tscholak commited on
Commit
3b8574b
1 Parent(s): 41e2ad7

add T5 model

Browse files
added_tokens.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {" <=": 32100, " <": 32101}
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:cda09ede19010acdf5597f9aa600f4e7de030a2fb5b3fd20c5a312f39b0c36c4
3
+ size 5806399
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f248d39a2070137ad39fc426945d89524f2d80fc04996b4f401900de1db4c50b
3
+ size 3132653069
rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:88f0e84e8273d3f7008540ca21fb9d37b7930d993c585e8857bab8c013c5da40
3
+ size 14721
rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c3b472889188304cfe344914f068b4b9670209a206fa32e85e459df3511708f
3
+ size 14721
rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3aa5cf1f38ada54d8de53e89488070d5d47251d5cc8de1b188aebf5c8a4cdb0e
3
+ size 14721
rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:255c35cd32d56a8434e721c188783f62f2b7f0b7f63677faa4ae15f9f5ef3c46
3
+ size 14721
rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f741dddda7bb9ac6a6af1835a33db77f8dbd26c8049bf1c2afac1850d80c3ab3
3
+ size 14721
rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69f7437b0296cc28d278112ba1765fff9f238d2e74b0d0442ea2eeea95657e85
3
+ size 14721
rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d22ac590498dc65c324efd2933f59bc163624000cc11b5058abb4737b7f0a8e
3
+ size 14721
rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1d1c920374887bfa54954776d2eb2567215e3efc9e0e243be3699c6dc08f5df
3
+ size 14721
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:421ecc568bf23d96aeacaa97c29ffaf32856737b947d548836c0a8a8e9048d5d
3
+ size 623
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>"]}
spiece.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
3
+ size 791656
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 +1,3067 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.5253227408142999,
3
+ "best_model_checkpoint": "/experiments/cosql/1ccc8b19-d4cc-42ca-b70f-7d796303d2e5/checkpoint-1856",
4
+ "epoch": 231.9922480620155,
5
+ "global_step": 1856,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.12,
12
+ "learning_rate": 0.0001,
13
+ "loss": 3.2653,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.5,
18
+ "learning_rate": 0.0001,
19
+ "loss": 2.2224,
20
+ "step": 4
21
+ },
22
+ {
23
+ "epoch": 0.99,
24
+ "learning_rate": 0.0001,
25
+ "loss": 1.1652,
26
+ "step": 8
27
+ },
28
+ {
29
+ "epoch": 1.5,
30
+ "learning_rate": 0.0001,
31
+ "loss": 0.8035,
32
+ "step": 12
33
+ },
34
+ {
35
+ "epoch": 1.99,
36
+ "learning_rate": 0.0001,
37
+ "loss": 0.642,
38
+ "step": 16
39
+ },
40
+ {
41
+ "epoch": 2.5,
42
+ "learning_rate": 0.0001,
43
+ "loss": 0.537,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 2.99,
48
+ "learning_rate": 0.0001,
49
+ "loss": 0.4739,
50
+ "step": 24
51
+ },
52
+ {
53
+ "epoch": 3.5,
54
+ "learning_rate": 0.0001,
55
+ "loss": 0.4251,
56
+ "step": 28
57
+ },
58
+ {
59
+ "epoch": 3.99,
60
+ "learning_rate": 0.0001,
61
+ "loss": 0.3945,
62
+ "step": 32
63
+ },
64
+ {
65
+ "epoch": 4.5,
66
+ "learning_rate": 0.0001,
67
+ "loss": 0.352,
68
+ "step": 36
69
+ },
70
+ {
71
+ "epoch": 4.99,
72
+ "learning_rate": 0.0001,
73
+ "loss": 0.3245,
74
+ "step": 40
75
+ },
76
+ {
77
+ "epoch": 5.5,
78
+ "learning_rate": 0.0001,
79
+ "loss": 0.2948,
80
+ "step": 44
81
+ },
82
+ {
83
+ "epoch": 5.99,
84
+ "learning_rate": 0.0001,
85
+ "loss": 0.287,
86
+ "step": 48
87
+ },
88
+ {
89
+ "epoch": 6.5,
90
+ "learning_rate": 0.0001,
91
+ "loss": 0.2686,
92
+ "step": 52
93
+ },
94
+ {
95
+ "epoch": 6.99,
96
+ "learning_rate": 0.0001,
97
+ "loss": 0.2745,
98
+ "step": 56
99
+ },
100
+ {
101
+ "epoch": 7.5,
102
+ "learning_rate": 0.0001,
103
+ "loss": 0.2495,
104
+ "step": 60
105
+ },
106
+ {
107
+ "epoch": 7.99,
108
+ "learning_rate": 0.0001,
109
+ "loss": 0.2298,
110
+ "step": 64
111
+ },
112
+ {
113
+ "epoch": 7.99,
114
+ "eval_exact_match": 0.3644488579940417,
115
+ "eval_exec": 0.3902681231380338,
116
+ "eval_loss": 0.23653237521648407,
117
+ "eval_runtime": 724.5845,
118
+ "eval_samples_per_second": 1.794,
119
+ "step": 64
120
+ },
121
+ {
122
+ "epoch": 8.5,
123
+ "learning_rate": 0.0001,
124
+ "loss": 0.2242,
125
+ "step": 68
126
+ },
127
+ {
128
+ "epoch": 8.99,
129
+ "learning_rate": 0.0001,
130
+ "loss": 0.205,
131
+ "step": 72
132
+ },
133
+ {
134
+ "epoch": 9.5,
135
+ "learning_rate": 0.0001,
136
+ "loss": 0.1932,
137
+ "step": 76
138
+ },
139
+ {
140
+ "epoch": 9.99,
141
+ "learning_rate": 0.0001,
142
+ "loss": 0.1937,
143
+ "step": 80
144
+ },
145
+ {
146
+ "epoch": 10.5,
147
+ "learning_rate": 0.0001,
148
+ "loss": 0.1978,
149
+ "step": 84
150
+ },
151
+ {
152
+ "epoch": 10.99,
153
+ "learning_rate": 0.0001,
154
+ "loss": 0.1747,
155
+ "step": 88
156
+ },
157
+ {
158
+ "epoch": 11.5,
159
+ "learning_rate": 0.0001,
160
+ "loss": 0.1669,
161
+ "step": 92
162
+ },
163
+ {
164
+ "epoch": 11.99,
165
+ "learning_rate": 0.0001,
166
+ "loss": 0.1734,
167
+ "step": 96
168
+ },
169
+ {
170
+ "epoch": 12.5,
171
+ "learning_rate": 0.0001,
172
+ "loss": 0.1615,
173
+ "step": 100
174
+ },
175
+ {
176
+ "epoch": 12.99,
177
+ "learning_rate": 0.0001,
178
+ "loss": 0.1551,
179
+ "step": 104
180
+ },
181
+ {
182
+ "epoch": 13.5,
183
+ "learning_rate": 0.0001,
184
+ "loss": 0.1519,
185
+ "step": 108
186
+ },
187
+ {
188
+ "epoch": 13.99,
189
+ "learning_rate": 0.0001,
190
+ "loss": 0.1476,
191
+ "step": 112
192
+ },
193
+ {
194
+ "epoch": 14.5,
195
+ "learning_rate": 0.0001,
196
+ "loss": 0.1483,
197
+ "step": 116
198
+ },
199
+ {
200
+ "epoch": 14.99,
201
+ "learning_rate": 0.0001,
202
+ "loss": 0.1478,
203
+ "step": 120
204
+ },
205
+ {
206
+ "epoch": 15.5,
207
+ "learning_rate": 0.0001,
208
+ "loss": 0.1292,
209
+ "step": 124
210
+ },
211
+ {
212
+ "epoch": 15.99,
213
+ "learning_rate": 0.0001,
214
+ "loss": 0.1285,
215
+ "step": 128
216
+ },
217
+ {
218
+ "epoch": 15.99,
219
+ "eval_exact_match": 0.4170804369414101,
220
+ "eval_exec": 0.45878848063555117,
221
+ "eval_loss": 0.20163151621818542,
222
+ "eval_runtime": 846.5253,
223
+ "eval_samples_per_second": 1.536,
224
+ "step": 128
225
+ },
226
+ {
227
+ "epoch": 16.5,
228
+ "learning_rate": 0.0001,
229
+ "loss": 0.1407,
230
+ "step": 132
231
+ },
232
+ {
233
+ "epoch": 16.99,
234
+ "learning_rate": 0.0001,
235
+ "loss": 0.1228,
236
+ "step": 136
237
+ },
238
+ {
239
+ "epoch": 17.5,
240
+ "learning_rate": 0.0001,
241
+ "loss": 0.1324,
242
+ "step": 140
243
+ },
244
+ {
245
+ "epoch": 17.99,
246
+ "learning_rate": 0.0001,
247
+ "loss": 0.1214,
248
+ "step": 144
249
+ },
250
+ {
251
+ "epoch": 18.5,
252
+ "learning_rate": 0.0001,
253
+ "loss": 0.1224,
254
+ "step": 148
255
+ },
256
+ {
257
+ "epoch": 18.99,
258
+ "learning_rate": 0.0001,
259
+ "loss": 0.1128,
260
+ "step": 152
261
+ },
262
+ {
263
+ "epoch": 19.5,
264
+ "learning_rate": 0.0001,
265
+ "loss": 0.1016,
266
+ "step": 156
267
+ },
268
+ {
269
+ "epoch": 19.99,
270
+ "learning_rate": 0.0001,
271
+ "loss": 0.107,
272
+ "step": 160
273
+ },
274
+ {
275
+ "epoch": 20.5,
276
+ "learning_rate": 0.0001,
277
+ "loss": 0.1116,
278
+ "step": 164
279
+ },
280
+ {
281
+ "epoch": 20.99,
282
+ "learning_rate": 0.0001,
283
+ "loss": 0.0997,
284
+ "step": 168
285
+ },
286
+ {
287
+ "epoch": 21.5,
288
+ "learning_rate": 0.0001,
289
+ "loss": 0.1025,
290
+ "step": 172
291
+ },
292
+ {
293
+ "epoch": 21.99,
294
+ "learning_rate": 0.0001,
295
+ "loss": 0.0973,
296
+ "step": 176
297
+ },
298
+ {
299
+ "epoch": 22.5,
300
+ "learning_rate": 0.0001,
301
+ "loss": 0.0996,
302
+ "step": 180
303
+ },
304
+ {
305
+ "epoch": 22.99,
306
+ "learning_rate": 0.0001,
307
+ "loss": 0.0917,
308
+ "step": 184
309
+ },
310
+ {
311
+ "epoch": 23.5,
312
+ "learning_rate": 0.0001,
313
+ "loss": 0.0899,
314
+ "step": 188
315
+ },
316
+ {
317
+ "epoch": 23.99,
318
+ "learning_rate": 0.0001,
319
+ "loss": 0.0896,
320
+ "step": 192
321
+ },
322
+ {
323
+ "epoch": 23.99,
