Hugging Face's logo

shubhexists
/
sql-joins

Transformers
Safetensors
t5
text2text-generation
text-generation-inference
Model card Files
xet
Community
shubhexists commited on
Commit
eb11bb7
·
verified ·
1 Parent(s): e9c563a

added config files

Browse files
Files changed (7) hide show
  1. config.json +31 -0
  2. optimizer.pt +3 -0
  3. special_tokens_map.json +1 -0
  4. tokenizer.json +0 -0
  5. tokenizer_config.json +1 -0
  6. trainer_state.json +4327 -0
  7. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/safekeeping/t5-1.1.lm100k/t5.1.1.lm100k.large/pytorch_model/",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 2816,
7
+ "d_kv": 64,
8
+ "d_model": 1024,
9
+ "decoder_start_token_id": 0,
10
+ "diversity_penalty": null,
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_encoder_decoder": true,
17
+ "layer_norm_epsilon": 1e-06,
18
+ "max_length": 512,
19
+ "model_type": "t5",
20
+ "num_beams": 4,
21
+ "num_decoder_layers": 24,
22
+ "num_heads": 16,
23
+ "num_layers": 24,
24
+ "output_past": true,
25
+ "pad_token_id": 0,
26
+ "relative_attention_num_buckets": 32,
27
+ "tie_word_embeddings": false,
28
+ "transformers_version": "4.8.2",
29
+ "use_cache": true,
30
+ "vocab_size": 32102
31
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:56251e0eddf14abd107689c4b5e4146054329e8909c20d9f4095485a2419fb6b
3
+ size 5806399
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"]}
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 100, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"], "model_max_length": 512, "name_or_path": "/safekeeping/t5-1.1.lm100k/t5.1.1.lm100k.large/pytorch_model/", "special_tokens_map_file": "/safekeeping/t5-1.1.lm100k/t5.1.1.lm100k.large/pytorch_model/special_tokens_map.json", "sp_model_kwargs": {}, "tokenizer_class": "T5Tokenizer"}
trainer_state.json ADDED
@@ -0,0 +1,4327 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7117988394584139,
3
+ "best_model_checkpoint": "/experiments/spider/d6368230-e8a6-4e57-923c-f0938a9bcf7b/checkpoint-2624",
4
+ "epoch": 874.5844748858448,
5
+ "global_step": 2624,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.29,
12
+ "learning_rate": 0.0001,
13
+ "loss": 3.2183,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 1.29,
18
+ "learning_rate": 0.0001,
19
+ "loss": 2.451,
20
+ "step": 4
21
+ },
22
+ {
23
+ "epoch": 2.58,
24
+ "learning_rate": 0.0001,
25
+ "loss": 1.2943,
26
+ "step": 8
27
+ },
28
+ {
29
+ "epoch": 3.88,
30
+ "learning_rate": 0.0001,
31
+ "loss": 0.8542,
32
+ "step": 12
33
+ },
34
+ {
35
+ "epoch": 5.29,
36
+ "learning_rate": 0.0001,
37
+ "loss": 0.7239,
38
+ "step": 16
39
+ },
40
+ {
41
+ "epoch": 6.58,
42
+ "learning_rate": 0.0001,
43
+ "loss": 0.5427,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 7.88,
48
+ "learning_rate": 0.0001,
49
+ "loss": 0.4532,
50
+ "step": 24
51
+ },
52
+ {
53
+ "epoch": 9.29,
54
+ "learning_rate": 0.0001,
55
+ "loss": 0.4275,
56
+ "step": 28
57
+ },
58
+ {
59
+ "epoch": 10.58,
60
+ "learning_rate": 0.0001,
61
+ "loss": 0.3646,
62
+ "step": 32
63
+ },
64
+ {
65
+ "epoch": 11.88,
66
+ "learning_rate": 0.0001,
67
+ "loss": 0.3155,
68
+ "step": 36
69
+ },
70
+ {
71
+ "epoch": 13.29,
72
+ "learning_rate": 0.0001,
73
+ "loss": 0.3455,
74
+ "step": 40
75
+ },
76
+ {
77
+ "epoch": 14.58,
78
+ "learning_rate": 0.0001,
79
+ "loss": 0.2804,
80
+ "step": 44
81
+ },
82
+ {
83
+ "epoch": 15.88,
84
+ "learning_rate": 0.0001,
85
+ "loss": 0.2518,
86
+ "step": 48
87
+ },
88
+ {
89
+ "epoch": 17.29,
90
+ "learning_rate": 0.0001,
91
+ "loss": 0.248,
92
+ "step": 52
93
+ },
94
+ {
95
+ "epoch": 18.58,
96
+ "learning_rate": 0.0001,
97
+ "loss": 0.2021,
98
+ "step": 56
99
+ },
100
+ {
101
+ "epoch": 19.88,
102
+ "learning_rate": 0.0001,
103
+ "loss": 0.212,
104
+ "step": 60
105
+ },
106
+ {
107
+ "epoch": 21.29,
108
+ "learning_rate": 0.0001,
109
+ "loss": 0.203,
110
+ "step": 64
111
+ },
112
+ {
113
+ "epoch": 21.29,
114
+ "eval_exact_match": 0.42359767891682787,
115
+ "eval_exec": 0.4284332688588008,
116
+ "eval_loss": 0.1833169162273407,
117
+ "eval_runtime": 682.1232,
118
+ "eval_samples_per_second": 1.516,
119
+ "step": 64
120
+ },
121
+ {
122
+ "epoch": 22.58,
123
+ "learning_rate": 0.0001,
124
+ "loss": 0.1895,
125
+ "step": 68
126
+ },
127
+ {
128
+ "epoch": 23.88,
129
+ "learning_rate": 0.0001,
130
+ "loss": 0.166,
131
+ "step": 72
132
+ },
133
+ {
134
+ "epoch": 25.29,
135
+ "learning_rate": 0.0001,
136
+ "loss": 0.1833,
137
+ "step": 76
138
+ },
139
+ {
140
+ "epoch": 26.58,
141
+ "learning_rate": 0.0001,
142
+ "loss": 0.1503,
143
+ "step": 80
144
+ },
145
+ {
146
+ "epoch": 27.88,
147
+ "learning_rate": 0.0001,
148
+ "loss": 0.1485,
149
+ "step": 84
150
+ },
151
+ {
152
+ "epoch": 29.29,
153
+ "learning_rate": 0.0001,
154
+ "loss": 0.1574,
155
+ "step": 88
156
+ },
157
+ {
158
+ "epoch": 30.58,
159
+ "learning_rate": 0.0001,
160
+ "loss": 0.1295,
161
+ "step": 92
162
+ },
163
+ {
164
+ "epoch": 31.88,
165
+ "learning_rate": 0.0001,
166
+ "loss": 0.1199,
167
+ "step": 96
168
+ },
169
+ {
170
+ "epoch": 33.29,
171
+ "learning_rate": 0.0001,
172
+ "loss": 0.1262,
173
+ "step": 100
174
+ },
175
+ {
176
+ "epoch": 34.58,
177
+ "learning_rate": 0.0001,
178
+ "loss": 0.1153,
179
+ "step": 104
180
+ },
181
+ {
182
+ "epoch": 35.88,
183
+ "learning_rate": 0.0001,
184
+ "loss": 0.1067,
185
+ "step": 108
186
+ },
187
+ {
188
+ "epoch": 37.29,
189
+ "learning_rate": 0.0001,
190
+ "loss": 0.1086,
191
+ "step": 112
192
+ },
193
+ {
194
+ "epoch": 38.58,
195
+ "learning_rate": 0.0001,
196
+ "loss": 0.0934,
197
+ "step": 116
198
+ },
199
+ {
200
+ "epoch": 39.88,
201
+ "learning_rate": 0.0001,
202
+ "loss": 0.096,
203
+ "step": 120
204
+ },
205
+ {
206
+ "epoch": 41.29,
207
+ "learning_rate": 0.0001,
208
+ "loss": 0.1075,
209
+ "step": 124
210
+ },
211
+ {
212
+ "epoch": 42.58,
213
+ "learning_rate": 0.0001,
214
+ "loss": 0.0984,
215
+ "step": 128
216
+ },
217
+ {
218
+ "epoch": 42.58,
219
+ "eval_exact_match": 0.5938104448742747,
220
+ "eval_exec": 0.5764023210831721,
221
+ "eval_loss": 0.16683875024318695,
222
+ "eval_runtime": 703.3328,
223
+ "eval_samples_per_second": 1.47,
224
+ "step": 128
225
+ },
226
+ {
227
+ "epoch": 43.88,
228
+ "learning_rate": 0.0001,
229
+ "loss": 0.0842,
230
+ "step": 132
231
+ },
232
+ {
233
+ "epoch": 45.29,
234
+ "learning_rate": 0.0001,
235
+ "loss": 0.0954,
236
+ "step": 136
237
+ },
238
+ {
239
+ "epoch": 46.58,
240
+ "learning_rate": 0.0001,
241
+ "loss": 0.0842,
242
+ "step": 140
243
+ },
244
+ {
245
+ "epoch": 47.88,
246
+ "learning_rate": 0.0001,
247
+ "loss": 0.0724,
248
+ "step": 144
249
+ },
250
+ {
251
+ "epoch": 49.29,
252
+ "learning_rate": 0.0001,
253
+ "loss": 0.0805,
254
+ "step": 148
255
+ },
256
+ {
257
+ "epoch": 50.58,
258
+ "learning_rate": 0.0001,
259
+ "loss": 0.0745,
260
+ "step": 152
261
+ },
262
+ {
263
+ "epoch": 51.88,
264
+ "learning_rate": 0.0001,
265
+ "loss": 0.0685,
266
+ "step": 156
267
+ },
268
+ {
269
+ "epoch": 53.29,
270
+ "learning_rate": 0.0001,
271
+ "loss": 0.0709,
272
+ "step": 160
273
+ },
274
+ {
275
+ "epoch": 54.58,
276
+ "learning_rate": 0.0001,
277
+ "loss": 0.0744,
278
+ "step": 164
279
+ },
280
+ {
281
+ "epoch": 55.88,
282
+ "learning_rate": 0.0001,
283
+ "loss": 0.0603,
284
+ "step": 168
285
+ },
286
+ {
287
+ "epoch": 57.29,
288
+ "learning_rate": 0.0001,
289
+ "loss": 0.0607,
290
+ "step": 172
291
+ },
292
+ {
293
+ "epoch": 58.58,
294
+ "learning_rate": 0.0001,
295
+ "loss": 0.0652,
296
+ "step": 176
297
+ },
298
+ {
299
+ "epoch": 59.88,
300
+ "learning_rate": 0.0001,
301
+ "loss": 0.0592,
302
+ "step": 180
303
+ },
304
+ {
305
+ "epoch": 61.29,
306
+ "learning_rate": 0.0001,
307
+ "loss": 0.0584,
308
+ "step": 184
309
+ },
310
+ {
311
+ "epoch": 62.58,
312
+ "learning_rate": 0.0001,
313
+ "loss": 0.0507,
314
+ "step": 188
315
+ },
316
+ {
317
+ "epoch": 63.88,
318
+ "learning_rate": 0.0001,
319
+ "loss": 0.0523,
320
+ "step": 192
321
+ },
322
+ {
323
+ "epoch": 63.88,
324
+ "eval_exact_match": 0.6266924564796905,
325
+ "eval_exec": 0.620889748549323,
326
+ "eval_loss": 0.16735364496707916,
327
+ "eval_runtime": 843.3561,
328
+ "eval_samples_per_second": 1.226,
329
+ "step": 192
330
+ },
331
+ {
332
+ "epoch": 65.29,
333
+ "learning_rate": 0.0001,
334
+ "loss": 0.0576,
335
+ "step": 196
336
+ },
337
+ {
338
+ "epoch": 66.58,
339
+ "learning_rate": 0.0001,
340
+ "loss": 0.0539,
341
+ "step": 200
342
+ },
343
+ {
344
+ "epoch": 67.88,
345
+ "learning_rate": 0.0001,
346
+ "loss": 0.0443,
347
+ "step": 204
348
+ },
349
+ {
350
+ "epoch": 69.29,
351
+ "learning_rate": 0.0001,
352
+ "loss": 0.0518,
353
+ "step": 208
354
+ },
355
+ {
356
+ "epoch": 70.58,
357
+ "learning_rate": 0.0001,
358
+ "loss": 0.0435,
359
+ "step": 212
360
+ },
361
+ {
362
+ "epoch": 71.88,
363
+ "learning_rate": 0.0001,
364
+ "loss": 0.0428,
365
+ "step": 216
366
+ },
367
+ {
368
+ "epoch": 73.29,
369
+ "learning_rate": 0.0001,
370
+ "loss": 0.043,
371
+ "step": 220
372
+ },
373
+ {
374
+ "epoch": 74.58,
375
+ "learning_rate": 0.0001,
376
+ "loss": 0.0383,
377
+ "step": 224
378
+ },
379
+ {
380
+ "epoch": 75.88,
381
+ "learning_rate": 0.0001,
382
+ "loss": 0.0399,
383
+ "step": 228
384
+ },
385
+ {
386
+ "epoch": 77.29,
387
+ "learning_rate": 0.0001,
388
+ "loss": 0.0422,
389
+ "step": 232
390
+ },
391
+ {
392
+ "epoch": 78.58,
393
+ "learning_rate": 0.0001,
394
+ "loss": 0.0357,
395
+ "step": 236
396
+ },
397
+ {
398
+ "epoch": 79.88,
399
+ "learning_rate": 0.0001,
400
+ "loss": 0.0346,
401
+ "step": 240
402
+ },
403
+ {
404
+ "epoch": 81.29,
405
+ "learning_rate": 0.0001,
406
+ "loss": 0.0433,
407
+ "step": 244
408
+ },
409
+ {
410
+ "epoch": 82.58,
411
+ "learning_rate": 0.0001,
412
+ "loss": 0.0339,
413
+ "step": 248
414
+ },
415
+ {
416
+ "epoch": 83.88,
417
+ "learning_rate": 0.0001,
418
+ "loss": 0.0353,
419
+ "step": 252
420
+ },
421
+ {
422
+ "epoch": 85.29,
423
+ "learning_rate": 0.0001,
424
+ "loss": 0.0326,
425
+ "step": 256
426
+ },
427
+ {
428
+ "epoch": 85.29,
429
+ "eval_exact_match": 0.6586073500967118,
430
+ "eval_exec": 0.6450676982591876,
431
+ "eval_loss": 0.21498563885688782,
432
+ "eval_runtime": 824.5584,
433
+ "eval_samples_per_second": 1.254,
434
+ "step": 256
435
+ },
436
+ {
437
+ "epoch": 86.58,
438
+ "learning_rate": 0.0001,
439
+ "loss": 0.0301,
440
+ "step": 260
441
+ },
442
+ {
443
+ "epoch": 87.88,
444
+ "learning_rate": 0.0001,
445
+ "loss": 0.0302,
446
+ "step": 264
447
+ },
448
+ {
449
+ "epoch": 89.29,
450
+ "learning_rate": 0.0001,
451
+ "loss": 0.0322,
452
+ "step": 268
453
+ },
454
+ {
455
+ "epoch": 90.58,
456
+ "learning_rate": 0.0001,
457
+ "loss": 0.0248,
458
+ "step": 272
459
+ },
460
+ {
461
+ "epoch": 91.88,
462
+ "learning_rate": 0.0001,
463
+ "loss": 0.026,
464
+ "step": 276
465
+ },
466
+ {
467
+ "epoch": 93.29,
468
+ "learning_rate": 0.0001,
469
+ "loss": 0.0297,
470
+ "step": 280
471
+ },
472
+ {
473
+ "epoch": 94.58,
474
+ "learning_rate": 0.0001,
475
+ "loss": 0.0276,
476
+ "step": 284
477
+ },
478
+ {
479
+ "epoch": 95.88,
480
+ "learning_rate": 0.0001,
481
+ "loss": 0.0253,
482
+ "step": 288
483
+ },
484
+ {
485
+ "epoch": 97.29,
486
+ "learning_rate": 0.0001,
487
+ "loss": 0.0282,
488
+ "step": 292
489
+ },
490
+ {
491
