lck0328 commited on
Commit
dad2fa9
·
verified ·
1 Parent(s): 3e249d4

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. config.json +59 -0
  2. optimizer.pt +3 -0
  3. pytorch_model.bin +3 -0
  4. rng_state.pth +3 -0
  5. scaler.pt +3 -0
  6. scheduler.pt +3 -0
  7. trainer_state.json +2424 -0
  8. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "models/sft_actor/",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "d_ff": 4096,
8
+ "d_kv": 64,
9
+ "d_model": 1024,
10
+ "decoder_start_token_id": 0,
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 2,
13
+ "feed_forward_proj": "relu",
14
+ "initializer_factor": 1.0,
15
+ "is_encoder_decoder": true,
16
+ "layer_norm_epsilon": 1e-06,
17
+ "model_type": "t5",
18
+ "n_positions": 512,
19
+ "num_decoder_layers": 24,
20
+ "num_heads": 16,
21
+ "num_layers": 24,
22
+ "output_past": true,
23
+ "pad_token_id": 0,
24
+ "relative_attention_max_distance": 128,
25
+ "relative_attention_num_buckets": 32,
26
+ "task_specific_params": {
27
+ "summarization": {
28
+ "early_stopping": true,
29
+ "length_penalty": 2.0,
30
+ "max_length": 200,
31
+ "min_length": 30,
32
+ "no_repeat_ngram_size": 3,
33
+ "num_beams": 4,
34
+ "prefix": "summarize: "
35
+ },
36
+ "translation_en_to_de": {
37
+ "early_stopping": true,
38
+ "max_length": 300,
39
+ "num_beams": 4,
40
+ "prefix": "translate English to German: "
41
+ },
42
+ "translation_en_to_fr": {
43
+ "early_stopping": true,
44
+ "max_length": 300,
45
+ "num_beams": 4,
46
+ "prefix": "translate English to French: "
47
+ },
48
+ "translation_en_to_ro": {
49
+ "early_stopping": true,
50
+ "max_length": 300,
51
+ "num_beams": 4,
52
+ "prefix": "translate English to Romanian: "
53
+ }
54
+ },
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.16.1",
57
+ "use_cache": true,
58
+ "vocab_size": 32100
59
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e821a1fb85ef3281527f19fb4f97acf920be595f2fbbbcebbc1f306a548a453e
3
+ size 5901421892
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26e7c83386e2c9b0aa1b819ca4b5e80cf20188ad27b8956ff72846879d9c697b
3
+ size 2950794110
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15185b3ba6bb896d8f37cefd066ddb20b42965cd8bd9db642cef7e4eba1dd18b
3
+ size 14244
scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a534c81b078841f39052452a213056eb89c801491a333cde25bdfe50814d7929
3
+ size 988
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:460338ac735a27a355cf65794e2b7ff0888458069f6e11987ff9aca351d291d7
3
+ size 1064
trainer_state.json ADDED
@@ -0,0 +1,2424 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.0013335833802173,
5
+ "global_step": 3000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "acc": 0.0,
12
+ "epoch": 0.0,
13
+ "learning_rate": 2e-05,
14
+ "loss": 0.1583,
15
+ "rl_loss": 0.0773,
16
+ "step": 1
17
+ },
18
+ {
19
+ "acc": 0.0,
20
+ "epoch": 0.01,
21
+ "learning_rate": 2e-05,
22
+ "loss": 0.1955,
23
+ "rl_loss": -0.0025,
24
+ "step": 10
25
+ },
26
+ {
27
+ "acc": 0.0,
28
+ "epoch": 0.01,
29
+ "learning_rate": 2e-05,
30
+ "loss": 0.1993,
31
+ "rl_loss": -0.0069,
32
+ "step": 20
33
+ },
34
+ {
35
+ "acc": 0.0,
36
+ "epoch": 0.02,
37
+ "learning_rate": 2e-05,
38
+ "loss": 0.2039,
39
+ "rl_loss": -0.0095,
40
+ "step": 30
41
+ },
42
+ {
43
+ "acc": 0.0,
44
+ "epoch": 0.03,
45
+ "learning_rate": 2e-05,
46
+ "loss": 0.1866,
47
+ "rl_loss": -0.0069,
48
+ "step": 40
49
+ },
50
+ {
51
+ "acc": 0.0,
52
+ "epoch": 0.03,
53
+ "learning_rate": 2e-05,
54
+ "loss": 0.2016,
55
+ "rl_loss": -0.012,
56
+ "step": 50
57
+ },
58
+ {
59
+ "acc": 0.0,
60
+ "epoch": 0.04,
61
+ "learning_rate": 2e-05,
62
+ "loss": 0.1938,
63
+ "rl_loss": -0.0042,
64
+ "step": 60
65
+ },
66
+ {
67
+ "acc": 0.0,
68
+ "epoch": 0.05,
69
+ "learning_rate": 2e-05,
70
+ "loss": 0.1973,
71
+ "rl_loss": -0.0091,
72
+ "step": 70
73
+ },
74
+ {
75
+ "acc": 0.0,
76
+ "epoch": 0.05,
77
+ "learning_rate": 2e-05,
78
+ "loss": 0.1937,
79
+ "rl_loss": -0.0073,
80
+ "step": 80
81
+ },
82
+ {
83
+ "acc": 0.0,
84
+ "epoch": 0.06,
85
+ "learning_rate": 2e-05,
86
+ "loss": 0.1995,
87
+ "rl_loss": -0.0092,
88
+ "step": 90
89
+ },
90
+ {
91
+ "acc": 0.0,
92
+ "epoch": 0.07,
93
+ "learning_rate": 2e-05,
94
+ "loss": 0.1954,
95
+ "rl_loss": -0.0138,
96
+ "step": 100
97
+ },
98
+ {
99
+ "acc": 0.0,
100
+ "epoch": 0.07,
101
+ "learning_rate": 2e-05,
102
+ "loss": 0.1963,
103
+ "rl_loss": -0.009,
104
+ "step": 110
105
+ },
106
+ {
107
+ "acc": 0.0,
108
+ "epoch": 0.08,
109
+ "learning_rate": 2e-05,
110
+ "loss": 0.2045,
111
+ "rl_loss": -0.0015,
112
+ "step": 120
113
+ },
114
+ {
115
+ "acc": 0.0,
116
+ "epoch": 0.09,
117
+ "learning_rate": 2e-05,
118
+ "loss": 0.1878,
119
+ "rl_loss": -0.0053,
120
+ "step": 130
121
+ },
122
+ {
123
+ "acc": 0.0,
124
+ "epoch": 0.09,
125
+ "learning_rate": 2e-05,
126
+ "loss": 0.1941,
127
+ "rl_loss": -0.0131,
128
+ "step": 140
129
+ },
130
+ {
131
+ "acc": 0.0,
132
+ "epoch": 0.1,
133
+ "learning_rate": 2e-05,
134
+ "loss": 0.1934,
135
+ "rl_loss": -0.0095,
136
+ "step": 150
137
+ },
138
+ {
139
+ "acc": 0.0,
140
+ "epoch": 0.11,
141
+ "learning_rate": 2e-05,
142
+ "loss": 0.1897,
143
+ "rl_loss": -0.0079,
144
+ "step": 160
145
+ },
146
+ {
147
+ "acc": 0.0,
148
+ "epoch": 0.11,
149
+ "learning_rate": 2e-05,
150
+ "loss": 0.1909,
151
+ "rl_loss": -0.0124,
152
+ "step": 170
153
+ },
154
+ {
155
+ "acc": 0.0,
156
+ "epoch": 0.12,
157
+ "learning_rate": 2e-05,
158
+ "loss": 0.192,
159
+ "rl_loss": -0.0121,
160
+ "step": 180
161
+ },
162
+ {
163
+ "acc": 0.0,
164
+ "epoch": 0.13,
165
+ "learning_rate": 2e-05,
166
+ "loss": 0.1972,
167
+ "rl_loss": 0.0008,
168
+ "step": 190
169
+ },
170
+ {
171
+ "acc": 0.0,
172
+ "epoch": 0.13,
173
+ "learning_rate": 2e-05,
174
+ "loss": 0.2,
175
+ "rl_loss": -0.0046,
176
+ "step": 200
177
+ },
178
+ {
179
+ "acc": 0.0,
180
+ "epoch": 0.14,
181
+ "learning_rate": 2e-05,
182
+ "loss": 0.198,
183
+ "rl_loss": -0.0104,
184
+ "step": 210
185
+ },
186
+ {
187
+ "acc": 0.0,
188
+ "epoch": 0.15,
189
+ "learning_rate": 2e-05,
190
+ "loss": 0.1969,
191
+ "rl_loss": -0.0194,
192
+ "step": 220
193
+ },
194
+ {
195
+ "acc": 0.0,
196
+ "epoch": 0.15,
197
+ "learning_rate": 2e-05,
198
+ "loss": 0.1933,
199
+ "rl_loss": -0.0105,
200
+ "step": 230
201
+ },
202
+ {
203
+ "acc": 0.0,
204
+ "epoch": 0.16,
205