324
+ "eval_exact_match": 0.4816285998013903,
325
+ "eval_exec": 0.5104270109235353,
326
+ "eval_loss": 0.19154271483421326,
327
+ "eval_runtime": 788.1761,
328
+ "eval_samples_per_second": 1.649,
329
+ "step": 192
330
+ },
331
+ {
332
+ "epoch": 24.5,
333
+ "learning_rate": 0.0001,
334
+ "loss": 0.0909,
335
+ "step": 196
336
+ },
337
+ {
338
+ "epoch": 24.99,
339
+ "learning_rate": 0.0001,
340
+ "loss": 0.0867,
341
+ "step": 200
342
+ },
343
+ {
344
+ "epoch": 25.5,
345
+ "learning_rate": 0.0001,
346
+ "loss": 0.089,
347
+ "step": 204
348
+ },
349
+ {
350
+ "epoch": 25.99,
351
+ "learning_rate": 0.0001,
352
+ "loss": 0.0888,
353
+ "step": 208
354
+ },
355
+ {
356
+ "epoch": 26.5,
357
+ "learning_rate": 0.0001,
358
+ "loss": 0.0946,
359
+ "step": 212
360
+ },
361
+ {
362
+ "epoch": 26.99,
363
+ "learning_rate": 0.0001,
364
+ "loss": 0.0803,
365
+ "step": 216
366
+ },
367
+ {
368
+ "epoch": 27.5,
369
+ "learning_rate": 0.0001,
370
+ "loss": 0.0769,
371
+ "step": 220
372
+ },
373
+ {
374
+ "epoch": 27.99,
375
+ "learning_rate": 0.0001,
376
+ "loss": 0.078,
377
+ "step": 224
378
+ },
379
+ {
380
+ "epoch": 28.5,
381
+ "learning_rate": 0.0001,
382
+ "loss": 0.084,
383
+ "step": 228
384
+ },
385
+ {
386
+ "epoch": 28.99,
387
+ "learning_rate": 0.0001,
388
+ "loss": 0.07,
389
+ "step": 232
390
+ },
391
+ {
392
+ "epoch": 29.5,
393
+ "learning_rate": 0.0001,
394
+ "loss": 0.0727,
395
+ "step": 236
396
+ },
397
+ {
398
+ "epoch": 29.99,
399
+ "learning_rate": 0.0001,
400
+ "loss": 0.0719,
401
+ "step": 240
402
+ },
403
+ {
404
+ "epoch": 30.5,
405
+ "learning_rate": 0.0001,
406
+ "loss": 0.0727,
407
+ "step": 244
408
+ },
409
+ {
410
+ "epoch": 30.99,
411
+ "learning_rate": 0.0001,
412
+ "loss": 0.0759,
413
+ "step": 248
414
+ },
415
+ {
416
+ "epoch": 31.5,
417
+ "learning_rate": 0.0001,
418
+ "loss": 0.0708,
419
+ "step": 252
420
+ },
421
+ {
422
+ "epoch": 31.99,
423
+ "learning_rate": 0.0001,
424
+ "loss": 0.0674,
425
+ "step": 256
426
+ },
427
+ {
428
+ "epoch": 31.99,
429
+ "eval_exact_match": 0.5014895729890765,
430
+ "eval_exec": 0.5382323733862959,
431
+ "eval_loss": 0.19477179646492004,
432
+ "eval_runtime": 786.6121,
433
+ "eval_samples_per_second": 1.653,
434
+ "step": 256
435
+ },
436
+ {
437
+ "epoch": 32.5,
438
+ "learning_rate": 0.0001,
439
+ "loss": 0.0673,
440
+ "step": 260
441
+ },
442
+ {
443
+ "epoch": 32.99,
444
+ "learning_rate": 0.0001,
445
+ "loss": 0.0716,
446
+ "step": 264
447
+ },
448
+ {
449
+ "epoch": 33.5,
450
+ "learning_rate": 0.0001,
451
+ "loss": 0.0658,
452
+ "step": 268
453
+ },
454
+ {
455
+ "epoch": 33.99,
456
+ "learning_rate": 0.0001,
457
+ "loss": 0.0564,
458
+ "step": 272
459
+ },
460
+ {
461
+ "epoch": 34.5,
462
+ "learning_rate": 0.0001,
463
+ "loss": 0.0604,
464
+ "step": 276
465
+ },
466
+ {
467
+ "epoch": 34.99,
468
+ "learning_rate": 0.0001,
469
+ "loss": 0.0632,
470
+ "step": 280
471
+ },
472
+ {
473
+ "epoch": 35.5,
474
+ "learning_rate": 0.0001,
475
+ "loss": 0.0611,
476
+ "step": 284
477
+ },
478
+ {
479
+ "epoch": 35.99,
480
+ "learning_rate": 0.0001,
481
+ "loss": 0.0577,
482
+ "step": 288
483
+ },
484
+ {
485
+ "epoch": 36.5,
486
+ "learning_rate": 0.0001,
487
+ "loss": 0.0602,
488
+ "step": 292
489
+ },
490
+ {
491
+ "epoch": 36.99,
492
+ "learning_rate": 0.0001,
493
+ "loss": 0.0556,
494
+ "step": 296
495
+ },
496
+ {
497
+ "epoch": 37.5,
498
+ "learning_rate": 0.0001,
499
+ "loss": 0.0588,
500
+ "step": 300
501
+ },
502
+ {
503
+ "epoch": 37.99,
504
+ "learning_rate": 0.0001,
505
+ "loss": 0.0584,
506
+ "step": 304
507
+ },
508
+ {
509
+ "epoch": 38.5,
510
+ "learning_rate": 0.0001,
511
+ "loss": 0.0537,
512
+ "step": 308
513
+ },
514
+ {
515
+ "epoch": 38.99,
516
+ "learning_rate": 0.0001,
517
+ "loss": 0.0538,
518
+ "step": 312
519
+ },
520
+ {
521
+ "epoch": 39.5,
522
+ "learning_rate": 0.0001,
523
+ "loss": 0.0517,
524
+ "step": 316
525
+ },
526
+ {
527
+ "epoch": 39.99,
528
+ "learning_rate": 0.0001,
529
+ "loss": 0.0542,
530
+ "step": 320
531
+ },
532
+ {
533
+ "epoch": 39.99,
534
+ "eval_exact_match": 0.5173783515392254,
535
+ "eval_exec": 0.5501489572989077,
536
+ "eval_loss": 0.21476374566555023,
537
+ "eval_runtime": 722.7734,
538
+ "eval_samples_per_second": 1.799,
539
+ "step": 320
540
+ },
541
+ {
542
+ "epoch": 40.5,
543
+ "learning_rate": 0.0001,
544
+ "loss": 0.052,
545
+ "step": 324
546
+ },
547
+ {
548
+ "epoch": 40.99,
549
+ "learning_rate": 0.0001,
550
+ "loss": 0.0552,
551
+ "step": 328
552
+ },
553
+ {
554
+ "epoch": 41.5,
555
+ "learning_rate": 0.0001,
556
+ "loss": 0.0474,
557
+ "step": 332
558
+ },
559
+ {
560
+ "epoch": 41.99,
561
+ "learning_rate": 0.0001,
562
+ "loss": 0.0456,
563
+ "step": 336
564
+ },
565
+ {
566
+ "epoch": 42.5,
567
+ "learning_rate": 0.0001,
568
+ "loss": 0.0517,
569
+ "step": 340
570
+ },
571
+ {
572
+ "epoch": 42.99,
573
+ "learning_rate": 0.0001,
574
+ "loss": 0.0484,
575
+ "step": 344
576
+ },
577
+ {
578
+ "epoch": 43.5,
579
+ "learning_rate": 0.0001,
580
+ "loss": 0.0432,
581
+ "step": 348
582
+ },
583
+ {
584
+ "epoch": 43.99,
585
+ "learning_rate": 0.0001,
586
+ "loss": 0.0446,
587
+ "step": 352
588
+ },
589
+ {
590
+ "epoch": 44.5,
591
+ "learning_rate": 0.0001,
592
+ "loss": 0.0443,
593
+ "step": 356
594
+ },
595
+ {
596
+ "epoch": 44.99,
597
+ "learning_rate": 0.0001,
598
+ "loss": 0.0425,
599
+ "step": 360
600
+ },
601
+ {
602
+ "epoch": 45.5,
603
+ "learning_rate": 0.0001,
604
+ "loss": 0.0452,
605
+ "step": 364
606
+ },
607
+ {
608
+ "epoch": 45.99,
609
+ "learning_rate": 0.0001,
610
+ "loss": 0.0429,
611
+ "step": 368
612
+ },
613
+ {
614
+ "epoch": 46.5,
615
+ "learning_rate": 0.0001,
616
+ "loss": 0.0471,
617
+ "step": 372
618
+ },
619
+ {
620
+ "epoch": 46.99,
621
+ "learning_rate": 0.0001,
622
+ "loss": 0.0433,
623
+ "step": 376
624
+ },
625
+ {
626
+ "epoch": 47.5,
627
+ "learning_rate": 0.0001,
628
+ "loss": 0.0416,
629
+ "step": 380
630
+ },
631
+ {
632
+ "epoch": 47.99,
633
+ "learning_rate": 0.0001,
634
+ "loss": 0.0397,
635
+ "step": 384
636
+ },
637
+ {
638
+ "epoch": 47.99,
639
+ "eval_exact_match": 0.5143992055610725,
640
+ "eval_exec": 0.5491559086395233,
641
+ "eval_loss": 0.22859126329421997,
642
+ "eval_runtime": 751.9885,
643
+ "eval_samples_per_second": 1.729,
644
+ "step": 384
645
+ },
646
+ {
647
+ "epoch": 48.5,
648
+ "learning_rate": 0.0001,
649
+ "loss": 0.0449,
650
+ "step": 388
651
+ },
652
+ {
653
+ "epoch": 48.99,
654
+ "learning_rate": 0.0001,
655
+ "loss": 0.0404,
656
+ "step": 392
657
+ },
658
+ {
659
+ "epoch": 49.5,
660
+ "learning_rate": 0.0001,
661
+ "loss": 0.0384,
662
+ "step": 396
663
+ },
664
+ {
665
+ "epoch": 49.99,
666
+ "learning_rate": 0.0001,
667
+ "loss": 0.0378,
668
+ "step": 400
669
+ },
670
+ {
671
+ "epoch": 50.5,
672
+ "learning_rate": 0.0001,
673
+ "loss": 0.0405,
674
+ "step": 404
675
+ },
676
+ {
677
+ "epoch": 50.99,
678
+ "learning_rate": 0.0001,
679
+ "loss": 0.0391,
680
+ "step": 408
681
+ },
682
+ {
683
+ "epoch": 51.5,
684
+ "learning_rate": 0.0001,
685
+ "loss": 0.0377,
686
+ "step": 412
687
+ },
688
+ {
689
+ "epoch": 51.99,
690
+ "learning_rate": 0.0001,
691
+ "loss": 0.0349,
692
+ "step": 416
693
+ },
694
+ {
695
+ "epoch": 52.5,
696
+ "learning_rate": 0.0001,
697
+ "loss": 0.0328,
698
+ "step": 420
699
+ },
700
+ {
701
+ "epoch": 52.99,
702
+ "learning_rate": 0.0001,
703
+ "loss": 0.0363,
704
+ "step": 424
705
+ },
706
+ {
707
+ "epoch": 53.5,
708
+ "learning_rate": 0.0001,
709
+ "loss": 0.0355,
710
+ "step": 428
711
+ },
712
+ {
713
+ "epoch": 53.99,
714
+ "learning_rate": 0.0001,
715
+ "loss": 0.0317,
716
+ "step": 432
717
+ },
718
+ {
719
+ "epoch": 54.5,
720
+ "learning_rate": 0.0001,
721
+ "loss": 0.0308,
722
+ "step": 436
723
+ },
724
+ {
725
+ "epoch": 54.99,
726
+ "learning_rate": 0.0001,
727
+ "loss": 0.0303,
728
+ "step": 440
729
+ },
730
+ {
731
+ "epoch": 55.5,
732
+ "learning_rate": 0.0001,
733
+ "loss": 0.0313,
734
+ "step": 444
735
+ },
736
