Hugging Face's logo

Tristan
/
pythia-70m_arc_easy

Text Generation
Transformers
TensorBoard
Safetensors
gpt_neox
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
Tristan commited on
Commit
7e4b855
•
1 Parent(s): c11d8a9

Training in progress, epoch 0

Browse files
Files changed (6) hide show
  1. config.json +30 -0
  2. eval_job_output.txt +42 -0
  3. logs/events.out.tfevents.1715303586.sphinx2 +3 -0
  4. model.safetensors +3 -0
  5. train_job_output.txt +115 -0
  6. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "EleutherAI/pythia-70m",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "attention_bias": true,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
+ "eos_token_id": 0,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 512,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 2048,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 2048,
18
+ "model_type": "gpt_neox",
19
+ "num_attention_heads": 8,
20
+ "num_hidden_layers": 6,
21
+ "rope_scaling": null,
22
+ "rotary_emb_base": 10000,
23
+ "rotary_pct": 0.25,
24
+ "tie_word_embeddings": false,
25
+ "torch_dtype": "float32",
26
+ "transformers_version": "4.39.3",
27
+ "use_cache": true,
28
+ "use_parallel_residual": true,
29
+ "vocab_size": 50304
30
+ }
eval_job_output.txt ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ slurm submission log: 2024-05-09 15:03:34.997660
2
+ created following sbatch script:
3
+
4
+ ###############################
5
+
6
+ #!/bin/bash
7
+
8
+ #SBATCH --account=nlp
9
+ #SBATCH --cpus-per-task=16
10
+ #SBATCH --dependency=afterok:7592327
11
+ #SBATCH --gres=gpu:1
12
+ #SBATCH --job-name=tthrush-job-3419847
13
+ #SBATCH --mem=60G
14
+ #SBATCH --nodelist=sphinx2
15
+ #SBATCH --open-mode=append
16
+ #SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy/eval_job_output.txt
17
+ #SBATCH --partition=sphinx
18
+ #SBATCH --time=14-0
19
+
20
+ # activate your desired anaconda environment
21
+ . /nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection
22
+
23
+ # cd to working directory
24
+ cd .
25
+
26
+ # launch commands
27
+ srun --unbuffered run_as_child_processes 'lm_eval --model hf --model_args pretrained=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy,revision=main,dtype=float16,trust_remote_code=True --tasks xnli_en,xnli_fr,sciq,piqa,lambada,arc_easy --device cuda --output_path /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy/perf'
28
+
29
+ ###############################
30
+
31
+ submission to slurm complete!
32
+
33
+
34
+ ###############################
35
+ slurm submission output
36
+
37
+ Submitted batch job 7592328
38
+
39
+
40
+
41
+ ###############################
42
+
logs/events.out.tfevents.1715303586.sphinx2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f658efc67807807ef486f03ff2fd5e5d1cc086948f28be1b505c517ded6a15a5
3
+ size 10961
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d46a30876d12cd0f74159979646a60198f767b341492d103e8ca59f3a294d350
3
+ size 281715176
train_job_output.txt ADDED
@@ -0,0 +1,115 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0
  0%| | 0/10682 [00:00<?, ?it/s][rank0]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
 
 
1
  0%| | 1/10682 [00:24<72:50:43, 24.55s/it]
2
  0%| | 2/10682 [00:30<40:22:35, 13.61s/it]
3
  0%| | 3/10682 [00:34<26:55:28, 9.08s/it]
4
  0%| | 4/10682 [00:37<19:47:10, 6.67s/it]
5
  0%| | 5/10682 [00:39<15:01:55, 5.07s/it]
6
  0%| | 6/10682 [00:41<11:32:34, 3.89s/it]
7
  0%| | 7/10682 [00:42<9:31:09, 3.21s/it]
8
  0%| | 8/10682 [00:44<7:38:58, 2.58s/it]
9
  0%| | 9/10682 [00:45<6:17:33, 2.12s/it]
10
  0%| | 10/10682 [00:46<5:17:31, 1.79s/it]
11
  0%| | 11/10682 [00:47<4:44:47, 1.60s/it]
12
  0%| | 12/10682 [00:48<4:02:04, 1.36s/it]
13
  0%| | 13/10682 [00:48<3:27:39, 1.17s/it]
14
  0%| | 14/10682 [00:49<3:05:04, 1.04s/it]
15
  0%| | 15/10682 [00:50<2:50:37, 1.04it/s]
16
  0%| | 16/10682 [00:51<2:39:54, 1.11it/s]
17
  0%| | 17/10682 [00:51<2:32:45, 1.16it/s]
18
  0%| | 18/10682 [00:52<2:19:43, 1.27it/s]
19
  0%| | 19/10682 [00:53<2:11:50, 1.35it/s]
20
  0%| | 20/10682 [00:53<2:06:18, 1.41it/s]
21
  0%| | 21/10682 [00:54<2:00:31, 1.47it/s]
22
  0%| | 22/10682 [00:55<1:57:33, 1.51it/s]
23
  0%| | 23/10682 [00:55<1:54:06, 1.56it/s]
24
  0%| | 24/10682 [00:56<1:48:28, 1.64it/s]
25
  0%| | 25/10682 [00:57<2:03:47, 1.43it/s]
26
 
 
27
  0%| | 25/10682 [00:57<2:03:47, 1.43it/s]
28
  0%| | 26/10682 [00:57<1:55:38, 1.54it/s]
29
  0%| | 27/10682 [00:58<1:50:06, 1.61it/s]
30
  0%| | 28/10682 [00:58<1:44:53, 1.69it/s]
31
  0%| | 29/10682 [00:59<1:41:47, 1.74it/s]
32
  0%| | 30/10682 [00:59<1:39:14, 1.79it/s]
33
  0%| | 31/10682 [01:00<1:36:57, 1.83it/s]
34
  0%| | 32/10682 [01:00<1:36:40, 1.84it/s]
35
  0%| | 33/10682 [01:01<1:35:13, 1.86it/s]
36
  0%| | 34/10682 [01:01<1:34:03, 1.89it/s]
37
  0%| | 35/10682 [01:02<1:33:14, 1.90it/s]
38
  0%| | 36/10682 [01:02<1:32:17, 1.92it/s]
39
  0%| | 37/10682 [01:03<1:31:49, 1.93it/s]
40
  0%| | 38/10682 [01:03<1:31:51, 1.93it/s]
41
  0%| | 39/10682 [01:04<1:31:50, 1.93it/s]
42
  0%| | 40/10682 [01:04<1:31:23, 1.94it/s]
43
  0%| | 41/10682 [01:05<1:31:33, 1.94it/s]
44
  0%| | 42/10682 [01:05<1:31:05, 1.95it/s]
45
  0%| | 43/10682 [01:06<1:30:43, 1.95it/s]
46
  0%| | 44/10682 [01:06<1:30:24, 1.96it/s]
47
  0%| | 45/10682 [01:07<1:30:14, 1.96it/s]
48
  0%| | 46/10682 [01:07<1:30:06, 1.97it/s]
49
  0%| | 47/10682 [01:08<1:29:58, 1.97it/s]
50
  0%| | 48/10682 [01:09<1:29:53, 1.97it/s]
51
  0%| | 49/10682 [01:09<1:29:42, 1.98it/s]
52
  0%| | 50/10682 [01:10<1:29:48, 1.97it/s]
53
  {'loss': 9.7768, 'grad_norm': 1.579921841621399, 'learning_rate': 4.677268475210477e-05, 'epoch': 0.07}
 
54
  0%| | 50/10682 [01:10<1:29:48, 1.97it/s]
55
  0%| | 51/10682 [01:10<1:29:53, 1.97it/s]
56
  0%| | 52/10682 [01:11<1:29:35, 1.98it/s]
57
  0%| | 53/10682 [01:11<1:29:25, 1.98it/s]
58
  1%| | 54/10682 [01:12<1:29:08, 1.99it/s]
59
  1%| | 55/10682 [01:12<1:28:56, 1.99it/s]
60
  1%| | 56/10682 [01:13<1:28:54, 1.99it/s]
61
  1%| | 57/10682 [01:13<1:29:03, 1.99it/s]
62
  1%| | 58/10682 [01:14<1:29:10, 1.99it/s]
63
  1%| | 59/10682 [01:14<1:29:06, 1.99it/s]
64
  1%| | 60/10682 [01:15<1:28:57, 1.99it/s]
65
  1%| | 61/10682 [01:15<1:28:53, 1.99it/s]
66
  1%| | 62/10682 [01:16<1:28:53, 1.99it/s]
67
  1%| | 63/10682 [01:16<1:28:51, 1.99it/s]
68
  1%| | 64/10682 [01:17<1:28:54, 1.99it/s]
69
  1%| | 65/10682 [01:17<1:28:42, 1.99it/s]
70
  1%| | 66/10682 [01:18<1:28:43, 1.99it/s]
71
  1%| | 67/10682 [01:18<1:28:31, 2.00it/s]
72
  1%| | 68/10682 [01:19<1:28:32, 2.00it/s]
73
  1%| | 69/10682 [01:19<1:28:31, 2.00it/s]
74
  1%| | 70/10682 [01:20<1:28:29, 2.00it/s]
75
  1%| | 71/10682 [01:20<1:28:20, 2.00it/s]
76
  1%| | 72/10682 [01:21<1:28:12, 2.00it/s]
77
  1%| | 73/10682 [01:21<1:28:18, 2.00it/s]
78
  1%| | 74/10682 [01:22<1:41:07, 1.75it/s]
79
  1%| | 75/10682 [01:22<1:37:16, 1.82it/s]
80
 
 
81
  1%| | 75/10682 [01:22<1:37:16, 1.82it/s]
82
  1%| | 76/10682 [01:23<1:34:38, 1.87it/s]
83
  1%| | 77/10682 [01:23<1:32:38, 1.91it/s]
84
  1%| | 78/10682 [01:24<1:31:11, 1.94it/s]
85
  1%| | 79/10682 [01:24<1:30:20, 1.96it/s]
86
  1%| | 80/10682 [01:25<1:29:41, 1.97it/s]
87
  1%| | 81/10682 [01:25<1:29:13, 1.98it/s]
88
  1%| | 82/10682 [01:26<1:28:50, 1.99it/s]
89
  1%| | 83/10682 [01:26<1:28:30, 2.00it/s]
90
  1%| | 84/10682 [01:27<1:28:18, 2.00it/s]
91
  1%| | 85/10682 [01:27<1:28:13, 2.00it/s]
92
  1%| | 86/10682 [01:28<1:28:07, 2.00it/s]
93
  1%| | 87/10682 [01:28<1:28:07, 2.00it/s]
94
  1%| | 88/10682 [01:29<1:27:49, 2.01it/s]
95
  1%| | 89/10682 [01:29<1:27:44, 2.01it/s]
96
  1%| | 90/10682 [01:30<1:27:37, 2.01it/s]
97
  1%| | 91/10682 [01:30<1:27:39, 2.01it/s]
98
  1%| | 92/10682 [01:31<1:27:40, 2.01it/s]
99
  1%| | 93/10682 [01:31<1:27:30, 2.02it/s]
100
  1%| | 94/10682 [01:32<1:27:35, 2.01it/s]
101
  1%| | 95/10682 [01:32<1:27:32, 2.02it/s]
102
  1%| | 96/10682 [01:33<1:27:32, 2.02it/s]
103
  1%| | 97/10682 [01:33<1:27:35, 2.01it/s]
104
  1%| | 98/10682 [01:34<1:27:28, 2.02it/s]
105
  1%| | 99/10682 [01:34<1:27:35, 2.01it/s]
106
  1%| | 100/10682 [01:35<1:27:28, 2.02it/s]{'loss': 8.2552, 'grad_norm': 1.013319969177246, 'learning_rate': 9.354536950420954e-05, 'epoch': 0.13}
107
 
 
108
  1%| | 100/10682 [01:35<1:27:28, 2.02it/s]
109
  1%| | 101/10682 [01:35<1:27:49, 2.01it/s]
110
  1%| | 102/10682 [01:36<1:27:45, 2.01it/s]
111
  1%| | 103/10682 [01:36<1:27:36, 2.01it/s]
112
  1%| | 104/10682 [01:37<1:27:28, 2.02it/s]
113
  1%| | 105/10682 [01:37<1:27:21, 2.02it/s]
114
  1%| | 106/10682 [01:38<1:27:28, 2.02it/s]
115
  1%| | 107/10682 [01:38<1:27:20, 2.02it/s]
116
  1%| | 108/10682 [01:39<1:27:21, 2.02it/s]
117
  1%| | 109/10682 [01:39<1:27:18, 2.02it/s]
118
  1%| | 110/10682 [01:40<1:27:16, 2.02it/s]
119
  1%| | 111/10682 [01:40<1:27:13, 2.02it/s]
120
  1%| | 112/10682 [01:41<1:27:16, 2.02it/s]
121
  1%| | 113/10682 [01:41<1:27:18, 2.02it/s]
122
  1%| | 114/10682 [01:42<1:27:14, 2.02it/s]
123
  1%| | 115/10682 [01:42<1:27:17, 2.02it/s]
124
  1%| | 116/10682 [01:43<1:27:10, 2.02it/s]
125
  1%| | 117/10682 [01:43<1:27:12, 2.02it/s]
126
  1%| | 118/10682 [01:44<1:27:26, 2.01it/s]
127
  1%| | 119/10682 [01:44<1:27:29, 2.01it/s]
128
  1%| | 120/10682 [01:45<1:27:24, 2.01it/s]
129
  1%| | 121/10682 [01:45<1:27:20, 2.02it/s]
130
  1%| | 122/10682 [01:46<1:27:11, 2.02it/s]
131
  1%| | 123/10682 [01:46<1:27:11, 2.02it/s]
132
  1%| | 124/10682 [01:47<1:27:05, 2.02it/s]
133
  1%| | 125/10682 [01:47<1:27:06, 2.02it/s]
134
  {'loss': 7.5756, 'grad_norm': 0.5084755420684814, 'learning_rate': 0.00011693171188026193, 'epoch': 0.16}
 
135
  1%| | 125/10682 [01:47<1:27:06, 2.02it/s]
136
  1%| | 126/10682 [01:48<1:27:09, 2.02it/s]
137
  1%| | 127/10682 [01:48<1:27:06, 2.02it/s]
138
  1%| | 128/10682 [01:49<1:27:01, 2.02it/s]
139
  1%| | 129/10682 [01:49<1:27:01, 2.02it/s]
140
  1%| | 130/10682 [01:50<1:27:06, 2.02it/s]
141
  1%| | 131/10682 [01:50<1:27:10, 2.02it/s]
142
  1%| | 132/10682 [01:51<1:27:03, 2.02it/s]
143
  1%| | 133/10682 [01:51<1:27:05, 2.02it/s]
144
  1%|▏ | 134/10682 [01:52<1:27:03, 2.02it/s]
145
  1%|▏ | 135/10682 [01:52<1:27:10, 2.02it/s]
146
  1%|▏ | 136/10682 [01:53<1:27:10, 2.02it/s]
147
  1%|▏ | 137/10682 [01:53<1:27:17, 2.01it/s]
148
  1%|▏ | 138/10682 [01:54<1:27:07, 2.02it/s]
149
  1%|▏ | 139/10682 [01:54<1:27:05, 2.02it/s]
150
  1%|▏ | 140/10682 [01:55<1:27:03, 2.02it/s]
151
  1%|▏ | 141/10682 [01:55<1:27:07, 2.02it/s]
152
  1%|▏ | 142/10682 [01:56<1:27:08, 2.02it/s]
153
  1%|▏ | 143/10682 [01:56<1:27:04, 2.02it/s]
154
  1%|▏ | 144/10682 [01:57<1:27:06, 2.02it/s]
155
  1%|▏ | 145/10682 [01:57<1:26:59, 2.02it/s]
156
  1%|▏ | 146/10682 [01:58<1:26:59, 2.02it/s]
157
  1%|▏ | 147/10682 [01:58<1:26:54, 2.02it/s]
158
  1%|▏ | 148/10682 [01:59<1:26:55, 2.02it/s]
159
  1%|▏ | 149/10682 [01:59<1:26:48, 2.02it/s]
160
  1%|▏ | 150/10682 [02:00<1:26:54, 2.02it/s]{'loss': 7.1387, 'grad_norm': 0.3660012483596802, 'learning_rate': 0.0001403180542563143, 'epoch': 0.2}
161
 
 
162
  1%|▏ | 150/10682 [02:00<1:26:54, 2.02it/s]
163
  1%|▏ | 151/10682 [02:00<1:26:59, 2.02it/s]
164
  1%|▏ | 152/10682 [02:00<1:26:56, 2.02it/s]
165
  1%|▏ | 153/10682 [02:01<1:26:56, 2.02it/s]
166
  1%|▏ | 154/10682 [02:01<1:26:52, 2.02it/s]
167
  1%|▏ | 155/10682 [02:02<1:26:50, 2.02it/s]
168
  1%|▏ | 156/10682 [02:02<1:26:49, 2.02it/s]
169
  1%|▏ | 157/10682 [02:03<1:26:57, 2.02it/s]
170
  1%|▏ | 158/10682 [02:03<1:26:52, 2.02it/s]
171
  1%|▏ | 159/10682 [02:04<1:26:54, 2.02it/s]
172
  1%|▏ | 160/10682 [02:04<1:26:47, 2.02it/s]
173
  2%|▏ | 161/10682 [02:05<1:26:49, 2.02it/s]
174
  2%|▏ | 162/10682 [02:05<1:26:47, 2.02it/s]
175
  2%|▏ | 163/10682 [02:06<1:26:49, 2.02it/s]
176
  2%|▏ | 164/10682 [02:06<1:26:40, 2.02it/s]
177
  2%|▏ | 165/10682 [02:07<1:26:47, 2.02it/s]
178
  2%|▏ | 166/10682 [02:07<1:26:40, 2.02it/s]
179
  2%|▏ | 167/10682 [02:08<1:26:43, 2.02it/s]
180
  2%|▏ | 168/10682 [02:08<1:26:40, 2.02it/s]
181
  2%|▏ | 169/10682 [02:09<1:26:43, 2.02it/s]
182
  2%|▏ | 170/10682 [02:09<1:26:39, 2.02it/s]
183
  2%|▏ | 171/10682 [02:10<1:26:42, 2.02it/s]
184
  2%|▏ | 172/10682 [02:10<1:26:39, 2.02it/s]
185
  2%|▏ | 173/10682 [02:11<1:26:40, 2.02it/s]
186
  2%|▏ | 174/10682 [02:11<1:26:36, 2.02it/s]
187
  2%|▏ | 175/10682 [02:12<1:26:39, 2.02it/s]{'loss': 6.8255, 'grad_norm': 0.5133531093597412, 'learning_rate': 0.00016370439663236668, 'epoch': 0.23}
188
 
