salma-elshafey commited on
Commit
c073c39
1 Parent(s): f450c1f

First epoch - no testing yet

Browse files
checkpoint-55500/config.json ADDED
@@ -0,0 +1,182 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "EncoderDecoderModel"
4
+ ],
5
+ "decoder": {
6
+ "_name_or_path": "gpt2",
7
+ "activation_function": "gelu_new",
8
+ "add_cross_attention": true,
9
+ "architectures": [
10
+ "GPT2LMHeadModel"
11
+ ],
12
+ "attn_pdrop": 0.1,
13
+ "bad_words_ids": null,
14
+ "bos_token_id": 50256,
15
+ "chunk_size_feed_forward": 0,
16
+ "cross_attention_hidden_size": null,
17
+ "decoder_start_token_id": null,
18
+ "diversity_penalty": 0.0,
19
+ "do_sample": false,
20
+ "early_stopping": false,
21
+ "embd_pdrop": 0.1,
22
+ "encoder_no_repeat_ngram_size": 0,
23
+ "eos_token_id": 50256,
24
+ "exponential_decay_length_penalty": null,
25
+ "finetuning_task": null,
26
+ "forced_bos_token_id": null,
27
+ "forced_eos_token_id": null,
28
+ "id2label": {
29
+ "0": "LABEL_0",
30
+ "1": "LABEL_1"
31
+ },
32
+ "initializer_range": 0.02,
33
+ "is_decoder": true,
34
+ "is_encoder_decoder": false,
35
+ "label2id": {
36
+ "LABEL_0": 0,
37
+ "LABEL_1": 1
38
+ },
39
+ "layer_norm_epsilon": 1e-05,
40
+ "length_penalty": 1.0,
41
+ "max_length": 20,
42
+ "min_length": 0,
43
+ "model_type": "gpt2",
44
+ "n_ctx": 1024,
45
+ "n_embd": 768,
46
+ "n_head": 12,
47
+ "n_inner": null,
48
+ "n_layer": 12,
49
+ "n_positions": 1024,
50
+ "no_repeat_ngram_size": 0,
51
+ "num_beam_groups": 1,
52
+ "num_beams": 1,
53
+ "num_return_sequences": 1,
54
+ "output_attentions": false,
55
+ "output_hidden_states": false,
56
+ "output_scores": false,
57
+ "pad_token_id": null,
58
+ "prefix": null,
59
+ "problem_type": null,
60
+ "pruned_heads": {},
61
+ "remove_invalid_values": false,
62
+ "reorder_and_upcast_attn": false,
63
+ "repetition_penalty": 1.0,
64
+ "resid_pdrop": 0.1,
65
+ "return_dict": true,
66
+ "return_dict_in_generate": false,
67
+ "scale_attn_by_inverse_layer_idx": false,
68
+ "scale_attn_weights": true,
69
+ "sep_token_id": null,
70
+ "summary_activation": null,
71
+ "summary_first_dropout": 0.1,
72
+ "summary_proj_to_labels": true,
73
+ "summary_type": "cls_index",
74
+ "summary_use_proj": true,
75
+ "task_specific_params": {
76
+ "text-generation": {
77
+ "do_sample": true,
78
+ "max_length": 50
79
+ }
80
+ },
81
+ "temperature": 1.0,
82
+ "tie_encoder_decoder": false,
83
+ "tie_word_embeddings": true,
84
+ "tokenizer_class": null,
85
+ "top_k": 50,
86
+ "top_p": 1.0,
87
+ "torch_dtype": null,
88
+ "torchscript": false,
89
+ "transformers_version": "4.19.2",
90
+ "typical_p": 1.0,
91
+ "use_bfloat16": false,
92
+ "use_cache": false,
93
+ "vocab_size": 50257
94
+ },
95
+ "decoder_start_token_id": 50256,
96
+ "encoder": {
97
+ "_name_or_path": "aubmindlab/bert-base-arabertv02",
98
+ "add_cross_attention": false,
99
+ "architectures": [
100
+ "BertForMaskedLM"
101
+ ],
102
+ "attention_probs_dropout_prob": 0.1,
103
+ "bad_words_ids": null,
104
+ "bos_token_id": null,
105
+ "chunk_size_feed_forward": 0,
106
+ "classifier_dropout": null,
107
+ "cross_attention_hidden_size": null,
108
+ "decoder_start_token_id": null,
109
+ "diversity_penalty": 0.0,
110
+ "do_sample": false,
111
+ "early_stopping": false,
112
+ "encoder_no_repeat_ngram_size": 0,
113
+ "eos_token_id": null,
114
+ "exponential_decay_length_penalty": null,
115
+ "finetuning_task": null,
116
+ "forced_bos_token_id": null,
117
+ "forced_eos_token_id": null,
118
+ "hidden_act": "gelu",
119
+ "hidden_dropout_prob": 0.1,
120
+ "hidden_size": 768,
121
+ "id2label": {
122
+ "0": "LABEL_0",
123
+ "1": "LABEL_1"
124
+ },
125
+ "initializer_range": 0.02,
126
+ "intermediate_size": 3072,
127
+ "is_decoder": false,
128
+ "is_encoder_decoder": false,
129
+ "label2id": {
130
+ "LABEL_0": 0,
131
+ "LABEL_1": 1
132
+ },
133
+ "layer_norm_eps": 1e-12,
134
+ "length_penalty": 1.0,
135
+ "max_length": 20,
136
+ "max_position_embeddings": 512,
137
+ "min_length": 0,
138
+ "model_type": "bert",
139
+ "no_repeat_ngram_size": 0,
140
+ "num_attention_heads": 12,
141
+ "num_beam_groups": 1,
142
+ "num_beams": 1,
143
+ "num_hidden_layers": 12,
144
+ "num_return_sequences": 1,
145
+ "output_attentions": false,
146
+ "output_hidden_states": false,
147
+ "output_scores": false,
148
+ "pad_token_id": 0,
149
+ "position_embedding_type": "absolute",
150
+ "prefix": null,
151
+ "problem_type": null,
152
+ "pruned_heads": {},
153
+ "remove_invalid_values": false,
154
+ "repetition_penalty": 1.0,
155
+ "return_dict": true,
156
+ "return_dict_in_generate": false,
157
+ "sep_token_id": null,
158
+ "task_specific_params": null,
159
+ "temperature": 1.0,
160
+ "tie_encoder_decoder": false,
161
+ "tie_word_embeddings": true,
162
+ "tokenizer_class": null,
163
+ "top_k": 50,
164
+ "top_p": 1.0,
165
+ "torch_dtype": null,
166
+ "torchscript": false,
167
+ "transformers_version": "4.19.2",
168
+ "type_vocab_size": 2,
169
+ "typical_p": 1.0,
170
+ "use_bfloat16": false,
171
+ "use_cache": true,
172
+ "vocab_size": 64000
173
+ },
174
+ "eos_token_id": 50256,
175
+ "is_encoder_decoder": true,
176
+ "max_length": 142,
177
+ "min_length": 1,
178
+ "model_type": "encoder-decoder",
179
+ "no_repeat_ngram_size": 3,
180
+ "torch_dtype": "float32",
