silas.rudolf@nexoya.com commited on
Commit ·
1003893
1
Parent(s): cd5c12a
QMSUM trained transformer
Browse files- config.json +70 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +15 -0
- tokenizer.json +0 -0
- tokenizer_config.json +65 -0
- trainer_state.json +2173 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "philschmid/bart-large-cnn-samsum",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"activation_dropout": 0.0,
|
| 5 |
+
"activation_function": "gelu",
|
| 6 |
+
"add_final_layer_norm": false,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"BartForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"classif_dropout": 0.0,
|
| 13 |
+
"classifier_dropout": 0.0,
|
| 14 |
+
"d_model": 1024,
|
| 15 |
+
"decoder_attention_heads": 16,
|
| 16 |
+
"decoder_ffn_dim": 4096,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 12,
|
| 19 |
+
"decoder_start_token_id": 2,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"early_stopping": true,
|
| 22 |
+
"encoder_attention_heads": 16,
|
| 23 |
+
"encoder_ffn_dim": 4096,
|
| 24 |
+
"encoder_layerdrop": 0.0,
|
| 25 |
+
"encoder_layers": 12,
|
| 26 |
+
"eos_token_id": 2,
|
| 27 |
+
"force_bos_token_to_be_generated": true,
|
| 28 |
+
"forced_bos_token_id": 0,
|
| 29 |
+
"forced_eos_token_id": 2,
|
| 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": 2.0,
|
| 44 |
+
"max_length": 142,
|
| 45 |
+
"max_position_embeddings": 1024,
|
| 46 |
+
"min_length": 56,
|
| 47 |
+
"model_type": "bart",
|
| 48 |
+
"no_repeat_ngram_size": 3,
|
| 49 |
+
"normalize_before": false,
|
| 50 |
+
"num_beams": 4,
|
| 51 |
+
"num_hidden_layers": 12,
|
| 52 |
+
"output_past": true,
|
| 53 |
+
"pad_token_id": 1,
|
| 54 |
+
"prefix": " ",
|
| 55 |
+
"scale_embedding": false,
|
| 56 |
+
"task_specific_params": {
|
| 57 |
+
"summarization": {
|
| 58 |
+
"early_stopping": true,
|
| 59 |
+
"length_penalty": 2.0,
|
| 60 |
+
"max_length": 142,
|
| 61 |
+
"min_length": 56,
|
| 62 |
+
"no_repeat_ngram_size": 3,
|
| 63 |
+
"num_beams": 4
|
| 64 |
+
}
|
| 65 |
+
},
|
| 66 |
+
"torch_dtype": "float32",
|
| 67 |
+
"transformers_version": "4.20.0",
|
| 68 |
+
"use_cache": true,
|
| 69 |
+
"vocab_size": 50264
|
| 70 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b43afdea394686fd7bcdb6a425f5158ce4f77c82478e0234221befd29dcb601
|
| 3 |
+
size 3250636379
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95588277b705660f1665e181a440a98df2adf50a0430f870e333759f080fce47
|
| 3 |
+
size 1625539886
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dee740e86f1e23c34edf83cc48784e344ed26cce99d30d20265c70ccad0bb1ed
|
| 3 |
+
size 14649
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:239dfb5595953bbcce43228606d9e6fad8f23fd9bd31606b6dbcd52bb7c2e2c5
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db15c89b2e4c0bf35382b83a45f3cfb517ccfdab61e5e376924645140c3a962a
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": {
|
| 6 |
+
"content": "<mask>",
|
| 7 |
+
"lstrip": true,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false
|
| 11 |
+
},
|
| 12 |
+
"pad_token": "<pad>",
|
| 13 |
+
"sep_token": "</s>",
|
| 14 |
+
"unk_token": "<unk>"
|
| 15 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"bos_token": {
|
| 4 |
+
"__type": "AddedToken",
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false
|
| 10 |
+
},
|
| 11 |
+
"cls_token": {
|
| 12 |
+
"__type": "AddedToken",
|
| 13 |
+
"content": "<s>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false
|
| 18 |
+
},
|
| 19 |
+
"eos_token": {
|
| 20 |
+
"__type": "AddedToken",
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false
|
| 26 |
+
},
|
| 27 |
+
"errors": "replace",
|
| 28 |
+
"mask_token": {
|
| 29 |
+
"__type": "AddedToken",
|
| 30 |
+
"content": "<mask>",
|
| 31 |
+
"lstrip": true,
|
| 32 |
+
"normalized": true,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false
|
| 35 |
+
},
|
| 36 |
+
"model_max_length": 1024,
|
| 37 |
+
"name_or_path": "philschmid/bart-large-cnn-samsum",
|
| 38 |
+
"pad_token": {
|
| 39 |
+
"__type": "AddedToken",
|
| 40 |
+
"content": "<pad>",
|
| 41 |
+
"lstrip": false,
|
| 42 |
+
"normalized": true,
|
| 43 |
+
"rstrip": false,
|
| 44 |
+
"single_word": false
|
| 45 |
+
},
|
| 46 |
+
"sep_token": {
|
| 47 |
+
"__type": "AddedToken",
|
| 48 |
+
"content": "</s>",
|
| 49 |
+
"lstrip": false,
|
| 50 |
+
"normalized": true,
|
| 51 |
+
"rstrip": false,
|
| 52 |
+
"single_word": false
|
| 53 |
+
},
|
| 54 |
+
"special_tokens_map_file": null,
|
| 55 |
+
"tokenizer_class": "BartTokenizer",
|
| 56 |
+
"trim_offsets": true,
|
| 57 |
+
"unk_token": {
|
| 58 |
+
"__type": "AddedToken",
|
| 59 |
+
"content": "<unk>",
|
| 60 |
+
"lstrip": false,
|
| 61 |
+
"normalized": true,
|
| 62 |
+
"rstrip": false,
|
| 63 |
+
"single_word": false
|
| 64 |
+
}
|
| 65 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2173 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.789693392831438,
|
| 5 |
+
"global_step": 34000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 1.994471638353009e-05,
|
| 13 |
+
"loss": 1.1578,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.06,
|
| 18 |
+
"learning_rate": 1.98871292830406e-05,
|
| 19 |
+
"loss": 1.1739,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"learning_rate": 1.982954218255111e-05,
|
| 25 |
+
"loss": 1.1411,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.12,
|
| 30 |
+
"learning_rate": 1.977195508206162e-05,
|
| 31 |
+
"loss": 1.1673,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.14,
|
| 36 |
+
"learning_rate": 1.971436798157213e-05,
|
| 37 |
+
"loss": 1.1968,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.17,
|
| 42 |
+
"learning_rate": 1.965678088108264e-05,
|
| 43 |
+
"loss": 1.195,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.2,
|
| 48 |
+
"learning_rate": 1.9599193780593148e-05,
|
| 49 |
+
"loss": 1.1281,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.23,
|
| 54 |
+
"learning_rate": 1.954218255110855e-05,
|
| 55 |
+
"loss": 1.182,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.26,
|
| 60 |
+
"learning_rate": 1.9484595450619064e-05,
|
| 61 |
+
"loss": 1.1541,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.29,
|
| 66 |
+
"learning_rate": 1.9427008350129574e-05,
|
| 67 |
+
"loss": 1.1361,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.32,
|
| 72 |
+
"learning_rate": 1.936942124964008e-05,
|
| 73 |
+
"loss": 1.2445,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.35,
|
| 78 |
+
"learning_rate": 1.9311834149150593e-05,
|
| 79 |
+
"loss": 1.2288,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.37,
|
| 84 |
+
"learning_rate": 1.9254247048661102e-05,
|
| 85 |
+
"loss": 1.2621,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.4,
|
| 90 |
+
"learning_rate": 1.9196659948171612e-05,
|
| 91 |
+
"loss": 1.1978,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.43,
|
| 96 |
+
"learning_rate": 1.913907284768212e-05,
|
| 97 |
+
"loss": 1.176,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.46,
|
| 102 |
+
"learning_rate": 1.908148574719263e-05,
|
| 103 |
+
"loss": 1.165,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.49,
|
| 108 |
+
"learning_rate": 1.902389864670314e-05,
|
| 109 |
+
"loss": 1.1823,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.52,
|
| 114 |
+
"learning_rate": 1.896631154621365e-05,
|
| 115 |
+
"loss": 1.212,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.55,
|
| 120 |
+
"learning_rate": 1.890872444572416e-05,
|
| 121 |
+
"loss": 1.2491,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.58,
|
| 126 |
+
"learning_rate": 1.885113734523467e-05,
|
| 127 |
+
"loss": 1.1902,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.6,
|
| 132 |
+
"learning_rate": 1.8794126115750072e-05,
|
| 133 |
+
"loss": 1.22,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.63,
|
| 138 |
+
"learning_rate": 1.8736539015260585e-05,
|
| 139 |
+
"loss": 1.1474,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.66,
|
| 144 |
+
"learning_rate": 1.867895191477109e-05,
|
| 145 |
+
"loss": 1.1868,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.69,
|
| 150 |
+
"learning_rate": 1.86213648142816e-05,
|
| 151 |
+