 
189
  2%|▏ | 175/10682 [02:12<1:26:39, 2.02it/s]
190
  2%|▏ | 176/10682 [02:12<1:26:46, 2.02it/s]
191
  2%|▏ | 177/10682 [02:13<1:26:44, 2.02it/s]
192
  2%|▏ | 178/10682 [02:13<1:26:43, 2.02it/s]
193
  2%|▏ | 179/10682 [02:14<1:26:39, 2.02it/s]
194
  2%|▏ | 180/10682 [02:14<1:26:40, 2.02it/s]
195
  2%|▏ | 181/10682 [02:15<1:26:38, 2.02it/s]
196
  2%|▏ | 182/10682 [02:15<1:26:38, 2.02it/s]
197
  2%|▏ | 183/10682 [02:16<1:26:35, 2.02it/s]
198
  2%|▏ | 184/10682 [02:16<1:26:41, 2.02it/s]
199
  2%|▏ | 185/10682 [02:17<1:26:41, 2.02it/s]
200
  2%|▏ | 186/10682 [02:17<1:26:47, 2.02it/s]
201
  2%|▏ | 187/10682 [02:18<1:26:48, 2.01it/s]
202
  2%|▏ | 188/10682 [02:18<1:26:48, 2.01it/s]
203
  2%|▏ | 189/10682 [02:19<1:26:48, 2.01it/s]
204
  2%|▏ | 190/10682 [02:19<1:26:43, 2.02it/s]
205
  2%|▏ | 191/10682 [02:20<1:26:41, 2.02it/s]
206
  2%|▏ | 192/10682 [02:20<1:26:34, 2.02it/s]
207
  2%|▏ | 193/10682 [02:21<1:26:31, 2.02it/s]
208
  2%|▏ | 194/10682 [02:21<1:26:29, 2.02it/s]
209
  2%|▏ | 195/10682 [02:22<1:26:33, 2.02it/s]
210
  2%|▏ | 196/10682 [02:22<1:26:26, 2.02it/s]
211
  2%|▏ | 197/10682 [02:23<1:26:26, 2.02it/s]
212
  2%|▏ | 198/10682 [02:23<1:26:23, 2.02it/s]
213
  2%|▏ | 199/10682 [02:24<1:26:28, 2.02it/s]
214
  2%|▏ | 200/10682 [02:24<1:26:25, 2.02it/s]
215
  {'loss': 6.575, 'grad_norm': 0.5841809511184692, 'learning_rate': 0.00018709073900841907, 'epoch': 0.26}
 
216
  2%|▏ | 200/10682 [02:24<1:26:25, 2.02it/s]
217
  2%|▏ | 201/10682 [02:25<1:26:32, 2.02it/s]
218
  2%|▏ | 202/10682 [02:25<1:26:25, 2.02it/s]
219
  2%|▏ | 203/10682 [02:26<1:26:29, 2.02it/s]
220
  2%|▏ | 204/10682 [02:26<1:26:28, 2.02it/s]
221
  2%|▏ | 205/10682 [02:27<1:26:31, 2.02it/s]
222
  2%|▏ | 206/10682 [02:27<1:26:26, 2.02it/s]
223
  2%|▏ | 207/10682 [02:28<1:26:29, 2.02it/s]
224
  2%|▏ | 208/10682 [02:28<1:26:24, 2.02it/s]
225
  2%|▏ | 209/10682 [02:29<1:26:27, 2.02it/s]
226
  2%|▏ | 210/10682 [02:29<1:26:20, 2.02it/s]
227
  2%|▏ | 211/10682 [02:30<1:26:22, 2.02it/s]
228
  2%|▏ | 212/10682 [02:30<1:26:19, 2.02it/s]
229
  2%|▏ | 213/10682 [02:31<1:26:20, 2.02it/s]
230
  2%|▏ | 214/10682 [02:31<1:26:19, 2.02it/s]
231
  2%|▏ | 215/10682 [02:32<1:26:26, 2.02it/s]
232
  2%|▏ | 216/10682 [02:32<1:26:21, 2.02it/s]
233
  2%|▏ | 217/10682 [02:33<1:26:21, 2.02it/s]
234
  2%|▏ | 218/10682 [02:33<1:26:17, 2.02it/s]
235
  2%|▏ | 219/10682 [02:34<1:26:22, 2.02it/s]
236
  2%|▏ | 220/10682 [02:34<1:26:15, 2.02it/s]
237
  2%|▏ | 221/10682 [02:35<1:26:18, 2.02it/s]
238
  2%|▏ | 222/10682 [02:35<1:26:15, 2.02it/s]
239
  2%|▏ | 223/10682 [02:36<1:26:21, 2.02it/s]
240
  2%|▏ | 224/10682 [02:36<1:26:22, 2.02it/s]
241
  2%|▏ | 225/10682 [02:37<1:26:18, 2.02it/s]{'loss': 6.3833, 'grad_norm': 0.6055994629859924, 'learning_rate': 0.00021047708138447147, 'epoch': 0.29}
 
242
 
243
  2%|▏ | 225/10682 [02:37<1:26:18, 2.02it/s]
244
  2%|▏ | 226/10682 [02:37<1:26:30, 2.01it/s]
245
  2%|▏ | 227/10682 [02:38<1:26:25, 2.02it/s]
246
  2%|▏ | 228/10682 [02:38<1:26:22, 2.02it/s]
247
  2%|▏ | 229/10682 [02:39<1:26:15, 2.02it/s]
248
  2%|▏ | 230/10682 [02:39<1:26:13, 2.02it/s]
249
  2%|▏ | 231/10682 [02:40<1:26:08, 2.02it/s]
250
  2%|▏ | 232/10682 [02:40<1:26:18, 2.02it/s]
251
  2%|▏ | 233/10682 [02:41<1:26:12, 2.02it/s]
252
  2%|▏ | 234/10682 [02:41<1:26:14, 2.02it/s]
253
  2%|▏ | 235/10682 [02:42<1:26:09, 2.02it/s]
254
  2%|▏ | 236/10682 [02:42<1:26:12, 2.02it/s]
255
  2%|▏ | 237/10682 [02:43<1:26:08, 2.02it/s]
256
  2%|▏ | 238/10682 [02:43<1:26:07, 2.02it/s]
257
  2%|▏ | 239/10682 [02:44<1:26:07, 2.02it/s]
258
  2%|▏ | 240/10682 [02:44<1:26:08, 2.02it/s]
259
  2%|▏ | 241/10682 [02:45<1:26:10, 2.02it/s]
260
  2%|▏ | 242/10682 [02:45<1:26:06, 2.02it/s]
261
  2%|▏ | 243/10682 [02:46<1:26:03, 2.02it/s]
262
  2%|▏ | 244/10682 [02:46<1:26:05, 2.02it/s]
263
  2%|▏ | 245/10682 [02:47<1:26:01, 2.02it/s]
264
  2%|▏ | 246/10682 [02:47<1:26:06, 2.02it/s]
265
  2%|▏ | 247/10682 [02:48<1:25:59, 2.02it/s]
266
  2%|▏ | 248/10682 [02:48<1:26:03, 2.02it/s]
267
  2%|▏ | 249/10682 [02:49<1:25:58, 2.02it/s]
268
  2%|▏ | 250/10682 [02:49<1:26:01, 2.02it/s]
269
  {'loss': 6.2106, 'grad_norm': 0.6558924317359924, 'learning_rate': 0.00023386342376052386, 'epoch': 0.33}
 
270
  2%|▏ | 250/10682 [02:49<1:26:01, 2.02it/s]
271
  2%|▏ | 251/10682 [02:50<1:26:05, 2.02it/s]
272
  2%|▏ | 252/10682 [02:50<1:26:04, 2.02it/s]
273
  2%|▏ | 253/10682 [02:50<1:25:59, 2.02it/s]
274
  2%|▏ | 254/10682 [02:51<1:26:01, 2.02it/s]
275
  2%|▏ | 255/10682 [02:51<1:26:08, 2.02it/s]
276
  2%|▏ | 256/10682 [02:52<1:26:12, 2.02it/s]
277
  2%|▏ | 257/10682 [02:52<1:26:06, 2.02it/s]
278
  2%|▏ | 258/10682 [02:53<1:26:06, 2.02it/s]
279
  2%|▏ | 259/10682 [02:53<1:26:00, 2.02it/s]
280
  2%|▏ | 260/10682 [02:54<1:26:02, 2.02it/s]
281
  2%|▏ | 261/10682 [02:54<1:25:58, 2.02it/s]
282
  2%|▏ | 262/10682 [02:55<1:26:05, 2.02it/s]
283
  2%|▏ | 263/10682 [02:55<1:25:59, 2.02it/s]
284
  2%|▏ | 264/10682 [02:56<1:25:53, 2.02it/s]
285
  2%|▏ | 265/10682 [02:56<1:25:55, 2.02it/s]
286
  2%|▏ | 266/10682 [02:57<1:25:50, 2.02it/s]
287
  2%|▏ | 267/10682 [02:57<1:25:54, 2.02it/s]
288
  3%|▎ | 268/10682 [02:58<1:25:53, 2.02it/s]
289
  3%|▎ | 269/10682 [02:58<1:25:55, 2.02it/s]
290
  3%|▎ | 270/10682 [02:59<1:25:54, 2.02it/s]
291
  3%|▎ | 271/10682 [02:59<1:25:59, 2.02it/s]
292
  3%|▎ | 272/10682 [03:00<1:25:56, 2.02it/s]
293
  3%|▎ | 273/10682 [03:00<1:25:55, 2.02it/s]
294
  3%|▎ | 274/10682 [03:01<1:25:55, 2.02it/s]
295
  3%|▎ | 275/10682 [03:01<1:25:50, 2.02it/s]
296
  {'loss': 6.0743, 'grad_norm': 0.5830900073051453, 'learning_rate': 0.00025724976613657625, 'epoch': 0.36}
 
297
  3%|▎ | 275/10682 [03:01<1:25:50, 2.02it/s]
298
  3%|▎ | 276/10682 [03:02<1:25:52, 2.02it/s]
299
  3%|▎ | 277/10682 [03:02<1:25:54, 2.02it/s]
300
  3%|▎ | 278/10682 [03:03<1:25:51, 2.02it/s]
301
  3%|▎ | 279/10682 [03:03<1:25:50, 2.02it/s]
302
  3%|▎ | 280/10682 [03:04<1:25:44, 2.02it/s]
303
  3%|▎ | 281/10682 [03:04<1:25:47, 2.02it/s]
304
  3%|▎ | 282/10682 [03:05<1:25:45, 2.02it/s]
305
  3%|▎ | 283/10682 [03:05<1:25:46, 2.02it/s]
306
  3%|▎ | 284/10682 [03:06<1:25:44, 2.02it/s]
307
  3%|▎ | 285/10682 [03:06<1:25:45, 2.02it/s]
308
  3%|▎ | 286/10682 [03:07<1:25:41, 2.02it/s]
309
  3%|▎ | 287/10682 [03:07<1:25:42, 2.02it/s]
310
  3%|▎ | 288/10682 [03:08<1:25:37, 2.02it/s]
311
  3%|▎ | 289/10682 [03:08<1:25:40, 2.02it/s]
312
  3%|▎ | 290/10682 [03:09<1:25:36, 2.02it/s]
313
  3%|▎ | 291/10682 [03:09<1:25:39, 2.02it/s]
314
  3%|▎ | 292/10682 [03:10<1:25:39, 2.02it/s]
315
  3%|▎ | 293/10682 [03:10<1:25:41, 2.02it/s]
316
  3%|▎ | 294/10682 [03:11<1:25:39, 2.02it/s]
317
  3%|▎ | 295/10682 [03:11<1:25:39, 2.02it/s]
318
  3%|▎ | 296/10682 [03:12<1:25:43, 2.02it/s]
319
  3%|▎ | 297/10682 [03:12<1:25:43, 2.02it/s]
320
  3%|▎ | 298/10682 [03:13<1:25:41, 2.02it/s]
321
  3%|▎ | 299/10682 [03:13<1:25:35, 2.02it/s]
322
  3%|▎ | 300/10682 [03:14<1:25:40, 2.02it/s]
323
  {'loss': 5.9526, 'grad_norm': 0.7878416180610657, 'learning_rate': 0.0002806361085126286, 'epoch': 0.39}
 
324
  3%|▎ | 300/10682 [03:14<1:25:40, 2.02it/s]
325
  3%|▎ | 301/10682 [03:14<1:25:41, 2.02it/s]
326
  3%|▎ | 302/10682 [03:15<1:25:44, 2.02it/s]
327
  3%|▎ | 303/10682 [03:15<1:25:36, 2.02it/s]
328
  3%|▎ | 304/10682 [03:16<1:25:38, 2.02it/s]
329
  3%|▎ | 305/10682 [03:16<1:25:34, 2.02it/s]
330
  3%|▎ | 306/10682 [03:17<1:25:36, 2.02it/s]
331
  3%|▎ | 307/10682 [03:17<1:25:31, 2.02it/s]
332
  3%|▎ | 308/10682 [03:18<1:25:32, 2.02it/s]
333
  3%|▎ | 309/10682 [03:18<1:25:30, 2.02it/s]
334
  3%|▎ | 310/10682 [03:19<1:25:34, 2.02it/s]
335
  3%|▎ | 311/10682 [03:19<1:25:30, 2.02it/s]
336
  3%|▎ | 312/10682 [03:20<1:25:32, 2.02it/s]
337
  3%|▎ | 313/10682 [03:20<1:25:30, 2.02it/s]
338
  3%|▎ | 314/10682 [03:21<1:25:30, 2.02it/s]
339
  3%|▎ | 315/10682 [03:21<1:25:21, 2.02it/s]
340
  3%|▎ | 316/10682 [03:22<1:25:30, 2.02it/s]
341
  3%|▎ | 317/10682 [03:22<1:25:30, 2.02it/s]
342
  3%|▎ | 318/10682 [03:23<1:25:27, 2.02it/s]
343
  3%|▎ | 319/10682 [03:23<1:25:21, 2.02it/s]
344
  3%|▎ | 320/10682 [03:24<1:25:24, 2.02it/s]
345
  3%|▎ | 321/10682 [03:24<1:25:21, 2.02it/s]
346
  3%|▎ | 322/10682 [03:25<1:25:26, 2.02it/s]
347
  3%|▎ | 323/10682 [03:25<1:25:38, 2.02it/s]
348
  3%|▎ | 324/10682 [03:26<1:25:45, 2.01it/s]
349
  3%|▎ | 325/10682 [03:26<1:25:41, 2.01it/s]
350
  {'loss': 5.844, 'grad_norm': 0.8623385429382324, 'learning_rate': 0.00030402245088868103, 'epoch': 0.43}
 
351
  3%|▎ | 325/10682 [03:26<1:25:41, 2.01it/s]
352
  3%|▎ | 326/10682 [03:27<1:25:42, 2.01it/s]
353
  3%|▎ | 327/10682 [03:27<1:25:37, 2.02it/s]
354
  3%|▎ | 328/10682 [03:28<1:25:31, 2.02it/s]
355
  3%|▎ | 329/10682 [03:28<1:25:30, 2.02it/s]
356
  3%|▎ | 330/10682 [03:29<1:25:23, 2.02it/s]
357
  3%|▎ | 331/10682 [03:29<1:25:26, 2.02it/s]
358
  3%|▎ | 332/10682 [03:30<1:25:18, 2.02it/s]
359
  3%|▎ | 333/10682 [03:30<1:25:24, 2.02it/s]
360
  3%|▎ | 334/10682 [03:31<1:25:18, 2.02it/s]
361
  3%|▎ | 335/10682 [03:31<1:25:23, 2.02it/s]
362
  3%|▎ | 336/10682 [03:32<1:25:18, 2.02it/s]
363
  3%|▎ | 337/10682 [03:32<1:25:18, 2.02it/s]
364
  3%|▎ | 338/10682 [03:33<1:25:15, 2.02it/s]
365
  3%|▎ | 339/10682 [03:33<1:25:21, 2.02it/s]
366
  3%|▎ | 340/10682 [03:34<1:25:21, 2.02it/s]
367
  3%|▎ | 341/10682 [03:34<1:25:23, 2.02it/s]
368
  3%|▎ | 342/10682 [03:35<1:25:18, 2.02it/s]
369
  3%|▎ | 343/10682 [03:35<1:25:20, 2.02it/s]
370
  3%|▎ | 344/10682 [03:36<1:25:16, 2.02it/s]
371
  3%|▎ | 345/10682 [03:36<1:25:19, 2.02it/s]
372
  3%|▎ | 346/10682 [03:37<1:25:15, 2.02it/s]
373
  3%|▎ | 347/10682 [03:37<1:25:15, 2.02it/s]
374
  3%|▎ | 348/10682 [03:38<1:25:13, 2.02it/s]
375
  3%|▎ | 349/10682 [03:38<1:25:11, 2.02it/s]
376
  3%|▎ | 350/10682 [03:39<1:25:09, 2.02it/s]
377
 
 
378
  3%|▎ | 350/10682 [03:39<1:25:09, 2.02it/s]
379
  3%|▎ | 351/10682 [03:39<1:25:19, 2.02it/s]
380
  3%|▎ | 352/10682 [03:40<1:25:16, 2.02it/s]
381
  3%|▎ | 353/10682 [03:40<1:25:16, 2.02it/s]
382
  3%|▎ | 354/10682 [03:40<1:25:10, 2.02it/s]
383
  3%|▎ | 355/10682 [03:41<1:25:13, 2.02it/s]
384
  3%|▎ | 356/10682 [03:41<1:25:07, 2.02it/s]
385
  3%|▎ | 357/10682 [03:42<1:25:10, 2.02it/s]
386
  3%|▎ | 358/10682 [03:42<1:25:05, 2.02it/s]
387
  3%|▎ | 359/10682 [03:43<1:25:08, 2.02it/s]
388
  3%|▎ | 360/10682 [03:43<1:25:02, 2.02it/s]
389
  3%|▎ | 361/10682 [03:44<1:25:06, 2.02it/s]
390
  3%|▎ | 362/10682 [03:44<1:25:04, 2.02it/s]
391
  3%|▎ | 363/10682 [03:45<1:25:12, 2.02it/s]
392
  3%|▎ | 364/10682 [03:45<1:25:15, 2.02it/s]
393
  3%|▎ | 365/10682 [03:46<1:25:12, 2.02it/s]
394
  3%|▎ | 366/10682 [03:46<1:25:14, 2.02it/s]
395
  3%|▎ | 367/10682 [03:47<1:25:09, 2.02it/s]
396
  3%|▎ | 368/10682 [03:47<1:25:13, 2.02it/s]
397
  3%|▎ | 369/10682 [03:48<1:25:08, 2.02it/s]
398
  3%|▎ | 370/10682 [03:48<1:25:11, 2.02it/s]
399
  3%|▎ | 371/10682 [03:49<1:25:08, 2.02it/s]
400
  3%|▎ | 372/10682 [03:49<1:25:13, 2.02it/s]
401
  3%|▎ | 373/10682 [03:50<1:25:09, 2.02it/s]
402
  4%|▎ | 374/10682 [03:50<1:25:07, 2.02it/s]
403
  4%|▎ | 375/10682 [03:51<1:25:00, 2.02it/s]
404
  {'loss': 5.6455, 'grad_norm': 0.7050163745880127, 'learning_rate': 0.0003507951356407858, 'epoch': 0.49}
 
