Upload 6 files
Browse files- config.json +75 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- trainer_state.json +2828 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,75 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./exps/bart-large-chinese",
|
| 3 |
+
"activation_dropout": 0.1,
|
| 4 |
+
"activation_function": "gelu",
|
| 5 |
+
"add_bias_logits": false,
|
| 6 |
+
"add_final_layer_norm": false,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"BartForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.1,
|
| 11 |
+
"bos_token_id": 101,
|
| 12 |
+
"classif_dropout": 0.1,
|
| 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": 102,
|
| 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": 102,
|
| 27 |
+
"forced_eos_token_id": 102,
|
| 28 |
+
"gradient_checkpointing": false,
|
| 29 |
+
"id2label": {
|
| 30 |
+
"0": "LABEL_0",
|
| 31 |
+
"1": "LABEL_1",
|
| 32 |
+
"2": "LABEL_2"
|
| 33 |
+
},
|
| 34 |
+
"init_std": 0.02,
|
| 35 |
+
"is_encoder_decoder": true,
|
| 36 |
+
"label2id": {
|
| 37 |
+
"LABEL_0": 0,
|
| 38 |
+
"LABEL_1": 1,
|
| 39 |
+
"LABEL_2": 2
|
| 40 |
+
},
|
| 41 |
+
"max_length": 100,
|
| 42 |
+
"max_position_embeddings": 1024,
|
| 43 |
+
"model_type": "bart",
|
| 44 |
+
"no_repeat_ngram_size": 3,
|
| 45 |
+
"normalize_before": false,
|
| 46 |
+
"normalize_embedding": true,
|
| 47 |
+
"num_beams": 4,
|
| 48 |
+
"num_hidden_layers": 12,
|
| 49 |
+
"pad_token_id": 0,
|
| 50 |
+
"scale_embedding": false,
|
| 51 |
+
"task_specific_params": {
|
| 52 |
+
"summarization": {
|
| 53 |
+
"length_penalty": 1.0,
|
| 54 |
+
"max_length": 128,
|
| 55 |
+
"min_length": 12,
|
| 56 |
+
"num_beams": 4
|
| 57 |
+
},
|
| 58 |
+
"summarization_cnn": {
|
| 59 |
+
"length_penalty": 2.0,
|
| 60 |
+
"max_length": 142,
|
| 61 |
+
"min_length": 56,
|
| 62 |
+
"num_beams": 4
|
| 63 |
+
},
|
| 64 |
+
"summarization_xsum": {
|
| 65 |
+
"length_penalty": 1.0,
|
| 66 |
+
"max_length": 62,
|
| 67 |
+
"min_length": 11,
|
| 68 |
+
"num_beams": 6
|
| 69 |
+
}
|
| 70 |
+
},
|
| 71 |
+
"tokenizer_class": "BertTokenizer",
|
| 72 |
+
"transformers_version": "4.4.1",
|
| 73 |
+
"use_cache": true,
|
| 74 |
+
"vocab_size": 51271
|
| 75 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab39ac477f63cd35993b320d7102ce567e192801088feae4fe9fa698913c6acb
|
| 3 |
+
size 1629681921
|
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_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"do_lower_case": true, "do_basic_tokenize": true, "never_split": null, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "./exps/bart-large-chinese"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2828 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.999989076887786,
|
| 5 |
+
"global_step": 228870,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 2.993664525713287e-06,
|
| 13 |
+
"loss": 1.7501,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 2.987329051426574e-06,
|
| 19 |
+
"loss": 1.6451,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.07,
|
| 24 |
+
"learning_rate": 2.980993577139861e-06,
|
| 25 |
+
"loss": 1.6171,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.09,
|
| 30 |
+
"learning_rate": 2.974658102853148e-06,
|
| 31 |
+
"loss": 1.5995,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.11,
|
| 36 |
+
"learning_rate": 2.968322628566435e-06,
|
| 37 |
+
"loss": 1.5905,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.13,
|
| 42 |
+
"learning_rate": 2.9619871542797224e-06,
|
| 43 |
+
"loss": 1.5876,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15,
|
| 48 |
+
"learning_rate": 2.9556516799930094e-06,
|
| 49 |
+
"loss": 1.579,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 2.9493162057062964e-06,
|
| 55 |
+
"loss": 1.5714,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.2,
|
| 60 |
+
"learning_rate": 2.9429807314195834e-06,
|
| 61 |
+
"loss": 1.5668,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.22,
|
| 66 |
+
"learning_rate": 2.9366452571328704e-06,
|
| 67 |
+
"loss": 1.5624,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.24,
|
| 72 |
+
"learning_rate": 2.9303097828461574e-06,
|
| 73 |
+
"loss": 1.558,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.26,
|
| 78 |
+
"learning_rate": 2.9239743085594444e-06,
|
| 79 |
+
"loss": 1.5558,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.28,
|
| 84 |
+
"learning_rate": 2.9176388342727313e-06,
|
| 85 |
+
"loss": 1.5515,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.31,
|
| 90 |
+
"learning_rate": 2.9113033599860183e-06,
|
| 91 |
+
"loss": 1.5473,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.33,
|
| 96 |
+
"learning_rate": 2.9049678856993053e-06,
|
| 97 |
+
"loss": 1.5436,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.35,
|
| 102 |
+
"learning_rate": 2.8986324114125923e-06,
|
| 103 |
+
"loss": 1.5402,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.37,
|
| 108 |
+
"learning_rate": 2.8922969371258797e-06,
|
| 109 |
+
"loss": 1.539,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.39,
|
| 114 |
+
"learning_rate": 2.8859614628391667e-06,
|
| 115 |
+
"loss": 1.5364,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.42,
|
| 120 |
+
"learning_rate": 2.8796259885524533e-06,
|
| 121 |
+
"loss": 1.5345,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.44,
|
| 126 |
+
"learning_rate": 2.8732905142657403e-06,
|
| 127 |
+
"loss": 1.5297,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.46,
|
| 132 |
+
"learning_rate": 2.8669550399790273e-06,
|
| 133 |
+
"loss": 1.528,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.48,
|
| 138 |
+
"learning_rate": 2.8606195656923142e-06,
|
| 139 |
+
"loss": 1.5248,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.5,
|
| 144 |
+
"learning_rate": 2.8542840914056012e-06,
|
| 145 |
+
"loss": 1.5221,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.52,
|
| 150 |
+
"learning_rate": 2.8479486171188882e-06,
|
| 151 |
+
"loss": 1.5202,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.55,
|
| 156 |
+
"learning_rate": 2.8416131428321756e-06,
|
| 157 |
+
"loss": 1.5188,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.57,
|
| 162 |
+
"learning_rate": 2.8352776685454626e-06,
|
| 163 |
+
"loss": 1.5164,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.59,
|
| 168 |
+
"learning_rate": 2.8289421942587496e-06,
|
| 169 |
+
"loss": 1.5152,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.61,
|
| 174 |
+
"learning_rate": 2.8226067199720366e-06,
|
| 175 |
+
"loss": 1.5142,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.63,
|
| 180 |
+
"learning_rate": 2.8162712456853236e-06,
|
| 181 |
+
"loss": 1.5113,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.66,
|
| 186 |
+
"learning_rate": 2.8099357713986106e-06,
|
| 187 |
+
"loss": 1.5098,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.68,
|
| 192 |
+
"learning_rate": 2.8036002971118976e-06,
|
| 193 |
+
"loss": 1.508,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.7,
|
| 198 |
+
"learning_rate": 2.7972648228251845e-06,
|
| 199 |
+
"loss": 1.5063,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.72,
|
| 204 |
+
"learning_rate": 2.7909293485384715e-06,
|
| 205 |
+
"loss": 1.5051,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.74,
|
| 210 |
+
"learning_rate": 2.7845938742517585e-06,
|
| 211 |
+
"loss": 1.5044,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.76,
|
| 216 |
+
"learning_rate": 2.7782583999650455e-06,
|
| 217 |
+
"loss": 1.503,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.79,
|
| 222 |
+
"learning_rate": 2.771922925678333e-06,
|
| 223 |
+
"loss": 1.5013,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.81,
|
| 228 |
+
"learning_rate": 2.76558745139162e-06,
|
| 229 |
+
"loss": 1.5011,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.83,
|
| 234 |
+
"learning_rate": 2.759251977104907e-06,
|
| 235 |
+
"loss": 1.5004,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.85,
|
| 240 |
+
"learning_rate": 2.752916502818194e-06,
|
| 241 |
+
"loss": 1.4989,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.87,
|
| 246 |
+
"learning_rate": 2.746581028531481e-06,
|
| 247 |
+
"loss": 1.498,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.9,
|
| 252 |
+
"learning_rate": 2.740245554244768e-06,
|
| 253 |
+
"loss": 1.4968,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.92,
|
| 258 |
+
"learning_rate": 2.733910079958055e-06,
|
| 259 |
+
"loss": 1.4963,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.94,
|
| 264 |
+
"learning_rate": 2.727574605671342e-06,
|
| 265 |
+
"loss": 1.4961,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.96,
|
| 270 |
+
"learning_rate": 2.721239131384629e-06,
|
| 271 |
+
"loss": 1.4944,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.98,
|
| 276 |
+
"learning_rate": 2.714903657097916e-06,
|
| 277 |
+
"loss": 1.4936,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.0,
|
| 282 |
+
"eval_loss": 1.4786614179611206,
|
| 283 |
+
"eval_runtime": 11.8277,
|
| 284 |
+
"eval_samples_per_second": 83.364,
|
| 285 |
+
"step": 22887
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.0,
|
| 289 |
+
"learning_rate": 2.7085681828112032e-06,
|
| 290 |
+
"loss": 1.4936,