+ "learning_rate": 2e-05,
206
+ "loss": 0.181,
207
+ "rl_loss": -0.0166,
208
+ "step": 240
209
+ },
210
+ {
211
+ "acc": 0.0,
212
+ "epoch": 0.17,
213
+ "learning_rate": 2e-05,
214
+ "loss": 0.1853,
215
+ "rl_loss": -0.0174,
216
+ "step": 250
217
+ },
218
+ {
219
+ "acc": 0.0,
220
+ "epoch": 0.17,
221
+ "learning_rate": 2e-05,
222
+ "loss": 0.206,
223
+ "rl_loss": -0.0192,
224
+ "step": 260
225
+ },
226
+ {
227
+ "acc": 0.0,
228
+ "epoch": 0.18,
229
+ "learning_rate": 2e-05,
230
+ "loss": 0.1902,
231
+ "rl_loss": -0.0071,
232
+ "step": 270
233
+ },
234
+ {
235
+ "acc": 0.0,
236
+ "epoch": 0.19,
237
+ "learning_rate": 2e-05,
238
+ "loss": 0.2093,
239
+ "rl_loss": -0.0172,
240
+ "step": 280
241
+ },
242
+ {
243
+ "acc": 0.0,
244
+ "epoch": 0.19,
245
+ "learning_rate": 2e-05,
246
+ "loss": 0.1916,
247
+ "rl_loss": -0.0226,
248
+ "step": 290
249
+ },
250
+ {
251
+ "acc": 0.0,
252
+ "epoch": 0.2,
253
+ "learning_rate": 2e-05,
254
+ "loss": 0.1899,
255
+ "rl_loss": -0.0175,
256
+ "step": 300
257
+ },
258
+ {
259
+ "acc": 0.0,
260
+ "epoch": 0.21,
261
+ "learning_rate": 2e-05,
262
+ "loss": 0.2031,
263
+ "rl_loss": -0.0283,
264
+ "step": 310
265
+ },
266
+ {
267
+ "acc": 0.0,
268
+ "epoch": 0.21,
269
+ "learning_rate": 2e-05,
270
+ "loss": 0.2002,
271
+ "rl_loss": -0.0079,
272
+ "step": 320
273
+ },
274
+ {
275
+ "acc": 0.0,
276
+ "epoch": 0.22,
277
+ "learning_rate": 2e-05,
278
+ "loss": 0.2052,
279
+ "rl_loss": -0.0195,
280
+ "step": 330
281
+ },
282
+ {
283
+ "acc": 0.0,
284
+ "epoch": 0.23,
285
+ "learning_rate": 2e-05,
286
+ "loss": 0.207,
287
+ "rl_loss": -0.0204,
288
+ "step": 340
289
+ },
290
+ {
291
+ "acc": 0.0,
292
+ "epoch": 0.23,
293
+ "learning_rate": 2e-05,
294
+ "loss": 0.1994,
295
+ "rl_loss": -0.0027,
296
+ "step": 350
297
+ },
298
+ {
299
+ "acc": 0.0,
300
+ "epoch": 0.24,
301
+ "learning_rate": 2e-05,
302
+ "loss": 0.1969,
303
+ "rl_loss": -0.0091,
304
+ "step": 360
305
+ },
306
+ {
307
+ "acc": 0.0,
308
+ "epoch": 0.25,
309
+ "learning_rate": 2e-05,
310
+ "loss": 0.1933,
311
+ "rl_loss": -0.0141,
312
+ "step": 370
313
+ },
314
+ {
315
+ "acc": 0.0,
316
+ "epoch": 0.25,
317
+ "learning_rate": 2e-05,
318
+ "loss": 0.1981,
319
+ "rl_loss": -0.0188,
320
+ "step": 380
321
+ },
322
+ {
323
+ "acc": 0.0,
324
+ "epoch": 0.26,
325
+ "learning_rate": 2e-05,
326
+ "loss": 0.1954,
327
+ "rl_loss": -0.0101,
328
+ "step": 390
329
+ },
330
+ {
331
+ "acc": 0.0,
332
+ "epoch": 0.27,
333
+ "learning_rate": 2e-05,
334
+ "loss": 0.1946,
335
+ "rl_loss": -0.0218,
336
+ "step": 400
337
+ },
338
+ {
339
+ "acc": 0.0,
340
+ "epoch": 0.27,
341
+ "learning_rate": 2e-05,
342
+ "loss": 0.1913,
343
+ "rl_loss": -0.0112,
344
+ "step": 410
345
+ },
346
+ {
347
+ "acc": 0.0,
348
+ "epoch": 0.28,
349
+ "learning_rate": 2e-05,
350
+ "loss": 0.2008,
351
+ "rl_loss": -0.006,
352
+ "step": 420
353
+ },
354
+ {
355
+ "acc": 0.0,
356
+ "epoch": 0.29,
357
+ "learning_rate": 2e-05,
358
+ "loss": 0.1908,
359
+ "rl_loss": -0.0199,
360
+ "step": 430
361
+ },
362
+ {
363
+ "acc": 0.0,
364
+ "epoch": 0.29,
365
+ "learning_rate": 2e-05,
366
+ "loss": 0.192,
367
+ "rl_loss": -0.0201,
368
+ "step": 440
369
+ },
370
+ {
371
+ "acc": 0.0,
372
+ "epoch": 0.3,
373
+ "learning_rate": 2e-05,
374
+ "loss": 0.2064,
375
+ "rl_loss": -0.0144,
376
+ "step": 450
377
+ },
378
+ {
379
+ "acc": 0.0,
380
+ "epoch": 0.31,
381
+ "learning_rate": 2e-05,
382
+ "loss": 0.1998,
383
+ "rl_loss": -0.0295,
384
+ "step": 460
385
+ },
386
+ {
387
+ "acc": 0.0,
388
+ "epoch": 0.31,
389
+ "learning_rate": 2e-05,
390
+ "loss": 0.1996,
391
+ "rl_loss": -0.0151,
392
+ "step": 470
393
+ },
394
+ {
395
+ "acc": 0.0,
396
+ "epoch": 0.32,
397
+ "learning_rate": 2e-05,
398
+ "loss": 0.2179,
399
+ "rl_loss": -0.0062,
400
+ "step": 480
401
+ },
402
+ {
403
+ "acc": 0.0,
404
+ "epoch": 0.33,
405
+ "learning_rate": 2e-05,
406
+ "loss": 0.189,
407
+ "rl_loss": -0.014,
408
+ "step": 490
409
+ },
410
+ {
411
+ "acc": 0.0,
412
+ "epoch": 0.33,
413
+ "learning_rate": 2e-05,
414
+ "loss": 0.2153,
415
+ "rl_loss": -0.0176,
416
+ "step": 500
417
+ },
418
+ {
419
+ "acc": 0.0,
420
+ "epoch": 0.34,
421
+ "learning_rate": 2e-05,
422
+ "loss": 0.1954,
423
+ "rl_loss": -0.0051,
424
+ "step": 510
425
+ },
426
+ {
427
+ "acc": 0.0,
428
+ "epoch": 0.35,
429
+ "learning_rate": 2e-05,
430
+ "loss": 0.1877,
431
+ "rl_loss": -0.0055,
432
+ "step": 520
433
+ },
434
+ {
435
+ "acc": 0.0,
436
+ "epoch": 0.35,
437
+ "learning_rate": 2e-05,
438
+ "loss": 0.1952,
439
+ "rl_loss": -0.0128,
440
+ "step": 530
441
+ },
442
+ {
443
+ "acc": 0.0,
444
+ "epoch": 0.36,
445
+ "learning_rate": 2e-05,
446
+ "loss": 0.1946,
447
+ "rl_loss": -0.0079,
448
+ "step": 540
449
+ },
450
+ {
451
+ "acc": 0.0,
452
+ "epoch": 0.37,
453
+ "learning_rate": 2e-05,
454
+ "loss": 0.2009,
455
+ "rl_loss": -0.0594,
456
+ "step": 550
457
+ },
458
+ {
459
+ "acc": 0.0,
460
+ "epoch": 0.37,
461
+ "learning_rate": 2e-05,
462
+ "loss": 0.2025,
463
+ "rl_loss": -0.0103,
464
+ "step": 560
465
+ },
466
+ {
467
+ "acc": 0.0,
468
+ "epoch": 0.38,
469
+ "learning_rate": 2e-05,
470
+ "loss": 0.1932,
471
+ "rl_loss": -0.0185,
472
+ "step": 570
473
+ },
474
+ {
475
+ "acc": 0.0,
476
+ "epoch": 0.39,
477
+ "learning_rate": 2e-05,
478
+ "loss": 0.207,
479
+ "rl_loss": -0.0175,
480
+ "step": 580
481
+ },
482
+ {
483
+ "acc": 0.0,
484
+ "epoch": 0.39,
485
+ "learning_rate": 2e-05,
486
+ "loss": 0.1869,
487
+ "rl_loss": -0.0559,
488
+ "step": 590
489
+ },
490
+ {
491
+ "acc": 0.0,
492
+ "epoch": 0.4,
493
+ "learning_rate": 2e-05,
494
+ "loss": 0.2019,
495
+ "rl_loss": -0.009,
496
+ "step": 600
497
+ },
498
+ {
499
+ "acc": 0.0,
500
+ "epoch": 0.41,
501
+ "learning_rate": 2e-05,
502
+ "loss": 0.2078,
503
+ "rl_loss": -0.0115,
504
+ "step": 610
505
+ },
506
+ {
507
+ "acc": 0.0,
508
+ "epoch": 0.41,
509
+ "learning_rate": 2e-05,
510
+ "loss": 0.1918,
511
+ "rl_loss": -0.0094,
512
+ "step": 620
513
+ },
514
+ {
515
+ "acc": 0.0,
516
+ "epoch": 0.42,
517
+ "learning_rate": 2e-05,
518
+ "loss": 0.2049,
519
+ "rl_loss": -0.0436,
520
+ "step": 630
521
+ },
522
+ {
523
+ "acc": 0.0,
524
+ "epoch": 0.43,
525
+ "learning_rate": 2e-05,
526
+ "loss": 0.1891,
527
+ "rl_loss": -0.0128,
528
+ "step": 640
529
+ },
530
+ {
531
+ "acc": 0.0,
532
+ "epoch": 0.43,
533
+ "learning_rate": 2e-05,
534
+ "loss": 0.1971,
535