405
  4%|▎ | 375/10682 [03:51<1:25:00, 2.02it/s]
406
  4%|▎ | 376/10682 [03:51<1:25:03, 2.02it/s]
407
  4%|▎ | 377/10682 [03:52<1:24:59, 2.02it/s]
408
  4%|▎ | 378/10682 [03:52<1:25:00, 2.02it/s]
409
  4%|▎ | 379/10682 [03:53<1:24:53, 2.02it/s]
410
  4%|▎ | 380/10682 [03:53<1:24:56, 2.02it/s]
411
  4%|▎ | 381/10682 [03:54<1:24:56, 2.02it/s]
412
  4%|▎ | 382/10682 [03:54<1:24:57, 2.02it/s]
413
  4%|▎ | 383/10682 [03:55<1:24:55, 2.02it/s]
414
  4%|▎ | 384/10682 [03:55<1:24:57, 2.02it/s]
415
  4%|▎ | 385/10682 [03:56<1:24:59, 2.02it/s]
416
  4%|▎ | 386/10682 [03:56<1:25:00, 2.02it/s]
417
  4%|▎ | 387/10682 [03:57<1:24:56, 2.02it/s]
418
  4%|▎ | 388/10682 [03:57<1:24:59, 2.02it/s]
419
  4%|▎ | 389/10682 [03:58<1:24:53, 2.02it/s]
420
  4%|▎ | 390/10682 [03:58<1:24:59, 2.02it/s]
421
  4%|▎ | 391/10682 [03:59<1:24:52, 2.02it/s]
422
  4%|▎ | 392/10682 [03:59<1:24:52, 2.02it/s]
423
  4%|▎ | 393/10682 [04:00<1:24:51, 2.02it/s]
424
  4%|▎ | 394/10682 [04:00<1:25:10, 2.01it/s]
425
  4%|▎ | 395/10682 [04:01<1:24:59, 2.02it/s]
426
  4%|▎ | 396/10682 [04:01<1:25:00, 2.02it/s]
427
  4%|▎ | 397/10682 [04:02<1:24:55, 2.02it/s]
428
  4%|▎ | 398/10682 [04:02<1:25:06, 2.01it/s]
429
  4%|▎ | 399/10682 [04:03<1:25:01, 2.02it/s]
430
  4%|▎ | 400/10682 [04:03<1:24:58, 2.02it/s]
431
  {'loss': 5.5555, 'grad_norm': 0.6004089713096619, 'learning_rate': 0.00037418147801683815, 'epoch': 0.52}
 
432
  4%|▎ | 400/10682 [04:03<1:24:58, 2.02it/s]
433
  4%|▍ | 401/10682 [04:04<1:25:01, 2.02it/s]
434
  4%|▍ | 402/10682 [04:04<1:24:57, 2.02it/s]
435
  4%|▍ | 403/10682 [04:05<1:24:53, 2.02it/s]
436
  4%|▍ | 404/10682 [04:05<1:24:47, 2.02it/s]
437
  4%|▍ | 405/10682 [04:06<1:24:47, 2.02it/s]
438
  4%|▍ | 406/10682 [04:06<1:24:44, 2.02it/s]
439
  4%|▍ | 407/10682 [04:07<1:24:44, 2.02it/s]
440
  4%|▍ | 408/10682 [04:07<1:24:42, 2.02it/s]
441
  4%|▍ | 409/10682 [04:08<1:24:45, 2.02it/s]
442
  4%|▍ | 410/10682 [04:08<1:24:38, 2.02it/s]
443
  4%|▍ | 411/10682 [04:09<1:24:40, 2.02it/s]
444
  4%|▍ | 412/10682 [04:09<1:24:38, 2.02it/s]
445
  4%|▍ | 413/10682 [04:10<1:24:42, 2.02it/s]
446
  4%|▍ | 414/10682 [04:10<1:24:35, 2.02it/s]
447
  4%|▍ | 415/10682 [04:11<1:24:39, 2.02it/s]
448
  4%|▍ | 416/10682 [04:11<1:24:34, 2.02it/s]
449
  4%|▍ | 417/10682 [04:12<1:24:38, 2.02it/s]
450
  4%|▍ | 418/10682 [04:12<1:24:34, 2.02it/s]
451
  4%|▍ | 419/10682 [04:13<1:24:38, 2.02it/s]
452
  4%|▍ | 420/10682 [04:13<1:24:37, 2.02it/s]
453
  4%|▍ | 421/10682 [04:14<1:24:44, 2.02it/s]
454
  4%|▍ | 422/10682 [04:14<1:24:44, 2.02it/s]
455
  4%|▍ | 423/10682 [04:15<1:24:39, 2.02it/s]
456
  4%|▍ | 424/10682 [04:15<1:24:39, 2.02it/s]
457
  4%|▍ | 425/10682 [04:16<1:24:37, 2.02it/s]
458
 
 
459
  4%|▍ | 425/10682 [04:16<1:24:37, 2.02it/s]
460
  4%|▍ | 426/10682 [04:16<1:24:48, 2.02it/s]
461
  4%|▍ | 427/10682 [04:17<1:24:37, 2.02it/s]
462
  4%|▍ | 428/10682 [04:17<1:24:40, 2.02it/s]
463
  4%|▍ | 429/10682 [04:18<1:24:35, 2.02it/s]
464
  4%|▍ | 430/10682 [04:18<1:24:36, 2.02it/s]
465
  4%|▍ | 431/10682 [04:19<1:24:33, 2.02it/s]
466
  4%|▍ | 432/10682 [04:19<1:24:38, 2.02it/s]
467
  4%|▍ | 433/10682 [04:20<1:24:44, 2.02it/s]
468
  4%|▍ | 434/10682 [04:20<1:24:32, 2.02it/s]
469
  4%|▍ | 435/10682 [04:21<1:24:30, 2.02it/s]
470
  4%|▍ | 436/10682 [04:21<1:24:30, 2.02it/s]
471
  4%|▍ | 437/10682 [04:22<1:24:30, 2.02it/s]
472
  4%|▍ | 438/10682 [04:22<1:24:27, 2.02it/s]
473
  4%|▍ | 439/10682 [04:23<1:24:30, 2.02it/s]
474
  4%|▍ | 440/10682 [04:23<1:24:27, 2.02it/s]
475
  4%|▍ | 441/10682 [04:24<1:24:24, 2.02it/s]
476
  4%|▍ | 442/10682 [04:24<1:24:25, 2.02it/s]
477
  4%|▍ | 443/10682 [04:25<1:24:27, 2.02it/s]
478
  4%|▍ | 444/10682 [04:25<1:24:27, 2.02it/s]
479
  4%|▍ | 445/10682 [04:26<1:24:29, 2.02it/s]
480
  4%|▍ | 446/10682 [04:26<1:24:22, 2.02it/s]
481
  4%|▍ | 447/10682 [04:27<1:24:26, 2.02it/s]
482
  4%|▍ | 448/10682 [04:27<1:24:24, 2.02it/s]
483
  4%|▍ | 449/10682 [04:28<1:24:26, 2.02it/s]
484
  4%|▍ | 450/10682 [04:28<1:24:21, 2.02it/s]
485
  {'loss': 5.399, 'grad_norm': 0.6876851916313171, 'learning_rate': 0.00042095416276894293, 'epoch': 0.59}
 
486
  4%|▍ | 450/10682 [04:28<1:24:21, 2.02it/s]
487
  4%|▍ | 451/10682 [04:29<1:24:29, 2.02it/s]
488
  4%|▍ | 452/10682 [04:29<1:24:23, 2.02it/s]
489
  4%|▍ | 453/10682 [04:30<1:24:24, 2.02it/s]
490
  4%|▍ | 454/10682 [04:30<1:24:18, 2.02it/s]
491
  4%|▍ | 455/10682 [04:31<1:24:24, 2.02it/s]
492
  4%|▍ | 456/10682 [04:31<1:24:24, 2.02it/s]
493
  4%|▍ | 457/10682 [04:31<1:24:29, 2.02it/s]
494
  4%|▍ | 458/10682 [04:32<1:24:24, 2.02it/s]
495
  4%|▍ | 459/10682 [04:32<1:24:21, 2.02it/s]
496
  4%|▍ | 460/10682 [04:33<1:24:16, 2.02it/s]
497
  4%|▍ | 461/10682 [04:33<1:24:18, 2.02it/s]
498
  4%|▍ | 462/10682 [04:34<1:24:10, 2.02it/s]
499
  4%|▍ | 463/10682 [04:34<1:24:10, 2.02it/s]
500
  4%|▍ | 464/10682 [04:35<1:24:10, 2.02it/s]
501
  4%|▍ | 465/10682 [04:35<1:24:19, 2.02it/s]
502
  4%|▍ | 466/10682 [04:36<1:24:17, 2.02it/s]
503
  4%|▍ | 467/10682 [04:36<1:24:17, 2.02it/s]
504
  4%|▍ | 468/10682 [04:37<1:24:17, 2.02it/s]
505
  4%|▍ | 469/10682 [04:37<1:24:19, 2.02it/s]
506
  4%|▍ | 470/10682 [04:38<1:24:10, 2.02it/s]
507
  4%|▍ | 471/10682 [04:38<1:24:15, 2.02it/s]
508
  4%|▍ | 472/10682 [04:39<1:24:06, 2.02it/s]
509
  4%|▍ | 473/10682 [04:39<1:24:09, 2.02it/s]
510
  4%|▍ | 474/10682 [04:40<1:24:08, 2.02it/s]
511
  4%|▍ | 475/10682 [04:40<1:24:13, 2.02it/s]{'loss': 5.327, 'grad_norm': 0.6345270872116089, 'learning_rate': 0.0004443405051449954, 'epoch': 0.62}
512
 
 
513
  4%|▍ | 475/10682 [04:40<1:24:13, 2.02it/s]
514
  4%|▍ | 476/10682 [04:41<1:24:18, 2.02it/s]
515
  4%|▍ | 477/10682 [04:41<1:24:12, 2.02it/s]
516
  4%|▍ | 478/10682 [04:42<1:24:09, 2.02it/s]
517
  4%|▍ | 479/10682 [04:42<1:24:14, 2.02it/s]
518
  4%|▍ | 480/10682 [04:43<1:24:14, 2.02it/s]
519
  5%|▍ | 481/10682 [04:43<1:24:13, 2.02it/s]
520
  5%|▍ | 482/10682 [04:44<1:24:08, 2.02it/s]
521
  5%|▍ | 483/10682 [04:44<1:24:09, 2.02it/s]
522
  5%|▍ | 484/10682 [04:45<1:24:05, 2.02it/s]
523
  5%|▍ | 485/10682 [04:45<1:24:07, 2.02it/s]
524
  5%|▍ | 486/10682 [04:46<1:24:07, 2.02it/s]
525
  5%|▍ | 487/10682 [04:46<1:24:05, 2.02it/s]
526
  5%|▍ | 488/10682 [04:47<1:24:06, 2.02it/s]
527
  5%|▍ | 489/10682 [04:47<1:24:09, 2.02it/s]
528
  5%|▍ | 490/10682 [04:48<1:24:04, 2.02it/s]
529
  5%|▍ | 491/10682 [04:48<1:24:06, 2.02it/s]
530
  5%|▍ | 492/10682 [04:49<1:24:04, 2.02it/s]
531
  5%|▍ | 493/10682 [04:49<1:24:05, 2.02it/s]
532
  5%|▍ | 494/10682 [04:50<1:24:00, 2.02it/s]
533
  5%|▍ | 495/10682 [04:50<1:24:01, 2.02it/s]
534
  5%|▍ | 496/10682 [04:51<1:23:57, 2.02it/s]
535
  5%|▍ | 497/10682 [04:51<1:24:01, 2.02it/s]
536
  5%|▍ | 498/10682 [04:52<1:23:59, 2.02it/s]
537
  5%|▍ | 499/10682 [04:52<1:23:56, 2.02it/s]
538
  5%|▍ | 500/10682 [04:53<1:23:56, 2.02it/s]
539
 
 
540
  5%|▍ | 500/10682 [04:53<1:23:56, 2.02it/s]
541
  5%|▍ | 501/10682 [04:53<1:24:03, 2.02it/s]
542
  5%|▍ | 502/10682 [04:54<1:24:03, 2.02it/s]
543
  5%|▍ | 503/10682 [04:54<1:24:03, 2.02it/s]
544
  5%|▍ | 504/10682 [04:55<1:24:02, 2.02it/s]
545
  5%|▍ | 505/10682 [04:55<1:24:01, 2.02it/s]
546
  5%|▍ | 506/10682 [04:56<1:24:08, 2.02it/s]
547
  5%|▍ | 507/10682 [04:56<1:24:01, 2.02it/s]
548
  5%|▍ | 508/10682 [04:57<1:24:02, 2.02it/s]
549
  5%|▍ | 509/10682 [04:57<1:23:56, 2.02it/s]
550
  5%|▍ | 510/10682 [04:58<1:24:00, 2.02it/s]
551
  5%|▍ | 511/10682 [04:58<1:23:55, 2.02it/s]
552
  5%|▍ | 512/10682 [04:59<1:23:58, 2.02it/s]
553
  5%|▍ | 513/10682 [04:59<1:23:53, 2.02it/s]
554
  5%|▍ | 514/10682 [05:00<1:24:01, 2.02it/s]
555
  5%|▍ | 515/10682 [05:00<1:23:56, 2.02it/s]
556
  5%|▍ | 516/10682 [05:01<1:23:54, 2.02it/s]
557
  5%|▍ | 517/10682 [05:01<1:23:53, 2.02it/s]
558
  5%|▍ | 518/10682 [05:02<1:23:51, 2.02it/s]
559
  5%|▍ | 519/10682 [05:02<1:23:51, 2.02it/s]
560
  5%|▍ | 520/10682 [05:03<1:23:51, 2.02it/s]
561
  5%|▍ | 521/10682 [05:03<1:23:45, 2.02it/s]
562
  5%|▍ | 522/10682 [05:04<1:23:47, 2.02it/s]
563
  5%|▍ | 523/10682 [05:04<1:23:41, 2.02it/s]
564
  5%|▍ | 524/10682 [05:05<1:23:46, 2.02it/s]
565
  5%|▍ | 525/10682 [05:05<1:23:42, 2.02it/s]{'loss': 5.2021, 'grad_norm': 0.6268776059150696, 'learning_rate': 0.0004911131898971, 'epoch': 0.69}
 
566
 
567
  5%|▍ | 525/10682 [05:05<1:23:42, 2.02it/s]
568
  5%|▍ | 526/10682 [05:06<1:23:54, 2.02it/s]
569
  5%|▍ | 527/10682 [05:06<1:23:49, 2.02it/s]
570
  5%|▍ | 528/10682 [05:07<1:23:46, 2.02it/s]
571
  5%|▍ | 529/10682 [05:07<1:23:43, 2.02it/s]
572
  5%|▍ | 530/10682 [05:08<1:23:43, 2.02it/s]
573
  5%|▍ | 531/10682 [05:08<1:23:44, 2.02it/s]
574
  5%|▍ | 532/10682 [05:09<1:23:40, 2.02it/s]
575
  5%|▍ | 533/10682 [05:09<1:23:42, 2.02it/s]
576
  5%|▍ | 534/10682 [05:10<1:24:01, 2.01it/s]
577
  5%|▌ | 535/10682 [05:10<1:24:00, 2.01it/s]
578
  5%|▌ | 536/10682 [05:11<1:24:02, 2.01it/s]
579
  5%|▌ | 537/10682 [05:11<1:24:01, 2.01it/s]
580
  5%|▌ | 538/10682 [05:12<1:23:59, 2.01it/s]
581
  5%|▌ | 539/10682 [05:12<1:23:52, 2.02it/s]
582
  5%|▌ | 540/10682 [05:13<1:23:55, 2.01it/s]
583
  5%|▌ | 541/10682 [05:13<1:23:45, 2.02it/s]
584
  5%|▌ | 542/10682 [05:14<1:23:43, 2.02it/s]
585
  5%|▌ | 543/10682 [05:14<1:23:35, 2.02it/s]
586
  5%|▌ | 544/10682 [05:15<1:23:34, 2.02it/s]
587
  5%|▌ | 545/10682 [05:15<1:23:32, 2.02it/s]
588
  5%|▌ | 546/10682 [05:16<1:23:33, 2.02it/s]
589
  5%|▌ | 547/10682 [05:16<1:23:31, 2.02it/s]
590
  5%|▌ | 548/10682 [05:17<1:23:36, 2.02it/s]
591
  5%|▌ | 549/10682 [05:17<1:23:34, 2.02it/s]
592
  5%|▌ | 550/10682 [05:18<1:23:38, 2.02it/s]
593
  {'loss': 5.1481, 'grad_norm': 0.6789132952690125, 'learning_rate': 0.0005144995322731525, 'epoch': 0.72}
 
594
  5%|▌ | 550/10682 [05:18<1:23:38, 2.02it/s]
595
  5%|▌ | 551/10682 [05:18<1:23:44, 2.02it/s]
596
  5%|▌ | 552/10682 [05:19<1:23:41, 2.02it/s]
597
  5%|▌ | 553/10682 [05:19<1:23:36, 2.02it/s]
598
  5%|▌ | 554/10682 [05:20<1:23:36, 2.02it/s]
599
  5%|▌ | 555/10682 [05:20<1:23:33, 2.02it/s]
600
  5%|▌ | 556/10682 [05:21<1:23:33, 2.02it/s]
601
  5%|▌ | 557/10682 [05:21<1:23:37, 2.02it/s]
602
  5%|▌ | 558/10682 [05:22<1:23:32, 2.02it/s]
603
  5%|▌ | 559/10682 [05:22<1:23:33, 2.02it/s]
604
  5%|▌ | 560/10682 [05:23<1:23:31, 2.02it/s]
605
  5%|▌ | 561/10682 [05:23<1:23:35, 2.02it/s]
606
  5%|▌ | 562/10682 [05:23<1:23:29, 2.02it/s]
607
  5%|▌ | 563/10682 [05:24<1:23:32, 2.02it/s]
608
  5%|▌ | 564/10682 [05:24<1:23:24, 2.02it/s]
609
  5%|▌ | 565/10682 [05:25<1:23:28, 2.02it/s]
610
  5%|▌ | 566/10682 [05:25<1:23:22, 2.02it/s]
611
  5%|▌ | 567/10682 [05:26<1:23:25, 2.02it/s]
612
  5%|▌ | 568/10682 [05:26<1:23:22, 2.02it/s]
613
  5%|▌ | 569/10682 [05:27<1:23:23, 2.02it/s]
614
  5%|▌ | 570/10682 [05:27<1:23:21, 2.02it/s]
615
  5%|▌ | 571/10682 [05:28<1:23:21, 2.02it/s]
616
  5%|▌ | 572/10682 [05:28<1:23:26, 2.02it/s]
617
  5%|▌ | 573/10682 [05:29<1:23:26, 2.02it/s]
618
  5%|▌ | 574/10682 [05:29<1:23:24, 2.02it/s]
619
  5%|▌ | 575/10682 [05:30<1:23:21, 2.02it/s]{'loss': 5.0923, 'grad_norm': 0.5607177019119263, 'learning_rate': 0.0005378858746492049, 'epoch': 0.75}
 
620
 
621
  5%|▌ | 575/10682 [05:30<1:23:21, 2.02it/s]
622
  5%|▌ | 576/10682 [05:30<1:23:33, 2.02it/s]
623
  5%|▌ | 577/10682 [05:31<1:23:23, 2.02it/s]
624
  5%|▌ | 578/10682 [05:31<1:23:26, 2.02it/s]
625
  5%|▌ | 579/10682 [05:32<1:23:17, 2.02it/s]
626
  5%|▌ | 580/10682 [05:32<1:23:20, 2.02it/s]
627
  5%|▌ | 581/10682 [05:33<1:23:18, 2.02it/s]
628
  5%|▌ | 582/10682 [05:33<1:23:22, 2.02it/s]
629
  5%|▌ | 583/10682 [05:34<1:23:21, 2.02it/s]
630
  5%|▌ | 584/10682 [05:34<1:23:24, 2.02it/s]
631
  5%|▌ | 585/10682 [05:35<1:23:22, 2.02it/s]
632
  5%|▌ | 586/10682 [05:35<1:23:18, 2.02it/s]
633
  5%|▌ | 587/10682 [05:36<1:23:16, 2.02it/s]
634
  6%|▌ | 588/10682 [05:36<1:23:17, 2.02it/s]
635
  6%|▌ | 589/10682 [05:37<1:23:13, 2.02it/s]
636
  6%|▌ | 590/10682 [05:37<1:23:15, 2.02it/s]
637
  6%|▌ | 591/10682 [05:38<1:23:16, 2.02it/s]
638
  6%|▌ | 592/10682 [05:38<1:23:14, 2.02it/s]
639
  6%|▌ | 593/10682 [05:39<1:23:12, 2.02it/s]
640
  6%|▌ | 594/10682 [05:39<1:23:12, 2.02it/s]
641
  6%|▌ | 595/10682 [05:40<1:23:09, 2.02it/s]
642
  6%|▌ | 596/10682 [05:40<1:23:09, 2.02it/s]
643
  6%|▌ | 597/10682 [05:41<1:23:06, 2.02it/s]
644
  6%|▌ | 598/10682 [05:41<1:23:09, 2.02it/s]
645
  6%|▌ | 599/10682 [05:42<1:23:06, 2.02it/s]
646
  6%|▌ | 600/10682 [05:42<1:23:10, 2.02it/s]
647
 
