throughput': 1101.14}
372
-
373
- [INFO|callbacks.py:310] 2024-07-11 11:19:39,970 >> {'loss': 0.2703, 'learning_rate': 5.1667e-07, 'epoch': 0.80, 'throughput': 1100.64}
374
-
375
- [INFO|callbacks.py:310] 2024-07-11 11:19:53,514 >> {'loss': 0.2647, 'learning_rate': 5.2500e-07, 'epoch': 0.81, 'throughput': 1101.50}
376
-
377
- [INFO|callbacks.py:310] 2024-07-11 11:20:07,052 >> {'loss': 0.2288, 'learning_rate': 5.3333e-07, 'epoch': 0.82, 'throughput': 1100.95}
378
-
379
- [INFO|callbacks.py:310] 2024-07-11 11:20:20,590 >> {'loss': 0.2426, 'learning_rate': 5.4167e-07, 'epoch': 0.84, 'throughput': 1101.61}
380
-
381
- [INFO|callbacks.py:310] 2024-07-11 11:20:34,211 >> {'loss': 0.1936, 'learning_rate': 5.5000e-07, 'epoch': 0.85, 'throughput': 1102.11}
382
-
383
- [INFO|callbacks.py:310] 2024-07-11 11:20:47,778 >> {'loss': 0.1920, 'learning_rate': 5.5833e-07, 'epoch': 0.86, 'throughput': 1102.83}
384
-
385
- [INFO|callbacks.py:310] 2024-07-11 11:21:01,424 >> {'loss': 0.1956, 'learning_rate': 5.6667e-07, 'epoch': 0.87, 'throughput': 1102.10}
386
-
387
- [INFO|callbacks.py:310] 2024-07-11 11:21:15,063 >> {'loss': 0.1862, 'learning_rate': 5.7500e-07, 'epoch': 0.89, 'throughput': 1102.18}
388
-
389
- [INFO|callbacks.py:310] 2024-07-11 11:21:28,594 >> {'loss': 0.1950, 'learning_rate': 5.8333e-07, 'epoch': 0.90, 'throughput': 1101.34}
390
-
391
- [INFO|callbacks.py:310] 2024-07-11 11:21:42,130 >> {'loss': 0.1971, 'learning_rate': 5.9167e-07, 'epoch': 0.91, 'throughput': 1101.47}
392
-
393
- [INFO|callbacks.py:310] 2024-07-11 11:21:55,703 >> {'loss': 0.1683, 'learning_rate': 6.0000e-07, 'epoch': 0.93, 'throughput': 1101.79}
394
-
395
- [INFO|callbacks.py:310] 2024-07-11 11:22:09,285 >> {'loss': 0.2003, 'learning_rate': 6.0833e-07, 'epoch': 0.94, 'throughput': 1101.80}
396
-
397
- [INFO|callbacks.py:310] 2024-07-11 11:22:22,833 >> {'loss': 0.1543, 'learning_rate': 6.1667e-07, 'epoch': 0.95, 'throughput': 1101.69}
398
-
399
- [INFO|callbacks.py:310] 2024-07-11 11:22:36,430 >> {'loss': 0.1982, 'learning_rate': 6.2500e-07, 'epoch': 0.96, 'throughput': 1101.60}
400
-
401
- [INFO|callbacks.py:310] 2024-07-11 11:22:50,024 >> {'loss': 0.1545, 'learning_rate': 6.3333e-07, 'epoch': 0.98, 'throughput': 1101.78}
402
-
403
- [INFO|callbacks.py:310] 2024-07-11 11:23:03,557 >> {'loss': 0.1573, 'learning_rate': 6.4167e-07, 'epoch': 0.99, 'throughput': 1101.66}
404
-
405
- [INFO|callbacks.py:310] 2024-07-11 11:23:17,169 >> {'loss': 0.1788, 'learning_rate': 6.5000e-07, 'epoch': 1.00, 'throughput': 1102.22}
406
-
407
- [INFO|callbacks.py:310] 2024-07-11 11:23:30,798 >> {'loss': 0.1870, 'learning_rate': 6.5833e-07, 'epoch': 1.02, 'throughput': 1102.50}
408
-
409
- [INFO|callbacks.py:310] 2024-07-11 11:23:44,378 >> {'loss': 0.2009, 'learning_rate': 6.6667e-07, 'epoch': 1.03, 'throughput': 1101.97}
410
-
411
- [INFO|callbacks.py:310] 2024-07-11 11:23:57,975 >> {'loss': 0.2655, 'learning_rate': 6.7500e-07, 'epoch': 1.04, 'throughput': 1102.17}
412
-
413
- [INFO|callbacks.py:310] 2024-07-11 11:24:11,555 >> {'loss': 0.2041, 'learning_rate': 6.8333e-07, 'epoch': 1.05, 'throughput': 1102.46}
414
-
415
- [INFO|callbacks.py:310] 2024-07-11 11:24:25,136 >> {'loss': 0.1675, 'learning_rate': 6.9167e-07, 'epoch': 1.07, 'throughput': 1102.33}
416
-
417
- [INFO|callbacks.py:310] 2024-07-11 11:24:38,737 >> {'loss': 0.2097, 'learning_rate': 7.0000e-07, 'epoch': 1.08, 'throughput': 1102.93}
418
-
419
- [INFO|callbacks.py:310] 2024-07-11 11:24:52,408 >> {'loss': 0.2171, 'learning_rate': 7.0833e-07, 'epoch': 1.09, 'throughput': 1103.25}
420
-
421
- [INFO|callbacks.py:310] 2024-07-11 11:25:05,972 >> {'loss': 0.1702, 'learning_rate': 7.1667e-07, 'epoch': 1.11, 'throughput': 1103.45}
422
-
423
- [INFO|callbacks.py:310] 2024-07-11 11:25:19,464 >> {'loss': 0.1255, 'learning_rate': 7.2500e-07, 'epoch': 1.12, 'throughput': 1102.85}
424
-
425
- [INFO|callbacks.py:310] 2024-07-11 11:25:33,033 >> {'loss': 0.1826, 'learning_rate': 7.3333e-07, 'epoch': 1.13, 'throughput': 1103.48}
426
-
427
- [INFO|callbacks.py:310] 2024-07-11 11:25:46,607 >> {'loss': 0.2039, 'learning_rate': 7.4167e-07, 'epoch': 1.14, 'throughput': 1103.34}
428
-
429
- [INFO|callbacks.py:310] 2024-07-11 11:26:00,228 >> {'loss': 0.2337, 'learning_rate': 7.5000e-07, 'epoch': 1.16, 'throughput': 1103.79}
430
-
431
- [INFO|callbacks.py:310] 2024-07-11 11:26:13,832 >> {'loss': 0.1466, 'learning_rate': 7.5833e-07, 'epoch': 1.17, 'throughput': 1103.97}
432
-
433
- [INFO|callbacks.py:310] 2024-07-11 11:26:27,410 >> {'loss': 0.1295, 'learning_rate': 7.6667e-07, 'epoch': 1.18, 'throughput': 1104.29}
434
-
435
- [INFO|callbacks.py:310] 2024-07-11 11:26:40,924 >> {'loss': 0.1525, 'learning_rate': 7.7500e-07, 'epoch': 1.20, 'throughput': 1104.03}
436
-
437
- [INFO|callbacks.py:310] 2024-07-11 11:26:54,523 >> {'loss': 0.1735, 'learning_rate': 7.8333e-07, 'epoch': 1.21, 'throughput': 1104.42}
438
-
439
- [INFO|callbacks.py:310] 2024-07-11 11:27:08,160 >> {'loss': 0.1484, 'learning_rate': 7.9167e-07, 'epoch': 1.22, 'throughput': 1105.59}
440
-
441
- [INFO|callbacks.py:310] 2024-07-11 11:27:21,753 >> {'loss': 0.1517, 'learning_rate': 8.0000e-07, 'epoch': 1.23, 'throughput': 1105.48}
442
-
443
- [INFO|callbacks.py:310] 2024-07-11 11:27:35,370 >> {'loss': 0.1429, 'learning_rate': 8.0833e-07, 'epoch': 1.25, 'throughput': 1105.30}
444
-
445
- [INFO|callbacks.py:310] 2024-07-11 11:27:49,027 >> {'loss': 0.1510, 'learning_rate': 8.1667e-07, 'epoch': 1.26, 'throughput': 1105.57}
446
-
447
- [INFO|callbacks.py:310] 2024-07-11 11:28:02,582 >> {'loss': 0.1217, 'learning_rate': 8.2500e-07, 'epoch': 1.27, 'throughput': 1105.56}
448
-
449
- [INFO|callbacks.py:310] 2024-07-11 11:28:16,078 >> {'loss': 0.1366, 'learning_rate': 8.3333e-07, 'epoch': 1.29, 'throughput': 1104.97}
450
-
451
- [INFO|callbacks.py:310] 2024-07-11 11:28:29,637 >> {'loss': 0.1534, 'learning_rate': 8.4167e-07, 'epoch': 1.30, 'throughput': 1105.43}
452
-
453
- [INFO|callbacks.py:310] 2024-07-11 11:28:43,152 >> {'loss': 0.1410, 'learning_rate': 8.5000e-07, 'epoch': 1.31, 'throughput': 1105.25}
454
-
455
- [INFO|callbacks.py:310] 2024-07-11 11:28:56,697 >> {'loss': 0.1238, 'learning_rate': 8.5833e-07, 'epoch': 1.32, 'throughput': 1104.87}
456
-
457
- [INFO|callbacks.py:310] 2024-07-11 11:29:10,263 >> {'loss': 0.1241, 'learning_rate': 8.6667e-07, 'epoch': 1.34, 'throughput': 1104.63}
458
-
459
- [INFO|callbacks.py:310] 2024-07-11 11:29:23,827 >> {'loss': 0.1414, 'learning_rate': 8.7500e-07, 'epoch': 1.35, 'throughput': 1104.27}
460
-
461
- [INFO|callbacks.py:310] 2024-07-11 11:29:37,411 >> {'loss': 0.1296, 'learning_rate': 8.8333e-07, 'epoch': 1.36, 'throughput': 1104.25}
462
-
463
- [INFO|callbacks.py:310] 2024-07-11 11:29:50,947 >> {'loss': 0.1232, 'learning_rate': 8.9167e-07, 'epoch': 1.38, 'throughput': 1104.12}
464
-
465
- [INFO|callbacks.py:310] 2024-07-11 11:30:04,589 >> {'loss': 0.1625, 'learning_rate': 9.0000e-07, 'epoch': 1.39, 'throughput': 1104.23}
466
-
467
- [INFO|callbacks.py:310] 2024-07-11 11:30:18,108 >> {'loss': 0.1509, 'learning_rate': 9.0833e-07, 'epoch': 1.40, 'throughput': 1104.50}
468
-
469
- [INFO|callbacks.py:310] 2024-07-11 11:30:31,663 >> {'loss': 0.1416, 'learning_rate': 9.1667e-07, 'epoch': 1.41, 'throughput': 1104.70}
470
-
471
- [INFO|callbacks.py:310] 2024-07-11 11:30:45,219 >> {'loss': 0.1481, 'learning_rate': 9.2500e-07, 'epoch': 1.43, 'throughput': 1104.91}
472
-
473
- [INFO|callbacks.py:310] 2024-07-11 11:30:58,830 >> {'loss': 0.1303, 'learning_rate': 9.3333e-07, 'epoch': 1.44, 'throughput': 1104.87}
474
-
475
- [INFO|callbacks.py:310] 2024-07-11 11:31:12,462 >> {'loss': 0.1160, 'learning_rate': 9.4167e-07, 'epoch': 1.45, 'throughput': 1104.77}
476
-
477
- [INFO|callbacks.py:310] 2024-07-11 11:31:25,996 >> {'loss': 0.0981, 'learning_rate': 9.5000e-07, 'epoch': 1.47, 'throughput': 1104.48}