+ "epoch": 98.58,
492
+ "learning_rate": 0.0001,
493
+ "loss": 0.0252,
494
+ "step": 296
495
+ },
496
+ {
497
+ "epoch": 99.88,
498
+ "learning_rate": 0.0001,
499
+ "loss": 0.0234,
500
+ "step": 300
501
+ },
502
+ {
503
+ "epoch": 101.29,
504
+ "learning_rate": 0.0001,
505
+ "loss": 0.0225,
506
+ "step": 304
507
+ },
508
+ {
509
+ "epoch": 102.58,
510
+ "learning_rate": 0.0001,
511
+ "loss": 0.0254,
512
+ "step": 308
513
+ },
514
+ {
515
+ "epoch": 103.88,
516
+ "learning_rate": 0.0001,
517
+ "loss": 0.0201,
518
+ "step": 312
519
+ },
520
+ {
521
+ "epoch": 105.29,
522
+ "learning_rate": 0.0001,
523
+ "loss": 0.022,
524
+ "step": 316
525
+ },
526
+ {
527
+ "epoch": 106.58,
528
+ "learning_rate": 0.0001,
529
+ "loss": 0.0206,
530
+ "step": 320
531
+ },
532
+ {
533
+ "epoch": 106.58,
534
+ "eval_exact_match": 0.6644100580270793,
535
+ "eval_exec": 0.6499032882011605,
536
+ "eval_loss": 0.21441736817359924,
537
+ "eval_runtime": 835.4598,
538
+ "eval_samples_per_second": 1.238,
539
+ "step": 320
540
+ },
541
+ {
542
+ "epoch": 107.88,
543
+ "learning_rate": 0.0001,
544
+ "loss": 0.0178,
545
+ "step": 324
546
+ },
547
+ {
548
+ "epoch": 109.29,
549
+ "learning_rate": 0.0001,
550
+ "loss": 0.0172,
551
+ "step": 328
552
+ },
553
+ {
554
+ "epoch": 110.58,
555
+ "learning_rate": 0.0001,
556
+ "loss": 0.0175,
557
+ "step": 332
558
+ },
559
+ {
560
+ "epoch": 111.88,
561
+ "learning_rate": 0.0001,
562
+ "loss": 0.0155,
563
+ "step": 336
564
+ },
565
+ {
566
+ "epoch": 113.29,
567
+ "learning_rate": 0.0001,
568
+ "loss": 0.0187,
569
+ "step": 340
570
+ },
571
+ {
572
+ "epoch": 114.58,
573
+ "learning_rate": 0.0001,
574
+ "loss": 0.0142,
575
+ "step": 344
576
+ },
577
+ {
578
+ "epoch": 115.88,
579
+ "learning_rate": 0.0001,
580
+ "loss": 0.0149,
581
+ "step": 348
582
+ },
583
+ {
584
+ "epoch": 117.29,
585
+ "learning_rate": 0.0001,
586
+ "loss": 0.0155,
587
+ "step": 352
588
+ },
589
+ {
590
+ "epoch": 118.58,
591
+ "learning_rate": 0.0001,
592
+ "loss": 0.0148,
593
+ "step": 356
594
+ },
595
+ {
596
+ "epoch": 119.88,
597
+ "learning_rate": 0.0001,
598
+ "loss": 0.0138,
599
+ "step": 360
600
+ },
601
+ {
602
+ "epoch": 121.29,
603
+ "learning_rate": 0.0001,
604
+ "loss": 0.0154,
605
+ "step": 364
606
+ },
607
+ {
608
+ "epoch": 122.58,
609
+ "learning_rate": 0.0001,
610
+ "loss": 0.0145,
611
+ "step": 368
612
+ },
613
+ {
614
+ "epoch": 123.88,
615
+ "learning_rate": 0.0001,
616
+ "loss": 0.0172,
617
+ "step": 372
618
+ },
619
+ {
620
+ "epoch": 125.29,
621
+ "learning_rate": 0.0001,
622
+ "loss": 0.0146,
623
+ "step": 376
624
+ },
625
+ {
626
+ "epoch": 126.58,
627
+ "learning_rate": 0.0001,
628
+ "loss": 0.0156,
629
+ "step": 380
630
+ },
631
+ {
632
+ "epoch": 127.88,
633
+ "learning_rate": 0.0001,
634
+ "loss": 0.0177,
635
+ "step": 384
636
+ },
637
+ {
638
+ "epoch": 127.88,
639
+ "eval_exact_match": 0.6682785299806576,
640
+ "eval_exec": 0.6518375241779497,
641
+ "eval_loss": 0.24249066412448883,
642
+ "eval_runtime": 822.9024,
643
+ "eval_samples_per_second": 1.257,
644
+ "step": 384
645
+ },
646
+ {
647
+ "epoch": 129.29,
648
+ "learning_rate": 0.0001,
649
+ "loss": 0.0134,
650
+ "step": 388
651
+ },
652
+ {
653
+ "epoch": 130.58,
654
+ "learning_rate": 0.0001,
655
+ "loss": 0.0119,
656
+ "step": 392
657
+ },
658
+ {
659
+ "epoch": 131.88,
660
+ "learning_rate": 0.0001,
661
+ "loss": 0.01,
662
+ "step": 396
663
+ },
664
+ {
665
+ "epoch": 133.29,
666
+ "learning_rate": 0.0001,
667
+ "loss": 0.0105,
668
+ "step": 400
669
+ },
670
+ {
671
+ "epoch": 134.58,
672
+ "learning_rate": 0.0001,
673
+ "loss": 0.0086,
674
+ "step": 404
675
+ },
676
+ {
677
+ "epoch": 135.88,
678
+ "learning_rate": 0.0001,
679
+ "loss": 0.0118,
680
+ "step": 408
681
+ },
682
+ {
683
+ "epoch": 137.29,
684
+ "learning_rate": 0.0001,
685
+ "loss": 0.01,
686
+ "step": 412
687
+ },
688
+ {
689
+ "epoch": 138.58,
690
+ "learning_rate": 0.0001,
691
+ "loss": 0.0102,
692
+ "step": 416
693
+ },
694
+ {
695
+ "epoch": 139.88,
696
+ "learning_rate": 0.0001,
697
+ "loss": 0.0106,
698
+ "step": 420
699
+ },
700
+ {
701
+ "epoch": 141.29,
702
+ "learning_rate": 0.0001,
703
+ "loss": 0.0087,
704
+ "step": 424
705
+ },
706
+ {
707
+ "epoch": 142.58,
708
+ "learning_rate": 0.0001,
709
+ "loss": 0.0089,
710
+ "step": 428
711
+ },
712
+ {
713
+ "epoch": 143.88,
714
+ "learning_rate": 0.0001,
715
+ "loss": 0.0099,
716
+ "step": 432
717
+ },
718
+ {
719
+ "epoch": 145.29,
720
+ "learning_rate": 0.0001,
721
+ "loss": 0.0092,
722
+ "step": 436
723
+ },
724
+ {
725
+ "epoch": 146.58,
726
+ "learning_rate": 0.0001,
727
+ "loss": 0.0081,
728
+ "step": 440
729
+ },
730
+ {
731
+ "epoch": 147.88,
732
+ "learning_rate": 0.0001,
733
+ "loss": 0.0088,
734
+ "step": 444
735
+ },
736
+ {
737
+ "epoch": 149.29,
738
+ "learning_rate": 0.0001,
739
+ "loss": 0.0079,
740
+ "step": 448
741
+ },
742
+ {
743
+ "epoch": 149.29,
744
+ "eval_exact_match": 0.6499032882011605,
745
+ "eval_exec": 0.6353965183752418,
746
+ "eval_loss": 0.2822052538394928,
747
+ "eval_runtime": 813.617,
748
+ "eval_samples_per_second": 1.271,
749
+ "step": 448
750
+ },
751
+ {
752
+ "epoch": 150.58,
753
+ "learning_rate": 0.0001,
754
+ "loss": 0.0077,
755
+ "step": 452
756
+ },
757
+ {
758
+ "epoch": 151.88,
759
+ "learning_rate": 0.0001,
760
+ "loss": 0.0075,
761
+ "step": 456
762
+ },
763
+ {
764
+ "epoch": 153.29,
765
+ "learning_rate": 0.0001,
766
+ "loss": 0.0071,
767
+ "step": 460
768
+ },
769
+ {
770
+ "epoch": 154.58,
771
+ "learning_rate": 0.0001,
772
+ "loss": 0.008,
773
+ "step": 464
774
+ },
775
+ {
776
+ "epoch": 155.88,
777
+ "learning_rate": 0.0001,
778
+ "loss": 0.0063,
779
+ "step": 468
780
+ },
781
+ {
782
+ "epoch": 157.29,
783
+ "learning_rate": 0.0001,
784
+ "loss": 0.0078,
785
+ "step": 472
786
+ },
787
+ {
788
+ "epoch": 158.58,
789
+ "learning_rate": 0.0001,
790
+ "loss": 0.0069,
791
+ "step": 476
792
+ },
793
+ {
794
+ "epoch": 159.88,
795
+ "learning_rate": 0.0001,
796
+ "loss": 0.0073,
797
+ "step": 480
798
+ },
799
+ {
800
+ "epoch": 161.29,
801
+ "learning_rate": 0.0001,
802
+ "loss": 0.0111,
803
+ "step": 484
804
+ },
805
+ {
806
+ "epoch": 162.58,
807
+ "learning_rate": 0.0001,
808
+ "loss": 0.0063,
809
+ "step": 488
810
+ },
811
+ {
812
+ "epoch": 163.88,
813
+ "learning_rate": 0.0001,
814
+ "loss": 0.0078,
815
+ "step": 492
816
+ },
817
+ {
818
+ "epoch": 165.29,
819
+ "learning_rate": 0.0001,
820
+ "loss": 0.0121,
821
+ "step": 496
822
+ },
823
+ {
824
+ "epoch": 166.58,
825
+ "learning_rate": 0.0001,
826
+ "loss": 0.0086,
827
+ "step": 500
828
+ },
829
+ {
830
+ "epoch": 167.88,
831
+ "learning_rate": 0.0001,
832
+ "loss": 0.0066,
833
+ "step": 504
834
+ },
835
+ {
836
+ "epoch": 169.29,
837
+ "learning_rate": 0.0001,
838
+ "loss": 0.0064,
839
+ "step": 508
840
+ },
841
+ {
842
+ "epoch": 170.58,
843
+ "learning_rate": 0.0001,
844
+ "loss": 0.0046,
845
+ "step": 512
846
+ },
847
+ {
848
+ "epoch": 170.58,
849
+ "eval_exact_match": 0.660541586073501,
850
+ "eval_exec": 0.6479690522243714,
851
+ "eval_loss": 0.29858535528182983,
852
+ "eval_runtime": 852.1163,
853
+ "eval_samples_per_second": 1.213,
854
+ "step": 512
855
+ },
856
+ {
857
+ "epoch": 171.88,
858
+ "learning_rate": 0.0001,
859
+ "loss": 0.0051,
860
+ "step": 516
861
+ },
862
+ {
863
+ "epoch": 173.29,
864
+ "learning_rate": 0.0001,
865
+ "loss": 0.0053,
866
+ "step": 520
867
+ },
868
+ {
869
+ "epoch": 174.58,
870
+ "learning_rate": 0.0001,
871
+ "loss": 0.0054,
872
+ "step": 524
873
+ },
874
+ {
875
+ "epoch": 175.88,
876
+ "learning_rate": 0.0001,
877
+ "loss": 0.0052,
878
+ "step": 528
879
+ },
880
+ {
881
+ "epoch": 177.29,
882
+ "learning_rate": 0.0001,
883
+ "loss": 0.0051,
884
+ "step": 532
885
+ },
886
+ {
887
+ "epoch": 178.58,
888
+ "learning_rate": 0.0001,
889
+ "loss": 0.0043,
890
+ "step": 536
891
+ },
892
+ {
893
+ "epoch": 179.88,
894
+ "learning_rate": 0.0001,
895
+ "loss": 0.0047,
896
+ "step": 540
897
+ },
898
+ {
899
+ "epoch": 181.29,
900
+ "learning_rate": 0.0001,
901
+ "loss": 0.0051,
902
+ "step": 544
903
+ },
904
+ {
905
+ "epoch": 182.58,
906
+ "learning_rate": 0.0001,
907
+ "loss": 0.0164,
908
+ "step": 548
909
+ },
910
+ {
911
+ "epoch": 183.88,
912
+ "learning_rate": 0.0001,
913
+ "loss": 0.0044,
914
+ "step": 552
915
+ },
916
+ {
917
+ "epoch": 185.29,
918
+ "learning_rate": 0.0001,
919
+ "loss": 0.0048,
920
+ "step": 556
921
+ },
922
+ {
923
+ "epoch": 186.58,
924
+ "learning_rate": 0.0001,
925
+ "loss": 0.0038,
926
+ "step": 560
927
+ },
928
+ {
929
+ "epoch": 187.88,
930
+ "learning_rate": 0.0001,
931
+ "loss": 0.0033,
932
+ "step": 564
933
+ },
934
+ {
935
+ "epoch": 189.29,
936
+ "learning_rate": 0.0001,
937
+ "loss": 0.0041,
938
+ "step": 568
939
+ },
940
+ {
941
+ "epoch": 190.58,
942
+ "learning_rate": 0.0001,
943
+ "loss": 0.0038,
944
+ "step": 572
945
+ },
946
+ {
947
+ "epoch": 191.88,
948
+ "learning_rate": 0.0001,
949
+ "loss": 0.0044,
950
+ "step": 576
951
+ },
952
+ {
953
+ "epoch": 191.88,
954
+ "eval_exact_match": 0.6663442940038685,
955
+ "eval_exec": 0.6479690522243714,
956
+ "eval_loss": 0.3111841380596161,
957
+ "eval_runtime": 837.4186,
958
+ "eval_samples_per_second": 1.235,
959
+ "step": 576
960
+ },
961
+ {
962
+ "epoch": 193.29,
963
+ "learning_rate": 0.0001,
964
+ "loss": 0.0045,
965
+ "step": 580
966
+ },
967
+ {
968
+ "epoch": 194.58,
969
+ "learning_rate": 0.0001,
970
+ "loss": 0.0027,
971
+ "step": 584
972
+ },
973
+ {
974
+ "epoch": 195.88,
975
+ "learning_rate": 0.0001,
976
+ "loss": 0.0024,
977
+ "step": 588
978
+ },
979
+ {
980
+ "epoch": 197.29,
981
+ "learning_rate": 0.0001,
982
+ "loss": 0.0034,
983
+ "step": 592
984
+ },
985
+ {
986
+ "epoch": 198.58,
987
+ "learning_rate": 0.0001,
988
+ "loss": 0.0037,
989
+ "step": 596
990
+ },
991
+ {
992
+ "epoch": 199.88,
993
+ "learning_rate": 0.0001,
994
+ "loss": 0.0037,
995
+ "step": 600
996
+ },
997
+ {
998
+ "epoch": 201.29,
999
+ "learning_rate": 0.0001,
1000
+ "loss": 0.0036,
1001
+ "step": 604
1002
+ },
1003
+ {
1004
+ "epoch": 202.58,
1005
+ "learning_rate": 0.0001,
1006
+ "loss": 0.0029,
1007
+ "step": 608
1008
+ },
1009
+ {
1010
+ "epoch": 203.88,
1011
+ "learning_rate": 0.0001,
1012
+ "loss": 0.0019,
1013
+ "step": 612
1014
+ },
1015
+ {
1016
+ "epoch": 205.29,
1017
+ "learning_rate": 0.0001,
1018
+ "loss": 0.0028,
1019
+ "step": 616
1020
+ },
1021
+ {
1022
+ "epoch": 206.58,
1023
+ "learning_rate": 0.0001,
1024
+ "loss": 0.0038,
1025
+ "step": 620
1026
+ },
1027
+ {
1028
+ "epoch": 207.88,
1029
+ "learning_rate": 0.0001,
1030
+ "loss": 0.003,
1031
+ "step": 624
1032
+ },
1033
+ {
1034
+ "epoch": 209.29,
1035
+ "learning_rate": 0.0001,
1036
+ "loss": 0.0033,
1037
+ "step": 628
1038
+ },
1039
+ {
1040
+ "epoch": 210.58,
1041
+ "learning_rate": 0.0001,
1042
+ "loss": 0.0028,
1043
+ "step": 632
1044
+ },
1045
+ {
1046
+ "epoch": 211.88,
1047
+ "learning_rate": 0.0001,
1048
+ "loss": 0.0028,
1049
+ "step": 636
1050
+ },
1051
+ {
1052
+ "epoch": 213.29,
1053
+ "learning_rate": 0.0001,
1054
+ "loss": 0.0025,
1055
+ "step": 640
1056
+ },
1057
+ {
1058
+ "epoch": 213.29,
1059
+ "eval_exact_match": 0.660541586073501,
1060
+ "eval_exec": 0.648936170212766,
1061
+ "eval_loss": 0.3557709753513336,
1062
+ "eval_runtime": 807.1499,
1063