 
648
  6%|▌ | 600/10682 [05:42<1:23:10, 2.02it/s]
649
  6%|▌ | 601/10682 [05:43<1:23:21, 2.02it/s]
650
  6%|▌ | 602/10682 [05:43<1:23:12, 2.02it/s]
651
  6%|▌ | 603/10682 [05:44<1:23:08, 2.02it/s]
652
  6%|▌ | 604/10682 [05:44<1:23:12, 2.02it/s]
653
  6%|▌ | 605/10682 [05:45<1:23:25, 2.01it/s]
654
  6%|▌ | 606/10682 [05:45<1:23:21, 2.01it/s]
655
  6%|▌ | 607/10682 [05:46<1:23:20, 2.01it/s]
656
  6%|▌ | 608/10682 [05:46<1:23:11, 2.02it/s]
657
  6%|▌ | 609/10682 [05:47<1:30:09, 1.86it/s]
658
  6%|▌ | 610/10682 [05:47<1:27:59, 1.91it/s]
659
  6%|▌ | 611/10682 [05:48<1:26:34, 1.94it/s]
660
  6%|▌ | 612/10682 [05:48<1:25:27, 1.96it/s]
661
  6%|▌ | 613/10682 [05:49<1:24:45, 1.98it/s]
662
  6%|▌ | 614/10682 [05:49<1:24:14, 1.99it/s]
663
  6%|▌ | 615/10682 [05:50<1:23:50, 2.00it/s]
664
  6%|▌ | 616/10682 [05:51<1:30:30, 1.85it/s]
665
  6%|▌ | 617/10682 [05:51<1:28:14, 1.90it/s]
666
  6%|▌ | 618/10682 [05:51<1:26:43, 1.93it/s]
667
  6%|▌ | 619/10682 [05:52<1:25:37, 1.96it/s]
668
  6%|▌ | 620/10682 [05:52<1:24:49, 1.98it/s]
669
  6%|▌ | 621/10682 [05:53<1:24:13, 1.99it/s]
670
  6%|▌ | 622/10682 [05:53<1:23:52, 2.00it/s]
671
  6%|▌ | 623/10682 [05:54<1:23:33, 2.01it/s]
672
  6%|▌ | 624/10682 [05:54<1:23:24, 2.01it/s]
673
  6%|▌ | 625/10682 [05:55<1:23:16, 2.01it/s]
674
 
 
675
  6%|▌ | 625/10682 [05:55<1:23:16, 2.01it/s]
676
  6%|▌ | 626/10682 [05:55<1:23:16, 2.01it/s]
677
  6%|▌ | 627/10682 [05:56<1:23:10, 2.01it/s]
678
  6%|▌ | 628/10682 [05:56<1:23:04, 2.02it/s]
679
  6%|▌ | 629/10682 [05:57<1:23:12, 2.01it/s]
680
  6%|▌ | 630/10682 [05:57<1:23:09, 2.01it/s]
681
  6%|▌ | 631/10682 [05:58<1:23:07, 2.02it/s]
682
  6%|▌ | 632/10682 [05:58<1:23:05, 2.02it/s]
683
  6%|▌ | 633/10682 [05:59<1:23:00, 2.02it/s]
684
  6%|▌ | 634/10682 [05:59<1:22:54, 2.02it/s]
685
  6%|▌ | 635/10682 [06:00<1:22:56, 2.02it/s]
686
  6%|▌ | 636/10682 [06:00<1:22:53, 2.02it/s]
687
  6%|▌ | 637/10682 [06:01<1:22:54, 2.02it/s]
688
  6%|▌ | 638/10682 [06:01<1:22:50, 2.02it/s]
689
  6%|▌ | 639/10682 [06:02<1:22:48, 2.02it/s]
690
  6%|▌ | 640/10682 [06:02<1:22:46, 2.02it/s]
691
  6%|▌ | 641/10682 [06:03<1:22:50, 2.02it/s]
692
  6%|▌ | 642/10682 [06:03<1:22:47, 2.02it/s]
693
  6%|▌ | 643/10682 [06:04<1:22:49, 2.02it/s]
694
  6%|▌ | 644/10682 [06:04<1:22:45, 2.02it/s]
695
  6%|▌ | 645/10682 [06:05<1:22:46, 2.02it/s]
696
  6%|▌ | 646/10682 [06:05<1:22:46, 2.02it/s]
697
  6%|▌ | 647/10682 [06:06<1:22:45, 2.02it/s]
698
  6%|▌ | 648/10682 [06:06<1:22:45, 2.02it/s]
699
  6%|▌ | 649/10682 [06:07<1:22:48, 2.02it/s]
700
  6%|▌ | 650/10682 [06:07<1:22:43, 2.02it/s]
701
 
 
702
  6%|▌ | 650/10682 [06:07<1:22:43, 2.02it/s]
703
  6%|▌ | 651/10682 [06:08<1:22:50, 2.02it/s]
704
  6%|▌ | 652/10682 [06:08<1:22:52, 2.02it/s]
705
  6%|▌ | 653/10682 [06:09<1:22:49, 2.02it/s]
706
  6%|▌ | 654/10682 [06:09<1:22:47, 2.02it/s]
707
  6%|▌ | 655/10682 [06:10<1:22:43, 2.02it/s]
708
  6%|▌ | 656/10682 [06:10<1:22:44, 2.02it/s]
709
  6%|▌ | 657/10682 [06:11<1:22:40, 2.02it/s]
710
  6%|▌ | 658/10682 [06:11<1:22:43, 2.02it/s]
711
  6%|▌ | 659/10682 [06:12<1:22:38, 2.02it/s]
712
  6%|▌ | 660/10682 [06:12<1:22:36, 2.02it/s]
713
  6%|▌ | 661/10682 [06:13<1:22:39, 2.02it/s]
714
  6%|▌ | 662/10682 [06:13<1:22:36, 2.02it/s]
715
  6%|▌ | 663/10682 [06:14<1:22:40, 2.02it/s]
716
  6%|▌ | 664/10682 [06:14<1:22:33, 2.02it/s]
717
  6%|▌ | 665/10682 [06:15<1:22:36, 2.02it/s]
718
  6%|▌ | 666/10682 [06:15<1:22:33, 2.02it/s]
719
  6%|▌ | 667/10682 [06:16<1:22:38, 2.02it/s]
720
  6%|▋ | 668/10682 [06:16<1:22:38, 2.02it/s]
721
  6%|▋ | 669/10682 [06:17<1:22:36, 2.02it/s]
722
  6%|▋ | 670/10682 [06:17<1:22:36, 2.02it/s]
723
  6%|▋ | 671/10682 [06:18<1:22:34, 2.02it/s]
724
  6%|▋ | 672/10682 [06:18<1:22:36, 2.02it/s]
725
  6%|▋ | 673/10682 [06:19<1:22:32, 2.02it/s]
726
  6%|▋ | 674/10682 [06:19<1:22:48, 2.01it/s]
727
  6%|▋ | 675/10682 [06:20<1:22:57, 2.01it/s]
728
  {'loss': 4.9272, 'grad_norm': 0.5725471377372742, 'learning_rate': 0.0006314312441534145, 'epoch': 0.88}
 
729
  6%|▋ | 675/10682 [06:20<1:22:57, 2.01it/s]
730
  6%|▋ | 676/10682 [06:20<1:22:58, 2.01it/s]
731
  6%|▋ | 677/10682 [06:21<1:22:53, 2.01it/s]
732
  6%|▋ | 678/10682 [06:21<1:22:54, 2.01it/s]
733
  6%|▋ | 679/10682 [06:22<1:22:41, 2.02it/s]
734
  6%|▋ | 680/10682 [06:22<1:22:40, 2.02it/s]
735
  6%|▋ | 681/10682 [06:23<1:22:33, 2.02it/s]
736
  6%|▋ | 682/10682 [06:23<1:22:35, 2.02it/s]
737
  6%|▋ | 683/10682 [06:24<1:22:31, 2.02it/s]
738
  6%|▋ | 684/10682 [06:24<1:22:34, 2.02it/s]
739
  6%|▋ | 685/10682 [06:25<1:22:27, 2.02it/s]
740
  6%|▋ | 686/10682 [06:25<1:22:31, 2.02it/s]
741
  6%|▋ | 687/10682 [06:26<1:22:30, 2.02it/s]
742
  6%|▋ | 688/10682 [06:26<1:22:34, 2.02it/s]
743
  6%|▋ | 689/10682 [06:27<1:22:26, 2.02it/s]
744
  6%|▋ | 690/10682 [06:27<1:22:29, 2.02it/s]
745
  6%|▋ | 691/10682 [06:28<1:22:23, 2.02it/s]
746
  6%|▋ | 692/10682 [06:28<1:22:26, 2.02it/s]
747
  6%|▋ | 693/10682 [06:29<1:22:25, 2.02it/s]
748
  6%|▋ | 694/10682 [06:29<1:22:26, 2.02it/s]
749
  7%|▋ | 695/10682 [06:30<1:22:25, 2.02it/s]
750
  7%|▋ | 696/10682 [06:30<1:22:24, 2.02it/s]
751
  7%|▋ | 697/10682 [06:31<1:22:24, 2.02it/s]
752
  7%|▋ | 698/10682 [06:31<1:22:23, 2.02it/s]
753
  7%|▋ | 699/10682 [06:32<1:22:20, 2.02it/s]
754
  7%|▋ | 700/10682 [06:32<1:22:18, 2.02it/s]{'loss': 4.8834, 'grad_norm': 0.48604607582092285, 'learning_rate': 0.0006548175865294667, 'epoch': 0.92}
 
755
 
756
  7%|▋ | 700/10682 [06:32<1:22:18, 2.02it/s]
757
  7%|▋ | 701/10682 [06:33<1:22:27, 2.02it/s]
758
  7%|▋ | 702/10682 [06:33<1:22:23, 2.02it/s]
759
  7%|▋ | 703/10682 [06:34<1:22:24, 2.02it/s]
760
  7%|▋ | 704/10682 [06:34<1:22:19, 2.02it/s]
761
  7%|▋ | 705/10682 [06:35<1:22:24, 2.02it/s]
762
  7%|▋ | 706/10682 [06:35<1:22:19, 2.02it/s]
763
  7%|▋ | 707/10682 [06:36<1:22:20, 2.02it/s]
764
  7%|▋ | 708/10682 [06:36<1:22:14, 2.02it/s]
765
  7%|▋ | 709/10682 [06:37<1:22:18, 2.02it/s]
766
  7%|▋ | 710/10682 [06:37<1:22:17, 2.02it/s]
767
  7%|▋ | 711/10682 [06:38<1:22:21, 2.02it/s]
768
  7%|▋ | 712/10682 [06:38<1:22:16, 2.02it/s]
769
  7%|▋ | 713/10682 [06:39<1:22:17, 2.02it/s]
770
  7%|▋ | 714/10682 [06:39<1:22:16, 2.02it/s]
771
  7%|▋ | 715/10682 [06:40<1:22:14, 2.02it/s]
772
  7%|▋ | 716/10682 [06:40<1:22:15, 2.02it/s]
773
  7%|▋ | 717/10682 [06:41<1:22:14, 2.02it/s]
774
  7%|▋ | 718/10682 [06:41<1:22:18, 2.02it/s]
775
  7%|▋ | 719/10682 [06:42<1:22:13, 2.02it/s]
776
  7%|▋ | 720/10682 [06:42<1:22:12, 2.02it/s]
777
  7%|▋ | 721/10682 [06:43<1:22:09, 2.02it/s]
778
  7%|▋ | 722/10682 [06:43<1:22:13, 2.02it/s]
779
  7%|▋ | 723/10682 [06:43<1:22:13, 2.02it/s]
780
  7%|▋ | 724/10682 [06:44<1:22:16, 2.02it/s]
781
  7%|▋ | 725/10682 [06:44<1:22:13, 2.02it/s]{'loss': 4.8458, 'grad_norm': 0.43501177430152893, 'learning_rate': 0.0006782039289055192, 'epoch': 0.95}
 
782
 
783
  7%|▋ | 725/10682 [06:44<1:22:13, 2.02it/s]
784
  7%|▋ | 726/10682 [06:45<1:22:24, 2.01it/s]
785
  7%|▋ | 727/10682 [06:45<1:22:27, 2.01it/s]
786
  7%|▋ | 728/10682 [06:46<1:22:23, 2.01it/s]
787
  7%|▋ | 729/10682 [06:46<1:22:21, 2.01it/s]
788
  7%|▋ | 730/10682 [06:47<1:22:15, 2.02it/s]
789
  7%|▋ | 731/10682 [06:47<1:22:14, 2.02it/s]
790
  7%|▋ | 732/10682 [06:48<1:22:07, 2.02it/s]
791
  7%|▋ | 733/10682 [06:48<1:22:09, 2.02it/s]
792
  7%|▋ | 734/10682 [06:49<1:22:09, 2.02it/s]
793
  7%|▋ | 735/10682 [06:49<1:22:10, 2.02it/s]
794
  7%|▋ | 736/10682 [06:50<1:22:07, 2.02it/s]
795
  7%|▋ | 737/10682 [06:50<1:22:06, 2.02it/s]
796
  7%|▋ | 738/10682 [06:51<1:22:09, 2.02it/s]
797
  7%|▋ | 739/10682 [06:51<1:22:04, 2.02it/s]
798
  7%|▋ | 740/10682 [06:52<1:22:10, 2.02it/s]
799
  7%|▋ | 741/10682 [06:52<1:22:05, 2.02it/s]
800
  7%|▋ | 742/10682 [06:53<1:22:11, 2.02it/s]
801
  7%|▋ | 743/10682 [06:53<1:22:06, 2.02it/s]
802
  7%|▋ | 744/10682 [06:54<1:22:14, 2.01it/s]
803
  7%|▋ | 745/10682 [06:54<1:22:34, 2.01it/s]
804
  7%|▋ | 746/10682 [06:55<1:22:48, 2.00it/s]
805
  7%|▋ | 747/10682 [06:55<1:23:16, 1.99it/s]
806
  7%|▋ | 748/10682 [06:56<1:23:37, 1.98it/s]
807
  7%|▋ | 749/10682 [06:56<1:23:49, 1.97it/s]
808
  7%|▋ | 750/10682 [06:57<1:23:50, 1.97it/s]{'loss': 4.809, 'grad_norm': 0.434061199426651, 'learning_rate': 0.0007015902712815716, 'epoch': 0.98}
 
809
 
810
  7%|▋ | 750/10682 [06:57<1:23:50, 1.97it/s]
811
  7%|▋ | 751/10682 [06:57<1:23:46, 1.98it/s]
812
  7%|▋ | 752/10682 [06:58<1:23:32, 1.98it/s]
813
  7%|▋ | 753/10682 [06:58<1:23:05, 1.99it/s]
814
  7%|▋ | 754/10682 [06:59<1:22:43, 2.00it/s]
815
  7%|▋ | 755/10682 [06:59<1:22:34, 2.00it/s]
816
  7%|▋ | 756/10682 [07:00<1:22:18, 2.01it/s]
817
  7%|▋ | 757/10682 [07:00<1:22:12, 2.01it/s]
818
  7%|▋ | 758/10682 [07:01<1:22:03, 2.02it/s]
819
  7%|▋ | 759/10682 [07:01<1:22:04, 2.02it/s]
820
  7%|▋ | 760/10682 [07:02<1:21:55, 2.02it/s]
821
  7%|▋ | 761/10682 [07:02<1:21:56, 2.02it/s]
822
  7%|▋ | 762/10682 [07:03<1:21:50, 2.02it/s]
823
  7%|▋ | 763/10682 [07:03<1:24:02, 1.97it/s]
824
  7%|▋ | 764/10682 [07:36<28:16:56, 10.27s/it]
825
  7%|▋ | 765/10682 [07:37<20:12:22, 7.34s/it]
826
  7%|▋ | 766/10682 [07:37<14:33:39, 5.29s/it]
827
  7%|▋ | 767/10682 [07:38<10:36:06, 3.85s/it]
828
  7%|▋ | 768/10682 [07:38<7:49:41, 2.84s/it]
829
  7%|▋ | 769/10682 [07:39<5:53:19, 2.14s/it]
830
  7%|▋ | 770/10682 [07:39<4:31:48, 1.65s/it]
831
  7%|▋ | 771/10682 [07:40<3:34:48, 1.30s/it]
832
  7%|▋ | 772/10682 [07:40<2:54:46, 1.06s/it]
833
  7%|▋ | 773/10682 [07:41<2:26:52, 1.12it/s]
834
  7%|▋ | 774/10682 [07:41<2:07:16, 1.30it/s]
835
  7%|▋ | 775/10682 [07:42<1:53:45, 1.45it/s]{'loss': 4.7582, 'grad_norm': 0.39575064182281494, 'learning_rate': 0.0007249766136576241, 'epoch': 1.02}
 
836
 
837
  7%|▋ | 775/10682 [07:42<1:53:45, 1.45it/s]
838
  7%|▋ | 776/10682 [07:42<1:44:22, 1.58it/s]
839
  7%|▋ | 777/10682 [07:43<1:37:30, 1.69it/s]
840
  7%|▋ | 778/10682 [07:43<1:32:52, 1.78it/s]
841
  7%|▋ | 779/10682 [07:44<1:29:29, 1.84it/s]
842
  7%|▋ | 780/10682 [07:44<1:27:18, 1.89it/s]
843
  7%|▋ | 781/10682 [07:45<1:25:28, 1.93it/s]
844
  7%|▋ | 782/10682 [07:45<1:24:24, 1.95it/s]
845
  7%|▋ | 783/10682 [07:46<1:23:35, 1.97it/s]
846
  7%|▋ | 784/10682 [07:46<1:23:12, 1.98it/s]
847
  7%|▋ | 785/10682 [07:47<1:22:48, 1.99it/s]
848
  7%|▋ | 786/10682 [07:47<1:22:41, 1.99it/s]
849
  7%|▋ | 787/10682 [07:48<1:22:20, 2.00it/s]
850
  7%|▋ | 788/10682 [07:48<1:22:09, 2.01it/s]
851
  7%|▋ | 789/10682 [07:49<1:22:03, 2.01it/s]
852
  7%|▋ | 790/10682 [07:49<1:22:01, 2.01it/s]
853
  7%|▋ | 791/10682 [07:50<1:22:04, 2.01it/s]
854
  7%|▋ | 792/10682 [07:50<1:21:54, 2.01it/s]
855
  7%|▋ | 793/10682 [07:51<1:21:51, 2.01it/s]
856
  7%|▋ | 794/10682 [07:51<1:21:45, 2.02it/s]
 