|
| 291 |
+
"step": 23000
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.03,
|
| 295 |
+
"learning_rate": 2.7022327085244902e-06,
|
| 296 |
+
"loss": 1.4907,
|
| 297 |
+
"step": 23500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.05,
|
| 301 |
+
"learning_rate": 2.695897234237777e-06,
|
| 302 |
+
"loss": 1.4898,
|
| 303 |
+
"step": 24000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.07,
|
| 307 |
+
"learning_rate": 2.689561759951064e-06,
|
| 308 |
+
"loss": 1.4892,
|
| 309 |
+
"step": 24500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.09,
|
| 313 |
+
"learning_rate": 2.683226285664351e-06,
|
| 314 |
+
"loss": 1.488,
|
| 315 |
+
"step": 25000
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.11,
|
| 319 |
+
"learning_rate": 2.6768908113776377e-06,
|
| 320 |
+
"loss": 1.4877,
|
| 321 |
+
"step": 25500
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.14,
|
| 325 |
+
"learning_rate": 2.6705553370909247e-06,
|
| 326 |
+
"loss": 1.4874,
|
| 327 |
+
"step": 26000
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.16,
|
| 331 |
+
"learning_rate": 2.6642198628042117e-06,
|
| 332 |
+
"loss": 1.4866,
|
| 333 |
+
"step": 26500
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.18,
|
| 337 |
+
"learning_rate": 2.6578843885174987e-06,
|
| 338 |
+
"loss": 1.4856,
|
| 339 |
+
"step": 27000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.2,
|
| 343 |
+
"learning_rate": 2.651548914230786e-06,
|
| 344 |
+
"loss": 1.4851,
|
| 345 |
+
"step": 27500
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.22,
|
| 349 |
+
"learning_rate": 2.645213439944073e-06,
|
| 350 |
+
"loss": 1.4848,
|
| 351 |
+
"step": 28000
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.25,
|
| 355 |
+
"learning_rate": 2.63887796565736e-06,
|
| 356 |
+
"loss": 1.4843,
|
| 357 |
+
"step": 28500
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.27,
|
| 361 |
+
"learning_rate": 2.632542491370647e-06,
|
| 362 |
+
"loss": 1.4842,
|
| 363 |
+
"step": 29000
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.29,
|
| 367 |
+
"learning_rate": 2.626207017083934e-06,
|
| 368 |
+
"loss": 1.4836,
|
| 369 |
+
"step": 29500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.31,
|
| 373 |
+
"learning_rate": 2.619871542797221e-06,
|
| 374 |
+
"loss": 1.4829,
|
| 375 |
+
"step": 30000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.33,
|
| 379 |
+
"learning_rate": 2.613536068510508e-06,
|
| 380 |
+
"loss": 1.4823,
|
| 381 |
+
"step": 30500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.35,
|
| 385 |
+
"learning_rate": 2.607200594223795e-06,
|
| 386 |
+
"loss": 1.482,
|
| 387 |
+
"step": 31000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.38,
|
| 391 |
+
"learning_rate": 2.600865119937082e-06,
|
| 392 |
+
"loss": 1.4813,
|
| 393 |
+
"step": 31500
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.4,
|
| 397 |
+
"learning_rate": 2.594529645650369e-06,
|
| 398 |
+
"loss": 1.4812,
|
| 399 |
+
"step": 32000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.42,
|
| 403 |
+
"learning_rate": 2.5881941713636564e-06,
|
| 404 |
+
"loss": 1.4808,
|
| 405 |
+
"step": 32500
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.44,
|
| 409 |
+
"learning_rate": 2.5818586970769434e-06,
|
| 410 |
+
"loss": 1.48,
|
| 411 |
+
"step": 33000
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.46,
|
| 415 |
+
"learning_rate": 2.5755232227902304e-06,
|
| 416 |
+
"loss": 1.4797,
|
| 417 |
+
"step": 33500
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.49,
|
| 421 |
+
"learning_rate": 2.5691877485035174e-06,
|
| 422 |
+
"loss": 1.4788,
|
| 423 |
+
"step": 34000
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.51,
|
| 427 |
+
"learning_rate": 2.5628522742168044e-06,
|
| 428 |
+
"loss": 1.4792,
|
| 429 |
+
"step": 34500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.53,
|
| 433 |
+
"learning_rate": 2.5565167999300914e-06,
|
| 434 |
+
"loss": 1.4783,
|
| 435 |
+
"step": 35000
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.55,
|
| 439 |
+
"learning_rate": 2.5501813256433784e-06,
|
| 440 |
+
"loss": 1.4772,
|
| 441 |
+
"step": 35500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.57,
|
| 445 |
+
"learning_rate": 2.5438458513566653e-06,
|
| 446 |
+
"loss": 1.4778,
|
| 447 |
+
"step": 36000
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.59,
|
| 451 |
+
"learning_rate": 2.5375103770699523e-06,
|
| 452 |
+
"loss": 1.4772,
|
| 453 |
+
"step": 36500
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.62,
|
| 457 |
+
"learning_rate": 2.5311749027832393e-06,
|
| 458 |
+
"loss": 1.4767,
|
| 459 |
+
"step": 37000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.64,
|
| 463 |
+
"learning_rate": 2.5248394284965263e-06,
|
| 464 |
+
"loss": 1.4764,
|
| 465 |
+
"step": 37500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.66,
|
| 469 |
+
"learning_rate": 2.5185039542098137e-06,
|
| 470 |
+
"loss": 1.4758,
|
| 471 |
+
"step": 38000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.68,
|
| 475 |
+
"learning_rate": 2.5121684799231007e-06,
|
| 476 |
+
"loss": 1.4755,
|
| 477 |
+
"step": 38500
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.7,
|
| 481 |
+
"learning_rate": 2.5058330056363877e-06,
|
| 482 |
+
"loss": 1.4754,
|
| 483 |
+
"step": 39000
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.73,
|
| 487 |
+
"learning_rate": 2.4994975313496747e-06,
|
| 488 |
+
"loss": 1.4745,
|
| 489 |
+
"step": 39500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.75,
|
| 493 |
+
"learning_rate": 2.4931620570629617e-06,
|
| 494 |
+
"loss": 1.4747,
|
| 495 |
+
"step": 40000
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.77,
|
| 499 |
+
"learning_rate": 2.4868265827762487e-06,
|
| 500 |
+
"loss": 1.4741,
|
| 501 |
+
"step": 40500
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.79,
|
| 505 |
+
"learning_rate": 2.4804911084895356e-06,
|
| 506 |
+
"loss": 1.4737,
|
| 507 |
+
"step": 41000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.81,
|
| 511 |
+
"learning_rate": 2.4741556342028226e-06,
|
| 512 |
+
"loss": 1.4734,
|
| 513 |
+
"step": 41500
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.84,
|
| 517 |
+
"learning_rate": 2.4678201599161096e-06,
|
| 518 |
+
"loss": 1.4731,
|
| 519 |
+
"step": 42000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.86,
|
| 523 |
+
"learning_rate": 2.4614846856293966e-06,
|
| 524 |
+
"loss": 1.4732,
|
| 525 |
+
"step": 42500
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.88,
|
| 529 |
+
"learning_rate": 2.4551492113426836e-06,
|
| 530 |
+
"loss": 1.472,
|
| 531 |
+
"step": 43000
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.9,
|
| 535 |
+
"learning_rate": 2.448813737055971e-06,
|
| 536 |
+
"loss": 1.4717,
|
| 537 |
+
"step": 43500
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.92,
|
| 541 |
+
"learning_rate": 2.442478262769258e-06,
|
| 542 |
+
"loss": 1.4715,
|
| 543 |
+
"step": 44000
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.94,
|
| 547 |
+
"learning_rate": 2.436142788482545e-06,
|
| 548 |
+
"loss": 1.4712,
|
| 549 |
+
"step": 44500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.97,
|
| 553 |
+
"learning_rate": 2.429807314195832e-06,
|
| 554 |
+
"loss": 1.471,
|
| 555 |
+
"step": 45000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.99,
|
| 559 |
+
"learning_rate": 2.423471839909119e-06,
|
| 560 |
+
"loss": 1.4712,
|
| 561 |
+
"step": 45500
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.0,
|
| 565 |
+
"eval_loss": 1.4626506567001343,
|
| 566 |
+
"eval_runtime": 11.7267,
|
| 567 |
+
"eval_samples_per_second": 84.081,
|
| 568 |
+
"step": 45774
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 2.01,
|
| 572 |
+
"learning_rate": 2.417136365622406e-06,
|
| 573 |
+
"loss": 1.4703,
|
| 574 |
+
"step": 46000
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 2.03,
|
| 578 |
+
"learning_rate": 2.410800891335693e-06,
|
| 579 |
+
"loss": 1.4686,
|
| 580 |
+
"step": 46500
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.05,
|
| 584 |
+
"learning_rate": 2.4044654170489795e-06,
|
| 585 |
+
"loss": 1.4685,
|
| 586 |
+
"step": 47000
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 2.08,
|
| 590 |
+
"learning_rate": 2.398129942762267e-06,
|
| 591 |
+
"loss": 1.468,
|
| 592 |
+
"step": 47500
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.1,
|
| 596 |
+
"learning_rate": 2.391794468475554e-06,
|
| 597 |
+
"loss": 1.4677,
|
| 598 |
+
"step": 48000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 2.12,
|
| 602 |
+
"learning_rate": 2.385458994188841e-06,
|
| 603 |
+
"loss": 1.468,
|
| 604 |
+
"step": 48500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.14,
|
| 608 |
+
"learning_rate": 2.379123519902128e-06,
|
| 609 |
+
"loss": 1.467,
|
| 610 |
+
"step": 49000
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 2.16,
|
| 614 |
+
"learning_rate": 2.372788045615415e-06,
|
| 615 |
+
"loss": 1.4672,
|
| 616 |
+
"step": 49500
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 2.18,
|
| 620 |
+
"learning_rate": 2.366452571328702e-06,
|
| 621 |
+
"loss": 1.4672,
|
| 622 |
+
"step": 50000
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.21,