478
-
479
- [INFO|callbacks.py:310] 2024-07-11 11:31:39,542 >> {'loss': 0.1174, 'learning_rate': 9.5833e-07, 'epoch': 1.48, 'throughput': 1104.16}
480
-
481
- [INFO|callbacks.py:310] 2024-07-11 11:31:53,119 >> {'loss': 0.1458, 'learning_rate': 9.6667e-07, 'epoch': 1.49, 'throughput': 1103.74}
482
-
483
- [INFO|callbacks.py:310] 2024-07-11 11:32:06,699 >> {'loss': 0.0952, 'learning_rate': 9.7500e-07, 'epoch': 1.50, 'throughput': 1103.76}
484
-
485
- [INFO|callbacks.py:310] 2024-07-11 11:32:20,301 >> {'loss': 0.1233, 'learning_rate': 9.8333e-07, 'epoch': 1.52, 'throughput': 1103.43}
486
-
487
- [INFO|callbacks.py:310] 2024-07-11 11:32:33,923 >> {'loss': 0.1270, 'learning_rate': 9.9167e-07, 'epoch': 1.53, 'throughput': 1103.60}
488
-
489
- [INFO|callbacks.py:310] 2024-07-11 11:32:47,494 >> {'loss': 0.1121, 'learning_rate': 1.0000e-06, 'epoch': 1.54, 'throughput': 1103.79}
490
-
491
- [INFO|callbacks.py:310] 2024-07-11 11:33:01,024 >> {'loss': 0.1432, 'learning_rate': 1.0083e-06, 'epoch': 1.56, 'throughput': 1103.55}
492
-
493
- [INFO|callbacks.py:310] 2024-07-11 11:33:14,596 >> {'loss': 0.1446, 'learning_rate': 1.0167e-06, 'epoch': 1.57, 'throughput': 1103.53}
494
-
495
- [INFO|callbacks.py:310] 2024-07-11 11:33:28,145 >> {'loss': 0.1056, 'learning_rate': 1.0250e-06, 'epoch': 1.58, 'throughput': 1103.73}
496
-
497
- [INFO|callbacks.py:310] 2024-07-11 11:33:41,757 >> {'loss': 0.1193, 'learning_rate': 1.0333e-06, 'epoch': 1.59, 'throughput': 1103.80}
498
-
499
- [INFO|callbacks.py:310] 2024-07-11 11:33:55,372 >> {'loss': 0.1409, 'learning_rate': 1.0417e-06, 'epoch': 1.61, 'throughput': 1104.18}
500
-
501
- [INFO|callbacks.py:310] 2024-07-11 11:34:08,971 >> {'loss': 0.1154, 'learning_rate': 1.0500e-06, 'epoch': 1.62, 'throughput': 1104.44}
502
-
503
- [INFO|callbacks.py:310] 2024-07-11 11:34:22,537 >> {'loss': 0.1046, 'learning_rate': 1.0583e-06, 'epoch': 1.63, 'throughput': 1104.66}
504
-
505
- [INFO|callbacks.py:310] 2024-07-11 11:34:36,093 >> {'loss': 0.0900, 'learning_rate': 1.0667e-06, 'epoch': 1.65, 'throughput': 1104.90}
506
-
507
- [INFO|callbacks.py:310] 2024-07-11 11:34:49,618 >> {'loss': 0.0858, 'learning_rate': 1.0750e-06, 'epoch': 1.66, 'throughput': 1104.55}
508
-
509
- [INFO|callbacks.py:310] 2024-07-11 11:35:03,212 >> {'loss': 0.0782, 'learning_rate': 1.0833e-06, 'epoch': 1.67, 'throughput': 1104.82}
510
-
511
- [INFO|callbacks.py:310] 2024-07-11 11:35:16,738 >> {'loss': 0.1429, 'learning_rate': 1.0917e-06, 'epoch': 1.68, 'throughput': 1104.80}
512
-
513
- [INFO|callbacks.py:310] 2024-07-11 11:35:30,359 >> {'loss': 0.1121, 'learning_rate': 1.1000e-06, 'epoch': 1.70, 'throughput': 1104.72}
514
-
515
- [INFO|callbacks.py:310] 2024-07-11 11:35:43,915 >> {'loss': 0.0458, 'learning_rate': 1.1083e-06, 'epoch': 1.71, 'throughput': 1104.76}
516
-
517
- [INFO|callbacks.py:310] 2024-07-11 11:35:57,555 >> {'loss': 0.1217, 'learning_rate': 1.1167e-06, 'epoch': 1.72, 'throughput': 1104.43}
518
-
519
- [INFO|callbacks.py:310] 2024-07-11 11:36:11,110 >> {'loss': 0.1253, 'learning_rate': 1.1250e-06, 'epoch': 1.74, 'throughput': 1104.59}
520
-
521
- [INFO|callbacks.py:310] 2024-07-11 11:36:24,626 >> {'loss': 0.0770, 'learning_rate': 1.1333e-06, 'epoch': 1.75, 'throughput': 1104.56}
522
-
523
- [INFO|callbacks.py:310] 2024-07-11 11:36:38,205 >> {'loss': 0.0719, 'learning_rate': 1.1417e-06, 'epoch': 1.76, 'throughput': 1104.75}
524
-
525
- [INFO|callbacks.py:310] 2024-07-11 11:36:51,771 >> {'loss': 0.0916, 'learning_rate': 1.1500e-06, 'epoch': 1.77, 'throughput': 1104.65}
526
-
527
- [INFO|callbacks.py:310] 2024-07-11 11:37:05,307 >> {'loss': 0.0812, 'learning_rate': 1.1583e-06, 'epoch': 1.79, 'throughput': 1104.64}
528
-
529
- [INFO|callbacks.py:310] 2024-07-11 11:37:18,865 >> {'loss': 0.1176, 'learning_rate': 1.1667e-06, 'epoch': 1.80, 'throughput': 1104.64}
530
-
531
- [INFO|callbacks.py:310] 2024-07-11 11:37:32,443 >> {'loss': 0.0631, 'learning_rate': 1.1750e-06, 'epoch': 1.81, 'throughput': 1104.63}
532
-
533
- [INFO|callbacks.py:310] 2024-07-11 11:37:45,990 >> {'loss': 0.1137, 'learning_rate': 1.1833e-06, 'epoch': 1.83, 'throughput': 1104.44}
534
-
535
- [INFO|callbacks.py:310] 2024-07-11 11:37:59,512 >> {'loss': 0.0958, 'learning_rate': 1.1917e-06, 'epoch': 1.84, 'throughput': 1104.29}
536
-
537
- [INFO|callbacks.py:310] 2024-07-11 11:38:13,063 >> {'loss': 0.1343, 'learning_rate': 1.2000e-06, 'epoch': 1.85, 'throughput': 1104.23}
538
-
539
- [INFO|callbacks.py:310] 2024-07-11 11:38:26,593 >> {'loss': 0.1101, 'learning_rate': 1.2083e-06, 'epoch': 1.86, 'throughput': 1103.88}
540
-
541
- [INFO|callbacks.py:310] 2024-07-11 11:38:40,125 >> {'loss': 0.0914, 'learning_rate': 1.2167e-06, 'epoch': 1.88, 'throughput': 1103.82}
542
-
543
- [INFO|callbacks.py:310] 2024-07-11 11:38:53,738 >> {'loss': 0.1114, 'learning_rate': 1.2250e-06, 'epoch': 1.89, 'throughput': 1103.91}
544
-
545
- [INFO|callbacks.py:310] 2024-07-11 11:39:07,254 >> {'loss': 0.0830, 'learning_rate': 1.2333e-06, 'epoch': 1.90, 'throughput': 1103.76}
546
-
547
- [INFO|callbacks.py:310] 2024-07-11 11:39:20,835 >> {'loss': 0.1095, 'learning_rate': 1.2417e-06, 'epoch': 1.92, 'throughput': 1104.03}
548
-
549
- [INFO|callbacks.py:310] 2024-07-11 11:39:34,388 >> {'loss': 0.0662, 'learning_rate': 1.2500e-06, 'epoch': 1.93, 'throughput': 1104.06}
550
-
551
- [INFO|callbacks.py:310] 2024-07-11 11:39:47,946 >> {'loss': 0.0979, 'learning_rate': 1.2583e-06, 'epoch': 1.94, 'throughput': 1104.54}
552
-
553
- [INFO|callbacks.py:310] 2024-07-11 11:40:01,559 >> {'loss': 0.0847, 'learning_rate': 1.2667e-06, 'epoch': 1.95, 'throughput': 1104.58}
554
-
555
- [INFO|callbacks.py:310] 2024-07-11 11:40:15,082 >> {'loss': 0.0949, 'learning_rate': 1.2750e-06, 'epoch': 1.97, 'throughput': 1104.06}
556
-
557
- [INFO|callbacks.py:310] 2024-07-11 11:40:28,684 >> {'loss': 0.1206, 'learning_rate': 1.2833e-06, 'epoch': 1.98, 'throughput': 1103.69}
558
-
559
- [INFO|callbacks.py:310] 2024-07-11 11:40:42,245 >> {'loss': 0.1390, 'learning_rate': 1.2917e-06, 'epoch': 1.99, 'throughput': 1103.92}
560
-
561
- [INFO|callbacks.py:310] 2024-07-11 11:40:55,833 >> {'loss': 0.0700, 'learning_rate': 1.3000e-06, 'epoch': 2.01, 'throughput': 1103.71}
562
-
563
- [INFO|callbacks.py:310] 2024-07-11 11:41:09,423 >> {'loss': 0.0562, 'learning_rate': 1.3083e-06, 'epoch': 2.02, 'throughput': 1103.62}
564
-
565
- [INFO|callbacks.py:310] 2024-07-11 11:41:23,078 >> {'loss': 0.0456, 'learning_rate': 1.3167e-06, 'epoch': 2.03, 'throughput': 1103.51}
566
-
567
- [INFO|callbacks.py:310] 2024-07-11 11:41:36,635 >> {'loss': 0.0582, 'learning_rate': 1.3250e-06, 'epoch': 2.05, 'throughput': 1103.66}
568
-
569
- [INFO|callbacks.py:310] 2024-07-11 11:41:50,227 >> {'loss': 0.0452, 'learning_rate': 1.3333e-06, 'epoch': 2.06, 'throughput': 1103.37}
570
-
571
- [INFO|callbacks.py:310] 2024-07-11 11:42:03,798 >> {'loss': 0.0553, 'learning_rate': 1.3417e-06, 'epoch': 2.07, 'throughput': 1103.25}
572
-
573
- [INFO|callbacks.py:310] 2024-07-11 11:42:17,344 >> {'loss': 0.1108, 'learning_rate': 1.3500e-06, 'epoch': 2.08, 'throughput': 1103.01}
574
-
575
- [INFO|callbacks.py:310] 2024-07-11 11:42:30,912 >> {'loss': 0.0791, 'learning_rate': 1.3583e-06, 'epoch': 2.10, 'throughput': 1102.88}
576
-
577
- [INFO|callbacks.py:310] 2024-07-11 11:42:44,458 >> {'loss': 0.0637, 'learning_rate': 1.3667e-06, 'epoch': 2.11, 'throughput': 1102.77}
578
-
579
- [INFO|callbacks.py:310] 2024-07-11 11:42:57,976 >> {'loss': 0.0404, 'learning_rate': 1.3750e-06, 'epoch': 2.12, 'throughput': 1102.66}
580
-
581
- [INFO|callbacks.py:310] 2024-07-11 11:43:11,545 >> {'loss': 0.0372, 'learning_rate': 1.3833e-06, 'epoch': 2.14, 'throughput': 1102.62}
582
-
583
