Hugging Face's logo

edlee123
/
bridgetower

TensorBoard
Safetensors
Habana
bridgetower
Generated from Trainer
Model card Files Metrics Community
edlee123 commited on
Commit
deab0b8
•
1 Parent(s): c8d2d9a

End of training

Browse files
Files changed (13) hide show
  1. README.md +7 -2
  2. all_results.json +15 -0
  3. merges.txt +0 -0
  4. preprocessor_config.json +28 -0
  5. runs/Oct14_22-43-32_workload-ai-workshop/events.out.tfevents.1728946985.workload-ai-workshop.3947.1 +3 -0
  6. special_tokens_map.json +15 -0
  7. test_results.json +10 -0
  8. tokenizer.json +0 -0
  9. tokenizer_config.json +57 -0
  10. train_results.json +11 -0
  11. trainer_state.json +1265 -0
  12. validation_results.json +10 -0
  13. vocab.json +0 -0
README.md CHANGED
@@ -4,7 +4,7 @@ base_model: BridgeTower/bridgetower-large-itm-mlm-itc
4
  tags:
5
  - generated_from_trainer
6
  datasets:
7
- - newyorker_caption_contest
8
  model-index:
9
  - name: bridgetower
10
  results: []
@@ -15,7 +15,12 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # bridgetower
17
 
18
- This model is a fine-tuned version of [BridgeTower/bridgetower-large-itm-mlm-itc](https://huggingface.co/BridgeTower/bridgetower-large-itm-mlm-itc) on the newyorker_caption_contest dataset.
 
 
 
 
 
19
 
20
  ## Model description
21
 
 
4
  tags:
5
  - generated_from_trainer
6
  datasets:
7
+ - jmhessel/newyorker_caption_contest
8
  model-index:
9
  - name: bridgetower
10
  results: []
 
15
 
16
  # bridgetower
17
 
18
+ This model is a fine-tuned version of [BridgeTower/bridgetower-large-itm-mlm-itc](https://huggingface.co/BridgeTower/bridgetower-large-itm-mlm-itc) on the jmhessel/newyorker_caption_contest matching dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 0.1215
21
+ - Memory Allocated (gb): 51.27
22
+ - Max Memory Allocated (gb): 60.52
23
+ - Total Memory Available (gb): 94.62
24
 