+ "rl_loss": -0.0107,
536
+ "step": 650
537
+ },
538
+ {
539
+ "acc": 0.0,
540
+ "epoch": 0.44,
541
+ "learning_rate": 2e-05,
542
+ "loss": 0.2326,
543
+ "rl_loss": -0.0503,
544
+ "step": 660
545
+ },
546
+ {
547
+ "acc": 0.0,
548
+ "epoch": 0.45,
549
+ "learning_rate": 2e-05,
550
+ "loss": 0.2079,
551
+ "rl_loss": -0.0093,
552
+ "step": 670
553
+ },
554
+ {
555
+ "acc": 0.0,
556
+ "epoch": 0.45,
557
+ "learning_rate": 2e-05,
558
+ "loss": 0.1986,
559
+ "rl_loss": -0.0099,
560
+ "step": 680
561
+ },
562
+ {
563
+ "acc": 0.0,
564
+ "epoch": 0.46,
565
+ "learning_rate": 2e-05,
566
+ "loss": 0.2051,
567
+ "rl_loss": -0.0126,
568
+ "step": 690
569
+ },
570
+ {
571
+ "acc": 0.0,
572
+ "epoch": 0.47,
573
+ "learning_rate": 2e-05,
574
+ "loss": 0.2038,
575
+ "rl_loss": -0.0255,
576
+ "step": 700
577
+ },
578
+ {
579
+ "acc": 0.0,
580
+ "epoch": 0.47,
581
+ "learning_rate": 2e-05,
582
+ "loss": 0.1982,
583
+ "rl_loss": -0.1333,
584
+ "step": 710
585
+ },
586
+ {
587
+ "acc": 0.0,
588
+ "epoch": 0.48,
589
+ "learning_rate": 2e-05,
590
+ "loss": 0.1991,
591
+ "rl_loss": -0.0114,
592
+ "step": 720
593
+ },
594
+ {
595
+ "acc": 0.0,
596
+ "epoch": 0.49,
597
+ "learning_rate": 2e-05,
598
+ "loss": 0.1953,
599
+ "rl_loss": -0.0084,
600
+ "step": 730
601
+ },
602
+ {
603
+ "acc": 0.0,
604
+ "epoch": 0.49,
605
+ "learning_rate": 2e-05,
606
+ "loss": 0.1884,
607
+ "rl_loss": -0.0047,
608
+ "step": 740
609
+ },
610
+ {
611
+ "acc": 0.0,
612
+ "epoch": 0.5,
613
+ "learning_rate": 2e-05,
614
+ "loss": 0.2078,
615
+ "rl_loss": -0.0588,
616
+ "step": 750
617
+ },
618
+ {
619
+ "acc": 0.0,
620
+ "epoch": 0.51,
621
+ "learning_rate": 2e-05,
622
+ "loss": 0.2105,
623
+ "rl_loss": -0.0238,
624
+ "step": 760
625
+ },
626
+ {
627
+ "acc": 0.0,
628
+ "epoch": 0.51,
629
+ "learning_rate": 2e-05,
630
+ "loss": 0.1908,
631
+ "rl_loss": -0.0122,
632
+ "step": 770
633
+ },
634
+ {
635
+ "acc": 0.0,
636
+ "epoch": 0.52,
637
+ "learning_rate": 2e-05,
638
+ "loss": 0.1997,
639
+ "rl_loss": -0.039,
640
+ "step": 780
641
+ },
642
+ {
643
+ "acc": 0.0,
644
+ "epoch": 0.53,
645
+ "learning_rate": 2e-05,
646
+ "loss": 0.1942,
647
+ "rl_loss": -0.0666,
648
+ "step": 790
649
+ },
650
+ {
651
+ "acc": 0.0,
652
+ "epoch": 0.53,
653
+ "learning_rate": 2e-05,
654
+ "loss": 0.2019,
655
+ "rl_loss": -0.0694,
656
+ "step": 800
657
+ },
658
+ {
659
+ "acc": 0.0,
660
+ "epoch": 0.54,
661
+ "learning_rate": 2e-05,
662
+ "loss": 0.2407,
663
+ "rl_loss": -0.0509,
664
+ "step": 810
665
+ },
666
+ {
667
+ "acc": 0.0,
668
+ "epoch": 0.55,
669
+ "learning_rate": 2e-05,
670
+ "loss": 0.2672,
671
+ "rl_loss": 0.0018,
672
+ "step": 820
673
+ },
674
+ {
675
+ "acc": 0.0,
676
+ "epoch": 0.55,
677
+ "learning_rate": 2e-05,
678
+ "loss": 0.2007,
679
+ "rl_loss": -0.0075,
680
+ "step": 830
681
+ },
682
+ {
683
+ "acc": 0.0,
684
+ "epoch": 0.56,
685
+ "learning_rate": 2e-05,
686
+ "loss": 0.2021,
687
+ "rl_loss": -0.072,
688
+ "step": 840
689
+ },
690
+ {
691
+ "acc": 0.0,
692
+ "epoch": 0.57,
693
+ "learning_rate": 2e-05,
694
+ "loss": 0.2401,
695
+ "rl_loss": -0.0257,
696
+ "step": 850
697
+ },
698
+ {
699
+ "acc": 0.0,
700
+ "epoch": 0.57,
701
+ "learning_rate": 2e-05,
702
+ "loss": 0.2168,
703
+ "rl_loss": -0.0371,
704
+ "step": 860
705
+ },
706
+ {
707
+ "acc": 0.0,
708
+ "epoch": 0.58,
709
+ "learning_rate": 2e-05,
710
+ "loss": 0.2008,
711
+ "rl_loss": -0.0549,
712
+ "step": 870
713
+ },
714
+ {
715
+ "acc": 0.0,
716
+ "epoch": 0.59,
717
+ "learning_rate": 2e-05,
718
+ "loss": 0.1992,
719
+ "rl_loss": -0.0534,
720
+ "step": 880
721
+ },
722
+ {
723
+ "acc": 0.0,
724
+ "epoch": 0.59,
725
+ "learning_rate": 2e-05,
726
+ "loss": 0.2226,
727
+ "rl_loss": -0.0824,
728
+ "step": 890
729
+ },
730
+ {
731
+ "acc": 0.0,
732
+ "epoch": 0.6,
733
+ "learning_rate": 2e-05,
734
+ "loss": 0.2067,
735
+ "rl_loss": -0.0092,
736
+ "step": 900
737
+ },
738
+ {
739
+ "acc": 0.0,
740
+ "epoch": 0.61,
741
+ "learning_rate": 2e-05,
742
+ "loss": 0.2192,
743
+ "rl_loss": -0.0334,
744
+ "step": 910
745
+ },
746
+ {
747
+ "acc": 0.0,
748
+ "epoch": 0.61,
749
+ "learning_rate": 2e-05,
750
+ "loss": 0.2127,
751
+ "rl_loss": -0.0402,
752
+ "step": 920
753
+ },
754
+ {
755
+ "acc": 0.0,
756
+ "epoch": 0.62,
757
+ "learning_rate": 2e-05,
758
+ "loss": 0.2399,
759
+ "rl_loss": -0.0901,
760
+ "step": 930
761
+ },
762
+ {
763
+ "acc": 0.0,
764
+ "epoch": 0.63,
765
+ "learning_rate": 2e-05,
766
+ "loss": 0.2302,
767
+ "rl_loss": -0.0274,
768
+ "step": 940
769
+ },
770
+ {
771
+ "acc": 0.0,
772
+ "epoch": 0.63,
773
+ "learning_rate": 2e-05,
774
+ "loss": 0.2229,
775
+ "rl_loss": -0.0317,
776
+ "step": 950
777
+ },
778
+ {
779
+ "acc": 0.0,
780
+ "epoch": 0.64,
781
+ "learning_rate": 2e-05,
782
+ "loss": 0.1961,
783
+ "rl_loss": 0.0009,
784
+ "step": 960
785
+ },
786
+ {
787
+ "acc": 0.0,
788
+ "epoch": 0.65,
789
+ "learning_rate": 2e-05,
790
+ "loss": 0.1909,
791
+ "rl_loss": -0.015,
792
+ "step": 970
793
+ },
794
+ {
795
+ "acc": 0.0,
796
+ "epoch": 0.65,
797
+ "learning_rate": 2e-05,
798
+ "loss": 0.2071,
799
+ "rl_loss": -0.0675,
800
+ "step": 980
801
+ },
802
+ {
803
+ "acc": 0.0,
804
+ "epoch": 0.66,
805
+ "learning_rate": 2e-05,
806
+ "loss": 0.2073,
807
+ "rl_loss": -0.0286,
808
+ "step": 990
809
+ },
810
+ {
811
+ "acc": 0.0,
812
+ "epoch": 0.67,
813
+ "learning_rate": 2e-05,
814
+ "loss": 0.2251,
815
+ "rl_loss": 0.0071,
816
+ "step": 1000
817
+ },
818
+ {
819
+ "acc": 0.0,
820
+ "epoch": 0.67,
821
+ "learning_rate": 2e-05,
822
+ "loss": 0.1874,
823
+ "rl_loss": -0.0176,
824
+ "step": 1010
825
+ },
826
+ {
827
+ "acc": 0.0,
828
+ "epoch": 0.68,
829
+ "learning_rate": 2e-05,
830
+ "loss": 0.2342,
831
+ "rl_loss": -0.0224,
832
+ "step": 1020
833
+ },
834
+ {
835
+ "acc": 0.0,
836
+ "epoch": 0.69,
837
+ "learning_rate": 2e-05,
838
+ "loss": 0.2244,
839
+ "rl_loss": -0.0204,
840
+ "step": 1030
841
+ },
842
+ {
843
+ "acc": 0.0,
844
+ "epoch": 0.69,
845
+ "learning_rate": 2e-05,
846
+ "loss": 0.2059,
847
+ "rl_loss": -0.0206,
848
+ "step": 1040
849
+ },
850
+ {
851
+ "acc": 0.0,
852
+ "epoch": 0.7,
853
+ "learning_rate": 2e-05,
854
+ "loss": 0.1919,
855
+ "rl_loss": -0.0619,
856
+ "step": 1050
857
+ },
858
+ {
859
+ "acc": 0.0,
860
+ "epoch": 0.71,
861
+ "learning_rate": 2e-05,
862
+ "loss": 0.2103,
863
+ "rl_loss": -0.0141,
864
+ "step": 1060
865
+ },