181
+ "transformers_version": null
182
+ }
checkpoint-55500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b4ac0570b811dcb79f1b744982b87e48972bcbba423bafb9036da8bea96f45ea
3
+ size 2299531997
checkpoint-55500/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4bd966dba7d33e01eacd2ae8403b0b1b41cc63cbbb7d471c24af9bd525e6ed62
3
+ size 1177334377
checkpoint-55500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f390bd7ffaa6282f6cd50ba6375a6c98dc852d17637c72a0c1d5d3d1d0b70f6f
3
+ size 14503
checkpoint-55500/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:568dae8a745ddda84cb8be0775421d246b24c06c87c687a9b9136ae6762b0ca5
3
+ size 559
checkpoint-55500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b54558c1fae785147987dcbaeb6f42d954e1d903052b42ecab22a3ecf3c7d6c
3
+ size 623
checkpoint-55500/trainer_state.json ADDED
@@ -0,0 +1,1678 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9955156950672646,
5
+ "global_step": 55500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 1.993004484304933e-05,
13
+ "loss": 0.5359,
14
+ "step": 200
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 1.985829596412556e-05,
19
+ "loss": 0.0269,
20
+ "step": 400
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.9786547085201794e-05,
25
+ "loss": 0.0198,
26
+ "step": 600
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 1.9714798206278026e-05,
31
+ "loss": 0.0073,
32
+ "step": 800
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 1.9643049327354262e-05,
37
+ "loss": 0.0052,
38
+ "step": 1000
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 1.9571300448430494e-05,
43
+ "loss": 0.006,
44
+ "step": 1200
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 1.9499551569506727e-05,
49
+ "loss": 0.0038,
50
+ "step": 1400
51
+ },
52
+ {
53
+ "epoch": 0.03,
54
+ "learning_rate": 1.9427802690582963e-05,
55
+ "loss": 0.0026,
56
+ "step": 1600
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 1.9356053811659195e-05,
61
+ "loss": 0.0023,
62
+ "step": 1800
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 1.9284304932735428e-05,
67
+ "loss": 0.0021,
68
+ "step": 2000
69
+ },
70
+ {
71
+ "epoch": 0.04,
72
+ "learning_rate": 1.921255605381166e-05,
73
+ "loss": 0.0034,
74
+ "step": 2200
75
+ },
76
+ {
77
+ "epoch": 0.04,
78
+ "learning_rate": 1.9140807174887893e-05,
79
+ "loss": 0.003,
80
+ "step": 2400
81
+ },
82
+ {
83
+ "epoch": 0.05,
84
+ "learning_rate": 1.9069058295964125e-05,
85
+ "loss": 0.0018,
86
+ "step": 2600
87
+ },
88
+ {
89
+ "epoch": 0.05,
90
+ "learning_rate": 1.8997309417040358e-05,
91
+ "loss": 0.0013,
92
+ "step": 2800
93
+ },
94
+ {
95
+ "epoch": 0.05,
96
+ "learning_rate": 1.8925560538116594e-05,
97
+ "loss": 0.0032,
98
+ "step": 3000
99
+ },
100
+ {
101
+ "epoch": 0.06,
102
+ "learning_rate": 1.8854170403587446e-05,
103
+ "loss": 0.0092,
104
+ "step": 3200
105
+ },
106
+ {
107
+ "epoch": 0.06,
108
+ "learning_rate": 1.878242152466368e-05,
109
+ "loss": 0.0012,
110
+ "step": 3400
111
+ },
112
+ {
113
+ "epoch": 0.06,
114
+ "learning_rate": 1.871067264573991e-05,
115
+ "loss": 0.0016,
116
+ "step": 3600
117
+ },
118
+ {
119
+ "epoch": 0.07,
120
+ "learning_rate": 1.8638923766816147e-05,
121
+ "loss": 0.0027,
122
+ "step": 3800
123
+ },
124
+ {
125
+ "epoch": 0.07,
126
+ "learning_rate": 1.856717488789238e-05,
127
+ "loss": 0.002,
128
+ "step": 4000
129
+ },
130
+ {
131
+ "epoch": 0.08,
132
+ "learning_rate": 1.8495426008968612e-05,
133
+ "loss": 0.0031,
134
+ "step": 4200
135
+ },
136
+ {
137
+ "epoch": 0.08,
138
+ "learning_rate": 1.8423677130044845e-05,
139
+ "loss": 0.0006,
140
+ "step": 4400
141
+ },
142
+ {
143
+ "epoch": 0.08,
144
+ "learning_rate": 1.8351928251121077e-05,
145
+ "loss": 0.0009,
146
+ "step": 4600
147
+ },
148
+ {
149
+ "epoch": 0.09,
150
+ "learning_rate": 1.828017937219731e-05,
151
+ "loss": 0.0027,
152
+ "step": 4800
153
+ },
154
+ {
155
+ "epoch": 0.09,
156
+ "learning_rate": 1.8208430493273542e-05,
157
+ "loss": 0.0009,
158
+ "step": 5000
159
+ },
160
+ {
161
+ "epoch": 0.09,
162
+ "learning_rate": 1.8136681614349778e-05,
163
+ "loss": 0.0007,
164
+ "step": 5200
165
+ },
166
+ {
167
+ "epoch": 0.1,
168
+ "learning_rate": 1.806493273542601e-05,
169
+ "loss": 0.0011,
170
+ "step": 5400
171
+ },
172
+ {
173
+ "epoch": 0.1,
174
+ "learning_rate": 1.7993183856502243e-05,
175
+ "loss": 0.0023,
176
+ "step": 5600
177
+ },
178
+ {
179
+ "epoch": 0.1,
180
+ "learning_rate": 1.7921434977578475e-05,
181
+ "loss": 0.0015,
182
+ "step": 5800
183
+ },
184
+ {
185
+ "epoch": 0.11,
186
+ "learning_rate": 1.784968609865471e-05,
187
+ "loss": 0.0019,
188
+ "step": 6000
189
+ },
190
+ {
191
+ "epoch": 0.11,
192
+ "learning_rate": 1.7777937219730944e-05,
193
+ "loss": 0.0011,
194
+ "step": 6200
195
+ },
196
+ {
197
+ "epoch": 0.11,
198
+ "learning_rate": 1.7706188340807176e-05,
199
+ "loss": 0.0015,
200
+ "step": 6400
201
+ },
202
+ {
203
+ "epoch": 0.12,
204
+ "learning_rate": 1.763443946188341e-05,
205
+ "loss": 0.0007,
206
+ "step": 6600
207
+ },
208
+ {
209
+ "epoch": 0.12,
210
+ "learning_rate": 1.756269058295964e-05,
211
+ "loss": 0.0014,
212
+ "step": 6800
213
+ },
214
+ {
215
+ "epoch": 0.13,
216
+ "learning_rate": 1.7490941704035874e-05,
217
+ "loss": 0.0007,