25
  ## Model description
26
 
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 0.12152472138404846,
4
+ "eval_runtime": 3.7753,
5
+ "eval_samples_per_second": 141.222,
6
+ "eval_steps_per_second": 8.826,
7
+ "max_memory_allocated (GB)": 60.52,
8
+ "memory_allocated (GB)": 51.27,
9
+ "total_flos": 3.0598946525952e+16,
10
+ "total_memory_available (GB)": 94.62,
11
+ "train_loss": 0.06085505417415074,
12
+ "train_runtime": 1020.8061,
13
+ "train_samples_per_second": 55.51,
14
+ "train_steps_per_second": 1.389
15
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": null,
3
+ "do_center_crop": true,
4
+ "do_normalize": true,
5
+ "do_pad": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.48145466,
10
+ 0.4578275,
11
+ 0.40821073
12
+ ],
13
+ "image_processor_type": "BridgeTowerImageProcessor",
14
+ "image_std": [
15
+ 0.26862954,
16
+ 0.26130258,
17
+ 0.27577711
18
+ ],
19
+ "max_text_len": 50,
20
+ "resample": 3,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "shortest_edge": 294
24
+ },
25
+ "size_divisor": 32,
26
+ "tokenizer": "roberta-large",
27
+ "vocab_size": 50265
28
+ }
runs/Oct14_22-43-32_workload-ai-workshop/events.out.tfevents.1728946985.workload-ai-workshop.3947.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:074b81b769f81184541b5745477cfc1ab677173c8e33b12714f97e43ea66a9fb
3
+ size 998
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
test_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 0.12152472138404846,
4
+ "eval_runtime": 3.7753,
5
+ "eval_samples_per_second": 141.222,
6
+ "eval_steps_per_second": 8.826,
7
+ "max_memory_allocated (GB)": 60.52,
8
+ "memory_allocated (GB)": 51.27,
9
+ "total_memory_available (GB)": 94.62
10
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<s>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<pad>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "50264": {
37
+ "content": "<mask>",
38
+ "lstrip": true,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ }
44
+ },
45
+ "bos_token": "<s>",
46
+ "clean_up_tokenization_spaces": true,
47
+ "cls_token": "<s>",
48
+ "eos_token": "</s>",
49
+ "errors": "replace",
50
+ "mask_token": "<mask>",
51
+ "model_max_length": 1000000000000000019884624838656,
52
+ "pad_token": "<pad>",
53
+ "sep_token": "</s>",
54
+ "tokenizer_class": "RobertaTokenizer",
55
+ "trim_offsets": true,
56
+ "unk_token": "<unk>"
57
+ }
train_results.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "max_memory_allocated (GB)": 60.52,
4
+ "memory_allocated (GB)": 50.57,
5
+ "total_flos": 3.0598946525952e+16,
6
+ "total_memory_available (GB)": 94.62,
7
+ "train_loss": 0.06085505417415074,
8
+ "train_runtime": 1020.8061,
9
+ "train_samples_per_second": 55.51,
10
+ "train_steps_per_second": 1.389
11
+ }
trainer_state.json ADDED
@@ -0,0 +1,1265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1225,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04081632653061224,
13
+ "grad_norm": 9.589848518371582,
14
+ "learning_rate": 9.918367346938776e-06,
15
+ "loss": 0.2612,
16
+ "max_memory_allocated (GB)": 57.18,
17
+ "memory_allocated (GB)": 50.57,
18
+ "step": 10,
19
+ "total_memory_available (GB)": 94.62
20
+ },
21
+ {
22
+ "epoch": 0.08163265306122448,
23
+ "grad_norm": 6.701302528381348,
24
+ "learning_rate": 9.836734693877552e-06,
25
+ "loss": 0.154,
26
+ "max_memory_allocated (GB)": 57.18,
27
+ "memory_allocated (GB)": 50.57,
28
+ "step": 20,
29
+ "total_memory_available (GB)": 94.62
30
+ },
31
+ {
32
+ "epoch": 0.12244897959183673,
33
+ "grad_norm": 5.337311267852783,
34
+ "learning_rate": 9.755102040816327e-06,
35
+ "loss": 0.1235,
36
+ "max_memory_allocated (GB)": 57.18,
37
+ "memory_allocated (GB)": 50.57,
38
+ "step": 30,
39
+ "total_memory_available (GB)": 94.62
40
+ },
41
+ {
42
+ "epoch": 0.16326530612244897,
43
+ "grad_norm": 4.5042338371276855,
44
+ "learning_rate": 9.673469387755103e-06,
45
+ "loss": 0.1096,
46
+ "max_memory_allocated (GB)": 57.18,
47
+ "memory_allocated (GB)": 50.57,
48
+ "step": 40,
49
+ "total_memory_available (GB)": 94.62
50
+ },
51
+ {
52
+ "epoch": 0.20408163265306123,
53
+ "grad_norm": 4.461822032928467,
54
+ "learning_rate": 9.591836734693878e-06,
55
+ "loss": 0.1196,
56
+ "max_memory_allocated (GB)": 57.18,
57
+ "memory_allocated (GB)": 50.57,
58
+ "step": 50,
59
+ "total_memory_available (GB)": 94.62
60
+ },
61
+ {
62
+ "epoch": 0.24489795918367346,
63
+ "grad_norm": 2.2825701236724854,
64
+ "learning_rate": 9.510204081632653e-06,
65
+ "loss": 0.0805,
66
+ "max_memory_allocated (GB)": 57.18,
67
+ "memory_allocated (GB)": 50.57,
68
+ "step": 60,
69
+ "total_memory_available (GB)": 94.62
70
+ },
71
+ {
72
+ "epoch": 0.2857142857142857,
73
+ "grad_norm": 3.725268602371216,
74
+ "learning_rate": 9.42857142857143e-06,
75
+ "loss": 0.1026,
76
+ "max_memory_allocated (GB)": 57.18,
77
+ "memory_allocated (GB)": 50.57,
78
+ "step": 70,
79
+ "total_memory_available (GB)": 94.62
80
+ },
81
+ {
82
+ "epoch": 0.32653061224489793,
83
+ "grad_norm": 1.707739233970642,
84
+ "learning_rate": 9.346938775510204e-06,
85
+ "loss": 0.1111,
86
+ "max_memory_allocated (GB)": 57.18,
87
+ "memory_allocated (GB)": 50.57,
88
+ "step": 80,
89
+ "total_memory_available (GB)": 94.62
90
+ },
91
+ {
92
+ "epoch": 0.3673469387755102,
93
+ "grad_norm": 4.5863938331604,
94
+ "learning_rate": 9.26530612244898e-06,
95
+ "loss": 0.0856,
96
+ "max_memory_allocated (GB)": 57.18,
97
+ "memory_allocated (GB)": 50.57,
98
+ "step": 90,
99
+ "total_memory_available (GB)": 94.62
100
+ },
101
+ {
102
+ "epoch": 0.40816326530612246,
103
+ "grad_norm": 11.972647666931152,
104
+ "learning_rate": 9.183673469387756e-06,
105
+ "loss": 0.0759,
106
+ "max_memory_allocated (GB)": 57.18,
107
+ "memory_allocated (GB)": 50.57,
108
+ "step": 100,
109
+ "total_memory_available (GB)": 94.62
110
+ },
111
+ {
112
+ "epoch": 0.4489795918367347,
113
+ "grad_norm": 4.550654888153076,