+ {
737
+ "epoch": 55.99,
738
+ "learning_rate": 0.0001,
739
+ "loss": 0.0351,
740
+ "step": 448
741
+ },
742
+ {
743
+ "epoch": 55.99,
744
+ "eval_exact_match": 0.5243296921549155,
745
+ "eval_exec": 0.5561072492552135,
746
+ "eval_loss": 0.2355286180973053,
747
+ "eval_runtime": 746.4078,
748
+ "eval_samples_per_second": 1.742,
749
+ "step": 448
750
+ },
751
+ {
752
+ "epoch": 56.5,
753
+ "learning_rate": 0.0001,
754
+ "loss": 0.0307,
755
+ "step": 452
756
+ },
757
+ {
758
+ "epoch": 56.99,
759
+ "learning_rate": 0.0001,
760
+ "loss": 0.0296,
761
+ "step": 456
762
+ },
763
+ {
764
+ "epoch": 57.5,
765
+ "learning_rate": 0.0001,
766
+ "loss": 0.0295,
767
+ "step": 460
768
+ },
769
+ {
770
+ "epoch": 57.99,
771
+ "learning_rate": 0.0001,
772
+ "loss": 0.0316,
773
+ "step": 464
774
+ },
775
+ {
776
+ "epoch": 58.5,
777
+ "learning_rate": 0.0001,
778
+ "loss": 0.0283,
779
+ "step": 468
780
+ },
781
+ {
782
+ "epoch": 58.99,
783
+ "learning_rate": 0.0001,
784
+ "loss": 0.0321,
785
+ "step": 472
786
+ },
787
+ {
788
+ "epoch": 59.5,
789
+ "learning_rate": 0.0001,
790
+ "loss": 0.0296,
791
+ "step": 476
792
+ },
793
+ {
794
+ "epoch": 59.99,
795
+ "learning_rate": 0.0001,
796
+ "loss": 0.0262,
797
+ "step": 480
798
+ },
799
+ {
800
+ "epoch": 60.5,
801
+ "learning_rate": 0.0001,
802
+ "loss": 0.0259,
803
+ "step": 484
804
+ },
805
+ {
806
+ "epoch": 60.99,
807
+ "learning_rate": 0.0001,
808
+ "loss": 0.027,
809
+ "step": 488
810
+ },
811
+ {
812
+ "epoch": 61.5,
813
+ "learning_rate": 0.0001,
814
+ "loss": 0.0263,
815
+ "step": 492
816
+ },
817
+ {
818
+ "epoch": 61.99,
819
+ "learning_rate": 0.0001,
820
+ "loss": 0.0259,
821
+ "step": 496
822
+ },
823
+ {
824
+ "epoch": 62.5,
825
+ "learning_rate": 0.0001,
826
+ "loss": 0.0275,
827
+ "step": 500
828
+ },
829
+ {
830
+ "epoch": 62.99,
831
+ "learning_rate": 0.0001,
832
+ "loss": 0.0262,
833
+ "step": 504
834
+ },
835
+ {
836
+ "epoch": 63.5,
837
+ "learning_rate": 0.0001,
838
+ "loss": 0.0252,
839
+ "step": 508
840
+ },
841
+ {
842
+ "epoch": 63.99,
843
+ "learning_rate": 0.0001,
844
+ "loss": 0.025,
845
+ "step": 512
846
+ },
847
+ {
848
+ "epoch": 63.99,
849
+ "eval_exact_match": 0.5094339622641509,
850
+ "eval_exec": 0.5422045680238332,
851
+ "eval_loss": 0.2562254071235657,
852
+ "eval_runtime": 720.3713,
853
+ "eval_samples_per_second": 1.805,
854
+ "step": 512
855
+ },
856
+ {
857
+ "epoch": 64.5,
858
+ "learning_rate": 0.0001,
859
+ "loss": 0.0256,
860
+ "step": 516
861
+ },
862
+ {
863
+ "epoch": 64.99,
864
+ "learning_rate": 0.0001,
865
+ "loss": 0.0272,
866
+ "step": 520
867
+ },
868
+ {
869
+ "epoch": 65.5,
870
+ "learning_rate": 0.0001,
871
+ "loss": 0.0293,
872
+ "step": 524
873
+ },
874
+ {
875
+ "epoch": 65.99,
876
+ "learning_rate": 0.0001,
877
+ "loss": 0.0231,
878
+ "step": 528
879
+ },
880
+ {
881
+ "epoch": 66.5,
882
+ "learning_rate": 0.0001,
883
+ "loss": 0.0232,
884
+ "step": 532
885
+ },
886
+ {
887
+ "epoch": 66.99,
888
+ "learning_rate": 0.0001,
889
+ "loss": 0.0253,
890
+ "step": 536
891
+ },
892
+ {
893
+ "epoch": 67.5,
894
+ "learning_rate": 0.0001,
895
+ "loss": 0.0224,
896
+ "step": 540
897
+ },
898
+ {
899
+ "epoch": 67.99,
900
+ "learning_rate": 0.0001,
901
+ "loss": 0.021,
902
+ "step": 544
903
+ },
904
+ {
905
+ "epoch": 68.5,
906
+ "learning_rate": 0.0001,
907
+ "loss": 0.0206,
908
+ "step": 548
909
+ },
910
+ {
911
+ "epoch": 68.99,
912
+ "learning_rate": 0.0001,
913
+ "loss": 0.0229,
914
+ "step": 552
915
+ },
916
+ {
917
+ "epoch": 69.5,
918
+ "learning_rate": 0.0001,
919
+ "loss": 0.0243,
920
+ "step": 556
921
+ },
922
+ {
923
+ "epoch": 69.99,
924
+ "learning_rate": 0.0001,
925
+ "loss": 0.0211,
926
+ "step": 560
927
+ },
928
+ {
929
+ "epoch": 70.5,
930
+ "learning_rate": 0.0001,
931
+ "loss": 0.0208,
932
+ "step": 564
933
+ },
934
+ {
935
+ "epoch": 70.99,
936
+ "learning_rate": 0.0001,
937
+ "loss": 0.0225,
938
+ "step": 568
939
+ },
940
+ {
941
+ "epoch": 71.5,
942
+ "learning_rate": 0.0001,
943
+ "loss": 0.0209,
944
+ "step": 572
945
+ },
946
+ {
947
+ "epoch": 71.99,
948
+ "learning_rate": 0.0001,
949
+ "loss": 0.0225,
950
+ "step": 576
951
+ },
952
+ {
953
+ "epoch": 71.99,
954
+ "eval_exact_match": 0.5173783515392254,
955
+ "eval_exec": 0.5461767626613704,
956
+ "eval_loss": 0.2805687189102173,
957
+ "eval_runtime": 749.8642,
958
+ "eval_samples_per_second": 1.734,
959
+ "step": 576
960
+ },
961
+ {
962
+ "epoch": 72.5,
963
+ "learning_rate": 0.0001,
964
+ "loss": 0.0188,
965
+ "step": 580
966
+ },
967
+ {
968
+ "epoch": 72.99,
969
+ "learning_rate": 0.0001,
970
+ "loss": 0.0209,
971
+ "step": 584
972
+ },
973
+ {
974
+ "epoch": 73.5,
975
+ "learning_rate": 0.0001,
976
+ "loss": 0.0193,
977
+ "step": 588
978
+ },
979
+ {
980
+ "epoch": 73.99,
981
+ "learning_rate": 0.0001,
982
+ "loss": 0.0221,
983
+ "step": 592
984
+ },
985
+ {
986
+ "epoch": 74.5,
987
+ "learning_rate": 0.0001,
988
+ "loss": 0.0193,
989
+ "step": 596
990
+ },
991
+ {
992
+ "epoch": 74.99,
993
+ "learning_rate": 0.0001,
994
+ "loss": 0.0184,
995
+ "step": 600
996
+ },
997
+ {
998
+ "epoch": 75.5,
999
+ "learning_rate": 0.0001,
1000
+ "loss": 0.0169,
1001
+ "step": 604
1002
+ },
1003
+ {
1004
+ "epoch": 75.99,
1005
+ "learning_rate": 0.0001,
1006
+ "loss": 0.0177,
1007
+ "step": 608
1008
+ },
1009
+ {
1010
+ "epoch": 76.5,
1011
+ "learning_rate": 0.0001,
1012
+ "loss": 0.0189,
1013
+ "step": 612
1014
+ },
1015
+ {
1016
+ "epoch": 76.99,
1017
+ "learning_rate": 0.0001,
1018
+ "loss": 0.0185,
1019
+ "step": 616
1020
+ },
1021
+ {
1022
+ "epoch": 77.5,
1023
+ "learning_rate": 0.0001,
1024
+ "loss": 0.0166,
1025
+ "step": 620
1026
+ },
1027
+ {
1028
+ "epoch": 77.99,
1029
+ "learning_rate": 0.0001,
1030
+ "loss": 0.0167,
1031
+ "step": 624
1032
+ },
1033
+ {
1034
+ "epoch": 78.5,
1035
+ "learning_rate": 0.0001,
1036
+ "loss": 0.0145,
1037
+ "step": 628
1038
+ },
1039
+ {
1040
+ "epoch": 78.99,
1041
+ "learning_rate": 0.0001,
1042
+ "loss": 0.0162,
1043
+ "step": 632
1044
+ },
1045
+ {
1046
+ "epoch": 79.5,
1047
+ "learning_rate": 0.0001,
1048
+ "loss": 0.0168,
1049
+ "step": 636
1050
+ },
1051
+ {
1052
+ "epoch": 79.99,
1053
+ "learning_rate": 0.0001,
1054
+ "loss": 0.0172,
1055
+ "step": 640
1056
+ },
1057
+ {
1058
+ "epoch": 79.99,
1059
+ "eval_exact_match": 0.5134061569016882,
1060
+ "eval_exec": 0.5431976166832175,
1061
+ "eval_loss": 0.2822588086128235,
1062
+ "eval_runtime": 735.1307,
1063
+ "eval_samples_per_second": 1.768,
1064
+ "step": 640
1065
+ },
1066
+ {
1067
+ "epoch": 80.5,
1068
+ "learning_rate": 0.0001,
1069
+ "loss": 0.017,
1070
+ "step": 644
1071
+ },
1072
+ {
1073
+ "epoch": 80.99,
1074
+ "learning_rate": 0.0001,
1075
+ "loss": 0.0166,
1076
+ "step": 648
1077
+ },
1078
+ {
1079
+ "epoch": 81.5,
1080
+ "learning_rate": 0.0001,
1081
+ "loss": 0.0157,
1082
+ "step": 652
1083
+ },
1084
+ {
1085
+ "epoch": 81.99,
1086
+ "learning_rate": 0.0001,
1087
+ "loss": 0.0182,
1088
+ "step": 656
1089
+ },
1090
+ {
1091
+ "epoch": 82.5,
1092
+ "learning_rate": 0.0001,
1093
+ "loss": 0.0168,
1094
+ "step": 660
1095
+ },
1096
+ {
1097
+ "epoch": 82.99,
1098
+ "learning_rate": 0.0001,
1099
+ "loss": 0.0145,
1100
+ "step": 664
1101
+ },
1102
+ {
1103
+ "epoch": 83.5,
1104
+ "learning_rate": 0.0001,
1105
+ "loss": 0.0142,
1106
+ "step": 668
1107
+ },
1108
+ {
1109
+ "epoch": 83.99,
1110
+ "learning_rate": 0.0001,
1111
+ "loss": 0.0139,
1112
+ "step": 672
1113
+ },
1114
+ {
1115
+ "epoch": 84.5,
1116
+ "learning_rate": 0.0001,
1117
+ "loss": 0.0151,
1118
+ "step": 676
1119
+ },
1120
+ {
1121
+ "epoch": 84.99,
1122
+ "learning_rate": 0.0001,
1123
+ "loss": 0.0137,
1124
+ "step": 680
1125
+ },
1126
+ {
1127
+ "epoch": 85.5,
1128
+ "learning_rate": 0.0001,
1129
+ "loss": 0.0137,
1130
+ "step": 684
1131
+ },
1132
+ {
1133
+ "epoch": 85.99,
1134
+ "learning_rate": 0.0001,
1135
+ "loss": 0.0119,
1136
+ "step": 688
1137
+ },
1138
+ {
1139
+ "epoch": 86.5,
1140
+ "learning_rate": 0.0001,