218
+ "step": 7000
219
+ },
220
+ {
221
+ "epoch": 0.13,
222
+ "learning_rate": 1.741919282511211e-05,
223
+ "loss": 0.0004,
224
+ "step": 7200
225
+ },
226
+ {
227
+ "epoch": 0.13,
228
+ "learning_rate": 1.7347443946188342e-05,
229
+ "loss": 0.0007,
230
+ "step": 7400
231
+ },
232
+ {
233
+ "epoch": 0.14,
234
+ "learning_rate": 1.7275695067264575e-05,
235
+ "loss": 0.0009,
236
+ "step": 7600
237
+ },
238
+ {
239
+ "epoch": 0.14,
240
+ "learning_rate": 1.7203946188340807e-05,
241
+ "loss": 0.0009,
242
+ "step": 7800
243
+ },
244
+ {
245
+ "epoch": 0.14,
246
+ "learning_rate": 1.7132197309417043e-05,
247
+ "loss": 0.0007,
248
+ "step": 8000
249
+ },
250
+ {
251
+ "epoch": 0.15,
252
+ "learning_rate": 1.7060448430493275e-05,
253
+ "loss": 0.0004,
254
+ "step": 8200
255
+ },
256
+ {
257
+ "epoch": 0.15,
258
+ "learning_rate": 1.6988699551569508e-05,
259
+ "loss": 0.0009,
260
+ "step": 8400
261
+ },
262
+ {
263
+ "epoch": 0.15,
264
+ "learning_rate": 1.6916950672645744e-05,
265
+ "loss": 0.0007,
266
+ "step": 8600
267
+ },
268
+ {
269
+ "epoch": 0.16,
270
+ "learning_rate": 1.6845201793721976e-05,
271
+ "loss": 0.0007,
272
+ "step": 8800
273
+ },
274
+ {
275
+ "epoch": 0.16,
276
+ "learning_rate": 1.677345291479821e-05,
277
+ "loss": 0.0019,
278
+ "step": 9000
279
+ },
280
+ {
281
+ "epoch": 0.17,
282
+ "learning_rate": 1.670170403587444e-05,
283
+ "loss": 0.0004,
284
+ "step": 9200
285
+ },
286
+ {
287
+ "epoch": 0.17,
288
+ "learning_rate": 1.6629955156950674e-05,
289
+ "loss": 0.0004,
290
+ "step": 9400
291
+ },
292
+ {
293
+ "epoch": 0.17,
294
+ "learning_rate": 1.6558206278026906e-05,
295
+ "loss": 0.0002,
296
+ "step": 9600
297
+ },
298
+ {
299
+ "epoch": 0.18,
300
+ "learning_rate": 1.648645739910314e-05,
301
+ "loss": 0.0007,
302
+ "step": 9800
303
+ },
304
+ {
305
+ "epoch": 0.18,
306
+ "learning_rate": 1.6414708520179375e-05,
307
+ "loss": 0.0005,
308
+ "step": 10000
309
+ },
310
+ {
311
+ "epoch": 0.18,
312
+ "learning_rate": 1.6342959641255607e-05,
313
+ "loss": 0.0005,
314
+ "step": 10200
315
+ },
316
+ {
317
+ "epoch": 0.19,
318
+ "learning_rate": 1.627121076233184e-05,
319
+ "loss": 0.0007,
320
+ "step": 10400
321
+ },
322
+ {
323
+ "epoch": 0.19,
324
+ "learning_rate": 1.6199461883408075e-05,
325
+ "loss": 0.0007,
326
+ "step": 10600
327
+ },
328
+ {
329
+ "epoch": 0.19,
330
+ "learning_rate": 1.6127713004484308e-05,
331
+ "loss": 0.0011,
332
+ "step": 10800
333
+ },
334
+ {
335
+ "epoch": 0.2,
336
+ "learning_rate": 1.605596412556054e-05,
337
+ "loss": 0.0005,
338
+ "step": 11000
339
+ },
340
+ {
341
+ "epoch": 0.2,
342
+ "learning_rate": 1.5984215246636773e-05,
343
+ "loss": 0.0007,
344
+ "step": 11200
345
+ },
346
+ {
347
+ "epoch": 0.2,
348
+ "learning_rate": 1.5912466367713005e-05,
349
+ "loss": 0.0004,
350
+ "step": 11400
351
+ },
352
+ {
353
+ "epoch": 0.21,
354
+ "learning_rate": 1.5840717488789238e-05,
355
+ "loss": 0.0001,
356
+ "step": 11600
357
+ },
358
+ {
359
+ "epoch": 0.21,
360
+ "learning_rate": 1.576896860986547e-05,
361
+ "loss": 0.0002,
362
+ "step": 11800
363
+ },
364
+ {
365
+ "epoch": 0.22,
366
+ "learning_rate": 1.5697219730941706e-05,
367
+ "loss": 0.0004,
368
+ "step": 12000
369
+ },
370
+ {
371
+ "epoch": 0.22,
372
+ "learning_rate": 1.562547085201794e-05,
373
+ "loss": 0.0003,
374
+ "step": 12200
375
+ },
376
+ {
377
+ "epoch": 0.22,
378
+ "learning_rate": 1.555372197309417e-05,
379
+ "loss": 0.0005,
380
+ "step": 12400
381
+ },
382
+ {
383
+ "epoch": 0.23,
384
+ "learning_rate": 1.5481973094170404e-05,
385
+ "loss": 0.0001,
386
+ "step": 12600
387
+ },
388
+ {
389
+ "epoch": 0.23,
390
+ "learning_rate": 1.541022421524664e-05,
391
+ "loss": 0.0001,
392
+ "step": 12800
393
+ },
394
+ {
395
+ "epoch": 0.23,
396
+ "learning_rate": 1.5338834080717492e-05,
397
+ "loss": 0.0006,
398
+ "step": 13000
399
+ },
400
+ {
401
+ "epoch": 0.24,
402
+ "learning_rate": 1.5267085201793725e-05,
403
+ "loss": 0.0006,
404
+ "step": 13200
405
+ },
406
+ {
407
+ "epoch": 0.24,
408
+ "learning_rate": 1.5195336322869956e-05,
409
+ "loss": 0.0004,
410
+ "step": 13400
411
+ },
412
+ {
413
+ "epoch": 0.24,
414
+ "learning_rate": 1.512358744394619e-05,
415
+ "loss": 0.0003,
416
+ "step": 13600
417
+ },
418
+ {
419
+ "epoch": 0.25,
420
+ "learning_rate": 1.5052197309417043e-05,
421
+ "loss": 0.0009,
422
+ "step": 13800
423
+ },
424
+ {
425
+ "epoch": 0.25,
426
+ "learning_rate": 1.4980448430493275e-05,
427
+ "loss": 0.0002,
428
+ "step": 14000
429
+ },
430
+ {
431
+ "epoch": 0.25,
432
+ "learning_rate": 1.4908699551569508e-05,
433
+ "loss": 0.001,
434
+ "step": 14200
435
+ },
436
+ {
437
+ "epoch": 0.26,
438
+ "learning_rate": 1.483695067264574e-05,
439
+ "loss": 0.0007,
440
+ "step": 14400
441
+ },
442
+ {
443
+ "epoch": 0.26,
444
+ "learning_rate": 1.4765201793721974e-05,
445
+ "loss": 0.0006,
446
+ "step": 14600
447
+ },
448
+ {
449
+ "epoch": 0.27,
450
+ "learning_rate": 1.4693452914798207e-05,
451
+ "loss": 0.0004,
452
+ "step": 14800
453
+ },
454
+ {
455
+ "epoch": 0.27,
456
+ "learning_rate": 1.462170403587444e-05,
457
+ "loss": 0.0004,
458
+ "step": 15000
459
+ },
460
+ {
461
+ "epoch": 0.27,
462
+ "learning_rate": 1.4549955156950675e-05,
463
+ "loss": 0.0004,
464
+ "step": 15200
465
