Upload 12 files
Browse files- checkpoint-4000/config.json +60 -0
- checkpoint-4000/generation_config.json +12 -0
- checkpoint-4000/optimizer.pt +3 -0
- checkpoint-4000/pytorch_model.bin +3 -0
- checkpoint-4000/rng_state.pth +3 -0
- checkpoint-4000/scheduler.pt +3 -0
- checkpoint-4000/special_tokens_map.json +110 -0
- checkpoint-4000/spiece.model +3 -0
- checkpoint-4000/tokenizer.json +0 -0
- checkpoint-4000/tokenizer_config.json +117 -0
- checkpoint-4000/trainer_state.json +576 -0
- checkpoint-4000/training_args.bin +3 -0
checkpoint-4000/config.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "google/pegasus-x-large",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"add_bias_logits": false,
|
| 6 |
+
"add_final_layer_norm": true,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"PegasusXForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.1,
|
| 11 |
+
"block_size": 512,
|
| 12 |
+
"bos_token_id": 0,
|
| 13 |
+
"classif_dropout": 0.0,
|
| 14 |
+
"classifier_dropout": 0.0,
|
| 15 |
+
"d_model": 1024,
|
| 16 |
+
"decoder_attention_heads": 16,
|
| 17 |
+
"decoder_ffn_dim": 4096,
|
| 18 |
+
"decoder_layerdrop": 0.0,
|
| 19 |
+
"decoder_layers": 16,
|
| 20 |
+
"decoder_start_token_id": 0,
|
| 21 |
+
"dropout": 0.1,
|
| 22 |
+
"encoder_attention_heads": 16,
|
| 23 |
+
"encoder_ffn_dim": 4096,
|
| 24 |
+
"encoder_layerdrop": 0.0,
|
| 25 |
+
"encoder_layers": 16,
|
| 26 |
+
"eos_token_id": 1,
|
| 27 |
+
"extra_pos_embeddings": 1,
|
| 28 |
+
"force_bos_token_to_be_generated": false,
|
| 29 |
+
"forced_eos_token_id": 1,
|
| 30 |
+
"gradient_checkpointing": false,
|
| 31 |
+
"id2label": {
|
| 32 |
+
"0": "LABEL_0",
|
| 33 |
+
"1": "LABEL_1",
|
| 34 |
+
"2": "LABEL_2"
|
| 35 |
+
},
|
| 36 |
+
"init_std": 0.02,
|
| 37 |
+
"is_encoder_decoder": true,
|
| 38 |
+
"label2id": {
|
| 39 |
+
"LABEL_0": 0,
|
| 40 |
+
"LABEL_1": 1,
|
| 41 |
+
"LABEL_2": 2
|
| 42 |
+
},
|
| 43 |
+
"length_penalty": 0.8,
|
| 44 |
+
"max_length": 16384,
|
| 45 |
+
"max_position_embeddings": 16384,
|
| 46 |
+
"model_type": "pegasus_x",
|
| 47 |
+
"normalize_before": true,
|
| 48 |
+
"normalize_embedding": false,
|
| 49 |
+
"num_beams": 8,
|
| 50 |
+
"num_global_tokens": 128,
|
| 51 |
+
"num_hidden_layers": 16,
|
| 52 |
+
"pad_token_id": 0,
|
| 53 |
+
"scale_embedding": true,
|
| 54 |
+
"stagger_local_blocks": true,
|
| 55 |
+
"static_position_embeddings": true,
|
| 56 |
+
"torch_dtype": "float32",
|
| 57 |
+
"transformers_version": "4.27.4",
|
| 58 |
+
"use_cache": true,
|
| 59 |
+
"vocab_size": 96103
|
| 60 |
+
}
|
checkpoint-4000/generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"decoder_start_token_id": 0,
|
| 5 |
+
"eos_token_id": 1,
|
| 6 |
+
"forced_eos_token_id": 1,
|
| 7 |
+
"length_penalty": 0.8,
|
| 8 |
+
"max_length": 16384,
|
| 9 |
+
"num_beams": 8,
|
| 10 |
+
"pad_token_id": 0,
|
| 11 |
+
"transformers_version": "4.27.4"
|
| 12 |
+
}
|
checkpoint-4000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3dbef991e1dec31549bed0fdc66dba715c068d901d980c64250d2f60caa00c9d
|
| 3 |
+
size 4549646261
|
checkpoint-4000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b34ca2ca0d3ec5c0e317aa3f77f5299a6ffdbf0a38f1f7348a8346b766c9a377
|
| 3 |
+
size 2274845861
|
checkpoint-4000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72f0576613c85f811dd17320e32a76f84e81457422ba7f884392ba925249fe28