1141
+ "loss": 0.0128,
1142
+ "step": 692
1143
+ },
1144
+ {
1145
+ "epoch": 86.99,
1146
+ "learning_rate": 0.0001,
1147
+ "loss": 0.0134,
1148
+ "step": 696
1149
+ },
1150
+ {
1151
+ "epoch": 87.5,
1152
+ "learning_rate": 0.0001,
1153
+ "loss": 0.0143,
1154
+ "step": 700
1155
+ },
1156
+ {
1157
+ "epoch": 87.99,
1158
+ "learning_rate": 0.0001,
1159
+ "loss": 0.0189,
1160
+ "step": 704
1161
+ },
1162
+ {
1163
+ "epoch": 87.99,
1164
+ "eval_exact_match": 0.5134061569016882,
1165
+ "eval_exec": 0.5441906653426017,
1166
+ "eval_loss": 0.285916805267334,
1167
+ "eval_runtime": 740.5163,
1168
+ "eval_samples_per_second": 1.756,
1169
+ "step": 704
1170
+ },
1171
+ {
1172
+ "epoch": 88.5,
1173
+ "learning_rate": 0.0001,
1174
+ "loss": 0.0125,
1175
+ "step": 708
1176
+ },
1177
+ {
1178
+ "epoch": 88.99,
1179
+ "learning_rate": 0.0001,
1180
+ "loss": 0.0122,
1181
+ "step": 712
1182
+ },
1183
+ {
1184
+ "epoch": 89.5,
1185
+ "learning_rate": 0.0001,
1186
+ "loss": 0.0141,
1187
+ "step": 716
1188
+ },
1189
+ {
1190
+ "epoch": 89.99,
1191
+ "learning_rate": 0.0001,
1192
+ "loss": 0.0132,
1193
+ "step": 720
1194
+ },
1195
+ {
1196
+ "epoch": 90.5,
1197
+ "learning_rate": 0.0001,
1198
+ "loss": 0.0152,
1199
+ "step": 724
1200
+ },
1201
+ {
1202
+ "epoch": 90.99,
1203
+ "learning_rate": 0.0001,
1204
+ "loss": 0.0118,
1205
+ "step": 728
1206
+ },
1207
+ {
1208
+ "epoch": 91.5,
1209
+ "learning_rate": 0.0001,
1210
+ "loss": 0.0105,
1211
+ "step": 732
1212
+ },
1213
+ {
1214
+ "epoch": 91.99,
1215
+ "learning_rate": 0.0001,
1216
+ "loss": 0.0117,
1217
+ "step": 736
1218
+ },
1219
+ {
1220
+ "epoch": 92.5,
1221
+ "learning_rate": 0.0001,
1222
+ "loss": 0.0119,
1223
+ "step": 740
1224
+ },
1225
+ {
1226
+ "epoch": 92.99,
1227
+ "learning_rate": 0.0001,
1228
+ "loss": 0.0126,
1229
+ "step": 744
1230
+ },
1231
+ {
1232
+ "epoch": 93.5,
1233
+ "learning_rate": 0.0001,
1234
+ "loss": 0.0118,
1235
+ "step": 748
1236
+ },
1237
+ {
1238
+ "epoch": 93.99,
1239
+ "learning_rate": 0.0001,
1240
+ "loss": 0.0143,
1241
+ "step": 752
1242
+ },
1243
+ {
1244
+ "epoch": 94.5,
1245
+ "learning_rate": 0.0001,
1246
+ "loss": 0.0127,
1247
+ "step": 756
1248
+ },
1249
+ {
1250
+ "epoch": 94.99,
1251
+ "learning_rate": 0.0001,
1252
+ "loss": 0.0143,
1253
+ "step": 760
1254
+ },
1255
+ {
1256
+ "epoch": 95.5,
1257
+ "learning_rate": 0.0001,
1258
+ "loss": 0.0107,
1259
+ "step": 764
1260
+ },
1261
+ {
1262
+ "epoch": 95.99,
1263
+ "learning_rate": 0.0001,
1264
+ "loss": 0.0112,
1265
+ "step": 768
1266
+ },
1267
+ {
1268
+ "epoch": 95.99,
1269
+ "eval_exact_match": 0.506454816285998,
1270
+ "eval_exec": 0.5461767626613704,
1271
+ "eval_loss": 0.3350318670272827,
1272
+ "eval_runtime": 791.2409,
1273
+ "eval_samples_per_second": 1.643,
1274
+ "step": 768
1275
+ },
1276
+ {
1277
+ "epoch": 96.5,
1278
+ "learning_rate": 0.0001,
1279
+ "loss": 0.0113,
1280
+ "step": 772
1281
+ },
1282
+ {
1283
+ "epoch": 96.99,
1284
+ "learning_rate": 0.0001,
1285
+ "loss": 0.0107,
1286
+ "step": 776
1287
+ },
1288
+ {
1289
+ "epoch": 97.5,
1290
+ "learning_rate": 0.0001,
1291
+ "loss": 0.0112,
1292
+ "step": 780
1293
+ },
1294
+ {
1295
+ "epoch": 97.99,
1296
+ "learning_rate": 0.0001,
1297
+ "loss": 0.011,
1298
+ "step": 784
1299
+ },
1300
+ {
1301
+ "epoch": 98.5,
1302
+ "learning_rate": 0.0001,
1303
+ "loss": 0.012,
1304
+ "step": 788
1305
+ },
1306
+ {
1307
+ "epoch": 98.99,
1308
+ "learning_rate": 0.0001,
1309
+ "loss": 0.0122,
1310
+ "step": 792
1311
+ },
1312
+ {
1313
+ "epoch": 99.5,
1314
+ "learning_rate": 0.0001,
1315
+ "loss": 0.0113,
1316
+ "step": 796
1317
+ },
1318
+ {
1319
+ "epoch": 99.99,
1320
+ "learning_rate": 0.0001,
1321
+ "loss": 0.0101,
1322
+ "step": 800
1323
+ },
1324
+ {
1325
+ "epoch": 100.5,
1326
+ "learning_rate": 0.0001,
1327
+ "loss": 0.0089,
1328
+ "step": 804
1329
+ },
1330
+ {
1331
+ "epoch": 100.99,
1332
+ "learning_rate": 0.0001,
1333
+ "loss": 0.011,
1334
+ "step": 808
1335
+ },
1336
+ {
1337
+ "epoch": 101.5,
1338
+ "learning_rate": 0.0001,
1339
+ "loss": 0.0086,
1340
+ "step": 812
1341
+ },
1342
+ {
1343
+ "epoch": 101.99,
1344
+ "learning_rate": 0.0001,
1345
+ "loss": 0.0097,
1346
+ "step": 816
1347
+ },
1348
+ {
1349
+ "epoch": 102.5,
1350
+ "learning_rate": 0.0001,
1351
+ "loss": 0.011,
1352
+ "step": 820
1353
+ },
1354
+ {
1355
+ "epoch": 102.99,
1356
+ "learning_rate": 0.0001,
1357
+ "loss": 0.0085,
1358
+ "step": 824
1359
+ },
1360
+ {
1361
+ "epoch": 103.5,
1362
+ "learning_rate": 0.0001,
1363
+ "loss": 0.0086,
1364
+ "step": 828
1365
+ },
1366
+ {
1367
+ "epoch": 103.99,
1368
+ "learning_rate": 0.0001,
1369
+ "loss": 0.0094,
1370
+ "step": 832
1371
+ },
1372
+ {
1373
+ "epoch": 103.99,
1374
+ "eval_exact_match": 0.5183714001986097,
1375
+ "eval_exec": 0.5441906653426017,
1376
+ "eval_loss": 0.34131428599357605,
1377
+ "eval_runtime": 681.4487,
1378
+ "eval_samples_per_second": 1.908,
1379
+ "step": 832
1380
+ },
1381
+ {
1382
+ "epoch": 104.5,
1383
+ "learning_rate": 0.0001,
1384
+ "loss": 0.0111,
1385
+ "step": 836
1386
+ },
1387
+ {
1388
+ "epoch": 104.99,
1389
+ "learning_rate": 0.0001,
1390
+ "loss": 0.0095,
1391
+ "step": 840
1392
+ },
1393
+ {
1394
+ "epoch": 105.5,
1395
+ "learning_rate": 0.0001,
1396
+ "loss": 0.0083,
1397
+ "step": 844
1398
+ },
1399
+ {
1400
+ "epoch": 105.99,
1401
+ "learning_rate": 0.0001,
1402
+ "loss": 0.0083,
1403
+ "step": 848
1404
+ },
1405
+ {
1406
+ "epoch": 106.5,
1407
+ "learning_rate": 0.0001,
1408
+ "loss": 0.0085,
1409
+ "step": 852
1410
+ },
1411
+ {
1412
+ "epoch": 106.99,
1413
+ "learning_rate": 0.0001,
1414
+ "loss": 0.0086,
1415
+ "step": 856
1416
+ },
1417
+ {
1418
+ "epoch": 107.5,
1419
+ "learning_rate": 0.0001,
1420
+ "loss": 0.0085,
1421
+ "step": 860
1422
+ },
1423
+ {
1424
+ "epoch": 107.99,
1425
+ "learning_rate": 0.0001,
1426
+ "loss": 0.0088,
1427
+ "step": 864
1428
+ },
1429
+ {
1430
+ "epoch": 108.5,
1431
+ "learning_rate": 0.0001,
1432
+ "loss": 0.0094,
1433
+ "step": 868
1434
+ },
1435
+ {
1436
+ "epoch": 108.99,
1437
+ "learning_rate": 0.0001,
1438
+ "loss": 0.0084,
1439
+ "step": 872
1440
+ },
1441
+ {
1442
+ "epoch": 109.5,
1443
+ "learning_rate": 0.0001,
1444
+ "loss": 0.0075,
1445
+ "step": 876
1446
+ },
1447
+ {
1448
+ "epoch": 109.99,
1449
+ "learning_rate": 0.0001,
1450
+ "loss": 0.0093,
1451
+ "step": 880
1452
+ },
1453
+ {
1454
+ "epoch": 110.5,
1455
+ "learning_rate": 0.0001,
1456
+ "loss": 0.0091,
1457
+ "step": 884
1458
+ },
1459
+ {
1460
+ "epoch": 110.99,
1461
+ "learning_rate": 0.0001,
1462
+ "loss": 0.0088,
1463
+ "step": 888
1464
+ },
1465
+ {
1466
+ "epoch": 111.5,
1467
+ "learning_rate": 0.0001,
1468
+ "loss": 0.0083,
1469
+ "step": 892
1470
+ },
1471
+ {
1472
+ "epoch": 111.99,
1473
+ "learning_rate": 0.0001,
1474
+ "loss": 0.0086,
1475
+ "step": 896
1476
+ },
1477
+ {
1478
+ "epoch": 111.99,
1479
+ "eval_exact_match": 0.5243296921549155,
1480
+ "eval_exec": 0.5521350546176763,
1481
+ "eval_loss": 0.3550397753715515,
1482
+ "eval_runtime": 717.3828,
1483
+ "eval_samples_per_second": 1.812,
1484
+ "step": 896
1485
+ },
1486
+ {
1487
+ "epoch": 112.5,
1488
+ "learning_rate": 0.0001,
1489
+ "loss": 0.0086,
1490
+ "step": 900
1491
+ },
1492
+ {
1493
+ "epoch": 112.99,
1494
+ "learning_rate": 0.0001,
1495
+ "loss": 0.0085,
1496
+ "step": 904
1497
+ },
1498
+ {
1499
+ "epoch": 113.5,
1500
+ "learning_rate": 0.0001,
1501
+ "loss": 0.0076,
1502
+ "step": 908
1503
+ },
1504
+ {
1505
+ "epoch": 113.99,
1506
+ "learning_rate": 0.0001,
1507
+ "loss": 0.0093,
1508
+ "step": 912
1509
+ },
1510
+ {
1511
+ "epoch": 114.5,
1512
+ "learning_rate": 0.0001,
1513
+ "loss": 0.0068,
1514
+ "step": 916
1515
+ },
1516
+ {
1517
+ "epoch": 114.99,
1518
+ "learning_rate": 0.0001,
1519
+ "loss": 0.0075,
1520
+ "step": 920
1521
+ },
1522
+ {
1523
+ "epoch": 115.5,
1524
+ "learning_rate": 0.0001,
1525
+ "loss": 0.0073,
1526
+ "step": 924
1527
+ },
1528
+ {
1529
+ "epoch": 115.99,
1530
+ "learning_rate": 0.0001,