+ },
466
+ {
467
+ "epoch": 0.28,
468
+ "learning_rate": 1.4478206278026908e-05,
469
+ "loss": 0.0003,
470
+ "step": 15400
471
+ },
472
+ {
473
+ "epoch": 0.28,
474
+ "learning_rate": 1.440645739910314e-05,
475
+ "loss": 0.0002,
476
+ "step": 15600
477
+ },
478
+ {
479
+ "epoch": 0.28,
480
+ "learning_rate": 1.4334708520179373e-05,
481
+ "loss": 0.0003,
482
+ "step": 15800
483
+ },
484
+ {
485
+ "epoch": 0.29,
486
+ "learning_rate": 1.4262959641255607e-05,
487
+ "loss": 0.0015,
488
+ "step": 16000
489
+ },
490
+ {
491
+ "epoch": 0.29,
492
+ "learning_rate": 1.419121076233184e-05,
493
+ "loss": 0.0002,
494
+ "step": 16200
495
+ },
496
+ {
497
+ "epoch": 0.29,
498
+ "learning_rate": 1.4119461883408072e-05,
499
+ "loss": 0.0005,
500
+ "step": 16400
501
+ },
502
+ {
503
+ "epoch": 0.3,
504
+ "learning_rate": 1.4047713004484308e-05,
505
+ "loss": 0.0001,
506
+ "step": 16600
507
+ },
508
+ {
509
+ "epoch": 0.3,
510
+ "learning_rate": 1.397596412556054e-05,
511
+ "loss": 0.0009,
512
+ "step": 16800
513
+ },
514
+ {
515
+ "epoch": 0.3,
516
+ "learning_rate": 1.3904215246636773e-05,
517
+ "loss": 0.0001,
518
+ "step": 17000
519
+ },
520
+ {
521
+ "epoch": 0.31,
522
+ "learning_rate": 1.3832466367713007e-05,
523
+ "loss": 0.0001,
524
+ "step": 17200
525
+ },
526
+ {
527
+ "epoch": 0.31,
528
+ "learning_rate": 1.376071748878924e-05,
529
+ "loss": 0.0001,
530
+ "step": 17400
531
+ },
532
+ {
533
+ "epoch": 0.32,
534
+ "learning_rate": 1.3688968609865472e-05,
535
+ "loss": 0.0002,
536
+ "step": 17600
537
+ },
538
+ {
539
+ "epoch": 0.32,
540
+ "learning_rate": 1.3617219730941704e-05,
541
+ "loss": 0.0,
542
+ "step": 17800
543
+ },
544
+ {
545
+ "epoch": 0.32,
546
+ "learning_rate": 1.3545470852017938e-05,
547
+ "loss": 0.0,
548
+ "step": 18000
549
+ },
550
+ {
551
+ "epoch": 0.33,
552
+ "learning_rate": 1.347372197309417e-05,
553
+ "loss": 0.0002,
554
+ "step": 18200
555
+ },
556
+ {
557
+ "epoch": 0.33,
558
+ "learning_rate": 1.3402331838565024e-05,
559
+ "loss": 0.0003,
560
+ "step": 18400
561
+ },
562
+ {
563
+ "epoch": 0.33,
564
+ "learning_rate": 1.3330582959641256e-05,
565
+ "loss": 0.0004,
566
+ "step": 18600
567
+ },
568
+ {
569
+ "epoch": 0.34,
570
+ "learning_rate": 1.325883408071749e-05,
571
+ "loss": 0.0003,
572
+ "step": 18800
573
+ },
574
+ {
575
+ "epoch": 0.34,
576
+ "learning_rate": 1.3187085201793723e-05,
577
+ "loss": 0.0003,
578
+ "step": 19000
579
+ },
580
+ {
581
+ "epoch": 0.34,
582
+ "learning_rate": 1.3115336322869955e-05,
583
+ "loss": 0.0004,
584
+ "step": 19200
585
+ },
586
+ {
587
+ "epoch": 0.35,
588
+ "learning_rate": 1.3043587443946188e-05,
589
+ "loss": 0.0003,
590
+ "step": 19400
591
+ },
592
+ {
593
+ "epoch": 0.35,
594
+ "learning_rate": 1.2971838565022424e-05,
595
+ "loss": 0.0002,
596
+ "step": 19600
597
+ },
598
+ {
599
+ "epoch": 0.36,
600
+ "learning_rate": 1.2900089686098656e-05,
601
+ "loss": 0.0005,
602
+ "step": 19800
603
+ },
604
+ {
605
+ "epoch": 0.36,
606
+ "learning_rate": 1.2828340807174889e-05,
607
+ "loss": 0.0002,
608
+ "step": 20000
609
+ },
610
+ {
611
+ "epoch": 0.36,
612
+ "learning_rate": 1.2756591928251123e-05,
613
+ "loss": 0.0025,
614
+ "step": 20200
615
+ },
616
+ {
617
+ "epoch": 0.37,
618
+ "learning_rate": 1.2684843049327355e-05,
619
+ "loss": 0.0001,
620
+ "step": 20400
621
+ },
622
+ {
623
+ "epoch": 0.37,
624
+ "learning_rate": 1.2613094170403588e-05,
625
+ "loss": 0.0003,
626
+ "step": 20600
627
+ },
628
+ {
629
+ "epoch": 0.37,
630
+ "learning_rate": 1.254134529147982e-05,
631
+ "loss": 0.0001,
632
+ "step": 20800
633
+ },
634
+ {
635
+ "epoch": 0.38,
636
+ "learning_rate": 1.2469596412556056e-05,
637
+ "loss": 0.0004,
638
+ "step": 21000
639
+ },
640
+ {
641
+ "epoch": 0.38,
642
+ "learning_rate": 1.2397847533632289e-05,
643
+ "loss": 0.0001,
644
+ "step": 21200
645
+ },
646
+ {
647
+ "epoch": 0.38,
648
+ "learning_rate": 1.2326098654708521e-05,
649
+ "loss": 0.0002,
650
+ "step": 21400
651
+ },
652
+ {
653
+ "epoch": 0.39,
654
+ "learning_rate": 1.2254349775784755e-05,
655
+ "loss": 0.0,
656
+ "step": 21600
657
+ },
658
+ {
659
+ "epoch": 0.39,
660
+ "learning_rate": 1.2182600896860988e-05,
661
+ "loss": 0.0001,
662
+ "step": 21800
663
+ },
664
+ {
665
+ "epoch": 0.39,
666
+ "learning_rate": 1.211085201793722e-05,
667
+ "loss": 0.0002,
668
+ "step": 22000
669
+ },
670
+ {
671
+ "epoch": 0.4,
672
+ "learning_rate": 1.2039103139013454e-05,
673
+ "loss": 0.0002,
674
+ "step": 22200
675
+ },
676
+ {
677
+ "epoch": 0.4,
678
+ "learning_rate": 1.1967354260089687e-05,
679
+ "loss": 0.0003,
680
+ "step": 22400
681
+ },
682
+ {
683
+ "epoch": 0.41,
684
+ "learning_rate": 1.189560538116592e-05,
685
+ "loss": 0.0009,
686
+ "step": 22600
687
+ },
688
+ {
689
+ "epoch": 0.41,
690
+ "learning_rate": 1.1823856502242152e-05,
691
+ "loss": 0.0001,
692
+ "step": 22800
693
+ },
694
+ {
695
+ "epoch": 0.41,
696
+ "learning_rate": 1.1752107623318388e-05,
697
+ "loss": 0.0001,
698
+ "step": 23000
699
+ },
700
+ {
701
+ "epoch": 0.42,
702
+ "learning_rate": 1.168035874439462e-05,
703
+ "loss": 0.0,
704
+ "step": 23200
705
+ },
706
+ {
707
+ "epoch": 0.42,
708
+ "learning_rate": 1.1608609865470853e-05,
709
+ "loss": 0.0003,
710
+ "step": 23400
711
+ },
712
+ {
713