|
| 3 |
+
size 14575
|
checkpoint-4000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7f4b5865bdfabcedab071f6ca08201dd914f54a5b590846b5688db29c7216c5
|
| 3 |
+
size 627
|
checkpoint-4000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<mask_1>",
|
| 4 |
+
"<unk_2>",
|
| 5 |
+
"<unk_3>",
|
| 6 |
+
"<unk_4>",
|
| 7 |
+
"<unk_5>",
|
| 8 |
+
"<unk_6>",
|
| 9 |
+
"<unk_7>",
|
| 10 |
+
"<unk_8>",
|
| 11 |
+
"<unk_9>",
|
| 12 |
+
"<unk_10>",
|
| 13 |
+
"<unk_11>",
|
| 14 |
+
"<unk_12>",
|
| 15 |
+
"<unk_13>",
|
| 16 |
+
"<unk_14>",
|
| 17 |
+
"<unk_15>",
|
| 18 |
+
"<unk_16>",
|
| 19 |
+
"<unk_17>",
|
| 20 |
+
"<unk_18>",
|
| 21 |
+
"<unk_19>",
|
| 22 |
+
"<unk_20>",
|
| 23 |
+
"<unk_21>",
|
| 24 |
+
"<unk_22>",
|
| 25 |
+
"<unk_23>",
|
| 26 |
+
"<unk_24>",
|
| 27 |
+
"<unk_25>",
|
| 28 |
+
"<unk_26>",
|
| 29 |
+
"<unk_27>",
|
| 30 |
+
"<unk_28>",
|
| 31 |
+
"<unk_29>",
|
| 32 |
+
"<unk_30>",
|
| 33 |
+
"<unk_31>",
|
| 34 |
+
"<unk_32>",
|
| 35 |
+
"<unk_33>",
|
| 36 |
+
"<unk_34>",
|
| 37 |
+
"<unk_35>",
|
| 38 |
+
"<unk_36>",
|
| 39 |
+
"<unk_37>",
|
| 40 |
+
"<unk_38>",
|
| 41 |
+
"<unk_39>",
|
| 42 |
+
"<unk_40>",
|
| 43 |
+
"<unk_41>",
|
| 44 |
+
"<unk_42>",
|
| 45 |
+
"<unk_43>",
|
| 46 |
+
"<unk_44>",
|
| 47 |
+
"<unk_45>",
|
| 48 |
+
"<unk_46>",
|
| 49 |
+
"<unk_47>",
|
| 50 |
+
"<unk_48>",
|
| 51 |
+
"<unk_49>",
|
| 52 |
+
"<unk_50>",
|
| 53 |
+
"<unk_51>",
|
| 54 |
+
"<unk_52>",
|
| 55 |
+
"<unk_53>",
|
| 56 |
+
"<unk_54>",
|
| 57 |
+
"<unk_55>",
|
| 58 |
+
"<unk_56>",
|
| 59 |
+
"<unk_57>",
|
| 60 |
+
"<unk_58>",
|
| 61 |
+
"<unk_59>",
|
| 62 |
+
"<unk_60>",
|
| 63 |
+
"<unk_61>",
|
| 64 |
+
"<unk_62>",
|
| 65 |
+
"<unk_63>",
|
| 66 |
+
"<unk_64>",
|
| 67 |
+
"<unk_65>",
|
| 68 |
+
"<unk_66>",
|
| 69 |
+
"<unk_67>",
|
| 70 |
+
"<unk_68>",
|
| 71 |
+
"<unk_69>",
|
| 72 |
+
"<unk_70>",
|
| 73 |
+
"<unk_71>",
|
| 74 |
+
"<unk_72>",
|
| 75 |
+
"<unk_73>",
|
| 76 |
+
"<unk_74>",
|
| 77 |
+
"<unk_75>",
|
| 78 |
+
"<unk_76>",
|
| 79 |
+
"<unk_77>",
|
| 80 |
+
"<unk_78>",
|
| 81 |
+
"<unk_79>",
|
| 82 |
+
"<unk_80>",
|
| 83 |
+
"<unk_81>",
|
| 84 |
+
"<unk_82>",
|
| 85 |
+
"<unk_83>",
|
| 86 |
+
"<unk_84>",
|
| 87 |
+
"<unk_85>",
|
| 88 |
+
"<unk_86>",
|
| 89 |
+
"<unk_87>",
|
| 90 |
+
"<unk_88>",
|
| 91 |
+
"<unk_89>",
|
| 92 |
+
"<unk_90>",
|
| 93 |
+
"<unk_91>",
|
| 94 |
+
"<unk_92>",
|
| 95 |
+
"<unk_93>",
|
| 96 |
+
"<unk_94>",
|
| 97 |
+
"<unk_95>",
|
| 98 |
+
"<unk_96>",
|
| 99 |
+
"<unk_97>",
|
| 100 |
+
"<unk_98>",
|
| 101 |
+
"<unk_99>",
|
| 102 |
+
"<unk_100>",
|
| 103 |
+
"<unk_101>",
|
| 104 |
+
"<unk_102>"
|
| 105 |
+
],
|
| 106 |
+
"eos_token": "</s>",
|
| 107 |
+
"mask_token": "<mask_2>",
|
| 108 |
+
"pad_token": "<pad>",
|
| 109 |
+
"unk_token": "<unk>"
|
| 110 |
+
}
|
checkpoint-4000/spiece.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0015189ef36359283fec8b93cf6d9ce51bca37eb1101defc68a53b394913b96c
|
| 3 |
+
size 1912529
|
checkpoint-4000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-4000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,117 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<mask_1>",
|
| 4 |
+
"<unk_2>",
|
| 5 |
+
"<unk_3>",
|
| 6 |
+
"<unk_4>",
|
| 7 |
+
"<unk_5>",
|
| 8 |
+
"<unk_6>",
|
| 9 |
+
"<unk_7>",
|
| 10 |
+
"<unk_8>",
|
| 11 |
+
"<unk_9>",