866
+ {
867
+ "acc": 0.0,
868
+ "epoch": 0.71,
869
+ "learning_rate": 2e-05,
870
+ "loss": 0.2018,
871
+ "rl_loss": -0.0241,
872
+ "step": 1070
873
+ },
874
+ {
875
+ "acc": 0.0,
876
+ "epoch": 0.72,
877
+ "learning_rate": 2e-05,
878
+ "loss": 0.2551,
879
+ "rl_loss": -0.0541,
880
+ "step": 1080
881
+ },
882
+ {
883
+ "acc": 0.0,
884
+ "epoch": 0.73,
885
+ "learning_rate": 2e-05,
886
+ "loss": 0.201,
887
+ "rl_loss": -0.0098,
888
+ "step": 1090
889
+ },
890
+ {
891
+ "acc": 0.0,
892
+ "epoch": 0.73,
893
+ "learning_rate": 2e-05,
894
+ "loss": 0.2008,
895
+ "rl_loss": -0.0105,
896
+ "step": 1100
897
+ },
898
+ {
899
+ "acc": 0.0,
900
+ "epoch": 0.74,
901
+ "learning_rate": 2e-05,
902
+ "loss": 0.2028,
903
+ "rl_loss": -0.05,
904
+ "step": 1110
905
+ },
906
+ {
907
+ "acc": 0.0,
908
+ "epoch": 0.75,
909
+ "learning_rate": 2e-05,
910
+ "loss": 0.1966,
911
+ "rl_loss": -0.006,
912
+ "step": 1120
913
+ },
914
+ {
915
+ "acc": 0.0,
916
+ "epoch": 0.75,
917
+ "learning_rate": 2e-05,
918
+ "loss": 0.2089,
919
+ "rl_loss": -0.0066,
920
+ "step": 1130
921
+ },
922
+ {
923
+ "acc": 0.0,
924
+ "epoch": 0.76,
925
+ "learning_rate": 2e-05,
926
+ "loss": 0.1951,
927
+ "rl_loss": -0.0268,
928
+ "step": 1140
929
+ },
930
+ {
931
+ "acc": 0.0,
932
+ "epoch": 0.77,
933
+ "learning_rate": 2e-05,
934
+ "loss": 0.1885,
935
+ "rl_loss": -0.0478,
936
+ "step": 1150
937
+ },
938
+ {
939
+ "acc": 0.0,
940
+ "epoch": 0.77,
941
+ "learning_rate": 2e-05,
942
+ "loss": 0.2258,
943
+ "rl_loss": -0.0383,
944
+ "step": 1160
945
+ },
946
+ {
947
+ "acc": 0.0,
948
+ "epoch": 0.78,
949
+ "learning_rate": 2e-05,
950
+ "loss": 0.2002,
951
+ "rl_loss": -0.0201,
952
+ "step": 1170
953
+ },
954
+ {
955
+ "acc": 0.0,
956
+ "epoch": 0.79,
957
+ "learning_rate": 2e-05,
958
+ "loss": 0.2319,
959
+ "rl_loss": -0.1138,
960
+ "step": 1180
961
+ },
962
+ {
963
+ "acc": 0.0,
964
+ "epoch": 0.79,
965
+ "learning_rate": 2e-05,
966
+ "loss": 0.2828,
967
+ "rl_loss": -0.0367,
968
+ "step": 1190
969
+ },
970
+ {
971
+ "acc": 0.0,
972
+ "epoch": 0.8,
973
+ "learning_rate": 2e-05,
974
+ "loss": 0.2212,
975
+ "rl_loss": -0.1415,
976
+ "step": 1200
977
+ },
978
+ {
979
+ "acc": 0.0,
980
+ "epoch": 0.81,
981
+ "learning_rate": 2e-05,
982
+ "loss": 0.2619,
983
+ "rl_loss": -0.0687,
984
+ "step": 1210
985
+ },
986
+ {
987
+ "acc": 0.0,
988
+ "epoch": 0.81,
989
+ "learning_rate": 2e-05,
990
+ "loss": 0.2034,
991
+ "rl_loss": -0.0475,
992
+ "step": 1220
993
+ },
994
+ {
995
+ "acc": 0.0,
996
+ "epoch": 0.82,
997
+ "learning_rate": 2e-05,
998
+ "loss": 0.2583,
999
+ "rl_loss": -0.2057,
1000
+ "step": 1230
1001
+ },
1002
+ {
1003
+ "acc": 0.0,
1004
+ "epoch": 0.83,
1005
+ "learning_rate": 2e-05,
1006
+ "loss": 0.2258,
1007
+ "rl_loss": -0.0758,
1008
+ "step": 1240
1009
+ },
1010
+ {
1011
+ "acc": 0.0,
1012
+ "epoch": 0.83,
1013
+ "learning_rate": 2e-05,
1014
+ "loss": 0.2015,
1015
+ "rl_loss": -0.2222,
1016
+ "step": 1250
1017
+ },
1018
+ {
1019
+ "acc": 0.0,
1020
+ "epoch": 0.84,
1021
+ "learning_rate": 2e-05,
1022
+ "loss": 0.2088,
1023
+ "rl_loss": -0.0407,
1024
+ "step": 1260
1025
+ },
1026
+ {
1027
+ "acc": 0.0,
1028
+ "epoch": 0.85,
1029
+ "learning_rate": 2e-05,
1030
+ "loss": 0.1986,
1031
+ "rl_loss": -0.0952,
1032
+ "step": 1270
1033
+ },
1034
+ {
1035
+ "acc": 0.0,
1036
+ "epoch": 0.85,
1037
+ "learning_rate": 2e-05,
1038
+ "loss": 0.2644,
1039
+ "rl_loss": -0.0654,
1040
+ "step": 1280
1041
+ },
1042
+ {
1043
+ "acc": 0.0,
1044
+ "epoch": 0.86,
1045
+ "learning_rate": 2e-05,
1046
+ "loss": 0.2119,
1047
+ "rl_loss": -0.099,
1048
+ "step": 1290
1049
+ },
1050
+ {
1051
+ "acc": 0.0,
1052
+ "epoch": 0.87,
1053
+ "learning_rate": 2e-05,
1054
+ "loss": 0.2162,
1055
+ "rl_loss": 0.003,
1056
+ "step": 1300
1057
+ },
1058
+ {
1059
+ "acc": 0.0,
1060
+ "epoch": 0.87,
1061
+ "learning_rate": 2e-05,
1062
+ "loss": 0.2345,
1063
+ "rl_loss": -0.1278,
1064
+ "step": 1310
1065
+ },
1066
+ {
1067
+ "acc": 0.0,
1068
+ "epoch": 0.88,
1069
+ "learning_rate": 2e-05,
1070
+ "loss": 0.2147,
1071
+ "rl_loss": 0.0178,
1072
+ "step": 1320
1073
+ },
1074
+ {
1075
+ "acc": 0.0,
1076
+ "epoch": 0.89,
1077
+ "learning_rate": 2e-05,
1078
+ "loss": 0.3013,
1079
+ "rl_loss": -0.0425,
1080
+ "step": 1330
1081
+ },
1082
+ {
1083
+ "acc": 0.0,
1084
+ "epoch": 0.89,
1085
+ "learning_rate": 2e-05,
1086
+ "loss": 0.2049,
1087
+ "rl_loss": -0.0238,
1088
+ "step": 1340
1089
+ },
1090
+ {
1091
+ "acc": 0.0,
1092
+ "epoch": 0.9,
1093
+ "learning_rate": 2e-05,
1094
+ "loss": 0.2489,
1095
+ "rl_loss": -0.0539,
1096
+ "step": 1350
1097
+ },
1098
+ {
1099
+ "acc": 0.0,
1100
+ "epoch": 0.91,
1101
+ "learning_rate": 2e-05,
1102
+ "loss": 0.2009,
1103
+ "rl_loss": -0.0724,
1104
+ "step": 1360
1105
+ },
1106
+ {
1107
+ "acc": 0.0,
1108
+ "epoch": 0.91,
1109
+ "learning_rate": 2e-05,
1110
+ "loss": 0.2411,
1111
+ "rl_loss": -0.1009,
1112
+ "step": 1370
1113
+ },
1114
+ {
1115
+ "acc": 0.0,
1116
+ "epoch": 0.92,
1117
+ "learning_rate": 2e-05,
1118
+ "loss": 0.2153,
1119
+ "rl_loss": -0.059,
1120
+ "step": 1380
1121
+ },
1122
+ {
1123
+ "acc": 0.0,
1124
+ "epoch": 0.93,
1125
+ "learning_rate": 2e-05,
1126
+ "loss": 0.2806,
1127
+ "rl_loss": -0.0582,
1128
+ "step": 1390
1129
+ },
1130
+ {
1131
+ "acc": 0.0,
1132
+ "epoch": 0.93,
1133
+ "learning_rate": 2e-05,
1134
+ "loss": 0.211,
1135
+ "rl_loss": -0.0547,
1136
+ "step": 1400
1137
+ },
1138
+ {
1139
+ "acc": 0.0,
1140
+ "epoch": 0.94,
1141
+ "learning_rate": 2e-05,
1142
+ "loss": 0.2064,
1143
+ "rl_loss": -0.0155,
1144
+ "step": 1410
1145
+ },
1146
+ {
1147
+ "acc": 0.0,
1148
+ "epoch": 0.95,
1149
+ "learning_rate": 2e-05,
1150
+ "loss": 0.2573,
1151
+ "rl_loss": -0.206,
1152
+ "step": 1420
1153
+ },
1154
+ {
1155
+ "acc": 0.0,
1156
+ "epoch": 0.95,
1157
+ "learning_rate": 2e-05,
1158
+ "loss": 0.3427,
1159
+ "rl_loss": -0.4052,
1160
+ "step": 1430
1161
+ },
1162
+ {
1163
+ "acc": 0.0,
1164
+ "epoch": 0.96,
1165
+ "learning_rate": 2e-05,
1166
+ "loss": 0.2292,
1167
+ "rl_loss": -0.0413,
1168
+ "step": 1440
1169
+ },
1170
+ {
1171
+ "acc": 0.0,
1172
+ "epoch": 0.97,
1173
+ "learning_rate": 2e-05,
1174
+ "loss": 0.2241,
1175
+ "rl_loss": -0.0907,
1176
+ "step": 1450
1177
+ },
1178
+ {
1179
+ "acc": 0.0,
1180
+ "epoch": 0.97,
1181
+ "learning_rate": 2e-05,
1182
+ "loss": 0.2004,
1183
+ "rl_loss": -0.2518,
1184