+ "epoch": 0.42,
714
+ "learning_rate": 1.1536860986547087e-05,
715
+ "loss": 0.0001,
716
+ "step": 23600
717
+ },
718
+ {
719
+ "epoch": 0.43,
720
+ "learning_rate": 1.146511210762332e-05,
721
+ "loss": 0.0,
722
+ "step": 23800
723
+ },
724
+ {
725
+ "epoch": 0.43,
726
+ "learning_rate": 1.1393363228699552e-05,
727
+ "loss": 0.0001,
728
+ "step": 24000
729
+ },
730
+ {
731
+ "epoch": 0.43,
732
+ "learning_rate": 1.1321614349775784e-05,
733
+ "loss": 0.0001,
734
+ "step": 24200
735
+ },
736
+ {
737
+ "epoch": 0.44,
738
+ "learning_rate": 1.124986547085202e-05,
739
+ "loss": 0.0001,
740
+ "step": 24400
741
+ },
742
+ {
743
+ "epoch": 0.44,
744
+ "learning_rate": 1.1178116591928253e-05,
745
+ "loss": 0.0001,
746
+ "step": 24600
747
+ },
748
+ {
749
+ "epoch": 0.44,
750
+ "learning_rate": 1.1106367713004485e-05,
751
+ "loss": 0.0001,
752
+ "step": 24800
753
+ },
754
+ {
755
+ "epoch": 0.45,
756
+ "learning_rate": 1.103461883408072e-05,
757
+ "loss": 0.0001,
758
+ "step": 25000
759
+ },
760
+ {
761
+ "epoch": 0.45,
762
+ "learning_rate": 1.0962869955156952e-05,
763
+ "loss": 0.0001,
764
+ "step": 25200
765
+ },
766
+ {
767
+ "epoch": 0.46,
768
+ "learning_rate": 1.0891121076233184e-05,
769
+ "loss": 0.0003,
770
+ "step": 25400
771
+ },
772
+ {
773
+ "epoch": 0.46,
774
+ "learning_rate": 1.0819372197309419e-05,
775
+ "loss": 0.0001,
776
+ "step": 25600
777
+ },
778
+ {
779
+ "epoch": 0.46,
780
+ "learning_rate": 1.0747623318385651e-05,
781
+ "loss": 0.0001,
782
+ "step": 25800
783
+ },
784
+ {
785
+ "epoch": 0.47,
786
+ "learning_rate": 1.0676233183856504e-05,
787
+ "loss": 0.0005,
788
+ "step": 26000
789
+ },
790
+ {
791
+ "epoch": 0.47,
792
+ "learning_rate": 1.0604484304932736e-05,
793
+ "loss": 0.0001,
794
+ "step": 26200
795
+ },
796
+ {
797
+ "epoch": 0.47,
798
+ "learning_rate": 1.0532735426008969e-05,
799
+ "loss": 0.0001,
800
+ "step": 26400
801
+ },
802
+ {
803
+ "epoch": 0.48,
804
+ "learning_rate": 1.0460986547085203e-05,
805
+ "loss": 0.0001,
806
+ "step": 26600
807
+ },
808
+ {
809
+ "epoch": 0.48,
810
+ "learning_rate": 1.0389237668161435e-05,
811
+ "loss": 0.0002,
812
+ "step": 26800
813
+ },
814
+ {
815
+ "epoch": 0.48,
816
+ "learning_rate": 1.0317488789237668e-05,
817
+ "loss": 0.0,
818
+ "step": 27000
819
+ },
820
+ {
821
+ "epoch": 0.49,
822
+ "learning_rate": 1.0245739910313904e-05,
823
+ "loss": 0.0,
824
+ "step": 27200
825
+ },
826
+ {
827
+ "epoch": 0.49,
828
+ "learning_rate": 1.0173991031390136e-05,
829
+ "loss": 0.0001,
830
+ "step": 27400
831
+ },
832
+ {
833
+ "epoch": 0.5,
834
+ "learning_rate": 1.0102242152466369e-05,
835
+ "loss": 0.0002,
836
+ "step": 27600
837
+ },
838
+ {
839
+ "epoch": 0.5,
840
+ "learning_rate": 1.0030493273542601e-05,
841
+ "loss": 0.0001,
842
+ "step": 27800
843
+ },
844
+ {
845
+ "epoch": 0.5,
846
+ "learning_rate": 9.958744394618834e-06,
847
+ "loss": 0.0001,
848
+ "step": 28000
849
+ },
850
+ {
851
+ "epoch": 0.51,
852
+ "learning_rate": 9.887354260089686e-06,
853
+ "loss": 0.0003,
854
+ "step": 28200
855
+ },
856
+ {
857
+ "epoch": 0.51,
858
+ "learning_rate": 9.81560538116592e-06,
859
+ "loss": 0.0001,
860
+ "step": 28400
861
+ },
862
+ {
863
+ "epoch": 0.51,
864
+ "learning_rate": 9.743856502242153e-06,
865
+ "loss": 0.0,
866
+ "step": 28600
867
+ },
868
+ {
869
+ "epoch": 0.52,
870
+ "learning_rate": 9.672107623318386e-06,
871
+ "loss": 0.0,
872
+ "step": 28800
873
+ },
874
+ {
875
+ "epoch": 0.52,
876
+ "learning_rate": 9.60035874439462e-06,
877
+ "loss": 0.0,
878
+ "step": 29000
879
+ },
880
+ {
881
+ "epoch": 0.52,
882
+ "learning_rate": 9.528609865470852e-06,
883
+ "loss": 0.0001,
884
+ "step": 29200
885
+ },
886
+ {
887
+ "epoch": 0.53,
888
+ "learning_rate": 9.456860986547086e-06,
889
+ "loss": 0.0,
890
+ "step": 29400
891
+ },
892
+ {
893
+ "epoch": 0.53,
894
+ "learning_rate": 9.385112107623319e-06,
895
+ "loss": 0.0001,
896
+ "step": 29600
897
+ },
898
+ {
899
+ "epoch": 0.53,
900
+ "learning_rate": 9.313363228699553e-06,
901
+ "loss": 0.0001,
902
+ "step": 29800
903
+ },
904
+ {
905
+ "epoch": 0.54,
906
+ "learning_rate": 9.241614349775786e-06,
907
+ "loss": 0.0,
908
+ "step": 30000
909
+ },
910
+ {
911
+ "epoch": 0.54,
912
+ "learning_rate": 9.169865470852018e-06,
913
+ "loss": 0.0002,
914
+ "step": 30200
915
+ },
916
+ {
917
+ "epoch": 0.55,
918
+ "learning_rate": 9.098475336322871e-06,
919
+ "loss": 0.0003,
920
+ "step": 30400
921
+ },
922
+ {
923
+ "epoch": 0.55,
924
+ "learning_rate": 9.026726457399105e-06,
925
+ "loss": 0.0001,
926
+ "step": 30600
927
+ },
928
+ {
929
+ "epoch": 0.55,
930
+ "learning_rate": 8.954977578475338e-06,
931
+ "loss": 0.0001,
932
+ "step": 30800
933
+ },
934
+ {
935
+ "epoch": 0.56,
936
+ "learning_rate": 8.88322869955157e-06,
937
+ "loss": 0.0001,
938
+ "step": 31000
939
+ },
940
+ {
941
+ "epoch": 0.56,
942
+ "learning_rate": 8.811479820627803e-06,
943
+ "loss": 0.0,
944
+ "step": 31200
945
+ },
946
+ {
947
+ "epoch": 0.56,
948
+ "learning_rate": 8.739730941704037e-06,
949
+ "loss": 0.0003,
950
+ "step": 31400
951
+ },
952
+ {
953
+ "epoch": 0.57,
954
+ "learning_rate": 8.667982062780271e-06,
955
+ "loss": 0.0001,
956
+ "step": 31600
957
+ },
958
+ {
959
+ "epoch": 0.57,
960
+ "learning_rate": 8.596233183856503e-06,