114
+ "learning_rate": 9.102040816326532e-06,
115
+ "loss": 0.0717,
116
+ "max_memory_allocated (GB)": 57.18,
117
+ "memory_allocated (GB)": 50.57,
118
+ "step": 110,
119
+ "total_memory_available (GB)": 94.62
120
+ },
121
+ {
122
+ "epoch": 0.4897959183673469,
123
+ "grad_norm": 4.418276786804199,
124
+ "learning_rate": 9.020408163265307e-06,
125
+ "loss": 0.0717,
126
+ "max_memory_allocated (GB)": 57.18,
127
+ "memory_allocated (GB)": 50.57,
128
+ "step": 120,
129
+ "total_memory_available (GB)": 94.62
130
+ },
131
+ {
132
+ "epoch": 0.5306122448979592,
133
+ "grad_norm": 1.651443600654602,
134
+ "learning_rate": 8.938775510204082e-06,
135
+ "loss": 0.0581,
136
+ "max_memory_allocated (GB)": 57.18,
137
+ "memory_allocated (GB)": 50.57,
138
+ "step": 130,
139
+ "total_memory_available (GB)": 94.62
140
+ },
141
+ {
142
+ "epoch": 0.5714285714285714,
143
+ "grad_norm": 1.5251814126968384,
144
+ "learning_rate": 8.857142857142858e-06,
145
+ "loss": 0.0481,
146
+ "max_memory_allocated (GB)": 57.18,
147
+ "memory_allocated (GB)": 50.57,
148
+ "step": 140,
149
+ "total_memory_available (GB)": 94.62
150
+ },
151
+ {
152
+ "epoch": 0.6122448979591837,
153
+ "grad_norm": 1.7455183267593384,
154
+ "learning_rate": 8.775510204081633e-06,
155
+ "loss": 0.0625,
156
+ "max_memory_allocated (GB)": 57.18,
157
+ "memory_allocated (GB)": 50.57,
158
+ "step": 150,
159
+ "total_memory_available (GB)": 94.62
160
+ },
161
+ {
162
+ "epoch": 0.6530612244897959,
163
+ "grad_norm": 1.7588891983032227,
164
+ "learning_rate": 8.69387755102041e-06,
165
+ "loss": 0.0711,
166
+ "max_memory_allocated (GB)": 57.18,
167
+ "memory_allocated (GB)": 50.57,
168
+ "step": 160,
169
+ "total_memory_available (GB)": 94.62
170
+ },
171
+ {
172
+ "epoch": 0.6938775510204082,
173
+ "grad_norm": 2.7675328254699707,
174
+ "learning_rate": 8.612244897959184e-06,
175
+ "loss": 0.0747,
176
+ "max_memory_allocated (GB)": 57.18,
177
+ "memory_allocated (GB)": 50.57,
178
+ "step": 170,
179
+ "total_memory_available (GB)": 94.62
180
+ },
181
+ {
182
+ "epoch": 0.7346938775510204,
183
+ "grad_norm": 1.781469464302063,
184
+ "learning_rate": 8.530612244897961e-06,
185
+ "loss": 0.061,
186
+ "max_memory_allocated (GB)": 57.18,
187
+ "memory_allocated (GB)": 50.57,
188
+ "step": 180,
189
+ "total_memory_available (GB)": 94.62
190
+ },
191
+ {
192
+ "epoch": 0.7755102040816326,
193
+ "grad_norm": 2.3728435039520264,
194
+ "learning_rate": 8.448979591836736e-06,
195
+ "loss": 0.0588,
196
+ "max_memory_allocated (GB)": 57.18,
197
+ "memory_allocated (GB)": 50.57,
198
+ "step": 190,
199
+ "total_memory_available (GB)": 94.62
200
+ },
201
+ {
202
+ "epoch": 0.8163265306122449,
203
+ "grad_norm": 0.8711996674537659,
204
+ "learning_rate": 8.36734693877551e-06,
205
+ "loss": 0.062,
206
+ "max_memory_allocated (GB)": 57.18,
207
+ "memory_allocated (GB)": 50.57,
208
+ "step": 200,
209
+ "total_memory_available (GB)": 94.62
210
+ },
211
+ {
212
+ "epoch": 0.8571428571428571,
213
+ "grad_norm": 1.1986733675003052,
214
+ "learning_rate": 8.285714285714287e-06,
215
+ "loss": 0.0627,
216
+ "max_memory_allocated (GB)": 57.18,
217
+ "memory_allocated (GB)": 50.57,
218
+ "step": 210,
219
+ "total_memory_available (GB)": 94.62
220
+ },
221
+ {
222
+ "epoch": 0.8979591836734694,
223
+ "grad_norm": 2.8968520164489746,
224
+ "learning_rate": 8.204081632653062e-06,
225
+ "loss": 0.0604,
226
+ "max_memory_allocated (GB)": 57.18,
227
+ "memory_allocated (GB)": 50.57,
228
+ "step": 220,
229
+ "total_memory_available (GB)": 94.62
230
+ },
231
+ {
232
+ "epoch": 0.9387755102040817,
233
+ "grad_norm": 0.8414793610572815,
234
+ "learning_rate": 8.122448979591837e-06,
235
+ "loss": 0.0559,
236
+ "max_memory_allocated (GB)": 57.18,
237
+ "memory_allocated (GB)": 50.57,
238
+ "step": 230,
239
+ "total_memory_available (GB)": 94.62
240
+ },
241
+ {
242
+ "epoch": 0.9795918367346939,
243
+ "grad_norm": 0.7434167861938477,
244
+ "learning_rate": 8.040816326530613e-06,
245
+ "loss": 0.0498,
246
+ "max_memory_allocated (GB)": 57.18,
247
+ "memory_allocated (GB)": 50.57,
248
+ "step": 240,
249
+ "total_memory_available (GB)": 94.62
250
+ },
251
+ {
252
+ "epoch": 1.0204081632653061,
253
+ "grad_norm": 0.8703041076660156,
254
+ "learning_rate": 7.959183673469388e-06,
255
+ "loss": 0.0618,
256
+ "max_memory_allocated (GB)": 57.18,
257
+ "memory_allocated (GB)": 50.57,
258
+ "step": 250,
259
+ "total_memory_available (GB)": 94.62
260
+ },
261
+ {
262
+ "epoch": 1.0612244897959184,
263
+ "grad_norm": 1.0856379270553589,
264
+ "learning_rate": 7.877551020408164e-06,
265
+ "loss": 0.056,
266
+ "max_memory_allocated (GB)": 57.18,
267
+ "memory_allocated (GB)": 50.57,
268
+ "step": 260,
269
+ "total_memory_available (GB)": 94.62
270
+ },
271
+ {
272
+ "epoch": 1.1020408163265305,
273
+ "grad_norm": 0.8847401142120361,
274
+ "learning_rate": 7.79591836734694e-06,
275
+ "loss": 0.0625,
276
+ "max_memory_allocated (GB)": 57.18,
277
+ "memory_allocated (GB)": 50.57,
278
+ "step": 270,
279
+ "total_memory_available (GB)": 94.62
280
+ },
281
+ {
282
+ "epoch": 1.1428571428571428,
283
+ "grad_norm": 1.5929882526397705,
284
+ "learning_rate": 7.714285714285716e-06,
285
+ "loss": 0.0571,
286
+ "max_memory_allocated (GB)": 57.18,
287
+ "memory_allocated (GB)": 50.57,
288
+ "step": 280,
289
+ "total_memory_available (GB)": 94.62
290
+ },
291
+ {
292
+ "epoch": 1.183673469387755,
293
+ "grad_norm": 0.8007532954216003,
294
+ "learning_rate": 7.63265306122449e-06,
295
+ "loss": 0.0511,
296
+ "max_memory_allocated (GB)": 57.18,
297
+ "memory_allocated (GB)": 50.57,
298
+ "step": 290,
299
+ "total_memory_available (GB)": 94.62
300
+ },
301
+ {
302
+ "epoch": 1.2244897959183674,
303
+ "grad_norm": 1.2002859115600586,
304
+ "learning_rate": 7.551020408163265e-06,
305
+ "loss": 0.065,
306
+ "max_memory_allocated (GB)": 57.18,
307
+ "memory_allocated (GB)": 50.57,