1
+ slurm submission log: 2024-05-09 15:03:34.769148
2
+ created following sbatch script:
3
+
4
+ ###############################
5
+
6
+ #!/bin/bash
7
+
8
+ #SBATCH --account=nlp
9
+ #SBATCH --cpus-per-task=16
10
+ #SBATCH --gres=gpu:2
11
+ #SBATCH --job-name=tthrush-job-4546404
12
+ #SBATCH --mem=400G
13
+ #SBATCH --nodelist=sphinx2
14
+ #SBATCH --open-mode=append
15
+ #SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy/train_job_output.txt
16
+ #SBATCH --partition=sphinx
17
+ #SBATCH --time=14-0
18
+
19
+ # activate your desired anaconda environment
20
+ . /nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection
21
+
22
+ # cd to working directory
23
+ cd .
24
+
25
+ # launch commands
26
+ srun --unbuffered run_as_child_processes 'torchrun --master_port 29505 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/arc_easy --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy --output_hub_id pythia-70m_arc_easy --model_id EleutherAI/pythia-70m --num_train_epochs 14 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2'
27
+
28
+ ###############################
29
+
30
+ submission to slurm complete!
31
+
32
+
33
+ ###############################
34
+ slurm submission output
35
+
36
+ Submitted batch job 7592327
37
+
38
+
39
+
40
+ ###############################
41
+
42
+ /var/lib/slurm/slurmd/job7592327/slurm_script: line 15: /nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/etc/profile.d/conda.sh: No such file or directory
43
+
44
+ CommandNotFoundError: Your shell has not been properly configured to use 'conda activate'.
45
+ To initialize your shell, run
46
+
47
+ $ conda init <SHELL_NAME>
48
+
49
+ Currently supported shells are:
50
+ - bash
51
+ - fish
52
+ - tcsh
53
+ - xonsh
54
+ - zsh
55
+ - powershell
56
+
57
+ See 'conda init --help' for more information and options.
58
+
59
+ IMPORTANT: You may need to close and restart your shell after running 'conda init'.
60
+
61
+
62
+ ###############################
63
+ start time: 2024-05-09 18:12:52.469667
64
+ machine: sphinx2
65
+ conda env: pretraining-coreset-selection
66
+ ###############################
67
+ running following processes
68
+
69
+ torchrun --master_port 29505 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/arc_easy --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy --output_hub_id pythia-70m_arc_easy --model_id EleutherAI/pythia-70m --num_train_epochs 14 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2
70
+
71
+
72
+ ###############################
73
+ command outputs:
74
+
75
+
76
+ [2024-05-09 18:12:53,880] torch.distributed.run: [WARNING]
77
+ [2024-05-09 18:12:53,880] torch.distributed.run: [WARNING] *****************************************
78
+ [2024-05-09 18:12:53,880] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
79
+ [2024-05-09 18:12:53,880] torch.distributed.run: [WARNING] *****************************************
80
+ 05/09/2024 18:12:57 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/arc_easy', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy', output_hub_id='pythia-70m_arc_easy', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=14, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
81
+ 05/09/2024 18:12:57 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/arc_easy', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_arc_easy', output_hub_id='pythia-70m_arc_easy', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=14, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
82
+
83
  0%| | 0/10682 [00:00<?, ?it/s][rank0]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
84
+ [rank1]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration, which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
85
+
86
  0%| | 1/10682 [00:24<72:50:43, 24.55s/it]
87
  0%| | 2/10682 [00:30<40:22:35, 13.61s/it]
88
  0%| | 3/10682 [00:34<26:55:28, 9.08s/it]
89
  0%| | 4/10682 [00:37<19:47:10, 6.67s/it]
90
  0%| | 5/10682 [00:39<15:01:55, 5.07s/it]
91
  0%| | 6/10682 [00:41<11:32:34, 3.89s/it]
92
  0%| | 7/10682 [00:42<9:31:09, 3.21s/it]
93
  0%| | 8/10682 [00:44<7:38:58, 2.58s/it]
94
  0%| | 9/10682 [00:45<6:17:33, 2.12s/it]
95
  0%| | 10/10682 [00:46<5:17:31, 1.79s/it]
96
  0%| | 11/10682 [00:47<4:44:47, 1.60s/it]
97
  0%| | 12/10682 [00:48<4:02:04, 1.36s/it]
98
  0%| | 13/10682 [00:48<3:27:39, 1.17s/it]
99
  0%| | 14/10682 [00:49<3:05:04, 1.04s/it]
100
  0%| | 15/10682 [00:50<2:50:37, 1.04it/s]
101
  0%| | 16/10682 [00:51<2:39:54, 1.11it/s]
102
  0%| | 17/10682 [00:51<2:32:45, 1.16it/s]
103
  0%| | 18/10682 [00:52<2:19:43, 1.27it/s]
104
  0%| | 19/10682 [00:53<2:11:50, 1.35it/s]
105
  0%| | 20/10682 [00:53<2:06:18, 1.41it/s]
106
  0%| | 21/10682 [00:54<2:00:31, 1.47it/s]
107
  0%| | 22/10682 [00:55<1:57:33, 1.51it/s]
108
  0%| | 23/10682 [00:55<1:54:06, 1.56it/s]
109
  0%| | 24/10682 [00:56<1:48:28, 1.64it/s]
110
  0%| | 25/10682 [00:57<2:03:47, 1.43it/s]
111
 
112
+
113
  0%| | 25/10682 [00:57<2:03:47, 1.43it/s]
114
  0%| | 26/10682 [00:57<1:55:38, 1.54it/s]
115
  0%| | 27/10682 [00:58<1:50:06, 1.61it/s]
116
  0%| | 28/10682 [00:58<1:44:53, 1.69it/s]
117
  0%| | 29/10682 [00:59<1:41:47, 1.74it/s]
118
  0%| | 30/10682 [00:59<1:39:14, 1.79it/s]
119
  0%| | 31/10682 [01:00<1:36:57, 1.83it/s]
120
  0%| | 32/10682 [01:00<1:36:40, 1.84it/s]
121
  0%| | 33/10682 [01:01<1:35:13, 1.86it/s]
122
  0%| | 34/10682 [01:01<1:34:03, 1.89it/s]
123
  0%| | 35/10682 [01:02<1:33:14, 1.90it/s]
124
  0%| | 36/10682 [01:02<1:32:17, 1.92it/s]
125
  0%| | 37/10682 [01:03<1:31:49, 1.93it/s]
126
  0%| | 38/10682 [01:03<1:31:51, 1.93it/s]
127
  0%| | 39/10682 [01:04<1:31:50, 1.93it/s]
128
  0%| | 40/10682 [01:04<1:31:23, 1.94it/s]
129
  0%| | 41/10682 [01:05<1:31:33, 1.94it/s]
130
  0%| | 42/10682 [01:05<1:31:05, 1.95it/s]
131
  0%| | 43/10682 [01:06<1:30:43, 1.95it/s]
132
  0%| | 44/10682 [01:06<1:30:24, 1.96it/s]
133
  0%| | 45/10682 [01:07<1:30:14, 1.96it/s]
134
  0%| | 46/10682 [01:07<1:30:06, 1.97it/s]
135
  0%| | 47/10682 [01:08<1:29:58, 1.97it/s]
136
  0%| | 48/10682 [01:09<1:29:53, 1.97it/s]
137
  0%| | 49/10682 [01:09<1:29:42, 1.98it/s]
138
  0%| | 50/10682 [01:10<1:29:48, 1.97it/s]
139
  {'loss': 9.7768, 'grad_norm': 1.579921841621399, 'learning_rate': 4.677268475210477e-05, 'epoch': 0.07}
140
+
141
  0%| | 50/10682 [01:10<1:29:48, 1.97it/s]
142
  0%| | 51/10682 [01:10<1:29:53, 1.97it/s]
143
  0%| | 52/10682 [01:11<1:29:35, 1.98it/s]
144
  0%| | 53/10682 [01:11<1:29:25, 1.98it/s]
145
  1%| | 54/10682 [01:12<1:29:08, 1.99it/s]
146
  1%| | 55/10682 [01:12<1:28:56, 1.99it/s]
147
  1%| | 56/10682 [01:13<1:28:54, 1.99it/s]
148
  1%| | 57/10682 [01:13<1:29:03, 1.99it/s]
149
  1%| | 58/10682 [01:14<1:29:10, 1.99it/s]
150
  1%| | 59/10682 [01:14<1:29:06, 1.99it/s]
151
  1%| | 60/10682 [01:15<1:28:57, 1.99it/s]
152
  1%| | 61/10682 [01:15<1:28:53, 1.99it/s]
153
  1%| | 62/10682 [01:16<1:28:53, 1.99it/s]
154
  1%| | 63/10682 [01:16<1:28:51, 1.99it/s]
155
  1%| | 64/10682 [01:17<1:28:54, 1.99it/s]
156
  1%| | 65/10682 [01:17<1:28:42, 1.99it/s]
157
  1%| | 66/10682 [01:18<1:28:43, 1.99it/s]
158
  1%| | 67/10682 [01:18<1:28:31, 2.00it/s]
159
  1%| | 68/10682 [01:19<1:28:32, 2.00it/s]
160
  1%| | 69/10682 [01:19<1:28:31, 2.00it/s]
161
  1%| | 70/10682 [01:20<1:28:29, 2.00it/s]
162
  1%| | 71/10682 [01:20<1:28:20, 2.00it/s]
163
  1%| | 72/10682 [01:21<1:28:12, 2.00it/s]
164
  1%| | 73/10682 [01:21<1:28:18, 2.00it/s]
165
  1%| | 74/10682 [01:22<1:41:07, 1.75it/s]
166
  1%| | 75/10682 [01:22<1:37:16, 1.82it/s]
167
 
168
+
169
  1%| | 75/10682 [01:22<1:37:16, 1.82it/s]
170
  1%| | 76/10682 [01:23<1:34:38, 1.87it/s]
171
  1%| | 77/10682 [01:23<1:32:38, 1.91it/s]
172
  1%| | 78/10682 [01:24<1:31:11, 1.94it/s]
173
  1%| | 79/10682 [01:24<1:30:20, 1.96it/s]
174
  1%| | 80/10682 [01:25<1:29:41, 1.97it/s]
175
  1%| | 81/10682 [01:25<1:29:13, 1.98it/s]
176
  1%| | 82/10682 [01:26<1:28:50, 1.99it/s]
177
  1%| | 83/10682 [01:26<1:28:30, 2.00it/s]
178
  1%| | 84/10682 [01:27<1:28:18, 2.00it/s]
179
  1%| | 85/10682 [01:27<1:28:13, 2.00it/s]
180
  1%| | 86/10682 [01:28<1:28:07, 2.00it/s]
181
  1%| | 87/10682 [01:28<1:28:07, 2.00it/s]
182
  1%| | 88/10682 [01:29<1:27:49, 2.01it/s]
183
  1%| | 89/10682 [01:29<1:27:44, 2.01it/s]
184
  1%| | 90/10682 [01:30<1:27:37, 2.01it/s]
185
  1%| | 91/10682 [01:30<1:27:39, 2.01it/s]
186
  1%| | 92/10682 [01:31<1:27:40, 2.01it/s]
187
  1%| | 93/10682 [01:31<1:27:30, 2.02it/s]
188
  1%| | 94/10682 [01:32<1:27:35, 2.01it/s]
189
  1%| | 95/10682 [01:32<1:27:32, 2.02it/s]
190
  1%| | 96/10682 [01:33<1:27:32, 2.02it/s]
191
  1%| | 97/10682 [01:33<1:27:35, 2.01it/s]
192
  1%| | 98/10682 [01:34<1:27:28, 2.02it/s]
193
  1%| | 99/10682 [01:34<1:27:35, 2.01it/s]
194
  1%| | 100/10682 [01:35<1:27:28, 2.02it/s]{'loss': 8.2552, 'grad_norm': 1.013319969177246, 'learning_rate': 9.354536950420954e-05, 'epoch': 0.13}
195
 
196
+
197
  1%| | 100/10682 [01:35<1:27:28, 2.02it/s]
198
  1%| | 101/10682 [01:35<1:27:49, 2.01it/s]
199
  1%| | 102/10682 [01:36<1:27:45, 2.01it/s]
200
  1%| | 103/10682 [01:36<1:27:36, 2.01it/s]
201
  1%| | 104/10682 [01:37<1:27:28, 2.02it/s]
202
  1%| | 105/10682 [01:37<1:27:21, 2.02it/s]
203
  1%| | 106/10682 [01:38<1:27:28, 2.02it/s]
204
  1%| | 107/10682 [01:38<1:27:20, 2.02it/s]
205
  1%| | 108/10682 [01:39<1:27:21, 2.02it/s]
206
  1%| | 109/10682 [01:39<1:27:18, 2.02it/s]
207
  1%| | 110/10682 [01:40<1:27:16, 2.02it/s]
208
  1%| | 111/10682 [01:40<1:27:13, 2.02it/s]
209
  1%| | 112/10682 [01:41<1:27:16, 2.02it/s]
210
  1%| | 113/10682 [01:41<1:27:18, 2.02it/s]
211
  1%| | 114/10682 [01:42<1:27:14, 2.02it/s]
212
  1%| | 115/10682 [01:42<1:27:17, 2.02it/s]
213
  1%| | 116/10682 [01:43<1:27:10, 2.02it/s]
214
  1%| | 117/10682 [01:43<1:27:12, 2.02it/s]
215
  1%| | 118/10682 [01:44<1:27:26, 2.01it/s]
216
  1%| | 119/10682 [01:44<1:27:29, 2.01it/s]
217
  1%| | 120/10682 [01:45<1:27:24, 2.01it/s]
218
  1%| | 121/10682 [01:45<1:27:20, 2.02it/s]
219
  1%| | 122/10682 [01:46<1:27:11, 2.02it/s]
220
  1%| | 123/10682 [01:46<1:27:11, 2.02it/s]
221
  1%| | 124/10682 [01:47<1:27:05, 2.02it/s]
222
  1%| | 125/10682 [01:47<1:27:06, 2.02it/s]
223
  {'loss': 7.5756, 'grad_norm': 0.5084755420684814, 'learning_rate': 0.00011693171188026193, 'epoch': 0.16}
224
+
225
  1%| | 125/10682 [01:47<1:27:06, 2.02it/s]
226
  1%| | 126/10682 [01:48<1:27:09, 2.02it/s]
227
  1%| | 127/10682 [01:48<1:27:06, 2.02it/s]
228
  1%| | 128/10682 [01:49<1:27:01, 2.02it/s]
229
  1%| | 129/10682 [01:49<1:27:01, 2.02it/s]
230
  1%| | 130/10682 [01:50<1:27:06, 2.02it/s]
231
  1%| | 131/10682 [01:50<1:27:10, 2.02it/s]
232
  1%| | 132/10682 [01:51<1:27:03, 2.02it/s]
233
  1%| | 133/10682 [01:51<1:27:05, 2.02it/s]
234
  1%|▏ | 134/10682 [01:52<1:27:03, 2.02it/s]
235
  1%|▏ | 135/10682 [01:52<1:27:10, 2.02it/s]
236
  1%|▏ | 136/10682 [01:53<1:27:10, 2.02it/s]
237
  1%|▏ | 137/10682 [01:53<1:27:17, 2.01it/s]
238
  1%|▏ | 138/10682 [01:54<1:27:07, 2.02it/s]
239
  1%|▏ | 139/10682 [01:54<1:27:05, 2.02it/s]
240
  1%|▏ | 140/10682 [01:55<1:27:03, 2.02it/s]
241
  1%|▏ | 141/10682 [01:55<1:27:07, 2.02it/s]
242
  1%|▏ | 142/10682 [01:56<1:27:08, 2.02it/s]
243
  1%|▏ | 143/10682 [01:56<1:27:04, 2.02it/s]
244
  1%|▏ | 144/10682 [01:57<1:27:06, 2.02it/s]
245
  1%|▏ | 145/10682 [01:57<1:26:59, 2.02it/s]
246
  1%|▏ | 146/10682 [01:58<1:26:59, 2.02it/s]
247
  1%|▏ | 147/10682 [01:58<1:26:54, 2.02it/s]
248
  1%|▏ | 148/10682 [01:59<1:26:55, 2.02it/s]
249
  1%|▏ | 149/10682 [01:59<1:26:48, 2.02it/s]
250
  1%|▏ | 150/10682 [02:00<1:26:54, 2.02it/s]{'loss': 7.1387, 'grad_norm': 0.3660012483596802, 'learning_rate': 0.0001403180542563143, 'epoch': 0.2}
251
 
252
+
253
  1%|▏ | 150/10682 [02:00<1:26:54, 2.02it/s]
254
  1%|▏ | 151/10682 [02:00<1:26:59, 2.02it/s]
255
  1%|▏ | 152/10682 [02:00<1:26:56, 2.02it/s]
256
  1%|▏ | 153/10682 [02:01<1:26:56, 2.02it/s]
257
  1%|▏ | 154/10682 [02:01<1:26:52, 2.02it/s]
258
  1%|▏ | 155/10682 [02:02<1:26:50, 2.02it/s]
259
  1%|▏ | 156/10682 [02:02<1:26:49, 2.02it/s]
260
  1%|▏ | 157/10682 [02:03<1:26:57, 2.02it/s]
261
  1%|▏ | 158/10682 [02:03<1:26:52, 2.02it/s]
262
  1%|▏ | 159/10682 [02:04<1:26:54, 2.02it/s]
263
  1%|▏ | 160/10682 [02:04<1:26:47, 2.02it/s]
264
  2%|▏ | 161/10682 [02:05<1:26:49, 2.02it/s]
265
  2%|▏ | 162/10682 [02:05<1:26:47, 2.02it/s]
266
  2%|▏ | 163/10682 [02:06<1:26:49, 2.02it/s]
267
  2%|▏ | 164/10682 [02:06<1:26:40, 2.02it/s]
268
  2%|▏ | 165/10682 [02:07<1:26:47, 2.02it/s]
269
  2%|▏ | 166/10682 [02:07<1:26:40, 2.02it/s]
270
  2%|▏ | 167/10682 [02:08<1:26:43, 2.02it/s]
271
  2%|▏ | 168/10682 [02:08<1:26:40, 2.02it/s]
272
  2%|▏ | 169/10682 [02:09<1:26:43, 2.02it/s]
273
  2%|▏ | 170/10682 [02:09<1:26:39, 2.02it/s]
274
  2%|▏ | 171/10682 [02:10<1:26:42, 2.02it/s]
275
  2%|▏ | 172/10682 [02:10<1:26:39, 2.02it/s]
276
  2%|▏ | 173/10682 [02:11<1:26:40, 2.02it/s]
277
  2%|▏ | 174/10682 [02:11<1:26:36, 2.02it/s]
278
  2%|▏ | 175/10682 [02:12<1:26:39, 2.02it/s]{'loss': 6.8255, 'grad_norm': 0.5133531093597412, 'learning_rate': 0.00016370439663236668, 'epoch': 0.23}
279
 