+ "step": 1460
1185
+ },
1186
+ {
1187
+ "acc": 0.0,
1188
+ "epoch": 0.98,
1189
+ "learning_rate": 2e-05,
1190
+ "loss": 0.2882,
1191
+ "rl_loss": -0.1383,
1192
+ "step": 1470
1193
+ },
1194
+ {
1195
+ "acc": 0.0,
1196
+ "epoch": 0.99,
1197
+ "learning_rate": 2e-05,
1198
+ "loss": 0.4526,
1199
+ "rl_loss": -0.3262,
1200
+ "step": 1480
1201
+ },
1202
+ {
1203
+ "acc": 0.0,
1204
+ "epoch": 0.99,
1205
+ "learning_rate": 2e-05,
1206
+ "loss": 0.2733,
1207
+ "rl_loss": -0.078,
1208
+ "step": 1490
1209
+ },
1210
+ {
1211
+ "acc": 0.0,
1212
+ "epoch": 1.0,
1213
+ "learning_rate": 2e-05,
1214
+ "loss": 0.24,
1215
+ "rl_loss": -0.0143,
1216
+ "step": 1500
1217
+ },
1218
+ {
1219
+ "acc": 0.0,
1220
+ "epoch": 1.01,
1221
+ "learning_rate": 2e-05,
1222
+ "loss": 0.2716,
1223
+ "rl_loss": -0.1058,
1224
+ "step": 1510
1225
+ },
1226
+ {
1227
+ "acc": 0.0,
1228
+ "epoch": 1.01,
1229
+ "learning_rate": 2e-05,
1230
+ "loss": 0.3491,
1231
+ "rl_loss": -0.2102,
1232
+ "step": 1520
1233
+ },
1234
+ {
1235
+ "acc": 0.0,
1236
+ "epoch": 1.02,
1237
+ "learning_rate": 2e-05,
1238
+ "loss": 0.2348,
1239
+ "rl_loss": -0.1394,
1240
+ "step": 1530
1241
+ },
1242
+ {
1243
+ "acc": 0.0,
1244
+ "epoch": 1.03,
1245
+ "learning_rate": 2e-05,
1246
+ "loss": 0.2861,
1247
+ "rl_loss": -0.1298,
1248
+ "step": 1540
1249
+ },
1250
+ {
1251
+ "acc": 0.0,
1252
+ "epoch": 1.03,
1253
+ "learning_rate": 2e-05,
1254
+ "loss": 0.2077,
1255
+ "rl_loss": -0.1497,
1256
+ "step": 1550
1257
+ },
1258
+ {
1259
+ "acc": 0.0,
1260
+ "epoch": 1.04,
1261
+ "learning_rate": 2e-05,
1262
+ "loss": 0.2019,
1263
+ "rl_loss": -0.2238,
1264
+ "step": 1560
1265
+ },
1266
+ {
1267
+ "acc": 0.0,
1268
+ "epoch": 1.05,
1269
+ "learning_rate": 2e-05,
1270
+ "loss": 0.2567,
1271
+ "rl_loss": -0.1693,
1272
+ "step": 1570
1273
+ },
1274
+ {
1275
+ "acc": 0.0,
1276
+ "epoch": 1.05,
1277
+ "learning_rate": 2e-05,
1278
+ "loss": 0.2814,
1279
+ "rl_loss": -0.1605,
1280
+ "step": 1580
1281
+ },
1282
+ {
1283
+ "acc": 0.0,
1284
+ "epoch": 1.06,
1285
+ "learning_rate": 2e-05,
1286
+ "loss": 0.2215,
1287
+ "rl_loss": -0.1134,
1288
+ "step": 1590
1289
+ },
1290
+ {
1291
+ "acc": 0.0,
1292
+ "epoch": 1.07,
1293
+ "learning_rate": 2e-05,
1294
+ "loss": 0.2175,
1295
+ "rl_loss": -0.3902,
1296
+ "step": 1600
1297
+ },
1298
+ {
1299
+ "acc": 0.0,
1300
+ "epoch": 1.07,
1301
+ "learning_rate": 2e-05,
1302
+ "loss": 0.2196,
1303
+ "rl_loss": -0.301,
1304
+ "step": 1610
1305
+ },
1306
+ {
1307
+ "acc": 0.0,
1308
+ "epoch": 1.08,
1309
+ "learning_rate": 2e-05,
1310
+ "loss": 0.292,
1311
+ "rl_loss": -0.213,
1312
+ "step": 1620
1313
+ },
1314
+ {
1315
+ "acc": 0.0,
1316
+ "epoch": 1.09,
1317
+ "learning_rate": 2e-05,
1318
+ "loss": 0.2572,
1319
+ "rl_loss": -0.1575,
1320
+ "step": 1630
1321
+ },
1322
+ {
1323
+ "acc": 0.0,
1324
+ "epoch": 1.09,
1325
+ "learning_rate": 2e-05,
1326
+ "loss": 0.2362,
1327
+ "rl_loss": -0.1837,
1328
+ "step": 1640
1329
+ },
1330
+ {
1331
+ "acc": 0.0,
1332
+ "epoch": 1.1,
1333
+ "learning_rate": 2e-05,
1334
+ "loss": 0.4009,
1335
+ "rl_loss": -0.2471,
1336
+ "step": 1650
1337
+ },
1338
+ {
1339
+ "acc": 0.0,
1340
+ "epoch": 1.11,
1341
+ "learning_rate": 2e-05,
1342
+ "loss": 0.2692,
1343
+ "rl_loss": -0.1602,
1344
+ "step": 1660
1345
+ },
1346
+ {
1347
+ "acc": 0.0,
1348
+ "epoch": 1.11,
1349
+ "learning_rate": 2e-05,
1350
+ "loss": 0.3368,
1351
+ "rl_loss": -0.3032,
1352
+ "step": 1670
1353
+ },
1354
+ {
1355
+ "acc": 0.0,
1356
+ "epoch": 1.12,
1357
+ "learning_rate": 2e-05,
1358
+ "loss": 0.2074,
1359
+ "rl_loss": -0.1879,
1360
+ "step": 1680
1361
+ },
1362
+ {
1363
+ "acc": 0.0,
1364
+ "epoch": 1.13,
1365
+ "learning_rate": 2e-05,
1366
+ "loss": 0.2025,
1367
+ "rl_loss": -0.2248,
1368
+ "step": 1690
1369
+ },
1370
+ {
1371
+ "acc": 0.0,
1372
+ "epoch": 1.13,
1373
+ "learning_rate": 2e-05,
1374
+ "loss": 0.2245,
1375
+ "rl_loss": -0.1585,
1376
+ "step": 1700
1377
+ },
1378
+ {
1379
+ "acc": 0.0,
1380
+ "epoch": 1.14,
1381
+ "learning_rate": 2e-05,
1382
+ "loss": 0.2777,
1383
+ "rl_loss": -0.2912,
1384
+ "step": 1710
1385
+ },
1386
+ {
1387
+ "acc": 0.0,
1388
+ "epoch": 1.15,
1389
+ "learning_rate": 2e-05,
1390
+ "loss": 0.2496,
1391
+ "rl_loss": -0.3379,
1392
+ "step": 1720
1393
+ },
1394
+ {
1395
+ "acc": 0.0,
1396
+ "epoch": 1.15,
1397
+ "learning_rate": 2e-05,
1398
+ "loss": 0.2507,
1399
+ "rl_loss": -0.1577,
1400
+ "step": 1730
1401
+ },
1402
+ {
1403
+ "acc": 0.0,
1404
+ "epoch": 1.16,
1405
+ "learning_rate": 2e-05,
1406
+ "loss": 0.2275,
1407
+ "rl_loss": -0.2425,
1408
+ "step": 1740
1409
+ },
1410
+ {
1411
+ "acc": 0.0,
1412
+ "epoch": 1.17,
1413
+ "learning_rate": 2e-05,
1414
+ "loss": 0.2019,
1415
+ "rl_loss": -0.3212,
1416
+ "step": 1750
1417
+ },
1418
+ {
1419
+ "acc": 0.0,
1420
+ "epoch": 1.17,
1421
+ "learning_rate": 2e-05,
1422
+ "loss": 0.2525,
1423
+ "rl_loss": -0.1072,
1424
+ "step": 1760
1425
+ },
1426
+ {
1427
+ "acc": 0.0,
1428
+ "epoch": 1.18,
1429
+ "learning_rate": 2e-05,
1430
+ "loss": 0.224,
1431
+ "rl_loss": -0.3149,
1432
+ "step": 1770
1433
+ },
1434
+ {
1435
+ "acc": 0.0,
1436
+ "epoch": 1.19,
1437
+ "learning_rate": 2e-05,
1438
+ "loss": 0.3125,
1439
+ "rl_loss": -0.2268,
1440
+ "step": 1780
1441
+ },
1442
+ {
1443
+ "acc": 0.0,
1444
+ "epoch": 1.19,
1445
+ "learning_rate": 2e-05,
1446
+ "loss": 0.2726,
1447
+ "rl_loss": -0.3026,
1448
+ "step": 1790
1449
+ },
1450
+ {
1451
+ "acc": 0.0,
1452
+ "epoch": 1.2,
1453
+ "learning_rate": 2e-05,
1454
+ "loss": 0.2067,
1455
+ "rl_loss": -0.0679,
1456
+ "step": 1800
1457
+ },
1458
+ {
1459
+ "acc": 0.0,
1460
+ "epoch": 1.21,
1461
+ "learning_rate": 2e-05,
1462
+ "loss": 0.2192,
1463
+ "rl_loss": -0.2496,
1464
+ "step": 1810
1465
+ },
1466
+ {
1467
+ "acc": 0.0,
1468
+ "epoch": 1.21,
1469
+ "learning_rate": 2e-05,
1470
+ "loss": 0.2192,
1471
+ "rl_loss": -0.3184,
1472
+ "step": 1820
1473
+ },
1474
+ {
1475
+ "acc": 0.0,
1476
+ "epoch": 1.22,
1477
+ "learning_rate": 2e-05,
1478
+ "loss": 0.2152,
1479
+ "rl_loss": -0.0508,
1480
+ "step": 1830
1481
+ },
1482
+ {
1483
+ "acc": 0.0,
1484
+ "epoch": 1.23,
1485
+ "learning_rate": 2e-05,
1486
+ "loss": 0.3504,
1487
+ "rl_loss": -0.1447,
1488
+ "step": 1840
1489
+ },
1490
+ {
1491
+ "acc": 0.0,
1492
+ "epoch": 1.23,
1493
+ "learning_rate": 2e-05,
1494
+ "loss": 0.2511,
1495
+ "rl_loss": -0.1003,
1496