- [INFO|callbacks.py:310] 2024-07-11 11:43:25,078 >> {'loss': 0.0450, 'learning_rate': 1.3917e-06, 'epoch': 2.15, 'throughput': 1102.43}
584
-
585
- [INFO|callbacks.py:310] 2024-07-11 11:43:38,684 >> {'loss': 0.0964, 'learning_rate': 1.4000e-06, 'epoch': 2.16, 'throughput': 1102.42}
586
-
587
- [INFO|callbacks.py:310] 2024-07-11 11:43:52,267 >> {'loss': 0.0543, 'learning_rate': 1.4083e-06, 'epoch': 2.17, 'throughput': 1102.35}
588
-
589
- [INFO|callbacks.py:310] 2024-07-11 11:44:05,854 >> {'loss': 0.0710, 'learning_rate': 1.4167e-06, 'epoch': 2.19, 'throughput': 1102.39}
590
-
591
- [INFO|callbacks.py:310] 2024-07-11 11:44:19,373 >> {'loss': 0.0285, 'learning_rate': 1.4250e-06, 'epoch': 2.20, 'throughput': 1102.26}
592
-
593
- [INFO|callbacks.py:310] 2024-07-11 11:44:32,944 >> {'loss': 0.0399, 'learning_rate': 1.4333e-06, 'epoch': 2.21, 'throughput': 1102.32}
594
-
595
- [INFO|callbacks.py:310] 2024-07-11 11:44:46,571 >> {'loss': 0.0670, 'learning_rate': 1.4417e-06, 'epoch': 2.23, 'throughput': 1102.74}
596
-
597
- [INFO|callbacks.py:310] 2024-07-11 11:45:00,151 >> {'loss': 0.0436, 'learning_rate': 1.4500e-06, 'epoch': 2.24, 'throughput': 1102.78}
598
-
599
- [INFO|callbacks.py:310] 2024-07-11 11:45:13,718 >> {'loss': 0.0522, 'learning_rate': 1.4583e-06, 'epoch': 2.25, 'throughput': 1102.75}
600
-
601
- [INFO|callbacks.py:310] 2024-07-11 11:45:27,246 >> {'loss': 0.0521, 'learning_rate': 1.4667e-06, 'epoch': 2.26, 'throughput': 1102.78}
602
-
603
- [INFO|callbacks.py:310] 2024-07-11 11:45:40,843 >> {'loss': 0.0446, 'learning_rate': 1.4750e-06, 'epoch': 2.28, 'throughput': 1102.60}
604
-
605
- [INFO|callbacks.py:310] 2024-07-11 11:45:54,432 >> {'loss': 0.0378, 'learning_rate': 1.4833e-06, 'epoch': 2.29, 'throughput': 1102.76}
606
-
607
- [INFO|callbacks.py:310] 2024-07-11 11:46:08,019 >> {'loss': 0.0387, 'learning_rate': 1.4917e-06, 'epoch': 2.30, 'throughput': 1102.37}
608
-
609
- [INFO|callbacks.py:310] 2024-07-11 11:46:21,587 >> {'loss': 0.0360, 'learning_rate': 1.5000e-06, 'epoch': 2.32, 'throughput': 1102.33}
610
-
611
- [INFO|callbacks.py:310] 2024-07-11 11:46:35,138 >> {'loss': 0.0765, 'learning_rate': 1.5083e-06, 'epoch': 2.33, 'throughput': 1102.23}
612
-
613
- [INFO|callbacks.py:310] 2024-07-11 11:46:48,709 >> {'loss': 0.0884, 'learning_rate': 1.5167e-06, 'epoch': 2.34, 'throughput': 1102.34}
614
-
615
- [INFO|callbacks.py:310] 2024-07-11 11:47:02,246 >> {'loss': 0.0801, 'learning_rate': 1.5250e-06, 'epoch': 2.35, 'throughput': 1102.13}
616
-
617
- [INFO|callbacks.py:310] 2024-07-11 11:47:15,861 >> {'loss': 0.0276, 'learning_rate': 1.5333e-06, 'epoch': 2.37, 'throughput': 1102.42}
618
-
619
- [INFO|callbacks.py:310] 2024-07-11 11:47:29,458 >> {'loss': 0.0778, 'learning_rate': 1.5417e-06, 'epoch': 2.38, 'throughput': 1102.27}
620
-
621
- [INFO|callbacks.py:310] 2024-07-11 11:47:42,996 >> {'loss': 0.0726, 'learning_rate': 1.5500e-06, 'epoch': 2.39, 'throughput': 1102.38}
622
-
623
- [INFO|callbacks.py:310] 2024-07-11 11:47:56,560 >> {'loss': 0.1381, 'learning_rate': 1.5583e-06, 'epoch': 2.41, 'throughput': 1102.23}
624
-
625
- [INFO|callbacks.py:310] 2024-07-11 11:48:10,123 >> {'loss': 0.0408, 'learning_rate': 1.5667e-06, 'epoch': 2.42, 'throughput': 1102.37}
626
-
627
- [INFO|callbacks.py:310] 2024-07-11 11:48:23,689 >> {'loss': 0.1066, 'learning_rate': 1.5750e-06, 'epoch': 2.43, 'throughput': 1102.60}
628
-
629
- [INFO|callbacks.py:310] 2024-07-11 11:48:37,250 >> {'loss': 0.0686, 'learning_rate': 1.5833e-06, 'epoch': 2.44, 'throughput': 1102.78}
630
-
631
- [INFO|callbacks.py:310] 2024-07-11 11:48:50,791 >> {'loss': 0.0428, 'learning_rate': 1.5917e-06, 'epoch': 2.46, 'throughput': 1102.86}
632
-
633
- [INFO|callbacks.py:310] 2024-07-11 11:49:04,395 >> {'loss': 0.0387, 'learning_rate': 1.6000e-06, 'epoch': 2.47, 'throughput': 1102.85}
634
-
635
- [INFO|callbacks.py:310] 2024-07-11 11:49:17,985 >> {'loss': 0.0489, 'learning_rate': 1.6083e-06, 'epoch': 2.48, 'throughput': 1102.92}
636
-
637
- [INFO|callbacks.py:310] 2024-07-11 11:49:31,560 >> {'loss': 0.0621, 'learning_rate': 1.6167e-06, 'epoch': 2.50, 'throughput': 1102.86}
638
-
639
- [INFO|callbacks.py:310] 2024-07-11 11:49:45,187 >> {'loss': 0.0651, 'learning_rate': 1.6250e-06, 'epoch': 2.51, 'throughput': 1102.76}
640
-
641
- [INFO|callbacks.py:310] 2024-07-11 11:49:58,749 >> {'loss': 0.0398, 'learning_rate': 1.6333e-06, 'epoch': 2.52, 'throughput': 1102.95}
642
-
643
- [INFO|callbacks.py:310] 2024-07-11 11:50:12,256 >> {'loss': 0.0369, 'learning_rate': 1.6417e-06, 'epoch': 2.53, 'throughput': 1102.84}
644
-
645
- [INFO|callbacks.py:310] 2024-07-11 11:50:25,823 >> {'loss': 0.0582, 'learning_rate': 1.6500e-06, 'epoch': 2.55, 'throughput': 1102.81}
646
-
647
- [INFO|callbacks.py:310] 2024-07-11 11:50:39,396 >> {'loss': 0.0479, 'learning_rate': 1.6583e-06, 'epoch': 2.56, 'throughput': 1102.76}
648
-
649
- [INFO|callbacks.py:310] 2024-07-11 11:50:52,966 >> {'loss': 0.0561, 'learning_rate': 1.6667e-06, 'epoch': 2.57, 'throughput': 1102.65}
650
-
651
- [INFO|callbacks.py:310] 2024-07-11 11:51:06,538 >> {'loss': 0.0497, 'learning_rate': 1.6750e-06, 'epoch': 2.59, 'throughput': 1102.71}
652
-
653
- [INFO|callbacks.py:310] 2024-07-11 11:51:20,116 >> {'loss': 0.0630, 'learning_rate': 1.6833e-06, 'epoch': 2.60, 'throughput': 1102.74}
654
-
655
- [INFO|callbacks.py:310] 2024-07-11 11:51:33,695 >> {'loss': 0.0540, 'learning_rate': 1.6917e-06, 'epoch': 2.61, 'throughput': 1102.72}
656
-
657
- [INFO|callbacks.py:310] 2024-07-11 11:51:47,299 >> {'loss': 0.0729, 'learning_rate': 1.7000e-06, 'epoch': 2.62, 'throughput': 1102.94}
658
-
659
- [INFO|callbacks.py:310] 2024-07-11 11:52:00,917 >> {'loss': 0.0685, 'learning_rate': 1.7083e-06, 'epoch': 2.64, 'throughput': 1103.02}
660
-
661
- [INFO|callbacks.py:310] 2024-07-11 11:52:14,561 >> {'loss': 0.0823, 'learning_rate': 1.7167e-06, 'epoch': 2.65, 'throughput': 1103.02}
662
-
663
- [INFO|callbacks.py:310] 2024-07-11 11:52:28,087 >> {'loss': 0.0322, 'learning_rate': 1.7250e-06, 'epoch': 2.66, 'throughput': 1102.86}
664
-
665
- [INFO|callbacks.py:310] 2024-07-11 11:52:41,692 >> {'loss': 0.0821, 'learning_rate': 1.7333e-06, 'epoch': 2.68, 'throughput': 1103.07}
666
-
667
- [INFO|callbacks.py:310] 2024-07-11 11:52:55,309 >> {'loss': 0.0561, 'learning_rate': 1.7417e-06, 'epoch': 2.69, 'throughput': 1103.13}
668
-
669
- [INFO|callbacks.py:310] 2024-07-11 11:53:08,887 >> {'loss': 0.0468, 'learning_rate': 1.7500e-06, 'epoch': 2.70, 'throughput': 1103.09}
670
-
671
- [INFO|callbacks.py:310] 2024-07-11 11:53:22,397 >> {'loss': 0.0593, 'learning_rate': 1.7583e-06, 'epoch': 2.71, 'throughput': 1103.31}
672
-
673
- [INFO|callbacks.py:310] 2024-07-11 11:53:35,991 >> {'loss': 0.0403, 'learning_rate': 1.7667e-06, 'epoch': 2.73, 'throughput': 1103.22}
674
-
675
- [INFO|callbacks.py:310] 2024-07-11 11:53:49,564 >> {'loss': 0.0459, 'learning_rate': 1.7750e-06, 'epoch': 2.74, 'throughput': 1103.05}
676
-
677
- [INFO|callbacks.py:310] 2024-07-11 11:54:03,161 >> {'loss': 0.0509, 'learning_rate': 1.7833e-06, 'epoch': 2.75, 'throughput': 1102.91}
678
-
679
- [INFO|callbacks.py:310] 2024-07-11 11:54:16,779 >> {'loss': 0.0873, 'learning_rate': 1.7917e-06, 'epoch': 2.77, 'throughput': 1103.07}
680
-
681
- [INFO|callbacks.py:310] 2024-07-11 11:54:30,334 >> {'loss': 0.0618, 'learning_rate': 1.8000e-06, 'epoch': 2.78, 'throughput': 1103.10}
682
-
683
- [INFO|callbacks.py:310] 2024-07-11 11:54:43,886 >> {'loss': 0.1025, 'learning_rate': 1.8083e-06, 'epoch': 2.79, 'throughput': 1102.80}
684
-
685
- [INFO|callbacks.py:310] 2024-07-11 11:54:57,475 >> {'loss': 0.0669, 'learning_rate': 1.8167e-06, 'epoch': 2.80, 'throughput': 1102.74}
686
-
687
- [INFO|callbacks.py:310] 2024-07-11 11:55:11,084 >> {'loss': 0.0447, 'learning_rate': 1.8250e-06, 'epoch': 2.82, 'throughput': 1102.58}