280
+
281
  2%|▏ | 175/10682 [02:12<1:26:39, 2.02it/s]
282
  2%|▏ | 176/10682 [02:12<1:26:46, 2.02it/s]
283
  2%|▏ | 177/10682 [02:13<1:26:44, 2.02it/s]
284
  2%|▏ | 178/10682 [02:13<1:26:43, 2.02it/s]
285
  2%|▏ | 179/10682 [02:14<1:26:39, 2.02it/s]
286
  2%|▏ | 180/10682 [02:14<1:26:40, 2.02it/s]
287
  2%|▏ | 181/10682 [02:15<1:26:38, 2.02it/s]
288
  2%|▏ | 182/10682 [02:15<1:26:38, 2.02it/s]
289
  2%|▏ | 183/10682 [02:16<1:26:35, 2.02it/s]
290
  2%|▏ | 184/10682 [02:16<1:26:41, 2.02it/s]
291
  2%|▏ | 185/10682 [02:17<1:26:41, 2.02it/s]
292
  2%|▏ | 186/10682 [02:17<1:26:47, 2.02it/s]
293
  2%|▏ | 187/10682 [02:18<1:26:48, 2.01it/s]
294
  2%|▏ | 188/10682 [02:18<1:26:48, 2.01it/s]
295
  2%|▏ | 189/10682 [02:19<1:26:48, 2.01it/s]
296
  2%|▏ | 190/10682 [02:19<1:26:43, 2.02it/s]
297
  2%|▏ | 191/10682 [02:20<1:26:41, 2.02it/s]
298
  2%|▏ | 192/10682 [02:20<1:26:34, 2.02it/s]
299
  2%|▏ | 193/10682 [02:21<1:26:31, 2.02it/s]
300
  2%|▏ | 194/10682 [02:21<1:26:29, 2.02it/s]
301
  2%|▏ | 195/10682 [02:22<1:26:33, 2.02it/s]
302
  2%|▏ | 196/10682 [02:22<1:26:26, 2.02it/s]
303
  2%|▏ | 197/10682 [02:23<1:26:26, 2.02it/s]
304
  2%|▏ | 198/10682 [02:23<1:26:23, 2.02it/s]
305
  2%|▏ | 199/10682 [02:24<1:26:28, 2.02it/s]
306
  2%|▏ | 200/10682 [02:24<1:26:25, 2.02it/s]
307
  {'loss': 6.575, 'grad_norm': 0.5841809511184692, 'learning_rate': 0.00018709073900841907, 'epoch': 0.26}
308
+
309
  2%|▏ | 200/10682 [02:24<1:26:25, 2.02it/s]
310
  2%|▏ | 201/10682 [02:25<1:26:32, 2.02it/s]
311
  2%|▏ | 202/10682 [02:25<1:26:25, 2.02it/s]
312
  2%|▏ | 203/10682 [02:26<1:26:29, 2.02it/s]
313
  2%|▏ | 204/10682 [02:26<1:26:28, 2.02it/s]
314
  2%|▏ | 205/10682 [02:27<1:26:31, 2.02it/s]
315
  2%|▏ | 206/10682 [02:27<1:26:26, 2.02it/s]
316
  2%|▏ | 207/10682 [02:28<1:26:29, 2.02it/s]
317
  2%|▏ | 208/10682 [02:28<1:26:24, 2.02it/s]
318
  2%|▏ | 209/10682 [02:29<1:26:27, 2.02it/s]
319
  2%|▏ | 210/10682 [02:29<1:26:20, 2.02it/s]
320
  2%|▏ | 211/10682 [02:30<1:26:22, 2.02it/s]
321
  2%|▏ | 212/10682 [02:30<1:26:19, 2.02it/s]
322
  2%|▏ | 213/10682 [02:31<1:26:20, 2.02it/s]
323
  2%|▏ | 214/10682 [02:31<1:26:19, 2.02it/s]
324
  2%|▏ | 215/10682 [02:32<1:26:26, 2.02it/s]
325
  2%|▏ | 216/10682 [02:32<1:26:21, 2.02it/s]
326
  2%|▏ | 217/10682 [02:33<1:26:21, 2.02it/s]
327
  2%|▏ | 218/10682 [02:33<1:26:17, 2.02it/s]
328
  2%|▏ | 219/10682 [02:34<1:26:22, 2.02it/s]
329
  2%|▏ | 220/10682 [02:34<1:26:15, 2.02it/s]
330
  2%|▏ | 221/10682 [02:35<1:26:18, 2.02it/s]
331
  2%|▏ | 222/10682 [02:35<1:26:15, 2.02it/s]
332
  2%|▏ | 223/10682 [02:36<1:26:21, 2.02it/s]
333
  2%|▏ | 224/10682 [02:36<1:26:22, 2.02it/s]
334
  2%|▏ | 225/10682 [02:37<1:26:18, 2.02it/s]{'loss': 6.3833, 'grad_norm': 0.6055994629859924, 'learning_rate': 0.00021047708138447147, 'epoch': 0.29}
335
+
336
 
337
  2%|▏ | 225/10682 [02:37<1:26:18, 2.02it/s]
338
  2%|▏ | 226/10682 [02:37<1:26:30, 2.01it/s]
339
  2%|▏ | 227/10682 [02:38<1:26:25, 2.02it/s]
340
  2%|▏ | 228/10682 [02:38<1:26:22, 2.02it/s]
341
  2%|▏ | 229/10682 [02:39<1:26:15, 2.02it/s]
342
  2%|▏ | 230/10682 [02:39<1:26:13, 2.02it/s]
343
  2%|▏ | 231/10682 [02:40<1:26:08, 2.02it/s]
344
  2%|▏ | 232/10682 [02:40<1:26:18, 2.02it/s]
345
  2%|▏ | 233/10682 [02:41<1:26:12, 2.02it/s]
346
  2%|▏ | 234/10682 [02:41<1:26:14, 2.02it/s]
347
  2%|▏ | 235/10682 [02:42<1:26:09, 2.02it/s]
348
  2%|▏ | 236/10682 [02:42<1:26:12, 2.02it/s]
349
  2%|▏ | 237/10682 [02:43<1:26:08, 2.02it/s]
350
  2%|▏ | 238/10682 [02:43<1:26:07, 2.02it/s]
351
  2%|▏ | 239/10682 [02:44<1:26:07, 2.02it/s]
352
  2%|▏ | 240/10682 [02:44<1:26:08, 2.02it/s]
353
  2%|▏ | 241/10682 [02:45<1:26:10, 2.02it/s]
354
  2%|▏ | 242/10682 [02:45<1:26:06, 2.02it/s]
355
  2%|▏ | 243/10682 [02:46<1:26:03, 2.02it/s]
356
  2%|▏ | 244/10682 [02:46<1:26:05, 2.02it/s]
357
  2%|▏ | 245/10682 [02:47<1:26:01, 2.02it/s]
358
  2%|▏ | 246/10682 [02:47<1:26:06, 2.02it/s]
359
  2%|▏ | 247/10682 [02:48<1:25:59, 2.02it/s]
360
  2%|▏ | 248/10682 [02:48<1:26:03, 2.02it/s]
361
  2%|▏ | 249/10682 [02:49<1:25:58, 2.02it/s]
362
  2%|▏ | 250/10682 [02:49<1:26:01, 2.02it/s]
363
  {'loss': 6.2106, 'grad_norm': 0.6558924317359924, 'learning_rate': 0.00023386342376052386, 'epoch': 0.33}
364
+
365
  2%|▏ | 250/10682 [02:49<1:26:01, 2.02it/s]
366
  2%|▏ | 251/10682 [02:50<1:26:05, 2.02it/s]
367
  2%|▏ | 252/10682 [02:50<1:26:04, 2.02it/s]
368
  2%|▏ | 253/10682 [02:50<1:25:59, 2.02it/s]
369
  2%|▏ | 254/10682 [02:51<1:26:01, 2.02it/s]
370
  2%|▏ | 255/10682 [02:51<1:26:08, 2.02it/s]
371
  2%|▏ | 256/10682 [02:52<1:26:12, 2.02it/s]
372
  2%|▏ | 257/10682 [02:52<1:26:06, 2.02it/s]
373
  2%|▏ | 258/10682 [02:53<1:26:06, 2.02it/s]
374
  2%|▏ | 259/10682 [02:53<1:26:00, 2.02it/s]
375
  2%|▏ | 260/10682 [02:54<1:26:02, 2.02it/s]
376
  2%|▏ | 261/10682 [02:54<1:25:58, 2.02it/s]
377
  2%|▏ | 262/10682 [02:55<1:26:05, 2.02it/s]
378
  2%|▏ | 263/10682 [02:55<1:25:59, 2.02it/s]
379
  2%|▏ | 264/10682 [02:56<1:25:53, 2.02it/s]
380
  2%|▏ | 265/10682 [02:56<1:25:55, 2.02it/s]
381
  2%|▏ | 266/10682 [02:57<1:25:50, 2.02it/s]
382
  2%|▏ | 267/10682 [02:57<1:25:54, 2.02it/s]
383
  3%|▎ | 268/10682 [02:58<1:25:53, 2.02it/s]
384
  3%|▎ | 269/10682 [02:58<1:25:55, 2.02it/s]
385
  3%|▎ | 270/10682 [02:59<1:25:54, 2.02it/s]
386
  3%|▎ | 271/10682 [02:59<1:25:59, 2.02it/s]
387
  3%|▎ | 272/10682 [03:00<1:25:56, 2.02it/s]
388
  3%|▎ | 273/10682 [03:00<1:25:55, 2.02it/s]
389
  3%|▎ | 274/10682 [03:01<1:25:55, 2.02it/s]
390
  3%|▎ | 275/10682 [03:01<1:25:50, 2.02it/s]
391
  {'loss': 6.0743, 'grad_norm': 0.5830900073051453, 'learning_rate': 0.00025724976613657625, 'epoch': 0.36}
392
+
393
  3%|▎ | 275/10682 [03:01<1:25:50, 2.02it/s]
394
  3%|▎ | 276/10682 [03:02<1:25:52, 2.02it/s]
395
  3%|▎ | 277/10682 [03:02<1:25:54, 2.02it/s]
396
  3%|▎ | 278/10682 [03:03<1:25:51, 2.02it/s]
397
  3%|▎ | 279/10682 [03:03<1:25:50, 2.02it/s]
398
  3%|▎ | 280/10682 [03:04<1:25:44, 2.02it/s]
399
  3%|▎ | 281/10682 [03:04<1:25:47, 2.02it/s]
400
  3%|▎ | 282/10682 [03:05<1:25:45, 2.02it/s]
401
  3%|▎ | 283/10682 [03:05<1:25:46, 2.02it/s]
402
  3%|▎ | 284/10682 [03:06<1:25:44, 2.02it/s]
403
  3%|▎ | 285/10682 [03:06<1:25:45, 2.02it/s]
404
  3%|▎ | 286/10682 [03:07<1:25:41, 2.02it/s]
405
  3%|▎ | 287/10682 [03:07<1:25:42, 2.02it/s]
406
  3%|▎ | 288/10682 [03:08<1:25:37, 2.02it/s]
407
  3%|▎ | 289/10682 [03:08<1:25:40, 2.02it/s]
408
  3%|▎ | 290/10682 [03:09<1:25:36, 2.02it/s]
409
  3%|▎ | 291/10682 [03:09<1:25:39, 2.02it/s]
410
  3%|▎ | 292/10682 [03:10<1:25:39, 2.02it/s]
411
  3%|▎ | 293/10682 [03:10<1:25:41, 2.02it/s]
412
  3%|▎ | 294/10682 [03:11<1:25:39, 2.02it/s]
413
  3%|▎ | 295/10682 [03:11<1:25:39, 2.02it/s]
414
  3%|▎ | 296/10682 [03:12<1:25:43, 2.02it/s]
415
  3%|▎ | 297/10682 [03:12<1:25:43, 2.02it/s]
416
  3%|▎ | 298/10682 [03:13<1:25:41, 2.02it/s]
417
  3%|▎ | 299/10682 [03:13<1:25:35, 2.02it/s]
418
  3%|▎ | 300/10682 [03:14<1:25:40, 2.02it/s]
419
  {'loss': 5.9526, 'grad_norm': 0.7878416180610657, 'learning_rate': 0.0002806361085126286, 'epoch': 0.39}
420
+
421
  3%|▎ | 300/10682 [03:14<1:25:40, 2.02it/s]
422
  3%|▎ | 301/10682 [03:14<1:25:41, 2.02it/s]
423
  3%|▎ | 302/10682 [03:15<1:25:44, 2.02it/s]
424
  3%|▎ | 303/10682 [03:15<1:25:36, 2.02it/s]
425
  3%|▎ | 304/10682 [03:16<1:25:38, 2.02it/s]
426
  3%|▎ | 305/10682 [03:16<1:25:34, 2.02it/s]
427
  3%|▎ | 306/10682 [03:17<1:25:36, 2.02it/s]
428
  3%|▎ | 307/10682 [03:17<1:25:31, 2.02it/s]
429
  3%|▎ | 308/10682 [03:18<1:25:32, 2.02it/s]
430
  3%|▎ | 309/10682 [03:18<1:25:30, 2.02it/s]
431
  3%|▎ | 310/10682 [03:19<1:25:34, 2.02it/s]
432
  3%|▎ | 311/10682 [03:19<1:25:30, 2.02it/s]
433
  3%|▎ | 312/10682 [03:20<1:25:32, 2.02it/s]
434
  3%|▎ | 313/10682 [03:20<1:25:30, 2.02it/s]
435
  3%|▎ | 314/10682 [03:21<1:25:30, 2.02it/s]
436
  3%|▎ | 315/10682 [03:21<1:25:21, 2.02it/s]
437
  3%|▎ | 316/10682 [03:22<1:25:30, 2.02it/s]
438
  3%|▎ | 317/10682 [03:22<1:25:30, 2.02it/s]
439
  3%|▎ | 318/10682 [03:23<1:25:27, 2.02it/s]
440
  3%|▎ | 319/10682 [03:23<1:25:21, 2.02it/s]
441
  3%|▎ | 320/10682 [03:24<1:25:24, 2.02it/s]
442
  3%|▎ | 321/10682 [03:24<1:25:21, 2.02it/s]
443
  3%|▎ | 322/10682 [03:25<1:25:26, 2.02it/s]
444
  3%|▎ | 323/10682 [03:25<1:25:38, 2.02it/s]
445
  3%|▎ | 324/10682 [03:26<1:25:45, 2.01it/s]
446
  3%|▎ | 325/10682 [03:26<1:25:41, 2.01it/s]
447
  {'loss': 5.844, 'grad_norm': 0.8623385429382324, 'learning_rate': 0.00030402245088868103, 'epoch': 0.43}
448
+
449
  3%|▎ | 325/10682 [03:26<1:25:41, 2.01it/s]
450
  3%|▎ | 326/10682 [03:27<1:25:42, 2.01it/s]
451
  3%|▎ | 327/10682 [03:27<1:25:37, 2.02it/s]
452
  3%|▎ | 328/10682 [03:28<1:25:31, 2.02it/s]
453
  3%|▎ | 329/10682 [03:28<1:25:30, 2.02it/s]
454
  3%|▎ | 330/10682 [03:29<1:25:23, 2.02it/s]
455
  3%|▎ | 331/10682 [03:29<1:25:26, 2.02it/s]
456
  3%|▎ | 332/10682 [03:30<1:25:18, 2.02it/s]
457
  3%|▎ | 333/10682 [03:30<1:25:24, 2.02it/s]
458
  3%|▎ | 334/10682 [03:31<1:25:18, 2.02it/s]
459
  3%|▎ | 335/10682 [03:31<1:25:23, 2.02it/s]
460
  3%|▎ | 336/10682 [03:32<1:25:18, 2.02it/s]
461
  3%|▎ | 337/10682 [03:32<1:25:18, 2.02it/s]
462
  3%|▎ | 338/10682 [03:33<1:25:15, 2.02it/s]
463
  3%|▎ | 339/10682 [03:33<1:25:21, 2.02it/s]
464
  3%|▎ | 340/10682 [03:34<1:25:21, 2.02it/s]
465
  3%|▎ | 341/10682 [03:34<1:25:23, 2.02it/s]
466
  3%|▎ | 342/10682 [03:35<1:25:18, 2.02it/s]
467
  3%|▎ | 343/10682 [03:35<1:25:20, 2.02it/s]
468
  3%|▎ | 344/10682 [03:36<1:25:16, 2.02it/s]
469
  3%|▎ | 345/10682 [03:36<1:25:19, 2.02it/s]
470
  3%|▎ | 346/10682 [03:37<1:25:15, 2.02it/s]
471
  3%|▎ | 347/10682 [03:37<1:25:15, 2.02it/s]
472
  3%|▎ | 348/10682 [03:38<1:25:13, 2.02it/s]
473
  3%|▎ | 349/10682 [03:38<1:25:11, 2.02it/s]
474
  3%|▎ | 350/10682 [03:39<1:25:09, 2.02it/s]
475
 