+ "eval_samples_per_second": 1.281,
1064
+ "step": 640
1065
+ },
1066
+ {
1067
+ "epoch": 214.58,
1068
+ "learning_rate": 0.0001,
1069
+ "loss": 0.0024,
1070
+ "step": 644
1071
+ },
1072
+ {
1073
+ "epoch": 215.88,
1074
+ "learning_rate": 0.0001,
1075
+ "loss": 0.0034,
1076
+ "step": 648
1077
+ },
1078
+ {
1079
+ "epoch": 217.29,
1080
+ "learning_rate": 0.0001,
1081
+ "loss": 0.0023,
1082
+ "step": 652
1083
+ },
1084
+ {
1085
+ "epoch": 218.58,
1086
+ "learning_rate": 0.0001,
1087
+ "loss": 0.0021,
1088
+ "step": 656
1089
+ },
1090
+ {
1091
+ "epoch": 219.88,
1092
+ "learning_rate": 0.0001,
1093
+ "loss": 0.0026,
1094
+ "step": 660
1095
+ },
1096
+ {
1097
+ "epoch": 221.29,
1098
+ "learning_rate": 0.0001,
1099
+ "loss": 0.0035,
1100
+ "step": 664
1101
+ },
1102
+ {
1103
+ "epoch": 222.58,
1104
+ "learning_rate": 0.0001,
1105
+ "loss": 0.0022,
1106
+ "step": 668
1107
+ },
1108
+ {
1109
+ "epoch": 223.88,
1110
+ "learning_rate": 0.0001,
1111
+ "loss": 0.002,
1112
+ "step": 672
1113
+ },
1114
+ {
1115
+ "epoch": 225.29,
1116
+ "learning_rate": 0.0001,
1117
+ "loss": 0.0029,
1118
+ "step": 676
1119
+ },
1120
+ {
1121
+ "epoch": 226.58,
1122
+ "learning_rate": 0.0001,
1123
+ "loss": 0.0029,
1124
+ "step": 680
1125
+ },
1126
+ {
1127
+ "epoch": 227.88,
1128
+ "learning_rate": 0.0001,
1129
+ "loss": 0.0058,
1130
+ "step": 684
1131
+ },
1132
+ {
1133
+ "epoch": 229.29,
1134
+ "learning_rate": 0.0001,
1135
+ "loss": 0.0085,
1136
+ "step": 688
1137
+ },
1138
+ {
1139
+ "epoch": 230.58,
1140
+ "learning_rate": 0.0001,
1141
+ "loss": 0.0049,
1142
+ "step": 692
1143
+ },
1144
+ {
1145
+ "epoch": 231.88,
1146
+ "learning_rate": 0.0001,
1147
+ "loss": 0.0026,
1148
+ "step": 696
1149
+ },
1150
+ {
1151
+ "epoch": 233.29,
1152
+ "learning_rate": 0.0001,
1153
+ "loss": 0.0028,
1154
+ "step": 700
1155
+ },
1156
+ {
1157
+ "epoch": 234.58,
1158
+ "learning_rate": 0.0001,
1159
+ "loss": 0.0026,
1160
+ "step": 704
1161
+ },
1162
+ {
1163
+ "epoch": 234.58,
1164
+ "eval_exact_match": 0.6682785299806576,
1165
+ "eval_exec": 0.6528046421663443,
1166
+ "eval_loss": 0.34394705295562744,
1167
+ "eval_runtime": 890.2876,
1168
+ "eval_samples_per_second": 1.161,
1169
+ "step": 704
1170
+ },
1171
+ {
1172
+ "epoch": 235.88,
1173
+ "learning_rate": 0.0001,
1174
+ "loss": 0.0019,
1175
+ "step": 708
1176
+ },
1177
+ {
1178
+ "epoch": 237.29,
1179
+ "learning_rate": 0.0001,
1180
+ "loss": 0.003,
1181
+ "step": 712
1182
+ },
1183
+ {
1184
+ "epoch": 238.58,
1185
+ "learning_rate": 0.0001,
1186
+ "loss": 0.0027,
1187
+ "step": 716
1188
+ },
1189
+ {
1190
+ "epoch": 239.88,
1191
+ "learning_rate": 0.0001,
1192
+ "loss": 0.0021,
1193
+ "step": 720
1194
+ },
1195
+ {
1196
+ "epoch": 241.29,
1197
+ "learning_rate": 0.0001,
1198
+ "loss": 0.0023,
1199
+ "step": 724
1200
+ },
1201
+ {
1202
+ "epoch": 242.58,
1203
+ "learning_rate": 0.0001,
1204
+ "loss": 0.0017,
1205
+ "step": 728
1206
+ },
1207
+ {
1208
+ "epoch": 243.88,
1209
+ "learning_rate": 0.0001,
1210
+ "loss": 0.0018,
1211
+ "step": 732
1212
+ },
1213
+ {
1214
+ "epoch": 245.29,
1215
+ "learning_rate": 0.0001,
1216
+ "loss": 0.0028,
1217
+ "step": 736
1218
+ },
1219
+ {
1220
+ "epoch": 246.58,
1221
+ "learning_rate": 0.0001,
1222
+ "loss": 0.0058,
1223
+ "step": 740
1224
+ },
1225
+ {
1226
+ "epoch": 247.88,
1227
+ "learning_rate": 0.0001,
1228
+ "loss": 0.0021,
1229
+ "step": 744
1230
+ },
1231
+ {
1232
+ "epoch": 249.29,
1233
+ "learning_rate": 0.0001,
1234
+ "loss": 0.0023,
1235
+ "step": 748
1236
+ },
1237
+ {
1238
+ "epoch": 250.58,
1239
+ "learning_rate": 0.0001,
1240
+ "loss": 0.003,
1241
+ "step": 752
1242
+ },
1243
+ {
1244
+ "epoch": 251.88,
1245
+ "learning_rate": 0.0001,
1246
+ "loss": 0.0027,
1247
+ "step": 756
1248
+ },
1249
+ {
1250
+ "epoch": 253.29,
1251
+ "learning_rate": 0.0001,
1252
+ "loss": 0.0029,
1253
+ "step": 760
1254
+ },
1255
+ {
1256
+ "epoch": 254.58,
1257
+ "learning_rate": 0.0001,
1258
+ "loss": 0.0023,
1259
+ "step": 764
1260
+ },
1261
+ {
1262
+ "epoch": 255.88,
1263
+ "learning_rate": 0.0001,
1264
+ "loss": 0.0017,
1265
+ "step": 768
1266
+ },
1267
+ {
1268
+ "epoch": 255.88,
1269
+ "eval_exact_match": 0.6702127659574468,
1270
+ "eval_exec": 0.6576402321083172,
1271
+ "eval_loss": 0.3505505919456482,
1272
+ "eval_runtime": 849.5769,
1273
+ "eval_samples_per_second": 1.217,
1274
+ "step": 768
1275
+ },
1276
+ {
1277
+ "epoch": 257.29,
1278
+ "learning_rate": 0.0001,
1279
+ "loss": 0.0014,
1280
+ "step": 772
1281
+ },
1282
+ {
1283
+ "epoch": 258.58,
1284
+ "learning_rate": 0.0001,
1285
+ "loss": 0.0017,
1286
+ "step": 776
1287
+ },
1288
+ {
1289
+ "epoch": 259.88,
1290
+ "learning_rate": 0.0001,
1291
+ "loss": 0.0023,
1292
+ "step": 780
1293
+ },
1294
+ {
1295
+ "epoch": 261.29,
1296
+ "learning_rate": 0.0001,
1297
+ "loss": 0.0031,
1298
+ "step": 784
1299
+ },
1300
+ {
1301
+ "epoch": 262.58,
1302
+ "learning_rate": 0.0001,
1303
+ "loss": 0.0015,
1304
+ "step": 788
1305
+ },
1306
+ {
1307
+ "epoch": 263.88,
1308
+ "learning_rate": 0.0001,
1309
+ "loss": 0.0018,
1310
+ "step": 792
1311
+ },
1312
+ {
1313
+ "epoch": 265.29,
1314
+ "learning_rate": 0.0001,
1315
+ "loss": 0.0019,
1316
+ "step": 796
1317
+ },
1318
+ {
1319
+ "epoch": 266.58,
1320
+ "learning_rate": 0.0001,
1321
+ "loss": 0.0017,
1322
+ "step": 800
1323
+ },
1324
+ {
1325
+ "epoch": 267.88,
1326
+ "learning_rate": 0.0001,
1327
+ "loss": 0.0019,
1328
+ "step": 804
1329
+ },
1330
+ {
1331
+ "epoch": 269.29,
1332
+ "learning_rate": 0.0001,
1333
+ "loss": 0.0018,
1334
+ "step": 808
1335
+ },
1336
+ {
1337
+ "epoch": 270.58,
1338
+ "learning_rate": 0.0001,
1339
+ "loss": 0.0021,
1340
+ "step": 812
1341
+ },
1342
+ {
1343
+ "epoch": 271.88,
1344
+ "learning_rate": 0.0001,
1345
+ "loss": 0.0022,
1346
+ "step": 816
1347
+ },
1348
+ {
1349
+ "epoch": 273.29,
1350
+ "learning_rate": 0.0001,
1351
+ "loss": 0.0013,
1352
+ "step": 820
1353
+ },
1354
+ {
1355
+ "epoch": 274.58,
1356
+ "learning_rate": 0.0001,
1357
+ "loss": 0.0016,
1358
+ "step": 824
1359
+ },
1360
+ {
1361
+ "epoch": 275.88,
1362
+ "learning_rate": 0.0001,
1363
+ "loss": 0.0016,
1364
+ "step": 828
1365
+ },
1366
+ {
1367
+ "epoch": 277.29,
1368
+ "learning_rate": 0.0001,
1369
+ "loss": 0.0017,
1370
+ "step": 832
1371
+ },
1372
+ {
1373
+ "epoch": 277.29,
1374
+ "eval_exact_match": 0.6789168278529981,
1375
+ "eval_exec": 0.6615087040618955,
1376
+ "eval_loss": 0.36397507786750793,
1377
+ "eval_runtime": 827.3744,
1378
+ "eval_samples_per_second": 1.25,
1379
+ "step": 832
1380
+ },
1381
+ {
1382
+ "epoch": 278.58,
1383
+ "learning_rate": 0.0001,
1384
+ "loss": 0.0016,
1385
+ "step": 836
1386
+ },
1387
+ {
1388
+ "epoch": 279.88,
1389
+ "learning_rate": 0.0001,
1390
+ "loss": 0.0019,
1391
+ "step": 840
1392
+ },
1393
+ {
1394
+ "epoch": 281.29,
1395
+ "learning_rate": 0.0001,
1396
+ "loss": 0.002,
1397
+ "step": 844
1398
+ },
1399
+ {
1400
+ "epoch": 282.58,
1401
+ "learning_rate": 0.0001,
1402
+ "loss": 0.0011,
1403
+ "step": 848
1404
+ },
1405
+ {
1406
+ "epoch": 283.88,
1407
+ "learning_rate": 0.0001,
1408
+ "loss": 0.0014,
1409
+ "step": 852
1410
+ },
1411
+ {
1412
+ "epoch": 285.29,
1413
+ "learning_rate": 0.0001,
1414
+ "loss": 0.0017,
1415
+ "step": 856
1416
+ },
1417
+ {
1418
+ "epoch": 286.58,
1419
+ "learning_rate": 0.0001,
1420
+ "loss": 0.0016,
1421
+ "step": 860
1422
+ },
1423
+ {
1424
+ "epoch": 287.88,
1425
+ "learning_rate": 0.0001,
1426
+ "loss": 0.0075,
1427
+ "step": 864
1428
+ },
1429
+ {
1430
+ "epoch": 289.29,
1431
+ "learning_rate": 0.0001,
1432
+ "loss": 0.0018,
1433
+ "step": 868
1434
+ },
1435
+ {
1436
+ "epoch": 290.58,
1437
+ "learning_rate": 0.0001,
1438
+ "loss": 0.0013,
1439
+ "step": 872
1440
+ },
1441
+ {
1442
+ "epoch": 291.88,
1443
+ "learning_rate": 0.0001,
1444
+ "loss": 0.0021,
1445
+ "step": 876
1446
+ },
1447
+ {
1448
+ "epoch": 293.29,
1449
+ "learning_rate": 0.0001,
1450
+ "loss": 0.002,
1451
+ "step": 880
1452
+ },
1453
+ {
1454
+ "epoch": 294.58,
1455
+ "learning_rate": 0.0001,
1456
+ "loss": 0.0012,
1457
+ "step": 884
1458
+ },
1459
+ {
1460
+ "epoch": 295.88,
1461
+ "learning_rate": 0.0001,
1462
+ "loss": 0.0015,
1463
+ "step": 888
1464
+ },
1465
+ {
1466
+ "epoch": 297.29,
1467
+ "learning_rate": 0.0001,
1468
+ "loss": 0.0013,
1469
+ "step": 892
1470
+ },
1471
+ {
1472
+ "epoch": 298.58,
1473
+ "learning_rate": 0.0001,
1474
+ "loss": 0.0012,
1475
+ "step": 896
1476
+ },
1477
+ {
1478
+ "epoch": 298.58,
1479
+ "eval_exact_match": 0.6740812379110251,
1480
+ "eval_exec": 0.6576402321083172,
1481
+ "eval_loss": 0.39234504103660583,
1482
+ "eval_runtime": 785.534,
1483
+ "eval_samples_per_second": 1.316,
1484
+ "step": 896
1485
+ },
1486
+ {
1487
+ "epoch": 299.88,
1488
+ "learning_rate": 0.0001,
1489
+ "loss": 0.0014,
1490
+ "step": 900
1491
+ },
1492
+ {
1493
+ "epoch": 301.29,
1494
+ "learning_rate": 0.0001,
1495
+ "loss": 0.0069,
1496
+ "step": 904
1497
+ },
1498
+ {
1499
+ "epoch": 302.58,
1500
+ "learning_rate": 0.0001,
1501
+ "loss": 0.0018,
1502
+ "step": 908
1503
+ },
1504
+ {
1505
+ "epoch": 303.88,
1506
+ "learning_rate": 0.0001,
1507
+ "loss": 0.0009,
1508
+ "step": 912
1509
+ },
1510
+ {
1511
+ "epoch": 305.29,
1512
+ "learning_rate": 0.0001,
1513
+ "loss": 0.0015,
1514
+ "step": 916
1515
+ },
1516
+ {
1517
+ "epoch": 306.58,
1518
+ "learning_rate": 0.0001,
1519
+ "loss": 0.0012,
1520
+ "step": 920
1521
+ },
1522
+ {
1523
+ "epoch": 307.88,
1524
+ "learning_rate": 0.0001,
1525
+ "loss": 0.0013,
1526
+ "step": 924
1527
+ },
1528
+ {
1529
+ "epoch": 309.29,
1530
+ "learning_rate": 0.0001,
1531
+ "loss": 0.002,
1532
+ "step": 928
1533
+ },
1534
+ {
1535
+ "epoch": 310.58,
1536
+ "learning_rate": 0.0001,
1537
+ "loss": 0.0011,
1538
+ "step": 932
1539
+ },
1540
+ {
1541
+ "epoch": 311.88,
1542
+ "learning_rate": 0.0001,
1543
+ "loss": 0.0011,
1544
+ "step": 936
1545
+ },
1546
+ {
1547
+ "epoch": 313.29,
1548
+ "learning_rate": 0.0001,
1549
+ "loss": 0.0008,
1550
+ "step": 940
1551
+ },
1552
+ {
1553
+ "epoch": 314.58,
1554
+ "learning_rate": 0.0001,
1555
+ "loss": 0.0011,
1556
+ "step": 944
1557
+ },
1558
+ {
1559
+ "epoch": 315.88,
1560
+ "learning_rate": 0.0001,
1561
+ "loss": 0.0009,
1562
+ "step": 948
1563
+ },
1564
+ {
1565
+ "epoch": 317.29,
1566
+ "learning_rate": 0.0001,
1567
+ "loss": 0.0009,
1568
+ "step": 952
1569
+ },
1570
+ {
1571
+ "epoch": 318.58,
1572
+ "learning_rate": 0.0001,
1573
+ "loss": 0.0013,
1574
+ "step": 956
1575
+ },
1576
+ {
1577
+ "epoch": 319.88,
1578
+ "learning_rate": 0.0001,
1579
+ "loss": 0.0009,
1580
+ "step": 960
1581
+ },
1582
+ {
1583
+ "epoch": 319.88,
1584
+ "eval_exact_match": 0.6740812379110251,
1585
+ "eval_exec": 0.655705996131528,
1586
+ "eval_loss": 0.40166109800338745,
1587
+ "eval_runtime": 799.99,
1588
+ "eval_samples_per_second": 1.293,
1589
+ "step": 960
1590
+ },
1591
+ {
1592
+ "epoch": 321.29,
1593
+ "learning_rate": 0.0001,
1594
+ "loss": 0.0012,
1595
+ "step": 964
1596
+ },
1597
+ {
1598
+ "epoch": 322.58,
1599
+ "learning_rate": 0.0001,
1600
+ "loss": 0.0012,
1601
+ "step": 968
1602
+ },
1603
+ {
1604
+ "epoch": 323.88,
1605
+ "learning_rate": 0.0001,
1606
+ "loss": 0.0011,
1607
+ "step": 972
1608
+ },
1609
+ {
1610
+ "epoch": 325.29,
1611
+ "learning_rate": 0.0001,
1612
+ "loss": 0.0021,