688
-
689
- [INFO|callbacks.py:310] 2024-07-11 11:55:24,676 >> {'loss': 0.0635, 'learning_rate': 1.8333e-06, 'epoch': 2.83, 'throughput': 1102.68}
690
-
691
- [INFO|callbacks.py:310] 2024-07-11 11:55:38,261 >> {'loss': 0.0614, 'learning_rate': 1.8417e-06, 'epoch': 2.84, 'throughput': 1102.57}
692
-
693
- [INFO|callbacks.py:310] 2024-07-11 11:55:51,820 >> {'loss': 0.0646, 'learning_rate': 1.8500e-06, 'epoch': 2.86, 'throughput': 1102.56}
694
-
695
- [INFO|callbacks.py:310] 2024-07-11 11:56:05,370 >> {'loss': 0.0893, 'learning_rate': 1.8583e-06, 'epoch': 2.87, 'throughput': 1102.76}
696
-
697
- [INFO|callbacks.py:310] 2024-07-11 11:56:18,931 >> {'loss': 0.0653, 'learning_rate': 1.8667e-06, 'epoch': 2.88, 'throughput': 1102.83}
698
-
699
- [INFO|callbacks.py:310] 2024-07-11 11:56:32,505 >> {'loss': 0.0402, 'learning_rate': 1.8750e-06, 'epoch': 2.89, 'throughput': 1102.96}
700
-
701
- [INFO|callbacks.py:310] 2024-07-11 11:56:46,074 >> {'loss': 0.0407, 'learning_rate': 1.8833e-06, 'epoch': 2.91, 'throughput': 1103.17}
702
-
703
- [INFO|callbacks.py:310] 2024-07-11 11:56:59,621 >> {'loss': 0.0949, 'learning_rate': 1.8917e-06, 'epoch': 2.92, 'throughput': 1103.04}
704
-
705
- [INFO|callbacks.py:310] 2024-07-11 11:57:13,230 >> {'loss': 0.0789, 'learning_rate': 1.9000e-06, 'epoch': 2.93, 'throughput': 1103.14}
706
-
707
- [INFO|callbacks.py:310] 2024-07-11 11:57:26,812 >> {'loss': 0.0438, 'learning_rate': 1.9083e-06, 'epoch': 2.95, 'throughput': 1103.18}
708
-
709
- [INFO|callbacks.py:310] 2024-07-11 11:57:40,386 >> {'loss': 0.0905, 'learning_rate': 1.9167e-06, 'epoch': 2.96, 'throughput': 1103.48}
710
-
711
- [INFO|callbacks.py:310] 2024-07-11 11:57:53,898 >> {'loss': 0.0495, 'learning_rate': 1.9250e-06, 'epoch': 2.97, 'throughput': 1103.43}
712
-
713
- [INFO|callbacks.py:310] 2024-07-11 11:58:07,535 >> {'loss': 0.0642, 'learning_rate': 1.9333e-06, 'epoch': 2.98, 'throughput': 1103.64}
714
-
715
- [INFO|callbacks.py:310] 2024-07-11 11:58:21,083 >> {'loss': 0.0607, 'learning_rate': 1.9417e-06, 'epoch': 3.00, 'throughput': 1103.60}
716
-
717
- [INFO|callbacks.py:310] 2024-07-11 11:58:34,682 >> {'loss': 0.0363, 'learning_rate': 1.9500e-06, 'epoch': 3.01, 'throughput': 1103.54}
718
-
719
- [INFO|callbacks.py:310] 2024-07-11 11:58:48,313 >> {'loss': 0.0492, 'learning_rate': 1.9583e-06, 'epoch': 3.02, 'throughput': 1103.42}
720
-
721
- [INFO|callbacks.py:310] 2024-07-11 11:59:01,897 >> {'loss': 0.0373, 'learning_rate': 1.9667e-06, 'epoch': 3.04, 'throughput': 1103.50}
722
-
723
- [INFO|callbacks.py:310] 2024-07-11 11:59:15,458 >> {'loss': 0.0203, 'learning_rate': 1.9750e-06, 'epoch': 3.05, 'throughput': 1103.35}
724
-
725
- [INFO|callbacks.py:310] 2024-07-11 11:59:28,966 >> {'loss': 0.0175, 'learning_rate': 1.9833e-06, 'epoch': 3.06, 'throughput': 1103.44}
726
-
727
- [INFO|callbacks.py:310] 2024-07-11 11:59:42,557 >> {'loss': 0.0497, 'learning_rate': 1.9917e-06, 'epoch': 3.07, 'throughput': 1103.62}
728
-
729
- [INFO|callbacks.py:310] 2024-07-11 11:59:56,112 >> {'loss': 0.0361, 'learning_rate': 2.0000e-06, 'epoch': 3.09, 'throughput': 1103.61}
730
-
731
- [INFO|callbacks.py:310] 2024-07-11 12:00:09,694 >> {'loss': 0.0193, 'learning_rate': 2.0083e-06, 'epoch': 3.10, 'throughput': 1103.63}
732
-
733
- [INFO|callbacks.py:310] 2024-07-11 12:00:23,246 >> {'loss': 0.0142, 'learning_rate': 2.0167e-06, 'epoch': 3.11, 'throughput': 1103.37}
734
-
735
- [INFO|callbacks.py:310] 2024-07-11 12:00:36,823 >> {'loss': 0.0415, 'learning_rate': 2.0250e-06, 'epoch': 3.13, 'throughput': 1103.30}
736
-
737
- [INFO|callbacks.py:310] 2024-07-11 12:00:50,432 >> {'loss': 0.0178, 'learning_rate': 2.0333e-06, 'epoch': 3.14, 'throughput': 1103.37}
738
-
739
- [INFO|callbacks.py:310] 2024-07-11 12:01:03,966 >> {'loss': 0.0166, 'learning_rate': 2.0417e-06, 'epoch': 3.15, 'throughput': 1103.51}
740
-
741
- [INFO|callbacks.py:310] 2024-07-11 12:01:17,574 >> {'loss': 0.0424, 'learning_rate': 2.0500e-06, 'epoch': 3.16, 'throughput': 1103.59}
742
-
743
- [INFO|callbacks.py:310] 2024-07-11 12:01:31,112 >> {'loss': 0.0464, 'learning_rate': 2.0583e-06, 'epoch': 3.18, 'throughput': 1103.67}
744
-
745
- [INFO|callbacks.py:310] 2024-07-11 12:01:44,672 >> {'loss': 0.0235, 'learning_rate': 2.0667e-06, 'epoch': 3.19, 'throughput': 1103.53}
746
-
747
- [INFO|callbacks.py:310] 2024-07-11 12:01:58,233 >> {'loss': 0.0128, 'learning_rate': 2.0750e-06, 'epoch': 3.20, 'throughput': 1103.41}
748
-
749
- [INFO|callbacks.py:310] 2024-07-11 12:02:11,831 >> {'loss': 0.0319, 'learning_rate': 2.0833e-06, 'epoch': 3.22, 'throughput': 1103.70}
750
-
751
- [INFO|callbacks.py:310] 2024-07-11 12:02:25,358 >> {'loss': 0.0196, 'learning_rate': 2.0917e-06, 'epoch': 3.23, 'throughput': 1103.57}
752
-
753
- [INFO|callbacks.py:310] 2024-07-11 12:02:38,937 >> {'loss': 0.0326, 'learning_rate': 2.1000e-06, 'epoch': 3.24, 'throughput': 1103.58}
754
-
755
- [INFO|callbacks.py:310] 2024-07-11 12:02:52,507 >> {'loss': 0.0170, 'learning_rate': 2.1083e-06, 'epoch': 3.25, 'throughput': 1103.43}
756
-
757
- [INFO|callbacks.py:310] 2024-07-11 12:03:06,088 >> {'loss': 0.0372, 'learning_rate': 2.1167e-06, 'epoch': 3.27, 'throughput': 1103.48}
758
-
759
- [INFO|callbacks.py:310] 2024-07-11 12:03:19,699 >> {'loss': 0.0165, 'learning_rate': 2.1250e-06, 'epoch': 3.28, 'throughput': 1103.48}
760
-
761
- [INFO|callbacks.py:310] 2024-07-11 12:03:33,259 >> {'loss': 0.0142, 'learning_rate': 2.1333e-06, 'epoch': 3.29, 'throughput': 1103.67}
762
-
763
- [INFO|callbacks.py:310] 2024-07-11 12:03:46,842 >> {'loss': 0.0117, 'learning_rate': 2.1417e-06, 'epoch': 3.31, 'throughput': 1103.45}
764
-
765
- [INFO|callbacks.py:310] 2024-07-11 12:04:00,471 >> {'loss': 0.0264, 'learning_rate': 2.1500e-06, 'epoch': 3.32, 'throughput': 1103.60}
766
-
767
- [INFO|callbacks.py:310] 2024-07-11 12:04:14,075 >> {'loss': 0.0340, 'learning_rate': 2.1583e-06, 'epoch': 3.33, 'throughput': 1103.69}
768
-
769
- [INFO|callbacks.py:310] 2024-07-11 12:04:27,601 >> {'loss': 0.0310, 'learning_rate': 2.1667e-06, 'epoch': 3.34, 'throughput': 1103.68}
770
-
771
- [INFO|callbacks.py:310] 2024-07-11 12:04:41,161 >> {'loss': 0.0180, 'learning_rate': 2.1750e-06, 'epoch': 3.36, 'throughput': 1103.52}
772
-
773
- [INFO|callbacks.py:310] 2024-07-11 12:04:54,755 >> {'loss': 0.0461, 'learning_rate': 2.1833e-06, 'epoch': 3.37, 'throughput': 1103.57}
774
-
775
- [INFO|callbacks.py:310] 2024-07-11 12:05:08,257 >> {'loss': 0.0372, 'learning_rate': 2.1917e-06, 'epoch': 3.38, 'throughput': 1103.44}
776
-
777
- [INFO|callbacks.py:310] 2024-07-11 12:05:21,809 >> {'loss': 0.0471, 'learning_rate': 2.2000e-06, 'epoch': 3.40, 'throughput': 1103.29}
778
-
779
- [INFO|callbacks.py:310] 2024-07-11 12:05:35,395 >> {'loss': 0.0370, 'learning_rate': 2.2083e-06, 'epoch': 3.41, 'throughput': 1103.17}
780
-
781
- [INFO|callbacks.py:310] 2024-07-11 12:05:48,966 >> {'loss': 0.0245, 'learning_rate': 2.2167e-06, 'epoch': 3.42, 'throughput': 1103.04}
782
-
783
- [INFO|callbacks.py:310] 2024-07-11 12:06:02,486 >> {'loss': 0.0233, 'learning_rate': 2.2250e-06, 'epoch': 3.43, 'throughput': 1103.10}
784
-
785
- [INFO|callbacks.py:310] 2024-07-11 12:06:16,087 >> {'loss': 0.0256, 'learning_rate': 2.2333e-06, 'epoch': 3.45, 'throughput': 1103.15}
786
-
787
- [INFO|callbacks.py:310] 2024-07-11 12:06:29,677 >> {'loss': 0.0345, 'learning_rate': 2.2417e-06, 'epoch': 3.46, 'throughput': 1103.06}
788
-
789
- [INFO|callbacks.py:310] 2024-07-11 12:06:43,259 >> {'loss': 0.0656, 'learning_rate': 2.2500e-06, 'epoch': 3.47, 'throughput': 1103.11}
790
-
791
- [INFO|callbacks.py:310] 2024-07-11 12:06:56,838 >> {'loss': 0.0247, 'learning_rate': 2.2583e-06, 'epoch': 3.49, 'throughput': 1103.12}
792
-
793
- [INFO|callbacks.py:310] 2024-07-11 12:07:10,402 >> {'loss': 0.0510, 'learning_rate': 2.2667e-06, 'epoch': 3.50, 'throughput': 1103.18}