308
+ "step": 300,
309
+ "total_memory_available (GB)": 94.62
310
+ },
311
+ {
312
+ "epoch": 1.2653061224489797,
313
+ "grad_norm": 12.871713638305664,
314
+ "learning_rate": 7.469387755102041e-06,
315
+ "loss": 0.0664,
316
+ "max_memory_allocated (GB)": 57.18,
317
+ "memory_allocated (GB)": 50.57,
318
+ "step": 310,
319
+ "total_memory_available (GB)": 94.62
320
+ },
321
+ {
322
+ "epoch": 1.306122448979592,
323
+ "grad_norm": 2.46173357963562,
324
+ "learning_rate": 7.387755102040817e-06,
325
+ "loss": 0.0495,
326
+ "max_memory_allocated (GB)": 57.18,
327
+ "memory_allocated (GB)": 50.57,
328
+ "step": 320,
329
+ "total_memory_available (GB)": 94.62
330
+ },
331
+ {
332
+ "epoch": 1.346938775510204,
333
+ "grad_norm": 0.860598087310791,
334
+ "learning_rate": 7.306122448979592e-06,
335
+ "loss": 0.0603,
336
+ "max_memory_allocated (GB)": 57.18,
337
+ "memory_allocated (GB)": 50.57,
338
+ "step": 330,
339
+ "total_memory_available (GB)": 94.62
340
+ },
341
+ {
342
+ "epoch": 1.3877551020408163,
343
+ "grad_norm": 2.5583598613739014,
344
+ "learning_rate": 7.224489795918368e-06,
345
+ "loss": 0.0547,
346
+ "max_memory_allocated (GB)": 57.18,
347
+ "memory_allocated (GB)": 50.57,
348
+ "step": 340,
349
+ "total_memory_available (GB)": 94.62
350
+ },
351
+ {
352
+ "epoch": 1.4285714285714286,
353
+ "grad_norm": 0.37155964970588684,
354
+ "learning_rate": 7.1428571428571436e-06,
355
+ "loss": 0.048,
356
+ "max_memory_allocated (GB)": 57.18,
357
+ "memory_allocated (GB)": 50.57,
358
+ "step": 350,
359
+ "total_memory_available (GB)": 94.62
360
+ },
361
+ {
362
+ "epoch": 1.469387755102041,
363
+ "grad_norm": 1.808316707611084,
364
+ "learning_rate": 7.061224489795919e-06,
365
+ "loss": 0.0462,
366
+ "max_memory_allocated (GB)": 57.18,
367
+ "memory_allocated (GB)": 50.57,
368
+ "step": 360,
369
+ "total_memory_available (GB)": 94.62
370
+ },
371
+ {
372
+ "epoch": 1.510204081632653,
373
+ "grad_norm": 1.0183931589126587,
374
+ "learning_rate": 6.979591836734695e-06,
375
+ "loss": 0.0594,
376
+ "max_memory_allocated (GB)": 57.18,
377
+ "memory_allocated (GB)": 50.57,
378
+ "step": 370,
379
+ "total_memory_available (GB)": 94.62
380
+ },
381
+ {
382
+ "epoch": 1.5510204081632653,
383
+ "grad_norm": 0.5249583721160889,
384
+ "learning_rate": 6.8979591836734705e-06,
385
+ "loss": 0.0479,
386
+ "max_memory_allocated (GB)": 57.18,
387
+ "memory_allocated (GB)": 50.57,
388
+ "step": 380,
389
+ "total_memory_available (GB)": 94.62
390
+ },
391
+ {
392
+ "epoch": 1.5918367346938775,
393
+ "grad_norm": 1.1005572080612183,
394
+ "learning_rate": 6.816326530612245e-06,
395
+ "loss": 0.0649,
396
+ "max_memory_allocated (GB)": 57.18,
397
+ "memory_allocated (GB)": 50.57,
398
+ "step": 390,
399
+ "total_memory_available (GB)": 94.62
400
+ },
401
+ {
402
+ "epoch": 1.6326530612244898,
403
+ "grad_norm": 0.6047573089599609,
404
+ "learning_rate": 6.734693877551021e-06,
405
+ "loss": 0.0607,
406
+ "max_memory_allocated (GB)": 57.18,
407
+ "memory_allocated (GB)": 50.57,
408
+ "step": 400,
409
+ "total_memory_available (GB)": 94.62
410
+ },
411
+ {
412
+ "epoch": 1.6734693877551021,
413
+ "grad_norm": 0.7261654734611511,
414
+ "learning_rate": 6.653061224489797e-06,
415
+ "loss": 0.0606,
416
+ "max_memory_allocated (GB)": 57.18,
417
+ "memory_allocated (GB)": 50.57,
418
+ "step": 410,
419
+ "total_memory_available (GB)": 94.62
420
+ },
421
+ {
422
+ "epoch": 1.7142857142857144,
423
+ "grad_norm": 0.848527193069458,
424
+ "learning_rate": 6.571428571428572e-06,
425
+ "loss": 0.0532,
426
+ "max_memory_allocated (GB)": 57.18,
427
+ "memory_allocated (GB)": 50.57,
428
+ "step": 420,
429
+ "total_memory_available (GB)": 94.62
430
+ },
431
+ {
432
+ "epoch": 1.7551020408163265,
433
+ "grad_norm": 0.23483288288116455,
434
+ "learning_rate": 6.489795918367348e-06,
435
+ "loss": 0.068,
436
+ "max_memory_allocated (GB)": 57.18,
437
+ "memory_allocated (GB)": 50.57,
438
+ "step": 430,
439
+ "total_memory_available (GB)": 94.62
440
+ },
441
+ {
442
+ "epoch": 1.7959183673469388,
443
+ "grad_norm": 2.0767459869384766,
444
+ "learning_rate": 6.408163265306124e-06,
445
+ "loss": 0.0617,
446
+ "max_memory_allocated (GB)": 57.18,
447
+ "memory_allocated (GB)": 50.57,
448
+ "step": 440,
449
+ "total_memory_available (GB)": 94.62
450
+ },
451
+ {
452
+ "epoch": 1.836734693877551,
453
+ "grad_norm": 0.5654011368751526,
454
+ "learning_rate": 6.326530612244899e-06,
455
+ "loss": 0.044,
456
+ "max_memory_allocated (GB)": 57.18,
457
+ "memory_allocated (GB)": 50.57,
458
+ "step": 450,
459
+ "total_memory_available (GB)": 94.62
460
+ },
461
+ {
462
+ "epoch": 1.8775510204081631,
463
+ "grad_norm": 0.7382919788360596,
464
+ "learning_rate": 6.244897959183675e-06,
465
+ "loss": 0.0537,
466
+ "max_memory_allocated (GB)": 57.18,
467
+ "memory_allocated (GB)": 50.57,
468
+ "step": 460,
469
+ "total_memory_available (GB)": 94.62
470
+ },
471
+ {
472
+ "epoch": 1.9183673469387754,
473
+ "grad_norm": 1.3547204732894897,
474
+ "learning_rate": 6.163265306122449e-06,
475
+ "loss": 0.0432,
476
+ "max_memory_allocated (GB)": 57.18,
477
+ "memory_allocated (GB)": 50.57,
478
+ "step": 470,
479
+ "total_memory_available (GB)": 94.62
480
+ },
481
+ {
482
+ "epoch": 1.9591836734693877,
483
+ "grad_norm": 0.19681082665920258,
484
+ "learning_rate": 6.0816326530612245e-06,
485
+ "loss": 0.0498,
486
+ "max_memory_allocated (GB)": 57.18,
487
+ "memory_allocated (GB)": 50.57,
488
+ "step": 480,
489
+ "total_memory_available (GB)": 94.62
490
+ },
491
+ {
492
+ "epoch": 2.0,
493
+ "grad_norm": 1.109737515449524,
494
+ "learning_rate": 6e-06,
495
+ "loss": 0.0639,
496
+ "max_memory_allocated (GB)": 57.18,
497
+ "memory_allocated (GB)": 50.57,
498
+ "step": 490,
499
+ "total_memory_available (GB)": 94.62
500
+ },
501
+ {
502
+ "epoch": 2.0408163265306123,
503
+ "grad_norm": 0.5894625782966614,
504