476
+
477
  3%|▎ | 350/10682 [03:39<1:25:09, 2.02it/s]
478
  3%|▎ | 351/10682 [03:39<1:25:19, 2.02it/s]
479
  3%|▎ | 352/10682 [03:40<1:25:16, 2.02it/s]
480
  3%|▎ | 353/10682 [03:40<1:25:16, 2.02it/s]
481
  3%|▎ | 354/10682 [03:40<1:25:10, 2.02it/s]
482
  3%|▎ | 355/10682 [03:41<1:25:13, 2.02it/s]
483
  3%|▎ | 356/10682 [03:41<1:25:07, 2.02it/s]
484
  3%|▎ | 357/10682 [03:42<1:25:10, 2.02it/s]
485
  3%|▎ | 358/10682 [03:42<1:25:05, 2.02it/s]
486
  3%|▎ | 359/10682 [03:43<1:25:08, 2.02it/s]
487
  3%|▎ | 360/10682 [03:43<1:25:02, 2.02it/s]
488
  3%|▎ | 361/10682 [03:44<1:25:06, 2.02it/s]
489
  3%|▎ | 362/10682 [03:44<1:25:04, 2.02it/s]
490
  3%|▎ | 363/10682 [03:45<1:25:12, 2.02it/s]
491
  3%|▎ | 364/10682 [03:45<1:25:15, 2.02it/s]
492
  3%|▎ | 365/10682 [03:46<1:25:12, 2.02it/s]
493
  3%|▎ | 366/10682 [03:46<1:25:14, 2.02it/s]
494
  3%|▎ | 367/10682 [03:47<1:25:09, 2.02it/s]
495
  3%|▎ | 368/10682 [03:47<1:25:13, 2.02it/s]
496
  3%|▎ | 369/10682 [03:48<1:25:08, 2.02it/s]
497
  3%|▎ | 370/10682 [03:48<1:25:11, 2.02it/s]
498
  3%|▎ | 371/10682 [03:49<1:25:08, 2.02it/s]
499
  3%|▎ | 372/10682 [03:49<1:25:13, 2.02it/s]
500
  3%|▎ | 373/10682 [03:50<1:25:09, 2.02it/s]
501
  4%|▎ | 374/10682 [03:50<1:25:07, 2.02it/s]
502
  4%|▎ | 375/10682 [03:51<1:25:00, 2.02it/s]
503
  {'loss': 5.6455, 'grad_norm': 0.7050163745880127, 'learning_rate': 0.0003507951356407858, 'epoch': 0.49}
504
+
505
  4%|▎ | 375/10682 [03:51<1:25:00, 2.02it/s]
506
  4%|▎ | 376/10682 [03:51<1:25:03, 2.02it/s]
507
  4%|▎ | 377/10682 [03:52<1:24:59, 2.02it/s]
508
  4%|▎ | 378/10682 [03:52<1:25:00, 2.02it/s]
509
  4%|▎ | 379/10682 [03:53<1:24:53, 2.02it/s]
510
  4%|▎ | 380/10682 [03:53<1:24:56, 2.02it/s]
511
  4%|▎ | 381/10682 [03:54<1:24:56, 2.02it/s]
512
  4%|▎ | 382/10682 [03:54<1:24:57, 2.02it/s]
513
  4%|▎ | 383/10682 [03:55<1:24:55, 2.02it/s]
514
  4%|▎ | 384/10682 [03:55<1:24:57, 2.02it/s]
515
  4%|▎ | 385/10682 [03:56<1:24:59, 2.02it/s]
516
  4%|▎ | 386/10682 [03:56<1:25:00, 2.02it/s]
517
  4%|▎ | 387/10682 [03:57<1:24:56, 2.02it/s]
518
  4%|▎ | 388/10682 [03:57<1:24:59, 2.02it/s]
519
  4%|▎ | 389/10682 [03:58<1:24:53, 2.02it/s]
520
  4%|▎ | 390/10682 [03:58<1:24:59, 2.02it/s]
521
  4%|▎ | 391/10682 [03:59<1:24:52, 2.02it/s]
522
  4%|▎ | 392/10682 [03:59<1:24:52, 2.02it/s]
523
  4%|▎ | 393/10682 [04:00<1:24:51, 2.02it/s]
524
  4%|▎ | 394/10682 [04:00<1:25:10, 2.01it/s]
525
  4%|▎ | 395/10682 [04:01<1:24:59, 2.02it/s]
526
  4%|▎ | 396/10682 [04:01<1:25:00, 2.02it/s]
527
  4%|▎ | 397/10682 [04:02<1:24:55, 2.02it/s]
528
  4%|▎ | 398/10682 [04:02<1:25:06, 2.01it/s]
529
  4%|▎ | 399/10682 [04:03<1:25:01, 2.02it/s]
530
  4%|▎ | 400/10682 [04:03<1:24:58, 2.02it/s]
531
  {'loss': 5.5555, 'grad_norm': 0.6004089713096619, 'learning_rate': 0.00037418147801683815, 'epoch': 0.52}
532
+
533
  4%|▎ | 400/10682 [04:03<1:24:58, 2.02it/s]
534
  4%|▍ | 401/10682 [04:04<1:25:01, 2.02it/s]
535
  4%|▍ | 402/10682 [04:04<1:24:57, 2.02it/s]
536
  4%|▍ | 403/10682 [04:05<1:24:53, 2.02it/s]
537
  4%|▍ | 404/10682 [04:05<1:24:47, 2.02it/s]
538
  4%|▍ | 405/10682 [04:06<1:24:47, 2.02it/s]
539
  4%|▍ | 406/10682 [04:06<1:24:44, 2.02it/s]
540
  4%|▍ | 407/10682 [04:07<1:24:44, 2.02it/s]
541
  4%|▍ | 408/10682 [04:07<1:24:42, 2.02it/s]
542
  4%|▍ | 409/10682 [04:08<1:24:45, 2.02it/s]
543
  4%|▍ | 410/10682 [04:08<1:24:38, 2.02it/s]
544
  4%|▍ | 411/10682 [04:09<1:24:40, 2.02it/s]
545
  4%|▍ | 412/10682 [04:09<1:24:38, 2.02it/s]
546
  4%|▍ | 413/10682 [04:10<1:24:42, 2.02it/s]
547
  4%|▍ | 414/10682 [04:10<1:24:35, 2.02it/s]
548
  4%|▍ | 415/10682 [04:11<1:24:39, 2.02it/s]
549
  4%|▍ | 416/10682 [04:11<1:24:34, 2.02it/s]
550
  4%|▍ | 417/10682 [04:12<1:24:38, 2.02it/s]
551
  4%|▍ | 418/10682 [04:12<1:24:34, 2.02it/s]
552
  4%|▍ | 419/10682 [04:13<1:24:38, 2.02it/s]
553
  4%|▍ | 420/10682 [04:13<1:24:37, 2.02it/s]
554
  4%|▍ | 421/10682 [04:14<1:24:44, 2.02it/s]
555
  4%|▍ | 422/10682 [04:14<1:24:44, 2.02it/s]
556
  4%|▍ | 423/10682 [04:15<1:24:39, 2.02it/s]
557
  4%|▍ | 424/10682 [04:15<1:24:39, 2.02it/s]
558
  4%|▍ | 425/10682 [04:16<1:24:37, 2.02it/s]
559
 
560
+
561
  4%|▍ | 425/10682 [04:16<1:24:37, 2.02it/s]
562
  4%|▍ | 426/10682 [04:16<1:24:48, 2.02it/s]
563
  4%|▍ | 427/10682 [04:17<1:24:37, 2.02it/s]
564
  4%|▍ | 428/10682 [04:17<1:24:40, 2.02it/s]
565
  4%|▍ | 429/10682 [04:18<1:24:35, 2.02it/s]
566
  4%|▍ | 430/10682 [04:18<1:24:36, 2.02it/s]
567
  4%|▍ | 431/10682 [04:19<1:24:33, 2.02it/s]
568
  4%|▍ | 432/10682 [04:19<1:24:38, 2.02it/s]
569
  4%|▍ | 433/10682 [04:20<1:24:44, 2.02it/s]
570
  4%|▍ | 434/10682 [04:20<1:24:32, 2.02it/s]
571
  4%|▍ | 435/10682 [04:21<1:24:30, 2.02it/s]
572
  4%|▍ | 436/10682 [04:21<1:24:30, 2.02it/s]
573
  4%|▍ | 437/10682 [04:22<1:24:30, 2.02it/s]
574
  4%|▍ | 438/10682 [04:22<1:24:27, 2.02it/s]
575
  4%|▍ | 439/10682 [04:23<1:24:30, 2.02it/s]
576
  4%|▍ | 440/10682 [04:23<1:24:27, 2.02it/s]
577
  4%|▍ | 441/10682 [04:24<1:24:24, 2.02it/s]
578
  4%|▍ | 442/10682 [04:24<1:24:25, 2.02it/s]
579
  4%|▍ | 443/10682 [04:25<1:24:27, 2.02it/s]
580
  4%|▍ | 444/10682 [04:25<1:24:27, 2.02it/s]
581
  4%|▍ | 445/10682 [04:26<1:24:29, 2.02it/s]
582
  4%|▍ | 446/10682 [04:26<1:24:22, 2.02it/s]
583
  4%|▍ | 447/10682 [04:27<1:24:26, 2.02it/s]
584
  4%|▍ | 448/10682 [04:27<1:24:24, 2.02it/s]
585
  4%|▍ | 449/10682 [04:28<1:24:26, 2.02it/s]
586
  4%|▍ | 450/10682 [04:28<1:24:21, 2.02it/s]
587
  {'loss': 5.399, 'grad_norm': 0.6876851916313171, 'learning_rate': 0.00042095416276894293, 'epoch': 0.59}
588
+
589
  4%|▍ | 450/10682 [04:28<1:24:21, 2.02it/s]
590
  4%|▍ | 451/10682 [04:29<1:24:29, 2.02it/s]
591
  4%|▍ | 452/10682 [04:29<1:24:23, 2.02it/s]
592
  4%|▍ | 453/10682 [04:30<1:24:24, 2.02it/s]
593
  4%|▍ | 454/10682 [04:30<1:24:18, 2.02it/s]
594
  4%|▍ | 455/10682 [04:31<1:24:24, 2.02it/s]
595
  4%|▍ | 456/10682 [04:31<1:24:24, 2.02it/s]
596
  4%|▍ | 457/10682 [04:31<1:24:29, 2.02it/s]
597
  4%|▍ | 458/10682 [04:32<1:24:24, 2.02it/s]
598
  4%|▍ | 459/10682 [04:32<1:24:21, 2.02it/s]
599
  4%|▍ | 460/10682 [04:33<1:24:16, 2.02it/s]
600
  4%|▍ | 461/10682 [04:33<1:24:18, 2.02it/s]
601
  4%|▍ | 462/10682 [04:34<1:24:10, 2.02it/s]
602
  4%|▍ | 463/10682 [04:34<1:24:10, 2.02it/s]
603
  4%|▍ | 464/10682 [04:35<1:24:10, 2.02it/s]
604
  4%|▍ | 465/10682 [04:35<1:24:19, 2.02it/s]
605
  4%|▍ | 466/10682 [04:36<1:24:17, 2.02it/s]
606
  4%|▍ | 467/10682 [04:36<1:24:17, 2.02it/s]
607
  4%|▍ | 468/10682 [04:37<1:24:17, 2.02it/s]
608
  4%|▍ | 469/10682 [04:37<1:24:19, 2.02it/s]
609
  4%|▍ | 470/10682 [04:38<1:24:10, 2.02it/s]
610
  4%|▍ | 471/10682 [04:38<1:24:15, 2.02it/s]
611
  4%|▍ | 472/10682 [04:39<1:24:06, 2.02it/s]
612
  4%|▍ | 473/10682 [04:39<1:24:09, 2.02it/s]
613
  4%|▍ | 474/10682 [04:40<1:24:08, 2.02it/s]
614
  4%|▍ | 475/10682 [04:40<1:24:13, 2.02it/s]{'loss': 5.327, 'grad_norm': 0.6345270872116089, 'learning_rate': 0.0004443405051449954, 'epoch': 0.62}
615
 
616
+
617
  4%|▍ | 475/10682 [04:40<1:24:13, 2.02it/s]
618
  4%|▍ | 476/10682 [04:41<1:24:18, 2.02it/s]
619
  4%|▍ | 477/10682 [04:41<1:24:12, 2.02it/s]
620
  4%|▍ | 478/10682 [04:42<1:24:09, 2.02it/s]
621
  4%|▍ | 479/10682 [04:42<1:24:14, 2.02it/s]
622
  4%|▍ | 480/10682 [04:43<1:24:14, 2.02it/s]
623
  5%|▍ | 481/10682 [04:43<1:24:13, 2.02it/s]
624
  5%|▍ | 482/10682 [04:44<1:24:08, 2.02it/s]
625
  5%|▍ | 483/10682 [04:44<1:24:09, 2.02it/s]
626
  5%|▍ | 484/10682 [04:45<1:24:05, 2.02it/s]
627
  5%|▍ | 485/10682 [04:45<1:24:07, 2.02it/s]
628
  5%|▍ | 486/10682 [04:46<1:24:07, 2.02it/s]
629
  5%|▍ | 487/10682 [04:46<1:24:05, 2.02it/s]
630
  5%|▍ | 488/10682 [04:47<1:24:06, 2.02it/s]
631
  5%|▍ | 489/10682 [04:47<1:24:09, 2.02it/s]
632
  5%|▍ | 490/10682 [04:48<1:24:04, 2.02it/s]
633
  5%|▍ | 491/10682 [04:48<1:24:06, 2.02it/s]
634
  5%|▍ | 492/10682 [04:49<1:24:04, 2.02it/s]
635
  5%|▍ | 493/10682 [04:49<1:24:05, 2.02it/s]
636
  5%|▍ | 494/10682 [04:50<1:24:00, 2.02it/s]
637
  5%|▍ | 495/10682 [04:50<1:24:01, 2.02it/s]
638
  5%|▍ | 496/10682 [04:51<1:23:57, 2.02it/s]
639
  5%|▍ | 497/10682 [04:51<1:24:01, 2.02it/s]
640
  5%|▍ | 498/10682 [04:52<1:23:59, 2.02it/s]
641
  5%|▍ | 499/10682 [04:52<1:23:56, 2.02it/s]
642
  5%|▍ | 500/10682 [04:53<1:23:56, 2.02it/s]
643
 
644
+
645
  5%|▍ | 500/10682 [04:53<1:23:56, 2.02it/s]
646
  5%|▍ | 501/10682 [04:53<1:24:03, 2.02it/s]
647
  5%|▍ | 502/10682 [04:54<1:24:03, 2.02it/s]
648
  5%|▍ | 503/10682 [04:54<1:24:03, 2.02it/s]
649
  5%|▍ | 504/10682 [04:55<1:24:02, 2.02it/s]
650
  5%|▍ | 505/10682 [04:55<1:24:01, 2.02it/s]
651
  5%|▍ | 506/10682 [04:56<1:24:08, 2.02it/s]
652
  5%|▍ | 507/10682 [04:56<1:24:01, 2.02it/s]
653
  5%|▍ | 508/10682 [04:57<1:24:02, 2.02it/s]
654
  5%|▍ | 509/10682 [04:57<1:23:56, 2.02it/s]
655
  5%|▍ | 510/10682 [04:58<1:24:00, 2.02it/s]
656
  5%|▍ | 511/10682 [04:58<1:23:55, 2.02it/s]
657
  5%|▍ | 512/10682 [04:59<1:23:58, 2.02it/s]
658
  5%|▍ | 513/10682 [04:59<1:23:53, 2.02it/s]
659
  5%|▍ | 514/10682 [05:00<1:24:01, 2.02it/s]
660
  5%|▍ | 515/10682 [05:00<1:23:56, 2.02it/s]
661
  5%|▍ | 516/10682 [05:01<1:23:54, 2.02it/s]
662
  5%|▍ | 517/10682 [05:01<1:23:53, 2.02it/s]
663
  5%|▍ | 518/10682 [05:02<1:23:51, 2.02it/s]
664
  5%|▍ | 519/10682 [05:02<1:23:51, 2.02it/s]
665
  5%|▍ | 520/10682 [05:03<1:23:51, 2.02it/s]
666
  5%|▍ | 521/10682 [05:03<1:23:45, 2.02it/s]
667
  5%|▍ | 522/10682 [05:04<1:23:47, 2.02it/s]
668
  5%|▍ | 523/10682 [05:04<1:23:41, 2.02it/s]
669
  5%|▍ | 524/10682 [05:05<1:23:46, 2.02it/s]
670
  5%|▍ | 525/10682 [05:05<1:23:42, 2.02it/s]{'loss': 5.2021, 'grad_norm': 0.6268776059150696, 'learning_rate': 0.0004911131898971, 'epoch': 0.69}
671
+
672
 
673
  5%|▍ | 525/10682 [05:05<1:23:42, 2.02it/s]
674
  5%|▍ | 526/10682 [05:06<1:23:54, 2.02it/s]
675
  5%|▍ | 527/10682 [05:06<1:23:49, 2.02it/s]
676
  5%|▍ | 528/10682 [05:07<1:23:46, 2.02it/s]
677
  5%|▍ | 529/10682 [05:07<1:23:43, 2.02it/s]
678
  5%|▍ | 530/10682 [05:08<1:23:43, 2.02it/s]
679
  5%|▍ | 531/10682 [05:08<1:23:44, 2.02it/s]
680
  5%|▍ | 532/10682 [05:09<1:23:40, 2.02it/s]
681
  5%|▍ | 533/10682 [05:09<1:23:42, 2.02it/s]
682
  5%|▍ | 534/10682 [05:10<1:24:01, 2.01it/s]
683
  5%|▌ | 535/10682 [05:10<1:24:00, 2.01it/s]
684
  5%|▌ | 536/10682 [05:11<1:24:02, 2.01it/s]
685
  5%|▌ | 537/10682 [05:11<1:24:01, 2.01it/s]
686
  5%|▌ | 538/10682 [05:12<1:23:59, 2.01it/s]
687
  5%|▌ | 539/10682 [05:12<1:23:52, 2.02it/s]
688
  5%|▌ | 540/10682 [05:13<1:23:55, 2.01it/s]
689
  5%|▌ | 541/10682 [05:13<1:23:45, 2.02it/s]
690
  5%|▌ | 542/10682 [05:14<1:23:43, 2.02it/s]
691
  5%|▌ | 543/10682 [05:14<1:23:35, 2.02it/s]
692
  5%|▌ | 544/10682 [05:15<1:23:34, 2.02it/s]
693
  5%|▌ | 545/10682 [05:15<1:23:32, 2.02it/s]
694
  5%|▌ | 546/10682 [05:16<1:23:33, 2.02it/s]
695
  5%|▌ | 547/10682 [05:16<1:23:31, 2.02it/s]
696
  5%|▌ | 548/10682 [05:17<1:23:36, 2.02it/s]
697
  5%|▌ | 549/10682 [05:17<1:23:34, 2.02it/s]
698
  5%|▌ | 550/10682 [05:18<1:23:38, 2.02it/s]
699
  {'loss': 5.1481, 'grad_norm': 0.6789132952690125, 'learning_rate': 0.0005144995322731525, 'epoch': 0.72}
700
+
701
  5%|▌ | 550/10682 [05:18<1:23:38, 2.02it/s]
702
  5%|▌ | 551/10682 [05:18<1:23:44, 2.02it/s]
703
  5%|▌ | 552/10682 [05:19<1:23:41, 2.02it/s]
704
  5%|▌ | 553/10682 [05:19<1:23:36, 2.02it/s]
705
  5%|▌ | 554/10682 [05:20<1:23:36, 2.02it/s]
706
  5%|▌ | 555/10682 [05:20<1:23:33, 2.02it/s]
707
  5%|▌ | 556/10682 [05:21<1:23:33, 2.02it/s]
708
  5%|▌ | 557/10682 [05:21<1:23:37, 2.02it/s]
709
  5%|▌ | 558/10682 [05:22<1:23:32, 2.02it/s]
710
  5%|▌ | 559/10682 [05:22<1:23:33, 2.02it/s]
711
  5%|▌ | 560/10682 [05:23<1:23:31, 2.02it/s]
712
  5%|▌ | 561/10682 [05:23<1:23:35, 2.02it/s]
713
  5%|▌ | 562/10682 [05:23<1:23:29, 2.02it/s]
714
  5%|▌ | 563/10682 [05:24<1:23:32, 2.02it/s]
715
  5%|▌ | 564/10682 [05:24<1:23:24, 2.02it/s]
716
  5%|▌ | 565/10682 [05:25<1:23:28, 2.02it/s]
717
  5%|▌ | 566/10682 [05:25<1:23:22, 2.02it/s]
718
  5%|▌ | 567/10682 [05:26<1:23:25, 2.02it/s]
719
  5%|▌ | 568/10682 [05:26<1:23:22, 2.02it/s]
720
  5%|▌ | 569/10682 [05:27<1:23:23, 2.02it/s]
721
  5%|▌ | 570/10682 [05:27<1:23:21, 2.02it/s]
722
  5%|▌ | 571/10682 [05:28<1:23:21, 2.02it/s]
723
  5%|▌ | 572/10682 [05:28<1:23:26, 2.02it/s]
724
  5%|▌ | 573/10682 [05:29<1:23:26, 2.02it/s]
725
  5%|▌ | 574/10682 [05:29<1:23:24, 2.02it/s]
726
  5%|▌ | 575/10682 [05:30<1:23:21, 2.02it/s]{'loss': 5.0923, 'grad_norm': 0.5607177019119263, 'learning_rate': 0.0005378858746492049, 'epoch': 0.75}
727
+
728
 
729
  5%|▌ | 575/10682 [05:30<1:23:21, 2.02it/s]
730
  5%|▌ | 576/10682 [05:30<1:23:33, 2.02it/s]
731
  5%|▌ | 577/10682 [05:31<1:23:23, 2.02it/s]
732
  5%|▌ | 578/10682 [05:31<1:23:26, 2.02it/s]
733
  5%|▌ | 579/10682 [05:32<1:23:17, 2.02it/s]
734
  5%|▌ | 580/10682 [05:32<1:23:20, 2.02it/s]
735
  5%|▌ | 581/10682 [05:33<1:23:18, 2.02it/s]
736
  5%|▌ | 582/10682 [05:33<1:23:22, 2.02it/s]
737
  5%|▌ | 583/10682 [05:34<1:23:21, 2.02it/s]
738
  5%|▌ | 584/10682 [05:34<1:23:24, 2.02it/s]
739
  5%|▌ | 585/10682 [05:35<1:23:22, 2.02it/s]
740
  5%|▌ | 586/10682 [05:35<1:23:18, 2.02it/s]
741
  5%|▌ | 587/10682 [05:36<1:23:16, 2.02it/s]
742
  6%|▌ | 588/10682 [05:36<1:23:17, 2.02it/s]
743
  6%|▌ | 589/10682 [05:37<1:23:13, 2.02it/s]
744
  6%|▌ | 590/10682 [05:37<1:23:15, 2.02it/s]
745
  6%|▌ | 591/10682 [05:38<1:23:16, 2.02it/s]
746
  6%|▌ | 592/10682 [05:38<1:23:14, 2.02it/s]
747
  6%|▌ | 593/10682 [05:39<1:23:12, 2.02it/s]
748
  6%|▌ | 594/10682 [05:39<1:23:12, 2.02it/s]
749
  6%|▌ | 595/10682 [05:40<1:23:09, 2.02it/s]
750
  6%|▌ | 596/10682 [05:40<1:23:09, 2.02it/s]
751
  6%|▌ | 597/10682 [05:41<1:23:06, 2.02it/s]
752
  6%|▌ | 598/10682 [05:41<1:23:09, 2.02it/s]
753
  6%|▌ | 599/10682 [05:42<1:23:06, 2.02it/s]
754
  6%|▌ | 600/10682 [05:42<1:23:10, 2.02it/s]
755
 