"loss": 1.2253,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.72,
|
| 156 |
+
"learning_rate": 1.8563777713792114e-05,
|
| 157 |
+
"loss": 1.2124,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.75,
|
| 162 |
+
"learning_rate": 1.850619061330262e-05,
|
| 163 |
+
"loss": 1.1693,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.78,
|
| 168 |
+
"learning_rate": 1.8448603512813133e-05,
|
| 169 |
+
"loss": 1.16,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.81,
|
| 174 |
+
"learning_rate": 1.8391016412323642e-05,
|
| 175 |
+
"loss": 1.1936,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.83,
|
| 180 |
+
"learning_rate": 1.833342931183415e-05,
|
| 181 |
+
"loss": 1.215,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.86,
|
| 186 |
+
"learning_rate": 1.827584221134466e-05,
|
| 187 |
+
"loss": 1.1846,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.89,
|
| 192 |
+
"learning_rate": 1.821825511085517e-05,
|
| 193 |
+
"loss": 1.1746,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.92,
|
| 198 |
+
"learning_rate": 1.8160668010365677e-05,
|
| 199 |
+
"loss": 1.1871,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.95,
|
| 204 |
+
"learning_rate": 1.810308090987619e-05,
|
| 205 |
+
"loss": 1.1507,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.98,
|
| 210 |
+
"learning_rate": 1.80454938093867e-05,
|
| 211 |
+
"loss": 1.2005,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.0,
|
| 216 |
+
"eval_gen_len": 60.5187,
|
| 217 |
+
"eval_loss": 1.355536699295044,
|
| 218 |
+
"eval_rouge1": 39.5211,
|
| 219 |
+
"eval_rouge2": 20.7719,
|
| 220 |
+
"eval_rougeL": 30.7229,
|
| 221 |
+
"eval_rougeLsum": 36.5645,
|
| 222 |
+
"eval_runtime": 1874.0573,
|
| 223 |
+
"eval_samples_per_second": 1.857,
|
| 224 |
+
"eval_steps_per_second": 0.928,
|
| 225 |
+
"step": 3473
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.01,
|
| 229 |
+
"learning_rate": 1.798790670889721e-05,
|
| 230 |
+
"loss": 1.0801,
|
| 231 |
+
"step": 3500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.04,
|
| 235 |
+
"learning_rate": 1.793031960840772e-05,
|
| 236 |
+
"loss": 0.8483,
|
| 237 |
+
"step": 3600
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 1.07,
|
| 241 |
+
"learning_rate": 1.7872732507918228e-05,
|
| 242 |
+
"loss": 0.875,
|
| 243 |
+
"step": 3700
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.09,
|
| 247 |
+
"learning_rate": 1.7815145407428738e-05,
|
| 248 |
+
"loss": 0.8482,
|
| 249 |
+
"step": 3800
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.12,
|
| 253 |
+
"learning_rate": 1.7757558306939247e-05,
|
| 254 |
+
"loss": 0.8639,
|
| 255 |
+
"step": 3900
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.15,
|
| 259 |
+
"learning_rate": 1.7699971206449757e-05,
|
| 260 |
+
"loss": 0.8351,
|
| 261 |
+
"step": 4000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.18,
|
| 265 |
+
"learning_rate": 1.7642384105960266e-05,
|
| 266 |
+
"loss": 0.8661,
|
| 267 |
+
"step": 4100
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.21,
|
| 271 |
+
"learning_rate": 1.7584797005470776e-05,
|
| 272 |
+
"loss": 0.9151,
|
| 273 |
+
"step": 4200
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.24,
|
| 277 |
+
"learning_rate": 1.7527209904981285e-05,
|
| 278 |
+
"loss": 0.8422,
|
| 279 |
+
"step": 4300
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.27,
|
| 283 |
+
"learning_rate": 1.7469622804491795e-05,
|
| 284 |
+
"loss": 0.9115,
|
| 285 |
+
"step": 4400
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.3,
|
| 289 |
+
"learning_rate": 1.7412035704002304e-05,
|
| 290 |
+
"loss": 0.8703,
|
| 291 |
+
"step": 4500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.32,
|
| 295 |
+
"learning_rate": 1.7354448603512814e-05,
|
| 296 |
+
"loss": 0.8565,
|
| 297 |
+
"step": 4600
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.35,
|
| 301 |
+
"learning_rate": 1.7296861503023323e-05,
|
| 302 |
+
"loss": 0.8967,
|
| 303 |
+
"step": 4700
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.38,
|
| 307 |
+
"learning_rate": 1.7239274402533833e-05,
|
| 308 |
+
"loss": 0.8496,
|
| 309 |
+
"step": 4800
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.41,
|
| 313 |
+
"learning_rate": 1.7181687302044342e-05,
|
| 314 |
+
"loss": 0.9083,
|
| 315 |
+
"step": 4900
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.44,
|
| 319 |
+
"learning_rate": 1.7124100201554855e-05,
|
| 320 |
+
"loss": 0.8918,
|
| 321 |
+
"step": 5000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.47,
|
| 325 |
+
"learning_rate": 1.706651310106536e-05,
|
| 326 |
+
"loss": 0.8788,
|
| 327 |
+
"step": 5100
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.5,
|
| 331 |
+
"learning_rate": 1.700892600057587e-05,
|
| 332 |
+
"loss": 0.8718,
|
| 333 |
+
"step": 5200
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.53,
|
| 337 |
+
"learning_rate": 1.6951338900086384e-05,
|
| 338 |
+
"loss": 0.8925,
|
| 339 |
+
"step": 5300
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.55,
|
| 343 |
+
"learning_rate": 1.689375179959689e-05,
|
| 344 |
+
"loss": 0.8785,
|
| 345 |
+
"step": 5400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.58,
|
| 349 |
+
"learning_rate": 1.6836164699107403e-05,
|
| 350 |
+
"loss": 0.889,
|
| 351 |
+
"step": 5500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.61,
|
| 355 |
+
"learning_rate": 1.6778577598617912e-05,
|
| 356 |
+
"loss": 0.9017,
|
| 357 |
+
"step": 5600
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.64,
|
| 361 |
+
"learning_rate": 1.672099049812842e-05,
|
| 362 |
+
"loss": 0.9152,
|
| 363 |
+
"step": 5700
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.67,
|
| 367 |
+
"learning_rate": 1.666340339763893e-05,
|
| 368 |
+
"loss": 0.9084,
|
| 369 |
+
"step": 5800
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.7,
|
| 373 |
+
"learning_rate": 1.660581629714944e-05,
|
| 374 |
+
"loss": 0.9019,
|
| 375 |
+
"step": 5900
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.73,
|
| 379 |
+
"learning_rate": 1.6548229196659947e-05,
|
| 380 |
+
"loss": 0.9162,
|
| 381 |
+
"step": 6000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.76,
|
| 385 |
+
"learning_rate": 1.649064209617046e-05,
|
| 386 |
+
"loss": 0.8894,
|
| 387 |
+
"step": 6100
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.79,
|
| 391 |
+
"learning_rate": 1.643305499568097e-05,
|
| 392 |
+
"loss": 0.9131,
|
| 393 |
+
"step": 6200
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.81,
|
| 397 |
+
"learning_rate": 1.637546789519148e-05,
|
| 398 |
+
"loss": 0.9156,
|
| 399 |
+
"step": 6300
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.84,
|
| 403 |
+
"learning_rate": 1.6318456665706882e-05,
|
| 404 |
+
"loss": 0.8944,
|
| 405 |
+
"step": 6400
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.87,
|
| 409 |
+
"learning_rate": 1.6260869565217392e-05,
|
| 410 |
+
"loss": 0.8816,
|
| 411 |
+
"step": 6500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.9,
|
| 415 |
+
"learning_rate": 1.62032824647279e-05,
|
| 416 |
+
"loss": 0.8735,
|
| 417 |
+
"step": 6600
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.93,
|
| 421 |
+
"learning_rate": 1.614569536423841e-05,
|
| 422 |
+
"loss": 0.935,
|
| 423 |
+
"step": 6700
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.96,
|
| 427 |
+
"learning_rate": 1.608810826374892e-05,
|
| 428 |
+
"loss": 0.8962,
|
| 429 |
+
"step": 6800
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.99,
|
| 433 |
+
"learning_rate": 1.603052116325943e-05,
|
| 434 |
+
"loss": 0.9181,
|
| 435 |
+
"step": 6900
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 2.0,
|
| 439 |
+
"eval_gen_len": 59.9184,
|
| 440 |
+
"eval_loss": 1.3943339586257935,
|
| 441 |
+
"eval_rouge1": 39.5279,
|
| 442 |
+
"eval_rouge2": 20.2797,
|
| 443 |
+
"eval_rougeL": 30.5951,
|
| 444 |
+
"eval_rougeLsum": 36.6789,
|
| 445 |
+
"eval_runtime": 1843.6748,
|
| 446 |
+
"eval_samples_per_second": 1.888,
|
| 447 |
+
"eval_steps_per_second": 0.944,
|
| 448 |
+
"step": 6946
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 2.02,
|
| 452 |
+
"learning_rate": 1.597293406276994e-05,
|
| 453 |
+
"loss": 0.7443,
|
| 454 |
+
"step": 7000