961
+ "loss": 0.0003,
962
+ "step": 31800
963
+ },
964
+ {
965
+ "epoch": 0.57,
966
+ "learning_rate": 8.524484304932736e-06,
967
+ "loss": 0.0,
968
+ "step": 32000
969
+ },
970
+ {
971
+ "epoch": 0.58,
972
+ "learning_rate": 8.452735426008968e-06,
973
+ "loss": 0.0,
974
+ "step": 32200
975
+ },
976
+ {
977
+ "epoch": 0.58,
978
+ "learning_rate": 8.380986547085203e-06,
979
+ "loss": 0.0,
980
+ "step": 32400
981
+ },
982
+ {
983
+ "epoch": 0.58,
984
+ "learning_rate": 8.309237668161437e-06,
985
+ "loss": 0.0,
986
+ "step": 32600
987
+ },
988
+ {
989
+ "epoch": 0.59,
990
+ "learning_rate": 8.23748878923767e-06,
991
+ "loss": 0.0,
992
+ "step": 32800
993
+ },
994
+ {
995
+ "epoch": 0.59,
996
+ "learning_rate": 8.165739910313902e-06,
997
+ "loss": 0.0001,
998
+ "step": 33000
999
+ },
1000
+ {
1001
+ "epoch": 0.6,
1002
+ "learning_rate": 8.093991031390134e-06,
1003
+ "loss": 0.0003,
1004
+ "step": 33200
1005
+ },
1006
+ {
1007
+ "epoch": 0.6,
1008
+ "learning_rate": 8.022242152466368e-06,
1009
+ "loss": 0.0,
1010
+ "step": 33400
1011
+ },
1012
+ {
1013
+ "epoch": 0.6,
1014
+ "learning_rate": 7.9504932735426e-06,
1015
+ "loss": 0.0,
1016
+ "step": 33600
1017
+ },
1018
+ {
1019
+ "epoch": 0.61,
1020
+ "learning_rate": 7.878744394618835e-06,
1021
+ "loss": 0.0001,
1022
+ "step": 33800
1023
+ },
1024
+ {
1025
+ "epoch": 0.61,
1026
+ "learning_rate": 7.806995515695068e-06,
1027
+ "loss": 0.0,
1028
+ "step": 34000
1029
+ },
1030
+ {
1031
+ "epoch": 0.61,
1032
+ "learning_rate": 7.7352466367713e-06,
1033
+ "loss": 0.0002,
1034
+ "step": 34200
1035
+ },
1036
+ {
1037
+ "epoch": 0.62,
1038
+ "learning_rate": 7.663497757847534e-06,
1039
+ "loss": 0.0,
1040
+ "step": 34400
1041
+ },
1042
+ {
1043
+ "epoch": 0.62,
1044
+ "learning_rate": 7.591748878923767e-06,
1045
+ "loss": 0.0,
1046
+ "step": 34600
1047
+ },
1048
+ {
1049
+ "epoch": 0.62,
1050
+ "learning_rate": 7.520000000000001e-06,
1051
+ "loss": 0.0,
1052
+ "step": 34800
1053
+ },
1054
+ {
1055
+ "epoch": 0.63,
1056
+ "learning_rate": 7.448251121076234e-06,
1057
+ "loss": 0.0001,
1058
+ "step": 35000
1059
+ },
1060
+ {
1061
+ "epoch": 0.63,
1062
+ "learning_rate": 7.376502242152467e-06,
1063
+ "loss": 0.0,
1064
+ "step": 35200
1065
+ },
1066
+ {
1067
+ "epoch": 0.63,
1068
+ "learning_rate": 7.304753363228701e-06,
1069
+ "loss": 0.0,
1070
+ "step": 35400
1071
+ },
1072
+ {
1073
+ "epoch": 0.64,
1074
+ "learning_rate": 7.233004484304933e-06,
1075
+ "loss": 0.0,
1076
+ "step": 35600
1077
+ },
1078
+ {
1079
+ "epoch": 0.64,
1080
+ "learning_rate": 7.161255605381167e-06,
1081
+ "loss": 0.0,
1082
+ "step": 35800
1083
+ },
1084
+ {
1085
+ "epoch": 0.65,
1086
+ "learning_rate": 7.0895067264574e-06,
1087
+ "loss": 0.0,
1088
+ "step": 36000
1089
+ },
1090
+ {
1091
+ "epoch": 0.65,
1092
+ "learning_rate": 7.0177578475336325e-06,
1093
+ "loss": 0.0,
1094
+ "step": 36200
1095
+ },
1096
+ {
1097
+ "epoch": 0.65,
1098
+ "learning_rate": 6.946008968609867e-06,
1099
+ "loss": 0.0,
1100
+ "step": 36400
1101
+ },
1102
+ {
1103
+ "epoch": 0.66,
1104
+ "learning_rate": 6.874260089686099e-06,
1105
+ "loss": 0.0,
1106
+ "step": 36600
1107
+ },
1108
+ {
1109
+ "epoch": 0.66,
1110
+ "learning_rate": 6.8025112107623325e-06,
1111
+ "loss": 0.0,
1112
+ "step": 36800
1113
+ },
1114
+ {
1115
+ "epoch": 0.66,
1116
+ "learning_rate": 6.730762331838565e-06,
1117
+ "loss": 0.0,
1118
+ "step": 37000
1119
+ },
1120
+ {
1121
+ "epoch": 0.67,
1122
+ "learning_rate": 6.659013452914798e-06,
1123
+ "loss": 0.0,
1124
+ "step": 37200
1125
+ },
1126
+ {
1127
+ "epoch": 0.67,
1128
+ "learning_rate": 6.5872645739910325e-06,
1129
+ "loss": 0.0,
1130
+ "step": 37400
1131
+ },
1132
+ {
1133
+ "epoch": 0.67,
1134
+ "learning_rate": 6.515515695067265e-06,
1135
+ "loss": 0.0,
1136
+ "step": 37600
1137
+ },
1138
+ {
1139
+ "epoch": 0.68,
1140
+ "learning_rate": 6.443766816143498e-06,
1141
+ "loss": 0.0,
1142
+ "step": 37800
1143
+ },
1144
+ {
1145
+ "epoch": 0.68,
1146
+ "learning_rate": 6.372017937219731e-06,
1147
+ "loss": 0.0,
1148
+ "step": 38000
1149
+ },
1150
+ {
1151
+ "epoch": 0.69,
1152
+ "learning_rate": 6.300269058295965e-06,
1153
+ "loss": 0.0,
1154
+ "step": 38200
1155
+ },
1156
+ {
1157
+ "epoch": 0.69,
1158
+ "learning_rate": 6.2292376681614354e-06,
1159
+ "loss": 0.0,
1160
+ "step": 38400
1161
+ },
1162
+ {
1163
+ "epoch": 0.69,
1164
+ "learning_rate": 6.157488789237669e-06,
1165
+ "loss": 0.0,
1166
+ "step": 38600
1167
+ },
1168
+ {
1169
+ "epoch": 0.7,
1170
+ "learning_rate": 6.085739910313901e-06,
1171
+ "loss": 0.0,
1172
+ "step": 38800
1173
+ },
1174
+ {
1175
+ "epoch": 0.7,
1176
+ "learning_rate": 6.0139910313901354e-06,
1177
+ "loss": 0.0,
1178
+ "step": 39000
1179
+ },
1180
+ {
1181
+ "epoch": 0.7,
1182
+ "learning_rate": 5.942242152466369e-06,
1183
+ "loss": 0.0,
1184
+ "step": 39200
1185
+ },
1186
+ {
1187
+ "epoch": 0.71,
1188
+ "learning_rate": 5.870493273542601e-06,
1189
+ "loss": 0.0,
1190
+ "step": 39400
1191
+ },
1192
+ {
1193
+ "epoch": 0.71,
1194
+ "learning_rate": 5.798744394618835e-06,
1195
+ "loss": 0.0,
1196
+ "step": 39600
1197
+ },
1198
+ {
1199
+ "epoch": 0.71,
1200
+ "learning_rate": 5.726995515695067e-06,
1201
+ "loss": 0.0,
1202
+ "step": 39800
1203
+ },
1204
+ {
1205
