Commit
•
5fd36d9
1
Parent(s):
faa1222
Upload 12 files
Browse files- config.json +59 -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 +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +2094 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/mbart-large-cc25",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"activation_dropout": 0.0,
|
| 5 |
+
"activation_function": "gelu",
|
| 6 |
+
"add_bias_logits": false,
|
| 7 |
+
"add_final_layer_norm": true,
|
| 8 |
+
"architectures": [
|
| 9 |
+
"MBartForConditionalGeneration"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"bos_token_id": 0,
|
| 13 |
+
"classif_dropout": 0.0,
|
| 14 |
+
"classifier_dropout": 0.0,
|
| 15 |
+
"d_model": 1024,
|
| 16 |
+
"decoder_attention_heads": 16,
|
| 17 |
+
"decoder_ffn_dim": 4096,
|
| 18 |
+
"decoder_layerdrop": 0.0,
|
| 19 |
+
"decoder_layers": 12,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"encoder_attention_heads": 16,
|
| 22 |
+
"encoder_ffn_dim": 4096,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 12,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"id2label": {
|
| 28 |
+
"0": "LABEL_0",
|
| 29 |
+
"1": "LABEL_1",
|
| 30 |
+
"2": "LABEL_2"
|
| 31 |
+
},
|
| 32 |
+
"init_std": 0.02,
|
| 33 |
+
"is_encoder_decoder": true,
|
| 34 |
+
"label2id": {
|
| 35 |
+
"LABEL_0": 0,
|
| 36 |
+
"LABEL_1": 1,
|
| 37 |
+
"LABEL_2": 2
|
| 38 |
+
},
|
| 39 |
+
"max_length": 1024,
|
| 40 |
+
"max_position_embeddings": 1024,
|
| 41 |
+
"model_type": "mbart",
|
| 42 |
+
"normalize_before": true,
|
| 43 |
+
"normalize_embedding": true,
|
| 44 |
+
"num_beams": 5,
|
| 45 |
+
"num_hidden_layers": 12,
|
| 46 |
+
"output_past": true,
|
| 47 |
+
"pad_token_id": 1,
|
| 48 |
+
"scale_embedding": true,
|
| 49 |
+
"static_position_embeddings": false,
|
| 50 |
+
"task_specific_params": {
|
| 51 |
+
"translation_en_to_ro": {
|
| 52 |
+
"decoder_start_token_id": 250020
|
| 53 |
+
}
|
| 54 |
+
},
|
| 55 |
+
"torch_dtype": "float32",
|
| 56 |
+
"transformers_version": "4.18.0",
|
| 57 |
+
"use_cache": true,
|
| 58 |
+
"vocab_size": 250027
|
| 59 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7d816b0f869ca049d637925853421489e5b74b4205285064ed648d20aa6f842
|
| 3 |
+
size 4887126731
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80eff53e4b151b9fce990cd2a50988568b0f2a74cb1d5a941673d5c39fe82002
|
| 3 |
+
size 2444583325
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:992eabbab4a86910a7a6fb4233615c8d078972c4c2c24ab557899c20f4daccbb
|
| 3 |
+
size 14511
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae96f83e1baa352dc151efdb528a3e7144c2b559ebb36c02e3351c1a6db7dad9
|
| 3 |
+
size 557
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5560dba5efee17dd5934705ac260150c56c6020f8f1ed37dc17dd73e51ee412
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": false, "lowercase": false, "special_tokens_map_file": null, "name_or_path": "./Tokenizer", "do_basic_tokenize": true, "never_split": null, "tokenizer_class": "BertTokenizer"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2094 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9918182928318475,
|
| 5 |
+
"global_step": 171500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 4.985636946193791e-05,
|
| 13 |
+
"loss": 4.5112,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 4.9710994423413524e-05,
|
| 19 |
+
"loss": 2.8906,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.03,
|
| 24 |
+
"learning_rate": 4.956561938488914e-05,
|
| 25 |
+
"loss": 2.5728,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.03,
|
| 30 |
+
"learning_rate": 4.9420244346364754e-05,
|
| 31 |
+
"loss": 2.3159,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.04,
|
| 36 |
+
"learning_rate": 4.927516005791742e-05,
|
| 37 |
+
"loss": 2.0901,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.05,
|
| 42 |
+
"learning_rate": 4.9129785019393034e-05,
|
| 43 |
+
"loss": 1.9299,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06,
|
| 48 |
+
"learning_rate": 4.898440998086865e-05,
|
| 49 |
+
"loss": 1.6719,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.07,
|
| 54 |
+
"learning_rate": 4.883903494234426e-05,
|
| 55 |
+
"loss": 1.5558,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.08,
|
| 60 |
+
"learning_rate": 4.869365990381988e-05,
|
| 61 |
+
"loss": 1.4674,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.09,
|
| 66 |
+
"learning_rate": 4.854828486529549e-05,
|
| 67 |
+
"loss": 1.3837,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.1,
|
| 72 |
+
"learning_rate": 4.840290982677111e-05,
|
| 73 |
+
"loss": 1.2387,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.1,
|
| 78 |
+
"learning_rate": 4.825753478824672e-05,
|
| 79 |
+
"loss": 1.2115,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.11,
|
| 84 |
+
"learning_rate": 4.811215974972234e-05,
|
| 85 |
+
"loss": 1.1154,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.12,
|
| 90 |
+
"learning_rate": 4.7966784711197945e-05,
|
| 91 |
+
"loss": 1.0496,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.13,
|
| 96 |
+
"learning_rate": 4.782140967267357e-05,
|
| 97 |
+
"loss": 0.9716,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.14,
|
| 102 |
+
"learning_rate": 4.767603463414918e-05,
|
| 103 |
+
"loss": 0.907,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.15,
|
| 108 |
+
"learning_rate": 4.753095034570184e-05,
|
| 109 |
+
"loss": 0.9429,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.16,
|
| 114 |
+
"learning_rate": 4.738557530717746e-05,
|
| 115 |
+
"loss": 0.9419,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.17,
|
| 120 |
+
"learning_rate": 4.7240200268653076e-05,
|
| 121 |
+
"loss": 0.8054,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.17,
|
| 126 |
+
"learning_rate": 4.7094825230128684e-05,
|
| 127 |
+
"loss": 0.7875,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.18,
|
| 132 |
+
"learning_rate": 4.6949450191604306e-05,
|
| 133 |
+
"loss": 0.7689,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.19,
|
| 138 |
+
"learning_rate": 4.680407515307992e-05,
|
| 139 |
+
"loss": 0.7609,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.2,
|
| 144 |
+
"learning_rate": 4.665870011455553e-05,
|
| 145 |
+
"loss": 0.7068,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.21,
|
| 150 |
+
"learning_rate": 4.651390657618524e-05,
|
| 151 |
+
"loss": 0.7209,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.22,
|
| 156 |
+
"learning_rate": 4.636853153766086e-05,
|
| 157 |
+
"loss": 0.64,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.23,
|
| 162 |
+
"learning_rate": 4.622315649913647e-05,
|
| 163 |
+
"loss": 0.6096,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.24,
|
| 168 |
+
"learning_rate": 4.607778146061209e-05,
|
| 169 |
+
"loss": 0.6555,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.24,
|
| 174 |
+
"learning_rate": 4.59324064220877e-05,
|
| 175 |
+
"loss": 0.6192,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.25,
|
| 180 |
+
"learning_rate": 4.578703138356332e-05,
|
| 181 |
+
"loss": 0.6229,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.26,
|
| 186 |
+
"learning_rate": 4.564165634503894e-05,
|
| 187 |
+
"loss": 0.5818,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.27,
|
| 192 |
+
"learning_rate": 4.549628130651455e-05,
|
| 193 |
+