1531
+ "loss": 0.0071,
1532
+ "step": 928
1533
+ },
1534
+ {
1535
+ "epoch": 116.5,
1536
+ "learning_rate": 0.0001,
1537
+ "loss": 0.0077,
1538
+ "step": 932
1539
+ },
1540
+ {
1541
+ "epoch": 116.99,
1542
+ "learning_rate": 0.0001,
1543
+ "loss": 0.006,
1544
+ "step": 936
1545
+ },
1546
+ {
1547
+ "epoch": 117.5,
1548
+ "learning_rate": 0.0001,
1549
+ "loss": 0.0065,
1550
+ "step": 940
1551
+ },
1552
+ {
1553
+ "epoch": 117.99,
1554
+ "learning_rate": 0.0001,
1555
+ "loss": 0.008,
1556
+ "step": 944
1557
+ },
1558
+ {
1559
+ "epoch": 118.5,
1560
+ "learning_rate": 0.0001,
1561
+ "loss": 0.0069,
1562
+ "step": 948
1563
+ },
1564
+ {
1565
+ "epoch": 118.99,
1566
+ "learning_rate": 0.0001,
1567
+ "loss": 0.0082,
1568
+ "step": 952
1569
+ },
1570
+ {
1571
+ "epoch": 119.5,
1572
+ "learning_rate": 0.0001,
1573
+ "loss": 0.0058,
1574
+ "step": 956
1575
+ },
1576
+ {
1577
+ "epoch": 119.99,
1578
+ "learning_rate": 0.0001,
1579
+ "loss": 0.0072,
1580
+ "step": 960
1581
+ },
1582
+ {
1583
+ "epoch": 119.99,
1584
+ "eval_exact_match": 0.5104270109235353,
1585
+ "eval_exec": 0.5431976166832175,
1586
+ "eval_loss": 0.3599448800086975,
1587
+ "eval_runtime": 731.1991,
1588
+ "eval_samples_per_second": 1.778,
1589
+ "step": 960
1590
+ },
1591
+ {
1592
+ "epoch": 120.5,
1593
+ "learning_rate": 0.0001,
1594
+ "loss": 0.0062,
1595
+ "step": 964
1596
+ },
1597
+ {
1598
+ "epoch": 120.99,
1599
+ "learning_rate": 0.0001,
1600
+ "loss": 0.0066,
1601
+ "step": 968
1602
+ },
1603
+ {
1604
+ "epoch": 121.5,
1605
+ "learning_rate": 0.0001,
1606
+ "loss": 0.0054,
1607
+ "step": 972
1608
+ },
1609
+ {
1610
+ "epoch": 121.99,
1611
+ "learning_rate": 0.0001,
1612
+ "loss": 0.0055,
1613
+ "step": 976
1614
+ },
1615
+ {
1616
+ "epoch": 122.5,
1617
+ "learning_rate": 0.0001,
1618
+ "loss": 0.0054,
1619
+ "step": 980
1620
+ },
1621
+ {
1622
+ "epoch": 122.99,
1623
+ "learning_rate": 0.0001,
1624
+ "loss": 0.0066,
1625
+ "step": 984
1626
+ },
1627
+ {
1628
+ "epoch": 123.5,
1629
+ "learning_rate": 0.0001,
1630
+ "loss": 0.0068,
1631
+ "step": 988
1632
+ },
1633
+ {
1634
+ "epoch": 123.99,
1635
+ "learning_rate": 0.0001,
1636
+ "loss": 0.0059,
1637
+ "step": 992
1638
+ },
1639
+ {
1640
+ "epoch": 124.5,
1641
+ "learning_rate": 0.0001,
1642
+ "loss": 0.0065,
1643
+ "step": 996
1644
+ },
1645
+ {
1646
+ "epoch": 124.99,
1647
+ "learning_rate": 0.0001,
1648
+ "loss": 0.0067,
1649
+ "step": 1000
1650
+ },
1651
+ {
1652
+ "epoch": 125.5,
1653
+ "learning_rate": 0.0001,
1654
+ "loss": 0.0061,
1655
+ "step": 1004
1656
+ },
1657
+ {
1658
+ "epoch": 125.99,
1659
+ "learning_rate": 0.0001,
1660
+ "loss": 0.0059,
1661
+ "step": 1008
1662
+ },
1663
+ {
1664
+ "epoch": 126.5,
1665
+ "learning_rate": 0.0001,
1666
+ "loss": 0.0056,
1667
+ "step": 1012
1668
+ },
1669
+ {
1670
+ "epoch": 126.99,
1671
+ "learning_rate": 0.0001,
1672
+ "loss": 0.0059,
1673
+ "step": 1016
1674
+ },
1675
+ {
1676
+ "epoch": 127.5,
1677
+ "learning_rate": 0.0001,
1678
+ "loss": 0.0062,
1679
+ "step": 1020
1680
+ },
1681
+ {
1682
+ "epoch": 127.99,
1683
+ "learning_rate": 0.0001,
1684
+ "loss": 0.006,
1685
+ "step": 1024
1686
+ },
1687
+ {
1688
+ "epoch": 127.99,
1689
+ "eval_exact_match": 0.5203574975173784,
1690
+ "eval_exec": 0.5501489572989077,
1691
+ "eval_loss": 0.39113467931747437,
1692
+ "eval_runtime": 888.0535,
1693
+ "eval_samples_per_second": 1.464,
1694
+ "step": 1024
1695
+ },
1696
+ {
1697
+ "epoch": 128.5,
1698
+ "learning_rate": 0.0001,
1699
+ "loss": 0.0066,
1700
+ "step": 1028
1701
+ },
1702
+ {
1703
+ "epoch": 128.99,
1704
+ "learning_rate": 0.0001,
1705
+ "loss": 0.0065,
1706
+ "step": 1032
1707
+ },
1708
+ {
1709
+ "epoch": 129.5,
1710
+ "learning_rate": 0.0001,
1711
+ "loss": 0.0047,
1712
+ "step": 1036
1713
+ },
1714
+ {
1715
+ "epoch": 129.99,
1716
+ "learning_rate": 0.0001,
1717
+ "loss": 0.005,
1718
+ "step": 1040
1719
+ },
1720
+ {
1721
+ "epoch": 130.5,
1722
+ "learning_rate": 0.0001,
1723
+ "loss": 0.0048,
1724
+ "step": 1044
1725
+ },
1726
+ {
1727
+ "epoch": 130.99,
1728
+ "learning_rate": 0.0001,
1729
+ "loss": 0.0053,
1730
+ "step": 1048
1731
+ },
1732
+ {
1733
+ "epoch": 131.5,
1734
+ "learning_rate": 0.0001,
1735
+ "loss": 0.0063,
1736
+ "step": 1052
1737
+ },
1738
+ {
1739
+ "epoch": 131.99,
1740
+ "learning_rate": 0.0001,
1741
+ "loss": 0.0095,
1742
+ "step": 1056
1743
+ },
1744
+ {
1745
+ "epoch": 132.5,
1746
+ "learning_rate": 0.0001,
1747
+ "loss": 0.0058,
1748
+ "step": 1060
1749
+ },
1750
+ {
1751
+ "epoch": 132.99,
1752
+ "learning_rate": 0.0001,
1753
+ "loss": 0.0046,
1754
+ "step": 1064
1755
+ },
1756
+ {
1757
+ "epoch": 133.5,
1758
+ "learning_rate": 0.0001,
1759
+ "loss": 0.0067,
1760
+ "step": 1068
1761
+ },
1762
+ {
1763
+ "epoch": 133.99,
1764
+ "learning_rate": 0.0001,
1765
+ "loss": 0.0056,
1766
+ "step": 1072
1767
+ },
1768
+ {
1769
+ "epoch": 134.5,
1770
+ "learning_rate": 0.0001,
1771
+ "loss": 0.0048,
1772
+ "step": 1076
1773
+ },
1774
+ {
1775
+ "epoch": 134.99,
1776
+ "learning_rate": 0.0001,
1777
+ "loss": 0.0059,
1778
+ "step": 1080
1779
+ },
1780
+ {
1781
+ "epoch": 135.5,
1782
+ "learning_rate": 0.0001,
1783
+ "loss": 0.0049,
1784
+ "step": 1084
1785
+ },
1786
+ {
1787
+ "epoch": 135.99,
1788
+ "learning_rate": 0.0001,
1789
+ "loss": 0.0054,
1790
+ "step": 1088
1791
+ },
1792
+ {
1793
+ "epoch": 135.99,
1794
+ "eval_exact_match": 0.5084409136047666,
1795
+ "eval_exec": 0.535253227408143,
1796
+ "eval_loss": 0.3826310336589813,
1797
+ "eval_runtime": 944.2579,
1798
+ "eval_samples_per_second": 1.377,
1799
+ "step": 1088
1800
+ },
1801
+ {
1802
+ "epoch": 136.5,
1803
+ "learning_rate": 0.0001,
1804
+ "loss": 0.0048,
1805
+ "step": 1092
1806
+ },
1807
+ {
1808
+ "epoch": 136.99,
1809
+ "learning_rate": 0.0001,
1810
+ "loss": 0.0048,
1811
+ "step": 1096
1812
+ },
1813
+ {
1814
+ "epoch": 137.5,
1815
+ "learning_rate": 0.0001,
1816
+ "loss": 0.0055,
1817
+ "step": 1100
1818
+ },
1819
+ {
1820
+ "epoch": 137.99,
1821
+ "learning_rate": 0.0001,
1822
+ "loss": 0.0048,
1823
+ "step": 1104
1824
+ },
1825
+ {
1826
+ "epoch": 138.5,
1827
+ "learning_rate": 0.0001,
1828
+ "loss": 0.0054,
1829
+ "step": 1108
1830
+ },
1831
+ {
1832
+ "epoch": 138.99,
1833
+ "learning_rate": 0.0001,
1834
+ "loss": 0.0051,
1835
+ "step": 1112
1836
+ },
1837
+ {
1838
+ "epoch": 139.5,
1839
+ "learning_rate": 0.0001,
1840
+ "loss": 0.0047,
1841
+ "step": 1116
1842
+ },
1843
+ {
1844
+ "epoch": 139.99,
1845
+ "learning_rate": 0.0001,
1846
+ "loss": 0.0045,
1847
+ "step": 1120
1848
+ },
1849
+ {
1850
+ "epoch": 140.5,
1851
+ "learning_rate": 0.0001,
1852
+ "loss": 0.0087,
1853
+ "step": 1124
1854
+ },
1855
+ {
1856
+ "epoch": 140.99,
1857
+ "learning_rate": 0.0001,
1858
+ "loss": 0.0043,
1859
+ "step": 1128
1860
+ },
1861
+ {
1862
+ "epoch": 141.5,
1863
+ "learning_rate": 0.0001,
1864
+ "loss": 0.0041,
1865
+ "step": 1132
1866
+ },
1867
+ {
1868
+ "epoch": 141.99,
1869
+ "learning_rate": 0.0001,
1870
+ "loss": 0.0047,
1871
+ "step": 1136
1872
+ },
1873
+ {
1874
+ "epoch": 142.5,
1875
+ "learning_rate": 0.0001,
1876
+ "loss": 0.0046,
1877
+ "step": 1140
1878
+ },
1879
+ {
1880
+ "epoch": 142.99,
1881
+ "learning_rate": 0.0001,
1882
+ "loss": 0.0042,
1883
+ "step": 1144
1884
+ },
1885
+ {
1886
+ "epoch": 143.5,
1887
+ "learning_rate": 0.0001,
1888
+ "loss": 0.0066,
1889
+ "step": 1148
1890
+ },
1891
+ {
1892
+ "epoch": 143.99,
1893
+ "learning_rate": 0.0001,
1894
+ "loss": 0.0052,
1895
+ "step": 1152
1896
+ },
1897
+ {
1898
+ "epoch": 143.99,
1899
+ "eval_exact_match": 0.506454816285998,
1900
+ "eval_exec": 0.5402184707050646,
1901
+ "eval_loss": 0.39830923080444336,
1902
+ "eval_runtime": 975.7311,
1903
+ "eval_samples_per_second": 1.332,
1904
+ "step": 1152
1905
+ },
1906
+ {
1907
+ "epoch": 144.5,
1908
+ "learning_rate": 0.0001,
1909
+ "loss": 0.0048,
1910
+ "step": 1156
1911
+ },
1912
+ {
1913
+ "epoch": 144.99,
1914
+ "learning_rate": 0.0001,
1915
+ "loss": 0.0051,
1916