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.04,
|
| 458 |
+
"learning_rate": 1.5915346962280452e-05,
|
| 459 |
+
"loss": 0.5957,
|
| 460 |
+
"step": 7100
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 2.07,
|
| 464 |
+
"learning_rate": 1.585775986179096e-05,
|
| 465 |
+
"loss": 0.6329,
|
| 466 |
+
"step": 7200
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 2.1,
|
| 470 |
+
"learning_rate": 1.5800172761301468e-05,
|
| 471 |
+
"loss": 0.6256,
|
| 472 |
+
"step": 7300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 2.13,
|
| 476 |
+
"learning_rate": 1.574258566081198e-05,
|
| 477 |
+
"loss": 0.6239,
|
| 478 |
+
"step": 7400
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.16,
|
| 482 |
+
"learning_rate": 1.5684998560322487e-05,
|
| 483 |
+
"loss": 0.6247,
|
| 484 |
+
"step": 7500
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 2.19,
|
| 488 |
+
"learning_rate": 1.5627411459833e-05,
|
| 489 |
+
"loss": 0.6203,
|
| 490 |
+
"step": 7600
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 2.22,
|
| 494 |
+
"learning_rate": 1.556982435934351e-05,
|
| 495 |
+
"loss": 0.6359,
|
| 496 |
+
"step": 7700
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.25,
|
| 500 |
+
"learning_rate": 1.5512237258854016e-05,
|
| 501 |
+
"loss": 0.6419,
|
| 502 |
+
"step": 7800
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 2.27,
|
| 506 |
+
"learning_rate": 1.545465015836453e-05,
|
| 507 |
+
"loss": 0.6646,
|
| 508 |
+
"step": 7900
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 2.3,
|
| 512 |
+
"learning_rate": 1.5397063057875038e-05,
|
| 513 |
+
"loss": 0.662,
|
| 514 |
+
"step": 8000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 2.33,
|
| 518 |
+
"learning_rate": 1.5339475957385548e-05,
|
| 519 |
+
"loss": 0.6209,
|
| 520 |
+
"step": 8100
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.36,
|
| 524 |
+
"learning_rate": 1.5281888856896057e-05,
|
| 525 |
+
"loss": 0.6836,
|
| 526 |
+
"step": 8200
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 2.39,
|
| 530 |
+
"learning_rate": 1.5224301756406565e-05,
|
| 531 |
+
"loss": 0.6497,
|
| 532 |
+
"step": 8300
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 2.42,
|
| 536 |
+
"learning_rate": 1.5166714655917076e-05,
|
| 537 |
+
"loss": 0.6458,
|
| 538 |
+
"step": 8400
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.45,
|
| 542 |
+
"learning_rate": 1.5109127555427586e-05,
|
| 543 |
+
"loss": 0.6631,
|
| 544 |
+
"step": 8500
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 2.48,
|
| 548 |
+
"learning_rate": 1.5051540454938097e-05,
|
| 549 |
+
"loss": 0.6804,
|
| 550 |
+
"step": 8600
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 2.5,
|
| 554 |
+
"learning_rate": 1.49945292254535e-05,
|
| 555 |
+
"loss": 0.6577,
|
| 556 |
+
"step": 8700
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 2.53,
|
| 560 |
+
"learning_rate": 1.4936942124964008e-05,
|
| 561 |
+
"loss": 0.6427,
|
| 562 |
+
"step": 8800
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.56,
|
| 566 |
+
"learning_rate": 1.487935502447452e-05,
|
| 567 |
+
"loss": 0.6792,
|
| 568 |
+
"step": 8900
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 2.59,
|
| 572 |
+
"learning_rate": 1.4821767923985029e-05,
|
| 573 |
+
"loss": 0.6669,
|
| 574 |
+
"step": 9000
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 2.62,
|
| 578 |
+
"learning_rate": 1.4764180823495537e-05,
|
| 579 |
+
"loss": 0.6682,
|
| 580 |
+
"step": 9100
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.65,
|
| 584 |
+
"learning_rate": 1.4706593723006048e-05,
|
| 585 |
+
"loss": 0.6558,
|
| 586 |
+
"step": 9200
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 2.68,
|
| 590 |
+
"learning_rate": 1.4649006622516557e-05,
|
| 591 |
+
"loss": 0.636,
|
| 592 |
+
"step": 9300
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.71,
|
| 596 |
+
"learning_rate": 1.4591419522027069e-05,
|
| 597 |
+
"loss": 0.6373,
|
| 598 |
+
"step": 9400
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 2.74,
|
| 602 |
+
"learning_rate": 1.4534408292542472e-05,
|
| 603 |
+
"loss": 0.633,
|
| 604 |
+
"step": 9500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.76,
|
| 608 |
+
"learning_rate": 1.4476821192052982e-05,
|
| 609 |
+
"loss": 0.7137,
|
| 610 |
+
"step": 9600
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 2.79,
|
| 614 |
+
"learning_rate": 1.441923409156349e-05,
|
| 615 |
+
"loss": 0.6387,
|
| 616 |
+
"step": 9700
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 2.82,
|
| 620 |
+
"learning_rate": 1.4361646991074e-05,
|
| 621 |
+
"loss": 0.6408,
|
| 622 |
+
"step": 9800
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.85,
|
| 626 |
+
"learning_rate": 1.430405989058451e-05,
|
| 627 |
+
"loss": 0.6785,
|
| 628 |
+
"step": 9900
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.88,
|
| 632 |
+
"learning_rate": 1.424647279009502e-05,
|
| 633 |
+
"loss": 0.6866,
|
| 634 |
+
"step": 10000
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.91,
|
| 638 |
+
"learning_rate": 1.418888568960553e-05,
|
| 639 |
+
"loss": 0.6642,
|
| 640 |
+
"step": 10100
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 2.94,
|
| 644 |
+
"learning_rate": 1.4131298589116039e-05,
|
| 645 |
+
"loss": 0.6602,
|
| 646 |
+
"step": 10200
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.97,
|
| 650 |
+
"learning_rate": 1.4073711488626548e-05,
|
| 651 |
+
"loss": 0.6597,
|
| 652 |
+
"step": 10300
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 2.99,
|
| 656 |
+
"learning_rate": 1.4016124388137058e-05,
|
| 657 |
+
"loss": 0.6889,
|
| 658 |
+
"step": 10400
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 3.0,
|
| 662 |
+
"eval_gen_len": 60.1934,
|
| 663 |
+
"eval_loss": 1.4120277166366577,
|
| 664 |
+
"eval_rouge1": 39.6648,
|
| 665 |
+
"eval_rouge2": 20.1994,
|
| 666 |
+
"eval_rougeL": 30.7884,
|
| 667 |
+
"eval_rougeLsum": 36.5151,
|
| 668 |
+
"eval_runtime": 1868.7911,
|
| 669 |
+
"eval_samples_per_second": 1.862,
|
| 670 |
+
"eval_steps_per_second": 0.931,
|
| 671 |
+
"step": 10419
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 3.02,
|
| 675 |
+
"learning_rate": 1.3958537287647569e-05,
|
| 676 |
+
"loss": 0.4742,
|
| 677 |
+
"step": 10500
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 3.05,
|
| 681 |
+
"learning_rate": 1.3900950187158077e-05,
|
| 682 |
+
"loss": 0.4733,
|
| 683 |
+
"step": 10600
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 3.08,
|
| 687 |
+
"learning_rate": 1.3843363086668586e-05,
|
| 688 |
+
"loss": 0.4536,
|
| 689 |
+
"step": 10700
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 3.11,
|
| 693 |
+
"learning_rate": 1.3785775986179098e-05,
|
| 694 |
+
"loss": 0.4418,
|
| 695 |
+
"step": 10800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.14,
|
| 699 |
+
"learning_rate": 1.3728188885689605e-05,
|
| 700 |
+
"loss": 0.4492,
|
| 701 |
+
"step": 10900
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.17,
|
| 705 |
+
"learning_rate": 1.3670601785200117e-05,
|
| 706 |
+
"loss": 0.4429,
|
| 707 |
+
"step": 11000
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 3.2,
|
| 711 |
+
"learning_rate": 1.3613014684710626e-05,
|
| 712 |
+
"loss": 0.4516,
|
| 713 |
+
"step": 11100
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 3.22,
|
| 717 |
+
"learning_rate": 1.3555427584221134e-05,
|
| 718 |
+
"loss": 0.4559,
|
| 719 |
+
"step": 11200
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 3.25,
|
| 723 |
+
"learning_rate": 1.3497840483731645e-05,
|
| 724 |
+
"loss": 0.4701,
|
| 725 |
+
"step": 11300
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 3.28,
|
| 729 |
+
"learning_rate": 1.3440253383242155e-05,
|
| 730 |
+
"loss": 0.4586,
|
| 731 |
+
"step": 11400
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 3.31,
|
| 735 |
+
"learning_rate": 1.3382666282752666e-05,
|
| 736 |
+
"loss": 0.487,
|
| 737 |
+
"step": 11500
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.34,
|
| 741 |
+
"learning_rate": 1.3325079182263174e-05,
|
| 742 |
+
"loss": 0.4554,
|
| 743 |
+
"step": 11600
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.37,
|
| 747 |
+
"learning_rate": 1.3267492081773683e-05,