"loss": 0.5736,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.28,
|
| 198 |
+
"learning_rate": 4.535090626799016e-05,
|
| 199 |
+
"loss": 0.5501,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.29,
|
| 204 |
+
"learning_rate": 4.520553122946578e-05,
|
| 205 |
+
"loss": 0.5234,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3,
|
| 210 |
+
"learning_rate": 4.506044694101844e-05,
|
| 211 |
+
"loss": 0.5175,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.31,
|
| 216 |
+
"learning_rate": 4.4915071902494056e-05,
|
| 217 |
+
"loss": 0.5241,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.31,
|
| 222 |
+
"learning_rate": 4.476969686396967e-05,
|
| 223 |
+
"loss": 0.4967,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.32,
|
| 228 |
+
"learning_rate": 4.4624321825445286e-05,
|
| 229 |
+
"loss": 0.4896,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.33,
|
| 234 |
+
"learning_rate": 4.44789467869209e-05,
|
| 235 |
+
"loss": 0.5281,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.34,
|
| 240 |
+
"learning_rate": 4.4333571748396515e-05,
|
| 241 |
+
"loss": 0.4756,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.35,
|
| 246 |
+
"learning_rate": 4.418819670987213e-05,
|
| 247 |
+
"loss": 0.4733,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.36,
|
| 252 |
+
"learning_rate": 4.4042821671347745e-05,
|
| 253 |
+
"loss": 0.4575,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.37,
|
| 258 |
+
"learning_rate": 4.389744663282336e-05,
|
| 259 |
+
"loss": 0.4333,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.38,
|
| 264 |
+
"learning_rate": 4.3752071594298975e-05,
|
| 265 |
+
"loss": 0.4247,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.38,
|
| 270 |
+
"learning_rate": 4.360669655577459e-05,
|
| 271 |
+
"loss": 0.4739,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.39,
|
| 276 |
+
"learning_rate": 4.3461321517250204e-05,
|
| 277 |
+
"loss": 0.4469,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.4,
|
| 282 |
+
"learning_rate": 4.331623722880287e-05,
|
| 283 |
+
"loss": 0.3953,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.41,
|
| 288 |
+
"learning_rate": 4.3170862190278484e-05,
|
| 289 |
+
"loss": 0.4534,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.42,
|
| 294 |
+
"learning_rate": 4.30254871517541e-05,
|
| 295 |
+
"loss": 0.4207,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.43,
|
| 300 |
+
"learning_rate": 4.288040286330676e-05,
|
| 301 |
+
"loss": 0.3691,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.44,
|
| 306 |
+
"learning_rate": 4.273502782478238e-05,
|
| 307 |
+
"loss": 0.3827,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.44,
|
| 312 |
+
"learning_rate": 4.2589652786257986e-05,
|
| 313 |
+
"loss": 0.4101,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.45,
|
| 318 |
+
"learning_rate": 4.244427774773361e-05,
|
| 319 |
+
"loss": 0.379,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.46,
|
| 324 |
+
"learning_rate": 4.229890270920922e-05,
|
| 325 |
+
"loss": 0.3616,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.47,
|
| 330 |
+
"learning_rate": 4.215381842076188e-05,
|
| 331 |
+
"loss": 0.4308,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.48,
|
| 336 |
+
"learning_rate": 4.20084433822375e-05,
|
| 337 |
+
"loss": 0.354,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.49,
|
| 342 |
+
"learning_rate": 4.186306834371311e-05,
|
| 343 |
+
"loss": 0.3633,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.5,
|
| 348 |
+
"learning_rate": 4.1717693305188725e-05,
|
| 349 |
+
"loss": 0.4011,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.51,
|
| 354 |
+
"learning_rate": 4.157231826666435e-05,
|
| 355 |
+
"loss": 0.3489,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.51,
|
| 360 |
+
"learning_rate": 4.142694322813996e-05,
|
| 361 |
+
"loss": 0.3721,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.52,
|
| 366 |
+
"learning_rate": 4.128185893969262e-05,
|
| 367 |
+
"loss": 0.3636,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.53,
|
| 372 |
+
"learning_rate": 4.1137065401322334e-05,
|
| 373 |
+
"loss": 0.3521,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.54,
|
| 378 |
+
"learning_rate": 4.099169036279795e-05,
|
| 379 |
+
"loss": 0.3484,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.55,
|
| 384 |
+
"learning_rate": 4.0846315324273564e-05,
|
| 385 |
+
"loss": 0.3303,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.56,
|
| 390 |
+
"learning_rate": 4.070094028574918e-05,
|
| 391 |
+
"loss": 0.3119,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.57,
|
| 396 |
+
"learning_rate": 4.055556524722479e-05,
|
| 397 |
+
"loss": 0.3495,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.58,
|
| 402 |
+
"learning_rate": 4.041019020870041e-05,
|
| 403 |
+
"loss": 0.3297,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.58,
|
| 408 |
+
"learning_rate": 4.026481517017602e-05,
|
| 409 |
+
"loss": 0.318,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.59,
|
| 414 |
+
"learning_rate": 4.011944013165164e-05,
|
| 415 |
+
"loss": 0.3036,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.6,
|
| 420 |
+
"learning_rate": 3.997406509312725e-05,
|
| 421 |
+
"loss": 0.3131,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.61,
|
| 426 |
+
"learning_rate": 3.982869005460287e-05,
|
| 427 |
+
"loss": 0.2966,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.62,
|
| 432 |
+
"learning_rate": 3.968331501607848e-05,
|
| 433 |
+
"loss": 0.3322,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.63,
|
| 438 |
+
"learning_rate": 3.953823072763114e-05,
|
| 439 |
+
"loss": 0.3059,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.64,
|
| 444 |
+
"learning_rate": 3.939285568910676e-05,
|
| 445 |
+
"loss": 0.3008,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.65,
|
| 450 |
+
"learning_rate": 3.9247480650582376e-05,
|
| 451 |
+
"loss": 0.2686,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.65,
|
| 456 |
+
"learning_rate": 3.9102105612057984e-05,
|
| 457 |
+
"loss": 0.3025,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.66,
|
| 462 |
+
"learning_rate": 3.8956730573533606e-05,
|
| 463 |
+
"loss": 0.295,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.67,
|
| 468 |
+
"learning_rate": 3.881135553500922e-05,
|
| 469 |
+
"loss": 0.3083,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.68,
|
| 474 |
+
"learning_rate": 3.8665980496484836e-05,
|
| 475 |
+
"loss": 0.3016,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.69,
|
| 480 |
+
"learning_rate": 3.8520605457960444e-05,
|
| 481 |
+
"loss": 0.3099,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.7,
|
| 486 |
+
"learning_rate": 3.8375230419436065e-05,
|
| 487 |
+
"loss": 0.2942,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.71,
|
| 492 |
+
"learning_rate": 3.822985538091168e-05,
|
| 493 |
+
"loss": 0.2642,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.72,
|
| 498 |
+
"learning_rate": 3.808448034238729e-05,
|
| 499 |
+
"loss": 0.268,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.72,
|
| 504 |
+
"learning_rate": 3.793910530386291e-05,
|
| 505 |
+
"loss": 0.2829,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.73,
|
| 510 |
+
"learning_rate": 3.779431176549262e-05,
|
| 511 |
+