+ "step": 1160
1917
+ },
1918
+ {
1919
+ "epoch": 145.5,
1920
+ "learning_rate": 0.0001,
1921
+ "loss": 0.0036,
1922
+ "step": 1164
1923
+ },
1924
+ {
1925
+ "epoch": 145.99,
1926
+ "learning_rate": 0.0001,
1927
+ "loss": 0.004,
1928
+ "step": 1168
1929
+ },
1930
+ {
1931
+ "epoch": 146.5,
1932
+ "learning_rate": 0.0001,
1933
+ "loss": 0.0034,
1934
+ "step": 1172
1935
+ },
1936
+ {
1937
+ "epoch": 146.99,
1938
+ "learning_rate": 0.0001,
1939
+ "loss": 0.0041,
1940
+ "step": 1176
1941
+ },
1942
+ {
1943
+ "epoch": 147.5,
1944
+ "learning_rate": 0.0001,
1945
+ "loss": 0.0048,
1946
+ "step": 1180
1947
+ },
1948
+ {
1949
+ "epoch": 147.99,
1950
+ "learning_rate": 0.0001,
1951
+ "loss": 0.0043,
1952
+ "step": 1184
1953
+ },
1954
+ {
1955
+ "epoch": 148.5,
1956
+ "learning_rate": 0.0001,
1957
+ "loss": 0.0041,
1958
+ "step": 1188
1959
+ },
1960
+ {
1961
+ "epoch": 148.99,
1962
+ "learning_rate": 0.0001,
1963
+ "loss": 0.0048,
1964
+ "step": 1192
1965
+ },
1966
+ {
1967
+ "epoch": 149.5,
1968
+ "learning_rate": 0.0001,
1969
+ "loss": 0.0038,
1970
+ "step": 1196
1971
+ },
1972
+ {
1973
+ "epoch": 149.99,
1974
+ "learning_rate": 0.0001,
1975
+ "loss": 0.004,
1976
+ "step": 1200
1977
+ },
1978
+ {
1979
+ "epoch": 150.5,
1980
+ "learning_rate": 0.0001,
1981
+ "loss": 0.0046,
1982
+ "step": 1204
1983
+ },
1984
+ {
1985
+ "epoch": 150.99,
1986
+ "learning_rate": 0.0001,
1987
+ "loss": 0.0063,
1988
+ "step": 1208
1989
+ },
1990
+ {
1991
+ "epoch": 151.5,
1992
+ "learning_rate": 0.0001,
1993
+ "loss": 0.0043,
1994
+ "step": 1212
1995
+ },
1996
+ {
1997
+ "epoch": 151.99,
1998
+ "learning_rate": 0.0001,
1999
+ "loss": 0.0084,
2000
+ "step": 1216
2001
+ },
2002
+ {
2003
+ "epoch": 151.99,
2004
+ "eval_exact_match": 0.44885799404170806,
2005
+ "eval_exec": 0.5014895729890765,
2006
+ "eval_loss": 0.34375911951065063,
2007
+ "eval_runtime": 974.1005,
2008
+ "eval_samples_per_second": 1.335,
2009
+ "step": 1216
2010
+ },
2011
+ {
2012
+ "epoch": 152.5,
2013
+ "learning_rate": 0.0001,
2014
+ "loss": 0.0114,
2015
+ "step": 1220
2016
+ },
2017
+ {
2018
+ "epoch": 152.99,
2019
+ "learning_rate": 0.0001,
2020
+ "loss": 0.0044,
2021
+ "step": 1224
2022
+ },
2023
+ {
2024
+ "epoch": 153.5,
2025
+ "learning_rate": 0.0001,
2026
+ "loss": 0.0037,
2027
+ "step": 1228
2028
+ },
2029
+ {
2030
+ "epoch": 153.99,
2031
+ "learning_rate": 0.0001,
2032
+ "loss": 0.0045,
2033
+ "step": 1232
2034
+ },
2035
+ {
2036
+ "epoch": 154.5,
2037
+ "learning_rate": 0.0001,
2038
+ "loss": 0.0034,
2039
+ "step": 1236
2040
+ },
2041
+ {
2042
+ "epoch": 154.99,
2043
+ "learning_rate": 0.0001,
2044
+ "loss": 0.0038,
2045
+ "step": 1240
2046
+ },
2047
+ {
2048
+ "epoch": 155.5,
2049
+ "learning_rate": 0.0001,
2050
+ "loss": 0.0027,
2051
+ "step": 1244
2052
+ },
2053
+ {
2054
+ "epoch": 155.99,
2055
+ "learning_rate": 0.0001,
2056
+ "loss": 0.0028,
2057
+ "step": 1248
2058
+ },
2059
+ {
2060
+ "epoch": 156.5,
2061
+ "learning_rate": 0.0001,
2062
+ "loss": 0.0039,
2063
+ "step": 1252
2064
+ },
2065
+ {
2066
+ "epoch": 156.99,
2067
+ "learning_rate": 0.0001,
2068
+ "loss": 0.0037,
2069
+ "step": 1256
2070
+ },
2071
+ {
2072
+ "epoch": 157.5,
2073
+ "learning_rate": 0.0001,
2074
+ "loss": 0.0029,
2075
+ "step": 1260
2076
+ },
2077
+ {
2078
+ "epoch": 157.99,
2079
+ "learning_rate": 0.0001,
2080
+ "loss": 0.0028,
2081
+ "step": 1264
2082
+ },
2083
+ {
2084
+ "epoch": 158.5,
2085
+ "learning_rate": 0.0001,
2086
+ "loss": 0.0048,
2087
+ "step": 1268
2088
+ },
2089
+ {
2090
+ "epoch": 158.99,
2091
+ "learning_rate": 0.0001,
2092
+ "loss": 0.0034,
2093
+ "step": 1272
2094
+ },
2095
+ {
2096
+ "epoch": 159.5,
2097
+ "learning_rate": 0.0001,
2098
+ "loss": 0.0033,
2099
+ "step": 1276
2100
+ },
2101
+ {
2102
+ "epoch": 159.99,
2103
+ "learning_rate": 0.0001,
2104
+ "loss": 0.0036,
2105
+ "step": 1280
2106
+ },
2107
+ {
2108
+ "epoch": 159.99,
2109
+ "eval_exact_match": 0.5153922542204568,
2110
+ "eval_exec": 0.5441906653426017,
2111
+ "eval_loss": 0.41516825556755066,
2112
+ "eval_runtime": 976.5716,
2113
+ "eval_samples_per_second": 1.331,
2114
+ "step": 1280
2115
+ },
2116
+ {
2117
+ "epoch": 160.5,
2118
+ "learning_rate": 0.0001,
2119
+ "loss": 0.0035,
2120
+ "step": 1284
2121
+ },
2122
+ {
2123
+ "epoch": 160.99,
2124
+ "learning_rate": 0.0001,
2125
+ "loss": 0.0032,
2126
+ "step": 1288
2127
+ },
2128
+ {
2129
+ "epoch": 161.5,
2130
+ "learning_rate": 0.0001,
2131
+ "loss": 0.0028,
2132
+ "step": 1292
2133
+ },
2134
+ {
2135
+ "epoch": 161.99,
2136
+ "learning_rate": 0.0001,
2137
+ "loss": 0.0039,
2138
+ "step": 1296
2139
+ },
2140
+ {
2141
+ "epoch": 162.5,
2142
+ "learning_rate": 0.0001,
2143
+ "loss": 0.0034,
2144
+ "step": 1300
2145
+ },
2146
+ {
2147
+ "epoch": 162.99,
2148
+ "learning_rate": 0.0001,
2149
+ "loss": 0.0037,
2150
+ "step": 1304
2151
+ },
2152
+ {
2153
+ "epoch": 163.5,
2154
+ "learning_rate": 0.0001,
2155
+ "loss": 0.0035,
2156
+ "step": 1308
2157
+ },
2158
+ {
2159
+ "epoch": 163.99,
2160
+ "learning_rate": 0.0001,
2161
+ "loss": 0.0038,
2162
+ "step": 1312
2163
+ },
2164
+ {
2165
+ "epoch": 164.5,
2166
+ "learning_rate": 0.0001,
2167
+ "loss": 0.0046,
2168
+ "step": 1316
2169
+ },
2170
+ {
2171
+ "epoch": 164.99,
2172
+ "learning_rate": 0.0001,
2173
+ "loss": 0.0035,
2174
+ "step": 1320
2175
+ },
2176
+ {
2177
+ "epoch": 165.5,
2178
+ "learning_rate": 0.0001,
2179
+ "loss": 0.0035,
2180
+ "step": 1324
2181
+ },
2182
+ {
2183
+ "epoch": 165.99,
2184
+ "learning_rate": 0.0001,
2185
+ "loss": 0.0033,
2186
+ "step": 1328
2187
+ },
2188
+ {
2189
+ "epoch": 166.5,
2190
+ "learning_rate": 0.0001,
2191
+ "loss": 0.0039,
2192
+ "step": 1332
2193
+ },
2194
+ {
2195
+ "epoch": 166.99,
2196
+ "learning_rate": 0.0001,
2197
+ "loss": 0.0029,
2198
+ "step": 1336
2199
+ },
2200
+ {
2201
+ "epoch": 167.5,
2202
+ "learning_rate": 0.0001,
2203
+ "loss": 0.0031,
2204
+ "step": 1340
2205
+ },
2206
+ {
2207
+ "epoch": 167.99,
2208
+ "learning_rate": 0.0001,
2209
+ "loss": 0.0029,
2210
+ "step": 1344
2211
+ },
2212
+ {
2213
+ "epoch": 167.99,
2214
+ "eval_exact_match": 0.5143992055610725,
2215
+ "eval_exec": 0.5382323733862959,
2216
+ "eval_loss": 0.4463094472885132,
2217
+ "eval_runtime": 955.823,
2218
+ "eval_samples_per_second": 1.36,
2219
+ "step": 1344
2220
+ },
2221
+ {
2222
+ "epoch": 168.5,
2223
+ "learning_rate": 0.0001,
2224
+ "loss": 0.003,
2225
+ "step": 1348
2226
+ },
2227
+ {
2228
+ "epoch": 168.99,
2229
+ "learning_rate": 0.0001,
2230
+ "loss": 0.0049,
2231
+ "step": 1352
2232
+ },
2233
+ {
2234
+ "epoch": 169.5,
2235
+ "learning_rate": 0.0001,
2236
+ "loss": 0.0042,
2237
+ "step": 1356
2238
+ },
2239
+ {
2240
+ "epoch": 169.99,
2241
+ "learning_rate": 0.0001,
2242
+ "loss": 0.0037,
2243
+ "step": 1360
2244
+ },
2245
+ {
2246
+ "epoch": 170.5,
2247
+ "learning_rate": 0.0001,
2248
+ "loss": 0.0038,
2249
+ "step": 1364
2250
+ },
2251
+ {
2252
+ "epoch": 170.99,
2253
+ "learning_rate": 0.0001,
2254
+ "loss": 0.0022,
2255
+ "step": 1368
2256
+ },
2257
+ {
2258
+ "epoch": 171.5,
2259
+ "learning_rate": 0.0001,
2260
+ "loss": 0.0031,
2261
+ "step": 1372
2262
+ },
2263
+ {
2264
+ "epoch": 171.99,
2265
+ "learning_rate": 0.0001,
2266
+ "loss": 0.0032,
2267
+ "step": 1376
2268
+ },
2269
+ {
2270
+ "epoch": 172.5,
2271
+ "learning_rate": 0.0001,
2272
+ "loss": 0.0024,
2273
+ "step": 1380
2274
+ },
2275
+ {
2276
+ "epoch": 172.99,
2277
+ "learning_rate": 0.0001,
2278
+ "loss": 0.0024,
2279
+ "step": 1384
2280
+ },
2281
+ {
2282
+ "epoch": 173.5,
2283
+ "learning_rate": 0.0001,
2284
+ "loss": 0.003,
2285
+ "step": 1388
2286
+ },
2287
+ {
2288
+ "epoch": 173.99,
2289
+ "learning_rate": 0.0001,
2290
+ "loss": 0.0035,
2291
+ "step": 1392
2292
+ },
2293
+ {
2294
+ "epoch": 174.5,
2295
+ "learning_rate": 0.0001,
2296
+ "loss": 0.0035,
2297
+ "step": 1396
2298
+ },
2299
+ {
2300
+ "epoch": 174.99,
2301
+ "learning_rate": 0.0001,
2302
+ "loss": 0.0029,
2303
+ "step": 1400
2304
+ },
2305
+ {
2306