+ "epoch": 0.72,
1206
+ "learning_rate": 5.655246636771301e-06,
1207
+ "loss": 0.0,
1208
+ "step": 40000
1209
+ },
1210
+ {
1211
+ "epoch": 0.72,
1212
+ "learning_rate": 5.583497757847534e-06,
1213
+ "loss": 0.0,
1214
+ "step": 40200
1215
+ },
1216
+ {
1217
+ "epoch": 0.72,
1218
+ "learning_rate": 5.511748878923767e-06,
1219
+ "loss": 0.0,
1220
+ "step": 40400
1221
+ },
1222
+ {
1223
+ "epoch": 0.73,
1224
+ "learning_rate": 5.440358744394619e-06,
1225
+ "loss": 0.0,
1226
+ "step": 40600
1227
+ },
1228
+ {
1229
+ "epoch": 0.73,
1230
+ "learning_rate": 5.368609865470853e-06,
1231
+ "loss": 0.0,
1232
+ "step": 40800
1233
+ },
1234
+ {
1235
+ "epoch": 0.74,
1236
+ "learning_rate": 5.296860986547086e-06,
1237
+ "loss": 0.0,
1238
+ "step": 41000
1239
+ },
1240
+ {
1241
+ "epoch": 0.74,
1242
+ "learning_rate": 5.225112107623319e-06,
1243
+ "loss": 0.0,
1244
+ "step": 41200
1245
+ },
1246
+ {
1247
+ "epoch": 0.74,
1248
+ "learning_rate": 5.1533632286995515e-06,
1249
+ "loss": 0.0,
1250
+ "step": 41400
1251
+ },
1252
+ {
1253
+ "epoch": 0.75,
1254
+ "learning_rate": 5.081973094170403e-06,
1255
+ "loss": 0.0,
1256
+ "step": 41600
1257
+ },
1258
+ {
1259
+ "epoch": 0.75,
1260
+ "learning_rate": 5.0102242152466375e-06,
1261
+ "loss": 0.0002,
1262
+ "step": 41800
1263
+ },
1264
+ {
1265
+ "epoch": 0.75,
1266
+ "learning_rate": 4.938834080717489e-06,
1267
+ "loss": 0.0001,
1268
+ "step": 42000
1269
+ },
1270
+ {
1271
+ "epoch": 0.76,
1272
+ "learning_rate": 4.867085201793723e-06,
1273
+ "loss": 0.0,
1274
+ "step": 42200
1275
+ },
1276
+ {
1277
+ "epoch": 0.76,
1278
+ "learning_rate": 4.795336322869955e-06,
1279
+ "loss": 0.0,
1280
+ "step": 42400
1281
+ },
1282
+ {
1283
+ "epoch": 0.76,
1284
+ "learning_rate": 4.723587443946189e-06,
1285
+ "loss": 0.0,
1286
+ "step": 42600
1287
+ },
1288
+ {
1289
+ "epoch": 0.77,
1290
+ "learning_rate": 4.651838565022422e-06,
1291
+ "loss": 0.0,
1292
+ "step": 42800
1293
+ },
1294
+ {
1295
+ "epoch": 0.77,
1296
+ "learning_rate": 4.580089686098655e-06,
1297
+ "loss": 0.0,
1298
+ "step": 43000
1299
+ },
1300
+ {
1301
+ "epoch": 0.77,
1302
+ "learning_rate": 4.508340807174889e-06,
1303
+ "loss": 0.0001,
1304
+ "step": 43200
1305
+ },
1306
+ {
1307
+ "epoch": 0.78,
1308
+ "learning_rate": 4.436591928251122e-06,
1309
+ "loss": 0.0001,
1310
+ "step": 43400
1311
+ },
1312
+ {
1313
+ "epoch": 0.78,
1314
+ "learning_rate": 4.364843049327354e-06,
1315
+ "loss": 0.0,
1316
+ "step": 43600
1317
+ },
1318
+ {
1319
+ "epoch": 0.79,
1320
+ "learning_rate": 4.293094170403588e-06,
1321
+ "loss": 0.0,
1322
+ "step": 43800
1323
+ },
1324
+ {
1325
+ "epoch": 0.79,
1326
+ "learning_rate": 4.221345291479821e-06,
1327
+ "loss": 0.0,
1328
+ "step": 44000
1329
+ },
1330
+ {
1331
+ "epoch": 0.79,
1332
+ "learning_rate": 4.1495964125560536e-06,
1333
+ "loss": 0.0,
1334
+ "step": 44200
1335
+ },
1336
+ {
1337
+ "epoch": 0.8,
1338
+ "learning_rate": 4.077847533632288e-06,
1339
+ "loss": 0.0,
1340
+ "step": 44400
1341
+ },
1342
+ {
1343
+ "epoch": 0.8,
1344
+ "learning_rate": 4.00609865470852e-06,
1345
+ "loss": 0.0,
1346
+ "step": 44600
1347
+ },
1348
+ {
1349
+ "epoch": 0.8,
1350
+ "learning_rate": 3.934349775784754e-06,
1351
+ "loss": 0.0,
1352
+ "step": 44800
1353
+ },
1354
+ {
1355
+ "epoch": 0.81,
1356
+ "learning_rate": 3.862600896860987e-06,
1357
+ "loss": 0.0,
1358
+ "step": 45000
1359
+ },
1360
+ {
1361
+ "epoch": 0.81,
1362
+ "learning_rate": 3.79085201793722e-06,
1363
+ "loss": 0.0,
1364
+ "step": 45200
1365
+ },
1366
+ {
1367
+ "epoch": 0.81,
1368
+ "learning_rate": 3.719103139013453e-06,
1369
+ "loss": 0.0,
1370
+ "step": 45400
1371
+ },
1372
+ {
1373
+ "epoch": 0.82,
1374
+ "learning_rate": 3.6473542600896865e-06,
1375
+ "loss": 0.0,
1376
+ "step": 45600
1377
+ },
1378
+ {
1379
+ "epoch": 0.82,
1380
+ "learning_rate": 3.57560538116592e-06,
1381
+ "loss": 0.0002,
1382
+ "step": 45800
1383
+ },
1384
+ {
1385
+ "epoch": 0.83,
1386
+ "learning_rate": 3.5038565022421527e-06,
1387
+ "loss": 0.0,
1388
+ "step": 46000
1389
+ },
1390
+ {
1391
+ "epoch": 0.83,
1392
+ "learning_rate": 3.432107623318386e-06,
1393
+ "loss": 0.0001,
1394
+ "step": 46200
1395
+ },
1396
+ {
1397
+ "epoch": 0.83,
1398
+ "learning_rate": 3.360717488789238e-06,
1399
+ "loss": 0.0001,
1400
+ "step": 46400
1401
+ },
1402
+ {
1403
+ "epoch": 0.84,
1404
+ "learning_rate": 3.288968609865471e-06,
1405
+ "loss": 0.0,
1406
+ "step": 46600
1407
+ },
1408
+ {
1409
+ "epoch": 0.84,
1410
+ "learning_rate": 3.217219730941704e-06,
1411
+ "loss": 0.0,
1412
+ "step": 46800
1413
+ },
1414
+ {
1415
+ "epoch": 0.84,
1416
+ "learning_rate": 3.145470852017937e-06,
1417
+ "loss": 0.0001,
1418
+ "step": 47000
1419
+ },
1420
+ {
1421
+ "epoch": 0.85,
1422
+ "learning_rate": 3.073721973094171e-06,
1423
+ "loss": 0.0,
1424
+ "step": 47200
1425
+ },
1426
+ {
1427
+ "epoch": 0.85,
1428
+ "learning_rate": 3.001973094170404e-06,
1429
+ "loss": 0.0,
1430
+ "step": 47400
1431
+ },
1432
+ {
1433
+ "epoch": 0.85,
1434
+ "learning_rate": 2.930224215246637e-06,
1435
+ "loss": 0.0,
1436
+ "step": 47600
1437
+ },
1438
+ {
1439
+ "epoch": 0.86,
1440
+ "learning_rate": 2.85847533632287e-06,
1441
+ "loss": 0.0,
1442
+ "step": 47800
1443
+ },
1444
+ {
1445
+ "epoch": 0.86,
1446