794
-
795
- [INFO|callbacks.py:310] 2024-07-11 12:07:23,955 >> {'loss': 0.0417, 'learning_rate': 2.2750e-06, 'epoch': 3.51, 'throughput': 1103.19}
796
-
797
- [INFO|callbacks.py:310] 2024-07-11 12:07:37,561 >> {'loss': 0.0176, 'learning_rate': 2.2833e-06, 'epoch': 3.52, 'throughput': 1103.12}
798
-
799
- [INFO|callbacks.py:310] 2024-07-11 12:07:51,125 >> {'loss': 0.0155, 'learning_rate': 2.2917e-06, 'epoch': 3.54, 'throughput': 1102.99}
800
-
801
- [INFO|callbacks.py:310] 2024-07-11 12:08:04,721 >> {'loss': 0.0135, 'learning_rate': 2.3000e-06, 'epoch': 3.55, 'throughput': 1103.13}
802
-
803
- [INFO|callbacks.py:310] 2024-07-11 12:08:18,303 >> {'loss': 0.0222, 'learning_rate': 2.3083e-06, 'epoch': 3.56, 'throughput': 1103.23}
804
-
805
- [INFO|callbacks.py:310] 2024-07-11 12:08:31,900 >> {'loss': 0.0289, 'learning_rate': 2.3167e-06, 'epoch': 3.58, 'throughput': 1103.41}
806
-
807
- [INFO|callbacks.py:310] 2024-07-11 12:08:45,544 >> {'loss': 0.0314, 'learning_rate': 2.3250e-06, 'epoch': 3.59, 'throughput': 1103.39}
808
-
809
- [INFO|callbacks.py:310] 2024-07-11 12:08:59,085 >> {'loss': 0.0582, 'learning_rate': 2.3333e-06, 'epoch': 3.60, 'throughput': 1103.12}
810
-
811
- [INFO|callbacks.py:310] 2024-07-11 12:09:12,702 >> {'loss': 0.0339, 'learning_rate': 2.3417e-06, 'epoch': 3.61, 'throughput': 1103.28}
812
-
813
- [INFO|callbacks.py:310] 2024-07-11 12:09:26,307 >> {'loss': 0.0656, 'learning_rate': 2.3500e-06, 'epoch': 3.63, 'throughput': 1103.32}
814
-
815
- [INFO|callbacks.py:310] 2024-07-11 12:09:39,915 >> {'loss': 0.0355, 'learning_rate': 2.3583e-06, 'epoch': 3.64, 'throughput': 1103.41}
816
-
817
- [INFO|callbacks.py:310] 2024-07-11 12:09:53,473 >> {'loss': 0.0370, 'learning_rate': 2.3667e-06, 'epoch': 3.65, 'throughput': 1103.36}
818
-
819
- [INFO|callbacks.py:310] 2024-07-11 12:10:07,008 >> {'loss': 0.0206, 'learning_rate': 2.3750e-06, 'epoch': 3.67, 'throughput': 1103.45}
820
-
821
- [INFO|callbacks.py:310] 2024-07-11 12:10:20,629 >> {'loss': 0.0286, 'learning_rate': 2.3833e-06, 'epoch': 3.68, 'throughput': 1103.36}
822
-
823
- [INFO|callbacks.py:310] 2024-07-11 12:10:34,205 >> {'loss': 0.0316, 'learning_rate': 2.3917e-06, 'epoch': 3.69, 'throughput': 1103.44}
824
-
825
- [INFO|callbacks.py:310] 2024-07-11 12:10:47,785 >> {'loss': 0.0263, 'learning_rate': 2.4000e-06, 'epoch': 3.70, 'throughput': 1103.55}
826
-
827
- [INFO|callbacks.py:310] 2024-07-11 12:11:01,368 >> {'loss': 0.0265, 'learning_rate': 2.4083e-06, 'epoch': 3.72, 'throughput': 1103.75}
828
-
829
- [INFO|callbacks.py:310] 2024-07-11 12:11:14,942 >> {'loss': 0.0251, 'learning_rate': 2.4167e-06, 'epoch': 3.73, 'throughput': 1103.60}
830
-
831
- [INFO|callbacks.py:310] 2024-07-11 12:11:28,479 >> {'loss': 0.0293, 'learning_rate': 2.4250e-06, 'epoch': 3.74, 'throughput': 1103.51}
832
-
833
- [INFO|callbacks.py:310] 2024-07-11 12:11:42,009 >> {'loss': 0.0278, 'learning_rate': 2.4333e-06, 'epoch': 3.76, 'throughput': 1103.50}
834
-
835
- [INFO|callbacks.py:310] 2024-07-11 12:11:55,561 >> {'loss': 0.0350, 'learning_rate': 2.4417e-06, 'epoch': 3.77, 'throughput': 1103.52}
836
-
837
- [INFO|callbacks.py:310] 2024-07-11 12:12:09,121 >> {'loss': 0.0455, 'learning_rate': 2.4500e-06, 'epoch': 3.78, 'throughput': 1103.38}
838
-
839
- [INFO|callbacks.py:310] 2024-07-11 12:12:22,668 >> {'loss': 0.0429, 'learning_rate': 2.4583e-06, 'epoch': 3.79, 'throughput': 1103.20}
840
-
841
- [INFO|callbacks.py:310] 2024-07-11 12:12:36,239 >> {'loss': 0.0560, 'learning_rate': 2.4667e-06, 'epoch': 3.81, 'throughput': 1103.28}
842
-
843
- [INFO|callbacks.py:310] 2024-07-11 12:12:49,764 >> {'loss': 0.0325, 'learning_rate': 2.4750e-06, 'epoch': 3.82, 'throughput': 1103.35}
844
-
845
- [INFO|callbacks.py:310] 2024-07-11 12:13:03,351 >> {'loss': 0.0401, 'learning_rate': 2.4833e-06, 'epoch': 3.83, 'throughput': 1103.39}
846
-
847
- [INFO|callbacks.py:310] 2024-07-11 12:13:16,918 >> {'loss': 0.0182, 'learning_rate': 2.4917e-06, 'epoch': 3.85, 'throughput': 1103.37}
848
-
849
- [INFO|callbacks.py:310] 2024-07-11 12:13:30,518 >> {'loss': 0.0344, 'learning_rate': 2.5000e-06, 'epoch': 3.86, 'throughput': 1103.37}
850
-
851
- [INFO|callbacks.py:310] 2024-07-11 12:13:44,052 >> {'loss': 0.0111, 'learning_rate': 2.5083e-06, 'epoch': 3.87, 'throughput': 1103.37}
852
-
853
- [INFO|callbacks.py:310] 2024-07-11 12:13:57,592 >> {'loss': 0.0088, 'learning_rate': 2.5167e-06, 'epoch': 3.88, 'throughput': 1103.52}
854
-
855
- [INFO|callbacks.py:310] 2024-07-11 12:14:11,256 >> {'loss': 0.0130, 'learning_rate': 2.5250e-06, 'epoch': 3.90, 'throughput': 1103.56}
856
-
857
- [INFO|callbacks.py:310] 2024-07-11 12:14:24,831 >> {'loss': 0.0433, 'learning_rate': 2.5333e-06, 'epoch': 3.91, 'throughput': 1103.64}
858
-
859
- [INFO|callbacks.py:310] 2024-07-11 12:14:38,383 >> {'loss': 0.0214, 'learning_rate': 2.5417e-06, 'epoch': 3.92, 'throughput': 1103.57}
860
-
861
- [INFO|callbacks.py:310] 2024-07-11 12:14:51,973 >> {'loss': 0.0416, 'learning_rate': 2.5500e-06, 'epoch': 3.94, 'throughput': 1103.50}
862
-
863
- [INFO|callbacks.py:310] 2024-07-11 12:15:05,507 >> {'loss': 0.0321, 'learning_rate': 2.5583e-06, 'epoch': 3.95, 'throughput': 1103.41}
864
-
865
- [INFO|callbacks.py:310] 2024-07-11 12:15:19,098 >> {'loss': 0.0526, 'learning_rate': 2.5667e-06, 'epoch': 3.96, 'throughput': 1103.50}
866
-
867
- [INFO|callbacks.py:310] 2024-07-11 12:15:32,712 >> {'loss': 0.0250, 'learning_rate': 2.5750e-06, 'epoch': 3.97, 'throughput': 1103.53}
868
-
869
- [INFO|callbacks.py:310] 2024-07-11 12:15:46,305 >> {'loss': 0.0787, 'learning_rate': 2.5833e-06, 'epoch': 3.99, 'throughput': 1103.77}
870
-
871
- [INFO|callbacks.py:310] 2024-07-11 12:15:59,841 >> {'loss': 0.0559, 'learning_rate': 2.5917e-06, 'epoch': 4.00, 'throughput': 1103.76}
872
-
873
- [INFO|callbacks.py:310] 2024-07-11 12:16:13,395 >> {'loss': 0.0259, 'learning_rate': 2.6000e-06, 'epoch': 4.01, 'throughput': 1103.73}
874
-
875
- [INFO|callbacks.py:310] 2024-07-11 12:16:27,011 >> {'loss': 0.0370, 'learning_rate': 2.6083e-06, 'epoch': 4.03, 'throughput': 1103.60}
876
-
877
- [INFO|callbacks.py:310] 2024-07-11 12:16:40,583 >> {'loss': 0.0293, 'learning_rate': 2.6167e-06, 'epoch': 4.04, 'throughput': 1103.59}
878
-
879
- [INFO|callbacks.py:310] 2024-07-11 12:16:54,092 >> {'loss': 0.0153, 'learning_rate': 2.6250e-06, 'epoch': 4.05, 'throughput': 1103.45}
880
-
881
- [INFO|callbacks.py:310] 2024-07-11 12:17:07,743 >> {'loss': 0.0234, 'learning_rate': 2.6333e-06, 'epoch': 4.06, 'throughput': 1103.75}
882
-
883
- [INFO|callbacks.py:310] 2024-07-11 12:17:21,282 >> {'loss': 0.0212, 'learning_rate': 2.6417e-06, 'epoch': 4.08, 'throughput': 1103.71}
884
-
885
- [INFO|callbacks.py:310] 2024-07-11 12:17:34,845 >> {'loss': 0.0187, 'learning_rate': 2.6500e-06, 'epoch': 4.09, 'throughput': 1103.79}
886
-
887
- [INFO|callbacks.py:310] 2024-07-11 12:17:48,409 >> {'loss': 0.0245, 'learning_rate': 2.6583e-06, 'epoch': 4.10, 'throughput': 1103.86}
888
-
889
- [INFO|callbacks.py:310] 2024-07-11 12:18:01,997 >> {'loss': 0.0032, 'learning_rate': 2.6667e-06, 'epoch': 4.12, 'throughput': 1103.80}
890
-
891
- [INFO|callbacks.py:310] 2024-07-11 12:18:15,570 >> {'loss': 0.0367, 'learning_rate': 2.6750e-06, 'epoch': 4.13, 'throughput': 1103.65}
892
-
893
- [INFO|callbacks.py:310] 2024-07-11 12:18:29,133 >> {'loss': 0.0380, 'learning_rate': 2.6833e-06, 'epoch': 4.14, 'throughput': 1103.71}
894
-
895
- [INFO|callbacks.py:310] 2024-07-11 12:18:42,699 >> {'loss': 0.0194, 'learning_rate': 2.6917e-06, 'epoch': 4.15, 'throughput': 1103.55}
896
-
897
- [INFO|callbacks.py:310] 2024-07-11 12:18:56,286 >> {'loss': 0.0262, 'learning_rate': 2.7000e-06, 'epoch': 4.17, 'throughput': 1103.64}
898
-
899