|
| 748 |
+
"loss": 0.4505,
|
| 749 |
+
"step": 11700
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 3.4,
|
| 753 |
+
"learning_rate": 1.3209904981284194e-05,
|
| 754 |
+
"loss": 0.4586,
|
| 755 |
+
"step": 11800
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 3.43,
|
| 759 |
+
"learning_rate": 1.3152317880794702e-05,
|
| 760 |
+
"loss": 0.4697,
|
| 761 |
+
"step": 11900
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 3.46,
|
| 765 |
+
"learning_rate": 1.3094730780305214e-05,
|
| 766 |
+
"loss": 0.4477,
|
| 767 |
+
"step": 12000
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 3.48,
|
| 771 |
+
"learning_rate": 1.3037143679815723e-05,
|
| 772 |
+
"loss": 0.4631,
|
| 773 |
+
"step": 12100
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 3.51,
|
| 777 |
+
"learning_rate": 1.2979556579326231e-05,
|
| 778 |
+
"loss": 0.4827,
|
| 779 |
+
"step": 12200
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.54,
|
| 783 |
+
"learning_rate": 1.2921969478836742e-05,
|
| 784 |
+
"loss": 0.4524,
|
| 785 |
+
"step": 12300
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.57,
|
| 789 |
+
"learning_rate": 1.2864382378347252e-05,
|
| 790 |
+
"loss": 0.484,
|
| 791 |
+
"step": 12400
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 3.6,
|
| 795 |
+
"learning_rate": 1.2806795277857763e-05,
|
| 796 |
+
"loss": 0.4786,
|
| 797 |
+
"step": 12500
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 3.63,
|
| 801 |
+
"learning_rate": 1.274920817736827e-05,
|
| 802 |
+
"loss": 0.4699,
|
| 803 |
+
"step": 12600
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 3.66,
|
| 807 |
+
"learning_rate": 1.269162107687878e-05,
|
| 808 |
+
"loss": 0.4754,
|
| 809 |
+
"step": 12700
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 3.69,
|
| 813 |
+
"learning_rate": 1.2634033976389291e-05,
|
| 814 |
+
"loss": 0.4851,
|
| 815 |
+
"step": 12800
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 3.71,
|
| 819 |
+
"learning_rate": 1.25764468758998e-05,
|
| 820 |
+
"loss": 0.4556,
|
| 821 |
+
"step": 12900
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.74,
|
| 825 |
+
"learning_rate": 1.2519435646415203e-05,
|
| 826 |
+
"loss": 0.4796,
|
| 827 |
+
"step": 13000
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.77,
|
| 831 |
+
"learning_rate": 1.2461848545925714e-05,
|
| 832 |
+
"loss": 0.4667,
|
| 833 |
+
"step": 13100
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 3.8,
|
| 837 |
+
"learning_rate": 1.2404261445436223e-05,
|
| 838 |
+
"loss": 0.4604,
|
| 839 |
+
"step": 13200
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 3.83,
|
| 843 |
+
"learning_rate": 1.2346674344946731e-05,
|
| 844 |
+
"loss": 0.4585,
|
| 845 |
+
"step": 13300
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 3.86,
|
| 849 |
+
"learning_rate": 1.2289087244457242e-05,
|
| 850 |
+
"loss": 0.4569,
|
| 851 |
+
"step": 13400
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 3.89,
|
| 855 |
+
"learning_rate": 1.2231500143967752e-05,
|
| 856 |
+
"loss": 0.481,
|
| 857 |
+
"step": 13500
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 3.92,
|
| 861 |
+
"learning_rate": 1.2173913043478263e-05,
|
| 862 |
+
"loss": 0.4908,
|
| 863 |
+
"step": 13600
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.94,
|
| 867 |
+
"learning_rate": 1.2116325942988771e-05,
|
| 868 |
+
"loss": 0.503,
|
| 869 |
+
"step": 13700
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.97,
|
| 873 |
+
"learning_rate": 1.205873884249928e-05,
|
| 874 |
+
"loss": 0.5112,
|
| 875 |
+
"step": 13800
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 4.0,
|
| 879 |
+
"eval_gen_len": 60.4152,
|
| 880 |
+
"eval_loss": 1.665522575378418,
|
| 881 |
+
"eval_rouge1": 39.8086,
|
| 882 |
+
"eval_rouge2": 19.9927,
|
| 883 |
+
"eval_rougeL": 30.635,
|
| 884 |
+
"eval_rougeLsum": 36.8366,
|
| 885 |
+
"eval_runtime": 1876.2894,
|
| 886 |
+
"eval_samples_per_second": 1.855,
|
| 887 |
+
"eval_steps_per_second": 0.927,
|
| 888 |
+
"step": 13892
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 4.0,
|
| 892 |
+
"learning_rate": 1.2001151742009792e-05,
|
| 893 |
+
"loss": 0.4853,
|
| 894 |
+
"step": 13900
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.03,
|
| 898 |
+
"learning_rate": 1.19435646415203e-05,
|
| 899 |
+
"loss": 0.316,
|
| 900 |
+
"step": 14000
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 4.06,
|
| 904 |
+
"learning_rate": 1.188597754103081e-05,
|
| 905 |
+
"loss": 0.3047,
|
| 906 |
+
"step": 14100
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.09,
|
| 910 |
+
"learning_rate": 1.182839044054132e-05,
|
| 911 |
+
"loss": 0.2988,
|
| 912 |
+
"step": 14200
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.12,
|
| 916 |
+
"learning_rate": 1.1770803340051828e-05,
|
| 917 |
+
"loss": 0.3301,
|
| 918 |
+
"step": 14300
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 4.15,
|
| 922 |
+
"learning_rate": 1.1713792110567235e-05,
|
| 923 |
+
"loss": 0.3166,
|
| 924 |
+
"step": 14400
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 4.18,
|
| 928 |
+
"learning_rate": 1.1656205010077743e-05,
|
| 929 |
+
"loss": 0.3229,
|
| 930 |
+
"step": 14500
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 4.2,
|
| 934 |
+
"learning_rate": 1.1598617909588252e-05,
|
| 935 |
+
"loss": 0.3222,
|
| 936 |
+
"step": 14600
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 4.23,
|
| 940 |
+
"learning_rate": 1.1541030809098764e-05,
|
| 941 |
+
"loss": 0.3276,
|
| 942 |
+
"step": 14700
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 4.26,
|
| 946 |
+
"learning_rate": 1.1483443708609271e-05,
|
| 947 |
+
"loss": 0.3232,
|
| 948 |
+
"step": 14800
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.29,
|
| 952 |
+
"learning_rate": 1.1425856608119783e-05,
|
| 953 |
+
"loss": 0.3192,
|
| 954 |
+
"step": 14900
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.32,
|
| 958 |
+
"learning_rate": 1.1368269507630292e-05,
|
| 959 |
+
"loss": 0.334,
|
| 960 |
+
"step": 15000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 4.35,
|
| 964 |
+
"learning_rate": 1.13106824071408e-05,
|
| 965 |
+
"loss": 0.3217,
|
| 966 |
+
"step": 15100
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 4.38,
|
| 970 |
+
"learning_rate": 1.1253095306651311e-05,
|
| 971 |
+
"loss": 0.3363,
|
| 972 |
+
"step": 15200
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 4.41,
|
| 976 |
+
"learning_rate": 1.119550820616182e-05,
|
| 977 |
+
"loss": 0.336,
|
| 978 |
+
"step": 15300
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.43,
|
| 982 |
+
"learning_rate": 1.1137921105672332e-05,
|
| 983 |
+
"loss": 0.3283,
|
| 984 |
+
"step": 15400
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 4.46,
|
| 988 |
+
"learning_rate": 1.108033400518284e-05,
|
| 989 |
+
"loss": 0.323,
|
| 990 |
+
"step": 15500
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.49,
|
| 994 |
+
"learning_rate": 1.102274690469335e-05,
|
| 995 |
+
"loss": 0.3122,
|
| 996 |
+
"step": 15600
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.52,
|
| 1000 |
+
"learning_rate": 1.096515980420386e-05,
|
| 1001 |
+
"loss": 0.3443,
|
| 1002 |
+
"step": 15700
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 4.55,
|
| 1006 |
+
"learning_rate": 1.0907572703714368e-05,
|
| 1007 |
+
"loss": 0.329,
|
| 1008 |
+
"step": 15800
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 4.58,
|
| 1012 |
+
"learning_rate": 1.084998560322488e-05,
|
| 1013 |
+
"loss": 0.3214,
|
| 1014 |
+
"step": 15900
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 4.61,
|
| 1018 |
+
"learning_rate": 1.0792398502735389e-05,
|
| 1019 |
+
"loss": 0.3444,
|
| 1020 |
+
"step": 16000
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 4.64,
|
| 1024 |
+
"learning_rate": 1.0734811402245897e-05,
|
| 1025 |
+
"loss": 0.3389,
|
| 1026 |
+
"step": 16100
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 4.66,
|
| 1030 |
+
"learning_rate": 1.0677224301756408e-05,
|
| 1031 |
+
"loss": 0.3404,
|
| 1032 |
+
"step": 16200
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 4.69,
|
| 1036 |
+
"learning_rate": 1.0619637201266918e-05,
|
| 1037 |
+
"loss": 0.3405,
|
| 1038 |
+
