Upload folder using huggingface_hub
Browse files- added_tokens.json +2 -0
- config.json +3 -3
- generation_config.json +8 -0
- pytorch_model.bin +2 -2
- tokenizer.json +14 -28
- tokenizer_config.json +1 -2
- trainer_state.json +403 -0
- training_args.bin +2 -2
added_tokens.json
CHANGED
|
@@ -1,4 +1,6 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
| 2 |
"<mask>": 60000,
|
| 3 |
"<text>": 60001
|
| 4 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"<lang:ca>": 60002,
|
| 3 |
+
"<lang:es>": 60003,
|
| 4 |
"<mask>": 60000,
|
| 5 |
"<text>": 60001
|
| 6 |
}
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "./
|
| 3 |
"activation_dropout": 0.0,
|
| 4 |
"activation_function": "gelu",
|
| 5 |
"architectures": [
|
|
@@ -42,8 +42,8 @@
|
|
| 42 |
"pad_token_id": 0,
|
| 43 |
"scale_embedding": false,
|
| 44 |
"torch_dtype": "float32",
|
| 45 |
-
"transformers_version": "4.
|
| 46 |
"unk_token_id": 3,
|
| 47 |
"use_cache": true,
|
| 48 |
-
"vocab_size":
|
| 49 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "./pre-nasca",
|
| 3 |
"activation_dropout": 0.0,
|
| 4 |
"activation_function": "gelu",
|
| 5 |
"architectures": [
|
|
|
|
| 42 |
"pad_token_id": 0,
|
| 43 |
"scale_embedding": false,
|
| 44 |
"torch_dtype": "float32",
|
| 45 |
+
"transformers_version": "4.27.4",
|
| 46 |
"unk_token_id": 3,
|
| 47 |
"use_cache": true,
|
| 48 |
+
"vocab_size": 60004
|
| 49 |
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 1,
|
| 3 |
+
"decoder_start_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"forced_eos_token_id": 2,
|
| 6 |
+
"pad_token_id": 0,
|
| 7 |
+
"transformers_version": "4.27.4"
|
| 8 |
+
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9df303c733e557cd238216950c2f323fed37a202206277c807099d6026e1b433
|
| 3 |
+
size 1661277005
|
tokenizer.json
CHANGED
|
@@ -1,21 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
| 4 |
-
|
| 5 |
-
"max_length": 512,
|
| 6 |
-
"strategy": "LongestFirst",
|
| 7 |
-
"stride": 0
|
| 8 |
-
},
|
| 9 |
-
"padding": {
|
| 10 |
-
"strategy": {
|
| 11 |
-
"Fixed": 512
|
| 12 |
-
},
|
| 13 |
-
"direction": "Right",
|
| 14 |
-
"pad_to_multiple_of": 8,
|
| 15 |
-
"pad_id": 0,
|
| 16 |
-
"pad_type_id": 0,
|
| 17 |
-
"pad_token": "<pad>"
|
| 18 |
-
},
|
| 19 |
"added_tokens": [
|
| 20 |
{
|
| 21 |
"id": 0,
|
|
@@ -54,17 +40,17 @@
|
|
| 54 |
"special": true
|
| 55 |
},
|
| 56 |
{
|
| 57 |
-
"id":
|
| 58 |
-
"content": "
|
| 59 |
"single_word": false,
|
| 60 |
-
"lstrip":
|
| 61 |
"rstrip": false,
|
| 62 |
-
"normalized":
|
| 63 |
"special": true
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"id":
|
| 67 |
-
"content": "
|
| 68 |
"single_word": false,
|
| 69 |
"lstrip": false,
|
| 70 |
"rstrip": false,
|
|
@@ -72,17 +58,17 @@
|
|
| 72 |
"special": true
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"id":
|
| 76 |
-
"content": "<
|
| 77 |
"single_word": false,
|
| 78 |
-
"lstrip":
|
| 79 |
"rstrip": false,
|
| 80 |
-
"normalized":
|
| 81 |
"special": true
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"id":
|
| 85 |
-
"content": "<
|
| 86 |
"single_word": false,
|
| 87 |
"lstrip": false,
|
| 88 |
"rstrip": false,
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
| 7 |
"id": 0,
|
|
|
|
| 40 |
"special": true
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"id": 60000,
|
| 44 |
+
"content": "<mask>",
|
| 45 |
"single_word": false,
|
| 46 |
+
"lstrip": true,
|
| 47 |
"rstrip": false,
|
| 48 |
+
"normalized": true,
|
| 49 |
"special": true
|
| 50 |
},
|
| 51 |
{
|
| 52 |
+
"id": 60001,
|
| 53 |
+
"content": "<text>",
|
| 54 |
"single_word": false,
|
| 55 |
"lstrip": false,
|
| 56 |
"rstrip": false,
|
|
|
|
| 58 |
"special": true
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"id": 60002,
|
| 62 |
+
"content": "<lang:ca>",
|
| 63 |
"single_word": false,
|
| 64 |
+
"lstrip": false,
|
| 65 |
"rstrip": false,
|
| 66 |
+
"normalized": false,
|
| 67 |
"special": true
|
| 68 |
},
|
| 69 |
{
|
| 70 |
+