|
| 12 |
+
"<unk_10>",
|
| 13 |
+
"<unk_11>",
|
| 14 |
+
"<unk_12>",
|
| 15 |
+
"<unk_13>",
|
| 16 |
+
"<unk_14>",
|
| 17 |
+
"<unk_15>",
|
| 18 |
+
"<unk_16>",
|
| 19 |
+
"<unk_17>",
|
| 20 |
+
"<unk_18>",
|
| 21 |
+
"<unk_19>",
|
| 22 |
+
"<unk_20>",
|
| 23 |
+
"<unk_21>",
|
| 24 |
+
"<unk_22>",
|
| 25 |
+
"<unk_23>",
|
| 26 |
+
"<unk_24>",
|
| 27 |
+
"<unk_25>",
|
| 28 |
+
"<unk_26>",
|
| 29 |
+
"<unk_27>",
|
| 30 |
+
"<unk_28>",
|
| 31 |
+
"<unk_29>",
|
| 32 |
+
"<unk_30>",
|
| 33 |
+
"<unk_31>",
|
| 34 |
+
"<unk_32>",
|
| 35 |
+
"<unk_33>",
|
| 36 |
+
"<unk_34>",
|
| 37 |
+
"<unk_35>",
|
| 38 |
+
"<unk_36>",
|
| 39 |
+
"<unk_37>",
|
| 40 |
+
"<unk_38>",
|
| 41 |
+
"<unk_39>",
|
| 42 |
+
"<unk_40>",
|
| 43 |
+
"<unk_41>",
|
| 44 |
+
"<unk_42>",
|
| 45 |
+
"<unk_43>",
|
| 46 |
+
"<unk_44>",
|
| 47 |
+
"<unk_45>",
|
| 48 |
+
"<unk_46>",
|
| 49 |
+
"<unk_47>",
|
| 50 |
+
"<unk_48>",
|
| 51 |
+
"<unk_49>",
|
| 52 |
+
"<unk_50>",
|
| 53 |
+
"<unk_51>",
|
| 54 |
+
"<unk_52>",
|
| 55 |
+
"<unk_53>",
|
| 56 |
+
"<unk_54>",
|
| 57 |
+
"<unk_55>",
|
| 58 |
+
"<unk_56>",
|
| 59 |
+
"<unk_57>",
|
| 60 |
+
"<unk_58>",
|
| 61 |
+
"<unk_59>",
|
| 62 |
+
"<unk_60>",
|
| 63 |
+
"<unk_61>",
|
| 64 |
+
"<unk_62>",
|
| 65 |
+
"<unk_63>",
|
| 66 |
+
"<unk_64>",
|
| 67 |
+
"<unk_65>",
|
| 68 |
+
"<unk_66>",
|
| 69 |
+
"<unk_67>",
|
| 70 |
+
"<unk_68>",
|
| 71 |
+
"<unk_69>",
|
| 72 |
+
"<unk_70>",
|
| 73 |
+
"<unk_71>",
|
| 74 |
+
"<unk_72>",
|
| 75 |
+
"<unk_73>",
|
| 76 |
+
"<unk_74>",
|
| 77 |
+
"<unk_75>",
|
| 78 |
+
"<unk_76>",
|
| 79 |
+
"<unk_77>",
|
| 80 |
+
"<unk_78>",
|
| 81 |
+
"<unk_79>",
|
| 82 |
+
"<unk_80>",
|
| 83 |
+
"<unk_81>",
|
| 84 |
+
"<unk_82>",
|
| 85 |
+
"<unk_83>",
|
| 86 |
+
"<unk_84>",
|
| 87 |
+
"<unk_85>",
|
| 88 |
+
"<unk_86>",
|
| 89 |
+
"<unk_87>",
|
| 90 |
+
"<unk_88>",
|
| 91 |
+
"<unk_89>",
|
| 92 |
+
"<unk_90>",
|
| 93 |
+
"<unk_91>",
|
| 94 |
+
"<unk_92>",
|
| 95 |
+
"<unk_93>",
|
| 96 |
+
"<unk_94>",
|
| 97 |
+
"<unk_95>",
|
| 98 |
+
"<unk_96>",
|
| 99 |
+
"<unk_97>",
|
| 100 |
+
"<unk_98>",
|
| 101 |
+
"<unk_99>",
|
| 102 |
+
"<unk_100>",
|
| 103 |
+
"<unk_101>",
|
| 104 |
+
"<unk_102>"
|
| 105 |
+
],
|
| 106 |
+
"eos_token": "</s>",
|
| 107 |
+
"full_tokenizer_file": null,
|
| 108 |
+
"mask_token": "<mask_2>",
|
| 109 |
+
"mask_token_sent": "<mask_1>",
|
| 110 |
+
"model_max_length": 1024,
|
| 111 |
+
"offset": 103,
|
| 112 |
+
"pad_token": "<pad>",
|
| 113 |
+
"sp_model_kwargs": {},
|
| 114 |
+
"special_tokens_map_file": null,
|
| 115 |
+
"tokenizer_class": "PegasusTokenizer",
|
| 116 |
+
"unk_token": "<unk>"
|
| 117 |
+
}
|
checkpoint-4000/trainer_state.json
ADDED
|
@@ -0,0 +1,576 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 7.952286282306163,
|
| 5 |
+
"global_step": 4000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.2,
|
| 12 |
+
"learning_rate": 1e-05,
|
| 13 |
+
"loss": 9.1779,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.2,
|
| 18 |
+
"eval_loss": 3.894397258758545,
|
| 19 |
+
"eval_runtime": 33.2081,
|
| 20 |
+
"eval_samples_per_second": 11.774,
|
| 21 |
+
"eval_steps_per_second": 1.686,
|
| 22 |
+
"step": 100
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0.4,
|
| 26 |
+
"learning_rate": 2e-05,
|
| 27 |
+
"loss": 1.9391,
|
| 28 |
+
"step": 200
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0.4,