1613
+ "step": 976
1614
+ },
1615
+ {
1616
+ "epoch": 326.58,
1617
+ "learning_rate": 0.0001,
1618
+ "loss": 0.0011,
1619
+ "step": 980
1620
+ },
1621
+ {
1622
+ "epoch": 327.88,
1623
+ "learning_rate": 0.0001,
1624
+ "loss": 0.0019,
1625
+ "step": 984
1626
+ },
1627
+ {
1628
+ "epoch": 329.29,
1629
+ "learning_rate": 0.0001,
1630
+ "loss": 0.001,
1631
+ "step": 988
1632
+ },
1633
+ {
1634
+ "epoch": 330.58,
1635
+ "learning_rate": 0.0001,
1636
+ "loss": 0.0008,
1637
+ "step": 992
1638
+ },
1639
+ {
1640
+ "epoch": 331.88,
1641
+ "learning_rate": 0.0001,
1642
+ "loss": 0.0009,
1643
+ "step": 996
1644
+ },
1645
+ {
1646
+ "epoch": 333.29,
1647
+ "learning_rate": 0.0001,
1648
+ "loss": 0.0012,
1649
+ "step": 1000
1650
+ },
1651
+ {
1652
+ "epoch": 334.58,
1653
+ "learning_rate": 0.0001,
1654
+ "loss": 0.0011,
1655
+ "step": 1004
1656
+ },
1657
+ {
1658
+ "epoch": 335.88,
1659
+ "learning_rate": 0.0001,
1660
+ "loss": 0.0013,
1661
+ "step": 1008
1662
+ },
1663
+ {
1664
+ "epoch": 337.29,
1665
+ "learning_rate": 0.0001,
1666
+ "loss": 0.0006,
1667
+ "step": 1012
1668
+ },
1669
+ {
1670
+ "epoch": 338.58,
1671
+ "learning_rate": 0.0001,
1672
+ "loss": 0.0015,
1673
+ "step": 1016
1674
+ },
1675
+ {
1676
+ "epoch": 339.88,
1677
+ "learning_rate": 0.0001,
1678
+ "loss": 0.0012,
1679
+ "step": 1020
1680
+ },
1681
+ {
1682
+ "epoch": 341.29,
1683
+ "learning_rate": 0.0001,
1684
+ "loss": 0.0098,
1685
+ "step": 1024
1686
+ },
1687
+ {
1688
+ "epoch": 341.29,
1689
+ "eval_exact_match": 0.6963249516441006,
1690
+ "eval_exec": 0.6721470019342359,
1691
+ "eval_loss": 0.3228262662887573,
1692
+ "eval_runtime": 823.0842,
1693
+ "eval_samples_per_second": 1.256,
1694
+ "step": 1024
1695
+ },
1696
+ {
1697
+ "epoch": 342.58,
1698
+ "learning_rate": 0.0001,
1699
+ "loss": 0.0016,
1700
+ "step": 1028
1701
+ },
1702
+ {
1703
+ "epoch": 343.88,
1704
+ "learning_rate": 0.0001,
1705
+ "loss": 0.0012,
1706
+ "step": 1032
1707
+ },
1708
+ {
1709
+ "epoch": 345.29,
1710
+ "learning_rate": 0.0001,
1711
+ "loss": 0.0007,
1712
+ "step": 1036
1713
+ },
1714
+ {
1715
+ "epoch": 346.58,
1716
+ "learning_rate": 0.0001,
1717
+ "loss": 0.0017,
1718
+ "step": 1040
1719
+ },
1720
+ {
1721
+ "epoch": 347.88,
1722
+ "learning_rate": 0.0001,
1723
+ "loss": 0.0008,
1724
+ "step": 1044
1725
+ },
1726
+ {
1727
+ "epoch": 349.29,
1728
+ "learning_rate": 0.0001,
1729
+ "loss": 0.0007,
1730
+ "step": 1048
1731
+ },
1732
+ {
1733
+ "epoch": 350.58,
1734
+ "learning_rate": 0.0001,
1735
+ "loss": 0.0007,
1736
+ "step": 1052
1737
+ },
1738
+ {
1739
+ "epoch": 351.88,
1740
+ "learning_rate": 0.0001,
1741
+ "loss": 0.0011,
1742
+ "step": 1056
1743
+ },
1744
+ {
1745
+ "epoch": 353.29,
1746
+ "learning_rate": 0.0001,
1747
+ "loss": 0.0011,
1748
+ "step": 1060
1749
+ },
1750
+ {
1751
+ "epoch": 354.58,
1752
+ "learning_rate": 0.0001,
1753
+ "loss": 0.0011,
1754
+ "step": 1064
1755
+ },
1756
+ {
1757
+ "epoch": 355.88,
1758
+ "learning_rate": 0.0001,
1759
+ "loss": 0.0008,
1760
+ "step": 1068
1761
+ },
1762
+ {
1763
+ "epoch": 357.29,
1764
+ "learning_rate": 0.0001,
1765
+ "loss": 0.0014,
1766
+ "step": 1072
1767
+ },
1768
+ {
1769
+ "epoch": 358.58,
1770
+ "learning_rate": 0.0001,
1771
+ "loss": 0.0011,
1772
+ "step": 1076
1773
+ },
1774
+ {
1775
+ "epoch": 359.88,
1776
+ "learning_rate": 0.0001,
1777
+ "loss": 0.001,
1778
+ "step": 1080
1779
+ },
1780
+ {
1781
+ "epoch": 361.29,
1782
+ "learning_rate": 0.0001,
1783
+ "loss": 0.0006,
1784
+ "step": 1084
1785
+ },
1786
+ {
1787
+ "epoch": 362.58,
1788
+ "learning_rate": 0.0001,
1789
+ "loss": 0.0008,
1790
+ "step": 1088
1791
+ },
1792
+ {
1793
+ "epoch": 362.58,
1794
+ "eval_exact_match": 0.6924564796905223,
1795
+ "eval_exec": 0.6721470019342359,
1796
+ "eval_loss": 0.4092560410499573,
1797
+ "eval_runtime": 834.3948,
1798
+ "eval_samples_per_second": 1.239,
1799
+ "step": 1088
1800
+ },
1801
+ {
1802
+ "epoch": 363.88,
1803
+ "learning_rate": 0.0001,
1804
+ "loss": 0.0005,
1805
+ "step": 1092
1806
+ },
1807
+ {
1808
+ "epoch": 365.29,
1809
+ "learning_rate": 0.0001,
1810
+ "loss": 0.0013,
1811
+ "step": 1096
1812
+ },
1813
+ {
1814
+ "epoch": 366.58,
1815
+ "learning_rate": 0.0001,
1816
+ "loss": 0.0008,
1817
+ "step": 1100
1818
+ },
1819
+ {
1820
+ "epoch": 367.88,
1821
+ "learning_rate": 0.0001,
1822
+ "loss": 0.001,
1823
+ "step": 1104
1824
+ },
1825
+ {
1826
+ "epoch": 369.29,
1827
+ "learning_rate": 0.0001,
1828
+ "loss": 0.001,
1829
+ "step": 1108
1830
+ },
1831
+ {
1832
+ "epoch": 370.58,
1833
+ "learning_rate": 0.0001,
1834
+ "loss": 0.0007,
1835
+ "step": 1112
1836
+ },
1837
+ {
1838
+ "epoch": 371.88,
1839
+ "learning_rate": 0.0001,
1840
+ "loss": 0.0006,
1841
+ "step": 1116
1842
+ },
1843
+ {
1844
+ "epoch": 373.29,
1845
+ "learning_rate": 0.0001,
1846
+ "loss": 0.001,
1847
+ "step": 1120
1848
+ },
1849
+ {
1850
+ "epoch": 374.58,
1851
+ "learning_rate": 0.0001,
1852
+ "loss": 0.0054,
1853
+ "step": 1124
1854
+ },
1855
+ {
1856
+ "epoch": 375.88,
1857
+ "learning_rate": 0.0001,
1858
+ "loss": 0.0009,
1859
+ "step": 1128
1860
+ },
1861
+ {
1862
+ "epoch": 377.29,
1863
+ "learning_rate": 0.0001,
1864
+ "loss": 0.0011,
1865
+ "step": 1132
1866
+ },
1867
+ {
1868
+ "epoch": 378.58,
1869
+ "learning_rate": 0.0001,
1870
+ "loss": 0.0006,
1871
+ "step": 1136
1872
+ },
1873
+ {
1874
+ "epoch": 379.88,
1875
+ "learning_rate": 0.0001,
1876
+ "loss": 0.0008,
1877
+ "step": 1140
1878
+ },
1879
+ {
1880
+ "epoch": 381.29,
1881
+ "learning_rate": 0.0001,
1882
+ "loss": 0.0006,
1883
+ "step": 1144
1884
+ },
1885
+ {
1886
+ "epoch": 382.58,
1887
+ "learning_rate": 0.0001,
1888
+ "loss": 0.0006,
1889
+ "step": 1148
1890
+ },
1891
+ {
1892
+ "epoch": 383.88,
1893
+ "learning_rate": 0.0001,
1894
+ "loss": 0.0013,
1895
+ "step": 1152
1896
+ },
1897
+ {
1898
+ "epoch": 383.88,
1899
+ "eval_exact_match": 0.6934235976789168,
1900
+ "eval_exec": 0.6818181818181818,
1901
+ "eval_loss": 0.43188637495040894,
1902
+ "eval_runtime": 827.3576,
1903
+ "eval_samples_per_second": 1.25,
1904
+ "step": 1152
1905
+ },
1906
+ {
1907
+ "epoch": 385.29,
1908
+ "learning_rate": 0.0001,
1909
+ "loss": 0.0037,
1910
+ "step": 1156
1911
+ },
1912
+ {
1913
+ "epoch": 386.58,
1914
+ "learning_rate": 0.0001,
1915
+ "loss": 0.0021,
1916
+ "step": 1160
1917
+ },
1918
+ {
1919
+ "epoch": 387.88,
1920
+ "learning_rate": 0.0001,
1921
+ "loss": 0.0009,
1922
+ "step": 1164
1923
+ },
1924
+ {
1925
+ "epoch": 389.29,
1926
+ "learning_rate": 0.0001,
1927
+ "loss": 0.0013,
1928
+ "step": 1168
1929
+ },
1930
+ {
1931
+ "epoch": 390.58,
1932
+ "learning_rate": 0.0001,
1933
+ "loss": 0.0006,
1934
+ "step": 1172
1935
+ },
1936
+ {
1937
+ "epoch": 391.88,
1938
+ "learning_rate": 0.0001,
1939
+ "loss": 0.0008,
1940
+ "step": 1176
1941
+ },
1942
+ {
1943
+ "epoch": 393.29,
1944
+ "learning_rate": 0.0001,
1945
+ "loss": 0.0013,
1946
+ "step": 1180
1947
+ },
1948
+ {
1949
+ "epoch": 394.58,
1950
+ "learning_rate": 0.0001,
1951
+ "loss": 0.0007,
1952
+ "step": 1184
1953
+ },
1954
+ {
1955
+ "epoch": 395.88,
1956
+ "learning_rate": 0.0001,
1957
+ "loss": 0.0005,
1958
+ "step": 1188
1959
+ },
1960
+ {
1961
+ "epoch": 397.29,
1962
+ "learning_rate": 0.0001,
1963
+ "loss": 0.0007,
1964
+ "step": 1192
1965
+ },
1966
+ {
1967
+ "epoch": 398.58,
1968
+ "learning_rate": 0.0001,
1969
+ "loss": 0.001,
1970
+ "step": 1196
1971
+ },
1972
+ {
1973
+ "epoch": 399.88,
1974
+ "learning_rate": 0.0001,
1975
+ "loss": 0.0004,
1976
+ "step": 1200
1977
+ },
1978
+ {
1979
+ "epoch": 401.29,
1980
+ "learning_rate": 0.0001,
1981
+ "loss": 0.0007,
1982
+ "step": 1204
1983
+ },
1984
+ {
1985
+ "epoch": 402.58,
1986
+ "learning_rate": 0.0001,
1987
+ "loss": 0.0014,
1988
+ "step": 1208
1989
+ },
1990
+ {
1991
+ "epoch": 403.88,
1992
+ "learning_rate": 0.0001,
1993
+ "loss": 0.001,
1994
+ "step": 1212
1995
+ },
1996
+ {
1997
+ "epoch": 405.29,
1998
+ "learning_rate": 0.0001,
1999
+ "loss": 0.0008,
2000
+ "step": 1216
2001
+ },
2002
+ {
2003
+ "epoch": 405.29,
2004
+ "eval_exact_match": 0.6750483558994197,
2005
+ "eval_exec": 0.6624758220502901,
2006
+ "eval_loss": 0.38173577189445496,
2007
+ "eval_runtime": 863.9436,
2008
+ "eval_samples_per_second": 1.197,
2009
+ "step": 1216
2010
+ },
2011
+ {
2012
+ "epoch": 406.58,
2013
+ "learning_rate": 0.0001,
2014
+ "loss": 0.0005,
2015
+ "step": 1220
2016
+ },
2017
+ {
2018
+ "epoch": 407.88,
2019
+ "learning_rate": 0.0001,
2020
+ "loss": 0.0005,
2021
+ "step": 1224
2022
+ },
2023
+ {
2024
+ "epoch": 409.29,
2025
+ "learning_rate": 0.0001,
2026
+ "loss": 0.0018,
2027
+ "step": 1228
2028
+ },
2029
+ {
2030
+ "epoch": 410.58,
2031
+ "learning_rate": 0.0001,
2032
+ "loss": 0.0005,
2033
+ "step": 1232
2034
+ },
2035
+ {
2036
+ "epoch": 411.88,
2037
+ "learning_rate": 0.0001,
2038
+ "loss": 0.0006,
2039
+ "step": 1236
2040
+ },
2041
+ {
2042
+ "epoch": 413.29,
2043
+ "learning_rate": 0.0001,
2044
+ "loss": 0.0006,
2045
+ "step": 1240
2046
+ },
2047
+ {
2048
+ "epoch": 414.58,
2049
+ "learning_rate": 0.0001,
2050
+ "loss": 0.0007,
2051
+ "step": 1244
2052
+ },
2053
+ {
2054
+ "epoch": 415.88,
2055
+ "learning_rate": 0.0001,
2056
+ "loss": 0.0006,
2057
+ "step": 1248
2058
+ },
2059
+ {
2060
+ "epoch": 417.29,
2061
+ "learning_rate": 0.0001,
2062
+ "loss": 0.0007,
2063
+ "step": 1252
2064
+ },
2065
+ {
2066
+ "epoch": 418.58,
2067
+ "learning_rate": 0.0001,
2068
+ "loss": 0.0007,
2069
+ "step": 1256
2070
+ },
2071
+ {
2072
+ "epoch": 419.88,
2073
+ "learning_rate": 0.0001,
2074
+ "loss": 0.0006,
2075
+ "step": 1260
2076
+ },
2077
+ {
2078
+ "epoch": 421.29,
2079
+ "learning_rate": 0.0001,
2080
+ "loss": 0.0008,
2081
+ "step": 1264
2082
+ },
2083
+ {
2084
+ "epoch": 422.58,
2085
+ "learning_rate": 0.0001,
2086
+ "loss": 0.0007,
2087
+ "step": 1268
2088
+ },
2089
+ {
2090
+ "epoch": 423.88,
2091
+ "learning_rate": 0.0001,
2092
+ "loss": 0.0005,
2093
+ "step": 1272
2094
+ },
2095
+ {
2096
+ "epoch": 425.29,
2097
+ "learning_rate": 0.0001,
2098
+ "loss": 0.0005,
2099
+ "step": 1276
2100
+ },
2101
+ {
2102
+ "epoch": 426.58,
2103
+ "learning_rate": 0.0001,
2104
+ "loss": 0.0006,
2105
+ "step": 1280
2106
+ },
2107
+ {
2108
+ "epoch": 426.58,
2109
+ "eval_exact_match": 0.6808510638297872,
2110
+ "eval_exec": 0.6644100580270793,
2111
+ "eval_loss": 0.42931750416755676,
2112
+ "eval_runtime": 845.9587,
2113
+ "eval_samples_per_second": 1.222,
2114
+ "step": 1280
2115
+ },
2116
+ {
2117
+ "epoch": 427.88,
2118
+ "learning_rate": 0.0001,
2119
+ "loss": 0.0003,
2120
+ "step": 1284
2121
+ },
2122
+ {
2123
+ "epoch": 429.29,
2124
+ "learning_rate": 0.0001,
2125
+ "loss": 0.0007,
2126
+ "step": 1288
2127
+ },
2128
+ {
2129
+ "epoch": 430.58,
2130
+ "learning_rate": 0.0001,
2131
+ "loss": 0.0011,
2132
+ "step": 1292
2133
+ },
2134
+ {
2135
+ "epoch": 431.88,
2136
+ "learning_rate": 0.0001,
2137
+ "loss": 0.0007,
2138
+ "step": 1296
2139
+ },
2140
+ {
2141
+ "epoch": 433.29,
2142
+ "learning_rate": 0.0001,
2143
+ "loss": 0.001,
2144
+ "step": 1300
2145
+ },
2146
+ {
2147
+ "epoch": 434.58,
2148
+ "learning_rate": 0.0001,
2149
+ "loss": 0.0005,
2150
+ "step": 1304
2151
+ },
2152
+ {
2153
+ "epoch": 435.88,
2154
+ "learning_rate": 0.0001,
2155
+ "loss": 0.0009,
2156
+ "step": 1308
2157
+ },
2158
+ {
2159