+ "step": 1850
1497
+ },
1498
+ {
1499
+ "acc": 0.0,
1500
+ "epoch": 1.24,
1501
+ "learning_rate": 2e-05,
1502
+ "loss": 0.2549,
1503
+ "rl_loss": -0.2655,
1504
+ "step": 1860
1505
+ },
1506
+ {
1507
+ "acc": 0.0,
1508
+ "epoch": 1.25,
1509
+ "learning_rate": 2e-05,
1510
+ "loss": 0.2788,
1511
+ "rl_loss": -0.1448,
1512
+ "step": 1870
1513
+ },
1514
+ {
1515
+ "acc": 0.0,
1516
+ "epoch": 1.25,
1517
+ "learning_rate": 2e-05,
1518
+ "loss": 0.2365,
1519
+ "rl_loss": -0.3471,
1520
+ "step": 1880
1521
+ },
1522
+ {
1523
+ "acc": 0.0,
1524
+ "epoch": 1.26,
1525
+ "learning_rate": 2e-05,
1526
+ "loss": 0.2466,
1527
+ "rl_loss": -0.2494,
1528
+ "step": 1890
1529
+ },
1530
+ {
1531
+ "acc": 0.0,
1532
+ "epoch": 1.27,
1533
+ "learning_rate": 2e-05,
1534
+ "loss": 0.2206,
1535
+ "rl_loss": -0.4574,
1536
+ "step": 1900
1537
+ },
1538
+ {
1539
+ "acc": 0.0,
1540
+ "epoch": 1.27,
1541
+ "learning_rate": 2e-05,
1542
+ "loss": 0.3148,
1543
+ "rl_loss": -0.1643,
1544
+ "step": 1910
1545
+ },
1546
+ {
1547
+ "acc": 0.0,
1548
+ "epoch": 1.28,
1549
+ "learning_rate": 2e-05,
1550
+ "loss": 0.262,
1551
+ "rl_loss": -0.2892,
1552
+ "step": 1920
1553
+ },
1554
+ {
1555
+ "acc": 0.0,
1556
+ "epoch": 1.29,
1557
+ "learning_rate": 2e-05,
1558
+ "loss": 0.339,
1559
+ "rl_loss": -0.0317,
1560
+ "step": 1930
1561
+ },
1562
+ {
1563
+ "acc": 0.0,
1564
+ "epoch": 1.29,
1565
+ "learning_rate": 2e-05,
1566
+ "loss": 0.2181,
1567
+ "rl_loss": -0.2939,
1568
+ "step": 1940
1569
+ },
1570
+ {
1571
+ "acc": 0.0,
1572
+ "epoch": 1.3,
1573
+ "learning_rate": 2e-05,
1574
+ "loss": 0.3218,
1575
+ "rl_loss": -0.1217,
1576
+ "step": 1950
1577
+ },
1578
+ {
1579
+ "acc": 0.0,
1580
+ "epoch": 1.31,
1581
+ "learning_rate": 2e-05,
1582
+ "loss": 0.2099,
1583
+ "rl_loss": -0.4231,
1584
+ "step": 1960
1585
+ },
1586
+ {
1587
+ "acc": 0.0,
1588
+ "epoch": 1.31,
1589
+ "learning_rate": 2e-05,
1590
+ "loss": 0.3301,
1591
+ "rl_loss": -0.1967,
1592
+ "step": 1970
1593
+ },
1594
+ {
1595
+ "acc": 0.0,
1596
+ "epoch": 1.32,
1597
+ "learning_rate": 2e-05,
1598
+ "loss": 0.3348,
1599
+ "rl_loss": -0.2999,
1600
+ "step": 1980
1601
+ },
1602
+ {
1603
+ "acc": 0.0,
1604
+ "epoch": 1.33,
1605
+ "learning_rate": 2e-05,
1606
+ "loss": 0.2197,
1607
+ "rl_loss": -0.2697,
1608
+ "step": 1990
1609
+ },
1610
+ {
1611
+ "acc": 0.0,
1612
+ "epoch": 1.33,
1613
+ "learning_rate": 2e-05,
1614
+ "loss": 0.2636,
1615
+ "rl_loss": -0.1534,
1616
+ "step": 2000
1617
+ },
1618
+ {
1619
+ "acc": 0.0,
1620
+ "epoch": 1.34,
1621
+ "learning_rate": 2e-05,
1622
+ "loss": 0.3831,
1623
+ "rl_loss": -0.3402,
1624
+ "step": 2010
1625
+ },
1626
+ {
1627
+ "acc": 0.0,
1628
+ "epoch": 1.35,
1629
+ "learning_rate": 2e-05,
1630
+ "loss": 0.2709,
1631
+ "rl_loss": -0.2815,
1632
+ "step": 2020
1633
+ },
1634
+ {
1635
+ "acc": 0.0,
1636
+ "epoch": 1.35,
1637
+ "learning_rate": 2e-05,
1638
+ "loss": 0.4735,
1639
+ "rl_loss": -0.1193,
1640
+ "step": 2030
1641
+ },
1642
+ {
1643
+ "acc": 0.0,
1644
+ "epoch": 1.36,
1645
+ "learning_rate": 2e-05,
1646
+ "loss": 0.3152,
1647
+ "rl_loss": -0.3396,
1648
+ "step": 2040
1649
+ },
1650
+ {
1651
+ "acc": 0.0,
1652
+ "epoch": 1.37,
1653
+ "learning_rate": 2e-05,
1654
+ "loss": 0.213,
1655
+ "rl_loss": -0.1602,
1656
+ "step": 2050
1657
+ },
1658
+ {
1659
+ "acc": 0.0,
1660
+ "epoch": 1.37,
1661
+ "learning_rate": 2e-05,
1662
+ "loss": 0.2118,
1663
+ "rl_loss": -0.2494,
1664
+ "step": 2060
1665
+ },
1666
+ {
1667
+ "acc": 0.0,
1668
+ "epoch": 1.38,
1669
+ "learning_rate": 2e-05,
1670
+ "loss": 0.2877,
1671
+ "rl_loss": -0.2309,
1672
+ "step": 2070
1673
+ },
1674
+ {
1675
+ "acc": 0.0,
1676
+ "epoch": 1.39,
1677
+ "learning_rate": 2e-05,
1678
+ "loss": 0.2999,
1679
+ "rl_loss": -0.2919,
1680
+ "step": 2080
1681
+ },
1682
+ {
1683
+ "acc": 0.0,
1684
+ "epoch": 1.39,
1685
+ "learning_rate": 2e-05,
1686
+ "loss": 0.3178,
1687
+ "rl_loss": -0.3203,
1688
+ "step": 2090
1689
+ },
1690
+ {
1691
+ "acc": 0.0,
1692
+ "epoch": 1.4,
1693
+ "learning_rate": 2e-05,
1694
+ "loss": 0.2346,
1695
+ "rl_loss": -0.3028,
1696
+ "step": 2100
1697
+ },
1698
+ {
1699
+ "acc": 0.0,
1700
+ "epoch": 1.41,
1701
+ "learning_rate": 2e-05,
1702
+ "loss": 0.2955,
1703
+ "rl_loss": -0.2407,
1704
+ "step": 2110
1705
+ },
1706
+ {
1707
+ "acc": 0.0,
1708
+ "epoch": 1.41,
1709
+ "learning_rate": 2e-05,
1710
+ "loss": 0.3054,
1711
+ "rl_loss": -0.3303,
1712
+ "step": 2120
1713
+ },
1714
+ {
1715
+ "acc": 0.0,
1716
+ "epoch": 1.42,
1717
+ "learning_rate": 2e-05,
1718
+ "loss": 0.221,
1719
+ "rl_loss": -0.0344,
1720
+ "step": 2130
1721
+ },
1722
+ {
1723
+ "acc": 0.0,
1724
+ "epoch": 1.43,
1725
+ "learning_rate": 2e-05,
1726
+ "loss": 0.2423,
1727
+ "rl_loss": -0.4771,
1728
+ "step": 2140
1729
+ },
1730
+ {
1731
+ "acc": 0.0,
1732
+ "epoch": 1.43,
1733
+ "learning_rate": 2e-05,
1734
+ "loss": 0.3343,
1735
+ "rl_loss": -0.7419,
1736
+ "step": 2150
1737
+ },
1738
+ {
1739
+ "acc": 0.0,
1740
+ "epoch": 1.44,
1741
+ "learning_rate": 2e-05,
1742
+ "loss": 0.2997,
1743
+ "rl_loss": -0.183,
1744
+ "step": 2160
1745
+ },
1746
+ {
1747
+ "acc": 0.0,
1748
+ "epoch": 1.45,
1749
+ "learning_rate": 2e-05,
1750
+ "loss": 0.2399,
1751
+ "rl_loss": -0.2342,
1752
+ "step": 2170
1753
+ },
1754
+ {
1755
+ "acc": 0.0,
1756
+ "epoch": 1.45,
1757
+ "learning_rate": 2e-05,
1758
+ "loss": 0.2391,
1759
+ "rl_loss": -0.3045,
1760
+ "step": 2180
1761
+ },
1762
+ {
1763
+ "acc": 0.0,
1764
+ "epoch": 1.46,
1765
+ "learning_rate": 2e-05,
1766
+ "loss": 0.4321,
1767
+ "rl_loss": -0.3114,
1768
+ "step": 2190
1769
+ },
1770
+ {
1771
+ "acc": 0.0,
1772
+ "epoch": 1.47,
1773
+ "learning_rate": 2e-05,
1774
+ "loss": 0.2387,
1775
+ "rl_loss": -0.2203,
1776
+ "step": 2200
1777
+ },
1778
+ {
1779
+ "acc": 0.0,
1780
+ "epoch": 1.47,
1781
+ "learning_rate": 2e-05,
1782
+ "loss": 0.2195,
1783
+ "rl_loss": -0.1261,
1784
+ "step": 2210
1785
+ },
1786
+ {
1787
+ "acc": 0.0,
1788
+ "epoch": 1.48,
1789
+ "learning_rate": 2e-05,
1790
+ "loss": 0.2744,
1791
+ "rl_loss": -0.2352,
1792
+ "step": 2220
1793
+ },
1794
+ {
1795
+ "acc": 0.0,
1796
+ "epoch": 1.49,
1797
+ "learning_rate": 2e-05,
1798
+ "loss": 0.56,
1799
+ "rl_loss": -0.4791,
1800
+ "step": 2230
1801
+ },
1802
+ {
1803
+ "acc": 0.0,
1804
+ "epoch": 1.49,
1805
+ "learning_rate": 2e-05,
1806
+ "loss": 0.2127,
1807
+ "rl_loss": -0.0956,
1808