+ "epoch": 175.5,
2307
+ "learning_rate": 0.0001,
2308
+ "loss": 0.0034,
2309
+ "step": 1404
2310
+ },
2311
+ {
2312
+ "epoch": 175.99,
2313
+ "learning_rate": 0.0001,
2314
+ "loss": 0.0029,
2315
+ "step": 1408
2316
+ },
2317
+ {
2318
+ "epoch": 175.99,
2319
+ "eval_exact_match": 0.5074478649453823,
2320
+ "eval_exec": 0.5342601787487586,
2321
+ "eval_loss": 0.4188894033432007,
2322
+ "eval_runtime": 941.4278,
2323
+ "eval_samples_per_second": 1.381,
2324
+ "step": 1408
2325
+ },
2326
+ {
2327
+ "epoch": 176.5,
2328
+ "learning_rate": 0.0001,
2329
+ "loss": 0.0036,
2330
+ "step": 1412
2331
+ },
2332
+ {
2333
+ "epoch": 176.99,
2334
+ "learning_rate": 0.0001,
2335
+ "loss": 0.0024,
2336
+ "step": 1416
2337
+ },
2338
+ {
2339
+ "epoch": 177.5,
2340
+ "learning_rate": 0.0001,
2341
+ "loss": 0.0025,
2342
+ "step": 1420
2343
+ },
2344
+ {
2345
+ "epoch": 177.99,
2346
+ "learning_rate": 0.0001,
2347
+ "loss": 0.0029,
2348
+ "step": 1424
2349
+ },
2350
+ {
2351
+ "epoch": 178.5,
2352
+ "learning_rate": 0.0001,
2353
+ "loss": 0.0021,
2354
+ "step": 1428
2355
+ },
2356
+ {
2357
+ "epoch": 178.99,
2358
+ "learning_rate": 0.0001,
2359
+ "loss": 0.0038,
2360
+ "step": 1432
2361
+ },
2362
+ {
2363
+ "epoch": 179.5,
2364
+ "learning_rate": 0.0001,
2365
+ "loss": 0.0057,
2366
+ "step": 1436
2367
+ },
2368
+ {
2369
+ "epoch": 179.99,
2370
+ "learning_rate": 0.0001,
2371
+ "loss": 0.0033,
2372
+ "step": 1440
2373
+ },
2374
+ {
2375
+ "epoch": 180.5,
2376
+ "learning_rate": 0.0001,
2377
+ "loss": 0.0025,
2378
+ "step": 1444
2379
+ },
2380
+ {
2381
+ "epoch": 180.99,
2382
+ "learning_rate": 0.0001,
2383
+ "loss": 0.0027,
2384
+ "step": 1448
2385
+ },
2386
+ {
2387
+ "epoch": 181.5,
2388
+ "learning_rate": 0.0001,
2389
+ "loss": 0.0024,
2390
+ "step": 1452
2391
+ },
2392
+ {
2393
+ "epoch": 181.99,
2394
+ "learning_rate": 0.0001,
2395
+ "loss": 0.003,
2396
+ "step": 1456
2397
+ },
2398
+ {
2399
+ "epoch": 182.5,
2400
+ "learning_rate": 0.0001,
2401
+ "loss": 0.0049,
2402
+ "step": 1460
2403
+ },
2404
+ {
2405
+ "epoch": 182.99,
2406
+ "learning_rate": 0.0001,
2407
+ "loss": 0.0051,
2408
+ "step": 1464
2409
+ },
2410
+ {
2411
+ "epoch": 183.5,
2412
+ "learning_rate": 0.0001,
2413
+ "loss": 0.0025,
2414
+ "step": 1468
2415
+ },
2416
+ {
2417
+ "epoch": 183.99,
2418
+ "learning_rate": 0.0001,
2419
+ "loss": 0.0031,
2420
+ "step": 1472
2421
+ },
2422
+ {
2423
+ "epoch": 183.99,
2424
+ "eval_exact_match": 0.519364448857994,
2425
+ "eval_exec": 0.5441906653426017,
2426
+ "eval_loss": 0.422220915555954,
2427
+ "eval_runtime": 973.3607,
2428
+ "eval_samples_per_second": 1.336,
2429
+ "step": 1472
2430
+ },
2431
+ {
2432
+ "epoch": 184.5,
2433
+ "learning_rate": 0.0001,
2434
+ "loss": 0.0025,
2435
+ "step": 1476
2436
+ },
2437
+ {
2438
+ "epoch": 184.99,
2439
+ "learning_rate": 0.0001,
2440
+ "loss": 0.0026,
2441
+ "step": 1480
2442
+ },
2443
+ {
2444
+ "epoch": 185.5,
2445
+ "learning_rate": 0.0001,
2446
+ "loss": 0.0028,
2447
+ "step": 1484
2448
+ },
2449
+ {
2450
+ "epoch": 185.99,
2451
+ "learning_rate": 0.0001,
2452
+ "loss": 0.0023,
2453
+ "step": 1488
2454
+ },
2455
+ {
2456
+ "epoch": 186.5,
2457
+ "learning_rate": 0.0001,
2458
+ "loss": 0.0037,
2459
+ "step": 1492
2460
+ },
2461
+ {
2462
+ "epoch": 186.99,
2463
+ "learning_rate": 0.0001,
2464
+ "loss": 0.0038,
2465
+ "step": 1496
2466
+ },
2467
+ {
2468
+ "epoch": 187.5,
2469
+ "learning_rate": 0.0001,
2470
+ "loss": 0.0029,
2471
+ "step": 1500
2472
+ },
2473
+ {
2474
+ "epoch": 187.99,
2475
+ "learning_rate": 0.0001,
2476
+ "loss": 0.0025,
2477
+ "step": 1504
2478
+ },
2479
+ {
2480
+ "epoch": 188.5,
2481
+ "learning_rate": 0.0001,
2482
+ "loss": 0.0026,
2483
+ "step": 1508
2484
+ },
2485
+ {
2486
+ "epoch": 188.99,
2487
+ "learning_rate": 0.0001,
2488
+ "loss": 0.0026,
2489
+ "step": 1512
2490
+ },
2491
+ {
2492
+ "epoch": 189.5,
2493
+ "learning_rate": 0.0001,
2494
+ "loss": 0.002,
2495
+ "step": 1516
2496
+ },
2497
+ {
2498
+ "epoch": 189.99,
2499
+ "learning_rate": 0.0001,
2500
+ "loss": 0.0021,
2501
+ "step": 1520
2502
+ },
2503
+ {
2504
+ "epoch": 190.5,
2505
+ "learning_rate": 0.0001,
2506
+ "loss": 0.0019,
2507
+ "step": 1524
2508
+ },
2509
+ {
2510
+ "epoch": 190.99,
2511
+ "learning_rate": 0.0001,
2512
+ "loss": 0.0025,
2513
+ "step": 1528
2514
+ },
2515
+ {
2516
+ "epoch": 191.5,
2517
+ "learning_rate": 0.0001,
2518
+ "loss": 0.0027,
2519
+ "step": 1532
2520
+ },
2521
+ {
2522
+ "epoch": 191.99,
2523
+ "learning_rate": 0.0001,
2524
+ "loss": 0.0034,
2525
+ "step": 1536
2526
+ },
2527
+ {
2528
+ "epoch": 191.99,
2529
+ "eval_exact_match": 0.5203574975173784,
2530
+ "eval_exec": 0.5491559086395233,
2531
+ "eval_loss": 0.44573745131492615,
2532
+ "eval_runtime": 974.7523,
2533
+ "eval_samples_per_second": 1.334,
2534
+ "step": 1536
2535
+ },
2536
+ {
2537
+ "epoch": 192.5,
2538
+ "learning_rate": 0.0001,
2539
+ "loss": 0.0026,
2540
+ "step": 1540
2541
+ },
2542
+ {
2543
+ "epoch": 192.99,
2544
+ "learning_rate": 0.0001,
2545
+ "loss": 0.0027,
2546
+ "step": 1544
2547
+ },
2548
+ {
2549
+ "epoch": 193.5,
2550
+ "learning_rate": 0.0001,
2551
+ "loss": 0.003,
2552
+ "step": 1548
2553
+ },
2554
+ {
2555
+ "epoch": 193.99,
2556
+ "learning_rate": 0.0001,
2557
+ "loss": 0.0026,
2558
+ "step": 1552
2559
+ },
2560
+ {
2561
+ "epoch": 194.5,
2562
+ "learning_rate": 0.0001,
2563
+ "loss": 0.0029,
2564
+ "step": 1556
2565
+ },
2566
+ {
2567
+ "epoch": 194.99,
2568
+ "learning_rate": 0.0001,
2569
+ "loss": 0.0026,
2570
+ "step": 1560
2571
+ },
2572
+ {
2573
+ "epoch": 195.5,
2574
+ "learning_rate": 0.0001,
2575
+ "loss": 0.002,
2576
+ "step": 1564
2577
+ },
2578
+ {
2579
+ "epoch": 195.99,
2580
+ "learning_rate": 0.0001,
2581
+ "loss": 0.002,
2582
+ "step": 1568
2583
+ },
2584
+ {
2585
+ "epoch": 196.5,
2586
+ "learning_rate": 0.0001,
2587
+ "loss": 0.0027,
2588
+ "step": 1572
2589
+ },
2590
+ {
2591
+ "epoch": 196.99,
2592
+ "learning_rate": 0.0001,
2593
+ "loss": 0.002,
2594
+ "step": 1576
2595
+ },
2596
+ {
2597
+ "epoch": 197.5,
2598
+ "learning_rate": 0.0001,
2599
+ "loss": 0.0028,
2600
+ "step": 1580
2601
+ },
2602
+ {
2603
+ "epoch": 197.99,
2604
+ "learning_rate": 0.0001,
2605
+ "loss": 0.003,
2606
+ "step": 1584
2607
+ },
2608
+ {
2609
+ "epoch": 198.5,
2610
+ "learning_rate": 0.0001,
2611
+ "loss": 0.0022,
2612
+ "step": 1588
2613
+ },
2614
+ {
2615
+ "epoch": 198.99,
2616
+ "learning_rate": 0.0001,
2617
+ "loss": 0.0022,
2618
+ "step": 1592
2619
+ },
2620
+ {
2621
+ "epoch": 199.5,
2622
+ "learning_rate": 0.0001,
2623
+ "loss": 0.0022,
2624
+ "step": 1596
2625
+ },
2626
+ {
2627
+ "epoch": 199.99,
2628
+ "learning_rate": 0.0001,
2629
+ "loss": 0.0022,
2630
+ "step": 1600
2631
+ },
2632
+ {
2633
+ "epoch": 199.99,
2634
+ "eval_exact_match": 0.5114200595829196,
2635
+ "eval_exec": 0.5521350546176763,
2636
+ "eval_loss": 0.4328438639640808,
2637
+ "eval_runtime": 959.4133,
2638
+ "eval_samples_per_second": 1.355,
2639
+ "step": 1600
2640
+ },
2641
+ {
2642
+ "epoch": 200.5,
2643
+ "learning_rate": 0.0001,
2644
+ "loss": 0.0022,
2645
+ "step": 1604
2646
+ },
2647
+ {
2648
+ "epoch": 200.99,
2649
+ "learning_rate": 0.0001,
2650
+ "loss": 0.0019,
2651
+ "step": 1608
2652
+ },
2653
+ {
2654
+ "epoch": 201.5,
2655
+ "learning_rate": 0.0001,
2656
+ "loss": 0.0024,
2657
+ "step": 1612
2658
+ },
2659
+ {
2660
+ "epoch": 201.99,
2661
+ "learning_rate": 0.0001,
2662
+ "loss": 0.0021,
2663
+ "step": 1616
2664
+ },
2665
+ {
2666
+ "epoch": 202.5,
2667
+ "learning_rate": 0.0001,
2668
+ "loss": 0.0029,
2669
+ "step": 1620
2670
+ },
2671
+ {
2672
+ "epoch": 202.99,
2673
+ "learning_rate": 0.0001,
2674
+ "loss": 0.0024,
2675
+ "step": 1624
2676
+ },
2677
+ {
2678
+ "epoch": 203.5,
2679
+ "learning_rate": 0.0001,
2680
+ "loss": 0.0026,
2681
+ "step": 1628
2682
+ },
2683
+ {
2684
+ "epoch": 203.99,
2685
+ "learning_rate": 0.0001,
2686
+ "loss": 0.0019,
2687
+ "step": 1632
2688
+ },
2689
+ {
2690
+ "epoch": 204.5,
2691