+ "learning_rate": 2.7867264573991034e-06,
1447
+ "loss": 0.0,
1448
+ "step": 48000
1449
+ },
1450
+ {
1451
+ "epoch": 0.86,
1452
+ "learning_rate": 2.7149775784753363e-06,
1453
+ "loss": 0.0,
1454
+ "step": 48200
1455
+ },
1456
+ {
1457
+ "epoch": 0.87,
1458
+ "learning_rate": 2.64322869955157e-06,
1459
+ "loss": 0.0,
1460
+ "step": 48400
1461
+ },
1462
+ {
1463
+ "epoch": 0.87,
1464
+ "learning_rate": 2.571479820627803e-06,
1465
+ "loss": 0.0,
1466
+ "step": 48600
1467
+ },
1468
+ {
1469
+ "epoch": 0.88,
1470
+ "learning_rate": 2.499730941704036e-06,
1471
+ "loss": 0.0,
1472
+ "step": 48800
1473
+ },
1474
+ {
1475
+ "epoch": 0.88,
1476
+ "learning_rate": 2.427982062780269e-06,
1477
+ "loss": 0.0001,
1478
+ "step": 49000
1479
+ },
1480
+ {
1481
+ "epoch": 0.88,
1482
+ "learning_rate": 2.3562331838565025e-06,
1483
+ "loss": 0.0,
1484
+ "step": 49200
1485
+ },
1486
+ {
1487
+ "epoch": 0.89,
1488
+ "learning_rate": 2.2844843049327355e-06,
1489
+ "loss": 0.0001,
1490
+ "step": 49400
1491
+ },
1492
+ {
1493
+ "epoch": 0.89,
1494
+ "learning_rate": 2.2127354260089688e-06,
1495
+ "loss": 0.0,
1496
+ "step": 49600
1497
+ },
1498
+ {
1499
+ "epoch": 0.89,
1500
+ "learning_rate": 2.140986547085202e-06,
1501
+ "loss": 0.0,
1502
+ "step": 49800
1503
+ },
1504
+ {
1505
+ "epoch": 0.9,
1506
+ "learning_rate": 2.069237668161435e-06,
1507
+ "loss": 0.0,
1508
+ "step": 50000
1509
+ },
1510
+ {
1511
+ "epoch": 0.9,
1512
+ "learning_rate": 1.9974887892376684e-06,
1513
+ "loss": 0.0,
1514
+ "step": 50200
1515
+ },
1516
+ {
1517
+ "epoch": 0.9,
1518
+ "learning_rate": 1.9257399103139017e-06,
1519
+ "loss": 0.0,
1520
+ "step": 50400
1521
+ },
1522
+ {
1523
+ "epoch": 0.91,
1524
+ "learning_rate": 1.8543497757847534e-06,
1525
+ "loss": 0.0001,
1526
+ "step": 50600
1527
+ },
1528
+ {
1529
+ "epoch": 0.91,
1530
+ "learning_rate": 1.7826008968609867e-06,
1531
+ "loss": 0.0,
1532
+ "step": 50800
1533
+ },
1534
+ {
1535
+ "epoch": 0.91,
1536
+ "learning_rate": 1.7108520179372198e-06,
1537
+ "loss": 0.0,
1538
+ "step": 51000
1539
+ },
1540
+ {
1541
+ "epoch": 0.92,
1542
+ "learning_rate": 1.639103139013453e-06,
1543
+ "loss": 0.0,
1544
+ "step": 51200
1545
+ },
1546
+ {
1547
+ "epoch": 0.92,
1548
+ "learning_rate": 1.5673542600896863e-06,
1549
+ "loss": 0.0,
1550
+ "step": 51400
1551
+ },
1552
+ {
1553
+ "epoch": 0.93,
1554
+ "learning_rate": 1.4956053811659194e-06,
1555
+ "loss": 0.0,
1556
+ "step": 51600
1557
+ },
1558
+ {
1559
+ "epoch": 0.93,
1560
+ "learning_rate": 1.4238565022421528e-06,
1561
+ "loss": 0.0,
1562
+ "step": 51800
1563
+ },
1564
+ {
1565
+ "epoch": 0.93,
1566
+ "learning_rate": 1.3521076233183859e-06,
1567
+ "loss": 0.0,
1568
+ "step": 52000
1569
+ },
1570
+ {
1571
+ "epoch": 0.94,
1572
+ "learning_rate": 1.2803587443946188e-06,
1573
+ "loss": 0.0,
1574
+ "step": 52200
1575
+ },
1576
+ {
1577
+ "epoch": 0.94,
1578
+ "learning_rate": 1.2089686098654709e-06,
1579
+ "loss": 0.0,
1580
+ "step": 52400
1581
+ },
1582
+ {
1583
+ "epoch": 0.94,
1584
+ "learning_rate": 1.137219730941704e-06,
1585
+ "loss": 0.0,
1586
+ "step": 52600
1587
+ },
1588
+ {
1589
+ "epoch": 0.95,
1590
+ "learning_rate": 1.0654708520179373e-06,
1591
+ "loss": 0.0,
1592
+ "step": 52800
1593
+ },
1594
+ {
1595
+ "epoch": 0.95,
1596
+ "learning_rate": 9.937219730941705e-07,
1597
+ "loss": 0.0,
1598
+ "step": 53000
1599
+ },
1600
+ {
1601
+ "epoch": 0.95,
1602
+ "learning_rate": 9.219730941704037e-07,
1603
+ "loss": 0.0,
1604
+ "step": 53200
1605
+ },
1606
+ {
1607
+ "epoch": 0.96,
1608
+ "learning_rate": 8.502242152466368e-07,
1609
+ "loss": 0.0,
1610
+ "step": 53400
1611
+ },
1612
+ {
1613
+ "epoch": 0.96,
1614
+ "learning_rate": 7.7847533632287e-07,
1615
+ "loss": 0.0,
1616
+ "step": 53600
1617
+ },
1618
+ {
1619
+ "epoch": 0.97,
1620
+ "learning_rate": 7.067264573991033e-07,
1621
+ "loss": 0.0,
1622
+ "step": 53800
1623
+ },
1624
+ {
1625
+ "epoch": 0.97,
1626
+ "learning_rate": 6.349775784753363e-07,
1627
+ "loss": 0.0,
1628
+ "step": 54000
1629
+ },
1630
+ {
1631
+ "epoch": 0.97,
1632
+ "learning_rate": 5.632286995515695e-07,
1633
+ "loss": 0.0,
1634
+ "step": 54200
1635
+ },
1636
+ {
1637
+ "epoch": 0.98,
1638
+ "learning_rate": 4.914798206278028e-07,
1639
+ "loss": 0.0,
1640
+ "step": 54400
1641
+ },
1642
+ {
1643
+ "epoch": 0.98,
1644
+ "learning_rate": 4.1973094170403593e-07,
1645
+ "loss": 0.0,
1646
+ "step": 54600
1647
+ },
1648
+ {
1649
+ "epoch": 0.98,
1650
+ "learning_rate": 3.4798206278026905e-07,
1651
+ "loss": 0.0,
1652
+ "step": 54800
1653
+ },
1654
+ {
1655
+ "epoch": 0.99,
1656
+ "learning_rate": 2.762331838565023e-07,
1657
+ "loss": 0.0,
1658
+ "step": 55000
1659
+ },
1660
+ {
1661
+ "epoch": 0.99,
1662
+ "learning_rate": 2.0448430493273546e-07,
1663
+ "loss": 0.0,
1664
+ "step": 55200
1665
+ },
1666
+ {
1667
+ "epoch": 0.99,
1668
+ "learning_rate": 1.327354260089686e-07,
1669
+ "loss": 0.0,
1670
+ "step": 55400
1671
+ }
1672
+ ],
1673
+ "max_steps": 55750,
1674
+ "num_train_epochs": 1,
1675
+ "total_flos": 6.788108648448e+16,
1676
+ "trial_name": null,
1677
+ "trial_params": null
1678
+ }
checkpoint-55500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:978270834e3f82890d9fc59e32b8d0822f371d114dd06ed8aebecb51859893bc
3
+ size 3247