|
| 32 |
+
"eval_loss": 1.7052955627441406,
|
| 33 |
+
"eval_runtime": 33.2014,
|
| 34 |
+
"eval_samples_per_second": 11.777,
|
| 35 |
+
"eval_steps_per_second": 1.687,
|
| 36 |
+
"step": 200
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.6,
|
| 40 |
+
"learning_rate": 3e-05,
|
| 41 |
+
"loss": 1.4896,
|
| 42 |
+
"step": 300
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.6,
|
| 46 |
+
"eval_loss": 1.597261667251587,
|
| 47 |
+
"eval_runtime": 33.2105,
|
| 48 |
+
"eval_samples_per_second": 11.773,
|
| 49 |
+
"eval_steps_per_second": 1.686,
|
| 50 |
+
"step": 300
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.8,
|
| 54 |
+
"learning_rate": 4e-05,
|
| 55 |
+
"loss": 1.4279,
|
| 56 |
+
"step": 400
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.8,
|
| 60 |
+
"eval_loss": 1.559727668762207,
|
| 61 |
+
"eval_runtime": 33.1972,
|
| 62 |
+
"eval_samples_per_second": 11.778,
|
| 63 |
+
"eval_steps_per_second": 1.687,
|
| 64 |
+
"step": 400
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.99,
|
| 68 |
+
"learning_rate": 5e-05,
|
| 69 |
+
"loss": 1.3676,
|
| 70 |
+
"step": 500
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.99,
|
| 74 |
+
"eval_loss": 1.5377521514892578,
|
| 75 |
+
"eval_runtime": 33.1973,
|
| 76 |
+
"eval_samples_per_second": 11.778,
|
| 77 |
+
"eval_steps_per_second": 1.687,
|
| 78 |
+
"step": 500
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.19,
|
| 82 |
+
"learning_rate": 4.965729952021933e-05,
|
| 83 |
+
"loss": 1.3666,
|
| 84 |
+
"step": 600
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.19,
|
| 88 |
+
"eval_loss": 1.5202970504760742,
|
| 89 |
+
"eval_runtime": 33.2,
|
| 90 |
+
"eval_samples_per_second": 11.777,
|
| 91 |
+
"eval_steps_per_second": 1.687,
|
| 92 |
+
"step": 600
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 1.39,
|
| 96 |
+
"learning_rate": 4.9314599040438655e-05,
|
| 97 |
+
"loss": 1.3432,
|
| 98 |
+
"step": 700
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.39,
|
| 102 |
+
"eval_loss": 1.5091770887374878,
|
| 103 |
+
"eval_runtime": 33.2069,
|
| 104 |
+
"eval_samples_per_second": 11.775,
|
| 105 |
+
"eval_steps_per_second": 1.686,
|
| 106 |
+
"step": 700
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 1.59,
|
| 110 |
+
"learning_rate": 4.8971898560657985e-05,
|
| 111 |
+
"loss": 1.3226,
|
| 112 |
+
"step": 800
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.59,
|
| 116 |
+
"eval_loss": 1.5007234811782837,
|
| 117 |
+
"eval_runtime": 33.2119,
|
| 118 |
+
"eval_samples_per_second": 11.773,
|
| 119 |
+
"eval_steps_per_second": 1.686,
|
| 120 |
+
"step": 800
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 1.79,
|
| 124 |
+
"learning_rate": 4.8629198080877314e-05,
|
| 125 |
+
"loss": 1.3145,
|
| 126 |
+
"step": 900
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.79,
|
| 130 |
+
"eval_loss": 1.4917149543762207,
|
| 131 |
+
"eval_runtime": 33.2096,
|
| 132 |
+
"eval_samples_per_second": 11.774,
|
| 133 |
+
"eval_steps_per_second": 1.686,
|
| 134 |
+
"step": 900
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.99,
|
| 138 |
+
"learning_rate": 4.8286497601096644e-05,
|
| 139 |
+
"loss": 1.2897,
|
| 140 |
+
"step": 1000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.99,
|
| 144 |
+
"eval_loss": 1.487337350845337,