+ "learning_rate": 5.918367346938776e-06,
505
+ "loss": 0.0593,
506
+ "max_memory_allocated (GB)": 57.18,
507
+ "memory_allocated (GB)": 50.57,
508
+ "step": 500,
509
+ "total_memory_available (GB)": 94.62
510
+ },
511
+ {
512
+ "epoch": 2.0816326530612246,
513
+ "grad_norm": 0.7122555375099182,
514
+ "learning_rate": 5.8367346938775515e-06,
515
+ "loss": 0.0498,
516
+ "max_memory_allocated (GB)": 57.18,
517
+ "memory_allocated (GB)": 50.57,
518
+ "step": 510,
519
+ "total_memory_available (GB)": 94.62
520
+ },
521
+ {
522
+ "epoch": 2.122448979591837,
523
+ "grad_norm": 0.8958902955055237,
524
+ "learning_rate": 5.755102040816327e-06,
525
+ "loss": 0.0457,
526
+ "max_memory_allocated (GB)": 57.18,
527
+ "memory_allocated (GB)": 50.57,
528
+ "step": 520,
529
+ "total_memory_available (GB)": 94.62
530
+ },
531
+ {
532
+ "epoch": 2.163265306122449,
533
+ "grad_norm": 11.620415687561035,
534
+ "learning_rate": 5.673469387755103e-06,
535
+ "loss": 0.0626,
536
+ "max_memory_allocated (GB)": 57.18,
537
+ "memory_allocated (GB)": 50.57,
538
+ "step": 530,
539
+ "total_memory_available (GB)": 94.62
540
+ },
541
+ {
542
+ "epoch": 2.204081632653061,
543
+ "grad_norm": 0.3538230061531067,
544
+ "learning_rate": 5.591836734693878e-06,
545
+ "loss": 0.0584,
546
+ "max_memory_allocated (GB)": 57.18,
547
+ "memory_allocated (GB)": 50.57,
548
+ "step": 540,
549
+ "total_memory_available (GB)": 94.62
550
+ },
551
+ {
552
+ "epoch": 2.2448979591836733,
553
+ "grad_norm": 1.5313146114349365,
554
+ "learning_rate": 5.510204081632653e-06,
555
+ "loss": 0.0627,
556
+ "max_memory_allocated (GB)": 57.18,
557
+ "memory_allocated (GB)": 50.57,
558
+ "step": 550,
559
+ "total_memory_available (GB)": 94.62
560
+ },
561
+ {
562
+ "epoch": 2.2857142857142856,
563
+ "grad_norm": 1.3519809246063232,
564
+ "learning_rate": 5.428571428571429e-06,
565
+ "loss": 0.0572,
566
+ "max_memory_allocated (GB)": 57.18,
567
+ "memory_allocated (GB)": 50.57,
568
+ "step": 560,
569
+ "total_memory_available (GB)": 94.62
570
+ },
571
+ {
572
+ "epoch": 2.326530612244898,
573
+ "grad_norm": 1.0263270139694214,
574
+ "learning_rate": 5.3469387755102045e-06,
575
+ "loss": 0.0585,
576
+ "max_memory_allocated (GB)": 57.18,
577
+ "memory_allocated (GB)": 50.57,
578
+ "step": 570,
579
+ "total_memory_available (GB)": 94.62
580
+ },
581
+ {
582
+ "epoch": 2.36734693877551,
583
+ "grad_norm": 0.8926671147346497,
584
+ "learning_rate": 5.26530612244898e-06,
585
+ "loss": 0.0673,
586
+ "max_memory_allocated (GB)": 57.18,
587
+ "memory_allocated (GB)": 50.57,
588
+ "step": 580,
589
+ "total_memory_available (GB)": 94.62
590
+ },
591
+ {
592
+ "epoch": 2.4081632653061225,
593
+ "grad_norm": 0.3185974955558777,
594
+ "learning_rate": 5.183673469387756e-06,
595
+ "loss": 0.0537,
596
+ "max_memory_allocated (GB)": 57.18,
597
+ "memory_allocated (GB)": 50.57,
598
+ "step": 590,
599
+ "total_memory_available (GB)": 94.62
600
+ },
601
+ {
602
+ "epoch": 2.4489795918367347,
603
+ "grad_norm": 0.944624662399292,
604
+ "learning_rate": 5.1020408163265315e-06,
605
+ "loss": 0.0442,
606
+ "max_memory_allocated (GB)": 57.18,
607
+ "memory_allocated (GB)": 50.57,
608
+ "step": 600,
609
+ "total_memory_available (GB)": 94.62
610
+ },
611
+ {
612
+ "epoch": 2.489795918367347,
613
+ "grad_norm": 0.32796111702919006,
614
+ "learning_rate": 5.020408163265307e-06,
615
+ "loss": 0.0413,
616
+ "max_memory_allocated (GB)": 57.18,
617
+ "memory_allocated (GB)": 50.57,
618
+ "step": 610,
619
+ "total_memory_available (GB)": 94.62
620
+ },
621
+ {
622
+ "epoch": 2.5306122448979593,
623
+ "grad_norm": 0.7929801940917969,
624
+ "learning_rate": 4.938775510204082e-06,
625
+ "loss": 0.0428,
626
+ "max_memory_allocated (GB)": 57.18,
627
+ "memory_allocated (GB)": 50.57,
628
+ "step": 620,
629
+ "total_memory_available (GB)": 94.62
630
+ },
631
+ {
632
+ "epoch": 2.571428571428571,
633
+ "grad_norm": 0.910254955291748,
634
+ "learning_rate": 4.857142857142858e-06,
635
+ "loss": 0.0813,
636
+ "max_memory_allocated (GB)": 57.18,
637
+ "memory_allocated (GB)": 50.57,
638
+ "step": 630,
639
+ "total_memory_available (GB)": 94.62
640
+ },
641
+ {
642
+ "epoch": 2.612244897959184,
643
+ "grad_norm": 1.101942539215088,
644
+ "learning_rate": 4.775510204081633e-06,
645
+ "loss": 0.0495,
646
+ "max_memory_allocated (GB)": 57.18,
647
+ "memory_allocated (GB)": 50.57,
648
+ "step": 640,
649
+ "total_memory_available (GB)": 94.62
650
+ },
651
+ {
652
+ "epoch": 2.6530612244897958,
653
+ "grad_norm": 0.7182526588439941,
654
+ "learning_rate": 4.693877551020409e-06,
655
+ "loss": 0.0471,
656
+ "max_memory_allocated (GB)": 57.18,
657
+ "memory_allocated (GB)": 50.57,
658
+ "step": 650,
659
+ "total_memory_available (GB)": 94.62
660
+ },
661
+ {
662
+ "epoch": 2.693877551020408,
663
+ "grad_norm": 0.8068158626556396,
664
+ "learning_rate": 4.612244897959184e-06,
665
+ "loss": 0.0469,
666
+ "max_memory_allocated (GB)": 57.18,
667
+ "memory_allocated (GB)": 50.57,
668
+ "step": 660,
669
+ "total_memory_available (GB)": 94.62
670
+ },
671
+ {
672
+ "epoch": 2.7346938775510203,
673
+ "grad_norm": 1.2375913858413696,
674
+ "learning_rate": 4.530612244897959e-06,
675
+ "loss": 0.0857,
676
+ "max_memory_allocated (GB)": 57.18,
677
+ "memory_allocated (GB)": 50.57,
678
+ "step": 670,
679
+ "total_memory_available (GB)": 94.62
680
+ },
681
+ {
682
+ "epoch": 2.7755102040816326,
683
+ "grad_norm": 1.1524357795715332,
684
+ "learning_rate": 4.448979591836735e-06,
685
+ "loss": 0.0488,
686
+ "max_memory_allocated (GB)": 57.18,
687
+ "memory_allocated (GB)": 50.57,
688
+ "step": 680,
689
+ "total_memory_available (GB)": 94.62
690
+ },
691
+ {
692
+ "epoch": 2.816326530612245,
693
+ "grad_norm": 0.3913586437702179,
694
+ "learning_rate": 4.367346938775511e-06,
695
+ "loss": 0.0451,
696
+ "max_memory_allocated (GB)": 57.18,
697
+ "memory_allocated (GB)": 50.57,
698
+ "step": 690,