+ "step": 2240
1809
+ },
1810
+ {
1811
+ "acc": 0.0,
1812
+ "epoch": 1.5,
1813
+ "learning_rate": 2e-05,
1814
+ "loss": 0.3765,
1815
+ "rl_loss": -0.1513,
1816
+ "step": 2250
1817
+ },
1818
+ {
1819
+ "acc": 0.0,
1820
+ "epoch": 1.51,
1821
+ "learning_rate": 2e-05,
1822
+ "loss": 0.2626,
1823
+ "rl_loss": -0.0945,
1824
+ "step": 2260
1825
+ },
1826
+ {
1827
+ "acc": 0.0,
1828
+ "epoch": 1.51,
1829
+ "learning_rate": 2e-05,
1830
+ "loss": 0.2584,
1831
+ "rl_loss": -0.2644,
1832
+ "step": 2270
1833
+ },
1834
+ {
1835
+ "acc": 0.0,
1836
+ "epoch": 1.52,
1837
+ "learning_rate": 2e-05,
1838
+ "loss": 0.3124,
1839
+ "rl_loss": -0.1856,
1840
+ "step": 2280
1841
+ },
1842
+ {
1843
+ "acc": 0.0,
1844
+ "epoch": 1.53,
1845
+ "learning_rate": 2e-05,
1846
+ "loss": 0.3314,
1847
+ "rl_loss": -0.1497,
1848
+ "step": 2290
1849
+ },
1850
+ {
1851
+ "acc": 0.0,
1852
+ "epoch": 1.53,
1853
+ "learning_rate": 2e-05,
1854
+ "loss": 0.222,
1855
+ "rl_loss": -0.2127,
1856
+ "step": 2300
1857
+ },
1858
+ {
1859
+ "acc": 0.0,
1860
+ "epoch": 1.54,
1861
+ "learning_rate": 2e-05,
1862
+ "loss": 0.2053,
1863
+ "rl_loss": -0.2785,
1864
+ "step": 2310
1865
+ },
1866
+ {
1867
+ "acc": 0.0,
1868
+ "epoch": 1.55,
1869
+ "learning_rate": 2e-05,
1870
+ "loss": 0.2848,
1871
+ "rl_loss": -0.1348,
1872
+ "step": 2320
1873
+ },
1874
+ {
1875
+ "acc": 0.0,
1876
+ "epoch": 1.55,
1877
+ "learning_rate": 2e-05,
1878
+ "loss": 0.2008,
1879
+ "rl_loss": -0.3603,
1880
+ "step": 2330
1881
+ },
1882
+ {
1883
+ "acc": 0.0,
1884
+ "epoch": 1.56,
1885
+ "learning_rate": 2e-05,
1886
+ "loss": 0.334,
1887
+ "rl_loss": -0.1207,
1888
+ "step": 2340
1889
+ },
1890
+ {
1891
+ "acc": 0.0,
1892
+ "epoch": 1.57,
1893
+ "learning_rate": 2e-05,
1894
+ "loss": 0.2366,
1895
+ "rl_loss": -0.556,
1896
+ "step": 2350
1897
+ },
1898
+ {
1899
+ "acc": 0.0,
1900
+ "epoch": 1.57,
1901
+ "learning_rate": 2e-05,
1902
+ "loss": 0.3089,
1903
+ "rl_loss": -0.1413,
1904
+ "step": 2360
1905
+ },
1906
+ {
1907
+ "acc": 0.0,
1908
+ "epoch": 1.58,
1909
+ "learning_rate": 2e-05,
1910
+ "loss": 0.2942,
1911
+ "rl_loss": -0.1651,
1912
+ "step": 2370
1913
+ },
1914
+ {
1915
+ "acc": 0.0,
1916
+ "epoch": 1.59,
1917
+ "learning_rate": 2e-05,
1918
+ "loss": 0.2984,
1919
+ "rl_loss": -0.3665,
1920
+ "step": 2380
1921
+ },
1922
+ {
1923
+ "acc": 0.0,
1924
+ "epoch": 1.59,
1925
+ "learning_rate": 2e-05,
1926
+ "loss": 0.2923,
1927
+ "rl_loss": -0.1287,
1928
+ "step": 2390
1929
+ },
1930
+ {
1931
+ "acc": 0.0,
1932
+ "epoch": 1.6,
1933
+ "learning_rate": 2e-05,
1934
+ "loss": 0.2936,
1935
+ "rl_loss": -0.2642,
1936
+ "step": 2400
1937
+ },
1938
+ {
1939
+ "acc": 0.0,
1940
+ "epoch": 1.61,
1941
+ "learning_rate": 2e-05,
1942
+ "loss": 0.2817,
1943
+ "rl_loss": -0.0497,
1944
+ "step": 2410
1945
+ },
1946
+ {
1947
+ "acc": 0.0,
1948
+ "epoch": 1.61,
1949
+ "learning_rate": 2e-05,
1950
+ "loss": 0.2176,
1951
+ "rl_loss": -0.1427,
1952
+ "step": 2420
1953
+ },
1954
+ {
1955
+ "acc": 0.0,
1956
+ "epoch": 1.62,
1957
+ "learning_rate": 2e-05,
1958
+ "loss": 0.2181,
1959
+ "rl_loss": -0.2153,
1960
+ "step": 2430
1961
+ },
1962
+ {
1963
+ "acc": 0.0,
1964
+ "epoch": 1.63,
1965
+ "learning_rate": 2e-05,
1966
+ "loss": 0.353,
1967
+ "rl_loss": -0.0923,
1968
+ "step": 2440
1969
+ },
1970
+ {
1971
+ "acc": 0.0,
1972
+ "epoch": 1.63,
1973
+ "learning_rate": 2e-05,
1974
+ "loss": 0.2725,
1975
+ "rl_loss": -0.1337,
1976
+ "step": 2450
1977
+ },
1978
+ {
1979
+ "acc": 0.0,
1980
+ "epoch": 1.64,
1981
+ "learning_rate": 2e-05,
1982
+ "loss": 0.3012,
1983
+ "rl_loss": -0.2073,
1984
+ "step": 2460
1985
+ },
1986
+ {
1987
+ "acc": 0.0,
1988
+ "epoch": 1.65,
1989
+ "learning_rate": 2e-05,
1990
+ "loss": 0.4572,
1991
+ "rl_loss": -0.2452,
1992
+ "step": 2470
1993
+ },
1994
+ {
1995
+ "acc": 0.0,
1996
+ "epoch": 1.65,
1997
+ "learning_rate": 2e-05,
1998
+ "loss": 0.3193,
1999
+ "rl_loss": -0.322,
2000
+ "step": 2480
2001
+ },
2002
+ {
2003
+ "acc": 0.0,
2004
+ "epoch": 1.66,
2005
+ "learning_rate": 2e-05,
2006
+ "loss": 0.2718,
2007
+ "rl_loss": -0.1434,
2008
+ "step": 2490
2009
+ },
2010
+ {
2011
+ "acc": 0.0,
2012
+ "epoch": 1.67,
2013
+ "learning_rate": 2e-05,
2014
+ "loss": 0.2396,
2015
+ "rl_loss": -0.3019,
2016
+ "step": 2500
2017
+ },
2018
+ {
2019
+ "acc": 0.0,
2020
+ "epoch": 1.67,
2021
+ "learning_rate": 2e-05,
2022
+ "loss": 0.2117,
2023
+ "rl_loss": -0.1138,
2024
+ "step": 2510
2025
+ },
2026
+ {
2027
+ "acc": 0.0,
2028
+ "epoch": 1.68,
2029
+ "learning_rate": 2e-05,
2030
+ "loss": 0.305,
2031
+ "rl_loss": -0.1993,
2032
+ "step": 2520
2033
+ },
2034
+ {
2035
+ "acc": 0.0,
2036
+ "epoch": 1.69,
2037
+ "learning_rate": 2e-05,
2038
+ "loss": 0.2979,
2039
+ "rl_loss": -0.1591,
2040
+ "step": 2530
2041
+ },
2042
+ {
2043
+ "acc": 0.0,
2044
+ "epoch": 1.69,
2045
+ "learning_rate": 2e-05,
2046
+ "loss": 0.2312,
2047
+ "rl_loss": -0.0862,
2048
+ "step": 2540
2049
+ },
2050
+ {
2051
+ "acc": 0.0,
2052
+ "epoch": 1.7,
2053
+ "learning_rate": 2e-05,
2054
+ "loss": 0.2726,
2055
+ "rl_loss": -0.3785,
2056
+ "step": 2550
2057
+ },
2058
+ {
2059
+ "acc": 0.0,
2060
+ "epoch": 1.71,
2061
+ "learning_rate": 2e-05,
2062
+ "loss": 0.3373,
2063
+ "rl_loss": -0.0622,
2064
+ "step": 2560
2065
+ },
2066
+ {
2067
+ "acc": 0.0,
2068
+ "epoch": 1.71,
2069
+ "learning_rate": 2e-05,
2070
+ "loss": 0.2486,
2071
+ "rl_loss": -0.0971,
2072
+ "step": 2570
2073
+ },
2074
+ {
2075
+ "acc": 0.0,
2076
+ "epoch": 1.72,
2077
+ "learning_rate": 2e-05,
2078
+ "loss": 0.2828,
2079
+ "rl_loss": -0.1688,
2080
+ "step": 2580
2081
+ },
2082
+ {
2083
+ "acc": 0.0,
2084
+ "epoch": 1.73,
2085
+ "learning_rate": 2e-05,
2086
+ "loss": 0.2374,
2087
+ "rl_loss": -0.1622,
2088
+ "step": 2590
2089
+ },
2090
+ {
2091
+ "acc": 0.0,
2092
+ "epoch": 1.73,
2093
+ "learning_rate": 2e-05,
2094
+ "loss": 0.428,
2095
+ "rl_loss": -0.2429,
2096
+ "step": 2600
2097
+ },
2098
+ {
2099
+ "acc": 0.0,
2100
+ "epoch": 1.74,
2101
+ "learning_rate": 2e-05,
2102
+ "loss": 0.265,
2103
+ "rl_loss": -0.314,
2104
+ "step": 2610
2105
+ },
2106
+ {
2107
+ "acc": 0.0,
2108
+ "epoch": 1.75,
2109
+ "learning_rate": 2e-05,
2110
+ "loss": 0.3008,
2111
+ "rl_loss": -0.2175,
2112
+ "step": 2620
2113
+ },
2114
+ {
2115
+ "acc": 0.0,
2116
+ "epoch": 1.75,
2117
+ "learning_rate": 2e-05,
2118
+ "loss": 0.212,
2119
+ "rl_loss": -0.2465,
2120