|
| 145 |
+
"eval_runtime": 33.2063,
|
| 146 |
+
"eval_samples_per_second": 11.775,
|
| 147 |
+
"eval_steps_per_second": 1.686,
|
| 148 |
+
"step": 1000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.19,
|
| 152 |
+
"learning_rate": 4.794379712131597e-05,
|
| 153 |
+
"loss": 1.2956,
|
| 154 |
+
"step": 1100
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.19,
|
| 158 |
+
"eval_loss": 1.481351613998413,
|
| 159 |
+
"eval_runtime": 33.2118,
|
| 160 |
+
"eval_samples_per_second": 11.773,
|
| 161 |
+
"eval_steps_per_second": 1.686,
|
| 162 |
+
"step": 1100
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 2.39,
|
| 166 |
+
"learning_rate": 4.76010966415353e-05,
|
| 167 |
+
"loss": 1.2791,
|
| 168 |
+
"step": 1200
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 2.39,
|
| 172 |
+
"eval_loss": 1.4768741130828857,
|
| 173 |
+
"eval_runtime": 33.2069,
|
| 174 |
+
"eval_samples_per_second": 11.775,
|
| 175 |
+
"eval_steps_per_second": 1.686,
|
| 176 |
+
"step": 1200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 2.58,
|
| 180 |
+
"learning_rate": 4.725839616175463e-05,
|
| 181 |
+
"loss": 1.2679,
|
| 182 |
+
"step": 1300
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 2.58,
|
| 186 |
+
"eval_loss": 1.4751813411712646,
|
| 187 |
+
"eval_runtime": 33.1963,
|
| 188 |
+
"eval_samples_per_second": 11.778,
|
| 189 |
+
"eval_steps_per_second": 1.687,
|
| 190 |
+
"step": 1300
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 2.78,
|
| 194 |
+
"learning_rate": 4.6915695681973956e-05,
|
| 195 |
+
"loss": 1.2583,
|
| 196 |
+
"step": 1400
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 2.78,
|
| 200 |
+
"eval_loss": 1.4700709581375122,
|
| 201 |
+
"eval_runtime": 33.2157,
|
| 202 |
+
"eval_samples_per_second": 11.772,
|
| 203 |
+
"eval_steps_per_second": 1.686,
|
| 204 |
+
"step": 1400
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 2.98,
|
| 208 |
+
"learning_rate": 4.6572995202193286e-05,
|
| 209 |
+
"loss": 1.2813,
|
| 210 |
+
"step": 1500
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 2.98,
|
| 214 |
+
"eval_loss": 1.4654401540756226,
|
| 215 |
+
"eval_runtime": 33.205,
|
| 216 |
+
"eval_samples_per_second": 11.775,
|
| 217 |
+
"eval_steps_per_second": 1.686,
|
| 218 |
+
"step": 1500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.18,
|
| 222 |
+
"learning_rate": 4.6230294722412615e-05,
|
| 223 |
+
"loss": 1.25,
|
| 224 |
+
"step": 1600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 3.18,
|
| 228 |
+
"eval_loss": 1.4636775255203247,
|
| 229 |
+
"eval_runtime": 33.207,
|
| 230 |
+
"eval_samples_per_second": 11.775,
|
| 231 |
+
"eval_steps_per_second": 1.686,
|
| 232 |
+
"step": 1600
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.38,
|
| 236 |
+
"learning_rate": 4.5887594242631945e-05,
|
| 237 |
+
"loss": 1.2393,
|
| 238 |
+
"step": 1700
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 3.38,
|
| 242 |
+
"eval_loss": 1.4608376026153564,
|
| 243 |
+
"eval_runtime": 33.2094,
|
| 244 |
+
"eval_samples_per_second": 11.774,
|
| 245 |
+
"eval_steps_per_second": 1.686,
|
| 246 |
+
"step": 1700
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.58,
|
| 250 |
+
"learning_rate": 4.554489376285127e-05,
|
| 251 |
+
"loss": 1.2599,
|
| 252 |
+
"step": 1800