|
| 626 |
+
"learning_rate": 2.360117097041989e-06,
|
| 627 |
+
"loss": 1.4668,
|
| 628 |
+
"step": 50500
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.23,
|
| 632 |
+
"learning_rate": 2.353781622755276e-06,
|
| 633 |
+
"loss": 1.4665,
|
| 634 |
+
"step": 51000
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 2.25,
|
| 638 |
+
"learning_rate": 2.347446148468563e-06,
|
| 639 |
+
"loss": 1.4661,
|
| 640 |
+
"step": 51500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 2.27,
|
| 644 |
+
"learning_rate": 2.34111067418185e-06,
|
| 645 |
+
"loss": 1.4658,
|
| 646 |
+
"step": 52000
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.29,
|
| 650 |
+
"learning_rate": 2.334775199895137e-06,
|
| 651 |
+
"loss": 1.4658,
|
| 652 |
+
"step": 52500
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 2.32,
|
| 656 |
+
"learning_rate": 2.328439725608424e-06,
|
| 657 |
+
"loss": 1.4657,
|
| 658 |
+
"step": 53000
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 2.34,
|
| 662 |
+
"learning_rate": 2.322104251321711e-06,
|
| 663 |
+
"loss": 1.4659,
|
| 664 |
+
"step": 53500
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.36,
|
| 668 |
+
"learning_rate": 2.315768777034998e-06,
|
| 669 |
+
"loss": 1.4653,
|
| 670 |
+
"step": 54000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 2.38,
|
| 674 |
+
"learning_rate": 2.309433302748285e-06,
|
| 675 |
+
"loss": 1.4654,
|
| 676 |
+
"step": 54500
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 2.4,
|
| 680 |
+
"learning_rate": 2.303097828461572e-06,
|
| 681 |
+
"loss": 1.4649,
|
| 682 |
+
"step": 55000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 2.42,
|
| 686 |
+
"learning_rate": 2.296762354174859e-06,
|
| 687 |
+
"loss": 1.4646,
|
| 688 |
+
"step": 55500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.45,
|
| 692 |
+
"learning_rate": 2.290426879888146e-06,
|
| 693 |
+
"loss": 1.4646,
|
| 694 |
+
"step": 56000
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 2.47,
|
| 698 |
+
"learning_rate": 2.284091405601433e-06,
|
| 699 |
+
"loss": 1.4639,
|
| 700 |
+
"step": 56500
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 2.49,
|
| 704 |
+
"learning_rate": 2.27775593131472e-06,
|
| 705 |
+
"loss": 1.464,
|
| 706 |
+
"step": 57000
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.51,
|
| 710 |
+
"learning_rate": 2.271420457028007e-06,
|
| 711 |
+
"loss": 1.4638,
|
| 712 |
+
"step": 57500
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 2.53,
|
| 716 |
+
"learning_rate": 2.265084982741294e-06,
|
| 717 |
+
"loss": 1.4634,
|
| 718 |
+
"step": 58000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 2.56,
|
| 722 |
+
"learning_rate": 2.258749508454581e-06,
|
| 723 |
+
"loss": 1.4629,
|
| 724 |
+
"step": 58500
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 2.58,
|
| 728 |
+
"learning_rate": 2.252414034167868e-06,
|
| 729 |
+
"loss": 1.4629,
|
| 730 |
+
"step": 59000
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.6,
|
| 734 |
+
"learning_rate": 2.246078559881155e-06,
|
| 735 |
+
"loss": 1.4627,
|
| 736 |
+
"step": 59500
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.62,
|
| 740 |
+
"learning_rate": 2.239743085594442e-06,
|
| 741 |
+
"loss": 1.463,
|
| 742 |
+
"step": 60000
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 2.64,
|
| 746 |
+
"learning_rate": 2.233407611307729e-06,
|
| 747 |
+
"loss": 1.4625,
|
| 748 |
+
"step": 60500
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.67,
|
| 752 |
+
"learning_rate": 2.227072137021016e-06,
|
| 753 |
+
"loss": 1.4622,
|
| 754 |
+
"step": 61000
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 2.69,
|
| 758 |
+
"learning_rate": 2.220736662734303e-06,
|
| 759 |
+
"loss": 1.4621,
|
| 760 |
+
"step": 61500
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 2.71,
|
| 764 |
+
"learning_rate": 2.21440118844759e-06,
|
| 765 |
+
"loss": 1.4619,
|
| 766 |
+
"step": 62000
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.73,
|
| 770 |
+
"learning_rate": 2.2080657141608774e-06,
|
| 771 |
+
"loss": 1.4618,
|
| 772 |
+
"step": 62500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.75,
|
| 776 |
+
"learning_rate": 2.2017302398741644e-06,
|
| 777 |
+
"loss": 1.4614,
|
| 778 |
+
"step": 63000
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 2.77,
|
| 782 |
+
"learning_rate": 2.1953947655874514e-06,
|
| 783 |
+
"loss": 1.4615,
|
| 784 |
+
"step": 63500
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 2.8,
|
| 788 |
+
"learning_rate": 2.1890592913007384e-06,
|
| 789 |
+
"loss": 1.4609,
|
| 790 |
+
"step": 64000
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.82,
|
| 794 |
+
"learning_rate": 2.1827238170140254e-06,
|
| 795 |
+
"loss": 1.4609,
|
| 796 |
+
"step": 64500
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 2.84,
|
| 800 |
+
"learning_rate": 2.1763883427273123e-06,
|
| 801 |
+
"loss": 1.461,
|
| 802 |
+
"step": 65000
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.86,
|
| 806 |
+
"learning_rate": 2.1700528684405993e-06,
|
| 807 |
+
"loss": 1.4609,
|
| 808 |
+
"step": 65500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.88,
|
| 812 |
+
"learning_rate": 2.1637173941538863e-06,
|
| 813 |
+
"loss": 1.4608,
|
| 814 |
+
"step": 66000
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.91,
|
| 818 |
+
"learning_rate": 2.1573819198671733e-06,
|
| 819 |
+
"loss": 1.4603,
|
| 820 |
+
"step": 66500
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 2.93,
|
| 824 |
+
"learning_rate": 2.1510464455804603e-06,
|
| 825 |
+
"loss": 1.4604,
|
| 826 |
+
"step": 67000
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 2.95,
|
| 830 |
+
"learning_rate": 2.1447109712937477e-06,
|
| 831 |
+
"loss": 1.4602,
|
| 832 |
+
"step": 67500
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.97,
|
| 836 |
+
"learning_rate": 2.1383754970070347e-06,
|
| 837 |
+
"loss": 1.4604,
|
| 838 |
+
"step": 68000
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.99,
|
| 842 |
+
"learning_rate": 2.1320400227203217e-06,
|
| 843 |
+
"loss": 1.4597,
|
| 844 |
+
"step": 68500
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 3.0,
|
| 848 |
+
"eval_loss": 1.4553998708724976,
|
| 849 |
+
"eval_runtime": 12.0973,
|
| 850 |
+
"eval_samples_per_second": 81.506,
|
| 851 |
+
"step": 68661
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 3.01,
|
| 855 |
+
"learning_rate": 2.1257045484336087e-06,
|
| 856 |
+
"loss": 1.4592,
|
| 857 |
+
"step": 69000
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 3.04,
|
| 861 |
+
"learning_rate": 2.1193690741468957e-06,
|
| 862 |
+
"loss": 1.458,
|
| 863 |
+
"step": 69500
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.06,
|
| 867 |
+
"learning_rate": 2.1130335998601827e-06,
|
| 868 |
+
"loss": 1.4579,
|
| 869 |
+
"step": 70000
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.08,
|
| 873 |
+
"learning_rate": 2.1066981255734696e-06,
|
| 874 |
+
"loss": 1.4581,
|
| 875 |
+
"step": 70500
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 3.1,
|
| 879 |
+
"learning_rate": 2.1003626512867566e-06,
|
| 880 |
+
"loss": 1.4579,
|
| 881 |
+
"step": 71000
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 3.12,
|
| 885 |
+
"learning_rate": 2.0940271770000436e-06,
|
| 886 |
+
"loss": 1.4576,
|
| 887 |
+
"step": 71500
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 3.15,
|
| 891 |
+
"learning_rate": 2.0876917027133306e-06,
|
| 892 |
+
"loss": 1.4573,
|
| 893 |
+
"step": 72000
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 3.17,
|
| 897 |
+
"learning_rate": 2.0813562284266176e-06,
|
| 898 |
+
"loss": 1.4574,
|
| 899 |
+
"step": 72500
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 3.19,
|
| 903 |
+
"learning_rate": 2.075020754139905e-06,
|
| 904 |
+
"loss": 1.457,
|
| 905 |
+
"step": 73000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.21,
|
| 909 |
+
"learning_rate": 2.068685279853192e-06,
|
| 910 |
+
"loss": 1.4571,
|
| 911 |
+
"step": 73500
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.23,
|
| 915 |
+
"learning_rate": 2.062349805566479e-06,
|
| 916 |
+
"loss": 1.4573,
|
| 917 |
+
"step": 74000
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 3.26,
|
| 921 |
+
"learning_rate": 2.056014331279766e-06,
|
| 922 |
+
"loss": 1.4569,
|
| 923 |
+
"step": 74500
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 3.28,
|
| 927 |
+
"learning_rate": 2.049678856993053e-06,
|
| 928 |
+
"loss": 1.4568,
|
| 929 |
+
"step": 75000
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.3,
|
| 933 |
+
"learning_rate": 2.04334338270634e-06,
|
| 934 |
+
"loss": 1.4569,
|
| 935 |
+
"step": 75500
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 3.32,
|
| 939 |
+
"learning_rate": 2.037007908419627e-06,
|
| 940 |
+
"loss": 1.4567,
|
| 941 |
+
"step": 76000
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 3.34,
|
| 945 |
+
"learning_rate": 2.030672434132914e-06,
|
| 946 |
+
"loss": 1.4568,
|
| 947 |
+
"step": 76500
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.36,
|
| 951 |
+
"learning_rate": 2.024336959846201e-06,
|
| 952 |
+
"loss": 1.4564,
|
| 953 |
+
"step": 77000
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.39,