699
+ "total_memory_available (GB)": 94.62
700
+ },
701
+ {
702
+ "epoch": 2.857142857142857,
703
+ "grad_norm": 0.47935113310813904,
704
+ "learning_rate": 4.2857142857142855e-06,
705
+ "loss": 0.0433,
706
+ "max_memory_allocated (GB)": 57.18,
707
+ "memory_allocated (GB)": 50.57,
708
+ "step": 700,
709
+ "total_memory_available (GB)": 94.62
710
+ },
711
+ {
712
+ "epoch": 2.8979591836734695,
713
+ "grad_norm": 0.8084143996238708,
714
+ "learning_rate": 4.204081632653061e-06,
715
+ "loss": 0.0548,
716
+ "max_memory_allocated (GB)": 57.18,
717
+ "memory_allocated (GB)": 50.57,
718
+ "step": 710,
719
+ "total_memory_available (GB)": 94.62
720
+ },
721
+ {
722
+ "epoch": 2.938775510204082,
723
+ "grad_norm": 1.7315497398376465,
724
+ "learning_rate": 4.122448979591837e-06,
725
+ "loss": 0.0587,
726
+ "max_memory_allocated (GB)": 57.18,
727
+ "memory_allocated (GB)": 50.57,
728
+ "step": 720,
729
+ "total_memory_available (GB)": 94.62
730
+ },
731
+ {
732
+ "epoch": 2.979591836734694,
733
+ "grad_norm": 0.20743349194526672,
734
+ "learning_rate": 4.040816326530612e-06,
735
+ "loss": 0.0342,
736
+ "max_memory_allocated (GB)": 57.18,
737
+ "memory_allocated (GB)": 50.57,
738
+ "step": 730,
739
+ "total_memory_available (GB)": 94.62
740
+ },
741
+ {
742
+ "epoch": 3.020408163265306,
743
+ "grad_norm": 0.8024761080741882,
744
+ "learning_rate": 3.959183673469388e-06,
745
+ "loss": 0.053,
746
+ "max_memory_allocated (GB)": 57.18,
747
+ "memory_allocated (GB)": 50.57,
748
+ "step": 740,
749
+ "total_memory_available (GB)": 94.62
750
+ },
751
+ {
752
+ "epoch": 3.061224489795918,
753
+ "grad_norm": 0.45326006412506104,
754
+ "learning_rate": 3.877551020408164e-06,
755
+ "loss": 0.0619,
756
+ "max_memory_allocated (GB)": 57.18,
757
+ "memory_allocated (GB)": 50.57,
758
+ "step": 750,
759
+ "total_memory_available (GB)": 94.62
760
+ },
761
+ {
762
+ "epoch": 3.1020408163265305,
763
+ "grad_norm": 0.6953087449073792,
764
+ "learning_rate": 3.795918367346939e-06,
765
+ "loss": 0.0527,
766
+ "max_memory_allocated (GB)": 57.18,
767
+ "memory_allocated (GB)": 50.57,
768
+ "step": 760,
769
+ "total_memory_available (GB)": 94.62
770
+ },
771
+ {
772
+ "epoch": 3.142857142857143,
773
+ "grad_norm": 1.2290390729904175,
774
+ "learning_rate": 3.7142857142857146e-06,
775
+ "loss": 0.0689,
776
+ "max_memory_allocated (GB)": 57.18,
777
+ "memory_allocated (GB)": 50.57,
778
+ "step": 770,
779
+ "total_memory_available (GB)": 94.62
780
+ },
781
+ {
782
+ "epoch": 3.183673469387755,
783
+ "grad_norm": 0.6281890869140625,
784
+ "learning_rate": 3.6326530612244903e-06,
785
+ "loss": 0.0647,
786
+ "max_memory_allocated (GB)": 57.18,
787
+ "memory_allocated (GB)": 50.57,
788
+ "step": 780,
789
+ "total_memory_available (GB)": 94.62
790
+ },
791
+ {
792
+ "epoch": 3.2244897959183674,
793
+ "grad_norm": 0.3096281588077545,
794
+ "learning_rate": 3.5510204081632655e-06,
795
+ "loss": 0.0522,
796
+ "max_memory_allocated (GB)": 57.18,
797
+ "memory_allocated (GB)": 50.57,
798
+ "step": 790,
799
+ "total_memory_available (GB)": 94.62
800
+ },
801
+ {
802
+ "epoch": 3.2653061224489797,
803
+ "grad_norm": 0.9390127062797546,
804
+ "learning_rate": 3.469387755102041e-06,
805
+ "loss": 0.0432,
806
+ "max_memory_allocated (GB)": 57.18,
807
+ "memory_allocated (GB)": 50.57,
808
+ "step": 800,
809
+ "total_memory_available (GB)": 94.62
810
+ },
811
+ {
812
+ "epoch": 3.306122448979592,
813
+ "grad_norm": 0.87565016746521,
814
+ "learning_rate": 3.3877551020408168e-06,
815
+ "loss": 0.0555,
816
+ "max_memory_allocated (GB)": 57.18,
817
+ "memory_allocated (GB)": 50.57,
818
+ "step": 810,
819
+ "total_memory_available (GB)": 94.62
820
+ },
821
+ {
822
+ "epoch": 3.3469387755102042,
823
+ "grad_norm": 1.0797837972640991,
824
+ "learning_rate": 3.3061224489795924e-06,
825
+ "loss": 0.0455,
826
+ "max_memory_allocated (GB)": 57.18,
827
+ "memory_allocated (GB)": 50.57,
828
+ "step": 820,
829
+ "total_memory_available (GB)": 94.62
830
+ },
831
+ {
832
+ "epoch": 3.387755102040816,
833
+ "grad_norm": 0.3658354878425598,
834
+ "learning_rate": 3.2244897959183672e-06,
835
+ "loss": 0.0487,
836
+ "max_memory_allocated (GB)": 57.18,
837
+ "memory_allocated (GB)": 50.57,
838
+ "step": 830,
839
+ "total_memory_available (GB)": 94.62
840
+ },
841
+ {
842
+ "epoch": 3.4285714285714284,
843
+ "grad_norm": 0.4766336977481842,
844
+ "learning_rate": 3.142857142857143e-06,
845
+ "loss": 0.053,
846
+ "max_memory_allocated (GB)": 57.18,
847
+ "memory_allocated (GB)": 50.57,
848
+ "step": 840,
849
+ "total_memory_available (GB)": 94.62
850
+ },
851
+ {
852
+ "epoch": 3.4693877551020407,
853
+ "grad_norm": 0.49318933486938477,
854
+ "learning_rate": 3.0612244897959185e-06,
855
+ "loss": 0.0812,
856
+ "max_memory_allocated (GB)": 57.18,
857
+ "memory_allocated (GB)": 50.57,
858
+ "step": 850,
859
+ "total_memory_available (GB)": 94.62
860
+ },
861
+ {
862
+ "epoch": 3.510204081632653,
863
+ "grad_norm": 1.3475311994552612,
864
+ "learning_rate": 2.979591836734694e-06,
865
+ "loss": 0.0451,
866
+ "max_memory_allocated (GB)": 57.18,
867
+ "memory_allocated (GB)": 50.57,
868
+ "step": 860,
869
+ "total_memory_available (GB)": 94.62
870
+ },
871
+ {
872
+ "epoch": 3.5510204081632653,
873
+ "grad_norm": 0.36763882637023926,
874
+ "learning_rate": 2.8979591836734694e-06,
875
+ "loss": 0.0646,
876
+ "max_memory_allocated (GB)": 57.18,
877
+ "memory_allocated (GB)": 50.57,
878
+ "step": 870,
879
+ "total_memory_available (GB)": 94.62
880
+ },
881
+ {
882
+ "epoch": 3.5918367346938775,
883
+ "grad_norm": 3.085198402404785,
884
+ "learning_rate": 2.816326530612245e-06,
885
+ "loss": 0.0439,
886
+ "max_memory_allocated (GB)": 57.18,
887
+ "memory_allocated (GB)": 50.57,
888
+ "step": 880,
889
+ "total_memory_available (GB)": 94.62
890
+ },
891
+ {
892
+ "epoch": 3.63265306122449,
893
+ "grad_norm": 0.17229312658309937,
894