"loss": 0.2487,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.74,
|
| 516 |
+
"learning_rate": 3.764893672696824e-05,
|
| 517 |
+
"loss": 0.2845,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.75,
|
| 522 |
+
"learning_rate": 3.750356168844385e-05,
|
| 523 |
+
"loss": 0.2894,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.76,
|
| 528 |
+
"learning_rate": 3.735818664991946e-05,
|
| 529 |
+
"loss": 0.2812,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.77,
|
| 534 |
+
"learning_rate": 3.721281161139508e-05,
|
| 535 |
+
"loss": 0.2448,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.78,
|
| 540 |
+
"learning_rate": 3.70674365728707e-05,
|
| 541 |
+
"loss": 0.246,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.79,
|
| 546 |
+
"learning_rate": 3.6922061534346307e-05,
|
| 547 |
+
"loss": 0.2405,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.79,
|
| 552 |
+
"learning_rate": 3.677668649582192e-05,
|
| 553 |
+
"loss": 0.259,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.8,
|
| 558 |
+
"learning_rate": 3.663131145729754e-05,
|
| 559 |
+
"loss": 0.2679,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.81,
|
| 564 |
+
"learning_rate": 3.648593641877315e-05,
|
| 565 |
+
"loss": 0.2557,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.82,
|
| 570 |
+
"learning_rate": 3.6340561380248766e-05,
|
| 571 |
+
"loss": 0.2326,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.83,
|
| 576 |
+
"learning_rate": 3.619576784187848e-05,
|
| 577 |
+
"loss": 0.2394,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.84,
|
| 582 |
+
"learning_rate": 3.6050392803354095e-05,
|
| 583 |
+
"loss": 0.2612,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.85,
|
| 588 |
+
"learning_rate": 3.590501776482971e-05,
|
| 589 |
+
"loss": 0.2153,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.85,
|
| 594 |
+
"learning_rate": 3.5759642726305325e-05,
|
| 595 |
+
"loss": 0.2298,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.86,
|
| 600 |
+
"learning_rate": 3.561426768778094e-05,
|
| 601 |
+
"loss": 0.2299,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.87,
|
| 606 |
+
"learning_rate": 3.5468892649256555e-05,
|
| 607 |
+
"loss": 0.2372,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.88,
|
| 612 |
+
"learning_rate": 3.532351761073217e-05,
|
| 613 |
+
"loss": 0.2233,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.89,
|
| 618 |
+
"learning_rate": 3.5178142572207784e-05,
|
| 619 |
+
"loss": 0.2256,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.9,
|
| 624 |
+
"learning_rate": 3.50327675336834e-05,
|
| 625 |
+
"loss": 0.2409,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.91,
|
| 630 |
+
"learning_rate": 3.488739249515901e-05,
|
| 631 |
+
"loss": 0.2315,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.92,
|
| 636 |
+
"learning_rate": 3.474230820671168e-05,
|
| 637 |
+
"loss": 0.2129,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.92,
|
| 642 |
+
"learning_rate": 3.4596933168187287e-05,
|
| 643 |
+
"loss": 0.2215,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.93,
|
| 648 |
+
"learning_rate": 3.445155812966291e-05,
|
| 649 |
+
"loss": 0.22,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.94,
|
| 654 |
+
"learning_rate": 3.430618309113852e-05,
|
| 655 |
+
"loss": 0.2431,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.95,
|
| 660 |
+
"learning_rate": 3.416080805261414e-05,
|
| 661 |
+
"loss": 0.2271,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.96,
|
| 666 |
+
"learning_rate": 3.4015433014089746e-05,
|
| 667 |
+
"loss": 0.2015,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.97,
|
| 672 |
+
"learning_rate": 3.387034872564242e-05,
|
| 673 |
+
"loss": 0.2285,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.98,
|
| 678 |
+
"learning_rate": 3.3724973687118025e-05,
|
| 679 |
+
"loss": 0.227,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.99,
|
| 684 |
+
"learning_rate": 3.357959864859364e-05,
|
| 685 |
+
"loss": 0.2139,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.99,
|
| 690 |
+
"learning_rate": 3.343422361006926e-05,
|
| 691 |
+
"loss": 0.2038,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.0,
|
| 696 |
+
"eval_bleu": 0.0,
|
| 697 |
+
"eval_gen_len": 1.0,
|
| 698 |
+
"eval_loss": 0.13874725997447968,
|
| 699 |
+
"eval_runtime": 2091.8758,
|
| 700 |
+
"eval_samples_per_second": 12.179,
|
| 701 |
+
"eval_steps_per_second": 3.045,
|
| 702 |
+
"step": 57323
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.0,
|
| 706 |
+
"learning_rate": 3.328913932162192e-05,
|
| 707 |
+
"loss": 0.2153,
|
| 708 |
+
"step": 57500
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.01,
|
| 712 |
+
"learning_rate": 3.314376428309754e-05,
|
| 713 |
+
"loss": 0.1445,
|
| 714 |
+
"step": 58000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.02,
|
| 718 |
+
"learning_rate": 3.299838924457315e-05,
|
| 719 |
+
"loss": 0.1382,
|
| 720 |
+
"step": 58500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.03,
|
| 724 |
+
"learning_rate": 3.2853014206048764e-05,
|
| 725 |
+
"loss": 0.1251,
|
| 726 |
+
"step": 59000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.04,
|
| 730 |
+
"learning_rate": 3.270792991760143e-05,
|
| 731 |
+
"loss": 0.1249,
|
| 732 |
+
"step": 59500
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.05,
|
| 736 |
+
"learning_rate": 3.2562554879077044e-05,
|
| 737 |
+
"loss": 0.135,
|
| 738 |
+
"step": 60000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.06,
|
| 742 |
+
"learning_rate": 3.241747059062971e-05,
|
| 743 |
+
"loss": 0.1367,
|
| 744 |
+
"step": 60500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.06,
|
| 748 |
+
"learning_rate": 3.227209555210532e-05,
|
| 749 |
+
"loss": 0.1398,
|
| 750 |
+
"step": 61000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 1.07,
|
| 754 |
+
"learning_rate": 3.212672051358094e-05,
|
| 755 |
+
"loss": 0.1445,
|
| 756 |
+
"step": 61500
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.08,
|
| 760 |
+
"learning_rate": 3.198134547505655e-05,
|
| 761 |
+
"loss": 0.1367,
|
| 762 |
+
"step": 62000
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.09,
|
| 766 |
+
"learning_rate": 3.183597043653217e-05,
|
| 767 |
+
"loss": 0.1416,
|
| 768 |
+
"step": 62500
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.1,
|
| 772 |
+
"learning_rate": 3.169059539800778e-05,
|
| 773 |
+
"loss": 0.1378,
|
| 774 |
+
"step": 63000
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.11,
|
| 778 |
+
"learning_rate": 3.15452203594834e-05,
|
| 779 |
+
"loss": 0.1439,
|
| 780 |
+
"step": 63500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.12,
|
| 784 |
+
"learning_rate": 3.139984532095901e-05,
|
| 785 |
+
"loss": 0.1413,
|
| 786 |
+
"step": 64000
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.13,
|
| 790 |
+
"learning_rate": 3.125447028243463e-05,
|
| 791 |
+
"loss": 0.1345,
|
| 792 |
+
"step": 64500
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.13,
|
| 796 |
+
"learning_rate": 3.110938599398729e-05,
|
| 797 |
+
"loss": 0.1261,
|
| 798 |
+
"step": 65000
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 1.14,
|
| 802 |
+
"learning_rate": 3.09640109554629e-05,
|
| 803 |
+
"loss": 0.1403,
|
| 804 |
+
"step": 65500
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 1.15,
|
| 808 |
+