|
| 957 |
+
"learning_rate": 2.018001485559488e-06,
|
| 958 |
+
"loss": 1.4559,
|
| 959 |
+
"step": 77500
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 3.41,
|
| 963 |
+
"learning_rate": 2.011666011272775e-06,
|
| 964 |
+
"loss": 1.4558,
|
| 965 |
+
"step": 78000
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 3.43,
|
| 969 |
+
"learning_rate": 2.0053305369860623e-06,
|
| 970 |
+
"loss": 1.4561,
|
| 971 |
+
"step": 78500
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.45,
|
| 975 |
+
"learning_rate": 1.9989950626993493e-06,
|
| 976 |
+
"loss": 1.4559,
|
| 977 |
+
"step": 79000
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 3.47,
|
| 981 |
+
"learning_rate": 1.9926595884126363e-06,
|
| 982 |
+
"loss": 1.4559,
|
| 983 |
+
"step": 79500
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 3.5,
|
| 987 |
+
"learning_rate": 1.9863241141259233e-06,
|
| 988 |
+
"loss": 1.4563,
|
| 989 |
+
"step": 80000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.52,
|
| 993 |
+
"learning_rate": 1.9799886398392103e-06,
|
| 994 |
+
"loss": 1.4555,
|
| 995 |
+
"step": 80500
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.54,
|
| 999 |
+
"learning_rate": 1.9736531655524972e-06,
|
| 1000 |
+
"loss": 1.4555,
|
| 1001 |
+
"step": 81000
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 3.56,
|
| 1005 |
+
"learning_rate": 1.9673176912657842e-06,
|
| 1006 |
+
"loss": 1.4553,
|
| 1007 |
+
"step": 81500
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 3.58,
|
| 1011 |
+
"learning_rate": 1.9609822169790712e-06,
|
| 1012 |
+
"loss": 1.4553,
|
| 1013 |
+
"step": 82000
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.6,
|
| 1017 |
+
"learning_rate": 1.9546467426923586e-06,
|
| 1018 |
+
"loss": 1.4548,
|
| 1019 |
+
"step": 82500
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 3.63,
|
| 1023 |
+
"learning_rate": 1.9483112684056456e-06,
|
| 1024 |
+
"loss": 1.4551,
|
| 1025 |
+
"step": 83000
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 3.65,
|
| 1029 |
+
"learning_rate": 1.9419757941189326e-06,
|
| 1030 |
+
"loss": 1.4551,
|
| 1031 |
+
"step": 83500
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.67,
|
| 1035 |
+
"learning_rate": 1.9356403198322196e-06,
|
| 1036 |
+
"loss": 1.4546,
|
| 1037 |
+
"step": 84000
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 3.69,
|
| 1041 |
+
"learning_rate": 1.929304845545506e-06,
|
| 1042 |
+
"loss": 1.4548,
|
| 1043 |
+
"step": 84500
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 3.71,
|
| 1047 |
+
"learning_rate": 1.922969371258793e-06,
|
| 1048 |
+
"loss": 1.4545,
|
| 1049 |
+
"step": 85000
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 3.74,
|
| 1053 |
+
"learning_rate": 1.91663389697208e-06,
|
| 1054 |
+
"loss": 1.4545,
|
| 1055 |
+
"step": 85500
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.76,
|
| 1059 |
+
"learning_rate": 1.910298422685367e-06,
|
| 1060 |
+
"loss": 1.4546,
|
| 1061 |
+
"step": 86000
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 3.78,
|
| 1065 |
+
"learning_rate": 1.9039629483986545e-06,
|
| 1066 |
+
"loss": 1.4544,
|
| 1067 |
+
"step": 86500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.8,
|
| 1071 |
+
"learning_rate": 1.8976274741119413e-06,
|
| 1072 |
+
"loss": 1.454,
|
| 1073 |
+
"step": 87000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.82,
|
| 1077 |
+
"learning_rate": 1.8912919998252285e-06,
|
| 1078 |
+
"loss": 1.4548,
|
| 1079 |
+
"step": 87500
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 3.84,
|
| 1083 |
+
"learning_rate": 1.8849565255385155e-06,
|
| 1084 |
+
"loss": 1.4542,
|
| 1085 |
+
"step": 88000
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 3.87,
|
| 1089 |
+
"learning_rate": 1.8786210512518025e-06,
|
| 1090 |
+
"loss": 1.4544,
|
| 1091 |
+
"step": 88500
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 3.89,
|
| 1095 |
+
"learning_rate": 1.8722855769650897e-06,
|
| 1096 |
+
"loss": 1.4543,
|
| 1097 |
+
"step": 89000
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.91,
|
| 1101 |
+
"learning_rate": 1.8659501026783765e-06,
|
| 1102 |
+
"loss": 1.4539,
|
| 1103 |
+
"step": 89500
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 3.93,
|
| 1107 |
+
"learning_rate": 1.8596146283916634e-06,
|
| 1108 |
+
"loss": 1.4538,
|
| 1109 |
+
"step": 90000
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 3.95,
|
| 1113 |
+
"learning_rate": 1.8532791541049502e-06,
|
| 1114 |
+
"loss": 1.4541,
|
| 1115 |
+
"step": 90500
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.98,
|
| 1119 |
+
"learning_rate": 1.8469436798182374e-06,
|
| 1120 |
+
"loss": 1.4535,
|
| 1121 |
+
"step": 91000
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 4.0,
|
| 1125 |
+
"learning_rate": 1.8406082055315244e-06,
|
| 1126 |
+
"loss": 1.4536,
|
| 1127 |
+
"step": 91500
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 4.0,
|
| 1131 |
+
"eval_loss": 1.4519952535629272,
|
| 1132 |
+
"eval_runtime": 11.774,
|
| 1133 |
+
"eval_samples_per_second": 83.744,
|
| 1134 |
+
"step": 91548
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 4.02,
|
| 1138 |
+
"learning_rate": 1.834272731244812e-06,
|
| 1139 |
+
"loss": 1.4534,
|
| 1140 |
+
"step": 92000
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 4.04,
|
| 1144 |
+
"learning_rate": 1.8279372569580988e-06,
|
| 1145 |
+
"loss": 1.4526,
|
| 1146 |
+
"step": 92500
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 4.06,
|
| 1150 |
+
"learning_rate": 1.8216017826713858e-06,
|
| 1151 |
+
"loss": 1.4524,
|
| 1152 |
+
"step": 93000
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 4.09,
|
| 1156 |
+
"learning_rate": 1.8152663083846724e-06,
|
| 1157 |
+
"loss": 1.4523,
|
| 1158 |
+
"step": 93500
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 4.11,
|
| 1162 |
+
"learning_rate": 1.8089308340979591e-06,
|
| 1163 |
+
"loss": 1.4523,
|
| 1164 |
+
"step": 94000
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.13,
|
| 1168 |
+
"learning_rate": 1.8025953598112466e-06,
|
| 1169 |
+
"loss": 1.452,
|
| 1170 |
+
"step": 94500
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 4.15,
|
| 1174 |
+
"learning_rate": 1.7962598855245333e-06,
|
| 1175 |
+
"loss": 1.4523,
|
| 1176 |
+
"step": 95000
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 4.17,
|
| 1180 |
+
"learning_rate": 1.7899244112378203e-06,
|
| 1181 |
+
"loss": 1.4519,
|
| 1182 |
+
"step": 95500
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 4.19,
|
| 1186 |
+
"learning_rate": 1.7835889369511077e-06,
|
| 1187 |
+
"loss": 1.4521,
|
| 1188 |
+
"step": 96000
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 4.22,
|
| 1192 |
+
"learning_rate": 1.7772534626643947e-06,
|
| 1193 |
+
"loss": 1.452,
|
| 1194 |
+
"step": 96500
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 4.24,
|
| 1198 |
+
"learning_rate": 1.7709179883776815e-06,
|
| 1199 |
+
"loss": 1.4516,
|
| 1200 |
+
"step": 97000
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 4.26,
|
| 1204 |
+
"learning_rate": 1.764582514090969e-06,
|
| 1205 |
+
"loss": 1.4517,
|
| 1206 |
+
"step": 97500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.28,
|
| 1210 |
+
"learning_rate": 1.7582470398042557e-06,
|
| 1211 |
+
"loss": 1.4515,
|
| 1212 |
+
"step": 98000
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 4.3,
|
| 1216 |
+
"learning_rate": 1.7519115655175427e-06,
|
| 1217 |
+
"loss": 1.4513,
|
| 1218 |
+
"step": 98500
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 4.33,
|
| 1222 |
+
"learning_rate": 1.7455760912308294e-06,
|
| 1223 |
+
"loss": 1.451,
|
| 1224 |
+
"step": 99000
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 4.35,
|
| 1228 |
+
"learning_rate": 1.7392406169441166e-06,
|
| 1229 |
+
"loss": 1.4513,
|
| 1230 |
+
"step": 99500
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 4.37,
|
| 1234 |
+
"learning_rate": 1.7329051426574036e-06,
|
| 1235 |
+
"loss": 1.4509,
|
| 1236 |
+
"step": 100000
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 4.39,
|
| 1240 |
+
"learning_rate": 1.7265696683706906e-06,
|
| 1241 |
+
"loss": 1.4513,
|
| 1242 |
+
"step": 100500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 4.41,
|
| 1246 |
+
"learning_rate": 1.7202341940839778e-06,
|
| 1247 |
+
"loss": 1.4514,
|
| 1248 |
+
"step": 101000
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.43,
|
| 1252 |
+
"learning_rate": 1.713898719797265e-06,
|
| 1253 |
+
"loss": 1.4513,
|
| 1254 |
+
"step": 101500
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 4.46,
|
| 1258 |
+
"learning_rate": 1.7075632455105518e-06,
|
| 1259 |
+
"loss": 1.4509,
|
| 1260 |
+
"step": 102000
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 4.48,
|
| 1264 |
+
"learning_rate": 1.701227771223839e-06,
|
| 1265 |
+
"loss": 1.4509,
|
| 1266 |
+
"step": 102500
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 4.5,
|
| 1270 |
+
"learning_rate": 1.694892296937126e-06,
|
| 1271 |
+
"loss": 1.451,
|
| 1272 |
+
"step": 103000
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 4.52,
|
| 1276 |
+
"learning_rate": 1.688556822650413e-06,
|
| 1277 |
+
"loss": 1.4505,
|
| 1278 |
+
"step": 103500