+ "learning_rate": 2.7346938775510207e-06,
895
+ "loss": 0.0288,
896
+ "max_memory_allocated (GB)": 57.18,
897
+ "memory_allocated (GB)": 50.57,
898
+ "step": 890,
899
+ "total_memory_available (GB)": 94.62
900
+ },
901
+ {
902
+ "epoch": 3.673469387755102,
903
+ "grad_norm": 1.0760900974273682,
904
+ "learning_rate": 2.6530612244897964e-06,
905
+ "loss": 0.0514,
906
+ "max_memory_allocated (GB)": 57.18,
907
+ "memory_allocated (GB)": 50.57,
908
+ "step": 900,
909
+ "total_memory_available (GB)": 94.62
910
+ },
911
+ {
912
+ "epoch": 3.7142857142857144,
913
+ "grad_norm": 0.45855164527893066,
914
+ "learning_rate": 2.571428571428571e-06,
915
+ "loss": 0.0602,
916
+ "max_memory_allocated (GB)": 57.18,
917
+ "memory_allocated (GB)": 50.57,
918
+ "step": 910,
919
+ "total_memory_available (GB)": 94.62
920
+ },
921
+ {
922
+ "epoch": 3.7551020408163263,
923
+ "grad_norm": 0.15575875341892242,
924
+ "learning_rate": 2.489795918367347e-06,
925
+ "loss": 0.0543,
926
+ "max_memory_allocated (GB)": 57.18,
927
+ "memory_allocated (GB)": 50.57,
928
+ "step": 920,
929
+ "total_memory_available (GB)": 94.62
930
+ },
931
+ {
932
+ "epoch": 3.795918367346939,
933
+ "grad_norm": 0.779755175113678,
934
+ "learning_rate": 2.4081632653061225e-06,
935
+ "loss": 0.0497,
936
+ "max_memory_allocated (GB)": 57.18,
937
+ "memory_allocated (GB)": 50.57,
938
+ "step": 930,
939
+ "total_memory_available (GB)": 94.62
940
+ },
941
+ {
942
+ "epoch": 3.836734693877551,
943
+ "grad_norm": 0.7307060956954956,
944
+ "learning_rate": 2.326530612244898e-06,
945
+ "loss": 0.0486,
946
+ "max_memory_allocated (GB)": 57.18,
947
+ "memory_allocated (GB)": 50.57,
948
+ "step": 940,
949
+ "total_memory_available (GB)": 94.62
950
+ },
951
+ {
952
+ "epoch": 3.877551020408163,
953
+ "grad_norm": 1.062565803527832,
954
+ "learning_rate": 2.244897959183674e-06,
955
+ "loss": 0.0594,
956
+ "max_memory_allocated (GB)": 57.18,
957
+ "memory_allocated (GB)": 50.57,
958
+ "step": 950,
959
+ "total_memory_available (GB)": 94.62
960
+ },
961
+ {
962
+ "epoch": 3.9183673469387754,
963
+ "grad_norm": 0.3031039535999298,
964
+ "learning_rate": 2.1632653061224495e-06,
965
+ "loss": 0.0497,
966
+ "max_memory_allocated (GB)": 57.18,
967
+ "memory_allocated (GB)": 50.57,
968
+ "step": 960,
969
+ "total_memory_available (GB)": 94.62
970
+ },
971
+ {
972
+ "epoch": 3.9591836734693877,
973
+ "grad_norm": 2.310593843460083,
974
+ "learning_rate": 2.0816326530612247e-06,
975
+ "loss": 0.0746,
976
+ "max_memory_allocated (GB)": 57.18,
977
+ "memory_allocated (GB)": 50.57,
978
+ "step": 970,
979
+ "total_memory_available (GB)": 94.62
980
+ },
981
+ {
982
+ "epoch": 4.0,
983
+ "grad_norm": 0.6998704075813293,
984
+ "learning_rate": 2.0000000000000003e-06,
985
+ "loss": 0.0703,
986
+ "max_memory_allocated (GB)": 60.52,
987
+ "memory_allocated (GB)": 50.57,
988
+ "step": 980,
989
+ "total_memory_available (GB)": 94.62
990
+ },
991
+ {
992
+ "epoch": 4.040816326530612,
993
+ "grad_norm": 0.7492395639419556,
994
+ "learning_rate": 1.9183673469387756e-06,
995
+ "loss": 0.0486,
996
+ "max_memory_allocated (GB)": 60.52,
997
+ "memory_allocated (GB)": 50.57,
998
+ "step": 990,
999
+ "total_memory_available (GB)": 94.62
1000
+ },
1001
+ {
1002
+ "epoch": 4.081632653061225,
1003
+ "grad_norm": 0.7633445858955383,
1004
+ "learning_rate": 1.8367346938775512e-06,
1005
+ "loss": 0.0625,
1006
+ "max_memory_allocated (GB)": 60.52,
1007
+ "memory_allocated (GB)": 50.57,
1008
+ "step": 1000,
1009
+ "total_memory_available (GB)": 94.62
1010
+ },
1011
+ {
1012
+ "epoch": 4.122448979591836,
1013
+ "grad_norm": 0.6911561489105225,
1014
+ "learning_rate": 1.7551020408163267e-06,
1015
+ "loss": 0.0632,
1016
+ "max_memory_allocated (GB)": 60.52,
1017
+ "memory_allocated (GB)": 50.57,
1018
+ "step": 1010,
1019
+ "total_memory_available (GB)": 94.62
1020
+ },
1021
+ {
1022
+ "epoch": 4.163265306122449,
1023
+ "grad_norm": 0.33521902561187744,
1024
+ "learning_rate": 1.6734693877551023e-06,
1025
+ "loss": 0.0406,
1026
+ "max_memory_allocated (GB)": 60.52,
1027
+ "memory_allocated (GB)": 50.57,
1028
+ "step": 1020,
1029
+ "total_memory_available (GB)": 94.62
1030
+ },
1031
+ {
1032
+ "epoch": 4.204081632653061,
1033
+ "grad_norm": 0.7509037852287292,
1034
+ "learning_rate": 1.5918367346938775e-06,
1035
+ "loss": 0.0531,
1036
+ "max_memory_allocated (GB)": 60.52,
1037
+ "memory_allocated (GB)": 50.57,
1038
+ "step": 1030,
1039
+ "total_memory_available (GB)": 94.62
1040
+ },
1041
+ {
1042
+ "epoch": 4.244897959183674,
1043
+ "grad_norm": 0.5234070420265198,
1044
+ "learning_rate": 1.5102040816326532e-06,
1045
+ "loss": 0.0396,
1046
+ "max_memory_allocated (GB)": 60.52,
1047
+ "memory_allocated (GB)": 50.57,
1048
+ "step": 1040,
1049
+ "total_memory_available (GB)": 94.62
1050
+ },
1051
+ {
1052
+ "epoch": 4.285714285714286,
1053
+ "grad_norm": 0.7997304797172546,
1054
+ "learning_rate": 1.4285714285714286e-06,
1055
+ "loss": 0.05,
1056
+ "max_memory_allocated (GB)": 60.52,
1057
+ "memory_allocated (GB)": 50.57,
1058
+ "step": 1050,
1059
+ "total_memory_available (GB)": 94.62
1060
+ },
1061
+ {
1062
+ "epoch": 4.326530612244898,
1063
+ "grad_norm": 0.2255077213048935,
1064
+ "learning_rate": 1.3469387755102043e-06,
1065
+ "loss": 0.0457,
1066
+ "max_memory_allocated (GB)": 60.52,
1067
+ "memory_allocated (GB)": 50.57,
1068
+ "step": 1060,
1069
+ "total_memory_available (GB)": 94.62
1070
+ },
1071
+ {
1072
+ "epoch": 4.36734693877551,
1073
+ "grad_norm": 0.5182124376296997,
1074
+ "learning_rate": 1.2653061224489795e-06,
1075
+ "loss": 0.0485,
1076
+ "max_memory_allocated (GB)": 60.52,
1077
+ "memory_allocated (GB)": 50.57,
1078
+ "step": 1070,
1079
+ "total_memory_available (GB)": 94.62
1080
+ },
1081
+ {
1082
+ "epoch": 4.408163265306122,
1083
+ "grad_norm": 0.35046374797821045,
1084
+ "learning_rate": 1.1836734693877552e-06,
1085
+ "loss": 0.0519,