"learning_rate": 3.081863591693852e-05,
|
| 809 |
+
"loss": 0.1351,
|
| 810 |
+
"step": 66000
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.16,
|
| 814 |
+
"learning_rate": 3.0673260878414136e-05,
|
| 815 |
+
"loss": 0.1414,
|
| 816 |
+
"step": 66500
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.17,
|
| 820 |
+
"learning_rate": 3.0527885839889744e-05,
|
| 821 |
+
"loss": 0.1428,
|
| 822 |
+
"step": 67000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.18,
|
| 826 |
+
"learning_rate": 3.0382510801365366e-05,
|
| 827 |
+
"loss": 0.1404,
|
| 828 |
+
"step": 67500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.19,
|
| 832 |
+
"learning_rate": 3.023713576284098e-05,
|
| 833 |
+
"loss": 0.1611,
|
| 834 |
+
"step": 68000
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 1.19,
|
| 838 |
+
"learning_rate": 3.0091760724316592e-05,
|
| 839 |
+
"loss": 0.1342,
|
| 840 |
+
"step": 68500
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 1.2,
|
| 844 |
+
"learning_rate": 2.994638568579221e-05,
|
| 845 |
+
"loss": 0.1428,
|
| 846 |
+
"step": 69000
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.21,
|
| 850 |
+
"learning_rate": 2.9801010647267825e-05,
|
| 851 |
+
"loss": 0.1441,
|
| 852 |
+
"step": 69500
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1.22,
|
| 856 |
+
"learning_rate": 2.9655635608743436e-05,
|
| 857 |
+
"loss": 0.1204,
|
| 858 |
+
"step": 70000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.23,
|
| 862 |
+
"learning_rate": 2.9510551320296105e-05,
|
| 863 |
+
"loss": 0.132,
|
| 864 |
+
"step": 70500
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.24,
|
| 868 |
+
"learning_rate": 2.9365176281771716e-05,
|
| 869 |
+
"loss": 0.1363,
|
| 870 |
+
"step": 71000
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.25,
|
| 874 |
+
"learning_rate": 2.921980124324733e-05,
|
| 875 |
+
"loss": 0.1293,
|
| 876 |
+
"step": 71500
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 1.26,
|
| 880 |
+
"learning_rate": 2.9074426204722942e-05,
|
| 881 |
+
"loss": 0.1276,
|
| 882 |
+
"step": 72000
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 1.26,
|
| 886 |
+
"learning_rate": 2.892934191627561e-05,
|
| 887 |
+
"loss": 0.1413,
|
| 888 |
+
"step": 72500
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1.27,
|
| 892 |
+
"learning_rate": 2.8783966877751222e-05,
|
| 893 |
+
"loss": 0.1259,
|
| 894 |
+
"step": 73000
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 1.28,
|
| 898 |
+
"learning_rate": 2.863859183922684e-05,
|
| 899 |
+
"loss": 0.1265,
|
| 900 |
+
"step": 73500
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.29,
|
| 904 |
+
"learning_rate": 2.8493216800702455e-05,
|
| 905 |
+
"loss": 0.1382,
|
| 906 |
+
"step": 74000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.3,
|
| 910 |
+
"learning_rate": 2.8348132512255116e-05,
|
| 911 |
+
"loss": 0.1313,
|
| 912 |
+
"step": 74500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.31,
|
| 916 |
+
"learning_rate": 2.8202757473730734e-05,
|
| 917 |
+
"loss": 0.1366,
|
| 918 |
+
"step": 75000
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 1.32,
|
| 922 |
+
"learning_rate": 2.805738243520635e-05,
|
| 923 |
+
"loss": 0.1283,
|
| 924 |
+
"step": 75500
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 1.33,
|
| 928 |
+
"learning_rate": 2.7912298146759007e-05,
|
| 929 |
+
"loss": 0.128,
|
| 930 |
+
"step": 76000
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 1.33,
|
| 934 |
+
"learning_rate": 2.776692310823463e-05,
|
| 935 |
+
"loss": 0.124,
|
| 936 |
+
"step": 76500
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 1.34,
|
| 940 |
+
"learning_rate": 2.762154806971024e-05,
|
| 941 |
+
"loss": 0.1176,
|
| 942 |
+
"step": 77000
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.35,
|
| 946 |
+
"learning_rate": 2.7476173031185855e-05,
|
| 947 |
+
"loss": 0.1314,
|
| 948 |
+
"step": 77500
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.36,
|
| 952 |
+
"learning_rate": 2.7330797992661473e-05,
|
| 953 |
+
"loss": 0.1146,
|
| 954 |
+
"step": 78000
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.37,
|
| 958 |
+
"learning_rate": 2.7185422954137084e-05,
|
| 959 |
+
"loss": 0.1198,
|
| 960 |
+
"step": 78500
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 1.38,
|
| 964 |
+
"learning_rate": 2.70400479156127e-05,
|
| 965 |
+
"loss": 0.1187,
|
| 966 |
+
"step": 79000
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 1.39,
|
| 970 |
+
"learning_rate": 2.689467287708831e-05,
|
| 971 |
+
"loss": 0.1201,
|
| 972 |
+
"step": 79500
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 1.4,
|
| 976 |
+
"learning_rate": 2.674929783856393e-05,
|
| 977 |
+
"loss": 0.1169,
|
| 978 |
+
"step": 80000
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.4,
|
| 982 |
+
"learning_rate": 2.6603922800039544e-05,
|
| 983 |
+
"loss": 0.1097,
|
| 984 |
+
"step": 80500
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.41,
|
| 988 |
+
"learning_rate": 2.6458547761515155e-05,
|
| 989 |
+
"loss": 0.1263,
|
| 990 |
+
"step": 81000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.42,
|
| 994 |
+
"learning_rate": 2.6313172722990777e-05,
|
| 995 |
+
"loss": 0.1141,
|
| 996 |
+
"step": 81500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.43,
|
| 1000 |
+
"learning_rate": 2.6168088434543435e-05,
|
| 1001 |
+
"loss": 0.1406,
|
| 1002 |
+
"step": 82000
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 1.44,
|
| 1006 |
+
"learning_rate": 2.602271339601905e-05,
|
| 1007 |
+
"loss": 0.1233,
|
| 1008 |
+
"step": 82500
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 1.45,
|
| 1012 |
+
"learning_rate": 2.5877338357494668e-05,
|
| 1013 |
+
"loss": 0.108,
|
| 1014 |
+
"step": 83000
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 1.46,
|
| 1018 |
+
"learning_rate": 2.5731963318970283e-05,
|
| 1019 |
+
"loss": 0.1129,
|
| 1020 |
+
"step": 83500
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 1.47,
|
| 1024 |
+
"learning_rate": 2.5586588280445894e-05,
|
| 1025 |
+
"loss": 0.1181,
|
| 1026 |
+
"step": 84000
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.47,
|
| 1030 |
+
"learning_rate": 2.544121324192151e-05,
|
| 1031 |
+
"loss": 0.1305,
|
| 1032 |
+
"step": 84500
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.48,
|
| 1036 |
+
"learning_rate": 2.5295838203397127e-05,
|
| 1037 |
+
"loss": 0.1111,
|
| 1038 |
+
"step": 85000
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.49,
|
| 1042 |
+
"learning_rate": 2.515075391494979e-05,
|
| 1043 |
+
"loss": 0.1064,
|
| 1044 |
+
"step": 85500
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 1.5,
|
| 1048 |
+
"learning_rate": 2.5005378876425407e-05,
|
| 1049 |
+
"loss": 0.103,
|
| 1050 |
+
"step": 86000
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 1.51,
|
| 1054 |
+
"learning_rate": 2.4860003837901018e-05,
|
| 1055 |
+
"loss": 0.1156,
|
| 1056 |
+
"step": 86500
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 1.52,
|
| 1060 |
+
"learning_rate": 2.4714628799376633e-05,
|
| 1061 |
+
"loss": 0.1359,
|
| 1062 |
+
"step": 87000
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 1.53,
|
| 1066 |
+
"learning_rate": 2.4569253760852248e-05,
|
| 1067 |
+
"loss": 0.1015,
|
| 1068 |
+
"step": 87500
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.54,
|