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 4.54,
|
| 1282 |
+
"learning_rate": 1.6822213483637002e-06,
|
| 1283 |
+
"loss": 1.4506,
|
| 1284 |
+
"step": 104000
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 4.57,
|
| 1288 |
+
"learning_rate": 1.675885874076987e-06,
|
| 1289 |
+
"loss": 1.4507,
|
| 1290 |
+
"step": 104500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 4.59,
|
| 1294 |
+
"learning_rate": 1.669550399790274e-06,
|
| 1295 |
+
"loss": 1.4506,
|
| 1296 |
+
"step": 105000
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 4.61,
|
| 1300 |
+
"learning_rate": 1.6632149255035607e-06,
|
| 1301 |
+
"loss": 1.4504,
|
| 1302 |
+
"step": 105500
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 4.63,
|
| 1306 |
+
"learning_rate": 1.6568794512168481e-06,
|
| 1307 |
+
"loss": 1.4502,
|
| 1308 |
+
"step": 106000
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 4.65,
|
| 1312 |
+
"learning_rate": 1.650543976930135e-06,
|
| 1313 |
+
"loss": 1.4504,
|
| 1314 |
+
"step": 106500
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 4.68,
|
| 1318 |
+
"learning_rate": 1.644208502643422e-06,
|
| 1319 |
+
"loss": 1.4506,
|
| 1320 |
+
"step": 107000
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 4.7,
|
| 1324 |
+
"learning_rate": 1.6378730283567093e-06,
|
| 1325 |
+
"loss": 1.4502,
|
| 1326 |
+
"step": 107500
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 4.72,
|
| 1330 |
+
"learning_rate": 1.6315375540699963e-06,
|
| 1331 |
+
"loss": 1.4504,
|
| 1332 |
+
"step": 108000
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 4.74,
|
| 1336 |
+
"learning_rate": 1.625202079783283e-06,
|
| 1337 |
+
"loss": 1.4505,
|
| 1338 |
+
"step": 108500
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 4.76,
|
| 1342 |
+
"learning_rate": 1.6188666054965705e-06,
|
| 1343 |
+
"loss": 1.4502,
|
| 1344 |
+
"step": 109000
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 4.78,
|
| 1348 |
+
"learning_rate": 1.6125311312098573e-06,
|
| 1349 |
+
"loss": 1.4503,
|
| 1350 |
+
"step": 109500
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 4.81,
|
| 1354 |
+
"learning_rate": 1.6061956569231442e-06,
|
| 1355 |
+
"loss": 1.4501,
|
| 1356 |
+
"step": 110000
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 4.83,
|
| 1360 |
+
"learning_rate": 1.599860182636431e-06,
|
| 1361 |
+
"loss": 1.45,
|
| 1362 |
+
"step": 110500
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 4.85,
|
| 1366 |
+
"learning_rate": 1.5935247083497182e-06,
|
| 1367 |
+
"loss": 1.4501,
|
| 1368 |
+
"step": 111000
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 4.87,
|
| 1372 |
+
"learning_rate": 1.5871892340630052e-06,
|
| 1373 |
+
"loss": 1.4495,
|
| 1374 |
+
"step": 111500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 4.89,
|
| 1378 |
+
"learning_rate": 1.5808537597762928e-06,
|
| 1379 |
+
"loss": 1.4495,
|
| 1380 |
+
"step": 112000
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 4.92,
|
| 1384 |
+
"learning_rate": 1.5745182854895796e-06,
|
| 1385 |
+
"loss": 1.4495,
|
| 1386 |
+
"step": 112500
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 4.94,
|
| 1390 |
+
"learning_rate": 1.5681828112028666e-06,
|
| 1391 |
+
"loss": 1.4498,
|
| 1392 |
+
"step": 113000
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 4.96,
|
| 1396 |
+
"learning_rate": 1.5618473369161534e-06,
|
| 1397 |
+
"loss": 1.4498,
|
| 1398 |
+
"step": 113500
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 4.98,
|
| 1402 |
+
"learning_rate": 1.5555118626294406e-06,
|
| 1403 |
+
"loss": 1.45,
|
| 1404 |
+
"step": 114000
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 5.0,
|
| 1408 |
+
"eval_loss": 1.449528455734253,
|
| 1409 |
+
"eval_runtime": 13.3969,
|
| 1410 |
+
"eval_samples_per_second": 73.599,
|
| 1411 |
+
"step": 114435
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 5.0,
|
| 1415 |
+
"learning_rate": 1.5491763883427276e-06,
|
| 1416 |
+
"loss": 1.4501,
|
| 1417 |
+
"step": 114500
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 5.02,
|
| 1421 |
+
"learning_rate": 1.5428409140560146e-06,
|
| 1422 |
+
"loss": 1.4485,
|
| 1423 |
+
"step": 115000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.05,
|
| 1427 |
+
"learning_rate": 1.5365054397693011e-06,
|
| 1428 |
+
"loss": 1.4482,
|
| 1429 |
+
"step": 115500
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 5.07,
|
| 1433 |
+
"learning_rate": 1.5301699654825883e-06,
|
| 1434 |
+
"loss": 1.4484,
|
| 1435 |
+
"step": 116000
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 5.09,
|
| 1439 |
+
"learning_rate": 1.5238344911958755e-06,
|
| 1440 |
+
"loss": 1.4482,
|
| 1441 |
+
"step": 116500
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 5.11,
|
| 1445 |
+
"learning_rate": 1.5174990169091623e-06,
|
| 1446 |
+
"loss": 1.4487,
|
| 1447 |
+
"step": 117000
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 5.13,
|
| 1451 |
+
"learning_rate": 1.5111635426224497e-06,
|
| 1452 |
+
"loss": 1.4482,
|
| 1453 |
+
"step": 117500
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 5.16,
|
| 1457 |
+
"learning_rate": 1.5048280683357365e-06,
|
| 1458 |
+
"loss": 1.4484,
|
| 1459 |
+
"step": 118000
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 5.18,
|
| 1463 |
+
"learning_rate": 1.4984925940490235e-06,
|
| 1464 |
+
"loss": 1.4483,
|
| 1465 |
+
"step": 118500
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.2,
|
| 1469 |
+
"learning_rate": 1.4921571197623105e-06,
|
| 1470 |
+
"loss": 1.4482,
|
| 1471 |
+
"step": 119000
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 5.22,
|
| 1475 |
+
"learning_rate": 1.4858216454755974e-06,
|
| 1476 |
+
"loss": 1.4482,
|
| 1477 |
+
"step": 119500
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 5.24,
|
| 1481 |
+
"learning_rate": 1.4794861711888844e-06,
|
| 1482 |
+
"loss": 1.448,
|
| 1483 |
+
"step": 120000
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 5.26,
|
| 1487 |
+
"learning_rate": 1.4731506969021714e-06,
|
| 1488 |
+
"loss": 1.4481,
|
| 1489 |
+
"step": 120500
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 5.29,
|
| 1493 |
+
"learning_rate": 1.4668152226154584e-06,
|
| 1494 |
+
"loss": 1.448,
|
| 1495 |
+
"step": 121000
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 5.31,
|
| 1499 |
+
"learning_rate": 1.4604797483287458e-06,
|
| 1500 |
+
"loss": 1.4482,
|
| 1501 |
+
"step": 121500
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 5.33,
|
| 1505 |
+
"learning_rate": 1.4541442740420328e-06,
|
| 1506 |
+
"loss": 1.4481,
|
| 1507 |
+
"step": 122000
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.35,
|
| 1511 |
+
"learning_rate": 1.4478087997553198e-06,
|
| 1512 |
+
"loss": 1.4479,
|
| 1513 |
+
"step": 122500
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 5.37,
|
| 1517 |
+
"learning_rate": 1.4414733254686068e-06,
|
| 1518 |
+
"loss": 1.4478,
|
| 1519 |
+
"step": 123000
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 5.4,
|
| 1523 |
+
"learning_rate": 1.4351378511818938e-06,
|
| 1524 |
+
"loss": 1.448,
|
| 1525 |
+
"step": 123500
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 5.42,
|
| 1529 |
+
"learning_rate": 1.4288023768951808e-06,
|
| 1530 |
+
"loss": 1.4478,
|
| 1531 |
+
"step": 124000
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 5.44,
|
| 1535 |
+
"learning_rate": 1.4224669026084677e-06,
|
| 1536 |
+
"loss": 1.4478,
|
| 1537 |
+
"step": 124500
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 5.46,
|
| 1541 |
+
"learning_rate": 1.4161314283217547e-06,
|
| 1542 |
+
"loss": 1.4478,
|
| 1543 |
+
"step": 125000
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 5.48,
|
| 1547 |
+
"learning_rate": 1.4097959540350417e-06,
|
| 1548 |
+
"loss": 1.4477,
|
| 1549 |
+
"step": 125500
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 5.51,
|
| 1553 |
+
"learning_rate": 1.4034604797483287e-06,
|
| 1554 |
+
"loss": 1.448,
|
| 1555 |
+
"step": 126000
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 5.53,
|
| 1559 |
+
"learning_rate": 1.3971250054616157e-06,
|
| 1560 |
+
"loss": 1.4477,
|
| 1561 |
+
"step": 126500
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 5.55,
|
| 1565 |
+
"learning_rate": 1.3907895311749031e-06,
|
| 1566 |
+
"loss": 1.4478,
|
| 1567 |
+
"step": 127000
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 5.57,
|
| 1571 |
+
"learning_rate": 1.38445405688819e-06,
|
| 1572 |
+
"loss": 1.4476,
|
| 1573 |
+
"step": 127500
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 5.59,
|
| 1577 |
+
"learning_rate": 1.378118582601477e-06,
|
| 1578 |
+
"loss": 1.4478,
|
| 1579 |
+
"step": 128000
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 5.61,
|
| 1583 |
+
"learning_rate": 1.371783108314764e-06,
|
| 1584 |
+
"loss": 1.4474,
|
| 1585 |
+
"step": 128500
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 5.64,
|
| 1589 |
+
"learning_rate": 1.365447634028051e-06,
|
| 1590 |
+
"loss": 1.4477,
|
| 1591 |
+
"step": 129000
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 5.66,
|
| 1595 |
+
"learning_rate": 1.359112159741338e-06,
|
| 1596 |
+
"loss": 1.4473,
|
| 1597 |
+
"step": 129500
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 5.68,
|
| 1601 |
+