+ "epoch": 437.29,
2160
+ "learning_rate": 0.0001,
2161
+ "loss": 0.0013,
2162
+ "step": 1312
2163
+ },
2164
+ {
2165
+ "epoch": 438.58,
2166
+ "learning_rate": 0.0001,
2167
+ "loss": 0.0005,
2168
+ "step": 1316
2169
+ },
2170
+ {
2171
+ "epoch": 439.88,
2172
+ "learning_rate": 0.0001,
2173
+ "loss": 0.0006,
2174
+ "step": 1320
2175
+ },
2176
+ {
2177
+ "epoch": 441.29,
2178
+ "learning_rate": 0.0001,
2179
+ "loss": 0.0005,
2180
+ "step": 1324
2181
+ },
2182
+ {
2183
+ "epoch": 442.58,
2184
+ "learning_rate": 0.0001,
2185
+ "loss": 0.0004,
2186
+ "step": 1328
2187
+ },
2188
+ {
2189
+ "epoch": 443.88,
2190
+ "learning_rate": 0.0001,
2191
+ "loss": 0.0009,
2192
+ "step": 1332
2193
+ },
2194
+ {
2195
+ "epoch": 445.29,
2196
+ "learning_rate": 0.0001,
2197
+ "loss": 0.0003,
2198
+ "step": 1336
2199
+ },
2200
+ {
2201
+ "epoch": 446.58,
2202
+ "learning_rate": 0.0001,
2203
+ "loss": 0.0005,
2204
+ "step": 1340
2205
+ },
2206
+ {
2207
+ "epoch": 447.88,
2208
+ "learning_rate": 0.0001,
2209
+ "loss": 0.001,
2210
+ "step": 1344
2211
+ },
2212
+ {
2213
+ "epoch": 447.88,
2214
+ "eval_exact_match": 0.6518375241779497,
2215
+ "eval_exec": 0.6499032882011605,
2216
+ "eval_loss": 0.416789174079895,
2217
+ "eval_runtime": 773.9784,
2218
+ "eval_samples_per_second": 1.336,
2219
+ "step": 1344
2220
+ },
2221
+ {
2222
+ "epoch": 449.29,
2223
+ "learning_rate": 0.0001,
2224
+ "loss": 0.0007,
2225
+ "step": 1348
2226
+ },
2227
+ {
2228
+ "epoch": 450.58,
2229
+ "learning_rate": 0.0001,
2230
+ "loss": 0.0006,
2231
+ "step": 1352
2232
+ },
2233
+ {
2234
+ "epoch": 451.88,
2235
+ "learning_rate": 0.0001,
2236
+ "loss": 0.0007,
2237
+ "step": 1356
2238
+ },
2239
+ {
2240
+ "epoch": 453.29,
2241
+ "learning_rate": 0.0001,
2242
+ "loss": 0.0009,
2243
+ "step": 1360
2244
+ },
2245
+ {
2246
+ "epoch": 454.58,
2247
+ "learning_rate": 0.0001,
2248
+ "loss": 0.0006,
2249
+ "step": 1364
2250
+ },
2251
+ {
2252
+ "epoch": 455.88,
2253
+ "learning_rate": 0.0001,
2254
+ "loss": 0.0003,
2255
+ "step": 1368
2256
+ },
2257
+ {
2258
+ "epoch": 457.29,
2259
+ "learning_rate": 0.0001,
2260
+ "loss": 0.0004,
2261
+ "step": 1372
2262
+ },
2263
+ {
2264
+ "epoch": 458.58,
2265
+ "learning_rate": 0.0001,
2266
+ "loss": 0.0005,
2267
+ "step": 1376
2268
+ },
2269
+ {
2270
+ "epoch": 459.88,
2271
+ "learning_rate": 0.0001,
2272
+ "loss": 0.0005,
2273
+ "step": 1380
2274
+ },
2275
+ {
2276
+ "epoch": 461.29,
2277
+ "learning_rate": 0.0001,
2278
+ "loss": 0.0007,
2279
+ "step": 1384
2280
+ },
2281
+ {
2282
+ "epoch": 462.58,
2283
+ "learning_rate": 0.0001,
2284
+ "loss": 0.0005,
2285
+ "step": 1388
2286
+ },
2287
+ {
2288
+ "epoch": 463.88,
2289
+ "learning_rate": 0.0001,
2290
+ "loss": 0.0008,
2291
+ "step": 1392
2292
+ },
2293
+ {
2294
+ "epoch": 465.29,
2295
+ "learning_rate": 0.0001,
2296
+ "loss": 0.0009,
2297
+ "step": 1396
2298
+ },
2299
+ {
2300
+ "epoch": 466.58,
2301
+ "learning_rate": 0.0001,
2302
+ "loss": 0.0007,
2303
+ "step": 1400
2304
+ },
2305
+ {
2306
+ "epoch": 467.88,
2307
+ "learning_rate": 0.0001,
2308
+ "loss": 0.001,
2309
+ "step": 1404
2310
+ },
2311
+ {
2312
+ "epoch": 469.29,
2313
+ "learning_rate": 0.0001,
2314
+ "loss": 0.0022,
2315
+ "step": 1408
2316
+ },
2317
+ {
2318
+ "epoch": 469.29,
2319
+ "eval_exact_match": 0.6750483558994197,
2320
+ "eval_exec": 0.6673114119922631,
2321
+ "eval_loss": 0.36535608768463135,
2322
+ "eval_runtime": 954.0188,
2323
+ "eval_samples_per_second": 1.084,
2324
+ "step": 1408
2325
+ },
2326
+ {
2327
+ "epoch": 470.58,
2328
+ "learning_rate": 0.0001,
2329
+ "loss": 0.0006,
2330
+ "step": 1412
2331
+ },
2332
+ {
2333
+ "epoch": 471.88,
2334
+ "learning_rate": 0.0001,
2335
+ "loss": 0.0006,
2336
+ "step": 1416
2337
+ },
2338
+ {
2339
+ "epoch": 473.29,
2340
+ "learning_rate": 0.0001,
2341
+ "loss": 0.0004,
2342
+ "step": 1420
2343
+ },
2344
+ {
2345
+ "epoch": 474.58,
2346
+ "learning_rate": 0.0001,
2347
+ "loss": 0.001,
2348
+ "step": 1424
2349
+ },
2350
+ {
2351
+ "epoch": 475.88,
2352
+ "learning_rate": 0.0001,
2353
+ "loss": 0.008,
2354
+ "step": 1428
2355
+ },
2356
+ {
2357
+ "epoch": 477.29,
2358
+ "learning_rate": 0.0001,
2359
+ "loss": 0.0043,
2360
+ "step": 1432
2361
+ },
2362
+ {
2363
+ "epoch": 478.58,
2364
+ "learning_rate": 0.0001,
2365
+ "loss": 0.0007,
2366
+ "step": 1436
2367
+ },
2368
+ {
2369
+ "epoch": 479.88,
2370
+ "learning_rate": 0.0001,
2371
+ "loss": 0.0003,
2372
+ "step": 1440
2373
+ },
2374
+ {
2375
+ "epoch": 481.29,
2376
+ "learning_rate": 0.0001,
2377
+ "loss": 0.0003,
2378
+ "step": 1444
2379
+ },
2380
+ {
2381
+ "epoch": 482.58,
2382
+ "learning_rate": 0.0001,
2383
+ "loss": 0.0004,
2384
+ "step": 1448
2385
+ },
2386
+ {
2387
+ "epoch": 483.88,
2388
+ "learning_rate": 0.0001,
2389
+ "loss": 0.0003,
2390
+ "step": 1452
2391
+ },
2392
+ {
2393
+ "epoch": 485.29,
2394
+ "learning_rate": 0.0001,
2395
+ "loss": 0.0007,
2396
+ "step": 1456
2397
+ },
2398
+ {
2399
+ "epoch": 486.58,
2400
+ "learning_rate": 0.0001,
2401
+ "loss": 0.0003,
2402
+ "step": 1460
2403
+ },
2404
+ {
2405
+ "epoch": 487.88,
2406
+ "learning_rate": 0.0001,
2407
+ "loss": 0.0007,
2408
+ "step": 1464
2409
+ },
2410
+ {
2411
+ "epoch": 489.29,
2412
+ "learning_rate": 0.0001,
2413
+ "loss": 0.0004,
2414
+ "step": 1468
2415
+ },
2416
+ {
2417
+ "epoch": 490.58,
2418
+ "learning_rate": 0.0001,
2419
+ "loss": 0.0008,
2420
+ "step": 1472
2421
+ },
2422
+ {
2423
+ "epoch": 490.58,
2424
+ "eval_exact_match": 0.6818181818181818,
2425
+ "eval_exec": 0.6740812379110251,
2426
+ "eval_loss": 0.41243937611579895,
2427
+ "eval_runtime": 955.6213,
2428
+ "eval_samples_per_second": 1.082,
2429
+ "step": 1472
2430
+ },
2431
+ {
2432
+ "epoch": 491.88,
2433
+ "learning_rate": 0.0001,
2434
+ "loss": 0.0008,
2435
+ "step": 1476
2436
+ },
2437
+ {
2438
+ "epoch": 493.29,
2439
+ "learning_rate": 0.0001,
2440
+ "loss": 0.0012,
2441
+ "step": 1480
2442
+ },
2443
+ {
2444
+ "epoch": 494.58,
2445
+ "learning_rate": 0.0001,
2446
+ "loss": 0.0003,
2447
+ "step": 1484
2448
+ },
2449
+ {
2450
+ "epoch": 495.88,
2451
+ "learning_rate": 0.0001,
2452
+ "loss": 0.0008,
2453
+ "step": 1488
2454
+ },
2455
+ {
2456
+ "epoch": 497.29,
2457
+ "learning_rate": 0.0001,
2458
+ "loss": 0.0003,
2459
+ "step": 1492
2460
+ },
2461
+ {
2462
+ "epoch": 498.58,
2463
+ "learning_rate": 0.0001,
2464
+ "loss": 0.0008,
2465
+ "step": 1496
2466
+ },
2467
+ {
2468
+ "epoch": 499.88,
2469
+ "learning_rate": 0.0001,
2470
+ "loss": 0.0003,
2471
+ "step": 1500
2472
+ },
2473
+ {
2474
+ "epoch": 501.29,
2475
+ "learning_rate": 0.0001,
2476
+ "loss": 0.0006,
2477
+ "step": 1504
2478
+ },
2479
+ {
2480
+ "epoch": 502.58,
2481
+ "learning_rate": 0.0001,
2482
+ "loss": 0.0007,
2483
+ "step": 1508
2484
+ },
2485
+ {
2486
+ "epoch": 503.88,
2487
+ "learning_rate": 0.0001,
2488
+ "loss": 0.0005,
2489
+ "step": 1512
2490
+ },
2491
+ {
2492
+ "epoch": 505.29,
2493
+ "learning_rate": 0.0001,
2494
+ "loss": 0.0003,
2495
+ "step": 1516
2496
+ },
2497
+ {
2498
+ "epoch": 506.58,
2499
+ "learning_rate": 0.0001,
2500
+ "loss": 0.0006,
2501
+ "step": 1520
2502
+ },
2503
+ {
2504
+ "epoch": 507.88,
2505
+ "learning_rate": 0.0001,
2506
+ "loss": 0.0005,
2507
+ "step": 1524
2508
+ },
2509
+ {
2510
+ "epoch": 509.29,
2511
+ "learning_rate": 0.0001,
2512
+ "loss": 0.0008,
2513
+ "step": 1528
2514
+ },
2515
+ {
2516
+ "epoch": 510.58,
2517
+ "learning_rate": 0.0001,
2518
+ "loss": 0.0006,
2519
+ "step": 1532
2520
+ },
2521
+ {
2522
+ "epoch": 511.88,
2523
+ "learning_rate": 0.0001,
2524
+ "loss": 0.0008,
2525
+ "step": 1536
2526
+ },
2527
+ {
2528
+ "epoch": 511.88,
2529
+ "eval_exact_match": 0.6856866537717602,
2530
+ "eval_exec": 0.6798839458413927,
2531
+ "eval_loss": 0.4265560507774353,
2532
+ "eval_runtime": 1299.537,
2533
+ "eval_samples_per_second": 0.796,
2534
+ "step": 1536
2535
+ },
2536
+ {
2537
+ "epoch": 513.29,
2538
+ "learning_rate": 0.0001,
2539
+ "loss": 0.0004,
2540
+ "step": 1540
2541
+ },
2542
+ {
2543
+ "epoch": 514.58,
2544
+ "learning_rate": 0.0001,
2545
+ "loss": 0.0005,
2546
+ "step": 1544
2547
+ },
2548
+ {
2549
+ "epoch": 515.88,
2550
+ "learning_rate": 0.0001,
2551
+ "loss": 0.0008,
2552
+ "step": 1548
2553
+ },
2554
+ {
2555
+ "epoch": 517.29,
2556
+ "learning_rate": 0.0001,
2557
+ "loss": 0.0005,
2558
+ "step": 1552
2559
+ },
2560
+ {
2561
+ "epoch": 518.58,
2562
+ "learning_rate": 0.0001,
2563
+ "loss": 0.0006,
2564
+ "step": 1556
2565
+ },
2566
+ {
2567
+ "epoch": 519.88,
2568
+ "learning_rate": 0.0001,
2569
+ "loss": 0.0003,
2570
+ "step": 1560
2571
+ },
2572
+ {
2573
+ "epoch": 521.29,
2574
+ "learning_rate": 0.0001,
2575
+ "loss": 0.0003,
2576
+ "step": 1564
2577
+ },
2578
+ {
2579
+ "epoch": 522.58,
2580
+ "learning_rate": 0.0001,
2581
+ "loss": 0.0008,
2582
+ "step": 1568
2583
+ },
2584
+ {
2585
+ "epoch": 523.88,
2586
+ "learning_rate": 0.0001,
2587
+ "loss": 0.0004,
2588
+ "step": 1572
2589
+ },
2590
+ {
2591
+ "epoch": 525.29,
2592
+ "learning_rate": 0.0001,
2593
+ "loss": 0.0007,
2594
+ "step": 1576
2595
+ },
2596
+ {
2597
+ "epoch": 526.58,
2598
+ "learning_rate": 0.0001,
2599
+ "loss": 0.0004,
2600
+ "step": 1580
2601
+ },
2602
+ {
2603
+ "epoch": 527.88,
2604
+ "learning_rate": 0.0001,
2605
+ "loss": 0.0002,
2606
+ "step": 1584
2607
+ },
2608
+ {
2609
+ "epoch": 529.29,
2610
+ "learning_rate": 0.0001,
2611
+ "loss": 0.0005,
2612
+ "step": 1588
2613
+ },
2614
+ {
2615
+ "epoch": 530.58,
2616
+ "learning_rate": 0.0001,
2617
+ "loss": 0.0014,
2618
+ "step": 1592
2619
+ },
2620
+ {
2621
+ "epoch": 531.88,
2622
+ "learning_rate": 0.0001,
2623
+ "loss": 0.0005,
2624
+ "step": 1596
2625
+ },
2626
+ {
2627
+ "epoch": 533.29,
2628
+ "learning_rate": 0.0001,
2629
+ "loss": 0.0005,
2630
+ "step": 1600
2631
+ },
2632
+ {
2633
+ "epoch": 533.29,
2634
+ "eval_exact_match": 0.6856866537717602,
2635
+ "eval_exec": 0.6779497098646035,
2636
+ "eval_loss": 0.4029601216316223,
2637
+ "eval_runtime": 999.4683,
2638
+ "eval_samples_per_second": 1.035,
2639
+ "step": 1600
2640
+ },
2641
+ {
2642
+ "epoch": 534.58,
2643
+ "learning_rate": 0.0001,
2644
+ "loss": 0.0003,
2645
+ "step": 1604
2646
+ },
2647
+ {
2648
+ "epoch": 535.88,
2649
+ "learning_rate": 0.0001,
2650
+ "loss": 0.0005,
2651
+ "step": 1608
2652
+ },
2653
+ {
2654
+ "epoch": 537.29,
2655
+ "learning_rate": 0.0001,
2656
+ "loss": 0.0003,
2657
+ "step": 1612
2658
+ },
2659
+ {
2660
+ "epoch": 538.58,
2661
+ "learning_rate": 0.0001,
2662
+ "loss": 0.0003,
2663
+ "step": 1616
2664
+ },
2665
+ {
2666
+ "epoch": 539.88,
2667
+ "learning_rate": 0.0001,
2668
+ "loss": 0.0006,
2669
+ "step": 1620
2670
+ },
2671
+ {
2672
+ "epoch": 541.29,
2673
+ "learning_rate": 0.0001,
2674
+ "loss": 0.0005,
2675
+ "step": 1624
2676
+ },
2677
+ {
2678
+ "epoch": 542.58,
2679
+ "learning_rate": 0.0001,
2680
+ "loss": 0.0006,
2681
+ "step": 1628
2682
+ },
2683
+ {
2684
+ "epoch": 543.88,
2685
+ "learning_rate": 0.0001,
2686
+ "loss": 0.0013,
2687
+ "step": 1632
2688
+ },
2689
+ {
2690
+ "epoch": 545.29,
2691
+ "learning_rate": 0.0001,
2692
+ "loss": 0.0005,
2693
+ "step": 1636
2694
+ },
2695
+ {
2696
+ "epoch": 546.58,
2697
+ "learning_rate": 0.0001,
2698
+ "loss": 0.0002,
2699
+ "step": 1640
2700
+ },
2701
+ {
2702
+ "epoch": 547.88,
2703
+ "learning_rate": 0.0001,
2704