"id": 60003,
|
| 71 |
+
"content": "<lang:es>",
|
| 72 |
"single_word": false,
|
| 73 |
"lstrip": false,
|
| 74 |
"rstrip": false,
|
tokenizer_config.json
CHANGED
|
@@ -33,8 +33,7 @@
|
|
| 33 |
"rstrip": false,
|
| 34 |
"single_word": false
|
| 35 |
},
|
| 36 |
-
"model_max_length":
|
| 37 |
-
"name_or_path": "./NAS-bilingue",
|
| 38 |
"pad_token": {
|
| 39 |
"__type": "AddedToken",
|
| 40 |
"content": "<pad>",
|
|
|
|
| 33 |
"rstrip": false,
|
| 34 |
"single_word": false
|
| 35 |
},
|
| 36 |
+
"model_max_length": 1000000000000000019884624838656,
|
|
|
|
| 37 |
"pad_token": {
|
| 38 |
"__type": "AddedToken",
|
| 39 |
"content": "<pad>",
|
trainer_state.json
ADDED
|
@@ -0,0 +1,403 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.03879310344827587,
|
| 3 |
+
"best_model_checkpoint": "NAS-Bi-output/20231214-103634/checkpoint-96",
|
| 4 |
+
"epoch": 3.84,
|
| 5 |
+
"global_step": 120,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.06,
|
| 12 |
+
"learning_rate": 1.3774193548387096e-06,
|
| 13 |
+
"loss": 4.5801,
|
| 14 |
+
"step": 2
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.13,
|
| 18 |
+
"learning_rate": 1.3548387096774193e-06,
|
| 19 |
+
"loss": 4.5691,
|
| 20 |
+
"step": 4
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.19,
|
| 24 |
+
"learning_rate": 1.332258064516129e-06,
|
| 25 |
+
"loss": 4.5408,
|
| 26 |
+
"step": 6
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.26,
|
| 30 |
+
"learning_rate": 1.3096774193548385e-06,
|
| 31 |
+
"loss": 4.3296,
|
| 32 |
+
"step": 8
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.32,
|
| 36 |
+
"learning_rate": 1.2870967741935482e-06,
|
| 37 |
+
"loss": 4.0271,
|
| 38 |
+
"step": 10
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.38,
|
| 42 |
+
"learning_rate": 1.264516129032258e-06,
|
| 43 |
+
"loss": 4.3243,
|
| 44 |
+
"step": 12
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.45,
|
| 48 |
+
"learning_rate": 1.2419354838709678e-06,
|
| 49 |
+
"loss": 4.1559,
|
| 50 |
+
"step": 14
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.51,
|
| 54 |
+
"learning_rate": 1.2193548387096774e-06,
|
| 55 |
+
"loss": 4.0442,
|
| 56 |
+
"step": 16
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.58,
|
| 60 |
+
"learning_rate": 1.1967741935483871e-06,
|
| 61 |
+
"loss": 4.1364,
|
| 62 |
+
"step": 18
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.64,
|
| 66 |
+
"learning_rate": 1.1741935483870968e-06,
|
| 67 |
+
"loss": 3.9896,
|
| 68 |
+
"step": 20
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.7,
|
| 72 |
+
"learning_rate": 1.1516129032258065e-06,
|
| 73 |
+
"loss": 3.9312,
|
| 74 |
+
"step": 22
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.77,
|
| 78 |
+
"learning_rate": 1.129032258064516e-06,
|
| 79 |
+
"loss": 4.3079,
|
| 80 |
+
"step": 24
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.83,
|
| 84 |
+
"learning_rate": 1.1064516129032257e-06,
|
| 85 |
+
"loss": 3.873,
|
| 86 |
+
"step": 26
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.9,
|
| 90 |
+
"learning_rate": 1.0838709677419354e-06,
|
| 91 |
+
"loss": 4.1228,
|
| 92 |
+
"step": 28
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.96,
|
| 96 |
+
"learning_rate": 1.061290322580645e-06,
|
| 97 |
+
"loss": 3.8045,
|
| 98 |
+
"step": 30
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.02,
|
| 102 |
+
"learning_rate": 1.0387096774193547e-06,
|
| 103 |
+
"loss": 3.8668,
|
| 104 |
+
"step": 32
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.02,
|
| 108 |
+
"eval_loss": 3.629098415374756,
|
| 109 |
+
"eval_rougeLsum": 0.04047338477718224,
|
| 110 |
+
"eval_runtime": 1.6835,
|
| 111 |
+
"eval_samples_per_second": 5.94,
|
| 112 |
+
"eval_steps_per_second": 1.188,
|
| 113 |
+
"step": 32
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 1.09,
|
| 117 |
+
"learning_rate": 1.0161290322580644e-06,
|
| 118 |
+
"loss": 3.86,
|
| 119 |
+
"step": 34
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.15,