| 1072 |
+
"learning_rate": 2.4423878722327862e-05,
|
| 1073 |
+
"loss": 0.105,
|
| 1074 |
+
"step": 88000
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.54,
|
| 1078 |
+
"learning_rate": 2.4278794433880524e-05,
|
| 1079 |
+
"loss": 0.1157,
|
| 1080 |
+
"step": 88500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.55,
|
| 1084 |
+
"learning_rate": 2.4133419395356142e-05,
|
| 1085 |
+
"loss": 0.0975,
|
| 1086 |
+
"step": 89000
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 1.56,
|
| 1090 |
+
"learning_rate": 2.3988044356831757e-05,
|
| 1091 |
+
"loss": 0.1053,
|
| 1092 |
+
"step": 89500
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 1.57,
|
| 1096 |
+
"learning_rate": 2.384296006838442e-05,
|
| 1097 |
+
"loss": 0.0981,
|
| 1098 |
+
"step": 90000
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.58,
|
| 1102 |
+
"learning_rate": 2.3697585029860033e-05,
|
| 1103 |
+
"loss": 0.1031,
|
| 1104 |
+
"step": 90500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.59,
|
| 1108 |
+
"learning_rate": 2.355220999133565e-05,
|
| 1109 |
+
"loss": 0.1124,
|
| 1110 |
+
"step": 91000
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.6,
|
| 1114 |
+
"learning_rate": 2.3406834952811263e-05,
|
| 1115 |
+
"loss": 0.108,
|
| 1116 |
+
"step": 91500
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.6,
|
| 1120 |
+
"learning_rate": 2.3261459914286877e-05,
|
| 1121 |
+
"loss": 0.0937,
|
| 1122 |
+
"step": 92000
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.61,
|
| 1126 |
+
"learning_rate": 2.3116084875762492e-05,
|
| 1127 |
+
"loss": 0.1043,
|
| 1128 |
+
"step": 92500
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 1.62,
|
| 1132 |
+
"learning_rate": 2.2970709837238107e-05,
|
| 1133 |
+
"loss": 0.0982,
|
| 1134 |
+
"step": 93000
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 1.63,
|
| 1138 |
+
"learning_rate": 2.2825625548790772e-05,
|
| 1139 |
+
"loss": 0.1112,
|
| 1140 |
+
"step": 93500
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 1.64,
|
| 1144 |
+
"learning_rate": 2.2680250510266387e-05,
|
| 1145 |
+
"loss": 0.1197,
|
| 1146 |
+
"step": 94000
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 1.65,
|
| 1150 |
+
"learning_rate": 2.2534875471742e-05,
|
| 1151 |
+
"loss": 0.1042,
|
| 1152 |
+
"step": 94500
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.66,
|
| 1156 |
+
"learning_rate": 2.2389500433217616e-05,
|
| 1157 |
+
"loss": 0.0888,
|
| 1158 |
+
"step": 95000
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.67,
|
| 1162 |
+
"learning_rate": 2.224412539469323e-05,
|
| 1163 |
+
"loss": 0.1044,
|
| 1164 |
+
"step": 95500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.67,
|
| 1168 |
+
"learning_rate": 2.2098750356168846e-05,
|
| 1169 |
+
"loss": 0.0927,
|
| 1170 |
+
"step": 96000
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 1.68,
|
| 1174 |
+
"learning_rate": 2.195337531764446e-05,
|
| 1175 |
+
"loss": 0.0971,
|
| 1176 |
+
"step": 96500
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 1.69,
|
| 1180 |
+
"learning_rate": 2.1808291029197122e-05,
|
| 1181 |
+
"loss": 0.1112,
|
| 1182 |
+
"step": 97000
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.7,
|
| 1186 |
+
"learning_rate": 2.166291599067274e-05,
|
| 1187 |
+
"loss": 0.1028,
|
| 1188 |
+
"step": 97500
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 1.71,
|
| 1192 |
+
"learning_rate": 2.1517540952148355e-05,
|
| 1193 |
+
"loss": 0.0985,
|
| 1194 |
+
"step": 98000
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.72,
|
| 1198 |
+
"learning_rate": 2.1372165913623966e-05,
|
| 1199 |
+
"loss": 0.1159,
|
| 1200 |
+
"step": 98500
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.73,
|
| 1204 |
+
"learning_rate": 2.1226790875099585e-05,
|
| 1205 |
+
"loss": 0.0989,
|
| 1206 |
+
"step": 99000
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.74,
|
| 1210 |
+
"learning_rate": 2.1081415836575196e-05,
|
| 1211 |
+
"loss": 0.1067,
|
| 1212 |
+
"step": 99500
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 1.74,
|
| 1216 |
+
"learning_rate": 2.0936040798050814e-05,
|
| 1217 |
+
"loss": 0.1184,
|
| 1218 |
+
"step": 100000
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 1.75,
|
| 1222 |
+
"learning_rate": 2.0790665759526426e-05,
|
| 1223 |
+
"loss": 0.081,
|
| 1224 |
+
"step": 100500
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 1.76,
|
| 1228 |
+
"learning_rate": 2.064529072100204e-05,
|
| 1229 |
+
"loss": 0.0942,
|
| 1230 |
+
"step": 101000
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 1.77,
|
| 1234 |
+
"learning_rate": 2.049991568247766e-05,
|
| 1235 |
+
"loss": 0.1033,
|
| 1236 |
+
"step": 101500
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.78,
|
| 1240 |
+
"learning_rate": 2.035454064395327e-05,
|
| 1241 |
+
"loss": 0.0883,
|
| 1242 |
+
"step": 102000
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.79,
|
| 1246 |
+
"learning_rate": 2.020916560542889e-05,
|
| 1247 |
+
"loss": 0.1017,
|
| 1248 |
+
"step": 102500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.8,
|
| 1252 |
+
"learning_rate": 2.00643720670586e-05,
|
| 1253 |
+
"loss": 0.0915,
|
| 1254 |
+
"step": 103000
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 1.81,
|
| 1258 |
+
"learning_rate": 1.9918997028534214e-05,
|
| 1259 |
+
"loss": 0.1009,
|
| 1260 |
+
"step": 103500
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 1.81,
|
| 1264 |
+
"learning_rate": 1.9773621990009826e-05,
|
| 1265 |
+
"loss": 0.0905,
|
| 1266 |
+
"step": 104000
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.82,
|
| 1270 |
+
"learning_rate": 1.962853770156249e-05,
|
| 1271 |
+
"loss": 0.0989,
|
| 1272 |
+
"step": 104500
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 1.83,
|
| 1276 |
+
"learning_rate": 1.9483162663038105e-05,
|
| 1277 |
+
"loss": 0.1035,
|
| 1278 |
+
"step": 105000
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.84,
|
| 1282 |
+
"learning_rate": 1.933778762451372e-05,
|
| 1283 |
+
"loss": 0.0879,
|
| 1284 |
+
"step": 105500
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.85,
|
| 1288 |
+
"learning_rate": 1.9192412585989335e-05,
|
| 1289 |
+
"loss": 0.1027,
|
| 1290 |
+
"step": 106000
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.86,
|
| 1294 |
+
"learning_rate": 1.9047037547464953e-05,
|
| 1295 |
+
"loss": 0.0813,
|
| 1296 |
+
"step": 106500
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 1.87,
|
| 1300 |
+
"learning_rate": 1.8901662508940565e-05,
|
| 1301 |
+
"loss": 0.0839,
|
| 1302 |
+
"step": 107000
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 1.88,
|
| 1306 |
+
"learning_rate": 1.8756287470416183e-05,
|
| 1307 |
+
"loss": 0.0878,
|
| 1308 |
+
"step": 107500
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 1.88,
|
| 1312 |
+
"learning_rate": 1.8610912431891794e-05,
|
| 1313 |
+
"loss": 0.0826,
|
| 1314 |
+
"step": 108000
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 1.89,
|
| 1318 |
+
"learning_rate": 1.846553739336741e-05,
|
| 1319 |
+
"loss": 0.0934,
|
| 1320 |
+
"step": 108500
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.9,
|
| 1324 |
+
"learning_rate": 1.8320162354843027e-05,
|
| 1325 |
+
"loss": 0.0807,
|
| 1326 |
+
"step": 109000
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.91,
|
| 1330 |
+