- [INFO|callbacks.py:310] 2024-07-11 12:19:09,822 >> {'loss': 0.0151, 'learning_rate': 2.7083e-06, 'epoch': 4.18, 'throughput': 1103.50}
900
-
901
- [INFO|callbacks.py:310] 2024-07-11 12:19:23,377 >> {'loss': 0.0267, 'learning_rate': 2.7167e-06, 'epoch': 4.19, 'throughput': 1103.52}
902
-
903
- [INFO|callbacks.py:310] 2024-07-11 12:19:36,893 >> {'loss': 0.0041, 'learning_rate': 2.7250e-06, 'epoch': 4.21, 'throughput': 1103.49}
904
-
905
- [INFO|callbacks.py:310] 2024-07-11 12:19:50,423 >> {'loss': 0.0365, 'learning_rate': 2.7333e-06, 'epoch': 4.22, 'throughput': 1103.40}
906
-
907
- [INFO|callbacks.py:310] 2024-07-11 12:20:04,049 >> {'loss': 0.0151, 'learning_rate': 2.7417e-06, 'epoch': 4.23, 'throughput': 1103.57}
908
-
909
- [INFO|callbacks.py:310] 2024-07-11 12:20:17,604 >> {'loss': 0.0060, 'learning_rate': 2.7500e-06, 'epoch': 4.24, 'throughput': 1103.59}
910
-
911
- [INFO|callbacks.py:310] 2024-07-11 12:20:31,180 >> {'loss': 0.0247, 'learning_rate': 2.7583e-06, 'epoch': 4.26, 'throughput': 1103.51}
912
-
913
- [INFO|callbacks.py:310] 2024-07-11 12:20:44,824 >> {'loss': 0.0147, 'learning_rate': 2.7667e-06, 'epoch': 4.27, 'throughput': 1103.61}
914
-
915
- [INFO|callbacks.py:310] 2024-07-11 12:20:58,301 >> {'loss': 0.0413, 'learning_rate': 2.7750e-06, 'epoch': 4.28, 'throughput': 1103.45}
916
-
917
- [INFO|callbacks.py:310] 2024-07-11 12:21:11,864 >> {'loss': 0.0318, 'learning_rate': 2.7833e-06, 'epoch': 4.30, 'throughput': 1103.40}
918
-
919
- [INFO|callbacks.py:310] 2024-07-11 12:21:25,395 >> {'loss': 0.0513, 'learning_rate': 2.7917e-06, 'epoch': 4.31, 'throughput': 1103.30}
920
-
921
- [INFO|callbacks.py:310] 2024-07-11 12:21:38,939 >> {'loss': 0.0165, 'learning_rate': 2.8000e-06, 'epoch': 4.32, 'throughput': 1103.44}
922
-
923
- [INFO|callbacks.py:310] 2024-07-11 12:21:52,483 >> {'loss': 0.0131, 'learning_rate': 2.8083e-06, 'epoch': 4.33, 'throughput': 1103.48}
924
-
925
- [INFO|callbacks.py:310] 2024-07-11 12:22:06,128 >> {'loss': 0.0246, 'learning_rate': 2.8167e-06, 'epoch': 4.35, 'throughput': 1103.52}
926
-
927
- [INFO|callbacks.py:310] 2024-07-11 12:22:19,671 >> {'loss': 0.0465, 'learning_rate': 2.8250e-06, 'epoch': 4.36, 'throughput': 1103.46}
928
-
929
- [INFO|callbacks.py:310] 2024-07-11 12:22:33,289 >> {'loss': 0.0084, 'learning_rate': 2.8333e-06, 'epoch': 4.37, 'throughput': 1103.49}
930
-
931
- [INFO|callbacks.py:310] 2024-07-11 12:22:46,833 >> {'loss': 0.0105, 'learning_rate': 2.8417e-06, 'epoch': 4.39, 'throughput': 1103.35}
932
-
933
- [INFO|callbacks.py:310] 2024-07-11 12:23:00,393 >> {'loss': 0.0388, 'learning_rate': 2.8500e-06, 'epoch': 4.40, 'throughput': 1103.48}
934
-
935
- [INFO|callbacks.py:310] 2024-07-11 12:23:13,920 >> {'loss': 0.0486, 'learning_rate': 2.8583e-06, 'epoch': 4.41, 'throughput': 1103.54}
936
-
937
- [INFO|callbacks.py:310] 2024-07-11 12:23:27,450 >> {'loss': 0.0354, 'learning_rate': 2.8667e-06, 'epoch': 4.42, 'throughput': 1103.53}
938
-
939
- [INFO|callbacks.py:310] 2024-07-11 12:23:41,045 >> {'loss': 0.0464, 'learning_rate': 2.8750e-06, 'epoch': 4.44, 'throughput': 1103.38}
940
-
941
- [INFO|callbacks.py:310] 2024-07-11 12:23:54,593 >> {'loss': 0.0312, 'learning_rate': 2.8833e-06, 'epoch': 4.45, 'throughput': 1103.41}
942
-
943
- [INFO|callbacks.py:310] 2024-07-11 12:24:08,195 >> {'loss': 0.0352, 'learning_rate': 2.8917e-06, 'epoch': 4.46, 'throughput': 1103.50}
944
-
945
- [INFO|callbacks.py:310] 2024-07-11 12:24:21,721 >> {'loss': 0.0116, 'learning_rate': 2.9000e-06, 'epoch': 4.48, 'throughput': 1103.46}
946
-
947
- [INFO|callbacks.py:310] 2024-07-11 12:24:35,342 >> {'loss': 0.0118, 'learning_rate': 2.9083e-06, 'epoch': 4.49, 'throughput': 1103.57}
948
-
949
- [INFO|callbacks.py:310] 2024-07-11 12:24:48,908 >> {'loss': 0.0138, 'learning_rate': 2.9167e-06, 'epoch': 4.50, 'throughput': 1103.54}
950
-
951
- [INFO|callbacks.py:310] 2024-07-11 12:25:02,509 >> {'loss': 0.0344, 'learning_rate': 2.9250e-06, 'epoch': 4.51, 'throughput': 1103.54}
952
-
953
- [INFO|callbacks.py:310] 2024-07-11 12:25:16,091 >> {'loss': 0.0231, 'learning_rate': 2.9333e-06, 'epoch': 4.53, 'throughput': 1103.62}
954
-
955
- [INFO|callbacks.py:310] 2024-07-11 12:25:29,594 >> {'loss': 0.0267, 'learning_rate': 2.9417e-06, 'epoch': 4.54, 'throughput': 1103.62}
956
-
957
- [INFO|callbacks.py:310] 2024-07-11 12:25:43,090 >> {'loss': 0.0109, 'learning_rate': 2.9500e-06, 'epoch': 4.55, 'throughput': 1103.38}
958
-
959
- [INFO|callbacks.py:310] 2024-07-11 12:25:56,700 >> {'loss': 0.0444, 'learning_rate': 2.9583e-06, 'epoch': 4.57, 'throughput': 1103.26}
960
-
961
- [INFO|callbacks.py:310] 2024-07-11 12:26:10,257 >> {'loss': 0.0142, 'learning_rate': 2.9667e-06, 'epoch': 4.58, 'throughput': 1103.35}
962
-
963
- [INFO|callbacks.py:310] 2024-07-11 12:26:23,765 >> {'loss': 0.0361, 'learning_rate': 2.9750e-06, 'epoch': 4.59, 'throughput': 1103.37}
964
-
965
- [INFO|callbacks.py:310] 2024-07-11 12:26:37,356 >> {'loss': 0.0197, 'learning_rate': 2.9833e-06, 'epoch': 4.60, 'throughput': 1103.36}
966
-
967
- [INFO|callbacks.py:310] 2024-07-11 12:26:50,915 >> {'loss': 0.0373, 'learning_rate': 2.9917e-06, 'epoch': 4.62, 'throughput': 1103.44}
968
-
969
- [INFO|callbacks.py:310] 2024-07-11 12:27:04,473 >> {'loss': 0.0469, 'learning_rate': 3.0000e-06, 'epoch': 4.63, 'throughput': 1103.59}
970
-
971
- [INFO|callbacks.py:310] 2024-07-11 12:27:17,984 >> {'loss': 0.0283, 'learning_rate': 3.0083e-06, 'epoch': 4.64, 'throughput': 1103.61}
972
-
973
- [INFO|callbacks.py:310] 2024-07-11 12:27:31,534 >> {'loss': 0.0271, 'learning_rate': 3.0167e-06, 'epoch': 4.66, 'throughput': 1103.55}
974
-
975
- [INFO|callbacks.py:310] 2024-07-11 12:27:45,113 >> {'loss': 0.0303, 'learning_rate': 3.0250e-06, 'epoch': 4.67, 'throughput': 1103.54}
976
-
977
- [INFO|callbacks.py:310] 2024-07-11 12:27:58,741 >> {'loss': 0.0207, 'learning_rate': 3.0333e-06, 'epoch': 4.68, 'throughput': 1103.69}
978
-
979
- [INFO|callbacks.py:310] 2024-07-11 12:28:12,347 >> {'loss': 0.0157, 'learning_rate': 3.0417e-06, 'epoch': 4.69, 'throughput': 1103.80}
980
-
981
- [INFO|callbacks.py:310] 2024-07-11 12:28:25,988 >> {'loss': 0.0128, 'learning_rate': 3.0500e-06, 'epoch': 4.71, 'throughput': 1103.71}
982
-
983
- [INFO|callbacks.py:310] 2024-07-11 12:28:39,517 >> {'loss': 0.0293, 'learning_rate': 3.0583e-06, 'epoch': 4.72, 'throughput': 1103.92}
984
-
985
- [INFO|callbacks.py:310] 2024-07-11 12:28:53,124 >> {'loss': 0.0211, 'learning_rate': 3.0667e-06, 'epoch': 4.73, 'throughput': 1103.87}
986
-
987
- [INFO|callbacks.py:310] 2024-07-11 12:29:06,615 >> {'loss': 0.0929, 'learning_rate': 3.0750e-06, 'epoch': 4.75, 'throughput': 1103.90}
988
-
989
- [INFO|callbacks.py:310] 2024-07-11 12:29:20,204 >> {'loss': 0.0195, 'learning_rate': 3.0833e-06, 'epoch': 4.76, 'throughput': 1103.97}
990
-
991
- [INFO|callbacks.py:310] 2024-07-11 12:29:33,779 >> {'loss': 0.0114, 'learning_rate': 3.0917e-06, 'epoch': 4.77, 'throughput': 1104.05}
992
-
993
- [INFO|callbacks.py:310] 2024-07-11 12:29:47,300 >> {'loss': 0.0207, 'learning_rate': 3.1000e-06, 'epoch': 4.78, 'throughput': 1104.23}
994
-
995
- [INFO|callbacks.py:310] 2024-07-11 12:30:00,871 >> {'loss': 0.0144, 'learning_rate': 3.1083e-06, 'epoch': 4.80, 'throughput': 1104.22}
996
-
997
- [INFO|callbacks.py:310] 2024-07-11 12:30:14,485 >> {'loss': 0.0197, 'learning_rate': 3.1167e-06, 'epoch': 4.81, 'throughput': 1104.27}
998
-
999
- [INFO|callbacks.py:310] 2024-07-11 12:30:28,105 >> {'loss': 0.0669, 'learning_rate': 3.1250e-06, 'epoch': 4.82, 'throughput': 1104.49}
1000
-
1001
- [INFO|callbacks.py:310] 2024-07-11 12:30:41,666 >> {'loss': 0.0156, 'learning_rate': 3.1333e-06, 'epoch': 4.84, 'throughput': 1104.45}
1002
-
1003
- [INFO|callbacks.py:310] 2024-07-11 12:30:55,231 >> {'loss': 0.0180, 'learning_rate': 3.1417e-06, 'epoch': 4.85, 'throughput': 1104.57}
1004
 