"step": 16300
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.72,
|
| 1042 |
+
"learning_rate": 1.0562050100777425e-05,
|
| 1043 |
+
"loss": 0.324,
|
| 1044 |
+
"step": 16400
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 4.75,
|
| 1048 |
+
"learning_rate": 1.0504463000287937e-05,
|
| 1049 |
+
"loss": 0.3242,
|
| 1050 |
+
"step": 16500
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 4.78,
|
| 1054 |
+
"learning_rate": 1.0446875899798446e-05,
|
| 1055 |
+
"loss": 0.3236,
|
| 1056 |
+
"step": 16600
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 4.81,
|
| 1060 |
+
"learning_rate": 1.0389288799308956e-05,
|
| 1061 |
+
"loss": 0.3301,
|
| 1062 |
+
"step": 16700
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 4.84,
|
| 1066 |
+
"learning_rate": 1.0331701698819465e-05,
|
| 1067 |
+
"loss": 0.3501,
|
| 1068 |
+
"step": 16800
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 4.87,
|
| 1072 |
+
"learning_rate": 1.0274114598329975e-05,
|
| 1073 |
+
"loss": 0.3405,
|
| 1074 |
+
"step": 16900
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.89,
|
| 1078 |
+
"learning_rate": 1.0216527497840484e-05,
|
| 1079 |
+
"loss": 0.3388,
|
| 1080 |
+
"step": 17000
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.92,
|
| 1084 |
+
"learning_rate": 1.0158940397350994e-05,
|
| 1085 |
+
"loss": 0.3389,
|
| 1086 |
+
"step": 17100
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 4.95,
|
| 1090 |
+
"learning_rate": 1.0101353296861505e-05,
|
| 1091 |
+
"loss": 0.3563,
|
| 1092 |
+
"step": 17200
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 4.98,
|
| 1096 |
+
"learning_rate": 1.0043766196372013e-05,
|
| 1097 |
+
"loss": 0.3554,
|
| 1098 |
+
"step": 17300
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 5.0,
|
| 1102 |
+
"eval_gen_len": 59.7468,
|
| 1103 |
+
"eval_loss": 1.7994598150253296,
|
| 1104 |
+
"eval_rouge1": 40.9514,
|
| 1105 |
+
"eval_rouge2": 20.4703,
|
| 1106 |
+
"eval_rougeL": 31.7911,
|
| 1107 |
+
"eval_rougeLsum": 37.7023,
|
| 1108 |
+
"eval_runtime": 1848.9202,
|
| 1109 |
+
"eval_samples_per_second": 1.882,
|
| 1110 |
+
"eval_steps_per_second": 0.941,
|
| 1111 |
+
"step": 17365
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 5.01,
|
| 1115 |
+
"learning_rate": 9.986179095882524e-06,
|
| 1116 |
+
"loss": 0.2886,
|
| 1117 |
+
"step": 17400
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 5.04,
|
| 1121 |
+
"learning_rate": 9.928591995393033e-06,
|
| 1122 |
+
"loss": 0.2131,
|
| 1123 |
+
"step": 17500
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 5.07,
|
| 1127 |
+
"learning_rate": 9.871004894903541e-06,
|
| 1128 |
+
"loss": 0.2163,
|
| 1129 |
+
"step": 17600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.1,
|
| 1133 |
+
"learning_rate": 9.813417794414052e-06,
|
| 1134 |
+
"loss": 0.2184,
|
| 1135 |
+
"step": 17700
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 5.13,
|
| 1139 |
+
"learning_rate": 9.755830693924562e-06,
|
| 1140 |
+
"loss": 0.2381,
|
| 1141 |
+
"step": 17800
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 5.15,
|
| 1145 |
+
"learning_rate": 9.698243593435071e-06,
|
| 1146 |
+
"loss": 0.2174,
|
| 1147 |
+
"step": 17900
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 5.18,
|
| 1151 |
+
"learning_rate": 9.640656492945581e-06,
|
| 1152 |
+
"loss": 0.2226,
|
| 1153 |
+
"step": 18000
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.21,
|
| 1157 |
+
"learning_rate": 9.58306939245609e-06,
|
| 1158 |
+
"loss": 0.2383,
|
| 1159 |
+
"step": 18100
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 5.24,
|
| 1163 |
+
"learning_rate": 9.5254822919666e-06,
|
| 1164 |
+
"loss": 0.2366,
|
| 1165 |
+
"step": 18200
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 5.27,
|
| 1169 |
+
"learning_rate": 9.46789519147711e-06,
|
| 1170 |
+
"loss": 0.2206,
|
| 1171 |
+
"step": 18300
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.3,
|
| 1175 |
+
"learning_rate": 9.410308090987619e-06,
|
| 1176 |
+
"loss": 0.2302,
|
| 1177 |
+
"step": 18400
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 5.33,
|
| 1181 |
+
"learning_rate": 9.352720990498129e-06,
|
| 1182 |
+
"loss": 0.2204,
|
| 1183 |
+
"step": 18500
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 5.36,
|
| 1187 |
+
"learning_rate": 9.295133890008638e-06,
|
| 1188 |
+
"loss": 0.2414,
|
| 1189 |
+
"step": 18600
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 5.38,
|
| 1193 |
+
"learning_rate": 9.237546789519148e-06,
|
| 1194 |
+
"loss": 0.2326,
|
| 1195 |
+
"step": 18700
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 5.41,
|
| 1199 |
+
"learning_rate": 9.179959689029659e-06,
|
| 1200 |
+
"loss": 0.2387,
|
| 1201 |
+
"step": 18800
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 5.44,
|
| 1205 |
+
"learning_rate": 9.122372588540168e-06,
|
| 1206 |
+
"loss": 0.2247,
|
| 1207 |
+
"step": 18900
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 5.47,
|
| 1211 |
+
"learning_rate": 9.064785488050676e-06,
|
| 1212 |
+
"loss": 0.2306,
|
| 1213 |
+
"step": 19000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.5,
|
| 1217 |
+
"learning_rate": 9.007198387561187e-06,
|
| 1218 |
+
"loss": 0.2456,
|
| 1219 |
+
"step": 19100
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 5.53,
|
| 1223 |
+
"learning_rate": 8.949611287071697e-06,
|
| 1224 |
+
"loss": 0.2187,
|
| 1225 |
+
"step": 19200
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 5.56,
|
| 1229 |
+
"learning_rate": 8.892024186582206e-06,
|
| 1230 |
+
"loss": 0.2425,
|
| 1231 |
+
"step": 19300
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 5.59,
|
| 1235 |
+
"learning_rate": 8.834437086092716e-06,
|
| 1236 |
+
"loss": 0.2229,
|
| 1237 |
+
"step": 19400
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 5.61,
|
| 1241 |
+
"learning_rate": 8.776849985603225e-06,
|
| 1242 |
+
"loss": 0.2296,
|
| 1243 |
+
"step": 19500
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 5.64,
|
| 1247 |
+
"learning_rate": 8.71983875611863e-06,
|
| 1248 |
+
"loss": 0.2354,
|
| 1249 |
+
"step": 19600
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 5.67,
|
| 1253 |
+
"learning_rate": 8.66225165562914e-06,
|
| 1254 |
+
"loss": 0.2258,
|
| 1255 |
+
"step": 19700
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.7,
|
| 1259 |
+
"learning_rate": 8.60466455513965e-06,
|
| 1260 |
+
"loss": 0.2307,
|
| 1261 |
+
"step": 19800
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 5.73,
|
| 1265 |
+
"learning_rate": 8.54707745465016e-06,
|
| 1266 |
+
"loss": 0.2257,
|
| 1267 |
+
"step": 19900
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 5.76,
|
| 1271 |
+
"learning_rate": 8.489490354160669e-06,
|
| 1272 |
+
"loss": 0.2335,
|
| 1273 |
+
"step": 20000
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 5.79,
|
| 1277 |
+
"learning_rate": 8.431903253671178e-06,
|
| 1278 |
+
"loss": 0.2433,
|
| 1279 |
+
"step": 20100
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 5.82,
|
| 1283 |
+
"learning_rate": 8.374316153181688e-06,
|
| 1284 |
+
"loss": 0.2349,
|
| 1285 |
+
"step": 20200
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 5.84,
|
| 1289 |
+
"learning_rate": 8.316729052692197e-06,
|
| 1290 |
+
"loss": 0.2372,
|
| 1291 |
+
"step": 20300
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 5.87,
|
| 1295 |
+
"learning_rate": 8.259141952202707e-06,
|
| 1296 |
+
"loss": 0.2479,
|
| 1297 |
+
"step": 20400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.9,
|
| 1301 |
+
"learning_rate": 8.201554851713216e-06,
|
| 1302 |
+
"loss": 0.232,
|
| 1303 |
+
"step": 20500
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 5.93,
|
| 1307 |
+
"learning_rate": 8.143967751223728e-06,
|
| 1308 |
+
"loss": 0.2373,
|
| 1309 |
+
"step": 20600
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 5.96,
|
| 1313 |
+
"learning_rate": 8.086380650734235e-06,
|
| 1314 |
+
"loss": 0.2175,
|
| 1315 |
+
"step": 20700
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 5.99,
|
| 1319 |
+
"learning_rate": 8.028793550244745e-06,
|
| 1320 |
+
"loss": 0.2414,
|
| 1321 |
+
"step": 20800
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 6.0,
|
| 1325 |
+
"eval_gen_len": 59.9523,