+ "step": 2630
2121
+ },
2122
+ {
2123
+ "acc": 0.0,
2124
+ "epoch": 1.76,
2125
+ "learning_rate": 2e-05,
2126
+ "loss": 0.2471,
2127
+ "rl_loss": -0.0837,
2128
+ "step": 2640
2129
+ },
2130
+ {
2131
+ "acc": 0.0,
2132
+ "epoch": 1.77,
2133
+ "learning_rate": 2e-05,
2134
+ "loss": 0.2352,
2135
+ "rl_loss": -0.0888,
2136
+ "step": 2650
2137
+ },
2138
+ {
2139
+ "acc": 0.0,
2140
+ "epoch": 1.77,
2141
+ "learning_rate": 2e-05,
2142
+ "loss": 0.4229,
2143
+ "rl_loss": -0.3808,
2144
+ "step": 2660
2145
+ },
2146
+ {
2147
+ "acc": 0.0,
2148
+ "epoch": 1.78,
2149
+ "learning_rate": 2e-05,
2150
+ "loss": 0.2738,
2151
+ "rl_loss": -0.1199,
2152
+ "step": 2670
2153
+ },
2154
+ {
2155
+ "acc": 0.0,
2156
+ "epoch": 1.79,
2157
+ "learning_rate": 2e-05,
2158
+ "loss": 0.2162,
2159
+ "rl_loss": -0.0656,
2160
+ "step": 2680
2161
+ },
2162
+ {
2163
+ "acc": 0.0,
2164
+ "epoch": 1.79,
2165
+ "learning_rate": 2e-05,
2166
+ "loss": 0.246,
2167
+ "rl_loss": -0.2123,
2168
+ "step": 2690
2169
+ },
2170
+ {
2171
+ "acc": 0.0,
2172
+ "epoch": 1.8,
2173
+ "learning_rate": 2e-05,
2174
+ "loss": 0.2648,
2175
+ "rl_loss": -0.2322,
2176
+ "step": 2700
2177
+ },
2178
+ {
2179
+ "acc": 0.0,
2180
+ "epoch": 1.81,
2181
+ "learning_rate": 2e-05,
2182
+ "loss": 0.2433,
2183
+ "rl_loss": -0.0837,
2184
+ "step": 2710
2185
+ },
2186
+ {
2187
+ "acc": 0.0,
2188
+ "epoch": 1.81,
2189
+ "learning_rate": 2e-05,
2190
+ "loss": 0.269,
2191
+ "rl_loss": -0.0561,
2192
+ "step": 2720
2193
+ },
2194
+ {
2195
+ "acc": 0.0,
2196
+ "epoch": 1.82,
2197
+ "learning_rate": 2e-05,
2198
+ "loss": 0.2715,
2199
+ "rl_loss": -0.3806,
2200
+ "step": 2730
2201
+ },
2202
+ {
2203
+ "acc": 0.0,
2204
+ "epoch": 1.83,
2205
+ "learning_rate": 2e-05,
2206
+ "loss": 0.3139,
2207
+ "rl_loss": -0.0817,
2208
+ "step": 2740
2209
+ },
2210
+ {
2211
+ "acc": 0.0,
2212
+ "epoch": 1.83,
2213
+ "learning_rate": 2e-05,
2214
+ "loss": 0.2513,
2215
+ "rl_loss": -0.2584,
2216
+ "step": 2750
2217
+ },
2218
+ {
2219
+ "acc": 0.0,
2220
+ "epoch": 1.84,
2221
+ "learning_rate": 2e-05,
2222
+ "loss": 0.3049,
2223
+ "rl_loss": -0.2728,
2224
+ "step": 2760
2225
+ },
2226
+ {
2227
+ "acc": 0.0,
2228
+ "epoch": 1.85,
2229
+ "learning_rate": 2e-05,
2230
+ "loss": 0.228,
2231
+ "rl_loss": -0.165,
2232
+ "step": 2770
2233
+ },
2234
+ {
2235
+ "acc": 0.0,
2236
+ "epoch": 1.85,
2237
+ "learning_rate": 2e-05,
2238
+ "loss": 0.2464,
2239
+ "rl_loss": -0.2593,
2240
+ "step": 2780
2241
+ },
2242
+ {
2243
+ "acc": 0.0,
2244
+ "epoch": 1.86,
2245
+ "learning_rate": 2e-05,
2246
+ "loss": 0.3339,
2247
+ "rl_loss": -0.1801,
2248
+ "step": 2790
2249
+ },
2250
+ {
2251
+ "acc": 0.0,
2252
+ "epoch": 1.87,
2253
+ "learning_rate": 2e-05,
2254
+ "loss": 0.2498,
2255
+ "rl_loss": -0.0794,
2256
+ "step": 2800
2257
+ },
2258
+ {
2259
+ "acc": 0.0,
2260
+ "epoch": 1.87,
2261
+ "learning_rate": 2e-05,
2262
+ "loss": 0.2176,
2263
+ "rl_loss": -0.2592,
2264
+ "step": 2810
2265
+ },
2266
+ {
2267
+ "acc": 0.0,
2268
+ "epoch": 1.88,
2269
+ "learning_rate": 2e-05,
2270
+ "loss": 0.2339,
2271
+ "rl_loss": -0.1273,
2272
+ "step": 2820
2273
+ },
2274
+ {
2275
+ "acc": 0.0,
2276
+ "epoch": 1.89,
2277
+ "learning_rate": 2e-05,
2278
+ "loss": 0.3225,
2279
+ "rl_loss": -0.2312,
2280
+ "step": 2830
2281
+ },
2282
+ {
2283
+ "acc": 0.0,
2284
+ "epoch": 1.89,
2285
+ "learning_rate": 2e-05,
2286
+ "loss": 0.3005,
2287
+ "rl_loss": -0.2849,
2288
+ "step": 2840
2289
+ },
2290
+ {
2291
+ "acc": 0.0,
2292
+ "epoch": 1.9,
2293
+ "learning_rate": 2e-05,
2294
+ "loss": 0.3244,
2295
+ "rl_loss": -0.4541,
2296
+ "step": 2850
2297
+ },
2298
+ {
2299
+ "acc": 0.0,
2300
+ "epoch": 1.91,
2301
+ "learning_rate": 2e-05,
2302
+ "loss": 0.2361,
2303
+ "rl_loss": -0.4858,
2304
+ "step": 2860
2305
+ },
2306
+ {
2307
+ "acc": 0.0,
2308
+ "epoch": 1.91,
2309
+ "learning_rate": 2e-05,
2310
+ "loss": 0.2548,
2311
+ "rl_loss": -0.0967,
2312
+ "step": 2870
2313
+ },
2314
+ {
2315
+ "acc": 0.0,
2316
+ "epoch": 1.92,
2317
+ "learning_rate": 2e-05,
2318
+ "loss": 0.2556,
2319
+ "rl_loss": -0.4954,
2320
+ "step": 2880
2321
+ },
2322
+ {
2323
+ "acc": 0.0,
2324
+ "epoch": 1.93,
2325
+ "learning_rate": 2e-05,
2326
+ "loss": 0.2572,
2327
+ "rl_loss": -0.19,
2328
+ "step": 2890
2329
+ },
2330
+ {
2331
+ "acc": 0.0,
2332
+ "epoch": 1.93,
2333
+ "learning_rate": 2e-05,
2334
+ "loss": 0.3286,
2335
+ "rl_loss": -0.2709,
2336
+ "step": 2900
2337
+ },
2338
+ {
2339
+ "acc": 0.0,
2340
+ "epoch": 1.94,
2341
+ "learning_rate": 2e-05,
2342
+ "loss": 0.2768,
2343
+ "rl_loss": -0.1072,
2344
+ "step": 2910
2345
+ },
2346
+ {
2347
+ "acc": 0.0,
2348
+ "epoch": 1.95,
2349
+ "learning_rate": 2e-05,
2350
+ "loss": 0.2248,
2351
+ "rl_loss": -0.4267,
2352
+ "step": 2920
2353
+ },
2354
+ {
2355
+ "acc": 0.0,
2356
+ "epoch": 1.95,
2357
+ "learning_rate": 2e-05,
2358
+ "loss": 0.2976,
2359
+ "rl_loss": -0.3855,
2360
+ "step": 2930
2361
+ },
2362
+ {
2363
+ "acc": 0.0,
2364
+ "epoch": 1.96,
2365
+ "learning_rate": 2e-05,
2366
+ "loss": 0.3344,
2367
+ "rl_loss": -0.1849,
2368
+ "step": 2940
2369
+ },
2370
+ {
2371
+ "acc": 0.0,
2372
+ "epoch": 1.97,
2373
+ "learning_rate": 2e-05,
2374
+ "loss": 0.2909,
2375
+ "rl_loss": -0.4777,
2376
+ "step": 2950
2377
+ },
2378
+ {
2379
+ "acc": 0.0,
2380
+ "epoch": 1.97,
2381
+ "learning_rate": 2e-05,
2382
+ "loss": 0.368,
2383
+ "rl_loss": -0.5148,
2384
+ "step": 2960
2385
+ },
2386
+ {
2387
+ "acc": 0.0,
2388
+ "epoch": 1.98,
2389
+ "learning_rate": 2e-05,
2390
+ "loss": 0.2897,
2391
+ "rl_loss": -0.4314,
2392
+ "step": 2970
2393
+ },
2394
+ {
2395
+ "acc": 0.0,
2396
+ "epoch": 1.99,
2397
+ "learning_rate": 2e-05,
2398
+ "loss": 0.3902,
2399
+ "rl_loss": -0.3644,
2400
+ "step": 2980
2401
+ },
2402
+ {
2403
+ "acc": 0.0,
2404
+ "epoch": 1.99,
2405
+ "learning_rate": 2e-05,
2406
+ "loss": 0.2532,
2407
+ "rl_loss": -0.5152,
2408
+ "step": 2990
2409
+ },
2410
+ {
2411
+ "acc": 0.0,
2412
+ "epoch": 2.0,
2413
+ "learning_rate": 2e-05,
2414
+ "loss": 0.2424,
2415
+ "rl_loss": -0.534,
2416
+ "step": 3000
2417
+ }
2418
+ ],
2419
+ "max_steps": 14990,
2420
+ "num_train_epochs": 10,
2421
+ "total_flos": 4.873690607616e+17,
2422
+ "trial_name": null,
2423
+ "trial_params": null
2424
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:077abbba99d6ff961ac045478ca5d83afc7b06fad05985dc40caba50146f2610
3
+ size 3512