add model files
Browse files- config.json +34 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +151 -0
- tokenizer_config.json +160 -0
- trainer_state.json +1393 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "models/bygpt5-medium/en",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ByGPT5LMHeadModel"
|
| 5 |
+
],
|
| 6 |
+
"d_ff": 3840,
|
| 7 |
+
"d_kv": 64,
|
| 8 |
+
"d_model": 1536,
|
| 9 |
+
"decoder_start_token_id": 0,
|
| 10 |
+
"dense_act_fn": "gelu_new",
|
| 11 |
+
"dropout_rate": 0.1,
|
| 12 |
+
"eos_token_id": 1,
|
| 13 |
+
"feed_forward_proj": "gated-gelu",
|
| 14 |
+
"gradient_checkpointing": false,
|
| 15 |
+
"initializer_factor": 1.0,
|
| 16 |
+
"is_decoder": true,
|
| 17 |
+
"is_encoder_decoder": false,
|
| 18 |
+
"is_gated_act": true,
|
| 19 |
+
"layer_norm_epsilon": 1e-06,
|
| 20 |
+
"model_type": "bygpt5",
|
| 21 |
+
"num_decoder_layers": 12,
|
| 22 |
+
"num_heads": 16,
|
| 23 |
+
"num_layers": 12,
|
| 24 |
+
"output_past": true,
|
| 25 |
+
"pad_token_id": 0,
|
| 26 |
+
"relative_attention_max_distance": 128,
|
| 27 |
+
"relative_attention_num_buckets": 32,
|
| 28 |
+
"tie_word_embeddings": false,
|
| 29 |
+
"tokenizer_class": "ByT5Tokenizer",
|
| 30 |
+
"torch_dtype": "float32",
|
| 31 |
+
"transformers_version": "4.20.0",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 384
|
| 34 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2685152c53b78c7cf45b601a4d8185e5eeb287f0ab738d05afa9f39affce2f7
|
| 3 |
+
size 1156247841
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,151 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<extra_id_0>",
|
| 4 |
+
"<extra_id_1>",
|
| 5 |
+
"<extra_id_2>",
|
| 6 |
+
"<extra_id_3>",
|
| 7 |
+
"<extra_id_4>",
|
| 8 |
+
"<extra_id_5>",
|
| 9 |
+
"<extra_id_6>",
|
| 10 |
+
"<extra_id_7>",
|
| 11 |
+
"<extra_id_8>",
|
| 12 |
+
"<extra_id_9>",
|
| 13 |
+
"<extra_id_10>",
|
| 14 |
+
"<extra_id_11>",
|
| 15 |
+
"<extra_id_12>",
|
| 16 |
+
"<extra_id_13>",
|
| 17 |
+
"<extra_id_14>",
|
| 18 |
+
"<extra_id_15>",
|
| 19 |
+
"<extra_id_16>",
|
| 20 |
+
"<extra_id_17>",
|
| 21 |
+
"<extra_id_18>",
|
| 22 |
+
"<extra_id_19>",
|
| 23 |
+
"<extra_id_20>",
|
| 24 |
+
"<extra_id_21>",
|
| 25 |
+
"<extra_id_22>",
|
| 26 |
+
"<extra_id_23>",
|
| 27 |
+
"<extra_id_24>",
|
| 28 |
+
"<extra_id_25>",
|
| 29 |
+
"<extra_id_26>",
|
| 30 |
+
"<extra_id_27>",
|
| 31 |
+
"<extra_id_28>",
|
| 32 |
+
"<extra_id_29>",
|
| 33 |
+
"<extra_id_30>",
|
| 34 |
+
"<extra_id_31>",
|
| 35 |
+
"<extra_id_32>",
|
| 36 |
+
"<extra_id_33>",
|
| 37 |
+
"<extra_id_34>",
|
| 38 |
+
"<extra_id_35>",
|
| 39 |
+
"<extra_id_36>",
|
| 40 |
+
"<extra_id_37>",
|
| 41 |
+
"<extra_id_38>",
|
| 42 |
+
"<extra_id_39>",
|
| 43 |
+
"<extra_id_40>",
|
| 44 |
+
"<extra_id_41>",
|
| 45 |
+
"<extra_id_42>",
|
| 46 |
+
"<extra_id_43>",
|
| 47 |
+
"<extra_id_44>",
|
| 48 |
+
"<extra_id_45>",
|
| 49 |
+
"<extra_id_46>",
|
| 50 |
+
"<extra_id_47>",
|
| 51 |
+
"<extra_id_48>",
|
| 52 |
+
"<extra_id_49>",
|
| 53 |
+
"<extra_id_50>",
|
| 54 |
+
"<extra_id_51>",
|
| 55 |
+
"<extra_id_52>",
|
| 56 |
+
"<extra_id_53>",
|
| 57 |
+
"<extra_id_54>",
|
| 58 |
+
"<extra_id_55>",
|
| 59 |
+
"<extra_id_56>",
|
| 60 |
+
"<extra_id_57>",
|
| 61 |
+
"<extra_id_58>",
|
| 62 |
+
"<extra_id_59>",
|
| 63 |
+
"<extra_id_60>",
|
| 64 |
+
"<extra_id_61>",
|
| 65 |
+
"<extra_id_62>",
|
| 66 |
+
"<extra_id_63>",
|
| 67 |
+
"<extra_id_64>",
|
| 68 |
+
"<extra_id_65>",
|
| 69 |
+
"<extra_id_66>",
|
| 70 |
+
"<extra_id_67>",
|
| 71 |
+
"<extra_id_68>",
|
| 72 |
+
"<extra_id_69>",
|
| 73 |
+
"<extra_id_70>",
|
| 74 |
+
"<extra_id_71>",
|
| 75 |
+
"<extra_id_72>",
|
| 76 |
+
"<extra_id_73>",
|
| 77 |
+
"<extra_id_74>",
|
| 78 |
+
"<extra_id_75>",
|
| 79 |
+
"<extra_id_76>",
|
| 80 |
+
"<extra_id_77>",
|
| 81 |
+
"<extra_id_78>",
|
| 82 |
+
"<extra_id_79>",
|
| 83 |
+
"<extra_id_80>",
|
| 84 |
+
"<extra_id_81>",
|
| 85 |
+
"<extra_id_82>",
|
| 86 |
+
"<extra_id_83>",
|
| 87 |
+
"<extra_id_84>",
|
| 88 |
+
"<extra_id_85>",
|
| 89 |
+
"<extra_id_86>",
|
| 90 |
+
"<extra_id_87>",
|
| 91 |
+
"<extra_id_88>",
|
| 92 |
+
"<extra_id_89>",
|
| 93 |
+
"<extra_id_90>",
|
| 94 |
+
"<extra_id_91>",
|
| 95 |
+
"<extra_id_92>",
|
| 96 |
+
"<extra_id_93>",
|
| 97 |
+
"<extra_id_94>",
|
| 98 |
+
"<extra_id_95>",
|
| 99 |
+
"<extra_id_96>",
|
| 100 |
+
"<extra_id_97>",
|
| 101 |
+
"<extra_id_98>",
|
| 102 |
+
"<extra_id_99>",
|
| 103 |
+
"<extra_id_100>",
|
| 104 |
+
"<extra_id_101>",
|
| 105 |
+
"<extra_id_102>",
|
| 106 |
+
"<extra_id_103>",
|
| 107 |
+
"<extra_id_104>",
|
| 108 |
+
"<extra_id_105>",
|
| 109 |
+
"<extra_id_106>",
|
| 110 |
+
"<extra_id_107>",
|
| 111 |
+
"<extra_id_108>",
|
| 112 |
+
"<extra_id_109>",
|
| 113 |
+
"<extra_id_110>",
|
| 114 |
+
"<extra_id_111>",
|
| 115 |
+
"<extra_id_112>",
|
| 116 |
+
"<extra_id_113>",
|
| 117 |
+
"<extra_id_114>",
|
| 118 |
+
"<extra_id_115>",
|
| 119 |
+
"<extra_id_116>",
|
| 120 |
+
"<extra_id_117>",
|
| 121 |
+
"<extra_id_118>",
|
| 122 |
+
"<extra_id_119>",
|
| 123 |
+
"<extra_id_120>",
|
| 124 |
+
"<extra_id_121>",
|
| 125 |
+
"<extra_id_122>",
|
| 126 |
+
"<extra_id_123>",
|
| 127 |
+
"<extra_id_124>"
|
| 128 |
+
],
|
| 129 |
+
"bos_token": "</s>",
|
| 130 |
+
"eos_token": {
|
| 131 |
+
"content": "</s>",
|
| 132 |
+
"lstrip": false,
|
| 133 |
+
"normalized": true,
|
| 134 |
+
"rstrip": false,
|
| 135 |
+
"single_word": false
|
| 136 |
+
},
|
| 137 |
+
"pad_token": {
|
| 138 |
+
"content": "<pad>",
|
| 139 |
+
"lstrip": false,
|
| 140 |
+
"normalized": true,
|
| 141 |
+
"rstrip": false,
|
| 142 |
+
"single_word": false
|
| 143 |
+
},
|
| 144 |
+
"unk_token": {
|
| 145 |
+
"content": "<unk>",
|
| 146 |
+
"lstrip": false,
|
| 147 |
+
"normalized": true,
|
| 148 |
+
"rstrip": false,
|
| 149 |
+
"single_word": false
|
| 150 |
+
}
|
| 151 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,160 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": false,
|
| 5 |
+
"additional_special_tokens": [
|
| 6 |
+
"<extra_id_0>",
|
| 7 |
+
"<extra_id_1>",
|
| 8 |
+
"<extra_id_2>",
|
| 9 |
+
"<extra_id_3>",
|
| 10 |
+
"<extra_id_4>",
|
| 11 |
+
"<extra_id_5>",
|
| 12 |
+
"<extra_id_6>",
|
| 13 |
+
"<extra_id_7>",
|
| 14 |
+
"<extra_id_8>",
|
| 15 |
+
"<extra_id_9>",
|
| 16 |
+
"<extra_id_10>",
|
| 17 |
+
"<extra_id_11>",
|
| 18 |
+
"<extra_id_12>",
|
| 19 |
+
"<extra_id_13>",
|
| 20 |
+
"<extra_id_14>",
|
| 21 |
+
"<extra_id_15>",
|
| 22 |
+
"<extra_id_16>",
|
| 23 |
+
"<extra_id_17>",
|
| 24 |
+
"<extra_id_18>",
|
| 25 |
+
"<extra_id_19>",
|
| 26 |
+
"<extra_id_20>",
|
| 27 |
+
"<extra_id_21>",
|
| 28 |
+
"<extra_id_22>",
|
| 29 |
+
"<extra_id_23>",
|
| 30 |
+
"<extra_id_24>",
|
| 31 |
+
"<extra_id_25>",
|
| 32 |
+
"<extra_id_26>",
|
| 33 |
+
"<extra_id_27>",
|
| 34 |
+
"<extra_id_28>",
|
| 35 |
+
"<extra_id_29>",
|
| 36 |
+
"<extra_id_30>",
|
| 37 |
+
"<extra_id_31>",
|
| 38 |
+
"<extra_id_32>",
|
| 39 |
+
"<extra_id_33>",
|
| 40 |
+
"<extra_id_34>",
|
| 41 |
+
"<extra_id_35>",
|
| 42 |
+
"<extra_id_36>",
|
| 43 |
+
"<extra_id_37>",
|
| 44 |
+
"<extra_id_38>",
|
| 45 |
+
"<extra_id_39>",
|
| 46 |
+
"<extra_id_40>",
|
| 47 |
+
"<extra_id_41>",
|
| 48 |
+
"<extra_id_42>",
|
| 49 |
+
"<extra_id_43>",
|
| 50 |
+
"<extra_id_44>",
|
| 51 |
+
"<extra_id_45>",
|
| 52 |
+
"<extra_id_46>",
|
| 53 |
+
"<extra_id_47>",
|
| 54 |
+
"<extra_id_48>",
|
| 55 |
+
"<extra_id_49>",
|
| 56 |
+
"<extra_id_50>",
|
| 57 |
+
"<extra_id_51>",
|
| 58 |
+
"<extra_id_52>",
|
| 59 |
+
"<extra_id_53>",
|
| 60 |
+
"<extra_id_54>",
|
| 61 |
+
"<extra_id_55>",
|
| 62 |
+
"<extra_id_56>",
|
| 63 |
+
"<extra_id_57>",
|
| 64 |
+
"<extra_id_58>",
|
| 65 |
+
"<extra_id_59>",
|
| 66 |
+
"<extra_id_60>",
|
| 67 |
+
"<extra_id_61>",
|
| 68 |
+
"<extra_id_62>",
|
| 69 |
+
"<extra_id_63>",
|
| 70 |
+
"<extra_id_64>",
|
| 71 |
+
"<extra_id_65>",
|
| 72 |
+
"<extra_id_66>",
|
| 73 |
+
"<extra_id_67>",
|
| 74 |
+
"<extra_id_68>",
|
| 75 |
+
"<extra_id_69>",
|
| 76 |
+
"<extra_id_70>",
|
| 77 |
+
"<extra_id_71>",
|
| 78 |
+
"<extra_id_72>",
|
| 79 |
+
"<extra_id_73>",
|
| 80 |
+
"<extra_id_74>",
|
| 81 |
+
"<extra_id_75>",
|
| 82 |
+
"<extra_id_76>",
|
| 83 |
+
"<extra_id_77>",
|
| 84 |
+
"<extra_id_78>",
|
| 85 |
+
"<extra_id_79>",
|
| 86 |
+
"<extra_id_80>",
|
| 87 |
+
"<extra_id_81>",
|
| 88 |
+
"<extra_id_82>",
|
| 89 |
+
"<extra_id_83>",
|
| 90 |
+
"<extra_id_84>",
|
| 91 |
+
"<extra_id_85>",
|
| 92 |
+
"<extra_id_86>",
|
| 93 |
+
"<extra_id_87>",
|
| 94 |
+
"<extra_id_88>",
|
| 95 |
+
"<extra_id_89>",
|
| 96 |
+
"<extra_id_90>",
|
| 97 |
+
"<extra_id_91>",
|
| 98 |
+
"<extra_id_92>",
|
| 99 |
+
"<extra_id_93>",
|
| 100 |
+
"<extra_id_94>",
|
| 101 |
+
"<extra_id_95>",
|
| 102 |
+
"<extra_id_96>",
|
| 103 |
+
"<extra_id_97>",
|
| 104 |
+
"<extra_id_98>",
|
| 105 |
+
"<extra_id_99>",
|
| 106 |
+
"<extra_id_100>",
|
| 107 |
+
"<extra_id_101>",
|
| 108 |
+
"<extra_id_102>",
|
| 109 |
+
"<extra_id_103>",
|
| 110 |
+
"<extra_id_104>",
|
| 111 |
+
"<extra_id_105>",
|
| 112 |
+
"<extra_id_106>",
|
| 113 |
+
"<extra_id_107>",
|
| 114 |
+
"<extra_id_108>",
|
| 115 |
+
"<extra_id_109>",
|
| 116 |
+
"<extra_id_110>",
|
| 117 |
+
"<extra_id_111>",
|
| 118 |
+
"<extra_id_112>",
|
| 119 |
+
"<extra_id_113>",
|
| 120 |
+
"<extra_id_114>",
|
| 121 |
+
"<extra_id_115>",
|
| 122 |
+
"<extra_id_116>",
|
| 123 |
+
"<extra_id_117>",
|
| 124 |
+
"<extra_id_118>",
|
| 125 |
+
"<extra_id_119>",
|
| 126 |
+
"<extra_id_120>",
|
| 127 |
+
"<extra_id_121>",
|
| 128 |
+
"<extra_id_122>",
|
| 129 |
+
"<extra_id_123>",
|
| 130 |
+
"<extra_id_124>"
|
| 131 |
+
],
|
| 132 |
+
"eos_token": {
|
| 133 |
+
"__type": "AddedToken",
|
| 134 |
+
"content": "</s>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": true,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false
|
| 139 |
+
},
|
| 140 |
+
"extra_ids": 125,
|
| 141 |
+
"name_or_path": "models/bygpt5-medium/en",
|
| 142 |
+
"pad_token": {
|
| 143 |
+
"__type": "AddedToken",
|
| 144 |
+
"content": "<pad>",
|
| 145 |
+
"lstrip": false,
|
| 146 |
+
"normalized": true,
|
| 147 |
+
"rstrip": false,
|
| 148 |
+
"single_word": false
|
| 149 |
+
},
|
| 150 |
+
"special_tokens_map_file": "/storage/ukp/work/jbelouadi/.cache/huggingface/transformers/1d7794e7e6dcb9aff7f1b0f72b85d665041ce4cfc7c59cf78a0e83302f9ab8e3.063895353d5ef9e19a25220cb616c43abc5e84a2f11b1ffb71c29e097572a109",
|
| 151 |
+
"tokenizer_class": "ByGPT5Tokenizer",
|
| 152 |
+
"unk_token": {
|
| 153 |
+
"__type": "AddedToken",
|
| 154 |
+
"content": "<unk>",
|
| 155 |
+
"lstrip": false,
|
| 156 |
+
"normalized": true,
|
| 157 |
+
"rstrip": false,
|
| 158 |
+
"single_word": false
|
| 159 |
+
}
|
| 160 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1393 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.999915498738517,
|
| 5 |
+
"global_step": 51770,
|
| 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": 5.791505791505791e-07,
|
| 13 |
+
"loss": 2.2583,
|
| 14 |
+
"step": 1
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.05,
|
| 18 |
+
"learning_rate": 0.00014478764478764478,
|
| 19 |
+
"loss": 1.4754,
|
| 20 |
+
"step": 250
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.1,
|
| 24 |
+
"learning_rate": 0.00028957528957528956,
|
| 25 |
+
"loss": 1.2393,
|
| 26 |
+
"step": 500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.14,
|
| 30 |
+
"learning_rate": 0.0002999848327088156,
|
| 31 |
+
"loss": 1.2015,
|
| 32 |
+
"step": 750
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.19,
|
| 36 |
+
"learning_rate": 0.0002999345361024936,
|
| 37 |
+
"loss": 1.1787,
|
| 38 |
+
"step": 1000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.24,
|
| 42 |
+
"learning_rate": 0.0002998490306892274,
|
| 43 |
+
"loss": 1.1622,
|
| 44 |
+
"step": 1250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.29,
|
| 48 |
+
"learning_rate": 0.0002997283365480707,
|
| 49 |
+
"loss": 1.1493,
|
| 50 |
+
"step": 1500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.34,
|
| 54 |
+
"learning_rate": 0.0002995724820213708,
|
| 55 |
+
"loss": 1.1407,
|
| 56 |
+
"step": 1750
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.39,
|
| 60 |
+
"learning_rate": 0.00029938150370811345,
|
| 61 |
+
"loss": 1.1306,
|
| 62 |
+
"step": 2000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.43,
|
| 66 |
+
"learning_rate": 0.0002991554464553282,
|
| 67 |
+
"loss": 1.1228,
|
| 68 |
+
"step": 2250
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.48,
|
| 72 |
+
"learning_rate": 0.0002988943633475569,
|
| 73 |
+
"loss": 1.1142,
|
| 74 |
+
"step": 2500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.53,
|
| 78 |
+
"learning_rate": 0.0002985983156943883,
|
| 79 |
+
"loss": 1.1082,
|
| 80 |
+
"step": 2750
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.58,
|
| 84 |
+
"learning_rate": 0.0002982673730160606,
|
| 85 |
+
"loss": 1.1035,
|
| 86 |
+
"step": 3000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.63,
|
| 90 |
+
"learning_rate": 0.0002979016130271363,
|
| 91 |
+
"loss": 1.0949,
|
| 92 |
+
"step": 3250
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.68,
|
| 96 |
+
"learning_rate": 0.00029750112161825245,
|
| 97 |
+
"loss": 1.0909,
|
| 98 |
+
"step": 3500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.72,
|
| 102 |
+
"learning_rate": 0.00029706599283595153,
|
| 103 |
+
"loss": 1.0862,
|
| 104 |
+
"step": 3750
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.77,
|
| 108 |
+
"learning_rate": 0.0002965963288605962,
|
| 109 |
+
"loss": 1.0799,
|
| 110 |
+
"step": 4000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.82,
|
| 114 |
+
"learning_rate": 0.0002960922399823749,
|
| 115 |
+
"loss": 1.0752,
|
| 116 |
+
"step": 4250
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.87,
|
| 120 |
+
"learning_rate": 0.0002955538445754026,
|
| 121 |
+
"loss": 1.0717,
|
| 122 |
+
"step": 4500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.92,
|
| 126 |
+
"learning_rate": 0.00029498126906992285,
|
| 127 |
+
"loss": 1.0675,
|
| 128 |
+
"step": 4750
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.97,
|
| 132 |
+
"learning_rate": 0.00029437464792261906,
|
| 133 |
+
"loss": 1.0629,
|
| 134 |
+
"step": 5000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.0,
|
| 138 |
+
"eval_alliteration_score": 0.39176506206478956,
|
| 139 |
+
"eval_harmonic_meter_score": 0.10270844382058368,
|
| 140 |
+
"eval_harmonic_rhyme_score": 0.13433666525497734,
|
| 141 |
+
"eval_meter_score": 0.28353772535274874,
|
| 142 |
+
"eval_rhyme_score": 0.6209502922727462,
|
| 143 |
+
"eval_runtime": 2477.4307,
|
| 144 |
+
"eval_samples_per_second": 1.09,
|
| 145 |
+
"eval_steps_per_second": 0.136,
|
| 146 |
+
"step": 5177
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.01,
|
| 150 |
+
"learning_rate": 0.0002937341235850398,
|
| 151 |
+
"loss": 1.0558,
|
| 152 |
+
"step": 5250
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.06,
|
| 156 |
+
"learning_rate": 0.0002930598464701476,
|
| 157 |
+
"loss": 1.0356,
|
| 158 |
+
"step": 5500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.11,
|
| 162 |
+
"learning_rate": 0.00029235197491699733,
|
| 163 |
+
"loss": 1.0344,
|
| 164 |
+
"step": 5750
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.16,
|
| 168 |
+
"learning_rate": 0.0002916106751535544,
|
| 169 |
+
"loss": 1.0315,
|
| 170 |
+
"step": 6000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.21,
|
| 174 |
+
"learning_rate": 0.0002908361212576593,
|
| 175 |
+
"loss": 1.0306,
|
| 176 |
+
"step": 6250
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.26,
|
| 180 |
+
"learning_rate": 0.00029002849511614926,
|
| 181 |
+
"loss": 1.027,
|
| 182 |
+
"step": 6500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3,
|
| 186 |
+
"learning_rate": 0.0002891879863821466,
|
| 187 |
+
"loss": 1.0232,
|
| 188 |
+
"step": 6750
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.35,
|
| 192 |
+
"learning_rate": 0.00028831479243052253,
|
| 193 |
+
"loss": 1.021,
|
| 194 |
+
"step": 7000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.4,
|
| 198 |
+
"learning_rate": 0.0002874091183115481,
|
| 199 |
+
"loss": 1.0189,
|
| 200 |
+
"step": 7250
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.45,
|
| 204 |
+
"learning_rate": 0.00028647117670274266,
|
| 205 |
+
"loss": 1.0161,
|
| 206 |
+
"step": 7500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.5,
|
| 210 |
+
"learning_rate": 0.00028550118785893134,
|
| 211 |
+
"loss": 1.0132,
|
| 212 |
+
"step": 7750
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.55,
|
| 216 |
+
"learning_rate": 0.000284499379560523,
|
| 217 |
+
"loss": 1.0102,
|
| 218 |
+
"step": 8000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.59,
|
| 222 |
+
"learning_rate": 0.0002834659870600212,
|
| 223 |
+
"loss": 1.0088,
|
| 224 |
+
"step": 8250
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.64,
|
| 228 |
+
"learning_rate": 0.0002824012530267801,
|
| 229 |
+
"loss": 1.0055,
|
| 230 |
+
"step": 8500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.69,
|
| 234 |
+
"learning_rate": 0.00028130542749001906,
|
| 235 |
+
"loss": 1.0033,
|
| 236 |
+
"step": 8750
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.74,
|
| 240 |
+
"learning_rate": 0.0002801787677801088,
|
| 241 |
+
"loss": 1.001,
|
| 242 |
+
"step": 9000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.79,
|
| 246 |
+
"learning_rate": 0.00027902153846814335,
|
| 247 |
+
"loss": 0.9979,
|
| 248 |
+
"step": 9250
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.83,
|
| 252 |
+
"learning_rate": 0.0002778340113038107,
|
| 253 |
+
"loss": 0.994,
|
| 254 |
+
"step": 9500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.88,
|
| 258 |
+
"learning_rate": 0.00027661646515157913,
|
| 259 |
+
"loss": 0.9919,
|
| 260 |
+
"step": 9750
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.93,
|
| 264 |
+
"learning_rate": 0.0002753691859252116,
|
| 265 |
+
"loss": 0.9895,
|
| 266 |
+
"step": 10000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.98,
|
| 270 |
+
"learning_rate": 0.0002740924665206254,
|
| 271 |
+
"loss": 0.9881,
|
| 272 |
+
"step": 10250
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.0,
|
| 276 |
+
"eval_alliteration_score": 0.4102959309494451,
|
| 277 |
+
"eval_harmonic_meter_score": 0.11239968726166898,
|
| 278 |
+
"eval_harmonic_rhyme_score": 0.2758249186922835,
|
| 279 |
+
"eval_meter_score": 0.30497402024728615,
|
| 280 |
+
"eval_rhyme_score": 0.7041029209046453,
|
| 281 |
+
"eval_runtime": 2601.7347,
|
| 282 |
+
"eval_samples_per_second": 1.038,
|
| 283 |
+
"eval_steps_per_second": 0.13,
|
| 284 |
+
"step": 10354
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.03,
|
| 288 |
+
"learning_rate": 0.000272786606747112,
|
| 289 |
+
"loss": 0.9675,
|
| 290 |
+
"step": 10500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.08,
|
| 294 |
+
"learning_rate": 0.00027145191325693354,
|
| 295 |
+
"loss": 0.9528,
|
| 296 |
+
"step": 10750
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.12,
|
| 300 |
+
"learning_rate": 0.0002700886994733122,
|
| 301 |
+
"loss": 0.9521,
|
| 302 |
+
"step": 11000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.17,
|
| 306 |
+
"learning_rate": 0.0002686972855168298,
|
| 307 |
+
"loss": 0.952,
|
| 308 |
+
"step": 11250
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 2.22,
|
| 312 |
+
"learning_rate": 0.00026727799813025446,
|
| 313 |
+
"loss": 0.9508,
|
| 314 |
+
"step": 11500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.27,
|
| 318 |
+
"learning_rate": 0.0002658311706018126,
|
| 319 |
+
"loss": 0.9499,
|
| 320 |
+
"step": 11750
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.32,
|
| 324 |
+
"learning_rate": 0.000264357142686923,
|
| 325 |
+
"loss": 0.947,
|
| 326 |
+
"step": 12000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.37,
|
| 330 |
+
"learning_rate": 0.0002628562605284134,
|
| 331 |
+
"loss": 0.9458,
|
| 332 |
+
"step": 12250
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.41,
|
| 336 |
+
"learning_rate": 0.0002613288765752361,
|
| 337 |
+
"loss": 0.9463,
|
| 338 |
+
"step": 12500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.46,
|
| 342 |
+
"learning_rate": 0.0002597753494997032,
|
| 343 |
+
"loss": 0.9415,
|
| 344 |
+
"step": 12750
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.51,
|
| 348 |
+
"learning_rate": 0.00025819604411326055,
|
| 349 |
+
"loss": 0.9412,
|
| 350 |
+
"step": 13000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.56,
|
| 354 |
+
"learning_rate": 0.0002565913312808196,
|
| 355 |
+
"loss": 0.9381,
|
| 356 |
+
"step": 13250
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.61,
|
| 360 |
+
"learning_rate": 0.00025496158783366803,
|
| 361 |
+
"loss": 0.9364,
|
| 362 |
+
"step": 13500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.66,
|
| 366 |
+
"learning_rate": 0.00025330719648097947,
|
| 367 |
+
"loss": 0.9346,
|
| 368 |
+
"step": 13750
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.7,
|
| 372 |
+
"learning_rate": 0.0002516285457199425,
|
| 373 |
+
"loss": 0.9339,
|
| 374 |
+
"step": 14000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.75,
|
| 378 |
+
"learning_rate": 0.00024992602974453066,
|
| 379 |
+
"loss": 0.9323,
|
| 380 |
+
"step": 14250
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.8,
|
| 384 |
+
"learning_rate": 0.000248200048352935,
|
| 385 |
+
"loss": 0.9284,
|
| 386 |
+
"step": 14500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.85,
|
| 390 |
+
"learning_rate": 0.00024645100685368013,
|
| 391 |
+
"loss": 0.9269,
|
| 392 |
+
"step": 14750
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.9,
|
| 396 |
+
"learning_rate": 0.0002446793159704463,
|
| 397 |
+
"loss": 0.9252,
|
| 398 |
+
"step": 15000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.95,
|
| 402 |
+
"learning_rate": 0.00024288539174562046,
|
| 403 |
+
"loss": 0.9243,
|
| 404 |
+
"step": 15250
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.99,
|
| 408 |
+
"learning_rate": 0.00024106965544259755,
|
| 409 |
+
"loss": 0.9211,
|
| 410 |
+
"step": 15500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 3.0,
|
| 414 |
+
"eval_alliteration_score": 0.4366551398931154,
|
| 415 |
+
"eval_harmonic_meter_score": 0.11132001638651683,
|
| 416 |
+
"eval_harmonic_rhyme_score": 0.3526620304363614,
|
| 417 |
+
"eval_meter_score": 0.3150168760675491,
|
| 418 |
+
"eval_rhyme_score": 0.7517596634968854,
|
| 419 |
+
"eval_runtime": 2670.3356,
|
| 420 |
+
"eval_samples_per_second": 1.011,
|
| 421 |
+
"eval_steps_per_second": 0.127,
|
| 422 |
+
"step": 15531
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 3.04,
|
| 426 |
+
"learning_rate": 0.00023923253344685635,
|
| 427 |
+
"loss": 0.8862,
|
| 428 |
+
"step": 15750
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.09,
|
| 432 |
+
"learning_rate": 0.0002373744571658321,
|
| 433 |
+
"loss": 0.8819,
|
| 434 |
+
"step": 16000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.14,
|
| 438 |
+
"learning_rate": 0.00023549586292761015,
|
| 439 |
+
"loss": 0.8832,
|
| 440 |
+
"step": 16250
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 3.19,
|
| 444 |
+
"learning_rate": 0.00023359719187846398,
|
| 445 |
+
"loss": 0.88,
|
| 446 |
+
"step": 16500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 3.24,
|
| 450 |
+
"learning_rate": 0.00023167888987926164,
|
| 451 |
+
"loss": 0.881,
|
| 452 |
+
"step": 16750
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 3.28,
|
| 456 |
+
"learning_rate": 0.00022974140740076524,
|
| 457 |
+
"loss": 0.8792,
|
| 458 |
+
"step": 17000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.33,
|
| 462 |
+
"learning_rate": 0.00022778519941784797,
|
| 463 |
+
"loss": 0.8803,
|
| 464 |
+
"step": 17250
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.38,
|
| 468 |
+
"learning_rate": 0.00022581072530265315,
|
| 469 |
+
"loss": 0.8774,
|
| 470 |
+
"step": 17500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.43,
|
| 474 |
+
"learning_rate": 0.000223818448716721,
|
| 475 |
+
"loss": 0.8759,
|
| 476 |
+
"step": 17750
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.48,
|
| 480 |
+
"learning_rate": 0.000221808837502108,
|
| 481 |
+
"loss": 0.8737,
|
| 482 |
+
"step": 18000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 3.53,
|
| 486 |
+
"learning_rate": 0.00021978236357152417,
|
| 487 |
+
"loss": 0.8717,
|
| 488 |
+
"step": 18250
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.57,
|
| 492 |
+
"learning_rate": 0.00021773950279751545,
|
| 493 |
+
"loss": 0.871,
|
| 494 |
+
"step": 18500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 3.62,
|
| 498 |
+
"learning_rate": 0.0002156807349007151,
|
| 499 |
+
"loss": 0.8702,
|
| 500 |
+
"step": 18750
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.67,
|
| 504 |
+
"learning_rate": 0.00021360654333719197,
|
| 505 |
+
"loss": 0.8682,
|
| 506 |
+
"step": 19000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.72,
|
| 510 |
+
"learning_rate": 0.00021151741518492183,
|
| 511 |
+
"loss": 0.8659,
|
| 512 |
+
"step": 19250
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.77,
|
| 516 |
+
"learning_rate": 0.00020941384102940748,
|
| 517 |
+
"loss": 0.8658,
|
| 518 |
+
"step": 19500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.81,
|
| 522 |
+
"learning_rate": 0.00020729631484847592,
|
| 523 |
+
"loss": 0.8625,
|
| 524 |
+
"step": 19750
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 3.86,
|
| 528 |
+
"learning_rate": 0.00020516533389627854,
|
| 529 |
+
"loss": 0.8607,
|
| 530 |
+
"step": 20000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.91,
|
| 534 |
+
"learning_rate": 0.00020302139858652179,
|
| 535 |
+
"loss": 0.8604,
|
| 536 |
+
"step": 20250
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.96,
|
| 540 |
+
"learning_rate": 0.00020086501237495636,
|
| 541 |
+
"loss": 0.8578,
|
| 542 |
+
"step": 20500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 4.0,
|
| 546 |
+
"eval_alliteration_score": 0.4113367687177933,
|
| 547 |
+
"eval_harmonic_meter_score": 0.11086907691273673,
|
| 548 |
+
"eval_harmonic_rhyme_score": 0.4120432314176195,
|
| 549 |
+
"eval_meter_score": 0.30988666333117193,
|
| 550 |
+
"eval_rhyme_score": 0.7734423562173335,
|
| 551 |
+
"eval_runtime": 2384.5447,
|
| 552 |
+
"eval_samples_per_second": 1.132,
|
| 553 |
+
"eval_steps_per_second": 0.142,
|
| 554 |
+
"step": 20708
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 4.01,
|
| 558 |
+
"learning_rate": 0.00019869668164115161,
|
| 559 |
+
"loss": 0.8501,
|
| 560 |
+
"step": 20750
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 4.06,
|
| 564 |
+
"learning_rate": 0.00019651691556958343,
|
| 565 |
+
"loss": 0.8109,
|
| 566 |
+
"step": 21000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 4.1,
|
| 570 |
+
"learning_rate": 0.00019432622603006387,
|
| 571 |
+
"loss": 0.8126,
|
| 572 |
+
"step": 21250
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 4.15,
|
| 576 |
+
"learning_rate": 0.00019212512745753942,
|
| 577 |
+
"loss": 0.8134,
|
| 578 |
+
"step": 21500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 4.2,
|
| 582 |
+
"learning_rate": 0.00018991413673128776,
|
| 583 |
+
"loss": 0.8149,
|
| 584 |
+
"step": 21750
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.25,
|
| 588 |
+
"learning_rate": 0.0001876937730535397,
|
| 589 |
+
"loss": 0.8136,
|
| 590 |
+
"step": 22000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 4.3,
|
| 594 |
+
"learning_rate": 0.00018546455782755646,
|
| 595 |
+
"loss": 0.8118,
|
| 596 |
+
"step": 22250
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 4.35,
|
| 600 |
+
"learning_rate": 0.0001832270145351892,
|
| 601 |
+
"loss": 0.811,
|
| 602 |
+
"step": 22500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 4.39,
|
| 606 |
+
"learning_rate": 0.00018098166861395132,
|
| 607 |
+
"loss": 0.8101,
|
| 608 |
+
"step": 22750
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 4.44,
|
| 612 |
+
"learning_rate": 0.00017872904733363072,
|
| 613 |
+
"loss": 0.8102,
|
| 614 |
+
"step": 23000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.49,
|
| 618 |
+
"learning_rate": 0.00017646967967247255,
|
| 619 |
+
"loss": 0.8072,
|
| 620 |
+
"step": 23250
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 4.54,
|
| 624 |
+
"learning_rate": 0.00017420409619295997,
|
| 625 |
+
"loss": 0.8071,
|
| 626 |
+
"step": 23500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.59,
|
| 630 |
+
"learning_rate": 0.00017193282891722365,
|
| 631 |
+
"loss": 0.8049,
|
| 632 |
+
"step": 23750
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 4.64,
|
| 636 |
+
"learning_rate": 0.00016965641120210775,
|
| 637 |
+
"loss": 0.8035,
|
| 638 |
+
"step": 24000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 4.68,
|
| 642 |
+
"learning_rate": 0.00016737537761392346,
|
| 643 |
+
"loss": 0.8033,
|
| 644 |
+
"step": 24250
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 4.73,
|
| 648 |
+
"learning_rate": 0.00016509026380291723,
|
| 649 |
+
"loss": 0.7998,
|
| 650 |
+
"step": 24500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 4.78,
|
| 654 |
+
"learning_rate": 0.00016280160637748584,
|
| 655 |
+
"loss": 0.8009,
|
| 656 |
+
"step": 24750
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 4.83,
|
| 660 |
+
"learning_rate": 0.0001605099427781652,
|
| 661 |
+
"loss": 0.7986,
|
| 662 |
+
"step": 25000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 4.88,
|
| 666 |
+
"learning_rate": 0.00015821581115142477,
|
| 667 |
+
"loss": 0.7954,
|
| 668 |
+
"step": 25250
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.93,
|
| 672 |
+
"learning_rate": 0.00015591975022329567,
|
| 673 |
+
"loss": 0.7945,
|
| 674 |
+
"step": 25500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 4.97,
|
| 678 |
+
"learning_rate": 0.0001536222991728626,
|
| 679 |
+
"loss": 0.7938,
|
| 680 |
+
"step": 25750
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 5.0,
|
| 684 |
+
"eval_alliteration_score": 0.41413515189088657,
|
| 685 |
+
"eval_harmonic_meter_score": 0.1026455246284712,
|
| 686 |
+
"eval_harmonic_rhyme_score": 0.34333090390331705,
|
| 687 |
+
"eval_meter_score": 0.2882500313162144,
|
| 688 |
+
"eval_rhyme_score": 0.6791661142839286,
|
| 689 |
+
"eval_runtime": 2270.9559,
|
| 690 |
+
"eval_samples_per_second": 1.189,
|
| 691 |
+
"eval_steps_per_second": 0.149,
|
| 692 |
+
"step": 25885
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 5.02,
|
| 696 |
+
"learning_rate": 0.0001513239975056498,
|
| 697 |
+
"loss": 0.772,
|
| 698 |
+
"step": 26000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 5.07,
|
| 702 |
+
"learning_rate": 0.00014902538492693,
|
| 703 |
+
"loss": 0.747,
|
| 704 |
+
"step": 26250
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 5.12,
|
| 708 |
+
"learning_rate": 0.00014672700121498661,
|
| 709 |
+
"loss": 0.7489,
|
| 710 |
+
"step": 26500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 5.17,
|
| 714 |
+
"learning_rate": 0.00014442938609435875,
|
| 715 |
+
"loss": 0.748,
|
| 716 |
+
"step": 26750
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 5.22,
|
| 720 |
+
"learning_rate": 0.00014213307910909895,
|
| 721 |
+
"loss": 0.7485,
|
| 722 |
+
"step": 27000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 5.26,
|
| 726 |
+
"learning_rate": 0.000139838619496073,
|
| 727 |
+
"loss": 0.7493,
|
| 728 |
+
"step": 27250
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 5.31,
|
| 732 |
+
"learning_rate": 0.00013754654605833231,
|
| 733 |
+
"loss": 0.7471,
|
| 734 |
+
"step": 27500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 5.36,
|
| 738 |
+
"learning_rate": 0.0001352573970385877,
|
| 739 |
+
"loss": 0.7466,
|
| 740 |
+
"step": 27750
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 5.41,
|
| 744 |
+
"learning_rate": 0.00013297170999281534,
|
| 745 |
+
"loss": 0.7462,
|
| 746 |
+
"step": 28000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 5.46,
|
| 750 |
+
"learning_rate": 0.0001306900216640233,
|
| 751 |
+
"loss": 0.7451,
|
| 752 |
+
"step": 28250
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 5.51,
|
| 756 |
+
"learning_rate": 0.0001284128678562096,
|
| 757 |
+
"loss": 0.7454,
|
| 758 |
+
"step": 28500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.55,
|
| 762 |
+
"learning_rate": 0.0001261407833085402,
|
| 763 |
+
"loss": 0.7424,
|
| 764 |
+
"step": 28750
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 5.6,
|
| 768 |
+
"learning_rate": 0.00012387430156977778,
|
| 769 |
+
"loss": 0.7414,
|
| 770 |
+
"step": 29000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 5.65,
|
| 774 |
+
"learning_rate": 0.00012161395487298911,
|
| 775 |
+
"loss": 0.7405,
|
| 776 |
+
"step": 29250
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 5.7,
|
| 780 |
+
"learning_rate": 0.00011936027401056238,
|
| 781 |
+
"loss": 0.7379,
|
| 782 |
+
"step": 29500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 5.75,
|
| 786 |
+
"learning_rate": 0.00011711378820956193,
|
| 787 |
+
"loss": 0.738,
|
| 788 |
+
"step": 29750
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 5.79,
|
| 792 |
+
"learning_rate": 0.00011487502500745135,
|
| 793 |
+
"loss": 0.7363,
|
| 794 |
+
"step": 30000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 5.84,
|
| 798 |
+
"learning_rate": 0.00011264451012821272,
|
| 799 |
+
"loss": 0.7347,
|
| 800 |
+
"step": 30250
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 5.89,
|
| 804 |
+
"learning_rate": 0.00011042276735889216,
|
| 805 |
+
"loss": 0.7321,
|
| 806 |
+
"step": 30500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 5.94,
|
| 810 |
+
"learning_rate": 0.00010821031842659992,
|
| 811 |
+
"loss": 0.7332,
|
| 812 |
+
"step": 30750
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 5.99,
|
| 816 |
+
"learning_rate": 0.00010600768287599417,
|
| 817 |
+
"loss": 0.7313,
|
| 818 |
+
"step": 31000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 6.0,
|
| 822 |
+
"eval_alliteration_score": 0.431023911710607,
|
| 823 |
+
"eval_harmonic_meter_score": 0.10282598031904473,
|
| 824 |
+
"eval_harmonic_rhyme_score": 0.43195083122804273,
|
| 825 |
+
"eval_meter_score": 0.3169839487583337,
|
| 826 |
+
"eval_rhyme_score": 0.7959113397342684,
|
| 827 |
+
"eval_runtime": 2406.9132,
|
| 828 |
+
"eval_samples_per_second": 1.122,
|
| 829 |
+
"eval_steps_per_second": 0.14,
|
| 830 |
+
"step": 31062
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 6.04,
|
| 834 |
+
"learning_rate": 0.00010381537794727796,
|
| 835 |
+
"loss": 0.6992,
|
| 836 |
+
"step": 31250
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 6.08,
|
| 840 |
+
"learning_rate": 0.00010163391845473613,
|
| 841 |
+
"loss": 0.6893,
|
| 842 |
+
"step": 31500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 6.13,
|
| 846 |
+
"learning_rate": 9.946381666584314e-05,
|
| 847 |
+
"loss": 0.6898,
|
| 848 |
+
"step": 31750
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 6.18,
|
| 852 |
+
"learning_rate": 9.7305582180968e-05,
|
| 853 |
+
"loss": 0.69,
|
| 854 |
+
"step": 32000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 6.23,
|
| 858 |
+
"learning_rate": 9.51597218137062e-05,
|
| 859 |
+
"loss": 0.6907,
|
| 860 |
+
"step": 32250
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 6.28,
|
| 864 |
+
"learning_rate": 9.302673947186545e-05,
|
| 865 |
+
"loss": 0.6903,
|
| 866 |
+
"step": 32500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 6.33,
|
| 870 |
+
"learning_rate": 9.090713603913437e-05,
|
| 871 |
+
"loss": 0.69,
|
| 872 |
+
"step": 32750
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 6.37,
|
| 876 |
+
"learning_rate": 8.88014092574609e-05,
|
| 877 |
+
"loss": 0.688,
|
| 878 |
+
"step": 33000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 6.42,
|
| 882 |
+
"learning_rate": 8.671005361016862e-05,
|
| 883 |
+
"loss": 0.687,
|
| 884 |
+
"step": 33250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 6.47,
|
| 888 |
+
"learning_rate": 8.463356020583833e-05,
|
| 889 |
+
"loss": 0.6859,
|
| 890 |
+
"step": 33500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 6.52,
|
| 894 |
+
"learning_rate": 8.257241666298198e-05,
|
| 895 |
+
"loss": 0.6874,
|
| 896 |
+
"step": 33750
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 6.57,
|
| 900 |
+
"learning_rate": 8.05271069955365e-05,
|
| 901 |
+
"loss": 0.6849,
|
| 902 |
+
"step": 34000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 6.62,
|
| 906 |
+
"learning_rate": 7.849811149920353e-05,
|
| 907 |
+
"loss": 0.6831,
|
| 908 |
+
"step": 34250
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 6.66,
|
| 912 |
+
"learning_rate": 7.648590663866274e-05,
|
| 913 |
+
"loss": 0.6843,
|
| 914 |
+
"step": 34500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 6.71,
|
| 918 |
+
"learning_rate": 7.449096493568466e-05,
|
| 919 |
+
"loss": 0.6832,
|
| 920 |
+
"step": 34750
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 6.76,
|
| 924 |
+
"learning_rate": 7.25137548581695e-05,
|
| 925 |
+
"loss": 0.6816,
|
| 926 |
+
"step": 35000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 6.81,
|
| 930 |
+
"learning_rate": 7.055474071013765e-05,
|
| 931 |
+
"loss": 0.6814,
|
| 932 |
+
"step": 35250
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 6.86,
|
| 936 |
+
"learning_rate": 6.861438252269817e-05,
|
| 937 |
+
"loss": 0.6788,
|
| 938 |
+
"step": 35500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 6.91,
|
| 942 |
+
"learning_rate": 6.669313594602076e-05,
|
| 943 |
+
"loss": 0.681,
|
| 944 |
+
"step": 35750
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 6.95,
|
| 948 |
+
"learning_rate": 6.479145214233655e-05,
|
| 949 |
+
"loss": 0.6787,
|
| 950 |
+
"step": 36000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 7.0,
|
| 954 |
+
"eval_alliteration_score": 0.419920073778051,
|
| 955 |
+
"eval_harmonic_meter_score": 0.10124233945058883,
|
| 956 |
+
"eval_harmonic_rhyme_score": 0.4258480329916534,
|
| 957 |
+
"eval_meter_score": 0.3190283478146982,
|
| 958 |
+
"eval_rhyme_score": 0.7980873008918292,
|
| 959 |
+
"eval_runtime": 2453.1033,
|
| 960 |
+
"eval_samples_per_second": 1.101,
|
| 961 |
+
"eval_steps_per_second": 0.138,
|
| 962 |
+
"step": 36239
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 7.0,
|
| 966 |
+
"learning_rate": 6.290977767999224e-05,
|
| 967 |
+
"loss": 0.676,
|
| 968 |
+
"step": 36250
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 7.05,
|
| 972 |
+
"learning_rate": 6.10485544285837e-05,
|
| 973 |
+
"loss": 0.6422,
|
| 974 |
+
"step": 36500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 7.1,
|
| 978 |
+
"learning_rate": 5.920821945519259e-05,
|
| 979 |
+
"loss": 0.6424,
|
| 980 |
+
"step": 36750
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 7.15,
|
| 984 |
+
"learning_rate": 5.7389204921750964e-05,
|
| 985 |
+
"loss": 0.6446,
|
| 986 |
+
"step": 37000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 7.2,
|
| 990 |
+
"learning_rate": 5.5591937983557386e-05,
|
| 991 |
+
"loss": 0.6426,
|
| 992 |
+
"step": 37250
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 7.24,
|
| 996 |
+
"learning_rate": 5.381684068896904e-05,
|
| 997 |
+
"loss": 0.6429,
|
| 998 |
+
"step": 37500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 7.29,
|
| 1002 |
+
"learning_rate": 5.2064329880293066e-05,
|
| 1003 |
+
"loss": 0.6415,
|
| 1004 |
+
"step": 37750
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 7.34,
|
| 1008 |
+
"learning_rate": 5.0334817095900465e-05,
|
| 1009 |
+
"loss": 0.6416,
|
| 1010 |
+
"step": 38000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 7.39,
|
| 1014 |
+
"learning_rate": 4.862870847358517e-05,
|
| 1015 |
+
"loss": 0.6414,
|
| 1016 |
+
"step": 38250
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 7.44,
|
| 1020 |
+
"learning_rate": 4.694640465519168e-05,
|
| 1021 |
+
"loss": 0.6424,
|
| 1022 |
+
"step": 38500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 7.48,
|
| 1026 |
+
"learning_rate": 4.528830069253341e-05,
|
| 1027 |
+
"loss": 0.6407,
|
| 1028 |
+
"step": 38750
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 7.53,
|
| 1032 |
+
"learning_rate": 4.365478595462316e-05,
|
| 1033 |
+
"loss": 0.6414,
|
| 1034 |
+
"step": 39000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 7.58,
|
| 1038 |
+
"learning_rate": 4.2046244036238635e-05,
|
| 1039 |
+
"loss": 0.6394,
|
| 1040 |
+
"step": 39250
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 7.63,
|
| 1044 |
+
"learning_rate": 4.0463052667843626e-05,
|
| 1045 |
+
"loss": 0.6377,
|
| 1046 |
+
"step": 39500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 7.68,
|
| 1050 |
+
"learning_rate": 3.8905583626886546e-05,
|
| 1051 |
+
"loss": 0.6384,
|
| 1052 |
+
"step": 39750
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 7.73,
|
| 1056 |
+
"learning_rate": 3.7374202650496385e-05,
|
| 1057 |
+
"loss": 0.6387,
|
| 1058 |
+
"step": 40000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 7.77,
|
| 1062 |
+
"learning_rate": 3.586926934959779e-05,
|
| 1063 |
+
"loss": 0.6388,
|
| 1064 |
+
"step": 40250
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 7.82,
|
| 1068 |
+
"learning_rate": 3.439113712446411e-05,
|
| 1069 |
+
"loss": 0.6369,
|
| 1070 |
+
"step": 40500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 7.87,
|
| 1074 |
+
"learning_rate": 3.294015308172949e-05,
|
| 1075 |
+
"loss": 0.637,
|
| 1076 |
+
"step": 40750
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 7.92,
|
| 1080 |
+
"learning_rate": 3.151665795287821e-05,
|
| 1081 |
+
"loss": 0.6353,
|
| 1082 |
+
"step": 41000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 7.97,
|
| 1086 |
+
"learning_rate": 3.0120986014231658e-05,
|
| 1087 |
+
"loss": 0.636,
|
| 1088 |
+
"step": 41250
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 8.0,
|
| 1092 |
+
"eval_alliteration_score": 0.44240757714634893,
|
| 1093 |
+
"eval_harmonic_meter_score": 0.11214598745179055,
|
| 1094 |
+
"eval_harmonic_rhyme_score": 0.4118503741406305,
|
| 1095 |
+
"eval_meter_score": 0.3255113242746067,
|
| 1096 |
+
"eval_rhyme_score": 0.7902240726132346,
|
| 1097 |
+
"eval_runtime": 2501.3058,
|
| 1098 |
+
"eval_samples_per_second": 1.079,
|
| 1099 |
+
"eval_steps_per_second": 0.135,
|
| 1100 |
+
"step": 41416
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 8.02,
|
| 1104 |
+
"learning_rate": 2.8753465008450878e-05,
|
| 1105 |
+
"loss": 0.613,
|
| 1106 |
+
"step": 41500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 8.06,
|
| 1110 |
+
"learning_rate": 2.7414416067573497e-05,
|
| 1111 |
+
"loss": 0.6104,
|
| 1112 |
+
"step": 41750
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 8.11,
|
| 1116 |
+
"learning_rate": 2.6104153637602887e-05,
|
| 1117 |
+
"loss": 0.6121,
|
| 1118 |
+
"step": 42000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 8.16,
|
| 1122 |
+
"learning_rate": 2.4822985404667573e-05,
|
| 1123 |
+
"loss": 0.6116,
|
| 1124 |
+
"step": 42250
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 8.21,
|
| 1128 |
+
"learning_rate": 2.3571212222767847e-05,
|
| 1129 |
+
"loss": 0.6112,
|
| 1130 |
+
"step": 42500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 8.26,
|
| 1134 |
+
"learning_rate": 2.2349128043127078e-05,
|
| 1135 |
+
"loss": 0.61,
|
| 1136 |
+
"step": 42750
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 8.31,
|
| 1140 |
+
"learning_rate": 2.115701984516349e-05,
|
| 1141 |
+
"loss": 0.611,
|
| 1142 |
+
"step": 43000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 8.35,
|
| 1146 |
+
"learning_rate": 1.999516756909961e-05,
|
| 1147 |
+
"loss": 0.6095,
|
| 1148 |
+
"step": 43250
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 8.4,
|
| 1152 |
+
"learning_rate": 1.886384405022437e-05,
|
| 1153 |
+
"loss": 0.6114,
|
| 1154 |
+
"step": 43500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 8.45,
|
| 1158 |
+
"learning_rate": 1.7763314954823966e-05,
|
| 1159 |
+
"loss": 0.6113,
|
| 1160 |
+
"step": 43750
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 8.5,
|
| 1164 |
+
"learning_rate": 1.669383871779581e-05,
|
| 1165 |
+
"loss": 0.6121,
|
| 1166 |
+
"step": 44000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 8.55,
|
| 1170 |
+
"learning_rate": 1.5655666481960795e-05,
|
| 1171 |
+
"loss": 0.6104,
|
| 1172 |
+
"step": 44250
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 8.6,
|
| 1176 |
+
"learning_rate": 1.4649042039088254e-05,
|
| 1177 |
+
"loss": 0.6114,
|
| 1178 |
+
"step": 44500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 8.64,
|
| 1182 |
+
"learning_rate": 1.3674201772646525e-05,
|
| 1183 |
+
"loss": 0.6089,
|
| 1184 |
+
"step": 44750
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 8.69,
|
| 1188 |
+
"learning_rate": 1.2731374602293753e-05,
|
| 1189 |
+
"loss": 0.6079,
|
| 1190 |
+
"step": 45000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 8.74,
|
| 1194 |
+
"learning_rate": 1.1820781930120982e-05,
|
| 1195 |
+
"loss": 0.6103,
|
| 1196 |
+
"step": 45250
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 8.79,
|
| 1200 |
+
"learning_rate": 1.0942637588661024e-05,
|
| 1201 |
+
"loss": 0.6087,
|
| 1202 |
+
"step": 45500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 8.84,
|
| 1206 |
+
"learning_rate": 1.0097147790674365e-05,
|
| 1207 |
+
"loss": 0.6091,
|
| 1208 |
+
"step": 45750
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 8.89,
|
| 1212 |
+
"learning_rate": 9.28451108072481e-06,
|
| 1213 |
+
"loss": 0.6088,
|
| 1214 |
+
"step": 46000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 8.93,
|
| 1218 |
+
"learning_rate": 8.504918288555651e-06,
|
| 1219 |
+
"loss": 0.6089,
|
| 1220 |
+
"step": 46250
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 8.98,
|
| 1224 |
+
"learning_rate": 7.758552484277575e-06,
|
| 1225 |
+
"loss": 0.6089,
|
| 1226 |
+
"step": 46500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 9.0,
|
| 1230 |
+
"eval_alliteration_score": 0.4335748792270531,
|
| 1231 |
+
"eval_harmonic_meter_score": 0.1091721650313209,
|
| 1232 |
+
"eval_harmonic_rhyme_score": 0.4075133294099665,
|
| 1233 |
+
"eval_meter_score": 0.3241651625589595,
|
| 1234 |
+
"eval_rhyme_score": 0.7938227877904658,
|
| 1235 |
+
"eval_runtime": 2386.2391,
|
| 1236 |
+
"eval_samples_per_second": 1.131,
|
| 1237 |
+
"eval_steps_per_second": 0.142,
|
| 1238 |
+
"step": 46593
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 9.03,
|
| 1242 |
+
"learning_rate": 7.045588935378604e-06,
|
| 1243 |
+
"loss": 0.6036,
|
| 1244 |
+
"step": 46750
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 9.08,
|
| 1248 |
+
"learning_rate": 6.366195065566454e-06,
|
| 1249 |
+
"loss": 0.5958,
|
| 1250 |
+
"step": 47000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 9.13,
|
| 1254 |
+
"learning_rate": 5.7205304154527955e-06,
|
| 1255 |
+
"loss": 0.5959,
|
| 1256 |
+
"step": 47250
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 9.18,
|
| 1260 |
+
"learning_rate": 5.108746605088709e-06,
|
| 1261 |
+
"loss": 0.5972,
|
| 1262 |
+
"step": 47500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 9.22,
|
| 1266 |
+
"learning_rate": 4.530987298359967e-06,
|
| 1267 |
+
"loss": 0.5972,
|
| 1268 |
+
"step": 47750
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 9.27,
|
| 1272 |
+
"learning_rate": 3.987388169250816e-06,
|
| 1273 |
+
"loss": 0.599,
|
| 1274 |
+
"step": 48000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 9.32,
|
| 1278 |
+
"learning_rate": 3.4780768699839313e-06,
|
| 1279 |
+
"loss": 0.5964,
|
| 1280 |
+
"step": 48250
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 9.37,
|
| 1284 |
+
"learning_rate": 3.0031730010441957e-06,
|
| 1285 |
+
"loss": 0.5978,
|
| 1286 |
+
"step": 48500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 9.42,
|
| 1290 |
+
"learning_rate": 2.562788083092998e-06,
|
| 1291 |
+
"loss": 0.5971,
|
| 1292 |
+
"step": 48750
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 9.46,
|
| 1296 |
+
"learning_rate": 2.1570255307802486e-06,
|
| 1297 |
+
"loss": 0.5957,
|
| 1298 |
+
"step": 49000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 9.51,
|
| 1302 |
+
"learning_rate": 1.785980628459638e-06,
|
| 1303 |
+
"loss": 0.5962,
|
| 1304 |
+
"step": 49250
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 9.56,
|
| 1308 |
+
"learning_rate": 1.4497405078132517e-06,
|
| 1309 |
+
"loss": 0.5976,
|
| 1310 |
+
"step": 49500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 9.61,
|
| 1314 |
+
"learning_rate": 1.1483841273906391e-06,
|
| 1315 |
+
"loss": 0.5977,
|
| 1316 |
+
"step": 49750
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 9.66,
|
| 1320 |
+
"learning_rate": 8.819822540670762e-07,
|
| 1321 |
+
"loss": 0.596,
|
| 1322 |
+
"step": 50000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 9.71,
|
| 1326 |
+
"learning_rate": 6.505974464255936e-07,
|
| 1327 |
+
"loss": 0.5967,
|
| 1328 |
+
"step": 50250
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 9.75,
|
| 1332 |
+
"learning_rate": 4.542840400665104e-07,
|
| 1333 |
+
"loss": 0.5965,
|
| 1334 |
+
"step": 50500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 9.8,
|
| 1338 |
+
"learning_rate": 2.9308813484780755e-07,
|
| 1339 |
+
"loss": 0.5977,
|
| 1340 |
+
"step": 50750
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 9.85,
|
| 1344 |
+
"learning_rate": 1.670475840597041e-07,
|
| 1345 |
+
"loss": 0.5965,
|
| 1346 |
+
"step": 51000
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 9.9,
|
| 1350 |
+
"learning_rate": 7.619198553558414e-08,
|
| 1351 |
+
"loss": 0.5948,
|
| 1352 |
+
"step": 51250
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 9.95,
|
| 1356 |
+
"learning_rate": 2.0542674701656824e-08,
|
| 1357 |
+
"loss": 0.598,
|
| 1358 |
+
"step": 51500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 10.0,
|
| 1362 |
+
"learning_rate": 1.1271956673053472e-10,
|
| 1363 |
+
"loss": 0.5966,
|
| 1364 |
+
"step": 51750
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 10.0,
|
| 1368 |
+
"eval_alliteration_score": 0.43619909502262444,
|
| 1369 |
+
"eval_harmonic_meter_score": 0.10679777933598275,
|
| 1370 |
+
"eval_harmonic_rhyme_score": 0.4260903757100796,
|
| 1371 |
+
"eval_meter_score": 0.31873257964794116,
|
| 1372 |
+
"eval_rhyme_score": 0.8030446845582035,
|
| 1373 |
+
"eval_runtime": 2387.7503,
|
| 1374 |
+
"eval_samples_per_second": 1.131,
|
| 1375 |
+
"eval_steps_per_second": 0.142,
|
| 1376 |
+
"step": 51770
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 10.0,
|
| 1380 |
+
"step": 51770,
|
| 1381 |
+
"total_flos": 2.202815059032277e+18,
|
| 1382 |
+
"train_loss": 0.12071779469207651,
|
| 1383 |
+
"train_runtime": 12056.8595,
|
| 1384 |
+
"train_samples_per_second": 549.652,
|
| 1385 |
+
"train_steps_per_second": 4.294
|
| 1386 |
+
}
|
| 1387 |
+
],
|
| 1388 |
+
"max_steps": 51770,
|
| 1389 |
+
"num_train_epochs": 10,
|
| 1390 |
+
"total_flos": 2.202815059032277e+18,
|
| 1391 |
+
"trial_name": null,
|
| 1392 |
+
"trial_params": null
|
| 1393 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd9be3fbfbdfc20f9b8c6e1d3589e8345d82e4dab13cf5baae11a1ce758dcc17
|
| 3 |
+
size 3439
|