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 3.58,
|
| 256 |
+
"eval_loss": 1.4582923650741577,
|
| 257 |
+
"eval_runtime": 33.2003,
|
| 258 |
+
"eval_samples_per_second": 11.777,
|
| 259 |
+
"eval_steps_per_second": 1.687,
|
| 260 |
+
"step": 1800
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.78,
|
| 264 |
+
"learning_rate": 4.52021932830706e-05,
|
| 265 |
+
"loss": 1.2336,
|
| 266 |
+
"step": 1900
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 3.78,
|
| 270 |
+
"eval_loss": 1.455262303352356,
|
| 271 |
+
"eval_runtime": 33.2047,
|
| 272 |
+
"eval_samples_per_second": 11.775,
|
| 273 |
+
"eval_steps_per_second": 1.687,
|
| 274 |
+
"step": 1900
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.98,
|
| 278 |
+
"learning_rate": 4.485949280328993e-05,
|
| 279 |
+
"loss": 1.2374,
|
| 280 |
+
"step": 2000
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 3.98,
|
| 284 |
+
"eval_loss": 1.4536148309707642,
|
| 285 |
+
"eval_runtime": 33.2036,
|
| 286 |
+
"eval_samples_per_second": 11.776,
|
| 287 |
+
"eval_steps_per_second": 1.687,
|
| 288 |
+
"step": 2000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 4.17,
|
| 292 |
+
"learning_rate": 4.451679232350926e-05,
|
| 293 |
+
"loss": 1.2171,
|
| 294 |
+
"step": 2100
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 4.17,
|
| 298 |
+
"eval_loss": 1.4514210224151611,
|
| 299 |
+
"eval_runtime": 33.2199,
|
| 300 |
+
"eval_samples_per_second": 11.77,
|
| 301 |
+
"eval_steps_per_second": 1.686,
|
| 302 |
+
"step": 2100
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 4.37,
|
| 306 |
+
"learning_rate": 4.417409184372858e-05,
|
| 307 |
+
"loss": 1.2154,
|
| 308 |
+
"step": 2200
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 4.37,
|
| 312 |
+
"eval_loss": 1.450243592262268,
|
| 313 |
+
"eval_runtime": 33.2384,
|
| 314 |
+
"eval_samples_per_second": 11.763,
|
| 315 |
+
"eval_steps_per_second": 1.685,
|
| 316 |
+
"step": 2200
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 4.57,
|
| 320 |
+
"learning_rate": 4.383139136394791e-05,
|
| 321 |
+
"loss": 1.2247,
|
| 322 |
+
"step": 2300
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 4.57,
|
| 326 |
+
"eval_loss": 1.4490052461624146,
|
| 327 |
+
"eval_runtime": 33.2063,
|
| 328 |
+
"eval_samples_per_second": 11.775,
|
| 329 |
+
"eval_steps_per_second": 1.686,
|
| 330 |
+
"step": 2300
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 4.77,
|
| 334 |
+
"learning_rate": 4.348869088416724e-05,
|
| 335 |
+
"loss": 1.2234,
|
| 336 |
+
"step": 2400
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 4.77,
|
| 340 |
+
"eval_loss": 1.4472484588623047,
|
| 341 |
+
"eval_runtime": 33.2181,
|
| 342 |
+
"eval_samples_per_second": 11.771,
|
| 343 |
+
"eval_steps_per_second": 1.686,
|
| 344 |
+
"step": 2400
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 4.97,
|
| 348 |
+
"learning_rate": 4.314599040438657e-05,
|
| 349 |
+
"loss": 1.2152,
|
| 350 |
+
"step": 2500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 4.97,
|
| 354 |
+
"eval_loss": 1.4455540180206299,
|
| 355 |
+
"eval_runtime": 33.205,
|
| 356 |
+
"eval_samples_per_second": 11.775,
|
| 357 |
+
"eval_steps_per_second": 1.686,
|
| 358 |
+
"step": 2500
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 5.17,
|
| 362 |
+
"learning_rate": 4.280328992460589e-05,
|