"learning_rate": 1.817478731631864e-05,
|
| 1331 |
+
"loss": 0.0856,
|
| 1332 |
+
"step": 109500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.92,
|
| 1336 |
+
"learning_rate": 1.8029412277794257e-05,
|
| 1337 |
+
"loss": 0.0819,
|
| 1338 |
+
"step": 110000
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 1.93,
|
| 1342 |
+
"learning_rate": 1.788403723926987e-05,
|
| 1343 |
+
"loss": 0.0979,
|
| 1344 |
+
"step": 110500
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 1.94,
|
| 1348 |
+
"learning_rate": 1.7738662200745483e-05,
|
| 1349 |
+
"loss": 0.1042,
|
| 1350 |
+
"step": 111000
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.95,
|
| 1354 |
+
"learning_rate": 1.7593287162221098e-05,
|
| 1355 |
+
"loss": 0.0744,
|
| 1356 |
+
"step": 111500
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1.95,
|
| 1360 |
+
"learning_rate": 1.7447912123696713e-05,
|
| 1361 |
+
"loss": 0.0879,
|
| 1362 |
+
"step": 112000
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.96,
|
| 1366 |
+
"learning_rate": 1.730253708517233e-05,
|
| 1367 |
+
"loss": 0.0849,
|
| 1368 |
+
"step": 112500
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.97,
|
| 1372 |
+
"learning_rate": 1.7157452796724992e-05,
|
| 1373 |
+
"loss": 0.0767,
|
| 1374 |
+
"step": 113000
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.98,
|
| 1378 |
+
"learning_rate": 1.7012077758200607e-05,
|
| 1379 |
+
"loss": 0.0859,
|
| 1380 |
+
"step": 113500
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 1.99,
|
| 1384 |
+
"learning_rate": 1.6866702719676222e-05,
|
| 1385 |
+
"loss": 0.0818,
|
| 1386 |
+
"step": 114000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 2.0,
|
| 1390 |
+
"learning_rate": 1.6721618431228887e-05,
|
| 1391 |
+
"loss": 0.0827,
|
| 1392 |
+
"step": 114500
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.0,
|
| 1396 |
+
"eval_bleu": 0.0,
|
| 1397 |
+
"eval_gen_len": 1.0,
|
| 1398 |
+
"eval_loss": 0.04327930510044098,
|
| 1399 |
+
"eval_runtime": 2076.3927,
|
| 1400 |
+
"eval_samples_per_second": 12.27,
|
| 1401 |
+
"eval_steps_per_second": 3.068,
|
| 1402 |
+
"step": 114646
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.01,
|
| 1406 |
+
"learning_rate": 1.6576243392704498e-05,
|
| 1407 |
+
"loss": 0.0638,
|
| 1408 |
+
"step": 115000
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 2.01,
|
| 1412 |
+
"learning_rate": 1.6430868354180116e-05,
|
| 1413 |
+
"loss": 0.0321,
|
| 1414 |
+
"step": 115500
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.02,
|
| 1418 |
+
"learning_rate": 1.6285493315655728e-05,
|
| 1419 |
+
"loss": 0.0423,
|
| 1420 |
+
"step": 116000
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.03,
|
| 1424 |
+
"learning_rate": 1.6140118277131343e-05,
|
| 1425 |
+
"loss": 0.0364,
|
| 1426 |
+
"step": 116500
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.04,
|
| 1430 |
+
"learning_rate": 1.5995033988684007e-05,
|
| 1431 |
+
"loss": 0.0399,
|
| 1432 |
+
"step": 117000
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 2.05,
|
| 1436 |
+
"learning_rate": 1.5849658950159622e-05,
|
| 1437 |
+
"loss": 0.0411,
|
| 1438 |
+
"step": 117500
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.06,
|
| 1442 |
+
"learning_rate": 1.5704283911635237e-05,
|
| 1443 |
+
"loss": 0.034,
|
| 1444 |
+
"step": 118000
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.07,
|
| 1448 |
+
"learning_rate": 1.5558908873110852e-05,
|
| 1449 |
+
"loss": 0.0467,
|
| 1450 |
+
"step": 118500
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 2.08,
|
| 1454 |
+
"learning_rate": 1.5413533834586467e-05,
|
| 1455 |
+
"loss": 0.0395,
|
| 1456 |
+
"step": 119000
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.08,
|
| 1460 |
+
"learning_rate": 1.526815879606208e-05,
|
| 1461 |
+
"loss": 0.0421,
|
| 1462 |
+
"step": 119500
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.09,
|
| 1466 |
+
"learning_rate": 1.5122783757537695e-05,
|
| 1467 |
+
"loss": 0.0425,
|
| 1468 |
+
"step": 120000
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 2.1,
|
| 1472 |
+
"learning_rate": 1.4977408719013311e-05,
|
| 1473 |
+
"loss": 0.0426,
|
| 1474 |
+
"step": 120500
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 2.11,
|
| 1478 |
+
"learning_rate": 1.4832324430565974e-05,
|
| 1479 |
+
"loss": 0.0506,
|
| 1480 |
+
"step": 121000
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.12,
|
| 1484 |
+
"learning_rate": 1.468694939204159e-05,
|
| 1485 |
+
"loss": 0.0341,
|
| 1486 |
+
"step": 121500
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.13,
|
| 1490 |
+
"learning_rate": 1.4541574353517204e-05,
|
| 1491 |
+
"loss": 0.0429,
|
| 1492 |
+
"step": 122000
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 2.14,
|
| 1496 |
+
"learning_rate": 1.439619931499282e-05,
|
| 1497 |
+
"loss": 0.039,
|
| 1498 |
+
"step": 122500
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 2.15,
|
| 1502 |
+
"learning_rate": 1.4250824276468433e-05,
|
| 1503 |
+
"loss": 0.0445,
|
| 1504 |
+
"step": 123000
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.15,
|
| 1508 |
+
"learning_rate": 1.4105449237944048e-05,
|
| 1509 |
+
"loss": 0.0367,
|
| 1510 |
+
"step": 123500
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 2.16,
|
| 1514 |
+
"learning_rate": 1.3960074199419665e-05,
|
| 1515 |
+
"loss": 0.0408,
|
| 1516 |
+
"step": 124000
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 2.17,
|
| 1520 |
+
"learning_rate": 1.3814699160895278e-05,
|
| 1521 |
+
"loss": 0.0432,
|
| 1522 |
+
"step": 124500
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.18,
|
| 1526 |
+
"learning_rate": 1.3669614872447942e-05,
|
| 1527 |
+
"loss": 0.0319,
|
| 1528 |
+
"step": 125000
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.19,
|
| 1532 |
+
"learning_rate": 1.3524239833923557e-05,
|
| 1533 |
+
"loss": 0.0547,
|
| 1534 |
+
"step": 125500
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 2.2,
|
| 1538 |
+
"learning_rate": 1.337886479539917e-05,
|
| 1539 |
+
"loss": 0.0347,
|
| 1540 |
+
"step": 126000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 2.21,
|
| 1544 |
+
"learning_rate": 1.3233489756874787e-05,
|
| 1545 |
+
"loss": 0.0329,
|
| 1546 |
+
"step": 126500
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 2.22,
|
| 1550 |
+
"learning_rate": 1.30881147183504e-05,
|
| 1551 |
+
"loss": 0.0381,
|
| 1552 |
+
"step": 127000
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 2.22,
|
| 1556 |
+
"learning_rate": 1.2942739679826017e-05,
|
| 1557 |
+
"loss": 0.0407,
|
| 1558 |
+
"step": 127500
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 2.23,
|
| 1562 |
+
"learning_rate": 1.279765539137868e-05,
|
| 1563 |
+
"loss": 0.0441,
|
| 1564 |
+
"step": 128000
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.24,
|
| 1568 |
+
"learning_rate": 1.2652280352854296e-05,
|
| 1569 |
+
"loss": 0.0412,
|
| 1570 |
+
"step": 128500
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.25,
|
| 1574 |
+
"learning_rate": 1.250690531432991e-05,
|
| 1575 |
+
"loss": 0.0422,
|
| 1576 |
+
"step": 129000
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 2.26,
|
| 1580 |
+
"learning_rate": 1.2361530275805522e-05,
|
| 1581 |
+
"loss": 0.0308,
|
| 1582 |
+
"step": 129500
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 2.27,
|
| 1586 |
+
"learning_rate": 1.2216155237281139e-05,