756
+
757
  6%|▌ | 600/10682 [05:42<1:23:10, 2.02it/s]
758
  6%|▌ | 601/10682 [05:43<1:23:21, 2.02it/s]
759
  6%|▌ | 602/10682 [05:43<1:23:12, 2.02it/s]
760
  6%|▌ | 603/10682 [05:44<1:23:08, 2.02it/s]
761
  6%|▌ | 604/10682 [05:44<1:23:12, 2.02it/s]
762
  6%|▌ | 605/10682 [05:45<1:23:25, 2.01it/s]
763
  6%|▌ | 606/10682 [05:45<1:23:21, 2.01it/s]
764
  6%|▌ | 607/10682 [05:46<1:23:20, 2.01it/s]
765
  6%|▌ | 608/10682 [05:46<1:23:11, 2.02it/s]
766
  6%|▌ | 609/10682 [05:47<1:30:09, 1.86it/s]
767
  6%|▌ | 610/10682 [05:47<1:27:59, 1.91it/s]
768
  6%|▌ | 611/10682 [05:48<1:26:34, 1.94it/s]
769
  6%|▌ | 612/10682 [05:48<1:25:27, 1.96it/s]
770
  6%|▌ | 613/10682 [05:49<1:24:45, 1.98it/s]
771
  6%|▌ | 614/10682 [05:49<1:24:14, 1.99it/s]
772
  6%|▌ | 615/10682 [05:50<1:23:50, 2.00it/s]
773
  6%|▌ | 616/10682 [05:51<1:30:30, 1.85it/s]
774
  6%|▌ | 617/10682 [05:51<1:28:14, 1.90it/s]
775
  6%|▌ | 618/10682 [05:51<1:26:43, 1.93it/s]
776
  6%|▌ | 619/10682 [05:52<1:25:37, 1.96it/s]
777
  6%|▌ | 620/10682 [05:52<1:24:49, 1.98it/s]
778
  6%|▌ | 621/10682 [05:53<1:24:13, 1.99it/s]
779
  6%|▌ | 622/10682 [05:53<1:23:52, 2.00it/s]
780
  6%|▌ | 623/10682 [05:54<1:23:33, 2.01it/s]
781
  6%|▌ | 624/10682 [05:54<1:23:24, 2.01it/s]
782
  6%|▌ | 625/10682 [05:55<1:23:16, 2.01it/s]
783
 
784
+
785
  6%|▌ | 625/10682 [05:55<1:23:16, 2.01it/s]
786
  6%|▌ | 626/10682 [05:55<1:23:16, 2.01it/s]
787
  6%|▌ | 627/10682 [05:56<1:23:10, 2.01it/s]
788
  6%|▌ | 628/10682 [05:56<1:23:04, 2.02it/s]
789
  6%|▌ | 629/10682 [05:57<1:23:12, 2.01it/s]
790
  6%|▌ | 630/10682 [05:57<1:23:09, 2.01it/s]
791
  6%|▌ | 631/10682 [05:58<1:23:07, 2.02it/s]
792
  6%|▌ | 632/10682 [05:58<1:23:05, 2.02it/s]
793
  6%|▌ | 633/10682 [05:59<1:23:00, 2.02it/s]
794
  6%|▌ | 634/10682 [05:59<1:22:54, 2.02it/s]
795
  6%|▌ | 635/10682 [06:00<1:22:56, 2.02it/s]
796
  6%|▌ | 636/10682 [06:00<1:22:53, 2.02it/s]
797
  6%|▌ | 637/10682 [06:01<1:22:54, 2.02it/s]
798
  6%|▌ | 638/10682 [06:01<1:22:50, 2.02it/s]
799
  6%|▌ | 639/10682 [06:02<1:22:48, 2.02it/s]
800
  6%|▌ | 640/10682 [06:02<1:22:46, 2.02it/s]
801
  6%|▌ | 641/10682 [06:03<1:22:50, 2.02it/s]
802
  6%|▌ | 642/10682 [06:03<1:22:47, 2.02it/s]
803
  6%|▌ | 643/10682 [06:04<1:22:49, 2.02it/s]
804
  6%|▌ | 644/10682 [06:04<1:22:45, 2.02it/s]
805
  6%|▌ | 645/10682 [06:05<1:22:46, 2.02it/s]
806
  6%|▌ | 646/10682 [06:05<1:22:46, 2.02it/s]
807
  6%|▌ | 647/10682 [06:06<1:22:45, 2.02it/s]
808
  6%|▌ | 648/10682 [06:06<1:22:45, 2.02it/s]
809
  6%|▌ | 649/10682 [06:07<1:22:48, 2.02it/s]
810
  6%|▌ | 650/10682 [06:07<1:22:43, 2.02it/s]
811
 
812
+
813
  6%|▌ | 650/10682 [06:07<1:22:43, 2.02it/s]
814
  6%|▌ | 651/10682 [06:08<1:22:50, 2.02it/s]
815
  6%|▌ | 652/10682 [06:08<1:22:52, 2.02it/s]
816
  6%|▌ | 653/10682 [06:09<1:22:49, 2.02it/s]
817
  6%|▌ | 654/10682 [06:09<1:22:47, 2.02it/s]
818
  6%|▌ | 655/10682 [06:10<1:22:43, 2.02it/s]
819
  6%|▌ | 656/10682 [06:10<1:22:44, 2.02it/s]
820
  6%|▌ | 657/10682 [06:11<1:22:40, 2.02it/s]
821
  6%|▌ | 658/10682 [06:11<1:22:43, 2.02it/s]
822
  6%|▌ | 659/10682 [06:12<1:22:38, 2.02it/s]
823
  6%|▌ | 660/10682 [06:12<1:22:36, 2.02it/s]
824
  6%|▌ | 661/10682 [06:13<1:22:39, 2.02it/s]
825
  6%|▌ | 662/10682 [06:13<1:22:36, 2.02it/s]
826
  6%|▌ | 663/10682 [06:14<1:22:40, 2.02it/s]
827
  6%|▌ | 664/10682 [06:14<1:22:33, 2.02it/s]
828
  6%|▌ | 665/10682 [06:15<1:22:36, 2.02it/s]
829
  6%|▌ | 666/10682 [06:15<1:22:33, 2.02it/s]
830
  6%|▌ | 667/10682 [06:16<1:22:38, 2.02it/s]
831
  6%|▋ | 668/10682 [06:16<1:22:38, 2.02it/s]
832
  6%|▋ | 669/10682 [06:17<1:22:36, 2.02it/s]
833
  6%|▋ | 670/10682 [06:17<1:22:36, 2.02it/s]
834
  6%|▋ | 671/10682 [06:18<1:22:34, 2.02it/s]
835
  6%|▋ | 672/10682 [06:18<1:22:36, 2.02it/s]
836
  6%|▋ | 673/10682 [06:19<1:22:32, 2.02it/s]
837
  6%|▋ | 674/10682 [06:19<1:22:48, 2.01it/s]
838
  6%|▋ | 675/10682 [06:20<1:22:57, 2.01it/s]
839
  {'loss': 4.9272, 'grad_norm': 0.5725471377372742, 'learning_rate': 0.0006314312441534145, 'epoch': 0.88}
840
+
841
  6%|▋ | 675/10682 [06:20<1:22:57, 2.01it/s]
842
  6%|▋ | 676/10682 [06:20<1:22:58, 2.01it/s]
843
  6%|▋ | 677/10682 [06:21<1:22:53, 2.01it/s]
844
  6%|▋ | 678/10682 [06:21<1:22:54, 2.01it/s]
845
  6%|▋ | 679/10682 [06:22<1:22:41, 2.02it/s]
846
  6%|▋ | 680/10682 [06:22<1:22:40, 2.02it/s]
847
  6%|▋ | 681/10682 [06:23<1:22:33, 2.02it/s]
848
  6%|▋ | 682/10682 [06:23<1:22:35, 2.02it/s]
849
  6%|▋ | 683/10682 [06:24<1:22:31, 2.02it/s]
850
  6%|▋ | 684/10682 [06:24<1:22:34, 2.02it/s]
851
  6%|▋ | 685/10682 [06:25<1:22:27, 2.02it/s]
852
  6%|▋ | 686/10682 [06:25<1:22:31, 2.02it/s]
853
  6%|▋ | 687/10682 [06:26<1:22:30, 2.02it/s]
854
  6%|▋ | 688/10682 [06:26<1:22:34, 2.02it/s]
855
  6%|▋ | 689/10682 [06:27<1:22:26, 2.02it/s]
856
  6%|▋ | 690/10682 [06:27<1:22:29, 2.02it/s]
857
  6%|▋ | 691/10682 [06:28<1:22:23, 2.02it/s]
858
  6%|▋ | 692/10682 [06:28<1:22:26, 2.02it/s]
859
  6%|▋ | 693/10682 [06:29<1:22:25, 2.02it/s]
860
  6%|▋ | 694/10682 [06:29<1:22:26, 2.02it/s]
861
  7%|▋ | 695/10682 [06:30<1:22:25, 2.02it/s]
862
  7%|▋ | 696/10682 [06:30<1:22:24, 2.02it/s]
863
  7%|▋ | 697/10682 [06:31<1:22:24, 2.02it/s]
864
  7%|▋ | 698/10682 [06:31<1:22:23, 2.02it/s]
865
  7%|▋ | 699/10682 [06:32<1:22:20, 2.02it/s]
866
  7%|▋ | 700/10682 [06:32<1:22:18, 2.02it/s]{'loss': 4.8834, 'grad_norm': 0.48604607582092285, 'learning_rate': 0.0006548175865294667, 'epoch': 0.92}
867
+
868
 
869
  7%|▋ | 700/10682 [06:32<1:22:18, 2.02it/s]
870
  7%|▋ | 701/10682 [06:33<1:22:27, 2.02it/s]
871
  7%|▋ | 702/10682 [06:33<1:22:23, 2.02it/s]
872
  7%|▋ | 703/10682 [06:34<1:22:24, 2.02it/s]
873
  7%|▋ | 704/10682 [06:34<1:22:19, 2.02it/s]
874
  7%|▋ | 705/10682 [06:35<1:22:24, 2.02it/s]
875
  7%|▋ | 706/10682 [06:35<1:22:19, 2.02it/s]
876
  7%|▋ | 707/10682 [06:36<1:22:20, 2.02it/s]
877
  7%|▋ | 708/10682 [06:36<1:22:14, 2.02it/s]
878
  7%|▋ | 709/10682 [06:37<1:22:18, 2.02it/s]
879
  7%|▋ | 710/10682 [06:37<1:22:17, 2.02it/s]
880
  7%|▋ | 711/10682 [06:38<1:22:21, 2.02it/s]
881
  7%|▋ | 712/10682 [06:38<1:22:16, 2.02it/s]
882
  7%|▋ | 713/10682 [06:39<1:22:17, 2.02it/s]
883
  7%|▋ | 714/10682 [06:39<1:22:16, 2.02it/s]
884
  7%|▋ | 715/10682 [06:40<1:22:14, 2.02it/s]
885
  7%|▋ | 716/10682 [06:40<1:22:15, 2.02it/s]
886
  7%|▋ | 717/10682 [06:41<1:22:14, 2.02it/s]
887
  7%|▋ | 718/10682 [06:41<1:22:18, 2.02it/s]
888
  7%|▋ | 719/10682 [06:42<1:22:13, 2.02it/s]
889
  7%|▋ | 720/10682 [06:42<1:22:12, 2.02it/s]
890
  7%|▋ | 721/10682 [06:43<1:22:09, 2.02it/s]
891
  7%|▋ | 722/10682 [06:43<1:22:13, 2.02it/s]
892
  7%|▋ | 723/10682 [06:43<1:22:13, 2.02it/s]
893
  7%|▋ | 724/10682 [06:44<1:22:16, 2.02it/s]
894
  7%|▋ | 725/10682 [06:44<1:22:13, 2.02it/s]{'loss': 4.8458, 'grad_norm': 0.43501177430152893, 'learning_rate': 0.0006782039289055192, 'epoch': 0.95}
895
+
896
 
897
  7%|▋ | 725/10682 [06:44<1:22:13, 2.02it/s]
898
  7%|▋ | 726/10682 [06:45<1:22:24, 2.01it/s]
899
  7%|▋ | 727/10682 [06:45<1:22:27, 2.01it/s]
900
  7%|▋ | 728/10682 [06:46<1:22:23, 2.01it/s]
901
  7%|▋ | 729/10682 [06:46<1:22:21, 2.01it/s]
902
  7%|▋ | 730/10682 [06:47<1:22:15, 2.02it/s]
903
  7%|▋ | 731/10682 [06:47<1:22:14, 2.02it/s]
904
  7%|▋ | 732/10682 [06:48<1:22:07, 2.02it/s]
905
  7%|▋ | 733/10682 [06:48<1:22:09, 2.02it/s]
906
  7%|▋ | 734/10682 [06:49<1:22:09, 2.02it/s]
907
  7%|▋ | 735/10682 [06:49<1:22:10, 2.02it/s]
908
  7%|▋ | 736/10682 [06:50<1:22:07, 2.02it/s]
909
  7%|▋ | 737/10682 [06:50<1:22:06, 2.02it/s]
910
  7%|▋ | 738/10682 [06:51<1:22:09, 2.02it/s]
911
  7%|▋ | 739/10682 [06:51<1:22:04, 2.02it/s]
912
  7%|▋ | 740/10682 [06:52<1:22:10, 2.02it/s]
913
  7%|▋ | 741/10682 [06:52<1:22:05, 2.02it/s]
914
  7%|▋ | 742/10682 [06:53<1:22:11, 2.02it/s]
915
  7%|▋ | 743/10682 [06:53<1:22:06, 2.02it/s]
916
  7%|▋ | 744/10682 [06:54<1:22:14, 2.01it/s]
917
  7%|▋ | 745/10682 [06:54<1:22:34, 2.01it/s]
918
  7%|▋ | 746/10682 [06:55<1:22:48, 2.00it/s]
919
  7%|▋ | 747/10682 [06:55<1:23:16, 1.99it/s]
920
  7%|▋ | 748/10682 [06:56<1:23:37, 1.98it/s]
921
  7%|▋ | 749/10682 [06:56<1:23:49, 1.97it/s]
922
  7%|▋ | 750/10682 [06:57<1:23:50, 1.97it/s]{'loss': 4.809, 'grad_norm': 0.434061199426651, 'learning_rate': 0.0007015902712815716, 'epoch': 0.98}
923
+
924
 
925
  7%|▋ | 750/10682 [06:57<1:23:50, 1.97it/s]
926
  7%|▋ | 751/10682 [06:57<1:23:46, 1.98it/s]
927
  7%|▋ | 752/10682 [06:58<1:23:32, 1.98it/s]
928
  7%|▋ | 753/10682 [06:58<1:23:05, 1.99it/s]
929
  7%|▋ | 754/10682 [06:59<1:22:43, 2.00it/s]
930
  7%|▋ | 755/10682 [06:59<1:22:34, 2.00it/s]
931
  7%|▋ | 756/10682 [07:00<1:22:18, 2.01it/s]
932
  7%|▋ | 757/10682 [07:00<1:22:12, 2.01it/s]
933
  7%|▋ | 758/10682 [07:01<1:22:03, 2.02it/s]
934
  7%|▋ | 759/10682 [07:01<1:22:04, 2.02it/s]
935
  7%|▋ | 760/10682 [07:02<1:21:55, 2.02it/s]
936
  7%|▋ | 761/10682 [07:02<1:21:56, 2.02it/s]
937
  7%|▋ | 762/10682 [07:03<1:21:50, 2.02it/s]
938
  7%|▋ | 763/10682 [07:03<1:24:02, 1.97it/s]
939
  7%|▋ | 764/10682 [07:36<28:16:56, 10.27s/it]
940
  7%|▋ | 765/10682 [07:37<20:12:22, 7.34s/it]
941
  7%|▋ | 766/10682 [07:37<14:33:39, 5.29s/it]
942
  7%|▋ | 767/10682 [07:38<10:36:06, 3.85s/it]
943
  7%|▋ | 768/10682 [07:38<7:49:41, 2.84s/it]
944
  7%|▋ | 769/10682 [07:39<5:53:19, 2.14s/it]
945
  7%|▋ | 770/10682 [07:39<4:31:48, 1.65s/it]
946
  7%|▋ | 771/10682 [07:40<3:34:48, 1.30s/it]
947
  7%|▋ | 772/10682 [07:40<2:54:46, 1.06s/it]
948
  7%|▋ | 773/10682 [07:41<2:26:52, 1.12it/s]
949
  7%|▋ | 774/10682 [07:41<2:07:16, 1.30it/s]
950
  7%|▋ | 775/10682 [07:42<1:53:45, 1.45it/s]{'loss': 4.7582, 'grad_norm': 0.39575064182281494, 'learning_rate': 0.0007249766136576241, 'epoch': 1.02}
951
+
952
 
953
  7%|▋ | 775/10682 [07:42<1:53:45, 1.45it/s]
954
  7%|▋ | 776/10682 [07:42<1:44:22, 1.58it/s]
955
  7%|▋ | 777/10682 [07:43<1:37:30, 1.69it/s]
956
  7%|▋ | 778/10682 [07:43<1:32:52, 1.78it/s]
957
  7%|▋ | 779/10682 [07:44<1:29:29, 1.84it/s]
958
  7%|▋ | 780/10682 [07:44<1:27:18, 1.89it/s]
959
  7%|▋ | 781/10682 [07:45<1:25:28, 1.93it/s]
960
  7%|▋ | 782/10682 [07:45<1:24:24, 1.95it/s]
961
  7%|▋ | 783/10682 [07:46<1:23:35, 1.97it/s]
962
  7%|▋ | 784/10682 [07:46<1:23:12, 1.98it/s]
963
  7%|▋ | 785/10682 [07:47<1:22:48, 1.99it/s]
964
  7%|▋ | 786/10682 [07:47<1:22:41, 1.99it/s]
965
  7%|▋ | 787/10682 [07:48<1:22:20, 2.00it/s]
966
  7%|▋ | 788/10682 [07:48<1:22:09, 2.01it/s]
967
  7%|▋ | 789/10682 [07:49<1:22:03, 2.01it/s]
968
  7%|▋ | 790/10682 [07:49<1:22:01, 2.01it/s]
969
  7%|▋ | 791/10682 [07:50<1:22:04, 2.01it/s]
970
  7%|▋ | 792/10682 [07:50<1:21:54, 2.01it/s]
971
  7%|▋ | 793/10682 [07:51<1:21:51, 2.01it/s]
972
  7%|▋ | 794/10682 [07:51<1:21:45, 2.02it/s]
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a3e2a60cd14b3dbbaf3d866569ed94ac9b6a79055688c2eb14eeb2de944679d
3
+ size 5112