| 363 |
+
"loss": 1.1883,
|
| 364 |
+
"step": 2600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 5.17,
|
| 368 |
+
"eval_loss": 1.4456433057785034,
|
| 369 |
+
"eval_runtime": 33.2111,
|
| 370 |
+
"eval_samples_per_second": 11.773,
|
| 371 |
+
"eval_steps_per_second": 1.686,
|
| 372 |
+
"step": 2600
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 5.37,
|
| 376 |
+
"learning_rate": 4.246058944482523e-05,
|
| 377 |
+
"loss": 1.1947,
|
| 378 |
+
"step": 2700
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 5.37,
|
| 382 |
+
"eval_loss": 1.4448508024215698,
|
| 383 |
+
"eval_runtime": 33.1978,
|
| 384 |
+
"eval_samples_per_second": 11.778,
|
| 385 |
+
"eval_steps_per_second": 1.687,
|
| 386 |
+
"step": 2700
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 5.57,
|
| 390 |
+
"learning_rate": 4.211788896504456e-05,
|
| 391 |
+
"loss": 1.2127,
|
| 392 |
+
"step": 2800
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 5.57,
|
| 396 |
+
"eval_loss": 1.4433377981185913,
|
| 397 |
+
"eval_runtime": 33.2104,
|
| 398 |
+
"eval_samples_per_second": 11.773,
|
| 399 |
+
"eval_steps_per_second": 1.686,
|
| 400 |
+
"step": 2800
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 5.77,
|
| 404 |
+
"learning_rate": 4.177518848526388e-05,
|
| 405 |
+
"loss": 1.1959,
|
| 406 |
+
"step": 2900
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 5.77,
|
| 410 |
+
"eval_loss": 1.4412949085235596,
|
| 411 |
+
"eval_runtime": 33.2197,
|
| 412 |
+
"eval_samples_per_second": 11.77,
|
| 413 |
+
"eval_steps_per_second": 1.686,
|
| 414 |
+
"step": 2900
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 5.96,
|
| 418 |
+
"learning_rate": 4.143248800548321e-05,
|
| 419 |
+
"loss": 1.1913,
|
| 420 |
+
"step": 3000
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 5.96,
|
| 424 |
+
"eval_loss": 1.4401354789733887,
|
| 425 |
+
"eval_runtime": 33.21,
|
| 426 |
+
"eval_samples_per_second": 11.774,
|
| 427 |
+
"eval_steps_per_second": 1.686,
|
| 428 |
+
"step": 3000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 6.16,
|
| 432 |
+
"learning_rate": 4.108978752570254e-05,
|
| 433 |
+
"loss": 1.1966,
|
| 434 |
+
"step": 3100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 6.16,
|
| 438 |
+
"eval_loss": 1.4406956434249878,
|
| 439 |
+
"eval_runtime": 33.2007,
|
| 440 |
+
"eval_samples_per_second": 11.777,
|
| 441 |
+
"eval_steps_per_second": 1.687,
|
| 442 |
+
"step": 3100
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 6.36,
|
| 446 |
+
"learning_rate": 4.0747087045921863e-05,
|
| 447 |
+
"loss": 1.1798,
|
| 448 |
+
"step": 3200
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 6.36,
|
| 452 |
+
"eval_loss": 1.4382696151733398,
|
| 453 |
+
"eval_runtime": 33.2065,
|
| 454 |
+
"eval_samples_per_second": 11.775,
|
| 455 |
+
"eval_steps_per_second": 1.686,
|
| 456 |
+
"step": 3200
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 6.56,
|
| 460 |
+
"learning_rate": 4.040438656614119e-05,
|
| 461 |
+
"loss": 1.1736,
|
| 462 |
+
"step": 3300
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 6.56,
|
| 466 |
+
"eval_loss": 1.4385173320770264,
|
| 467 |
+
"eval_runtime": 33.2237,
|
| 468 |
+
"eval_samples_per_second": 11.769,
|
| 469 |
+
"eval_steps_per_second": 1.686,
|
| 470 |
+
"step": 3300
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 6.76,