+ "loss": 0.0003,
2705
+ "step": 1644
2706
+ },
2707
+ {
2708
+ "epoch": 549.29,
2709
+ "learning_rate": 0.0001,
2710
+ "loss": 0.0002,
2711
+ "step": 1648
2712
+ },
2713
+ {
2714
+ "epoch": 550.58,
2715
+ "learning_rate": 0.0001,
2716
+ "loss": 0.0002,
2717
+ "step": 1652
2718
+ },
2719
+ {
2720
+ "epoch": 551.88,
2721
+ "learning_rate": 0.0001,
2722
+ "loss": 0.0004,
2723
+ "step": 1656
2724
+ },
2725
+ {
2726
+ "epoch": 553.29,
2727
+ "learning_rate": 0.0001,
2728
+ "loss": 0.0004,
2729
+ "step": 1660
2730
+ },
2731
+ {
2732
+ "epoch": 554.58,
2733
+ "learning_rate": 0.0001,
2734
+ "loss": 0.0002,
2735
+ "step": 1664
2736
+ },
2737
+ {
2738
+ "epoch": 554.58,
2739
+ "eval_exact_match": 0.6934235976789168,
2740
+ "eval_exec": 0.6827852998065764,
2741
+ "eval_loss": 0.44639065861701965,
2742
+ "eval_runtime": 1031.7773,
2743
+ "eval_samples_per_second": 1.002,
2744
+ "step": 1664
2745
+ },
2746
+ {
2747
+ "epoch": 555.88,
2748
+ "learning_rate": 0.0001,
2749
+ "loss": 0.0003,
2750
+ "step": 1668
2751
+ },
2752
+ {
2753
+ "epoch": 557.29,
2754
+ "learning_rate": 0.0001,
2755
+ "loss": 0.0004,
2756
+ "step": 1672
2757
+ },
2758
+ {
2759
+ "epoch": 558.58,
2760
+ "learning_rate": 0.0001,
2761
+ "loss": 0.0025,
2762
+ "step": 1676
2763
+ },
2764
+ {
2765
+ "epoch": 559.88,
2766
+ "learning_rate": 0.0001,
2767
+ "loss": 0.0029,
2768
+ "step": 1680
2769
+ },
2770
+ {
2771
+ "epoch": 561.29,
2772
+ "learning_rate": 0.0001,
2773
+ "loss": 0.0004,
2774
+ "step": 1684
2775
+ },
2776
+ {
2777
+ "epoch": 562.58,
2778
+ "learning_rate": 0.0001,
2779
+ "loss": 0.0005,
2780
+ "step": 1688
2781
+ },
2782
+ {
2783
+ "epoch": 563.88,
2784
+ "learning_rate": 0.0001,
2785
+ "loss": 0.0007,
2786
+ "step": 1692
2787
+ },
2788
+ {
2789
+ "epoch": 565.29,
2790
+ "learning_rate": 0.0001,
2791
+ "loss": 0.0003,
2792
+ "step": 1696
2793
+ },
2794
+ {
2795
+ "epoch": 566.58,
2796
+ "learning_rate": 0.0001,
2797
+ "loss": 0.0003,
2798
+ "step": 1700
2799
+ },
2800
+ {
2801
+ "epoch": 567.88,
2802
+ "learning_rate": 0.0001,
2803
+ "loss": 0.0004,
2804
+ "step": 1704
2805
+ },
2806
+ {
2807
+ "epoch": 569.29,
2808
+ "learning_rate": 0.0001,
2809
+ "loss": 0.0007,
2810
+ "step": 1708
2811
+ },
2812
+ {
2813
+ "epoch": 570.58,
2814
+ "learning_rate": 0.0001,
2815
+ "loss": 0.0018,
2816
+ "step": 1712
2817
+ },
2818
+ {
2819
+ "epoch": 571.88,
2820
+ "learning_rate": 0.0001,
2821
+ "loss": 0.0011,
2822
+ "step": 1716
2823
+ },
2824
+ {
2825
+ "epoch": 573.29,
2826
+ "learning_rate": 0.0001,
2827
+ "loss": 0.0004,
2828
+ "step": 1720
2829
+ },
2830
+ {
2831
+ "epoch": 574.58,
2832
+ "learning_rate": 0.0001,
2833
+ "loss": 0.0005,
2834
+ "step": 1724
2835
+ },
2836
+ {
2837
+ "epoch": 575.88,
2838
+ "learning_rate": 0.0001,
2839
+ "loss": 0.0002,
2840
+ "step": 1728
2841
+ },
2842
+ {
2843
+ "epoch": 575.88,
2844
+ "eval_exact_match": 0.6876208897485493,
2845
+ "eval_exec": 0.6760154738878144,
2846
+ "eval_loss": 0.41115066409111023,
2847
+ "eval_runtime": 1007.2726,
2848
+ "eval_samples_per_second": 1.027,
2849
+ "step": 1728
2850
+ },
2851
+ {
2852
+ "epoch": 577.29,
2853
+ "learning_rate": 0.0001,
2854
+ "loss": 0.0003,
2855
+ "step": 1732
2856
+ },
2857
+ {
2858
+ "epoch": 578.58,
2859
+ "learning_rate": 0.0001,
2860
+ "loss": 0.0006,
2861
+ "step": 1736
2862
+ },
2863
+ {
2864
+ "epoch": 579.88,
2865
+ "learning_rate": 0.0001,
2866
+ "loss": 0.0002,
2867
+ "step": 1740
2868
+ },
2869
+ {
2870
+ "epoch": 581.29,
2871
+ "learning_rate": 0.0001,
2872
+ "loss": 0.0003,
2873
+ "step": 1744
2874
+ },
2875
+ {
2876
+ "epoch": 582.58,
2877
+ "learning_rate": 0.0001,
2878
+ "loss": 0.0002,
2879
+ "step": 1748
2880
+ },
2881
+ {
2882
+ "epoch": 583.88,
2883
+ "learning_rate": 0.0001,
2884
+ "loss": 0.0004,
2885
+ "step": 1752
2886
+ },
2887
+ {
2888
+ "epoch": 585.29,
2889
+ "learning_rate": 0.0001,
2890
+ "loss": 0.0029,
2891
+ "step": 1756
2892
+ },
2893
+ {
2894
+ "epoch": 586.58,
2895
+ "learning_rate": 0.0001,
2896
+ "loss": 0.0021,
2897
+ "step": 1760
2898
+ },
2899
+ {
2900
+ "epoch": 587.88,
2901
+ "learning_rate": 0.0001,
2902
+ "loss": 0.0006,
2903
+ "step": 1764
2904
+ },
2905
+ {
2906
+ "epoch": 589.29,
2907
+ "learning_rate": 0.0001,
2908
+ "loss": 0.0002,
2909
+ "step": 1768
2910
+ },
2911
+ {
2912
+ "epoch": 590.58,
2913
+ "learning_rate": 0.0001,
2914
+ "loss": 0.0004,
2915
+ "step": 1772
2916
+ },
2917
+ {
2918
+ "epoch": 591.88,
2919
+ "learning_rate": 0.0001,
2920
+ "loss": 0.0003,
2921
+ "step": 1776
2922
+ },
2923
+ {
2924
+ "epoch": 593.29,
2925
+ "learning_rate": 0.0001,
2926
+ "loss": 0.0007,
2927
+ "step": 1780
2928
+ },
2929
+ {
2930
+ "epoch": 594.58,
2931
+ "learning_rate": 0.0001,
2932
+ "loss": 0.0004,
2933
+ "step": 1784
2934
+ },
2935
+ {
2936
+ "epoch": 595.88,
2937
+ "learning_rate": 0.0001,
2938
+ "loss": 0.0003,
2939
+ "step": 1788
2940
+ },
2941
+ {
2942
+ "epoch": 597.29,
2943
+ "learning_rate": 0.0001,
2944
+ "loss": 0.0003,
2945
+ "step": 1792
2946
+ },
2947
+ {
2948
+ "epoch": 597.29,
2949
+ "eval_exact_match": 0.6943907156673114,
2950
+ "eval_exec": 0.6779497098646035,
2951
+ "eval_loss": 0.4605121910572052,
2952
+ "eval_runtime": 1298.121,
2953
+ "eval_samples_per_second": 0.797,
2954
+ "step": 1792
2955
+ },
2956
+ {
2957
+ "epoch": 598.58,
2958
+ "learning_rate": 0.0001,
2959
+ "loss": 0.0003,
2960
+ "step": 1796
2961
+ },
2962
+ {
2963
+ "epoch": 599.88,
2964
+ "learning_rate": 0.0001,
2965
+ "loss": 0.0005,
2966
+ "step": 1800
2967
+ },
2968
+ {
2969
+ "epoch": 601.29,
2970
+ "learning_rate": 0.0001,
2971
+ "loss": 0.0003,
2972
+ "step": 1804
2973
+ },
2974
+ {
2975
+ "epoch": 602.58,
2976
+ "learning_rate": 0.0001,
2977
+ "loss": 0.0008,
2978
+ "step": 1808
2979
+ },
2980
+ {
2981
+ "epoch": 603.88,
2982
+ "learning_rate": 0.0001,
2983
+ "loss": 0.0008,
2984
+ "step": 1812
2985
+ },
2986
+ {
2987
+ "epoch": 605.29,
2988
+ "learning_rate": 0.0001,
2989
+ "loss": 0.0041,
2990
+ "step": 1816
2991
+ },
2992
+ {
2993
+ "epoch": 606.58,
2994
+ "learning_rate": 0.0001,
2995
+ "loss": 0.0006,
2996
+ "step": 1820
2997
+ },
2998
+ {
2999
+ "epoch": 607.88,
3000
+ "learning_rate": 0.0001,
3001
+ "loss": 0.0006,
3002
+ "step": 1824
3003
+ },
3004
+ {
3005
+ "epoch": 609.29,
3006
+ "learning_rate": 0.0001,
3007
+ "loss": 0.0003,
3008
+ "step": 1828
3009
+ },
3010
+ {
3011
+ "epoch": 610.58,
3012
+ "learning_rate": 0.0001,
3013
+ "loss": 0.0004,
3014
+ "step": 1832
3015
+ },
3016
+ {
3017
+ "epoch": 611.88,
3018
+ "learning_rate": 0.0001,
3019
+ "loss": 0.0007,
3020
+ "step": 1836
3021
+ },
3022
+ {
3023
+ "epoch": 613.29,
3024
+ "learning_rate": 0.0001,
3025
+ "loss": 0.0004,
3026
+ "step": 1840
3027
+ },
3028
+ {
3029
+ "epoch": 614.58,
3030
+ "learning_rate": 0.0001,
3031
+ "loss": 0.0002,
3032
+ "step": 1844
3033
+ },
3034
+ {
3035
+ "epoch": 615.88,
3036
+ "learning_rate": 0.0001,
3037
+ "loss": 0.0007,
3038
+ "step": 1848
3039
+ },
3040
+ {
3041
+ "epoch": 617.29,
3042
+ "learning_rate": 0.0001,
3043
+ "loss": 0.0012,
3044
+ "step": 1852
3045
+ },
3046
+ {
3047
+ "epoch": 618.58,
3048
+ "learning_rate": 0.0001,
3049
+ "loss": 0.0004,
3050
+ "step": 1856
3051
+ },
3052
+ {
3053
+ "epoch": 618.58,
3054
+ "eval_exact_match": 0.6982591876208898,
3055
+ "eval_exec": 0.683752417794971,
3056
+ "eval_loss": 0.4248420000076294,
3057
+ "eval_runtime": 1371.1332,
3058
+ "eval_samples_per_second": 0.754,
3059
+ "step": 1856
3060
+ },
3061
+ {
3062
+ "epoch": 619.88,
3063
+ "learning_rate": 0.0001,
3064
+ "loss": 0.0003,
3065
+ "step": 1860
3066
+ },
3067
+ {
3068
+ "epoch": 621.29,
3069
+ "learning_rate": 0.0001,
3070
+ "loss": 0.0002,
3071
+ "step": 1864
3072
+ },
3073
+ {
3074
+ "epoch": 622.58,
3075
+ "learning_rate": 0.0001,
3076
+ "loss": 0.0002,
3077
+ "step": 1868
3078
+ },
3079
+ {
3080
+ "epoch": 623.88,
3081
+ "learning_rate": 0.0001,
3082
+ "loss": 0.0005,
3083
+ "step": 1872
3084
+ },
3085
+ {
3086
+ "epoch": 625.29,
3087
+ "learning_rate": 0.0001,
3088
+ "loss": 0.0002,
3089
+ "step": 1876
3090
+ },
3091
+ {
3092
+ "epoch": 626.58,
3093
+ "learning_rate": 0.0001,
3094
+ "loss": 0.0002,
3095
+ "step": 1880
3096
+ },
3097
+ {
3098
+ "epoch": 627.88,
3099
+ "learning_rate": 0.0001,
3100
+ "loss": 0.0002,
3101
+ "step": 1884
3102
+ },
3103
+ {
3104
+ "epoch": 629.29,
3105
+ "learning_rate": 0.0001,
3106
+ "loss": 0.0001,
3107
+ "step": 1888
3108
+ },
3109
+ {
3110
+ "epoch": 630.58,
3111
+ "learning_rate": 0.0001,
3112
+ "loss": 0.0004,
3113
+ "step": 1892
3114
+ },
3115
+ {
3116
+ "epoch": 631.88,
3117
+ "learning_rate": 0.0001,
3118
+ "loss": 0.0002,
3119
+ "step": 1896
3120
+ },
3121
+ {
3122
+ "epoch": 633.29,
3123
+ "learning_rate": 0.0001,
3124
+ "loss": 0.0003,
3125
+ "step": 1900
3126
+ },
3127
+ {
3128
+ "epoch": 634.58,
3129
+ "learning_rate": 0.0001,
3130
+ "loss": 0.0004,
3131
+ "step": 1904
3132
+ },
3133
+ {
3134
+ "epoch": 635.88,
3135
+ "learning_rate": 0.0001,
3136
+ "loss": 0.0004,
3137
+ "step": 1908
3138
+ },
3139
+ {
3140
+ "epoch": 637.29,
3141
+ "learning_rate": 0.0001,
3142
+ "loss": 0.0003,
3143
+ "step": 1912
3144
+ },
3145
+ {
3146
+ "epoch": 638.58,
3147
+ "learning_rate": 0.0001,
3148
+ "loss": 0.0002,
3149
+ "step": 1916
3150
+ },
3151
+ {
3152
+ "epoch": 639.88,
3153
+ "learning_rate": 0.0001,
3154
+ "loss": 0.0004,
3155
+ "step": 1920
3156
+ },
3157
+ {
3158
+ "epoch": 639.88,
3159
+ "eval_exact_match": 0.6847195357833655,
3160
+ "eval_exec": 0.6769825918762089,
3161
+ "eval_loss": 0.45460787415504456,
3162
+ "eval_runtime": 1038.9518,
3163
+ "eval_samples_per_second": 0.995,
3164
+ "step": 1920
3165
+ },
3166
+ {
3167
+ "epoch": 641.29,
3168
+ "learning_rate": 0.0001,
3169
+ "loss": 0.0003,
3170
+ "step": 1924
3171
+ },
3172
+ {
3173
+ "epoch": 642.58,
3174
+ "learning_rate": 0.0001,
3175
+ "loss": 0.0004,
3176
+ "step": 1928
3177
+ },
3178
+ {
3179
+ "epoch": 643.88,
3180
+ "learning_rate": 0.0001,
3181
+ "loss": 0.0003,
3182
+ "step": 1932
3183
+ },
3184
+ {
3185
+ "epoch": 645.29,
3186
+ "learning_rate": 0.0001,
3187
+ "loss": 0.0006,
3188
+ "step": 1936
3189
+ },
3190
+ {
3191
+ "epoch": 646.58,
3192
+ "learning_rate": 0.0001,
3193
+ "loss": 0.0006,
3194
+ "step": 1940
3195
+ },
3196
+ {
3197
+ "epoch": 647.88,
3198
+ "learning_rate": 0.0001,
3199
+ "loss": 0.0004,
3200
+ "step": 1944
3201
+ },
3202
+ {
3203
+ "epoch": 649.29,
3204
+ "learning_rate": 0.0001,
3205
+ "loss": 0.0002,
3206
+ "step": 1948
3207
+ },
3208
+ {
3209
+ "epoch": 650.58,
3210
+ "learning_rate": 0.0001,
3211
+ "loss": 0.0002,
3212
+ "step": 1952
3213
+ },
3214
+ {
3215
+ "epoch": 651.88,
3216
+ "learning_rate": 0.0001,
3217
+ "loss": 0.0005,
3218
+ "step": 1956
3219
+ },
3220
+ {
3221
+ "epoch": 653.29,
3222
+ "learning_rate": 0.0001,
3223
+ "loss": 0.0006,
3224
+ "step": 1960
3225
+ },
3226
+ {
3227
+ "epoch": 654.58,
3228
+ "learning_rate": 0.0001,
3229
+ "loss": 0.0003,
3230
+ "step": 1964
3231
+ },
3232
+ {
3233
+ "epoch": 655.88,
3234
+ "learning_rate": 0.0001,
3235
+ "loss": 0.0003,
3236
+ "step": 1968
3237
+ },
3238
+ {
3239
+ "epoch": 657.29,
3240
+ "learning_rate": 0.0001,
3241
+ "loss": 0.0003,
3242
+ "step": 1972
3243
+ },
3244
+ {
3245
+ "epoch": 658.58,
3246
+ "learning_rate": 0.0001,
3247
+ "loss": 0.0003,
3248
+ "step": 1976
3249