"learning_rate": 1.352776685454625e-06,
|
| 1602 |
+
"loss": 1.4474,
|
| 1603 |
+
"step": 130000
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 5.7,
|
| 1607 |
+
"learning_rate": 1.346441211167912e-06,
|
| 1608 |
+
"loss": 1.4478,
|
| 1609 |
+
"step": 130500
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 5.72,
|
| 1613 |
+
"learning_rate": 1.340105736881199e-06,
|
| 1614 |
+
"loss": 1.4475,
|
| 1615 |
+
"step": 131000
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 5.75,
|
| 1619 |
+
"learning_rate": 1.333770262594486e-06,
|
| 1620 |
+
"loss": 1.4473,
|
| 1621 |
+
"step": 131500
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 5.77,
|
| 1625 |
+
"learning_rate": 1.327434788307773e-06,
|
| 1626 |
+
"loss": 1.4473,
|
| 1627 |
+
"step": 132000
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 5.79,
|
| 1631 |
+
"learning_rate": 1.3210993140210604e-06,
|
| 1632 |
+
"loss": 1.4471,
|
| 1633 |
+
"step": 132500
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 5.81,
|
| 1637 |
+
"learning_rate": 1.314763839734347e-06,
|
| 1638 |
+
"loss": 1.4471,
|
| 1639 |
+
"step": 133000
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 5.83,
|
| 1643 |
+
"learning_rate": 1.308428365447634e-06,
|
| 1644 |
+
"loss": 1.447,
|
| 1645 |
+
"step": 133500
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 5.85,
|
| 1649 |
+
"learning_rate": 1.302092891160921e-06,
|
| 1650 |
+
"loss": 1.4471,
|
| 1651 |
+
"step": 134000
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 5.88,
|
| 1655 |
+
"learning_rate": 1.295757416874208e-06,
|
| 1656 |
+
"loss": 1.4473,
|
| 1657 |
+
"step": 134500
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 5.9,
|
| 1661 |
+
"learning_rate": 1.289421942587495e-06,
|
| 1662 |
+
"loss": 1.447,
|
| 1663 |
+
"step": 135000
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 5.92,
|
| 1667 |
+
"learning_rate": 1.283086468300782e-06,
|
| 1668 |
+
"loss": 1.447,
|
| 1669 |
+
"step": 135500
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 5.94,
|
| 1673 |
+
"learning_rate": 1.276750994014069e-06,
|
| 1674 |
+
"loss": 1.4468,
|
| 1675 |
+
"step": 136000
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 5.96,
|
| 1679 |
+
"learning_rate": 1.2704155197273563e-06,
|
| 1680 |
+
"loss": 1.4469,
|
| 1681 |
+
"step": 136500
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 5.99,
|
| 1685 |
+
"learning_rate": 1.2640800454406433e-06,
|
| 1686 |
+
"loss": 1.4469,
|
| 1687 |
+
"step": 137000
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 6.0,
|
| 1691 |
+
"eval_loss": 1.4480507373809814,
|
| 1692 |
+
"eval_runtime": 13.3322,
|
| 1693 |
+
"eval_samples_per_second": 73.956,
|
| 1694 |
+
"step": 137322
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 6.01,
|
| 1698 |
+
"learning_rate": 1.2577445711539303e-06,
|
| 1699 |
+
"loss": 1.4472,
|
| 1700 |
+
"step": 137500
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 6.03,
|
| 1704 |
+
"learning_rate": 1.2514090968672173e-06,
|
| 1705 |
+
"loss": 1.4465,
|
| 1706 |
+
"step": 138000
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 6.05,
|
| 1710 |
+
"learning_rate": 1.2450736225805043e-06,
|
| 1711 |
+
"loss": 1.446,
|
| 1712 |
+
"step": 138500
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 6.07,
|
| 1716 |
+
"learning_rate": 1.2387381482937913e-06,
|
| 1717 |
+
"loss": 1.4461,
|
| 1718 |
+
"step": 139000
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 6.1,
|
| 1722 |
+
"learning_rate": 1.2324026740070782e-06,
|
| 1723 |
+
"loss": 1.4462,
|
| 1724 |
+
"step": 139500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 6.12,
|
| 1728 |
+
"learning_rate": 1.2260671997203652e-06,
|
| 1729 |
+
"loss": 1.4461,
|
| 1730 |
+
"step": 140000
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 6.14,
|
| 1734 |
+
"learning_rate": 1.2197317254336522e-06,
|
| 1735 |
+
"loss": 1.4462,
|
| 1736 |
+
"step": 140500
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 6.16,
|
| 1740 |
+
"learning_rate": 1.2133962511469392e-06,
|
| 1741 |
+
"loss": 1.446,
|
| 1742 |
+
"step": 141000
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 6.18,
|
| 1746 |
+
"learning_rate": 1.2070607768602262e-06,
|
| 1747 |
+
"loss": 1.446,
|
| 1748 |
+
"step": 141500
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 6.2,
|
| 1752 |
+
"learning_rate": 1.2007253025735136e-06,
|
| 1753 |
+
"loss": 1.4461,
|
| 1754 |
+
"step": 142000
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 6.23,
|
| 1758 |
+
"learning_rate": 1.1943898282868006e-06,
|
| 1759 |
+
"loss": 1.446,
|
| 1760 |
+
"step": 142500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 6.25,
|
| 1764 |
+
"learning_rate": 1.1880543540000876e-06,
|
| 1765 |
+
"loss": 1.4458,
|
| 1766 |
+
"step": 143000
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 6.27,
|
| 1770 |
+
"learning_rate": 1.1817188797133746e-06,
|
| 1771 |
+
"loss": 1.4458,
|
| 1772 |
+
"step": 143500
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 6.29,
|
| 1776 |
+
"learning_rate": 1.1753834054266616e-06,
|
| 1777 |
+
"loss": 1.4454,
|
| 1778 |
+
"step": 144000
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 6.31,
|
| 1782 |
+
"learning_rate": 1.1690479311399485e-06,
|
| 1783 |
+
"loss": 1.4459,
|
| 1784 |
+
"step": 144500
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 6.34,
|
| 1788 |
+
"learning_rate": 1.1627124568532355e-06,
|
| 1789 |
+
"loss": 1.4453,
|
| 1790 |
+
"step": 145000
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 6.36,
|
| 1794 |
+
"learning_rate": 1.1563769825665225e-06,
|
| 1795 |
+
"loss": 1.4458,
|
| 1796 |
+
"step": 145500
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 6.38,
|
| 1800 |
+
"learning_rate": 1.1500415082798095e-06,
|
| 1801 |
+
"loss": 1.4458,
|
| 1802 |
+
"step": 146000
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 6.4,
|
| 1806 |
+
"learning_rate": 1.1437060339930965e-06,
|
| 1807 |
+
"loss": 1.4459,
|
| 1808 |
+
"step": 146500
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 6.42,
|
| 1812 |
+
"learning_rate": 1.1373705597063835e-06,
|
| 1813 |
+
"loss": 1.4456,
|
| 1814 |
+
"step": 147000
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 6.44,
|
| 1818 |
+
"learning_rate": 1.131035085419671e-06,
|
| 1819 |
+
"loss": 1.4456,
|
| 1820 |
+
"step": 147500
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 6.47,
|
| 1824 |
+
"learning_rate": 1.1246996111329579e-06,
|
| 1825 |
+
"loss": 1.4456,
|
| 1826 |
+
"step": 148000
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 6.49,
|
| 1830 |
+
"learning_rate": 1.1183641368462449e-06,
|
| 1831 |
+
"loss": 1.4454,
|
| 1832 |
+
"step": 148500
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 6.51,
|
| 1836 |
+
"learning_rate": 1.1120286625595319e-06,
|
| 1837 |
+
"loss": 1.4456,
|
| 1838 |
+
"step": 149000
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 6.53,
|
| 1842 |
+
"learning_rate": 1.1056931882728189e-06,
|
| 1843 |
+
"loss": 1.4455,
|
| 1844 |
+
"step": 149500
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 6.55,
|
| 1848 |
+
"learning_rate": 1.0993577139861058e-06,
|
| 1849 |
+
"loss": 1.4453,
|
| 1850 |
+
"step": 150000
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.58,
|
| 1854 |
+
"learning_rate": 1.0930222396993928e-06,
|
| 1855 |
+
"loss": 1.4454,
|
| 1856 |
+
"step": 150500
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 6.6,
|
| 1860 |
+
"learning_rate": 1.0866867654126794e-06,
|
| 1861 |
+
"loss": 1.4454,
|
| 1862 |
+
"step": 151000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 6.62,
|
| 1866 |
+
"learning_rate": 1.0803512911259668e-06,
|
| 1867 |
+
"loss": 1.4455,
|
| 1868 |
+
"step": 151500
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 6.64,
|
| 1872 |
+
"learning_rate": 1.0740158168392538e-06,
|
| 1873 |
+
"loss": 1.4455,
|
| 1874 |
+
"step": 152000
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 6.66,
|
| 1878 |
+
"learning_rate": 1.0676803425525408e-06,
|
| 1879 |
+
"loss": 1.4454,
|
| 1880 |
+
"step": 152500
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 6.69,
|
| 1884 |
+
"learning_rate": 1.0613448682658278e-06,
|
| 1885 |
+
"loss": 1.4453,
|
| 1886 |
+
"step": 153000
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 6.71,
|
| 1890 |
+
"learning_rate": 1.0550093939791148e-06,
|
| 1891 |
+
"loss": 1.4455,
|
| 1892 |
+
"step": 153500
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.73,
|
| 1896 |
+
"learning_rate": 1.0486739196924017e-06,
|
| 1897 |
+
"loss": 1.4454,
|
| 1898 |
+
"step": 154000
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.75,
|
| 1902 |
+
"learning_rate": 1.0423384454056887e-06,
|
| 1903 |
+
"loss": 1.4455,
|
| 1904 |
+
"step": 154500
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 6.77,
|
| 1908 |
+
"learning_rate": 1.0360029711189757e-06,
|
| 1909 |
+
"loss": 1.4453,
|
| 1910 |
+
"step": 155000
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 6.79,
|
| 1914 |
+
"learning_rate": 1.0296674968322627e-06,
|
| 1915 |
+
"loss": 1.4451,
|
| 1916 |
+
"step": 155500
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 6.82,
|
| 1920 |
+
"learning_rate": 1.02333202254555e-06,
|
| 1921 |
+
"loss": 1.4453,