+ "learning_rate": 0.0001,
2692
+ "loss": 0.0022,
2693
+ "step": 1636
2694
+ },
2695
+ {
2696
+ "epoch": 204.99,
2697
+ "learning_rate": 0.0001,
2698
+ "loss": 0.0021,
2699
+ "step": 1640
2700
+ },
2701
+ {
2702
+ "epoch": 205.5,
2703
+ "learning_rate": 0.0001,
2704
+ "loss": 0.0026,
2705
+ "step": 1644
2706
+ },
2707
+ {
2708
+ "epoch": 205.99,
2709
+ "learning_rate": 0.0001,
2710
+ "loss": 0.0023,
2711
+ "step": 1648
2712
+ },
2713
+ {
2714
+ "epoch": 206.5,
2715
+ "learning_rate": 0.0001,
2716
+ "loss": 0.002,
2717
+ "step": 1652
2718
+ },
2719
+ {
2720
+ "epoch": 206.99,
2721
+ "learning_rate": 0.0001,
2722
+ "loss": 0.0026,
2723
+ "step": 1656
2724
+ },
2725
+ {
2726
+ "epoch": 207.5,
2727
+ "learning_rate": 0.0001,
2728
+ "loss": 0.0027,
2729
+ "step": 1660
2730
+ },
2731
+ {
2732
+ "epoch": 207.99,
2733
+ "learning_rate": 0.0001,
2734
+ "loss": 0.0022,
2735
+ "step": 1664
2736
+ },
2737
+ {
2738
+ "epoch": 207.99,
2739
+ "eval_exact_match": 0.5153922542204568,
2740
+ "eval_exec": 0.5422045680238332,
2741
+ "eval_loss": 0.46325209736824036,
2742
+ "eval_runtime": 948.9363,
2743
+ "eval_samples_per_second": 1.37,
2744
+ "step": 1664
2745
+ },
2746
+ {
2747
+ "epoch": 208.5,
2748
+ "learning_rate": 0.0001,
2749
+ "loss": 0.0028,
2750
+ "step": 1668
2751
+ },
2752
+ {
2753
+ "epoch": 208.99,
2754
+ "learning_rate": 0.0001,
2755
+ "loss": 0.003,
2756
+ "step": 1672
2757
+ },
2758
+ {
2759
+ "epoch": 209.5,
2760
+ "learning_rate": 0.0001,
2761
+ "loss": 0.0026,
2762
+ "step": 1676
2763
+ },
2764
+ {
2765
+ "epoch": 209.99,
2766
+ "learning_rate": 0.0001,
2767
+ "loss": 0.0015,
2768
+ "step": 1680
2769
+ },
2770
+ {
2771
+ "epoch": 210.5,
2772
+ "learning_rate": 0.0001,
2773
+ "loss": 0.0021,
2774
+ "step": 1684
2775
+ },
2776
+ {
2777
+ "epoch": 210.99,
2778
+ "learning_rate": 0.0001,
2779
+ "loss": 0.0025,
2780
+ "step": 1688
2781
+ },
2782
+ {
2783
+ "epoch": 211.5,
2784
+ "learning_rate": 0.0001,
2785
+ "loss": 0.0014,
2786
+ "step": 1692
2787
+ },
2788
+ {
2789
+ "epoch": 211.99,
2790
+ "learning_rate": 0.0001,
2791
+ "loss": 0.0014,
2792
+ "step": 1696
2793
+ },
2794
+ {
2795
+ "epoch": 212.5,
2796
+ "learning_rate": 0.0001,
2797
+ "loss": 0.0015,
2798
+ "step": 1700
2799
+ },
2800
+ {
2801
+ "epoch": 212.99,
2802
+ "learning_rate": 0.0001,
2803
+ "loss": 0.0022,
2804
+ "step": 1704
2805
+ },
2806
+ {
2807
+ "epoch": 213.5,
2808
+ "learning_rate": 0.0001,
2809
+ "loss": 0.0024,
2810
+ "step": 1708
2811
+ },
2812
+ {
2813
+ "epoch": 213.99,
2814
+ "learning_rate": 0.0001,
2815
+ "loss": 0.0021,
2816
+ "step": 1712
2817
+ },
2818
+ {
2819
+ "epoch": 214.5,
2820
+ "learning_rate": 0.0001,
2821
+ "loss": 0.0022,
2822
+ "step": 1716
2823
+ },
2824
+ {
2825
+ "epoch": 214.99,
2826
+ "learning_rate": 0.0001,
2827
+ "loss": 0.0016,
2828
+ "step": 1720
2829
+ },
2830
+ {
2831
+ "epoch": 215.5,
2832
+ "learning_rate": 0.0001,
2833
+ "loss": 0.0016,
2834
+ "step": 1724
2835
+ },
2836
+ {
2837
+ "epoch": 215.99,
2838
+ "learning_rate": 0.0001,
2839
+ "loss": 0.002,
2840
+ "step": 1728
2841
+ },
2842
+ {
2843
+ "epoch": 215.99,
2844
+ "eval_exact_match": 0.5134061569016882,
2845
+ "eval_exec": 0.551142005958292,
2846
+ "eval_loss": 0.4671519100666046,
2847
+ "eval_runtime": 968.39,
2848
+ "eval_samples_per_second": 1.342,
2849
+ "step": 1728
2850
+ },
2851
+ {
2852
+ "epoch": 216.5,
2853
+ "learning_rate": 0.0001,
2854
+ "loss": 0.002,
2855
+ "step": 1732
2856
+ },
2857
+ {
2858
+ "epoch": 216.99,
2859
+ "learning_rate": 0.0001,
2860
+ "loss": 0.0027,
2861
+ "step": 1736
2862
+ },
2863
+ {
2864
+ "epoch": 217.5,
2865
+ "learning_rate": 0.0001,
2866
+ "loss": 0.0024,
2867
+ "step": 1740
2868
+ },
2869
+ {
2870
+ "epoch": 217.99,
2871
+ "learning_rate": 0.0001,
2872
+ "loss": 0.0026,
2873
+ "step": 1744
2874
+ },
2875
+ {
2876
+ "epoch": 218.5,
2877
+ "learning_rate": 0.0001,
2878
+ "loss": 0.0024,
2879
+ "step": 1748
2880
+ },
2881
+ {
2882
+ "epoch": 218.99,
2883
+ "learning_rate": 0.0001,
2884
+ "loss": 0.0022,
2885
+ "step": 1752
2886
+ },
2887
+ {
2888
+ "epoch": 219.5,
2889
+ "learning_rate": 0.0001,
2890
+ "loss": 0.002,
2891
+ "step": 1756
2892
+ },
2893
+ {
2894
+ "epoch": 219.99,
2895
+ "learning_rate": 0.0001,
2896
+ "loss": 0.002,
2897
+ "step": 1760
2898
+ },
2899
+ {
2900
+ "epoch": 220.5,
2901
+ "learning_rate": 0.0001,
2902
+ "loss": 0.0018,
2903
+ "step": 1764
2904
+ },
2905
+ {
2906
+ "epoch": 220.99,
2907
+ "learning_rate": 0.0001,
2908
+ "loss": 0.0014,
2909
+ "step": 1768
2910
+ },
2911
+ {
2912
+ "epoch": 221.5,
2913
+ "learning_rate": 0.0001,
2914
+ "loss": 0.0021,
2915
+ "step": 1772
2916
+ },
2917
+ {
2918
+ "epoch": 221.99,
2919
+ "learning_rate": 0.0001,
2920
+ "loss": 0.0019,
2921
+ "step": 1776
2922
+ },
2923
+ {
2924
+ "epoch": 222.5,
2925
+ "learning_rate": 0.0001,
2926
+ "loss": 0.0014,
2927
+ "step": 1780
2928
+ },
2929
+ {
2930
+ "epoch": 222.99,
2931
+ "learning_rate": 0.0001,
2932
+ "loss": 0.0012,
2933
+ "step": 1784
2934
+ },
2935
+ {
2936
+ "epoch": 223.5,
2937
+ "learning_rate": 0.0001,
2938
+ "loss": 0.0012,
2939
+ "step": 1788
2940
+ },
2941
+ {
2942
+ "epoch": 223.99,
2943
+ "learning_rate": 0.0001,
2944
+ "loss": 0.0017,
2945
+ "step": 1792
2946
+ },
2947
+ {
2948
+ "epoch": 223.99,
2949
+ "eval_exact_match": 0.5143992055610725,
2950
+ "eval_exec": 0.5402184707050646,
2951
+ "eval_loss": 0.4584057629108429,
2952
+ "eval_runtime": 959.2229,
2953
+ "eval_samples_per_second": 1.355,
2954
+ "step": 1792
2955
+ },
2956
+ {
2957
+ "epoch": 224.5,
2958
+ "learning_rate": 0.0001,
2959
+ "loss": 0.0018,
2960
+ "step": 1796
2961
+ },
2962
+ {
2963
+ "epoch": 224.99,
2964
+ "learning_rate": 0.0001,
2965
+ "loss": 0.0017,
2966
+ "step": 1800
2967
+ },
2968
+ {
2969
+ "epoch": 225.5,
2970
+ "learning_rate": 0.0001,
2971
+ "loss": 0.0015,
2972
+ "step": 1804
2973
+ },
2974
+ {
2975
+ "epoch": 225.99,
2976
+ "learning_rate": 0.0001,
2977
+ "loss": 0.0024,
2978
+ "step": 1808
2979
+ },
2980
+ {
2981
+ "epoch": 226.5,
2982
+ "learning_rate": 0.0001,
2983
+ "loss": 0.0018,
2984
+ "step": 1812
2985
+ },
2986
+ {
2987
+ "epoch": 226.99,
2988
+ "learning_rate": 0.0001,
2989
+ "loss": 0.0019,
2990
+ "step": 1816
2991
+ },
2992
+ {
2993
+ "epoch": 227.5,
2994
+ "learning_rate": 0.0001,
2995
+ "loss": 0.0028,
2996
+ "step": 1820
2997
+ },
2998
+ {
2999
+ "epoch": 227.99,
3000
+ "learning_rate": 0.0001,
3001
+ "loss": 0.0015,
3002
+ "step": 1824
3003
+ },
3004
+ {
3005
+ "epoch": 228.5,
3006
+ "learning_rate": 0.0001,
3007
+ "loss": 0.002,
3008
+ "step": 1828
3009
+ },
3010
+ {
3011
+ "epoch": 228.99,
3012
+ "learning_rate": 0.0001,
3013
+ "loss": 0.0025,
3014
+ "step": 1832
3015
+ },
3016
+ {
3017
+ "epoch": 229.5,
3018
+ "learning_rate": 0.0001,
3019
+ "loss": 0.0063,
3020
+ "step": 1836
3021
+ },
3022
+ {
3023
+ "epoch": 229.99,
3024
+ "learning_rate": 0.0001,
3025
+ "loss": 0.0021,
3026
+ "step": 1840
3027
+ },
3028
+ {
3029
+ "epoch": 230.5,
3030
+ "learning_rate": 0.0001,
3031
+ "loss": 0.0022,
3032
+ "step": 1844
3033
+ },
3034
+ {
3035
+ "epoch": 230.99,
3036
+ "learning_rate": 0.0001,
3037
+ "loss": 0.0019,
3038
+ "step": 1848
3039
+ },
3040
+ {
3041
+ "epoch": 231.5,
3042
+ "learning_rate": 0.0001,
3043
+ "loss": 0.0017,
3044
+ "step": 1852
3045
+ },
3046
+ {
3047
+ "epoch": 231.99,
3048
+ "learning_rate": 0.0001,
3049
+ "loss": 0.0019,
3050
+ "step": 1856
3051
+ },
3052
+ {
3053
+ "epoch": 231.99,
3054
+ "eval_exact_match": 0.5253227408142999,
3055
+ "eval_exec": 0.5561072492552135,
3056
+ "eval_loss": 0.44773775339126587,
3057
+ "eval_runtime": 980.451,
3058
+ "eval_samples_per_second": 1.326,
3059
+ "step": 1856
3060
+ }
3061
+ ],
3062
+ "max_steps": 24576,
3063
+ "num_train_epochs": 3072,
3064
+ "total_flos": 7.797838246859244e+18,
3065
+ "trial_name": null,
3066
+ "trial_params": null
3067
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02bb805abc0514b5e7633855b3689de319f73a8be713910355ff40013d0655af
3
+ size 2799