+ },
3250
+ {
3251
+ "epoch": 659.88,
3252
+ "learning_rate": 0.0001,
3253
+ "loss": 0.0003,
3254
+ "step": 1980
3255
+ },
3256
+ {
3257
+ "epoch": 661.29,
3258
+ "learning_rate": 0.0001,
3259
+ "loss": 0.0007,
3260
+ "step": 1984
3261
+ },
3262
+ {
3263
+ "epoch": 661.29,
3264
+ "eval_exact_match": 0.6856866537717602,
3265
+ "eval_exec": 0.6808510638297872,
3266
+ "eval_loss": 0.4657697379589081,
3267
+ "eval_runtime": 1097.6691,
3268
+ "eval_samples_per_second": 0.942,
3269
+ "step": 1984
3270
+ },
3271
+ {
3272
+ "epoch": 662.58,
3273
+ "learning_rate": 0.0001,
3274
+ "loss": 0.0004,
3275
+ "step": 1988
3276
+ },
3277
+ {
3278
+ "epoch": 663.88,
3279
+ "learning_rate": 0.0001,
3280
+ "loss": 0.0002,
3281
+ "step": 1992
3282
+ },
3283
+ {
3284
+ "epoch": 665.29,
3285
+ "learning_rate": 0.0001,
3286
+ "loss": 0.0002,
3287
+ "step": 1996
3288
+ },
3289
+ {
3290
+ "epoch": 666.58,
3291
+ "learning_rate": 0.0001,
3292
+ "loss": 0.0002,
3293
+ "step": 2000
3294
+ },
3295
+ {
3296
+ "epoch": 667.88,
3297
+ "learning_rate": 0.0001,
3298
+ "loss": 0.0001,
3299
+ "step": 2004
3300
+ },
3301
+ {
3302
+ "epoch": 669.29,
3303
+ "learning_rate": 0.0001,
3304
+ "loss": 0.0001,
3305
+ "step": 2008
3306
+ },
3307
+ {
3308
+ "epoch": 670.58,
3309
+ "learning_rate": 0.0001,
3310
+ "loss": 0.0002,
3311
+ "step": 2012
3312
+ },
3313
+ {
3314
+ "epoch": 671.88,
3315
+ "learning_rate": 0.0001,
3316
+ "loss": 0.0002,
3317
+ "step": 2016
3318
+ },
3319
+ {
3320
+ "epoch": 673.29,
3321
+ "learning_rate": 0.0001,
3322
+ "loss": 0.0002,
3323
+ "step": 2020
3324
+ },
3325
+ {
3326
+ "epoch": 674.58,
3327
+ "learning_rate": 0.0001,
3328
+ "loss": 0.0004,
3329
+ "step": 2024
3330
+ },
3331
+ {
3332
+ "epoch": 675.88,
3333
+ "learning_rate": 0.0001,
3334
+ "loss": 0.0001,
3335
+ "step": 2028
3336
+ },
3337
+ {
3338
+ "epoch": 677.29,
3339
+ "learning_rate": 0.0001,
3340
+ "loss": 0.0001,
3341
+ "step": 2032
3342
+ },
3343
+ {
3344
+ "epoch": 678.58,
3345
+ "learning_rate": 0.0001,
3346
+ "loss": 0.0003,
3347
+ "step": 2036
3348
+ },
3349
+ {
3350
+ "epoch": 679.88,
3351
+ "learning_rate": 0.0001,
3352
+ "loss": 0.0001,
3353
+ "step": 2040
3354
+ },
3355
+ {
3356
+ "epoch": 681.29,
3357
+ "learning_rate": 0.0001,
3358
+ "loss": 0.0003,
3359
+ "step": 2044
3360
+ },
3361
+ {
3362
+ "epoch": 682.58,
3363
+ "learning_rate": 0.0001,
3364
+ "loss": 0.0037,
3365
+ "step": 2048
3366
+ },
3367
+ {
3368
+ "epoch": 682.58,
3369
+ "eval_exact_match": 0.6914893617021277,
3370
+ "eval_exec": 0.690522243713733,
3371
+ "eval_loss": 0.4180769622325897,
3372
+ "eval_runtime": 1092.62,
3373
+ "eval_samples_per_second": 0.946,
3374
+ "step": 2048
3375
+ },
3376
+ {
3377
+ "epoch": 683.88,
3378
+ "learning_rate": 0.0001,
3379
+ "loss": 0.0004,
3380
+ "step": 2052
3381
+ },
3382
+ {
3383
+ "epoch": 685.29,
3384
+ "learning_rate": 0.0001,
3385
+ "loss": 0.0002,
3386
+ "step": 2056
3387
+ },
3388
+ {
3389
+ "epoch": 686.58,
3390
+ "learning_rate": 0.0001,
3391
+ "loss": 0.0003,
3392
+ "step": 2060
3393
+ },
3394
+ {
3395
+ "epoch": 687.88,
3396
+ "learning_rate": 0.0001,
3397
+ "loss": 0.0011,
3398
+ "step": 2064
3399
+ },
3400
+ {
3401
+ "epoch": 689.29,
3402
+ "learning_rate": 0.0001,
3403
+ "loss": 0.0014,
3404
+ "step": 2068
3405
+ },
3406
+ {
3407
+ "epoch": 690.58,
3408
+ "learning_rate": 0.0001,
3409
+ "loss": 0.0003,
3410
+ "step": 2072
3411
+ },
3412
+ {
3413
+ "epoch": 691.88,
3414
+ "learning_rate": 0.0001,
3415
+ "loss": 0.0002,
3416
+ "step": 2076
3417
+ },
3418
+ {
3419
+ "epoch": 693.29,
3420
+ "learning_rate": 0.0001,
3421
+ "loss": 0.0003,
3422
+ "step": 2080
3423
+ },
3424
+ {
3425
+ "epoch": 694.58,
3426
+ "learning_rate": 0.0001,
3427
+ "loss": 0.0005,
3428
+ "step": 2084
3429
+ },
3430
+ {
3431
+ "epoch": 695.88,
3432
+ "learning_rate": 0.0001,
3433
+ "loss": 0.0004,
3434
+ "step": 2088
3435
+ },
3436
+ {
3437
+ "epoch": 697.29,
3438
+ "learning_rate": 0.0001,
3439
+ "loss": 0.0003,
3440
+ "step": 2092
3441
+ },
3442
+ {
3443
+ "epoch": 698.58,
3444
+ "learning_rate": 0.0001,
3445
+ "loss": 0.0012,
3446
+ "step": 2096
3447
+ },
3448
+ {
3449
+ "epoch": 699.88,
3450
+ "learning_rate": 0.0001,
3451
+ "loss": 0.0007,
3452
+ "step": 2100
3453
+ },
3454
+ {
3455
+ "epoch": 701.29,
3456
+ "learning_rate": 0.0001,
3457
+ "loss": 0.0002,
3458
+ "step": 2104
3459
+ },
3460
+ {
3461
+ "epoch": 702.58,
3462
+ "learning_rate": 0.0001,
3463
+ "loss": 0.0004,
3464
+ "step": 2108
3465
+ },
3466
+ {
3467
+ "epoch": 703.88,
3468
+ "learning_rate": 0.0001,
3469
+ "loss": 0.0003,
3470
+ "step": 2112
3471
+ },
3472
+ {
3473
+ "epoch": 703.88,
3474
+ "eval_exact_match": 0.6943907156673114,
3475
+ "eval_exec": 0.6924564796905223,
3476
+ "eval_loss": 0.44465285539627075,
3477
+ "eval_runtime": 999.8273,
3478
+ "eval_samples_per_second": 1.034,
3479
+ "step": 2112
3480
+ },
3481
+ {
3482
+ "epoch": 705.29,
3483
+ "learning_rate": 0.0001,
3484
+ "loss": 0.0008,
3485
+ "step": 2116
3486
+ },
3487
+ {
3488
+ "epoch": 706.58,
3489
+ "learning_rate": 0.0001,
3490
+ "loss": 0.0005,
3491
+ "step": 2120
3492
+ },
3493
+ {
3494
+ "epoch": 707.88,
3495
+ "learning_rate": 0.0001,
3496
+ "loss": 0.0002,
3497
+ "step": 2124
3498
+ },
3499
+ {
3500
+ "epoch": 709.29,
3501
+ "learning_rate": 0.0001,
3502
+ "loss": 0.0002,
3503
+ "step": 2128
3504
+ },
3505
+ {
3506
+ "epoch": 710.58,
3507
+ "learning_rate": 0.0001,
3508
+ "loss": 0.0003,
3509
+ "step": 2132
3510
+ },
3511
+ {
3512
+ "epoch": 711.88,
3513
+ "learning_rate": 0.0001,
3514
+ "loss": 0.0001,
3515
+ "step": 2136
3516
+ },
3517
+ {
3518
+ "epoch": 713.29,
3519
+ "learning_rate": 0.0001,
3520
+ "loss": 0.0004,
3521
+ "step": 2140
3522
+ },
3523
+ {
3524
+ "epoch": 714.58,
3525
+ "learning_rate": 0.0001,
3526
+ "loss": 0.0053,
3527
+ "step": 2144
3528
+ },
3529
+ {
3530
+ "epoch": 715.88,
3531
+ "learning_rate": 0.0001,
3532
+ "loss": 0.0006,
3533
+ "step": 2148
3534
+ },
3535
+ {
3536
+ "epoch": 717.29,
3537
+ "learning_rate": 0.0001,
3538
+ "loss": 0.0002,
3539
+ "step": 2152
3540
+ },
3541
+ {
3542
+ "epoch": 718.58,
3543
+ "learning_rate": 0.0001,
3544
+ "loss": 0.0002,
3545
+ "step": 2156
3546
+ },
3547
+ {
3548
+ "epoch": 719.88,
3549
+ "learning_rate": 0.0001,
3550
+ "loss": 0.0002,
3551
+ "step": 2160
3552
+ },
3553
+ {
3554
+ "epoch": 721.29,
3555
+ "learning_rate": 0.0001,
3556
+ "loss": 0.0003,
3557
+ "step": 2164
3558
+ },
3559
+ {
3560
+ "epoch": 722.58,
3561
+ "learning_rate": 0.0001,
3562
+ "loss": 0.0013,
3563
+ "step": 2168
3564
+ },
3565
+ {
3566
+ "epoch": 723.88,
3567
+ "learning_rate": 0.0001,
3568
+ "loss": 0.0016,
3569
+ "step": 2172
3570
+ },
3571
+ {
3572
+ "epoch": 725.29,
3573
+ "learning_rate": 0.0001,
3574
+ "loss": 0.0003,
3575
+ "step": 2176
3576
+ },
3577
+ {
3578
+ "epoch": 725.29,
3579
+ "eval_exact_match": 0.6992263056092843,
3580
+ "eval_exec": 0.6876208897485493,
3581
+ "eval_loss": 0.4321630299091339,
3582
+ "eval_runtime": 878.8579,
3583
+ "eval_samples_per_second": 1.177,
3584
+ "step": 2176
3585
+ },
3586
+ {
3587
+ "epoch": 726.58,
3588
+ "learning_rate": 0.0001,
3589
+ "loss": 0.0002,
3590
+ "step": 2180
3591
+ },
3592
+ {
3593
+ "epoch": 727.88,
3594
+ "learning_rate": 0.0001,
3595
+ "loss": 0.0002,
3596
+ "step": 2184
3597
+ },
3598
+ {
3599
+ "epoch": 729.29,
3600
+ "learning_rate": 0.0001,
3601
+ "loss": 0.0002,
3602
+ "step": 2188
3603
+ },
3604
+ {
3605
+ "epoch": 730.58,
3606
+ "learning_rate": 0.0001,
3607
+ "loss": 0.0002,
3608
+ "step": 2192
3609
+ },
3610
+ {
3611
+ "epoch": 731.88,
3612
+ "learning_rate": 0.0001,
3613
+ "loss": 0.0001,
3614
+ "step": 2196
3615
+ },
3616
+ {
3617
+ "epoch": 733.29,
3618
+ "learning_rate": 0.0001,
3619
+ "loss": 0.0007,
3620
+ "step": 2200
3621
+ },
3622
+ {
3623
+ "epoch": 734.58,
3624
+ "learning_rate": 0.0001,
3625
+ "loss": 0.0002,
3626
+ "step": 2204
3627
+ },
3628
+ {
3629
+ "epoch": 735.88,
3630
+ "learning_rate": 0.0001,
3631
+ "loss": 0.0004,
3632
+ "step": 2208
3633
+ },
3634
+ {
3635
+ "epoch": 737.29,
3636
+ "learning_rate": 0.0001,
3637
+ "loss": 0.0002,
3638
+ "step": 2212
3639
+ },
3640
+ {
3641
+ "epoch": 738.58,
3642
+ "learning_rate": 0.0001,
3643
+ "loss": 0.0002,
3644
+ "step": 2216
3645
+ },
3646
+ {
3647
+ "epoch": 739.88,
3648
+ "learning_rate": 0.0001,
3649
+ "loss": 0.0002,
3650
+ "step": 2220
3651
+ },
3652
+ {
3653
+ "epoch": 741.29,
3654
+ "learning_rate": 0.0001,
3655
+ "loss": 0.0003,
3656
+ "step": 2224
3657
+ },
3658
+ {
3659
+ "epoch": 742.58,
3660
+ "learning_rate": 0.0001,
3661
+ "loss": 0.0003,
3662
+ "step": 2228
3663
+ },
3664
+ {
3665
+ "epoch": 743.88,
3666
+ "learning_rate": 0.0001,
3667
+ "loss": 0.0001,
3668
+ "step": 2232
3669
+ },
3670
+ {
3671
+ "epoch": 745.29,
3672
+ "learning_rate": 0.0001,
3673
+ "loss": 0.0001,
3674
+ "step": 2236
3675
+ },
3676
+ {
3677
+ "epoch": 746.58,
3678
+ "learning_rate": 0.0001,
3679
+ "loss": 0.0001,
3680
+ "step": 2240
3681
+ },
3682
+ {
3683
+ "epoch": 746.58,
3684
+ "eval_exact_match": 0.6963249516441006,
3685
+ "eval_exec": 0.6789168278529981,
3686
+ "eval_loss": 0.45437225699424744,
3687
+ "eval_runtime": 864.1019,
3688
+ "eval_samples_per_second": 1.197,
3689
+ "step": 2240
3690
+ },
3691
+ {
3692
+ "epoch": 747.88,
3693
+ "learning_rate": 0.0001,
3694
+ "loss": 0.0003,
3695
+ "step": 2244
3696
+ },
3697
+ {
3698
+ "epoch": 749.29,
3699
+ "learning_rate": 0.0001,
3700
+ "loss": 0.0003,
3701
+ "step": 2248
3702
+ },
3703
+ {
3704
+ "epoch": 750.58,
3705
+ "learning_rate": 0.0001,
3706
+ "loss": 0.0006,
3707
+ "step": 2252
3708
+ },
3709
+ {
3710
+ "epoch": 751.88,
3711
+ "learning_rate": 0.0001,
3712
+ "loss": 0.0001,
3713
+ "step": 2256
3714
+ },
3715
+ {
3716
+ "epoch": 753.29,
3717
+ "learning_rate": 0.0001,
3718
+ "loss": 0.0002,
3719
+ "step": 2260
3720
+ },
3721
+ {
3722
+ "epoch": 754.58,
3723
+ "learning_rate": 0.0001,
3724
+ "loss": 0.0001,
3725
+ "step": 2264
3726
+ },
3727
+ {
3728
+ "epoch": 755.88,
3729
+ "learning_rate": 0.0001,
3730
+ "loss": 0.0001,
3731
+ "step": 2268
3732
+ },
3733
+ {
3734
+ "epoch": 757.29,
3735
+ "learning_rate": 0.0001,
3736
+ "loss": 0.0003,
3737
+ "step": 2272
3738
+ },
3739
+ {
3740
+ "epoch": 758.58,
3741
+ "learning_rate": 0.0001,
3742
+ "loss": 0.0001,
3743
+ "step": 2276
3744
+ },
3745
+ {
3746
+ "epoch": 759.88,
3747
+ "learning_rate": 0.0001,
3748
+ "loss": 0.0001,
3749
+ "step": 2280
3750
+ },
3751
+ {
3752
+ "epoch": 761.29,
3753
+ "learning_rate": 0.0001,
3754
+ "loss": 0.0002,
3755
+ "step": 2284
3756
+ },
3757
+ {
3758
+ "epoch": 762.58,
3759
+ "learning_rate": 0.0001,
3760
+ "loss": 0.0001,
3761
+ "step": 2288
3762
+ },
3763
+ {
3764
+ "epoch": 763.88,
3765
+ "learning_rate": 0.0001,
3766
+ "loss": 0.0002,
3767
+ "step": 2292
3768
+ },
3769
+ {
3770
+ "epoch": 765.29,
3771
+ "learning_rate": 0.0001,
3772
+ "loss": 0.0003,
3773
+ "step": 2296
3774
+ },
3775
+ {
3776
+ "epoch": 766.58,
3777
+ "learning_rate": 0.0001,
3778
+ "loss": 0.0002,
3779
+ "step": 2300
3780
+ },
3781
+ {
3782
+ "epoch": 767.88,
3783
+ "learning_rate": 0.0001,
3784
+ "loss": 0.0002,
3785
+ "step": 2304
3786
+ },
3787
+ {
3788
+ "epoch": 767.88,
3789
+ "eval_exact_match": 0.6963249516441006,
3790
+ "eval_exec": 0.6866537717601547,
3791
+ "eval_loss": 0.4832020401954651,