1005
- [INFO|callbacks.py:310] 2024-07-11 12:31:08,788 >> {'loss': 0.0640, 'learning_rate': 3.1500e-06, 'epoch': 4.86, 'throughput': 1104.56}
1006
 
1007
- [INFO|callbacks.py:310] 2024-07-11 12:31:22,373 >> {'loss': 0.0263, 'learning_rate': 3.1583e-06, 'epoch': 4.87, 'throughput': 1104.51}
 
1008
 
1009
- [INFO|callbacks.py:310] 2024-07-11 12:31:35,946 >> {'loss': 0.0256, 'learning_rate': 3.1667e-06, 'epoch': 4.89, 'throughput': 1104.46}
1010
 
1011
- [INFO|callbacks.py:310] 2024-07-11 12:31:49,501 >> {'loss': 0.0269, 'learning_rate': 3.1750e-06, 'epoch': 4.90, 'throughput': 1104.40}
1012
 
1013
- [INFO|callbacks.py:310] 2024-07-11 12:32:03,102 >> {'loss': 0.0281, 'learning_rate': 3.1833e-06, 'epoch': 4.91, 'throughput': 1104.41}
1014
 
1015
- [INFO|callbacks.py:310] 2024-07-11 12:32:16,664 >> {'loss': 0.0199, 'learning_rate': 3.1917e-06, 'epoch': 4.93, 'throughput': 1104.39}
1016
 
1017
- [INFO|callbacks.py:310] 2024-07-11 12:32:30,171 >> {'loss': 0.0277, 'learning_rate': 3.2000e-06, 'epoch': 4.94, 'throughput': 1104.28}
1018
 
1019
- [INFO|callbacks.py:310] 2024-07-11 12:32:43,768 >> {'loss': 0.0158, 'learning_rate': 3.2083e-06, 'epoch': 4.95, 'throughput': 1104.47}
1020
 
1021
- [INFO|trainer.py:3478] 2024-07-11 12:32:51,369 >> Saving model checkpoint to saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/checkpoint-385
1022
 
1023
- [INFO|configuration_utils.py:472] 2024-07-11 12:32:51,372 >> Configuration saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/checkpoint-385/config.json
1024
 
1025
- [INFO|configuration_utils.py:769] 2024-07-11 12:32:51,373 >> Configuration saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/checkpoint-385/generation_config.json
1026
 
1027
- [INFO|modeling_utils.py:2698] 2024-07-11 12:33:07,622 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 4 checkpoint shards. You can find where each parameters has been saved in the index located at saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/checkpoint-385/model.safetensors.index.json.
1028
 
1029
- [INFO|tokenization_utils_base.py:2574] 2024-07-11 12:33:07,625 >> tokenizer config file saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/checkpoint-385/tokenizer_config.json
1030
 
1031
- [INFO|tokenization_utils_base.py:2583] 2024-07-11 12:33:07,626 >> Special tokens file saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/checkpoint-385/special_tokens_map.json
1032
 
1033
- [INFO|trainer.py:2383] 2024-07-11 12:33:44,158 >>
1034
 
1035
- Training completed. Do not forget to share your model on huggingface.co/models =)
1036
 
 
1037
 
 
1038
 
1039
- [INFO|trainer.py:3478] 2024-07-11 12:33:51,651 >> Saving model checkpoint to saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3
1040
 