|
| 1326 |
+
"eval_loss": 2.0208189487457275,
|
| 1327 |
+
"eval_rouge1": 40.7013,
|
| 1328 |
+
"eval_rouge2": 20.2748,
|
| 1329 |
+
"eval_rougeL": 31.3653,
|
| 1330 |
+
"eval_rougeLsum": 37.5646,
|
| 1331 |
+
"eval_runtime": 1863.4517,
|
| 1332 |
+
"eval_samples_per_second": 1.868,
|
| 1333 |
+
"eval_steps_per_second": 0.934,
|
| 1334 |
+
"step": 20838
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 6.02,
|
| 1338 |
+
"learning_rate": 7.971206449755256e-06,
|
| 1339 |
+
"loss": 0.1724,
|
| 1340 |
+
"step": 20900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 6.05,
|
| 1344 |
+
"learning_rate": 7.913619349265766e-06,
|
| 1345 |
+
"loss": 0.155,
|
| 1346 |
+
"step": 21000
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.08,
|
| 1350 |
+
"learning_rate": 7.856032248776275e-06,
|
| 1351 |
+
"loss": 0.1492,
|
| 1352 |
+
"step": 21100
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 6.1,
|
| 1356 |
+
"learning_rate": 7.798445148286785e-06,
|
| 1357 |
+
"loss": 0.1561,
|
| 1358 |
+
"step": 21200
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.13,
|
| 1362 |
+
"learning_rate": 7.740858047797294e-06,
|
| 1363 |
+
"loss": 0.1561,
|
| 1364 |
+
"step": 21300
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 6.16,
|
| 1368 |
+
"learning_rate": 7.683270947307804e-06,
|
| 1369 |
+
"loss": 0.1572,
|
| 1370 |
+
"step": 21400
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 6.19,
|
| 1374 |
+
"learning_rate": 7.625683846818314e-06,
|
| 1375 |
+
"loss": 0.1755,
|
| 1376 |
+
"step": 21500
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 6.22,
|
| 1380 |
+
"learning_rate": 7.568096746328823e-06,
|
| 1381 |
+
"loss": 0.1659,
|
| 1382 |
+
"step": 21600
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 6.25,
|
| 1386 |
+
"learning_rate": 7.510509645839332e-06,
|
| 1387 |
+
"loss": 0.1652,
|
| 1388 |
+
"step": 21700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.28,
|
| 1392 |
+
"learning_rate": 7.452922545349843e-06,
|
| 1393 |
+
"loss": 0.1605,
|
| 1394 |
+
"step": 21800
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 6.31,
|
| 1398 |
+
"learning_rate": 7.395335444860352e-06,
|
| 1399 |
+
"loss": 0.1558,
|
| 1400 |
+
"step": 21900
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 6.33,
|
| 1404 |
+
"learning_rate": 7.337748344370862e-06,
|
| 1405 |
+
"loss": 0.1591,
|
| 1406 |
+
"step": 22000
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 6.36,
|
| 1410 |
+
"learning_rate": 7.28016124388137e-06,
|
| 1411 |
+
"loss": 0.1502,
|
| 1412 |
+
"step": 22100
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 6.39,
|
| 1416 |
+
"learning_rate": 7.222574143391881e-06,
|
| 1417 |
+
"loss": 0.1598,
|
| 1418 |
+
"step": 22200
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 6.42,
|
| 1422 |
+
"learning_rate": 7.16498704290239e-06,
|
| 1423 |
+
"loss": 0.1536,
|
| 1424 |
+
"step": 22300
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 6.45,
|
| 1428 |
+
"learning_rate": 7.1073999424129005e-06,
|
| 1429 |
+
"loss": 0.1501,
|
| 1430 |
+
"step": 22400
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.48,
|
| 1434 |
+
"learning_rate": 7.04981284192341e-06,
|
| 1435 |
+
"loss": 0.1487,
|
| 1436 |
+
"step": 22500
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 6.51,
|
| 1440 |
+
"learning_rate": 6.9928016124388144e-06,
|
| 1441 |
+
"loss": 0.1563,
|
| 1442 |
+
"step": 22600
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 6.54,
|
| 1446 |
+
"learning_rate": 6.935214511949324e-06,
|
| 1447 |
+
"loss": 0.1536,
|
| 1448 |
+
"step": 22700
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 6.56,
|
| 1452 |
+
"learning_rate": 6.877627411459834e-06,
|
| 1453 |
+
"loss": 0.164,
|
| 1454 |
+
"step": 22800
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 6.59,
|
| 1458 |
+
"learning_rate": 6.820040310970343e-06,
|
| 1459 |
+
"loss": 0.1676,
|
| 1460 |
+
"step": 22900
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 6.62,
|
| 1464 |
+
"learning_rate": 6.7624532104808525e-06,
|
| 1465 |
+
"loss": 0.1513,
|
| 1466 |
+
"step": 23000
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 6.65,
|
| 1470 |
+
"learning_rate": 6.704866109991363e-06,
|
| 1471 |
+
"loss": 0.1566,
|
| 1472 |
+
"step": 23100
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.68,
|
| 1476 |
+
"learning_rate": 6.647279009501872e-06,
|
| 1477 |
+
"loss": 0.1511,
|
| 1478 |
+
"step": 23200
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 6.71,
|
| 1482 |
+
"learning_rate": 6.589691909012382e-06,
|
| 1483 |
+
"loss": 0.1608,
|
| 1484 |
+
"step": 23300
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 6.74,
|
| 1488 |
+
"learning_rate": 6.532104808522891e-06,
|
| 1489 |
+
"loss": 0.1633,
|
| 1490 |
+
"step": 23400
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.77,
|
| 1494 |
+
"learning_rate": 6.474517708033401e-06,
|
| 1495 |
+
"loss": 0.1588,
|
| 1496 |
+
"step": 23500
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 6.8,
|
| 1500 |
+
"learning_rate": 6.4169306075439104e-06,
|
| 1501 |
+
"loss": 0.1597,
|
| 1502 |
+
"step": 23600
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 6.82,
|
| 1506 |
+
"learning_rate": 6.359343507054421e-06,
|
| 1507 |
+
"loss": 0.1654,
|
| 1508 |
+
"step": 23700
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 6.85,
|
| 1512 |
+
"learning_rate": 6.3017564065649295e-06,
|
| 1513 |
+
"loss": 0.1615,
|
| 1514 |
+
"step": 23800
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.88,
|
| 1518 |
+
"learning_rate": 6.244169306075439e-06,
|
| 1519 |
+
"loss": 0.1612,
|
| 1520 |
+
"step": 23900
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 6.91,
|
| 1524 |
+
"learning_rate": 6.186582205585949e-06,
|
| 1525 |
+
"loss": 0.1532,
|
| 1526 |
+
"step": 24000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 6.94,
|
| 1530 |
+
"learning_rate": 6.128995105096459e-06,
|
| 1531 |
+
"loss": 0.148,
|
| 1532 |
+
"step": 24100
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 6.97,
|
| 1536 |
+
"learning_rate": 6.071408004606969e-06,
|
| 1537 |
+
"loss": 0.1671,
|
| 1538 |
+
"step": 24200
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 7.0,
|
| 1542 |
+
"learning_rate": 6.013820904117478e-06,
|
| 1543 |
+
"loss": 0.1529,
|
| 1544 |
+
"step": 24300
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 7.0,
|
| 1548 |
+
"eval_gen_len": 60.129,
|
| 1549 |
+
"eval_loss": 2.3271801471710205,
|
| 1550 |
+
"eval_rouge1": 40.6267,
|
| 1551 |
+
"eval_rouge2": 19.9817,
|
| 1552 |
+
"eval_rougeL": 31.2612,
|
| 1553 |
+
"eval_rougeLsum": 37.1969,
|
| 1554 |
+
"eval_runtime": 1881.9513,
|
| 1555 |
+
"eval_samples_per_second": 1.849,
|
| 1556 |
+
"eval_steps_per_second": 0.925,
|
| 1557 |
+
"step": 24311
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 7.03,
|
| 1561 |
+
"learning_rate": 5.956233803627987e-06,
|
| 1562 |
+
"loss": 0.1123,
|
| 1563 |
+
"step": 24400
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.05,
|
| 1567 |
+
"learning_rate": 5.898646703138498e-06,
|
| 1568 |
+
"loss": 0.1085,
|
| 1569 |
+
"step": 24500
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 7.08,
|
| 1573 |
+
"learning_rate": 5.841059602649007e-06,
|
| 1574 |
+
"loss": 0.1008,
|
| 1575 |
+
"step": 24600
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 7.11,
|
| 1579 |
+
"learning_rate": 5.783472502159518e-06,
|
| 1580 |
+
"loss": 0.1167,
|
| 1581 |
+
"step": 24700
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 7.14,
|
| 1585 |
+
"learning_rate": 5.725885401670026e-06,
|
| 1586 |
+
"loss": 0.1039,
|
| 1587 |
+
"step": 24800
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 7.17,
|
| 1591 |
+
"learning_rate": 5.668298301180536e-06,
|
| 1592 |
+
"loss": 0.1102,
|
| 1593 |
+
"step": 24900
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 7.2,
|
| 1597 |
+
"learning_rate": 5.610711200691046e-06,
|
| 1598 |
+
"loss": 0.1127,
|
| 1599 |
+
"step": 25000
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 7.23,
|
| 1603 |
+
"learning_rate": 5.55369997120645e-06,
|
| 1604 |
+
"loss": 0.1149,
|
| 1605 |
+
"step": 25100
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.26,
|
| 1609 |
+