1086
+ "max_memory_allocated (GB)": 60.52,
1087
+ "memory_allocated (GB)": 50.57,
1088
+ "step": 1080,
1089
+ "total_memory_available (GB)": 94.62
1090
+ },
1091
+ {
1092
+ "epoch": 4.448979591836735,
1093
+ "grad_norm": 0.3923434615135193,
1094
+ "learning_rate": 1.1020408163265308e-06,
1095
+ "loss": 0.0507,
1096
+ "max_memory_allocated (GB)": 60.52,
1097
+ "memory_allocated (GB)": 50.57,
1098
+ "step": 1090,
1099
+ "total_memory_available (GB)": 94.62
1100
+ },
1101
+ {
1102
+ "epoch": 4.489795918367347,
1103
+ "grad_norm": 0.23866137862205505,
1104
+ "learning_rate": 1.0204081632653063e-06,
1105
+ "loss": 0.0362,
1106
+ "max_memory_allocated (GB)": 60.52,
1107
+ "memory_allocated (GB)": 50.57,
1108
+ "step": 1100,
1109
+ "total_memory_available (GB)": 94.62
1110
+ },
1111
+ {
1112
+ "epoch": 4.530612244897959,
1113
+ "grad_norm": 0.15117916464805603,
1114
+ "learning_rate": 9.387755102040817e-07,
1115
+ "loss": 0.0464,
1116
+ "max_memory_allocated (GB)": 60.52,
1117
+ "memory_allocated (GB)": 50.57,
1118
+ "step": 1110,
1119
+ "total_memory_available (GB)": 94.62
1120
+ },
1121
+ {
1122
+ "epoch": 4.571428571428571,
1123
+ "grad_norm": 0.5993088483810425,
1124
+ "learning_rate": 8.571428571428572e-07,
1125
+ "loss": 0.0404,
1126
+ "max_memory_allocated (GB)": 60.52,
1127
+ "memory_allocated (GB)": 50.57,
1128
+ "step": 1120,
1129
+ "total_memory_available (GB)": 94.62
1130
+ },
1131
+ {
1132
+ "epoch": 4.612244897959184,
1133
+ "grad_norm": 0.30265432596206665,
1134
+ "learning_rate": 7.755102040816327e-07,
1135
+ "loss": 0.0545,
1136
+ "max_memory_allocated (GB)": 60.52,
1137
+ "memory_allocated (GB)": 50.57,
1138
+ "step": 1130,
1139
+ "total_memory_available (GB)": 94.62
1140
+ },
1141
+ {
1142
+ "epoch": 4.653061224489796,
1143
+ "grad_norm": 0.6385183334350586,
1144
+ "learning_rate": 6.938775510204082e-07,
1145
+ "loss": 0.0731,
1146
+ "max_memory_allocated (GB)": 60.52,
1147
+ "memory_allocated (GB)": 50.57,
1148
+ "step": 1140,
1149
+ "total_memory_available (GB)": 94.62
1150
+ },
1151
+ {
1152
+ "epoch": 4.6938775510204085,
1153
+ "grad_norm": 1.128566026687622,
1154
+ "learning_rate": 6.122448979591837e-07,
1155
+ "loss": 0.0516,
1156
+ "max_memory_allocated (GB)": 60.52,
1157
+ "memory_allocated (GB)": 50.57,
1158
+ "step": 1150,
1159
+ "total_memory_available (GB)": 94.62
1160
+ },
1161
+ {
1162
+ "epoch": 4.73469387755102,
1163
+ "grad_norm": 1.1660116910934448,
1164
+ "learning_rate": 5.306122448979592e-07,
1165
+ "loss": 0.0611,
1166
+ "max_memory_allocated (GB)": 60.52,
1167
+ "memory_allocated (GB)": 50.57,
1168
+ "step": 1160,
1169
+ "total_memory_available (GB)": 94.62
1170
+ },
1171
+ {
1172
+ "epoch": 4.775510204081632,
1173
+ "grad_norm": 0.5327439904212952,
1174
+ "learning_rate": 4.489795918367347e-07,
1175
+ "loss": 0.0549,
1176
+ "max_memory_allocated (GB)": 60.52,
1177
+ "memory_allocated (GB)": 50.57,
1178
+ "step": 1170,
1179
+ "total_memory_available (GB)": 94.62
1180
+ },
1181
+ {
1182
+ "epoch": 4.816326530612245,
1183
+ "grad_norm": 0.8764423131942749,
1184
+ "learning_rate": 3.6734693877551025e-07,
1185
+ "loss": 0.0441,
1186
+ "max_memory_allocated (GB)": 60.52,
1187
+ "memory_allocated (GB)": 50.57,
1188
+ "step": 1180,
1189
+ "total_memory_available (GB)": 94.62
1190
+ },
1191
+ {
1192
+ "epoch": 4.857142857142857,
1193
+ "grad_norm": 0.47835007309913635,
1194
+ "learning_rate": 2.8571428571428575e-07,
1195
+ "loss": 0.0541,
1196
+ "max_memory_allocated (GB)": 60.52,
1197
+ "memory_allocated (GB)": 50.57,
1198
+ "step": 1190,
1199
+ "total_memory_available (GB)": 94.62
1200
+ },
1201
+ {
1202
+ "epoch": 4.8979591836734695,
1203
+ "grad_norm": 1.048047661781311,
1204
+ "learning_rate": 2.0408163265306121e-07,
1205
+ "loss": 0.0731,
1206
+ "max_memory_allocated (GB)": 60.52,
1207
+ "memory_allocated (GB)": 50.57,
1208
+ "step": 1200,
1209
+ "total_memory_available (GB)": 94.62
1210
+ },
1211
+ {
1212
+ "epoch": 4.938775510204081,
1213
+ "grad_norm": 0.3101171851158142,
1214
+ "learning_rate": 1.2244897959183673e-07,
1215
+ "loss": 0.0648,
1216
+ "max_memory_allocated (GB)": 60.52,
1217
+ "memory_allocated (GB)": 50.57,
1218
+ "step": 1210,
1219
+ "total_memory_available (GB)": 94.62
1220
+ },
1221
+ {
1222
+ "epoch": 4.979591836734694,
1223
+ "grad_norm": 0.76802659034729,
1224
+ "learning_rate": 4.0816326530612253e-08,
1225
+ "loss": 0.0418,
1226
+ "max_memory_allocated (GB)": 60.52,
1227
+ "memory_allocated (GB)": 50.57,
1228
+ "step": 1220,
1229
+ "total_memory_available (GB)": 94.62
1230
+ },
1231
+ {
1232
+ "epoch": 5.0,
1233
+ "max_memory_allocated (GB)": 60.52,
1234
+ "memory_allocated (GB)": 50.57,
1235
+ "step": 1225,
1236
+ "total_flos": 3.0598946525952e+16,
1237
+ "total_memory_available (GB)": 94.62,
1238
+ "train_loss": 0.06085505417415074,
1239
+ "train_runtime": 1020.8061,
1240
+ "train_samples_per_second": 55.51,
1241
+ "train_steps_per_second": 1.389
1242
+ }
1243
+ ],
1244
+ "logging_steps": 10,
1245
+ "max_steps": 1225,
1246
+ "num_input_tokens_seen": 0,
1247
+ "num_train_epochs": 5,
1248
+ "save_steps": 500,
1249
+ "stateful_callbacks": {
1250
+ "TrainerControl": {
1251
+ "args": {
1252
+ "should_epoch_stop": false,
1253
+ "should_evaluate": false,
1254
+ "should_log": false,
1255
+ "should_save": true,
1256
+ "should_training_stop": true
1257
+ },
1258
+ "attributes": {}
1259
+ }
1260
+ },
1261
+ "total_flos": 3.0598946525952e+16,
1262
+ "train_batch_size": 40,
1263
+ "trial_name": null,
1264
+ "trial_params": null
1265
+ }
validation_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_loss": 0.1272137612104416,
4
+ "eval_runtime": 30.5176,
5
+ "eval_samples_per_second": 31.06,
6
+ "eval_steps_per_second": 1.994,
7
+ "max_memory_allocated (GB)": 60.52,
8
+ "memory_allocated (GB)": 51.27,
9
+ "total_memory_available (GB)": 94.62
10
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff