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-base/en",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ByGPT5LMHeadModel"
|
| 5 |
+
],
|
| 6 |
+
"d_ff": 3968,
|
| 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": 6,
|
| 22 |
+
"num_heads": 12,
|
| 23 |
+
"num_layers": 6,
|
| 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 sha256:947307b129a53136b2d4d2af15f4994a9a47fb7e68933a86aea631d0fb2e2111
|
| 3 |
+
size 556894071
|
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-base/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": "/home/jbelouadi/.cache/huggingface/transformers/f22e687c418a9ed3e651ca340d6a5880bbb312bc24bbe893f5cd47288891b89d.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": 9.999903428295509,
|
| 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": 1.1583011583011583e-06,
|
| 13 |
+
"loss": 2.3104,
|
| 14 |
+
"step": 1
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.05,
|
| 18 |
+
"learning_rate": 0.00028957528957528956,
|
| 19 |
+
"loss": 1.4338,
|
| 20 |
+
"step": 250
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.1,
|
| 24 |
+
"learning_rate": 0.0005791505791505791,
|
| 25 |
+
"loss": 1.2416,
|
| 26 |
+
"step": 500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.14,
|
| 30 |
+
"learning_rate": 0.0005999696654176312,
|
| 31 |
+
"loss": 1.2139,
|
| 32 |
+
"step": 750
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.19,
|
| 36 |
+
"learning_rate": 0.0005998690722049872,
|
| 37 |
+
"loss": 1.1936,
|
| 38 |
+
"step": 1000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.24,
|
| 42 |
+
"learning_rate": 0.0005996980613784548,
|
| 43 |
+
"loss": 1.1791,
|
| 44 |
+
"step": 1250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.29,
|
| 48 |
+
"learning_rate": 0.0005994566730961414,
|
| 49 |
+
"loss": 1.1672,
|
| 50 |
+
"step": 1500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.34,
|
| 54 |
+
"learning_rate": 0.0005991449640427416,
|
| 55 |
+
"loss": 1.1574,
|
| 56 |
+
"step": 1750
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.39,
|
| 60 |
+
"learning_rate": 0.0005987630074162269,
|
| 61 |
+
"loss": 1.1486,
|
| 62 |
+
"step": 2000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.43,
|
| 66 |
+
"learning_rate": 0.0005983108929106564,
|
| 67 |
+
"loss": 1.1427,
|
| 68 |
+
"step": 2250
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.48,
|
| 72 |
+
"learning_rate": 0.0005977887266951138,
|
| 73 |
+
"loss": 1.1356,
|
| 74 |
+
"step": 2500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.53,
|
| 78 |
+
"learning_rate": 0.0005971966313887766,
|
| 79 |
+
"loss": 1.1272,
|
| 80 |
+
"step": 2750
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.58,
|
| 84 |
+
"learning_rate": 0.0005965347460321212,
|
| 85 |
+
"loss": 1.1254,
|
| 86 |
+
"step": 3000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.63,
|
| 90 |
+
"learning_rate": 0.0005958032260542726,
|
| 91 |
+
"loss": 1.118,
|
| 92 |
+
"step": 3250
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.68,
|
| 96 |
+
"learning_rate": 0.0005950022432365049,
|
| 97 |
+
"loss": 1.1146,
|
| 98 |
+
"step": 3500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.72,
|
| 102 |
+
"learning_rate": 0.0005941319856719031,
|
| 103 |
+
"loss": 1.1097,
|
| 104 |
+
"step": 3750
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.77,
|
| 108 |
+
"learning_rate": 0.0005931926577211924,
|
| 109 |
+
"loss": 1.1063,
|
| 110 |
+
"step": 4000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.82,
|
| 114 |
+
"learning_rate": 0.0005921844799647499,
|
| 115 |
+
"loss": 1.1027,
|
| 116 |
+
"step": 4250
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.87,
|
| 120 |
+
"learning_rate": 0.0005911076891508052,
|
| 121 |
+
"loss": 1.0999,
|
| 122 |
+
"step": 4500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.92,
|
| 126 |
+
"learning_rate": 0.0005899625381398457,
|
| 127 |
+
"loss": 1.0966,
|
| 128 |
+
"step": 4750
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.97,
|
| 132 |
+
"learning_rate": 0.0005887492958452381,
|
| 133 |
+
"loss": 1.0931,
|
| 134 |
+
"step": 5000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.0,
|
| 138 |
+
"eval_alliteration_score": 0.3998726520216492,
|
| 139 |
+
"eval_harmonic_meter_score": 0.1010308935212264,
|
| 140 |
+
"eval_harmonic_rhyme_score": 0.35292231999777424,
|
| 141 |
+
"eval_meter_score": 0.3077031352625674,
|
| 142 |
+
"eval_rhyme_score": 0.7291149689204202,
|
| 143 |
+
"eval_runtime": 3230.3537,
|
| 144 |
+
"eval_samples_per_second": 0.836,
|
| 145 |
+
"eval_steps_per_second": 0.026,
|
| 146 |
+
"step": 5177
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.01,
|
| 150 |
+
"learning_rate": 0.0005874682471700796,
|
| 151 |
+
"loss": 1.0865,
|
| 152 |
+
"step": 5250
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.06,
|
| 156 |
+
"learning_rate": 0.0005861196929402952,
|
| 157 |
+
"loss": 1.0684,
|
| 158 |
+
"step": 5500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.11,
|
| 162 |
+
"learning_rate": 0.0005847039498339947,
|
| 163 |
+
"loss": 1.0705,
|
| 164 |
+
"step": 5750
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.16,
|
| 168 |
+
"learning_rate": 0.0005832213503071088,
|
| 169 |
+
"loss": 1.0694,
|
| 170 |
+
"step": 6000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.21,
|
| 174 |
+
"learning_rate": 0.0005816722425153186,
|
| 175 |
+
"loss": 1.0678,
|
| 176 |
+
"step": 6250
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.26,
|
| 180 |
+
"learning_rate": 0.0005800569902322985,
|
| 181 |
+
"loss": 1.0669,
|
| 182 |
+
"step": 6500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.3,
|
| 186 |
+
"learning_rate": 0.0005783759727642932,
|
| 187 |
+
"loss": 1.0634,
|
| 188 |
+
"step": 6750
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.35,
|
| 192 |
+
"learning_rate": 0.0005766295848610451,
|
| 193 |
+
"loss": 1.062,
|
| 194 |
+
"step": 7000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.4,
|
| 198 |
+
"learning_rate": 0.0005748182366230962,
|
| 199 |
+
"loss": 1.0602,
|
| 200 |
+
"step": 7250
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.45,
|
| 204 |
+
"learning_rate": 0.0005729423534054853,
|
| 205 |
+
"loss": 1.0587,
|
| 206 |
+
"step": 7500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.5,
|
| 210 |
+
"learning_rate": 0.0005710023757178627,
|
| 211 |
+
"loss": 1.0564,
|
| 212 |
+
"step": 7750
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.55,
|
| 216 |
+
"learning_rate": 0.000568998759121046,
|
| 217 |
+
"loss": 1.0547,
|
| 218 |
+
"step": 8000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.59,
|
| 222 |
+
"learning_rate": 0.0005669319741200425,
|
| 223 |
+
"loss": 1.0536,
|
| 224 |
+
"step": 8250
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.64,
|
| 228 |
+
"learning_rate": 0.0005648025060535602,
|
| 229 |
+
"loss": 1.0517,
|
| 230 |
+
"step": 8500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.69,
|
| 234 |
+
"learning_rate": 0.0005626108549800381,
|
| 235 |
+
"loss": 1.0498,
|
| 236 |
+
"step": 8750
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.74,
|
| 240 |
+
"learning_rate": 0.0005603575355602176,
|
| 241 |
+
"loss": 1.0482,
|
| 242 |
+
"step": 9000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.79,
|
| 246 |
+
"learning_rate": 0.0005580430769362867,
|
| 247 |
+
"loss": 1.0479,
|
| 248 |
+
"step": 9250
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.83,
|
| 252 |
+
"learning_rate": 0.0005556680226076214,
|
| 253 |
+
"loss": 1.0446,
|
| 254 |
+
"step": 9500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.88,
|
| 258 |
+
"learning_rate": 0.0005532329303031583,
|
| 259 |
+
"loss": 1.0444,
|
| 260 |
+
"step": 9750
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.93,
|
| 264 |
+
"learning_rate": 0.0005507383718504232,
|
| 265 |
+
"loss": 1.0416,
|
| 266 |
+
"step": 10000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.98,
|
| 270 |
+
"learning_rate": 0.0005481849330412508,
|
| 271 |
+
"loss": 1.0392,
|
| 272 |
+
"step": 10250
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.0,
|
| 276 |
+
"eval_alliteration_score": 0.43183984747378457,
|
| 277 |
+
"eval_harmonic_meter_score": 0.09296526825958593,
|
| 278 |
+
"eval_harmonic_rhyme_score": 0.5441505816703797,
|
| 279 |
+
"eval_meter_score": 0.30327833996094755,
|
| 280 |
+
"eval_rhyme_score": 0.8429996906126559,
|
| 281 |
+
"eval_runtime": 2339.9147,
|
| 282 |
+
"eval_samples_per_second": 1.154,
|
| 283 |
+
"eval_steps_per_second": 0.036,
|
| 284 |
+
"step": 10354
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.03,
|
| 288 |
+
"learning_rate": 0.000545573213494224,
|
| 289 |
+
"loss": 1.0256,
|
| 290 |
+
"step": 10500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.08,
|
| 294 |
+
"learning_rate": 0.0005429038265138671,
|
| 295 |
+
"loss": 1.0149,
|
| 296 |
+
"step": 10750
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.12,
|
| 300 |
+
"learning_rate": 0.0005401773989466244,
|
| 301 |
+
"loss": 1.0166,
|
| 302 |
+
"step": 11000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.17,
|
| 306 |
+
"learning_rate": 0.0005373945710336596,
|
| 307 |
+
"loss": 1.0157,
|
| 308 |
+
"step": 11250
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 2.22,
|
| 312 |
+
"learning_rate": 0.0005345559962605089,
|
| 313 |
+
"loss": 1.0155,
|
| 314 |
+
"step": 11500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.27,
|
| 318 |
+
"learning_rate": 0.0005316623412036252,
|
| 319 |
+
"loss": 1.0135,
|
| 320 |
+
"step": 11750
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.32,
|
| 324 |
+
"learning_rate": 0.000528714285373846,
|
| 325 |
+
"loss": 1.0134,
|
| 326 |
+
"step": 12000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.37,
|
| 330 |
+
"learning_rate": 0.0005257125210568268,
|
| 331 |
+
"loss": 1.0118,
|
| 332 |
+
"step": 12250
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.41,
|
| 336 |
+
"learning_rate": 0.0005226577531504722,
|
| 337 |
+
"loss": 1.012,
|
| 338 |
+
"step": 12500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.46,
|
| 342 |
+
"learning_rate": 0.0005195506989994064,
|
| 343 |
+
"loss": 1.0101,
|
| 344 |
+
"step": 12750
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.51,
|
| 348 |
+
"learning_rate": 0.0005163920882265211,
|
| 349 |
+
"loss": 1.0097,
|
| 350 |
+
"step": 13000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.56,
|
| 354 |
+
"learning_rate": 0.0005131826625616392,
|
| 355 |
+
"loss": 1.0085,
|
| 356 |
+
"step": 13250
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.61,
|
| 360 |
+
"learning_rate": 0.0005099231756673361,
|
| 361 |
+
"loss": 1.0052,
|
| 362 |
+
"step": 13500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.66,
|
| 366 |
+
"learning_rate": 0.0005066143929619589,
|
| 367 |
+
"loss": 1.0064,
|
| 368 |
+
"step": 13750
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.7,
|
| 372 |
+
"learning_rate": 0.000503257091439885,
|
| 373 |
+
"loss": 1.0044,
|
| 374 |
+
"step": 14000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.75,
|
| 378 |
+
"learning_rate": 0.0004998520594890613,
|
| 379 |
+
"loss": 1.0015,
|
| 380 |
+
"step": 14250
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.8,
|
| 384 |
+
"learning_rate": 0.00049640009670587,
|
| 385 |
+
"loss": 1.0014,
|
| 386 |
+
"step": 14500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.85,
|
| 390 |
+
"learning_rate": 0.0004929020137073603,
|
| 391 |
+
"loss": 1.0006,
|
| 392 |
+
"step": 14750
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.9,
|
| 396 |
+
"learning_rate": 0.0004893586319408926,
|
| 397 |
+
"loss": 0.9997,
|
| 398 |
+
"step": 15000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.95,
|
| 402 |
+
"learning_rate": 0.0004857707834912409,
|
| 403 |
+
"loss": 0.9991,
|
| 404 |
+
"step": 15250
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.99,
|
| 408 |
+
"learning_rate": 0.0004821393108851951,
|
| 409 |
+
"loss": 0.9969,
|
| 410 |
+
"step": 15500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 3.0,
|
| 414 |
+
"eval_alliteration_score": 0.40482822655524603,
|
| 415 |
+
"eval_harmonic_meter_score": 0.10445053009647749,
|
| 416 |
+
"eval_harmonic_rhyme_score": 0.49816467657522634,
|
| 417 |
+
"eval_meter_score": 0.3098783806475492,
|
| 418 |
+
"eval_rhyme_score": 0.8195044525640993,
|
| 419 |
+
"eval_runtime": 2332.5068,
|
| 420 |
+
"eval_samples_per_second": 1.158,
|
| 421 |
+
"eval_steps_per_second": 0.036,
|
| 422 |
+
"step": 15531
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 3.04,
|
| 426 |
+
"learning_rate": 0.0004784650668937127,
|
| 427 |
+
"loss": 0.9695,
|
| 428 |
+
"step": 15750
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.09,
|
| 432 |
+
"learning_rate": 0.0004747489143316642,
|
| 433 |
+
"loss": 0.9666,
|
| 434 |
+
"step": 16000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.14,
|
| 438 |
+
"learning_rate": 0.0004709917258552203,
|
| 439 |
+
"loss": 0.9684,
|
| 440 |
+
"step": 16250
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 3.19,
|
| 444 |
+
"learning_rate": 0.00046719438375692797,
|
| 445 |
+
"loss": 0.9701,
|
| 446 |
+
"step": 16500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 3.24,
|
| 450 |
+
"learning_rate": 0.0004633577797585233,
|
| 451 |
+
"loss": 0.9693,
|
| 452 |
+
"step": 16750
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 3.28,
|
| 456 |
+
"learning_rate": 0.0004594828148015305,
|
| 457 |
+
"loss": 0.9689,
|
| 458 |
+
"step": 17000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.33,
|
| 462 |
+
"learning_rate": 0.00045557039883569595,
|
| 463 |
+
"loss": 0.9675,
|
| 464 |
+
"step": 17250
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.38,
|
| 468 |
+
"learning_rate": 0.0004516214506053063,
|
| 469 |
+
"loss": 0.966,
|
| 470 |
+
"step": 17500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.43,
|
| 474 |
+
"learning_rate": 0.000447636897433442,
|
| 475 |
+
"loss": 0.9664,
|
| 476 |
+
"step": 17750
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.48,
|
| 480 |
+
"learning_rate": 0.000443617675004216,
|
| 481 |
+
"loss": 0.9645,
|
| 482 |
+
"step": 18000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 3.53,
|
| 486 |
+
"learning_rate": 0.00043956472714304834,
|
| 487 |
+
"loss": 0.9635,
|
| 488 |
+
"step": 18250
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.57,
|
| 492 |
+
"learning_rate": 0.0004354790055950309,
|
| 493 |
+
"loss": 0.9621,
|
| 494 |
+
"step": 18500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 3.62,
|
| 498 |
+
"learning_rate": 0.0004313614698014302,
|
| 499 |
+
"loss": 0.962,
|
| 500 |
+
"step": 18750
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.67,
|
| 504 |
+
"learning_rate": 0.00042721308667438394,
|
| 505 |
+
"loss": 0.9609,
|
| 506 |
+
"step": 19000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.72,
|
| 510 |
+
"learning_rate": 0.00042303483036984366,
|
| 511 |
+
"loss": 0.9596,
|
| 512 |
+
"step": 19250
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.77,
|
| 516 |
+
"learning_rate": 0.00041882768205881495,
|
| 517 |
+
"loss": 0.9578,
|
| 518 |
+
"step": 19500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.81,
|
| 522 |
+
"learning_rate": 0.00041459262969695184,
|
| 523 |
+
"loss": 0.9568,
|
| 524 |
+
"step": 19750
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 3.86,
|
| 528 |
+
"learning_rate": 0.0004103306677925571,
|
| 529 |
+
"loss": 0.9552,
|
| 530 |
+
"step": 20000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.91,
|
| 534 |
+
"learning_rate": 0.00040604279717304357,
|
| 535 |
+
"loss": 0.9534,
|
| 536 |
+
"step": 20250
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.96,
|
| 540 |
+
"learning_rate": 0.0004017300247499127,
|
| 541 |
+
"loss": 0.9541,
|
| 542 |
+
"step": 20500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 4.0,
|
| 546 |
+
"eval_alliteration_score": 0.43839452395768513,
|
| 547 |
+
"eval_harmonic_meter_score": 0.12086540507848588,
|
| 548 |
+
"eval_harmonic_rhyme_score": 0.4986143276050798,
|
| 549 |
+
"eval_meter_score": 0.338880767934961,
|
| 550 |
+
"eval_rhyme_score": 0.8207791218734892,
|
| 551 |
+
"eval_runtime": 2519.5282,
|
| 552 |
+
"eval_samples_per_second": 1.072,
|
| 553 |
+
"eval_steps_per_second": 0.034,
|
| 554 |
+
"step": 20708
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 4.01,
|
| 558 |
+
"learning_rate": 0.00039739336328230323,
|
| 559 |
+
"loss": 0.9477,
|
| 560 |
+
"step": 20750
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 4.06,
|
| 564 |
+
"learning_rate": 0.00039303383113916687,
|
| 565 |
+
"loss": 0.9123,
|
| 566 |
+
"step": 21000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 4.1,
|
| 570 |
+
"learning_rate": 0.00038865245206012774,
|
| 571 |
+
"loss": 0.9168,
|
| 572 |
+
"step": 21250
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 4.15,
|
| 576 |
+
"learning_rate": 0.00038425025491507883,
|
| 577 |
+
"loss": 0.9193,
|
| 578 |
+
"step": 21500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 4.2,
|
| 582 |
+
"learning_rate": 0.0003798282734625755,
|
| 583 |
+
"loss": 0.9201,
|
| 584 |
+
"step": 21750
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.25,
|
| 588 |
+
"learning_rate": 0.0003753875461070794,
|
| 589 |
+
"loss": 0.9188,
|
| 590 |
+
"step": 22000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 4.3,
|
| 594 |
+
"learning_rate": 0.0003709291156551129,
|
| 595 |
+
"loss": 0.9192,
|
| 596 |
+
"step": 22250
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 4.35,
|
| 600 |
+
"learning_rate": 0.0003664540290703784,
|
| 601 |
+
"loss": 0.9173,
|
| 602 |
+
"step": 22500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 4.39,
|
| 606 |
+
"learning_rate": 0.00036196333722790264,
|
| 607 |
+
"loss": 0.9166,
|
| 608 |
+
"step": 22750
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 4.44,
|
| 612 |
+
"learning_rate": 0.00035745809466726145,
|
| 613 |
+
"loss": 0.9145,
|
| 614 |
+
"step": 23000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 4.49,
|
| 618 |
+
"learning_rate": 0.0003529393593449451,
|
| 619 |
+
"loss": 0.9144,
|
| 620 |
+
"step": 23250
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 4.54,
|
| 624 |
+
"learning_rate": 0.00034840819238591994,
|
| 625 |
+
"loss": 0.9139,
|
| 626 |
+
"step": 23500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.59,
|
| 630 |
+
"learning_rate": 0.0003438656578344473,
|
| 631 |
+
"loss": 0.9126,
|
| 632 |
+
"step": 23750
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 4.64,
|
| 636 |
+
"learning_rate": 0.0003393128224042155,
|
| 637 |
+
"loss": 0.9119,
|
| 638 |
+
"step": 24000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 4.68,
|
| 642 |
+
"learning_rate": 0.0003347507552278469,
|
| 643 |
+
"loss": 0.9099,
|
| 644 |
+
"step": 24250
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 4.73,
|
| 648 |
+
"learning_rate": 0.00033018052760583447,
|
| 649 |
+
"loss": 0.9072,
|
| 650 |
+
"step": 24500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 4.78,
|
| 654 |
+
"learning_rate": 0.0003256032127549717,
|
| 655 |
+
"loss": 0.908,
|
| 656 |
+
"step": 24750
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 4.83,
|
| 660 |
+
"learning_rate": 0.0003210198855563304,
|
| 661 |
+
"loss": 0.9063,
|
| 662 |
+
"step": 25000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 4.88,
|
| 666 |
+
"learning_rate": 0.00031643162230284954,
|
| 667 |
+
"loss": 0.9036,
|
| 668 |
+
"step": 25250
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 4.93,
|
| 672 |
+
"learning_rate": 0.00031183950044659135,
|
| 673 |
+
"loss": 0.9039,
|
| 674 |
+
"step": 25500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 4.97,
|
| 678 |
+
"learning_rate": 0.0003072445983457252,
|
| 679 |
+
"loss": 0.9023,
|
| 680 |
+
"step": 25750
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 5.0,
|
| 684 |
+
"eval_alliteration_score": 0.432449105490438,
|
| 685 |
+
"eval_harmonic_meter_score": 0.11469247564220626,
|
| 686 |
+
"eval_harmonic_rhyme_score": 0.5565104008090618,
|
| 687 |
+
"eval_meter_score": 0.3233335011377191,
|
| 688 |
+
"eval_rhyme_score": 0.8467553072173599,
|
| 689 |
+
"eval_runtime": 2231.152,
|
| 690 |
+
"eval_samples_per_second": 1.21,
|
| 691 |
+
"eval_steps_per_second": 0.038,
|
| 692 |
+
"step": 25885
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 5.02,
|
| 696 |
+
"learning_rate": 0.0003026479950112996,
|
| 697 |
+
"loss": 0.8805,
|
| 698 |
+
"step": 26000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 5.07,
|
| 702 |
+
"learning_rate": 0.00029805076985386,
|
| 703 |
+
"loss": 0.8592,
|
| 704 |
+
"step": 26250
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 5.12,
|
| 708 |
+
"learning_rate": 0.00029345400242997323,
|
| 709 |
+
"loss": 0.8613,
|
| 710 |
+
"step": 26500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 5.17,
|
| 714 |
+
"learning_rate": 0.0002888587721887175,
|
| 715 |
+
"loss": 0.8621,
|
| 716 |
+
"step": 26750
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 5.22,
|
| 720 |
+
"learning_rate": 0.0002842661582181979,
|
| 721 |
+
"loss": 0.8628,
|
| 722 |
+
"step": 27000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 5.26,
|
| 726 |
+
"learning_rate": 0.000279677238992146,
|
| 727 |
+
"loss": 0.8614,
|
| 728 |
+
"step": 27250
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 5.31,
|
| 732 |
+
"learning_rate": 0.00027509309211666463,
|
| 733 |
+
"loss": 0.8617,
|
| 734 |
+
"step": 27500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 5.36,
|
| 738 |
+
"learning_rate": 0.0002705147940771754,
|
| 739 |
+
"loss": 0.8606,
|
| 740 |
+
"step": 27750
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 5.41,
|
| 744 |
+
"learning_rate": 0.0002659434199856307,
|
| 745 |
+
"loss": 0.8597,
|
| 746 |
+
"step": 28000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 5.46,
|
| 750 |
+
"learning_rate": 0.0002613800433280466,
|
| 751 |
+
"loss": 0.8589,
|
| 752 |
+
"step": 28250
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 5.51,
|
| 756 |
+
"learning_rate": 0.0002568257357124192,
|
| 757 |
+
"loss": 0.856,
|
| 758 |
+
"step": 28500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.55,
|
| 762 |
+
"learning_rate": 0.0002522815666170804,
|
| 763 |
+
"loss": 0.8542,
|
| 764 |
+
"step": 28750
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 5.6,
|
| 768 |
+
"learning_rate": 0.00024774860313955555,
|
| 769 |
+
"loss": 0.8544,
|
| 770 |
+
"step": 29000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 5.65,
|
| 774 |
+
"learning_rate": 0.00024322790974597822,
|
| 775 |
+
"loss": 0.8517,
|
| 776 |
+
"step": 29250
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 5.7,
|
| 780 |
+
"learning_rate": 0.00023872054802112475,
|
| 781 |
+
"loss": 0.8522,
|
| 782 |
+
"step": 29500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 5.75,
|
| 786 |
+
"learning_rate": 0.00023422757641912385,
|
| 787 |
+
"loss": 0.8512,
|
| 788 |
+
"step": 29750
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 5.79,
|
| 792 |
+
"learning_rate": 0.0002297500500149027,
|
| 793 |
+
"loss": 0.8495,
|
| 794 |
+
"step": 30000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 5.84,
|
| 798 |
+
"learning_rate": 0.00022528902025642543,
|
| 799 |
+
"loss": 0.8473,
|
| 800 |
+
"step": 30250
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 5.89,
|
| 804 |
+
"learning_rate": 0.00022084553471778432,
|
| 805 |
+
"loss": 0.8451,
|
| 806 |
+
"step": 30500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 5.94,
|
| 810 |
+
"learning_rate": 0.00021642063685319983,
|
| 811 |
+
"loss": 0.8414,
|
| 812 |
+
"step": 30750
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 5.99,
|
| 816 |
+
"learning_rate": 0.00021201536575198834,
|
| 817 |
+
"loss": 0.8411,
|
| 818 |
+
"step": 31000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 6.0,
|
| 822 |
+
"eval_alliteration_score": 0.43164362519201227,
|
| 823 |
+
"eval_harmonic_meter_score": 0.10449656298407087,
|
| 824 |
+
"eval_harmonic_rhyme_score": 0.5587655117502281,
|
| 825 |
+
"eval_meter_score": 0.3140831187981907,
|
| 826 |
+
"eval_rhyme_score": 0.8511610758760736,
|
| 827 |
+
"eval_runtime": 2100.1558,
|
| 828 |
+
"eval_samples_per_second": 1.286,
|
| 829 |
+
"eval_steps_per_second": 0.04,
|
| 830 |
+
"step": 31062
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 6.04,
|
| 834 |
+
"learning_rate": 0.00020763075589455592,
|
| 835 |
+
"loss": 0.8075,
|
| 836 |
+
"step": 31250
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 6.08,
|
| 840 |
+
"learning_rate": 0.00020326783690947226,
|
| 841 |
+
"loss": 0.796,
|
| 842 |
+
"step": 31500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 6.13,
|
| 846 |
+
"learning_rate": 0.00019892763333168628,
|
| 847 |
+
"loss": 0.7985,
|
| 848 |
+
"step": 31750
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 6.18,
|
| 852 |
+
"learning_rate": 0.000194611164361936,
|
| 853 |
+
"loss": 0.7995,
|
| 854 |
+
"step": 32000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 6.23,
|
| 858 |
+
"learning_rate": 0.0001903194436274124,
|
| 859 |
+
"loss": 0.7987,
|
| 860 |
+
"step": 32250
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 6.28,
|
| 864 |
+
"learning_rate": 0.0001860534789437309,
|
| 865 |
+
"loss": 0.7977,
|
| 866 |
+
"step": 32500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 6.33,
|
| 870 |
+
"learning_rate": 0.00018181427207826875,
|
| 871 |
+
"loss": 0.799,
|
| 872 |
+
"step": 32750
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 6.37,
|
| 876 |
+
"learning_rate": 0.0001776028185149218,
|
| 877 |
+
"loss": 0.797,
|
| 878 |
+
"step": 33000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 6.42,
|
| 882 |
+
"learning_rate": 0.00017342010722033724,
|
| 883 |
+
"loss": 0.799,
|
| 884 |
+
"step": 33250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 6.47,
|
| 888 |
+
"learning_rate": 0.00016926712041167666,
|
| 889 |
+
"loss": 0.7938,
|
| 890 |
+
"step": 33500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 6.52,
|
| 894 |
+
"learning_rate": 0.00016514483332596397,
|
| 895 |
+
"loss": 0.7928,
|
| 896 |
+
"step": 33750
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 6.57,
|
| 900 |
+
"learning_rate": 0.000161054213991073,
|
| 901 |
+
"loss": 0.7948,
|
| 902 |
+
"step": 34000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 6.62,
|
| 906 |
+
"learning_rate": 0.00015699622299840705,
|
| 907 |
+
"loss": 0.7913,
|
| 908 |
+
"step": 34250
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 6.66,
|
| 912 |
+
"learning_rate": 0.00015297181327732549,
|
| 913 |
+
"loss": 0.7911,
|
| 914 |
+
"step": 34500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 6.71,
|
| 918 |
+
"learning_rate": 0.00014898192987136932,
|
| 919 |
+
"loss": 0.7883,
|
| 920 |
+
"step": 34750
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 6.76,
|
| 924 |
+
"learning_rate": 0.000145027509716339,
|
| 925 |
+
"loss": 0.7865,
|
| 926 |
+
"step": 35000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 6.81,
|
| 930 |
+
"learning_rate": 0.0001411094814202753,
|
| 931 |
+
"loss": 0.7863,
|
| 932 |
+
"step": 35250
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 6.86,
|
| 936 |
+
"learning_rate": 0.00013722876504539635,
|
| 937 |
+
"loss": 0.7826,
|
| 938 |
+
"step": 35500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 6.91,
|
| 942 |
+
"learning_rate": 0.00013338627189204153,
|
| 943 |
+
"loss": 0.7829,
|
| 944 |
+
"step": 35750
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 6.95,
|
| 948 |
+
"learning_rate": 0.0001295829042846731,
|
| 949 |
+
"loss": 0.7816,
|
| 950 |
+
"step": 36000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 7.0,
|
| 954 |
+
"eval_alliteration_score": 0.44621513944223107,
|
| 955 |
+
"eval_harmonic_meter_score": 0.11710967524898745,
|
| 956 |
+
"eval_harmonic_rhyme_score": 0.5515777130298052,
|
| 957 |
+
"eval_meter_score": 0.33484803934701896,
|
| 958 |
+
"eval_rhyme_score": 0.8477574870320731,
|
| 959 |
+
"eval_runtime": 2122.0667,
|
| 960 |
+
"eval_samples_per_second": 1.272,
|
| 961 |
+
"eval_steps_per_second": 0.04,
|
| 962 |
+
"step": 36239
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 7.0,
|
| 966 |
+
"learning_rate": 0.00012581955535998448,
|
| 967 |
+
"loss": 0.7792,
|
| 968 |
+
"step": 36250
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 7.05,
|
| 972 |
+
"learning_rate": 0.0001220971088571674,
|
| 973 |
+
"loss": 0.734,
|
| 974 |
+
"step": 36500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 7.1,
|
| 978 |
+
"learning_rate": 0.00011841643891038518,
|
| 979 |
+
"loss": 0.7362,
|
| 980 |
+
"step": 36750
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 7.15,
|
| 984 |
+
"learning_rate": 0.00011477840984350193,
|
| 985 |
+
"loss": 0.7386,
|
| 986 |
+
"step": 37000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 7.2,
|
| 990 |
+
"learning_rate": 0.00011118387596711477,
|
| 991 |
+
"loss": 0.7384,
|
| 992 |
+
"step": 37250
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 7.24,
|
| 996 |
+
"learning_rate": 0.00010763368137793809,
|
| 997 |
+
"loss": 0.7364,
|
| 998 |
+
"step": 37500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 7.29,
|
| 1002 |
+
"learning_rate": 0.00010412865976058613,
|
| 1003 |
+
"loss": 0.7367,
|
| 1004 |
+
"step": 37750
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 7.34,
|
| 1008 |
+
"learning_rate": 0.00010066963419180093,
|
| 1009 |
+
"loss": 0.7365,
|
| 1010 |
+
"step": 38000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 7.39,
|
| 1014 |
+
"learning_rate": 9.725741694717035e-05,
|
| 1015 |
+
"loss": 0.7343,
|
| 1016 |
+
"step": 38250
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 7.44,
|
| 1020 |
+
"learning_rate": 9.389280931038336e-05,
|
| 1021 |
+
"loss": 0.737,
|
| 1022 |
+
"step": 38500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 7.48,
|
| 1026 |
+
"learning_rate": 9.057660138506682e-05,
|
| 1027 |
+
"loss": 0.7345,
|
| 1028 |
+
"step": 38750
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 7.53,
|
| 1032 |
+
"learning_rate": 8.730957190924632e-05,
|
| 1033 |
+
"loss": 0.7332,
|
| 1034 |
+
"step": 39000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 7.58,
|
| 1038 |
+
"learning_rate": 8.409248807247727e-05,
|
| 1039 |
+
"loss": 0.7335,
|
| 1040 |
+
"step": 39250
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 7.63,
|
| 1044 |
+
"learning_rate": 8.092610533568725e-05,
|
| 1045 |
+
"loss": 0.7302,
|
| 1046 |
+
"step": 39500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 7.68,
|
| 1050 |
+
"learning_rate": 7.781116725377309e-05,
|
| 1051 |
+
"loss": 0.7291,
|
| 1052 |
+
"step": 39750
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 7.73,
|
| 1056 |
+
"learning_rate": 7.474840530099277e-05,
|
| 1057 |
+
"loss": 0.7303,
|
| 1058 |
+
"step": 40000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 7.77,
|
| 1062 |
+
"learning_rate": 7.173853869919559e-05,
|
| 1063 |
+
"loss": 0.7288,
|
| 1064 |
+
"step": 40250
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 7.82,
|
| 1068 |
+
"learning_rate": 6.878227424892822e-05,
|
| 1069 |
+
"loss": 0.7283,
|
| 1070 |
+
"step": 40500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 7.87,
|
| 1074 |
+
"learning_rate": 6.588030616345898e-05,
|
| 1075 |
+
"loss": 0.726,
|
| 1076 |
+
"step": 40750
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 7.92,
|
| 1080 |
+
"learning_rate": 6.303331590575642e-05,
|
| 1081 |
+
"loss": 0.7246,
|
| 1082 |
+
"step": 41000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 7.97,
|
| 1086 |
+
"learning_rate": 6.0241972028463316e-05,
|
| 1087 |
+
"loss": 0.7243,
|
| 1088 |
+
"step": 41250
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 8.0,
|
| 1092 |
+
"eval_alliteration_score": 0.43950617283950616,
|
| 1093 |
+
"eval_harmonic_meter_score": 0.11078968787875117,
|
| 1094 |
+
"eval_harmonic_rhyme_score": 0.5599690603568903,
|
| 1095 |
+
"eval_meter_score": 0.3230506264598415,
|
| 1096 |
+
"eval_rhyme_score": 0.850570153117891,
|
| 1097 |
+
"eval_runtime": 2165.4131,
|
| 1098 |
+
"eval_samples_per_second": 1.247,
|
| 1099 |
+
"eval_steps_per_second": 0.039,
|
| 1100 |
+
"step": 41416
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 8.02,
|
| 1104 |
+
"learning_rate": 5.7506930016901755e-05,
|
| 1105 |
+
"loss": 0.7133,
|
| 1106 |
+
"step": 41500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 8.06,
|
| 1110 |
+
"learning_rate": 5.4828832135146994e-05,
|
| 1111 |
+
"loss": 0.6873,
|
| 1112 |
+
"step": 41750
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 8.11,
|
| 1116 |
+
"learning_rate": 5.2208307275205774e-05,
|
| 1117 |
+
"loss": 0.6887,
|
| 1118 |
+
"step": 42000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 8.16,
|
| 1122 |
+
"learning_rate": 4.9645970809335146e-05,
|
| 1123 |
+
"loss": 0.6883,
|
| 1124 |
+
"step": 42250
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 8.21,
|
| 1128 |
+
"learning_rate": 4.7142424445535695e-05,
|
| 1129 |
+
"loss": 0.6887,
|
| 1130 |
+
"step": 42500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 8.26,
|
| 1134 |
+
"learning_rate": 4.4698256086254156e-05,
|
| 1135 |
+
"loss": 0.688,
|
| 1136 |
+
"step": 42750
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 8.31,
|
| 1140 |
+
"learning_rate": 4.231403969032698e-05,
|
| 1141 |
+
"loss": 0.6895,
|
| 1142 |
+
"step": 43000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 8.35,
|
| 1146 |
+
"learning_rate": 3.999033513819922e-05,
|
| 1147 |
+
"loss": 0.6901,
|
| 1148 |
+
"step": 43250
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 8.4,
|
| 1152 |
+
"learning_rate": 3.772768810044874e-05,
|
| 1153 |
+
"loss": 0.6893,
|
| 1154 |
+
"step": 43500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 8.45,
|
| 1158 |
+
"learning_rate": 3.552662990964793e-05,
|
| 1159 |
+
"loss": 0.6873,
|
| 1160 |
+
"step": 43750
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 8.5,
|
| 1164 |
+
"learning_rate": 3.338767743559162e-05,
|
| 1165 |
+
"loss": 0.6895,
|
| 1166 |
+
"step": 44000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 8.55,
|
| 1170 |
+
"learning_rate": 3.131133296392159e-05,
|
| 1171 |
+
"loss": 0.6868,
|
| 1172 |
+
"step": 44250
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 8.6,
|
| 1176 |
+
"learning_rate": 2.929808407817651e-05,
|
| 1177 |
+
"loss": 0.6875,
|
| 1178 |
+
"step": 44500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 8.64,
|
| 1182 |
+
"learning_rate": 2.734840354529305e-05,
|
| 1183 |
+
"loss": 0.687,
|
| 1184 |
+
"step": 44750
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 8.69,
|
| 1188 |
+
"learning_rate": 2.5462749204587507e-05,
|
| 1189 |
+
"loss": 0.6864,
|
| 1190 |
+
"step": 45000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 8.74,
|
| 1194 |
+
"learning_rate": 2.3641563860241965e-05,
|
| 1195 |
+
"loss": 0.6841,
|
| 1196 |
+
"step": 45250
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 8.79,
|
| 1200 |
+
"learning_rate": 2.1885275177322048e-05,
|
| 1201 |
+
"loss": 0.6865,
|
| 1202 |
+
"step": 45500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 8.84,
|
| 1206 |
+
"learning_rate": 2.019429558134873e-05,
|
| 1207 |
+
"loss": 0.685,
|
| 1208 |
+
"step": 45750
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 8.89,
|
| 1212 |
+
"learning_rate": 1.856902216144962e-05,
|
| 1213 |
+
"loss": 0.6834,
|
| 1214 |
+
"step": 46000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 8.93,
|
| 1218 |
+
"learning_rate": 1.7009836577111302e-05,
|
| 1219 |
+
"loss": 0.6827,
|
| 1220 |
+
"step": 46250
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 8.98,
|
| 1224 |
+
"learning_rate": 1.551710496855515e-05,
|
| 1225 |
+
"loss": 0.6828,
|
| 1226 |
+
"step": 46500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 9.0,
|
| 1230 |
+
"eval_alliteration_score": 0.44789762340036565,
|
| 1231 |
+
"eval_harmonic_meter_score": 0.11082478811479961,
|
| 1232 |
+
"eval_harmonic_rhyme_score": 0.5453593958718314,
|
| 1233 |
+
"eval_meter_score": 0.328679747890422,
|
| 1234 |
+
"eval_rhyme_score": 0.8485504968592449,
|
| 1235 |
+
"eval_runtime": 2181.3379,
|
| 1236 |
+
"eval_samples_per_second": 1.238,
|
| 1237 |
+
"eval_steps_per_second": 0.039,
|
| 1238 |
+
"step": 46593
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 9.03,
|
| 1242 |
+
"learning_rate": 1.4091177870757209e-05,
|
| 1243 |
+
"loss": 0.673,
|
| 1244 |
+
"step": 46750
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 9.08,
|
| 1248 |
+
"learning_rate": 1.2732390131132907e-05,
|
| 1249 |
+
"loss": 0.6644,
|
| 1250 |
+
"step": 47000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 9.13,
|
| 1254 |
+
"learning_rate": 1.1441060830905591e-05,
|
| 1255 |
+
"loss": 0.6632,
|
| 1256 |
+
"step": 47250
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 9.18,
|
| 1260 |
+
"learning_rate": 1.0217493210177418e-05,
|
| 1261 |
+
"loss": 0.6641,
|
| 1262 |
+
"step": 47500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 9.22,
|
| 1266 |
+
"learning_rate": 9.061974596719934e-06,
|
| 1267 |
+
"loss": 0.6654,
|
| 1268 |
+
"step": 47750
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 9.27,
|
| 1272 |
+
"learning_rate": 7.974776338501631e-06,
|
| 1273 |
+
"loss": 0.6633,
|
| 1274 |
+
"step": 48000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 9.32,
|
| 1278 |
+
"learning_rate": 6.956153739967863e-06,
|
| 1279 |
+
"loss": 0.6656,
|
| 1280 |
+
"step": 48250
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 9.37,
|
| 1284 |
+
"learning_rate": 6.0063460020883915e-06,
|
| 1285 |
+
"loss": 0.664,
|
| 1286 |
+
"step": 48500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 9.42,
|
| 1290 |
+
"learning_rate": 5.125576166185996e-06,
|
| 1291 |
+
"loss": 0.6638,
|
| 1292 |
+
"step": 48750
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 9.46,
|
| 1296 |
+
"learning_rate": 4.314051061560497e-06,
|
| 1297 |
+
"loss": 0.6645,
|
| 1298 |
+
"step": 49000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 9.51,
|
| 1302 |
+
"learning_rate": 3.571961256919276e-06,
|
| 1303 |
+
"loss": 0.6624,
|
| 1304 |
+
"step": 49250
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 9.56,
|
| 1308 |
+
"learning_rate": 2.8994810156265035e-06,
|
| 1309 |
+
"loss": 0.663,
|
| 1310 |
+
"step": 49500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 9.61,
|
| 1314 |
+
"learning_rate": 2.2967682547812782e-06,
|
| 1315 |
+
"loss": 0.6631,
|
| 1316 |
+
"step": 49750
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 9.66,
|
| 1320 |
+
"learning_rate": 1.7639645081341524e-06,
|
| 1321 |
+
"loss": 0.663,
|
| 1322 |
+
"step": 50000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 9.71,
|
| 1326 |
+
"learning_rate": 1.3011948928511873e-06,
|
| 1327 |
+
"loss": 0.6641,
|
| 1328 |
+
"step": 50250
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 9.75,
|
| 1332 |
+
"learning_rate": 9.085680801330208e-07,
|
| 1333 |
+
"loss": 0.6634,
|
| 1334 |
+
"step": 50500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 9.8,
|
| 1338 |
+
"learning_rate": 5.861762696956151e-07,
|
| 1339 |
+
"loss": 0.6629,
|
| 1340 |
+
"step": 50750
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 9.85,
|
| 1344 |
+
"learning_rate": 3.340951681194082e-07,
|
| 1345 |
+
"loss": 0.6652,
|
| 1346 |
+
"step": 51000
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 9.9,
|
| 1350 |
+
"learning_rate": 1.523839710711683e-07,
|
| 1351 |
+
"loss": 0.663,
|
| 1352 |
+
"step": 51250
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 9.95,
|
| 1356 |
+
"learning_rate": 4.108534940331365e-08,
|
| 1357 |
+
"loss": 0.6623,
|
| 1358 |
+
"step": 51500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 10.0,
|
| 1362 |
+
"learning_rate": 2.2543913346106945e-10,
|
| 1363 |
+
"loss": 0.6634,
|
| 1364 |
+
"step": 51750
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 10.0,
|
| 1368 |
+
"eval_alliteration_score": 0.4406211936662607,
|
| 1369 |
+
"eval_harmonic_meter_score": 0.11891243404089967,
|
| 1370 |
+
"eval_harmonic_rhyme_score": 0.5297086635780841,
|
| 1371 |
+
"eval_meter_score": 0.33801080805418304,
|
| 1372 |
+
"eval_rhyme_score": 0.8432321768961265,
|
| 1373 |
+
"eval_runtime": 2127.802,
|
| 1374 |
+
"eval_samples_per_second": 1.269,
|
| 1375 |
+
"eval_steps_per_second": 0.04,
|
| 1376 |
+
"step": 51770
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 10.0,
|
| 1380 |
+
"step": 51770,
|
| 1381 |
+
"total_flos": 1.310953664443056e+18,
|
| 1382 |
+
"train_loss": 0.8821613311353097,
|
| 1383 |
+
"train_runtime": 68769.7958,
|
| 1384 |
+
"train_samples_per_second": 96.366,
|
| 1385 |
+
"train_steps_per_second": 0.753
|
| 1386 |
+
}
|
| 1387 |
+
],
|
| 1388 |
+
"max_steps": 51770,
|
| 1389 |
+
"num_train_epochs": 10,
|
| 1390 |
+
"total_flos": 1.310953664443056e+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 sha256:4528ef5b5a5280b2ed2ff9b76b089cd25d30609a5a2b811dd5d223f0ac6bfb8f
|
| 3 |
+
size 3439
|