"learning_rate": 5.49611287071696e-06,
|
| 1610 |
+
"loss": 0.1194,
|
| 1611 |
+
"step": 25200
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 7.28,
|
| 1615 |
+
"learning_rate": 5.43852577022747e-06,
|
| 1616 |
+
"loss": 0.106,
|
| 1617 |
+
"step": 25300
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 7.31,
|
| 1621 |
+
"learning_rate": 5.380938669737979e-06,
|
| 1622 |
+
"loss": 0.1123,
|
| 1623 |
+
"step": 25400
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 7.34,
|
| 1627 |
+
"learning_rate": 5.323351569248489e-06,
|
| 1628 |
+
"loss": 0.1158,
|
| 1629 |
+
"step": 25500
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 7.37,
|
| 1633 |
+
"learning_rate": 5.265764468758998e-06,
|
| 1634 |
+
"loss": 0.1069,
|
| 1635 |
+
"step": 25600
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 7.4,
|
| 1639 |
+
"learning_rate": 5.208177368269508e-06,
|
| 1640 |
+
"loss": 0.109,
|
| 1641 |
+
"step": 25700
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 7.43,
|
| 1645 |
+
"learning_rate": 5.150590267780018e-06,
|
| 1646 |
+
"loss": 0.1188,
|
| 1647 |
+
"step": 25800
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.46,
|
| 1651 |
+
"learning_rate": 5.0930031672905276e-06,
|
| 1652 |
+
"loss": 0.1077,
|
| 1653 |
+
"step": 25900
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 7.49,
|
| 1657 |
+
"learning_rate": 5.035416066801036e-06,
|
| 1658 |
+
"loss": 0.1108,
|
| 1659 |
+
"step": 26000
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 7.52,
|
| 1663 |
+
"learning_rate": 4.977828966311547e-06,
|
| 1664 |
+
"loss": 0.1136,
|
| 1665 |
+
"step": 26100
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 7.54,
|
| 1669 |
+
"learning_rate": 4.920817736826951e-06,
|
| 1670 |
+
"loss": 0.1054,
|
| 1671 |
+
"step": 26200
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 7.57,
|
| 1675 |
+
"learning_rate": 4.863230636337461e-06,
|
| 1676 |
+
"loss": 0.1104,
|
| 1677 |
+
"step": 26300
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 7.6,
|
| 1681 |
+
"learning_rate": 4.80564353584797e-06,
|
| 1682 |
+
"loss": 0.1079,
|
| 1683 |
+
"step": 26400
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 7.63,
|
| 1687 |
+
"learning_rate": 4.74805643535848e-06,
|
| 1688 |
+
"loss": 0.1047,
|
| 1689 |
+
"step": 26500
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.66,
|
| 1693 |
+
"learning_rate": 4.690469334868989e-06,
|
| 1694 |
+
"loss": 0.1083,
|
| 1695 |
+
"step": 26600
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 7.69,
|
| 1699 |
+
"learning_rate": 4.632882234379499e-06,
|
| 1700 |
+
"loss": 0.1065,
|
| 1701 |
+
"step": 26700
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 7.72,
|
| 1705 |
+
"learning_rate": 4.575295133890009e-06,
|
| 1706 |
+
"loss": 0.1095,
|
| 1707 |
+
"step": 26800
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 7.75,
|
| 1711 |
+
"learning_rate": 4.5177080334005185e-06,
|
| 1712 |
+
"loss": 0.1056,
|
| 1713 |
+
"step": 26900
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 7.77,
|
| 1717 |
+
"learning_rate": 4.460120932911029e-06,
|
| 1718 |
+
"loss": 0.1052,
|
| 1719 |
+
"step": 27000
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 7.8,
|
| 1723 |
+
"learning_rate": 4.4025338324215375e-06,
|
| 1724 |
+
"loss": 0.1107,
|
| 1725 |
+
"step": 27100
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 7.83,
|
| 1729 |
+
"learning_rate": 4.344946731932048e-06,
|
| 1730 |
+
"loss": 0.1049,
|
| 1731 |
+
"step": 27200
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.86,
|
| 1735 |
+
"learning_rate": 4.287359631442557e-06,
|
| 1736 |
+
"loss": 0.102,
|
| 1737 |
+
"step": 27300
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 7.89,
|
| 1741 |
+
"learning_rate": 4.229772530953067e-06,
|
| 1742 |
+
"loss": 0.1156,
|
| 1743 |
+
"step": 27400
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 7.92,
|
| 1747 |
+
"learning_rate": 4.172185430463576e-06,
|
| 1748 |
+
"loss": 0.0907,
|
| 1749 |
+
"step": 27500
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 7.95,
|
| 1753 |
+
"learning_rate": 4.114598329974086e-06,
|
| 1754 |
+
"loss": 0.1052,
|
| 1755 |
+
"step": 27600
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 7.98,
|
| 1759 |
+
"learning_rate": 4.057011229484596e-06,
|
| 1760 |
+
"loss": 0.1236,
|
| 1761 |
+
"step": 27700
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 8.0,
|
| 1765 |
+
"eval_gen_len": 59.8247,
|
| 1766 |
+
"eval_loss": 2.426095485687256,
|
| 1767 |
+
"eval_rouge1": 41.8276,
|
| 1768 |
+
"eval_rouge2": 20.7036,
|
| 1769 |
+
"eval_rougeL": 32.3998,
|
| 1770 |
+
"eval_rougeLsum": 38.5374,
|
| 1771 |
+
"eval_runtime": 1864.0588,
|
| 1772 |
+
"eval_samples_per_second": 1.867,
|
| 1773 |
+
"eval_steps_per_second": 0.933,
|
| 1774 |
+
"step": 27784
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 8.0,
|
| 1778 |
+
"learning_rate": 3.999424128995105e-06,
|
| 1779 |
+
"loss": 0.1053,
|
| 1780 |
+
"step": 27800
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 8.03,
|
| 1784 |
+
"learning_rate": 3.941837028505615e-06,
|
| 1785 |
+
"loss": 0.0783,
|
| 1786 |
+
"step": 27900
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 8.06,
|
| 1790 |
+
"learning_rate": 3.884249928016125e-06,
|
| 1791 |
+
"loss": 0.0753,
|
| 1792 |
+
"step": 28000
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 8.09,
|
| 1796 |
+
"learning_rate": 3.826662827526634e-06,
|
| 1797 |
+
"loss": 0.0759,
|
| 1798 |
+
"step": 28100
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 8.12,
|
| 1802 |
+
"learning_rate": 3.7690757270371443e-06,
|
| 1803 |
+
"loss": 0.0737,
|
| 1804 |
+
"step": 28200
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 8.15,
|
| 1808 |
+
"learning_rate": 3.7114886265476534e-06,
|
| 1809 |
+
"loss": 0.0733,
|
| 1810 |
+
"step": 28300
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 8.18,
|
| 1814 |
+
"learning_rate": 3.6539015260581633e-06,
|
| 1815 |
+
"loss": 0.0771,
|
| 1816 |
+
"step": 28400
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 8.21,
|
| 1820 |
+
"learning_rate": 3.596314425568673e-06,
|
| 1821 |
+
"loss": 0.081,
|
| 1822 |
+
"step": 28500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.23,
|
| 1826 |
+
"learning_rate": 3.5387273250791828e-06,
|
| 1827 |
+
"loss": 0.0757,
|
| 1828 |
+
"step": 28600
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 8.26,
|
| 1832 |
+
"learning_rate": 3.481140224589692e-06,
|
| 1833 |
+
"loss": 0.0692,
|
| 1834 |
+
"step": 28700
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 8.29,
|
| 1838 |
+
"learning_rate": 3.4241289951050967e-06,
|
| 1839 |
+
"loss": 0.0683,
|
| 1840 |
+
"step": 28800
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 8.32,
|
| 1844 |
+
"learning_rate": 3.3665418946156066e-06,
|
| 1845 |
+
"loss": 0.074,
|
| 1846 |
+
"step": 28900
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 8.35,
|
| 1850 |
+
"learning_rate": 3.3089547941261157e-06,
|
| 1851 |
+
"loss": 0.0728,
|
| 1852 |
+
"step": 29000
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 8.38,
|
| 1856 |
+
"learning_rate": 3.2513676936366256e-06,
|
| 1857 |
+
"loss": 0.0734,
|
| 1858 |
+
"step": 29100
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 8.41,
|
| 1862 |
+
"learning_rate": 3.1937805931471356e-06,
|
| 1863 |
+
"loss": 0.0707,
|
| 1864 |
+
"step": 29200
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.44,
|
| 1868 |
+
"learning_rate": 3.1361934926576447e-06,
|
| 1869 |
+
"loss": 0.0755,
|
| 1870 |
+
"step": 29300
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 8.47,
|
| 1874 |
+
"learning_rate": 3.0786063921681546e-06,
|
| 1875 |
+
"loss": 0.08,
|
| 1876 |
+
"step": 29400
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 8.49,
|
| 1880 |
+
"learning_rate": 3.021019291678664e-06,
|
| 1881 |
+
"loss": 0.075,
|
| 1882 |
+
"step": 29500
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 8.52,
|
| 1886 |
+
"learning_rate": 2.963432191189174e-06,
|
| 1887 |
+
"loss": 0.0709,
|
| 1888 |
+
"step": 29600
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 8.55,
|
| 1892 |
+
"learning_rate": 2.905845090699683e-06,
|
| 1893 |
+
"loss": 0.0749,
|
| 1894 |
+
"step": 29700