|
| 1922 |
+
"step": 156000
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 6.84,
|
| 1926 |
+
"learning_rate": 1.016996548258837e-06,
|
| 1927 |
+
"loss": 1.4449,
|
| 1928 |
+
"step": 156500
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 6.86,
|
| 1932 |
+
"learning_rate": 1.0106610739721239e-06,
|
| 1933 |
+
"loss": 1.4453,
|
| 1934 |
+
"step": 157000
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 6.88,
|
| 1938 |
+
"learning_rate": 1.004325599685411e-06,
|
| 1939 |
+
"loss": 1.4452,
|
| 1940 |
+
"step": 157500
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.9,
|
| 1944 |
+
"learning_rate": 9.97990125398698e-07,
|
| 1945 |
+
"loss": 1.4452,
|
| 1946 |
+
"step": 158000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 6.93,
|
| 1950 |
+
"learning_rate": 9.91654651111985e-07,
|
| 1951 |
+
"loss": 1.4453,
|
| 1952 |
+
"step": 158500
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 6.95,
|
| 1956 |
+
"learning_rate": 9.85319176825272e-07,
|
| 1957 |
+
"loss": 1.4449,
|
| 1958 |
+
"step": 159000
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 6.97,
|
| 1962 |
+
"learning_rate": 9.78983702538559e-07,
|
| 1963 |
+
"loss": 1.4448,
|
| 1964 |
+
"step": 159500
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 6.99,
|
| 1968 |
+
"learning_rate": 9.72648228251846e-07,
|
| 1969 |
+
"loss": 1.445,
|
| 1970 |
+
"step": 160000
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 7.0,
|
| 1974 |
+
"eval_loss": 1.4469441175460815,
|
| 1975 |
+
"eval_runtime": 13.3822,
|
| 1976 |
+
"eval_samples_per_second": 73.68,
|
| 1977 |
+
"step": 160209
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 7.01,
|
| 1981 |
+
"learning_rate": 9.66312753965133e-07,
|
| 1982 |
+
"loss": 1.4454,
|
| 1983 |
+
"step": 160500
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 7.03,
|
| 1987 |
+
"learning_rate": 9.5997727967842e-07,
|
| 1988 |
+
"loss": 1.4446,
|
| 1989 |
+
"step": 161000
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 7.06,
|
| 1993 |
+
"learning_rate": 9.536418053917071e-07,
|
| 1994 |
+
"loss": 1.4443,
|
| 1995 |
+
"step": 161500
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 7.08,
|
| 1999 |
+
"learning_rate": 9.473063311049942e-07,
|
| 2000 |
+
"loss": 1.4444,
|
| 2001 |
+
"step": 162000
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 7.1,
|
| 2005 |
+
"learning_rate": 9.409708568182811e-07,
|
| 2006 |
+
"loss": 1.4443,
|
| 2007 |
+
"step": 162500
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 7.12,
|
| 2011 |
+
"learning_rate": 9.346353825315684e-07,
|
| 2012 |
+
"loss": 1.4444,
|
| 2013 |
+
"step": 163000
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 7.14,
|
| 2017 |
+
"learning_rate": 9.282999082448553e-07,
|
| 2018 |
+
"loss": 1.4448,
|
| 2019 |
+
"step": 163500
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 7.17,
|
| 2023 |
+
"learning_rate": 9.219644339581422e-07,
|
| 2024 |
+
"loss": 1.4443,
|
| 2025 |
+
"step": 164000
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.19,
|
| 2029 |
+
"learning_rate": 9.156289596714293e-07,
|
| 2030 |
+
"loss": 1.4442,
|
| 2031 |
+
"step": 164500
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 7.21,
|
| 2035 |
+
"learning_rate": 9.092934853847164e-07,
|
| 2036 |
+
"loss": 1.4446,
|
| 2037 |
+
"step": 165000
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 7.23,
|
| 2041 |
+
"learning_rate": 9.029580110980034e-07,
|
| 2042 |
+
"loss": 1.4442,
|
| 2043 |
+
"step": 165500
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 7.25,
|
| 2047 |
+
"learning_rate": 8.966225368112904e-07,
|
| 2048 |
+
"loss": 1.4443,
|
| 2049 |
+
"step": 166000
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 7.27,
|
| 2053 |
+
"learning_rate": 8.902870625245774e-07,
|
| 2054 |
+
"loss": 1.4445,
|
| 2055 |
+
"step": 166500
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 7.3,
|
| 2059 |
+
"learning_rate": 8.839515882378645e-07,
|
| 2060 |
+
"loss": 1.4447,
|
| 2061 |
+
"step": 167000
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 7.32,
|
| 2065 |
+
"learning_rate": 8.776161139511516e-07,
|
| 2066 |
+
"loss": 1.4444,
|
| 2067 |
+
"step": 167500
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.34,
|
| 2071 |
+
"learning_rate": 8.712806396644386e-07,
|
| 2072 |
+
"loss": 1.4442,
|
| 2073 |
+
"step": 168000
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 7.36,
|
| 2077 |
+
"learning_rate": 8.649451653777251e-07,
|
| 2078 |
+
"loss": 1.4441,
|
| 2079 |
+
"step": 168500
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 7.38,
|
| 2083 |
+
"learning_rate": 8.586096910910122e-07,
|
| 2084 |
+
"loss": 1.444,
|
| 2085 |
+
"step": 169000
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 7.41,
|
| 2089 |
+
"learning_rate": 8.522742168042993e-07,
|
| 2090 |
+
"loss": 1.4441,
|
| 2091 |
+
"step": 169500
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 7.43,
|
| 2095 |
+
"learning_rate": 8.459387425175863e-07,
|
| 2096 |
+
"loss": 1.4443,
|
| 2097 |
+
"step": 170000
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 7.45,
|
| 2101 |
+
"learning_rate": 8.396032682308732e-07,
|
| 2102 |
+
"loss": 1.444,
|
| 2103 |
+
"step": 170500
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 7.47,
|
| 2107 |
+
"learning_rate": 8.332677939441605e-07,
|
| 2108 |
+
"loss": 1.4441,
|
| 2109 |
+
"step": 171000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.49,
|
| 2113 |
+
"learning_rate": 8.269323196574474e-07,
|
| 2114 |
+
"loss": 1.4441,
|
| 2115 |
+
"step": 171500
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 7.52,
|
| 2119 |
+
"learning_rate": 8.205968453707344e-07,
|
| 2120 |
+
"loss": 1.4442,
|
| 2121 |
+
"step": 172000
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 7.54,
|
| 2125 |
+
"learning_rate": 8.142613710840215e-07,
|
| 2126 |
+
"loss": 1.4441,
|
| 2127 |
+
"step": 172500
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 7.56,
|
| 2131 |
+
"learning_rate": 8.079258967973084e-07,
|
| 2132 |
+
"loss": 1.444,
|
| 2133 |
+
"step": 173000
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 7.58,
|
| 2137 |
+
"learning_rate": 8.015904225105957e-07,
|
| 2138 |
+
"loss": 1.4441,
|
| 2139 |
+
"step": 173500
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 7.6,
|
| 2143 |
+
"learning_rate": 7.952549482238825e-07,
|
| 2144 |
+
"loss": 1.4439,
|
| 2145 |
+
"step": 174000
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 7.62,
|
| 2149 |
+
"learning_rate": 7.889194739371695e-07,
|
| 2150 |
+
"loss": 1.4443,
|
| 2151 |
+
"step": 174500
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.65,
|
| 2155 |
+
"learning_rate": 7.825839996504566e-07,
|
| 2156 |
+
"loss": 1.4443,
|
| 2157 |
+
"step": 175000
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 7.67,
|
| 2161 |
+
"learning_rate": 7.762485253637437e-07,
|
| 2162 |
+
"loss": 1.444,
|
| 2163 |
+
"step": 175500
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 7.69,
|
| 2167 |
+
"learning_rate": 7.699130510770306e-07,
|
| 2168 |
+
"loss": 1.444,
|
| 2169 |
+
"step": 176000
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 7.71,
|
| 2173 |
+
"learning_rate": 7.635775767903177e-07,
|
| 2174 |
+
"loss": 1.4435,
|
| 2175 |
+
"step": 176500
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 7.73,
|
| 2179 |
+
"learning_rate": 7.572421025036047e-07,
|
| 2180 |
+
"loss": 1.4439,
|
| 2181 |
+
"step": 177000
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 7.76,
|
| 2185 |
+
"learning_rate": 7.509066282168918e-07,
|
| 2186 |
+
"loss": 1.4439,
|
| 2187 |
+
"step": 177500
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 7.78,
|
| 2191 |
+
"learning_rate": 7.445711539301788e-07,
|
| 2192 |
+
"loss": 1.4437,
|
| 2193 |
+
"step": 178000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 7.8,
|
| 2197 |
+
"learning_rate": 7.382356796434658e-07,
|
| 2198 |
+
"loss": 1.444,
|
| 2199 |
+
"step": 178500
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 7.82,
|
| 2203 |
+
"learning_rate": 7.319002053567528e-07,
|
| 2204 |
+
"loss": 1.4438,
|
| 2205 |
+
"step": 179000
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 7.84,
|
| 2209 |
+
"learning_rate": 7.255647310700398e-07,
|
| 2210 |
+
"loss": 1.4436,
|
| 2211 |
+
"step": 179500
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 7.86,
|
| 2215 |
+
"learning_rate": 7.192292567833268e-07,
|
| 2216 |
+
"loss": 1.4438,
|
| 2217 |
+
"step": 180000
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 7.89,
|
| 2221 |
+
"learning_rate": 7.128937824966139e-07,
|
| 2222 |
+
"loss": 1.4438,
|
| 2223 |
+
"step": 180500
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 7.91,
|
| 2227 |
+
"learning_rate": 7.065583082099009e-07,
|
| 2228 |
+
"loss": 1.4439,
|
| 2229 |
+
"step": 181000
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 7.93,
|
| 2233 |
+
"learning_rate": 7.00222833923188e-07,
|
| 2234 |
+
"loss": 1.4439,
|
| 2235 |
+
"step": 181500
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 7.95,
|
| 2239 |
+
"learning_rate": 6.93887359636475e-07,
|
| 2240 |
+
"loss": 1.4437,
|
| 2241 |
+
"step": 182000
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 7.97,
|
| 2245 |
+