|
| 123 |
+
"learning_rate": 9.935483870967741e-07,
|
| 124 |
+
"loss": 3.5875,
|
| 125 |
+
"step": 36
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 1.22,
|
| 129 |
+
"learning_rate": 9.709677419354838e-07,
|
| 130 |
+
"loss": 3.8769,
|
| 131 |
+
"step": 38
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.28,
|
| 135 |
+
"learning_rate": 9.483870967741934e-07,
|
| 136 |
+
"loss": 3.58,
|
| 137 |
+
"step": 40
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.34,
|
| 141 |
+
"learning_rate": 9.258064516129032e-07,
|
| 142 |
+
"loss": 3.6624,
|
| 143 |
+
"step": 42
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.41,
|
| 147 |
+
"learning_rate": 9.032258064516129e-07,
|
| 148 |
+
"loss": 3.7948,
|
| 149 |
+
"step": 44
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.47,
|
| 153 |
+
"learning_rate": 8.806451612903226e-07,
|
| 154 |
+
"loss": 3.6461,
|
| 155 |
+
"step": 46
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 1.54,
|
| 159 |
+
"learning_rate": 8.580645161290323e-07,
|
| 160 |
+
"loss": 3.8532,
|
| 161 |
+
"step": 48
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.6,
|
| 165 |
+
"learning_rate": 8.354838709677419e-07,
|
| 166 |
+
"loss": 3.8078,
|
| 167 |
+
"step": 50
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 1.66,
|
| 171 |
+
"learning_rate": 8.129032258064516e-07,
|
| 172 |
+
"loss": 3.6687,
|
| 173 |
+
"step": 52
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 1.73,
|
| 177 |
+
"learning_rate": 7.903225806451613e-07,
|
| 178 |
+
"loss": 3.6238,
|
| 179 |
+
"step": 54
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.79,
|
| 183 |
+
"learning_rate": 7.677419354838709e-07,
|
| 184 |
+
"loss": 3.3701,
|
| 185 |
+
"step": 56
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.86,
|
| 189 |
+
"learning_rate": 7.451612903225806e-07,
|
| 190 |
+
"loss": 3.6845,
|
| 191 |
+
"step": 58
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.92,
|
| 195 |
+
"learning_rate": 7.225806451612903e-07,
|
| 196 |
+
"loss": 3.5206,
|
| 197 |
+
"step": 60
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 1.98,
|
| 201 |
+
"learning_rate": 7e-07,
|
| 202 |
+
"loss": 3.6524,
|
| 203 |
+
"step": 62
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 2.05,
|
| 207 |
+
"learning_rate": 6.774193548387097e-07,
|
| 208 |
+
"loss": 3.5519,
|
| 209 |
+
"step": 64
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.05,
|
| 213 |
+
"eval_loss": 3.363889694213867,
|
| 214 |
+
"eval_rougeLsum": 0.03520105471324984,
|
| 215 |
+
"eval_runtime": 0.7217,
|
| 216 |
+
"eval_samples_per_second": 13.857,
|
| 217 |
+
"eval_steps_per_second": 2.771,
|
| 218 |
+
"step": 64
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.11,
|
| 222 |
+
"learning_rate": 6.548387096774192e-07,
|
| 223 |
+
"loss": 3.534,
|
| 224 |
+
"step": 66
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.18,
|
| 228 |
+
"learning_rate": 6.32258064516129e-07,
|
| 229 |
+
"loss": 3.4723,
|
| 230 |
+
"step": 68
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.24,
|
| 234 |
+
"learning_rate": 6.096774193548387e-07,
|
| 235 |
+
"loss": 3.4788,
|
| 236 |
+
"step": 70
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 2.3,
|
| 240 |
+
"learning_rate": 5.870967741935484e-07,
|
| 241 |
+
"loss": 3.5552,
|
| 242 |
+
"step": 72
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.37,
|
| 246 |
+
"learning_rate": 5.64516129032258e-07,
|
| 247 |
+
"loss": 3.6136,
|
| 248 |
+
"step": 74
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.43,
|
| 252 |
+
"learning_rate": 5.419354838709677e-07,
|
| 253 |
+
"loss": 3.4261,
|
| 254 |
+
"step": 76
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.5,
|
| 258 |
+
"learning_rate": 5.193548387096774e-07,
|
| 259 |
+
"loss": 3.2605,
|
| 260 |
+
"step": 78
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.56,
|
| 264 |
+
"learning_rate": 4.967741935483871e-07,
|
| 265 |
+
"loss": 3.4041,
|