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 8.58,
|
| 1898 |
+
"learning_rate": 2.848257990210193e-06,
|
| 1899 |
+
"loss": 0.0707,
|
| 1900 |
+
"step": 29800
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 8.61,
|
| 1904 |
+
"learning_rate": 2.790670889720703e-06,
|
| 1905 |
+
"loss": 0.0748,
|
| 1906 |
+
"step": 29900
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.64,
|
| 1910 |
+
"learning_rate": 2.733083789231212e-06,
|
| 1911 |
+
"loss": 0.0708,
|
| 1912 |
+
"step": 30000
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 8.67,
|
| 1916 |
+
"learning_rate": 2.675496688741722e-06,
|
| 1917 |
+
"loss": 0.0784,
|
| 1918 |
+
"step": 30100
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 8.7,
|
| 1922 |
+
"learning_rate": 2.6179095882522316e-06,
|
| 1923 |
+
"loss": 0.069,
|
| 1924 |
+
"step": 30200
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 8.72,
|
| 1928 |
+
"learning_rate": 2.5603224877627415e-06,
|
| 1929 |
+
"loss": 0.0766,
|
| 1930 |
+
"step": 30300
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 8.75,
|
| 1934 |
+
"learning_rate": 2.5027353872732515e-06,
|
| 1935 |
+
"loss": 0.0809,
|
| 1936 |
+
"step": 30400
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 8.78,
|
| 1940 |
+
"learning_rate": 2.4451482867837606e-06,
|
| 1941 |
+
"loss": 0.0739,
|
| 1942 |
+
"step": 30500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 8.81,
|
| 1946 |
+
"learning_rate": 2.38756118629427e-06,
|
| 1947 |
+
"loss": 0.0765,
|
| 1948 |
+
"step": 30600
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.84,
|
| 1952 |
+
"learning_rate": 2.3299740858047796e-06,
|
| 1953 |
+
"loss": 0.0728,
|
| 1954 |
+
"step": 30700
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 8.87,
|
| 1958 |
+
"learning_rate": 2.2723869853152895e-06,
|
| 1959 |
+
"loss": 0.0787,
|
| 1960 |
+
"step": 30800
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 8.9,
|
| 1964 |
+
"learning_rate": 2.2147998848257995e-06,
|
| 1965 |
+
"loss": 0.0847,
|
| 1966 |
+
"step": 30900
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 8.93,
|
| 1970 |
+
"learning_rate": 2.157212784336309e-06,
|
| 1971 |
+
"loss": 0.0791,
|
| 1972 |
+
"step": 31000
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 8.95,
|
| 1976 |
+
"learning_rate": 2.0996256838468185e-06,
|
| 1977 |
+
"loss": 0.0644,
|
| 1978 |
+
"step": 31100
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 8.98,
|
| 1982 |
+
"learning_rate": 2.042038583357328e-06,
|
| 1983 |
+
"loss": 0.078,
|
| 1984 |
+
"step": 31200
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 9.0,
|
| 1988 |
+
"eval_gen_len": 59.9862,
|
| 1989 |
+
"eval_loss": 2.554534673690796,
|
| 1990 |
+
"eval_rouge1": 41.439,
|
| 1991 |
+
"eval_rouge2": 20.612,
|
| 1992 |
+
"eval_rougeL": 32.0749,
|
| 1993 |
+
"eval_rougeLsum": 38.1027,
|
| 1994 |
+
"eval_runtime": 1862.3443,
|
| 1995 |
+
"eval_samples_per_second": 1.869,
|
| 1996 |
+
"eval_steps_per_second": 0.934,
|
| 1997 |
+
"step": 31257
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 9.01,
|
| 2001 |
+
"learning_rate": 1.9844514828678375e-06,
|
| 2002 |
+
"loss": 0.0672,
|
| 2003 |
+
"step": 31300
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 9.04,
|
| 2007 |
+
"learning_rate": 1.9268643823783475e-06,
|
| 2008 |
+
"loss": 0.0623,
|
| 2009 |
+
"step": 31400
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 9.07,
|
| 2013 |
+
"learning_rate": 1.8692772818888572e-06,
|
| 2014 |
+
"loss": 0.0516,
|
| 2015 |
+
"step": 31500
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 9.1,
|
| 2019 |
+
"learning_rate": 1.8116901813993667e-06,
|
| 2020 |
+
"loss": 0.054,
|
| 2021 |
+
"step": 31600
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 9.13,
|
| 2025 |
+
"learning_rate": 1.7541030809098765e-06,
|
| 2026 |
+
"loss": 0.0478,
|
| 2027 |
+
"step": 31700
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 9.16,
|
| 2031 |
+
"learning_rate": 1.696515980420386e-06,
|
| 2032 |
+
"loss": 0.0573,
|
| 2033 |
+
"step": 31800
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 9.19,
|
| 2037 |
+
"learning_rate": 1.6389288799308955e-06,
|
| 2038 |
+
"loss": 0.0518,
|
| 2039 |
+
"step": 31900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 9.21,
|
| 2043 |
+
"learning_rate": 1.5813417794414052e-06,
|
| 2044 |
+
"loss": 0.0531,
|
| 2045 |
+
"step": 32000
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 9.24,
|
| 2049 |
+
"learning_rate": 1.523754678951915e-06,
|
| 2050 |
+
"loss": 0.0471,
|
| 2051 |
+
"step": 32100
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 9.27,
|
| 2055 |
+
"learning_rate": 1.4661675784624247e-06,
|
| 2056 |
+
"loss": 0.0488,
|
| 2057 |
+
"step": 32200
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 9.3,
|
| 2061 |
+
"learning_rate": 1.4085804779729342e-06,
|
| 2062 |
+
"loss": 0.0477,
|
| 2063 |
+
"step": 32300
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 9.33,
|
| 2067 |
+
"learning_rate": 1.350993377483444e-06,
|
| 2068 |
+
"loss": 0.0561,
|
| 2069 |
+
"step": 32400
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 9.36,
|
| 2073 |
+
"learning_rate": 1.2934062769939534e-06,
|
| 2074 |
+
"loss": 0.0563,
|
| 2075 |
+
"step": 32500
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 9.39,
|
| 2079 |
+
"learning_rate": 1.235819176504463e-06,
|
| 2080 |
+
"loss": 0.0486,
|
| 2081 |
+
"step": 32600
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 9.42,
|
| 2085 |
+
"learning_rate": 1.1782320760149727e-06,
|
| 2086 |
+
"loss": 0.0568,
|
| 2087 |
+
"step": 32700
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 9.44,
|
| 2091 |
+
"learning_rate": 1.1206449755254824e-06,
|
| 2092 |
+
"loss": 0.0504,
|
| 2093 |
+
"step": 32800
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 9.47,
|
| 2097 |
+
"learning_rate": 1.063057875035992e-06,
|
| 2098 |
+
"loss": 0.0556,
|
| 2099 |
+
"step": 32900
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 9.5,
|
| 2103 |
+
"learning_rate": 1.0054707745465017e-06,
|
| 2104 |
+
"loss": 0.0525,
|
| 2105 |
+
"step": 33000
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 9.53,
|
| 2109 |
+
"learning_rate": 9.490354160668011e-07,
|
| 2110 |
+
"loss": 0.057,
|
| 2111 |
+
"step": 33100
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 9.56,
|
| 2115 |
+
"learning_rate": 8.914483155773107e-07,
|
| 2116 |
+
"loss": 0.0508,
|
| 2117 |
+
"step": 33200
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 9.59,
|
| 2121 |
+
"learning_rate": 8.338612150878205e-07,
|
| 2122 |
+
"loss": 0.0489,
|
| 2123 |
+
"step": 33300
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.62,
|
| 2127 |
+
"learning_rate": 7.762741145983301e-07,
|
| 2128 |
+
"loss": 0.0559,
|
| 2129 |
+
"step": 33400
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 9.65,
|
| 2133 |
+
"learning_rate": 7.186870141088396e-07,
|
| 2134 |
+
"loss": 0.0545,
|
| 2135 |
+
"step": 33500
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 9.67,
|
| 2139 |
+
"learning_rate": 6.610999136193493e-07,
|
| 2140 |
+
"loss": 0.0519,
|
| 2141 |
+
"step": 33600
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 9.7,
|
| 2145 |
+
"learning_rate": 6.03512813129859e-07,
|
| 2146 |
+
"loss": 0.0478,
|
| 2147 |
+
"step": 33700
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 9.73,
|
| 2151 |
+
"learning_rate": 5.459257126403686e-07,
|
| 2152 |
+
"loss": 0.0508,
|
| 2153 |
+
"step": 33800
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 9.76,
|
| 2157 |
+
"learning_rate": 4.883386121508782e-07,
|
| 2158 |
+
"loss": 0.0523,
|
| 2159 |
+
"step": 33900
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 9.79,
|
| 2163 |
+
"learning_rate": 4.307515116613879e-07,
|
| 2164 |
+
"loss": 0.0491,
|
| 2165 |
+
"step": 34000
|
| 2166 |
+
}
|
| 2167 |
+
],
|
| 2168 |
+
"max_steps": 34730,
|
| 2169 |
+
"num_train_epochs": 10,
|
| 2170 |
+
"total_flos": 8.56161528236114e+16,
|
| 2171 |
+
"trial_name": null,
|
| 2172 |
+
"trial_params": null
|
| 2173 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d5c6b0d248bc4f75e35afea863fc903ccb1cc5a2870820e14def07ea5fa1a37
|
| 3 |
+
size 3375
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|