1041
- [INFO|configuration_utils.py:472] 2024-07-11 12:33:51,653 >> Configuration saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/config.json
1042
 
1043
- [INFO|configuration_utils.py:769] 2024-07-11 12:33:51,654 >> Configuration saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/generation_config.json
1044
 
1045
- [INFO|modeling_utils.py:2698] 2024-07-11 12:34:08,270 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 4 checkpoint shards. You can find where each parameters has been saved in the index located at saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/model.safetensors.index.json.
1046
 
1047
- [INFO|tokenization_utils_base.py:2574] 2024-07-11 12:34:08,273 >> tokenizer config file saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/tokenizer_config.json
1048
 
1049
- [INFO|tokenization_utils_base.py:2583] 2024-07-11 12:34:08,273 >> Special tokens file saved in saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/special_tokens_map.json
1050
 
1051
- [WARNING|ploting.py:89] 2024-07-11 12:34:09,588 >> No metric eval_loss to plot.
1052
 
1053
- [WARNING|ploting.py:89] 2024-07-11 12:34:09,588 >> No metric eval_accuracy to plot.
1054
 
1055
- [INFO|modelcard.py:449] 2024-07-11 12:34:09,589 >> Dropping the following result as it does not have all the necessary fields:
1056
- {'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}}
1057
 
 
1
+ [INFO|parser.py:325] 2024-07-11 13:12:32,834 >> Process rank: 0, device: cuda:0, n_gpu: 1, distributed training: True, compute dtype: None
2
 
3
+ [INFO|tokenization_utils_base.py:2159] 2024-07-11 13:12:32,836 >> loading file tokenizer.json
4
 
5
+ 07/11/2024 13:12:32 - INFO - llamafactory.hparams.parser - Process rank: 5, device: cuda:5, n_gpu: 1, distributed training: True, compute dtype: None
6
 
7
+ 07/11/2024 13:12:32 - INFO - llamafactory.hparams.parser - Process rank: 4, device: cuda:4, n_gpu: 1, distributed training: True, compute dtype: None
8
 
9
+ 07/11/2024 13:12:32 - INFO - llamafactory.hparams.parser - Process rank: 6, device: cuda:6, n_gpu: 1, distributed training: True, compute dtype: None
10
 
11
+ 07/11/2024 13:12:32 - INFO - llamafactory.hparams.parser - Process rank: 3, device: cuda:3, n_gpu: 1, distributed training: True, compute dtype: None
12
 
13
+ 07/11/2024 13:12:33 - INFO - llamafactory.hparams.parser - Process rank: 2, device: cuda:2, n_gpu: 1, distributed training: True, compute dtype: None
14
 
15
+ 07/11/2024 13:12:33 - INFO - llamafactory.hparams.parser - Process rank: 1, device: cuda:1, n_gpu: 1, distributed training: True, compute dtype: None
16
 
17
+ [INFO|tokenization_utils_base.py:2159] 2024-07-11 13:12:32,836 >> loading file added_tokens.json
18
 
19
+ [INFO|tokenization_utils_base.py:2159] 2024-07-11 13:12:32,836 >> loading file special_tokens_map.json
20
 
21
+ [INFO|tokenization_utils_base.py:2159] 2024-07-11 13:12:32,836 >> loading file tokenizer_config.json
22
 
23
+ [WARNING|logging.py:313] 2024-07-11 13:12:33,150 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
24
 
25
+ [INFO|template.py:270] 2024-07-11 13:12:33,150 >> Replace eos token: <|eot_id|>
26
 
27
+ [INFO|loader.py:50] 2024-07-11 13:12:33,151 >> Loading dataset dev_output.json...
28
 
29
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
30
 
31
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
32
 
33
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
34
 
35
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
36
 
37
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
38
 
39
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
40
 
41
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
42
 
43
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
44
 
45
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
46
 
47
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
48
 
49
+ 07/11/2024 13:12:33 - INFO - llamafactory.hparams.parser - Process rank: 7, device: cuda:7, n_gpu: 1, distributed training: True, compute dtype: None
50
 
51
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
52
 
53
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
54
 
55
+ 07/11/2024 13:12:33 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
56
 
57
+ 07/11/2024 13:12:33 - INFO - llamafactory.data.template - Replace eos token: <|eot_id|>
58
 
59
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
60
 
61
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
62
 
63
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
64
 
65
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
66
 
67
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
68
 
69
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
70
 
71
+ 07/11/2024 13:12:34 - INFO - llamafactory.data.loader - Loading dataset dev_output.json...
72
 
73
+ [INFO|configuration_utils.py:731] 2024-07-11 13:12:38,208 >> loading configuration file saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/config.json
74
 
75
+ [INFO|configuration_utils.py:800] 2024-07-11 13:12:38,209 >> Model config LlamaConfig {
76
+ "_name_or_path": "saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
77
  "architectures": [
78
  "LlamaForCausalLM"
79
  ],
 
98
  "tie_word_embeddings": false,
99
  "torch_dtype": "bfloat16",
100
  "transformers_version": "4.42.3",
101
+ "use_cache": false,
102
  "vocab_size": 128256
103
  }
104
 
105
 
106
+ [INFO|patcher.py:81] 2024-07-11 13:12:38,210 >> Using KV cache for faster generation.
107
+
108
+ [INFO|modeling_utils.py:3553] 2024-07-11 13:12:38,232 >> loading weights file saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/model.safetensors.index.json
109
 
110
+ [INFO|modeling_utils.py:1531] 2024-07-11 13:12:38,232 >> Instantiating LlamaForCausalLM model under default dtype torch.bfloat16.
111
 
112
+ [INFO|configuration_utils.py:1000] 2024-07-11 13:12:38,233 >> Generate config GenerationConfig {
113
  "bos_token_id": 128000,
114
  "eos_token_id": 128009
115
  }
116
 
117
 
118
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
119
+
120
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
121
+
122
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
123
 
124
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
125
 
126
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
127
 
128
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
129
 
130
+ 07/11/2024 13:12:38 - INFO - llamafactory.model.patcher - Using KV cache for faster generation.
131
 
132
+ [INFO|modeling_utils.py:4364] 2024-07-11 13:12:42,279 >> All model checkpoint weights were used when initializing LlamaForCausalLM.
133
 
134
 
135
+ [INFO|modeling_utils.py:4372] 2024-07-11 13:12:42,279 >> All the weights of LlamaForCausalLM were initialized from the model checkpoint at saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3.
136
  If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
137
 
138
+ [INFO|configuration_utils.py:953] 2024-07-11 13:12:42,283 >> loading configuration file saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3/generation_config.json
139
 
140
+ [INFO|configuration_utils.py:1000] 2024-07-11 13:12:42,283 >> Generate config GenerationConfig {
141
  "bos_token_id": 128000,
142
  "do_sample": true,
143
  "eos_token_id": [
 
150
  }
151
 
152
 
153
+ [INFO|attention.py:80] 2024-07-11 13:12:42,289 >> Using torch SDPA for faster training and inference.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
154
 
155
+ [INFO|loader.py:196] 2024-07-11 13:12:42,294 >> all params: 8,030,261,248
156
 
157
+ [INFO|trainer.py:3788] 2024-07-11 13:12:42,401 >>
158
+ ***** Running Prediction *****
159
 
160
+ [INFO|trainer.py:3790] 2024-07-11 13:12:42,401 >> Num examples = 2554
161
 
162
+ [INFO|trainer.py:3793] 2024-07-11 13:12:42,401 >> Batch size = 2
163
 
164
+ [WARNING|logging.py:328] 2024-07-11 13:12:43,055 >> We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
165
 
166
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
167
 
168
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
169
 
170
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
171
 
172
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
173
 
174
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
175
 
176
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
177
 
178
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
179
 
180
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
181
 
182
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
183
 
184
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
185
 
186
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
187
 
188
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
189
 
190
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference.
191
 
192
+ 07/11/2024 13:12:43 - INFO - llamafactory.model.loader - all params: 8,030,261,248
193
 
194
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
195
 
196
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
197
 
198
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
199
 
200
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
201
 
202
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
203
 
204
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
205
 
206
+ 07/11/2024 13:12:44 - WARNING - transformers.models.llama.modeling_llama - We detected that you are passing `past_key_values` as a tuple and this is deprecated and will be removed in v4.43. Please use an appropriate `Cache` class (https://huggingface.co/docs/transformers/v4.41.3/en/internal/generation_utils#transformers.Cache)
207
 
208
+ [INFO|trainer.py:127] 2024-07-11 13:12:59,490 >> Saving prediction results to saves/LLaMA3-8B-Chat/full/eval_2024-07-11-10-49-45/generated_predictions.jsonl
 
209
 
trainer_log.jsonl CHANGED
The diff for this file is too large to render. See raw diff
 
training_args.yaml CHANGED
@@ -1,30 +1,18 @@
1
- bf16: true
2
  cutoff_len: 1024
3
- dataset: truth_train
4
  dataset_dir: data
5
- ddp_timeout: 180000000
6
- deepspeed: cache/ds_z2_config.json
7
- do_train: true
8
  finetuning_type: full
9
  flash_attn: auto
10
- gradient_accumulation_steps: 8
11
- include_num_input_tokens_seen: true
12
- learning_rate: 5.0e-06
13
- logging_steps: 1
14
- lr_scheduler_type: cosine
15
- max_grad_norm: 1.0
16
  max_samples: 100000
17
- model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
18
- num_train_epochs: 5.0
19
- optim: adamw_torch
20
- output_dir: saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3
21
- packing: false
22
- per_device_train_batch_size: 4
23
- plot_loss: true
24
  preprocessing_num_workers: 16
25
  quantization_method: bitsandbytes
26
- report_to: none
27
- save_steps: 1000
28
  stage: sft
 
29
  template: llama3
30
- warmup_steps: 600
 
 
1
  cutoff_len: 1024
2
+ dataset: truth_dev
3
  dataset_dir: data
4
+ do_predict: true
 
 
5
  finetuning_type: full
6
  flash_attn: auto
7
+ max_new_tokens: 512
 
 
 
 
 
8
  max_samples: 100000
9
+ model_name_or_path: saves/LLaMA3-8B-Chat/full/train_2024-07-11-10-49-45_inst_llama3
10
+ output_dir: saves/LLaMA3-8B-Chat/full/eval_2024-07-11-10-49-45
11
+ per_device_eval_batch_size: 2
12
+ predict_with_generate: true
 
 
 
13
  preprocessing_num_workers: 16
14
  quantization_method: bitsandbytes
 
 
15
  stage: sft
16
+ temperature: 0.95
17
  template: llama3
18
+ top_p: 0.7