| 266 |
+
"step": 80
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.62,
|
| 270 |
+
"learning_rate": 4.741935483870967e-07,
|
| 271 |
+
"loss": 3.4199,
|
| 272 |
+
"step": 82
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.69,
|
| 276 |
+
"learning_rate": 4.5161290322580644e-07,
|
| 277 |
+
"loss": 3.5787,
|
| 278 |
+
"step": 84
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.75,
|
| 282 |
+
"learning_rate": 4.2903225806451613e-07,
|
| 283 |
+
"loss": 3.2749,
|
| 284 |
+
"step": 86
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.82,
|
| 288 |
+
"learning_rate": 4.064516129032258e-07,
|
| 289 |
+
"loss": 3.2766,
|
| 290 |
+
"step": 88
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.88,
|
| 294 |
+
"learning_rate": 3.8387096774193545e-07,
|
| 295 |
+
"loss": 3.5915,
|
| 296 |
+
"step": 90
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.94,
|
| 300 |
+
"learning_rate": 3.6129032258064514e-07,
|
| 301 |
+
"loss": 3.4801,
|
| 302 |
+
"step": 92
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.01,
|
| 306 |
+
"learning_rate": 3.3870967741935483e-07,
|
| 307 |
+
"loss": 3.4917,
|
| 308 |
+
"step": 94
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 3.07,
|
| 312 |
+
"learning_rate": 3.161290322580645e-07,
|
| 313 |
+
"loss": 3.4979,
|
| 314 |
+
"step": 96
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 3.07,
|
| 318 |
+
"eval_loss": 3.2409491539001465,
|
| 319 |
+
"eval_rougeLsum": 0.03879310344827587,
|
| 320 |
+
"eval_runtime": 0.7453,
|
| 321 |
+
"eval_samples_per_second": 13.417,
|
| 322 |
+
"eval_steps_per_second": 2.683,
|
| 323 |
+
"step": 96
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.14,
|
| 327 |
+
"learning_rate": 2.935483870967742e-07,
|
| 328 |
+
"loss": 3.4648,
|
| 329 |
+
"step": 98
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 3.2,
|
| 333 |
+
"learning_rate": 2.7096774193548384e-07,
|
| 334 |
+
"loss": 3.4253,
|
| 335 |
+
"step": 100
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 3.26,
|
| 339 |
+
"learning_rate": 2.4838709677419353e-07,
|
| 340 |
+
"loss": 3.3574,
|
| 341 |
+
"step": 102
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 3.33,
|
| 345 |
+
"learning_rate": 2.2580645161290322e-07,
|
| 346 |
+
"loss": 3.4262,
|
| 347 |
+
"step": 104
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 3.39,
|
| 351 |
+
"learning_rate": 2.032258064516129e-07,
|
| 352 |
+
"loss": 3.2748,
|
| 353 |
+
"step": 106
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 3.46,
|
| 357 |
+
"learning_rate": 1.8064516129032257e-07,
|
| 358 |
+
"loss": 3.4467,
|
| 359 |
+
"step": 108
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.52,
|
| 363 |
+
"learning_rate": 1.5806451612903226e-07,
|
| 364 |
+
"loss": 3.3972,
|
| 365 |
+
"step": 110
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.58,
|
| 369 |
+
"learning_rate": 1.3548387096774192e-07,
|
| 370 |
+
"loss": 3.0868,
|
| 371 |
+
"step": 112
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 3.65,
|
| 375 |
+
"learning_rate": 1.1290322580645161e-07,
|
| 376 |
+
"loss": 3.4098,
|
| 377 |
+
"step": 114
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 3.71,
|
| 381 |
+
"learning_rate": 9.032258064516128e-08,
|
| 382 |
+
"loss": 3.3889,
|
| 383 |
+
"step": 116
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 3.78,
|
| 387 |
+
"learning_rate": 6.774193548387096e-08,
|
| 388 |
+
"loss": 3.4246,
|
| 389 |
+
"step": 118
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 3.84,
|
| 393 |
+
"learning_rate": 4.516129032258064e-08,
|
| 394 |
+
"loss": 3.1682,
|
| 395 |
+
"step": 120
|
| 396 |
+
}
|
| 397 |
+
],
|
| 398 |
+
"max_steps": 124,
|
| 399 |
+
"num_train_epochs": 4,
|
| 400 |
+
"total_flos": 4160840836055040.0,
|
| 401 |
+
"trial_name": null,
|
| 402 |
+
"trial_params": null
|
| 403 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d5ba13682c76280a9fb401c44134bf15d7c6a0453d3e2424624c79587d7b9a3
|
| 3 |
+
size 3707
|