3792
+ "eval_runtime": 816.387,
3793
+ "eval_samples_per_second": 1.267,
3794
+ "step": 2304
3795
+ },
3796
+ {
3797
+ "epoch": 769.29,
3798
+ "learning_rate": 0.0001,
3799
+ "loss": 0.0001,
3800
+ "step": 2308
3801
+ },
3802
+ {
3803
+ "epoch": 770.58,
3804
+ "learning_rate": 0.0001,
3805
+ "loss": 0.0005,
3806
+ "step": 2312
3807
+ },
3808
+ {
3809
+ "epoch": 771.88,
3810
+ "learning_rate": 0.0001,
3811
+ "loss": 0.0002,
3812
+ "step": 2316
3813
+ },
3814
+ {
3815
+ "epoch": 773.29,
3816
+ "learning_rate": 0.0001,
3817
+ "loss": 0.0006,
3818
+ "step": 2320
3819
+ },
3820
+ {
3821
+ "epoch": 774.58,
3822
+ "learning_rate": 0.0001,
3823
+ "loss": 0.0001,
3824
+ "step": 2324
3825
+ },
3826
+ {
3827
+ "epoch": 775.88,
3828
+ "learning_rate": 0.0001,
3829
+ "loss": 0.0009,
3830
+ "step": 2328
3831
+ },
3832
+ {
3833
+ "epoch": 777.29,
3834
+ "learning_rate": 0.0001,
3835
+ "loss": 0.0211,
3836
+ "step": 2332
3837
+ },
3838
+ {
3839
+ "epoch": 778.58,
3840
+ "learning_rate": 0.0001,
3841
+ "loss": 0.0001,
3842
+ "step": 2336
3843
+ },
3844
+ {
3845
+ "epoch": 779.88,
3846
+ "learning_rate": 0.0001,
3847
+ "loss": 0.0002,
3848
+ "step": 2340
3849
+ },
3850
+ {
3851
+ "epoch": 781.29,
3852
+ "learning_rate": 0.0001,
3853
+ "loss": 0.0001,
3854
+ "step": 2344
3855
+ },
3856
+ {
3857
+ "epoch": 782.58,
3858
+ "learning_rate": 0.0001,
3859
+ "loss": 0.0001,
3860
+ "step": 2348
3861
+ },
3862
+ {
3863
+ "epoch": 783.88,
3864
+ "learning_rate": 0.0001,
3865
+ "loss": 0.0001,
3866
+ "step": 2352
3867
+ },
3868
+ {
3869
+ "epoch": 785.29,
3870
+ "learning_rate": 0.0001,
3871
+ "loss": 0.0009,
3872
+ "step": 2356
3873
+ },
3874
+ {
3875
+ "epoch": 786.58,
3876
+ "learning_rate": 0.0001,
3877
+ "loss": 0.0002,
3878
+ "step": 2360
3879
+ },
3880
+ {
3881
+ "epoch": 787.88,
3882
+ "learning_rate": 0.0001,
3883
+ "loss": 0.0001,
3884
+ "step": 2364
3885
+ },
3886
+ {
3887
+ "epoch": 789.29,
3888
+ "learning_rate": 0.0001,
3889
+ "loss": 0.0002,
3890
+ "step": 2368
3891
+ },
3892
+ {
3893
+ "epoch": 789.29,
3894
+ "eval_exact_match": 0.6924564796905223,
3895
+ "eval_exec": 0.6779497098646035,
3896
+ "eval_loss": 0.47273918986320496,
3897
+ "eval_runtime": 826.2169,
3898
+ "eval_samples_per_second": 1.251,
3899
+ "step": 2368
3900
+ },
3901
+ {
3902
+ "epoch": 790.58,
3903
+ "learning_rate": 0.0001,
3904
+ "loss": 0.0001,
3905
+ "step": 2372
3906
+ },
3907
+ {
3908
+ "epoch": 791.88,
3909
+ "learning_rate": 0.0001,
3910
+ "loss": 0.0001,
3911
+ "step": 2376
3912
+ },
3913
+ {
3914
+ "epoch": 793.29,
3915
+ "learning_rate": 0.0001,
3916
+ "loss": 0.0004,
3917
+ "step": 2380
3918
+ },
3919
+ {
3920
+ "epoch": 794.58,
3921
+ "learning_rate": 0.0001,
3922
+ "loss": 0.0002,
3923
+ "step": 2384
3924
+ },
3925
+ {
3926
+ "epoch": 795.88,
3927
+ "learning_rate": 0.0001,
3928
+ "loss": 0.0002,
3929
+ "step": 2388
3930
+ },
3931
+ {
3932
+ "epoch": 797.29,
3933
+ "learning_rate": 0.0001,
3934
+ "loss": 0.0001,
3935
+ "step": 2392
3936
+ },
3937
+ {
3938
+ "epoch": 798.58,
3939
+ "learning_rate": 0.0001,
3940
+ "loss": 0.0001,
3941
+ "step": 2396
3942
+ },
3943
+ {
3944
+ "epoch": 799.88,
3945
+ "learning_rate": 0.0001,
3946
+ "loss": 0.0001,
3947
+ "step": 2400
3948
+ },
3949
+ {
3950
+ "epoch": 801.29,
3951
+ "learning_rate": 0.0001,
3952
+ "loss": 0.0003,
3953
+ "step": 2404
3954
+ },
3955
+ {
3956
+ "epoch": 802.58,
3957
+ "learning_rate": 0.0001,
3958
+ "loss": 0.0002,
3959
+ "step": 2408
3960
+ },
3961
+ {
3962
+ "epoch": 803.88,
3963
+ "learning_rate": 0.0001,
3964
+ "loss": 0.0001,
3965
+ "step": 2412
3966
+ },
3967
+ {
3968
+ "epoch": 805.29,
3969
+ "learning_rate": 0.0001,
3970
+ "loss": 0.0,
3971
+ "step": 2416
3972
+ },
3973
+ {
3974
+ "epoch": 806.58,
3975
+ "learning_rate": 0.0001,
3976
+ "loss": 0.0001,
3977
+ "step": 2420
3978
+ },
3979
+ {
3980
+ "epoch": 807.88,
3981
+ "learning_rate": 0.0001,
3982
+ "loss": 0.0,
3983
+ "step": 2424
3984
+ },
3985
+ {
3986
+ "epoch": 809.29,
3987
+ "learning_rate": 0.0001,
3988
+ "loss": 0.0003,
3989
+ "step": 2428
3990
+ },
3991
+ {
3992
+ "epoch": 810.58,
3993
+ "learning_rate": 0.0001,
3994
+ "loss": 0.0001,
3995
+ "step": 2432
3996
+ },
3997
+ {
3998
+ "epoch": 810.58,
3999
+ "eval_exact_match": 0.6808510638297872,
4000
+ "eval_exec": 0.6692456479690522,
4001
+ "eval_loss": 0.48608434200286865,
4002
+ "eval_runtime": 825.1274,
4003
+ "eval_samples_per_second": 1.253,
4004
+ "step": 2432
4005
+ },
4006
+ {
4007
+ "epoch": 811.88,
4008
+ "learning_rate": 0.0001,
4009
+ "loss": 0.0004,
4010
+ "step": 2436
4011
+ },
4012
+ {
4013
+ "epoch": 813.29,
4014
+ "learning_rate": 0.0001,
4015
+ "loss": 0.0002,
4016
+ "step": 2440
4017
+ },
4018
+ {
4019
+ "epoch": 814.58,
4020
+ "learning_rate": 0.0001,
4021
+ "loss": 0.0002,
4022
+ "step": 2444
4023
+ },
4024
+ {
4025
+ "epoch": 815.88,
4026
+ "learning_rate": 0.0001,
4027
+ "loss": 0.0003,
4028
+ "step": 2448
4029
+ },
4030
+ {
4031
+ "epoch": 817.29,
4032
+ "learning_rate": 0.0001,
4033
+ "loss": 0.0006,
4034
+ "step": 2452
4035
+ },
4036
+ {
4037
+ "epoch": 818.58,
4038
+ "learning_rate": 0.0001,
4039
+ "loss": 0.0001,
4040
+ "step": 2456
4041
+ },
4042
+ {
4043
+ "epoch": 819.88,
4044
+ "learning_rate": 0.0001,
4045
+ "loss": 0.0001,
4046
+ "step": 2460
4047
+ },
4048
+ {
4049
+ "epoch": 821.29,
4050
+ "learning_rate": 0.0001,
4051
+ "loss": 0.0003,
4052
+ "step": 2464
4053
+ },
4054
+ {
4055
+ "epoch": 822.58,
4056
+ "learning_rate": 0.0001,
4057
+ "loss": 0.0004,
4058
+ "step": 2468
4059
+ },
4060
+ {
4061
+ "epoch": 823.88,
4062
+ "learning_rate": 0.0001,
4063
+ "loss": 0.0001,
4064
+ "step": 2472
4065
+ },
4066
+ {
4067
+ "epoch": 825.29,
4068
+ "learning_rate": 0.0001,
4069
+ "loss": 0.0001,
4070
+ "step": 2476
4071
+ },
4072
+ {
4073
+ "epoch": 826.58,
4074
+ "learning_rate": 0.0001,
4075
+ "loss": 0.0001,
4076
+ "step": 2480
4077
+ },
4078
+ {
4079
+ "epoch": 827.88,
4080
+ "learning_rate": 0.0001,
4081
+ "loss": 0.0001,
4082
+ "step": 2484
4083
+ },
4084
+ {
4085
+ "epoch": 829.29,
4086
+ "learning_rate": 0.0001,
4087
+ "loss": 0.0001,
4088
+ "step": 2488
4089
+ },
4090
+ {
4091
+ "epoch": 830.58,
4092
+ "learning_rate": 0.0001,
4093
+ "loss": 0.0005,
4094
+ "step": 2492
4095
+ },
4096
+ {
4097
+ "epoch": 831.88,
4098
+ "learning_rate": 0.0001,
4099
+ "loss": 0.0003,
4100
+ "step": 2496
4101
+ },
4102
+ {
4103
+ "epoch": 831.88,
4104
+ "eval_exact_match": 0.6934235976789168,
4105
+ "eval_exec": 0.6750483558994197,
4106
+ "eval_loss": 0.47877049446105957,
4107
+ "eval_runtime": 803.416,
4108
+ "eval_samples_per_second": 1.287,
4109
+ "step": 2496
4110
+ },
4111
+ {
4112
+ "epoch": 833.29,
4113
+ "learning_rate": 0.0001,
4114
+ "loss": 0.0001,
4115
+ "step": 2500
4116
+ },
4117
+ {
4118
+ "epoch": 834.58,
4119
+ "learning_rate": 0.0001,
4120
+ "loss": 0.0002,
4121
+ "step": 2504
4122
+ },
4123
+ {
4124
+ "epoch": 835.88,
4125
+ "learning_rate": 0.0001,
4126
+ "loss": 0.0001,
4127
+ "step": 2508
4128
+ },
4129
+ {
4130
+ "epoch": 837.29,
4131
+ "learning_rate": 0.0001,
4132
+ "loss": 0.0005,
4133
+ "step": 2512
4134
+ },
4135
+ {
4136
+ "epoch": 838.58,
4137
+ "learning_rate": 0.0001,
4138
+ "loss": 0.0002,
4139
+ "step": 2516
4140
+ },
4141
+ {
4142
+ "epoch": 839.88,
4143
+ "learning_rate": 0.0001,
4144
+ "loss": 0.0002,
4145
+ "step": 2520
4146
+ },
4147
+ {
4148
+ "epoch": 841.29,
4149
+ "learning_rate": 0.0001,
4150
+ "loss": 0.0001,
4151
+ "step": 2524
4152
+ },
4153
+ {
4154
+ "epoch": 842.58,
4155
+ "learning_rate": 0.0001,
4156
+ "loss": 0.0002,
4157
+ "step": 2528
4158
+ },
4159
+ {
4160
+ "epoch": 843.88,
4161
+ "learning_rate": 0.0001,
4162
+ "loss": 0.0006,
4163
+ "step": 2532
4164
+ },
4165
+ {
4166
+ "epoch": 845.29,
4167
+ "learning_rate": 0.0001,
4168
+ "loss": 0.0002,
4169
+ "step": 2536
4170
+ },
4171
+ {
4172
+ "epoch": 846.58,
4173
+ "learning_rate": 0.0001,
4174
+ "loss": 0.0001,
4175
+ "step": 2540
4176
+ },
4177
+ {
4178
+ "epoch": 847.88,
4179
+ "learning_rate": 0.0001,
4180
+ "loss": 0.0004,
4181
+ "step": 2544
4182
+ },
4183
+ {
4184
+ "epoch": 849.29,
4185
+ "learning_rate": 0.0001,
4186
+ "loss": 0.0002,
4187
+ "step": 2548
4188
+ },
4189
+ {
4190
+ "epoch": 850.58,
4191
+ "learning_rate": 0.0001,
4192
+ "loss": 0.0001,
4193
+ "step": 2552
4194
+ },
4195
+ {
4196
+ "epoch": 851.88,
4197
+ "learning_rate": 0.0001,
4198
+ "loss": 0.0003,
4199
+ "step": 2556
4200
+ },
4201
+ {
4202
+ "epoch": 853.29,
4203
+ "learning_rate": 0.0001,
4204
+ "loss": 0.0001,
4205
+ "step": 2560
4206
+ },
4207
+ {
4208
+ "epoch": 853.29,
4209
+ "eval_exact_match": 0.6914893617021277,
4210
+ "eval_exec": 0.6779497098646035,
4211
+ "eval_loss": 0.4677638113498688,
4212
+ "eval_runtime": 845.5605,
4213
+ "eval_samples_per_second": 1.223,
4214
+ "step": 2560
4215
+ },
4216
+ {
4217
+ "epoch": 854.58,
4218
+ "learning_rate": 0.0001,
4219
+ "loss": 0.0005,
4220
+ "step": 2564
4221
+ },
4222
+ {
4223
+ "epoch": 855.88,
4224
+ "learning_rate": 0.0001,
4225
+ "loss": 0.0004,
4226
+ "step": 2568
4227
+ },
4228
+ {
4229
+ "epoch": 857.29,
4230
+ "learning_rate": 0.0001,
4231
+ "loss": 0.0014,
4232
+ "step": 2572
4233
+ },
4234
+ {
4235
+ "epoch": 858.58,
4236
+ "learning_rate": 0.0001,
4237
+ "loss": 0.0004,
4238
+ "step": 2576
4239
+ },
4240
+ {
4241
+ "epoch": 859.88,
4242
+ "learning_rate": 0.0001,
4243
+ "loss": 0.0001,
4244
+ "step": 2580
4245
+ },
4246
+ {
4247
+ "epoch": 861.29,
4248
+ "learning_rate": 0.0001,
4249
+ "loss": 0.0001,
4250
+ "step": 2584
4251
+ },
4252
+ {
4253
+ "epoch": 862.58,
4254
+ "learning_rate": 0.0001,
4255
+ "loss": 0.0002,
4256
+ "step": 2588
4257
+ },
4258
+ {
4259
+ "epoch": 863.88,
4260
+ "learning_rate": 0.0001,
4261
+ "loss": 0.0002,
4262
+ "step": 2592
4263
+ },
4264
+ {
4265
+ "epoch": 865.29,
4266
+ "learning_rate": 0.0001,
4267
+ "loss": 0.0001,
4268
+ "step": 2596
4269
+ },
4270
+ {
4271
+ "epoch": 866.58,
4272
+ "learning_rate": 0.0001,
4273
+ "loss": 0.0001,
4274
+ "step": 2600
4275
+ },
4276
+ {
4277
+ "epoch": 867.88,
4278
+ "learning_rate": 0.0001,
4279
+ "loss": 0.0003,
4280
+ "step": 2604
4281
+ },
4282
+ {
4283
+ "epoch": 869.29,
4284
+ "learning_rate": 0.0001,
4285
+ "loss": 0.0001,
4286
+ "step": 2608
4287
+ },
4288
+ {
4289
+ "epoch": 870.58,
4290
+ "learning_rate": 0.0001,
4291
+ "loss": 0.0001,
4292
+ "step": 2612
4293
+ },
4294
+ {
4295
+ "epoch": 871.88,
4296
+ "learning_rate": 0.0001,
4297
+ "loss": 0.0001,
4298
+ "step": 2616
4299
+ },
4300
+ {
4301
+ "epoch": 873.29,
4302
+ "learning_rate": 0.0001,
4303
+ "loss": 0.0001,
4304
+ "step": 2620
4305
+ },
4306
+ {
4307
+ "epoch": 874.58,
4308
+ "learning_rate": 0.0001,
4309
+ "loss": 0.0,
4310
+ "step": 2624
4311
+ },
4312
+ {
4313
+ "epoch": 874.58,
4314
+ "eval_exact_match": 0.7117988394584139,
4315
+ "eval_exec": 0.695357833655706,
4316
+ "eval_loss": 0.487076997756958,
4317
+ "eval_runtime": 1063.4572,
4318
+ "eval_samples_per_second": 0.972,
4319
+ "step": 2624
4320
+ }
4321
+ ],
4322
+ "max_steps": 9216,
4323
+ "num_train_epochs": 3072,
4324
+ "total_flos": 1.20676205882696e+19,
4325
+ "trial_name": null,
4326
+ "trial_params": null
4327
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0868dff65625d28408737b7217c68e86d01510fbe1d74a54d3f52c4282da8742
3
+ size 2799