|
| 1587 |
+
"loss": 0.0404,
|
| 1588 |
+
"step": 130000
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.28,
|
| 1592 |
+
"learning_rate": 1.2071070948833802e-05,
|
| 1593 |
+
"loss": 0.041,
|
| 1594 |
+
"step": 130500
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 2.29,
|
| 1598 |
+
"learning_rate": 1.1925695910309417e-05,
|
| 1599 |
+
"loss": 0.0331,
|
| 1600 |
+
"step": 131000
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 2.29,
|
| 1604 |
+
"learning_rate": 1.1780320871785032e-05,
|
| 1605 |
+
"loss": 0.041,
|
| 1606 |
+
"step": 131500
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.3,
|
| 1610 |
+
"learning_rate": 1.1634945833260646e-05,
|
| 1611 |
+
"loss": 0.0393,
|
| 1612 |
+
"step": 132000
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.31,
|
| 1616 |
+
"learning_rate": 1.1489570794736261e-05,
|
| 1617 |
+
"loss": 0.036,
|
| 1618 |
+
"step": 132500
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 2.32,
|
| 1622 |
+
"learning_rate": 1.1344486506288924e-05,
|
| 1623 |
+
"loss": 0.036,
|
| 1624 |
+
"step": 133000
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 2.33,
|
| 1628 |
+
"learning_rate": 1.1199111467764539e-05,
|
| 1629 |
+
"loss": 0.0406,
|
| 1630 |
+
"step": 133500
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 2.34,
|
| 1634 |
+
"learning_rate": 1.1053736429240154e-05,
|
| 1635 |
+
"loss": 0.0421,
|
| 1636 |
+
"step": 134000
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 2.35,
|
| 1640 |
+
"learning_rate": 1.090836139071577e-05,
|
| 1641 |
+
"loss": 0.0345,
|
| 1642 |
+
"step": 134500
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 2.36,
|
| 1646 |
+
"learning_rate": 1.0762986352191383e-05,
|
| 1647 |
+
"loss": 0.0489,
|
| 1648 |
+
"step": 135000
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.36,
|
| 1652 |
+
"learning_rate": 1.0617611313666998e-05,
|
| 1653 |
+
"loss": 0.0298,
|
| 1654 |
+
"step": 135500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.37,
|
| 1658 |
+
"learning_rate": 1.0472236275142613e-05,
|
| 1659 |
+
"loss": 0.0339,
|
| 1660 |
+
"step": 136000
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 2.38,
|
| 1664 |
+
"learning_rate": 1.0327151986695278e-05,
|
| 1665 |
+
"loss": 0.0362,
|
| 1666 |
+
"step": 136500
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 2.39,
|
| 1670 |
+
"learning_rate": 1.0181776948170891e-05,
|
| 1671 |
+
"loss": 0.0298,
|
| 1672 |
+
"step": 137000
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 2.4,
|
| 1676 |
+
"learning_rate": 1.0036401909646506e-05,
|
| 1677 |
+
"loss": 0.0433,
|
| 1678 |
+
"step": 137500
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 2.41,
|
| 1682 |
+
"learning_rate": 9.89102687112212e-06,
|
| 1683 |
+
"loss": 0.0362,
|
| 1684 |
+
"step": 138000
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 2.42,
|
| 1688 |
+
"learning_rate": 9.745651832597737e-06,
|
| 1689 |
+
"loss": 0.0365,
|
| 1690 |
+
"step": 138500
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 2.42,
|
| 1694 |
+
"learning_rate": 9.600276794073352e-06,
|
| 1695 |
+
"loss": 0.0404,
|
| 1696 |
+
"step": 139000
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.43,
|
| 1700 |
+
"learning_rate": 9.455192505626015e-06,
|
| 1701 |
+
"loss": 0.039,
|
| 1702 |
+
"step": 139500
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 2.44,
|
| 1706 |
+
"learning_rate": 9.30981746710163e-06,
|
| 1707 |
+
"loss": 0.0361,
|
| 1708 |
+
"step": 140000
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 2.45,
|
| 1712 |
+
"learning_rate": 9.164442428577245e-06,
|
| 1713 |
+
"loss": 0.0297,
|
| 1714 |
+
"step": 140500
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 2.46,
|
| 1718 |
+
"learning_rate": 9.01906739005286e-06,
|
| 1719 |
+
"loss": 0.0405,
|
| 1720 |
+
"step": 141000
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 2.47,
|
| 1724 |
+
"learning_rate": 8.873692351528472e-06,
|
| 1725 |
+
"loss": 0.0355,
|
| 1726 |
+
"step": 141500
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 2.48,
|
| 1730 |
+
"learning_rate": 8.728608063081137e-06,
|
| 1731 |
+
"loss": 0.0361,
|
| 1732 |
+
"step": 142000
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.49,
|
| 1736 |
+
"learning_rate": 8.583233024556752e-06,
|
| 1737 |
+
"loss": 0.0327,
|
| 1738 |
+
"step": 142500
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.49,
|
| 1742 |
+
"learning_rate": 8.437857986032367e-06,
|
| 1743 |
+
"loss": 0.0285,
|
| 1744 |
+
"step": 143000
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 2.5,
|
| 1748 |
+
"learning_rate": 8.292482947507982e-06,
|
| 1749 |
+
"loss": 0.0315,
|
| 1750 |
+
"step": 143500
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 2.51,
|
| 1754 |
+
"learning_rate": 8.147398659060646e-06,
|
| 1755 |
+
"loss": 0.0401,
|
| 1756 |
+
"step": 144000
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 2.52,
|
| 1760 |
+
"learning_rate": 8.00202362053626e-06,
|
| 1761 |
+
"loss": 0.0356,
|
| 1762 |
+
"step": 144500
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 2.53,
|
| 1766 |
+
"learning_rate": 7.856648582011874e-06,
|
| 1767 |
+
"loss": 0.0257,
|
| 1768 |
+
"step": 145000
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 2.54,
|
| 1772 |
+
"learning_rate": 7.711273543487489e-06,
|
| 1773 |
+
"loss": 0.0346,
|
| 1774 |
+
"step": 145500
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.55,
|
| 1778 |
+
"learning_rate": 7.566189255040153e-06,
|
| 1779 |
+
"loss": 0.0326,
|
| 1780 |
+
"step": 146000
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.56,
|
| 1784 |
+
"learning_rate": 7.420814216515768e-06,
|
| 1785 |
+
"loss": 0.0334,
|
| 1786 |
+
"step": 146500
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 2.56,
|
| 1790 |
+
"learning_rate": 7.27602067814548e-06,
|
| 1791 |
+
"loss": 0.0309,
|
| 1792 |
+
"step": 147000
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 2.57,
|
| 1796 |
+
"learning_rate": 7.1306456396210946e-06,
|
| 1797 |
+
"loss": 0.0439,
|
| 1798 |
+
"step": 147500
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 2.58,
|
| 1802 |
+
"learning_rate": 6.985270601096709e-06,
|
| 1803 |
+
"loss": 0.036,
|
| 1804 |
+
"step": 148000
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 2.59,
|
| 1808 |
+
"learning_rate": 6.839895562572323e-06,
|
| 1809 |
+
"loss": 0.0454,
|
| 1810 |
+
"step": 148500
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 2.6,
|
| 1814 |
+
"learning_rate": 6.69452052404794e-06,
|
| 1815 |
+
"loss": 0.0307,
|
| 1816 |
+
"step": 149000
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.61,
|
| 1820 |
+
"learning_rate": 6.549145485523554e-06,
|
| 1821 |
+
"loss": 0.031,
|
| 1822 |
+
"step": 149500
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.62,
|
| 1826 |
+
"learning_rate": 6.403770446999169e-06,
|
| 1827 |
+
"loss": 0.0357,
|
| 1828 |
+
"step": 150000
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 2.63,
|
| 1832 |
+
"learning_rate": 6.2583954084747835e-06,
|
| 1833 |
+
"loss": 0.0262,
|
| 1834 |
+
"step": 150500
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 2.63,
|
| 1838 |
+
"learning_rate": 6.113020369950398e-06,
|
| 1839 |
+
"loss": 0.0296,
|
| 1840 |
+
"step": 151000
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 2.64,
|
| 1844 |
+