|
| 474 |
+
"learning_rate": 4.006168608636052e-05,
|
| 475 |
+
"loss": 1.1661,
|
| 476 |
+
"step": 3400
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 6.76,
|
| 480 |
+
"eval_loss": 1.4370990991592407,
|
| 481 |
+
"eval_runtime": 33.2081,
|
| 482 |
+
"eval_samples_per_second": 11.774,
|
| 483 |
+
"eval_steps_per_second": 1.686,
|
| 484 |
+
"step": 3400
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 6.96,
|
| 488 |
+
"learning_rate": 3.971898560657985e-05,
|
| 489 |
+
"loss": 1.1802,
|
| 490 |
+
"step": 3500
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 6.96,
|
| 494 |
+
"eval_loss": 1.4352325201034546,
|
| 495 |
+
"eval_runtime": 33.2067,
|
| 496 |
+
"eval_samples_per_second": 11.775,
|
| 497 |
+
"eval_steps_per_second": 1.686,
|
| 498 |
+
"step": 3500
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 7.16,
|
| 502 |
+
"learning_rate": 3.9376285126799175e-05,
|
| 503 |
+
"loss": 1.1649,
|
| 504 |
+
"step": 3600
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 7.16,
|
| 508 |
+
"eval_loss": 1.436591386795044,
|
| 509 |
+
"eval_runtime": 33.2379,
|
| 510 |
+
"eval_samples_per_second": 11.764,
|
| 511 |
+
"eval_steps_per_second": 1.685,
|
| 512 |
+
"step": 3600
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 7.36,
|
| 516 |
+
"learning_rate": 3.9033584647018505e-05,
|
| 517 |
+
"loss": 1.1654,
|
| 518 |
+
"step": 3700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 7.36,
|
| 522 |
+
"eval_loss": 1.4353548288345337,
|
| 523 |
+
"eval_runtime": 33.2162,
|
| 524 |
+
"eval_samples_per_second": 11.771,
|
| 525 |
+
"eval_steps_per_second": 1.686,
|
| 526 |
+
"step": 3700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 7.55,
|
| 530 |
+
"learning_rate": 3.8690884167237835e-05,
|
| 531 |
+
"loss": 1.1604,
|
| 532 |
+
"step": 3800
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 7.55,
|
| 536 |
+
"eval_loss": 1.4346917867660522,
|
| 537 |
+
"eval_runtime": 33.2322,
|
| 538 |
+
"eval_samples_per_second": 11.766,
|
| 539 |
+
"eval_steps_per_second": 1.685,
|
| 540 |
+
"step": 3800
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 7.75,
|
| 544 |
+
"learning_rate": 3.8348183687457165e-05,
|
| 545 |
+
"loss": 1.1555,
|
| 546 |
+
"step": 3900
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 7.75,
|
| 550 |
+
"eval_loss": 1.4331263303756714,
|
| 551 |
+
"eval_runtime": 33.2124,
|
| 552 |
+
"eval_samples_per_second": 11.773,
|
| 553 |
+
"eval_steps_per_second": 1.686,
|
| 554 |
+
"step": 3900
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 7.95,
|
| 558 |
+
"learning_rate": 3.800548320767649e-05,
|
| 559 |
+
"loss": 1.1617,
|
| 560 |
+
"step": 4000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 7.95,
|
| 564 |
+
"eval_loss": 1.4325422048568726,
|
| 565 |
+
"eval_runtime": 33.2006,
|
| 566 |
+
"eval_samples_per_second": 11.777,
|
| 567 |
+
"eval_steps_per_second": 1.687,
|
| 568 |
+
"step": 4000
|
| 569 |
+
}
|
| 570 |
+
],
|
| 571 |
+
"max_steps": 15090,
|
| 572 |
+
"num_train_epochs": 30,
|
| 573 |
+
"total_flos": 8.075550264813158e+16,
|
| 574 |
+
"trial_name": null,
|
| 575 |
+
"trial_params": null
|
| 576 |
+
}
|
checkpoint-4000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73a3e18b94ee595e0628e947ea0948c27be0e6b9473c64a1eaa48716158a7d3e
|
| 3 |
+
size 3515
|