"learning_rate": 5.967645331426013e-06,
|
| 1845 |
+
"loss": 0.0312,
|
| 1846 |
+
"step": 151500
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 2.65,
|
| 1850 |
+
"learning_rate": 5.822270292901628e-06,
|
| 1851 |
+
"loss": 0.0313,
|
| 1852 |
+
"step": 152000
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 2.66,
|
| 1856 |
+
"learning_rate": 5.676895254377243e-06,
|
| 1857 |
+
"loss": 0.0237,
|
| 1858 |
+
"step": 152500
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.67,
|
| 1862 |
+
"learning_rate": 5.531520215852858e-06,
|
| 1863 |
+
"loss": 0.0208,
|
| 1864 |
+
"step": 153000
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.68,
|
| 1868 |
+
"learning_rate": 5.3861451773284725e-06,
|
| 1869 |
+
"loss": 0.0312,
|
| 1870 |
+
"step": 153500
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 2.69,
|
| 1874 |
+
"learning_rate": 5.240770138804087e-06,
|
| 1875 |
+
"loss": 0.0354,
|
| 1876 |
+
"step": 154000
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 2.7,
|
| 1880 |
+
"learning_rate": 5.095395100279702e-06,
|
| 1881 |
+
"loss": 0.0294,
|
| 1882 |
+
"step": 154500
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 2.7,
|
| 1886 |
+
"learning_rate": 4.950020061755316e-06,
|
| 1887 |
+
"loss": 0.0222,
|
| 1888 |
+
"step": 155000
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 2.71,
|
| 1892 |
+
"learning_rate": 4.804645023230931e-06,
|
| 1893 |
+
"loss": 0.0279,
|
| 1894 |
+
"step": 155500
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 2.72,
|
| 1898 |
+
"learning_rate": 4.6592699847065466e-06,
|
| 1899 |
+
"loss": 0.0437,
|
| 1900 |
+
"step": 156000
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.73,
|
| 1904 |
+
"learning_rate": 4.5141856962592096e-06,
|
| 1905 |
+
"loss": 0.0317,
|
| 1906 |
+
"step": 156500
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.74,
|
| 1910 |
+
"learning_rate": 4.368810657734824e-06,
|
| 1911 |
+
"loss": 0.0286,
|
| 1912 |
+
"step": 157000
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 2.75,
|
| 1916 |
+
"learning_rate": 4.223435619210439e-06,
|
| 1917 |
+
"loss": 0.0334,
|
| 1918 |
+
"step": 157500
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 2.76,
|
| 1922 |
+
"learning_rate": 4.078060580686054e-06,
|
| 1923 |
+
"loss": 0.0379,
|
| 1924 |
+
"step": 158000
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 2.77,
|
| 1928 |
+
"learning_rate": 3.932976292238718e-06,
|
| 1929 |
+
"loss": 0.0279,
|
| 1930 |
+
"step": 158500
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 2.77,
|
| 1934 |
+
"learning_rate": 3.7876012537143327e-06,
|
| 1935 |
+
"loss": 0.0385,
|
| 1936 |
+
"step": 159000
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 2.78,
|
| 1940 |
+
"learning_rate": 3.642226215189947e-06,
|
| 1941 |
+
"loss": 0.0287,
|
| 1942 |
+
"step": 159500
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.79,
|
| 1946 |
+
"learning_rate": 3.4968511766655615e-06,
|
| 1947 |
+
"loss": 0.0275,
|
| 1948 |
+
"step": 160000
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.8,
|
| 1952 |
+
"learning_rate": 3.3514761381411767e-06,
|
| 1953 |
+
"loss": 0.0321,
|
| 1954 |
+
"step": 160500
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 2.81,
|
| 1958 |
+
"learning_rate": 3.2061010996167916e-06,
|
| 1959 |
+
"loss": 0.0274,
|
| 1960 |
+
"step": 161000
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 2.82,
|
| 1964 |
+
"learning_rate": 3.0607260610924064e-06,
|
| 1965 |
+
"loss": 0.0357,
|
| 1966 |
+
"step": 161500
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 2.83,
|
| 1970 |
+
"learning_rate": 2.9153510225680212e-06,
|
| 1971 |
+
"loss": 0.0304,
|
| 1972 |
+
"step": 162000
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 2.83,
|
| 1976 |
+
"learning_rate": 2.7699759840436356e-06,
|
| 1977 |
+
"loss": 0.0209,
|
| 1978 |
+
"step": 162500
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.84,
|
| 1982 |
+
"learning_rate": 2.624600945519251e-06,
|
| 1983 |
+
"loss": 0.0216,
|
| 1984 |
+
"step": 163000
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.85,
|
| 1988 |
+
"learning_rate": 2.4792259069948657e-06,
|
| 1989 |
+
"loss": 0.0288,
|
| 1990 |
+
"step": 163500
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.86,
|
| 1994 |
+
"learning_rate": 2.33385086847048e-06,
|
| 1995 |
+
"loss": 0.0258,
|
| 1996 |
+
"step": 164000
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 2.87,
|
| 2000 |
+
"learning_rate": 2.1890573301001925e-06,
|
| 2001 |
+
"loss": 0.0294,
|
| 2002 |
+
"step": 164500
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 2.88,
|
| 2006 |
+
"learning_rate": 2.0436822915758073e-06,
|
| 2007 |
+
"loss": 0.0287,
|
| 2008 |
+
"step": 165000
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 2.89,
|
| 2012 |
+
"learning_rate": 1.898307253051422e-06,
|
| 2013 |
+
"loss": 0.0262,
|
| 2014 |
+
"step": 165500
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 2.9,
|
| 2018 |
+
"learning_rate": 1.7532229646040858e-06,
|
| 2019 |
+
"loss": 0.0359,
|
| 2020 |
+
"step": 166000
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 2.9,
|
| 2024 |
+
"learning_rate": 1.6078479260797006e-06,
|
| 2025 |
+
"loss": 0.0255,
|
| 2026 |
+
"step": 166500
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.91,
|
| 2030 |
+
"learning_rate": 1.4624728875553154e-06,
|
| 2031 |
+
"loss": 0.0241,
|
| 2032 |
+
"step": 167000
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.92,
|
| 2036 |
+
"learning_rate": 1.3173885991079788e-06,
|
| 2037 |
+
"loss": 0.025,
|
| 2038 |
+
"step": 167500
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 2.93,
|
| 2042 |
+
"learning_rate": 1.1720135605835934e-06,
|
| 2043 |
+
"loss": 0.0262,
|
| 2044 |
+
"step": 168000
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 2.94,
|
| 2048 |
+
"learning_rate": 1.0266385220592085e-06,
|
| 2049 |
+
"loss": 0.0251,
|
| 2050 |
+
"step": 168500
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 2.95,
|
| 2054 |
+
"learning_rate": 8.812634835348232e-07,
|
| 2055 |
+
"loss": 0.0264,
|
| 2056 |
+
"step": 169000
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 2.96,
|
| 2060 |
+
"learning_rate": 7.35888445010438e-07,
|
| 2061 |
+
"loss": 0.0284,
|
| 2062 |
+
"step": 169500
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 2.97,
|
| 2066 |
+
"learning_rate": 5.905134064860527e-07,
|
| 2067 |
+
"loss": 0.0246,
|
| 2068 |
+
"step": 170000
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.97,
|
| 2072 |
+
"learning_rate": 4.4513836796166756e-07,
|
| 2073 |
+
"loss": 0.0244,
|
| 2074 |
+
"step": 170500
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.98,
|
| 2078 |
+
"learning_rate": 2.9976332943728233e-07,
|
| 2079 |
+
"loss": 0.0306,
|
| 2080 |
+
"step": 171000
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 2.99,
|
| 2084 |
+
"learning_rate": 1.543882909128971e-07,
|
| 2085 |
+
"loss": 0.0237,
|
| 2086 |
+
"step": 171500
|
| 2087 |
+
}
|
| 2088 |
+
],
|
| 2089 |
+
"max_steps": 171969,
|
| 2090 |
+
"num_train_epochs": 3,
|
| 2091 |
+
"total_flos": 3.801898817534362e+16,
|
| 2092 |
+
"trial_name": null,
|
| 2093 |
+
"trial_params": null
|
| 2094 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7184488daa1a5e1e1b30fb4d60550e9662d09113a1848e7c4dda6665e818a9d3
|
| 3 |
+
size 3195
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|