"learning_rate": 6.87551885349762e-07,
|
| 2246 |
+
"loss": 1.4438,
|
| 2247 |
+
"step": 182500
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 8.0,
|
| 2251 |
+
"learning_rate": 6.812164110630491e-07,
|
| 2252 |
+
"loss": 1.4438,
|
| 2253 |
+
"step": 183000
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 8.0,
|
| 2257 |
+
"eval_loss": 1.446402907371521,
|
| 2258 |
+
"eval_runtime": 12.0912,
|
| 2259 |
+
"eval_samples_per_second": 81.547,
|
| 2260 |
+
"step": 183096
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 8.02,
|
| 2264 |
+
"learning_rate": 6.748809367763361e-07,
|
| 2265 |
+
"loss": 1.4439,
|
| 2266 |
+
"step": 183500
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 8.04,
|
| 2270 |
+
"learning_rate": 6.68545462489623e-07,
|
| 2271 |
+
"loss": 1.4432,
|
| 2272 |
+
"step": 184000
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 8.06,
|
| 2276 |
+
"learning_rate": 6.622099882029101e-07,
|
| 2277 |
+
"loss": 1.4433,
|
| 2278 |
+
"step": 184500
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 8.08,
|
| 2282 |
+
"learning_rate": 6.558745139161971e-07,
|
| 2283 |
+
"loss": 1.4431,
|
| 2284 |
+
"step": 185000
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 8.11,
|
| 2288 |
+
"learning_rate": 6.495390396294842e-07,
|
| 2289 |
+
"loss": 1.4431,
|
| 2290 |
+
"step": 185500
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 8.13,
|
| 2294 |
+
"learning_rate": 6.432035653427712e-07,
|
| 2295 |
+
"loss": 1.4432,
|
| 2296 |
+
"step": 186000
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 8.15,
|
| 2300 |
+
"learning_rate": 6.368680910560579e-07,
|
| 2301 |
+
"loss": 1.4433,
|
| 2302 |
+
"step": 186500
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"epoch": 8.17,
|
| 2306 |
+
"learning_rate": 6.30532616769345e-07,
|
| 2307 |
+
"loss": 1.4432,
|
| 2308 |
+
"step": 187000
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 8.19,
|
| 2312 |
+
"learning_rate": 6.24197142482632e-07,
|
| 2313 |
+
"loss": 1.4434,
|
| 2314 |
+
"step": 187500
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 8.21,
|
| 2318 |
+
"learning_rate": 6.17861668195919e-07,
|
| 2319 |
+
"loss": 1.4435,
|
| 2320 |
+
"step": 188000
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 8.24,
|
| 2324 |
+
"learning_rate": 6.11526193909206e-07,
|
| 2325 |
+
"loss": 1.4435,
|
| 2326 |
+
"step": 188500
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 8.26,
|
| 2330 |
+
"learning_rate": 6.05190719622493e-07,
|
| 2331 |
+
"loss": 1.4433,
|
| 2332 |
+
"step": 189000
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 8.28,
|
| 2336 |
+
"learning_rate": 5.9885524533578e-07,
|
| 2337 |
+
"loss": 1.443,
|
| 2338 |
+
"step": 189500
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 8.3,
|
| 2342 |
+
"learning_rate": 5.925197710490671e-07,
|
| 2343 |
+
"loss": 1.4433,
|
| 2344 |
+
"step": 190000
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 8.32,
|
| 2348 |
+
"learning_rate": 5.861842967623541e-07,
|
| 2349 |
+
"loss": 1.4431,
|
| 2350 |
+
"step": 190500
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 8.35,
|
| 2354 |
+
"learning_rate": 5.798488224756412e-07,
|
| 2355 |
+
"loss": 1.4432,
|
| 2356 |
+
"step": 191000
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 8.37,
|
| 2360 |
+
"learning_rate": 5.735133481889282e-07,
|
| 2361 |
+
"loss": 1.443,
|
| 2362 |
+
"step": 191500
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 8.39,
|
| 2366 |
+
"learning_rate": 5.671778739022152e-07,
|
| 2367 |
+
"loss": 1.443,
|
| 2368 |
+
"step": 192000
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 8.41,
|
| 2372 |
+
"learning_rate": 5.608423996155022e-07,
|
| 2373 |
+
"loss": 1.4434,
|
| 2374 |
+
"step": 192500
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 8.43,
|
| 2378 |
+
"learning_rate": 5.545069253287893e-07,
|
| 2379 |
+
"loss": 1.4432,
|
| 2380 |
+
"step": 193000
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 8.45,
|
| 2384 |
+
"learning_rate": 5.481714510420763e-07,
|
| 2385 |
+
"loss": 1.4429,
|
| 2386 |
+
"step": 193500
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 8.48,
|
| 2390 |
+
"learning_rate": 5.418359767553633e-07,
|
| 2391 |
+
"loss": 1.4432,
|
| 2392 |
+
"step": 194000
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 8.5,
|
| 2396 |
+
"learning_rate": 5.355005024686503e-07,
|
| 2397 |
+
"loss": 1.4432,
|
| 2398 |
+
"step": 194500
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 8.52,
|
| 2402 |
+
"learning_rate": 5.291650281819373e-07,
|
| 2403 |
+
"loss": 1.4429,
|
| 2404 |
+
"step": 195000
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 8.54,
|
| 2408 |
+
"learning_rate": 5.228295538952244e-07,
|
| 2409 |
+
"loss": 1.443,
|
| 2410 |
+
"step": 195500
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 8.56,
|
| 2414 |
+
"learning_rate": 5.164940796085115e-07,
|
| 2415 |
+
"loss": 1.4433,
|
| 2416 |
+
"step": 196000
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 8.59,
|
| 2420 |
+
"learning_rate": 5.101586053217985e-07,
|
| 2421 |
+
"loss": 1.4428,
|
| 2422 |
+
"step": 196500
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 8.61,
|
| 2426 |
+
"learning_rate": 5.038231310350855e-07,
|
| 2427 |
+
"loss": 1.4429,
|
| 2428 |
+
"step": 197000
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 8.63,
|
| 2432 |
+
"learning_rate": 4.974876567483725e-07,
|
| 2433 |
+
"loss": 1.4431,
|
| 2434 |
+
"step": 197500
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 8.65,
|
| 2438 |
+
"learning_rate": 4.911521824616596e-07,
|
| 2439 |
+
"loss": 1.4431,
|
| 2440 |
+
"step": 198000
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 8.67,
|
| 2444 |
+
"learning_rate": 4.848167081749465e-07,
|
| 2445 |
+
"loss": 1.443,
|
| 2446 |
+
"step": 198500
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 8.69,
|
| 2450 |
+
"learning_rate": 4.784812338882336e-07,
|
| 2451 |
+
"loss": 1.4433,
|
| 2452 |
+
"step": 199000
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 8.72,
|
| 2456 |
+
"learning_rate": 4.7214575960152063e-07,
|
| 2457 |
+
"loss": 1.4431,
|
| 2458 |
+
"step": 199500
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 8.74,
|
| 2462 |
+
"learning_rate": 4.6581028531480767e-07,
|
| 2463 |
+
"loss": 1.4431,
|
| 2464 |
+
"step": 200000
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 8.76,
|
| 2468 |
+
"learning_rate": 4.5947481102809466e-07,
|
| 2469 |
+
"loss": 1.4433,
|
| 2470 |
+
"step": 200500
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 8.78,
|
| 2474 |
+
"learning_rate": 4.531393367413817e-07,
|
| 2475 |
+
"loss": 1.4429,
|
| 2476 |
+
"step": 201000
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 8.8,
|
| 2480 |
+
"learning_rate": 4.468038624546688e-07,
|
| 2481 |
+
"loss": 1.4431,
|
| 2482 |
+
"step": 201500
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 8.83,
|
| 2486 |
+
"learning_rate": 4.404683881679558e-07,
|
| 2487 |
+
"loss": 1.4431,
|
| 2488 |
+
"step": 202000
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 8.85,
|
| 2492 |
+
"learning_rate": 4.341329138812427e-07,
|
| 2493 |
+
"loss": 1.4433,
|
| 2494 |
+
"step": 202500
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 8.87,
|
| 2498 |
+
"learning_rate": 4.277974395945298e-07,
|
| 2499 |
+
"loss": 1.4428,
|
| 2500 |
+
"step": 203000
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 8.89,
|
| 2504 |
+
"learning_rate": 4.2146196530781685e-07,
|
| 2505 |
+
"loss": 1.4429,
|
| 2506 |
+
"step": 203500
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 8.91,
|
| 2510 |
+
"learning_rate": 4.151264910211035e-07,
|
| 2511 |
+
"loss": 1.4428,
|
| 2512 |
+
"step": 204000
|
| 2513 |
+
},
|
| 2514 |
+
{
|
| 2515 |
+
"epoch": 8.94,
|
| 2516 |
+
"learning_rate": 4.0879101673439056e-07,
|
| 2517 |
+
"loss": 1.4427,
|
| 2518 |
+
"step": 204500
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 8.96,
|
| 2522 |
+
"learning_rate": 4.024555424476776e-07,
|
| 2523 |
+
"loss": 1.4428,
|
| 2524 |
+
"step": 205000
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 8.98,
|
| 2528 |
+
"learning_rate": 3.961200681609647e-07,
|
| 2529 |
+
"loss": 1.4428,
|
| 2530 |
+
"step": 205500
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 9.0,
|
| 2534 |
+
"eval_loss": 1.4459742307662964,
|
| 2535 |
+
"eval_runtime": 11.8598,
|
| 2536 |
+
"eval_samples_per_second": 83.138,
|
| 2537 |
+
"step": 205983
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 9.0,
|
| 2541 |
+
"learning_rate": 3.897845938742517e-07,
|
| 2542 |
+
"loss": 1.4438,
|
| 2543 |
+
"step": 206000
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 9.02,
|
| 2547 |
+
"learning_rate": 3.834491195875386e-07,
|
| 2548 |
+
"loss": 1.4426,
|
| 2549 |
+
"step": 206500
|
| 2550 |
+
},
|
| 2551 |
+
{
|
| 2552 |
+
"epoch": 9.04,
|
| 2553 |
+
"learning_rate": 3.771136453008257e-07,
|
| 2554 |
+
"loss": 1.4426,
|
| 2555 |
+
"step": 207000
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 9.07,
|
| 2559 |
+
"learning_rate": 3.7077817101411276e-07,
|
| 2560 |
+
"loss": 1.4426,
|
| 2561 |
+
"step": 207500
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 9.09,
|
| 2565 |
+
"learning_rate": 3.6444269672739975e-07,
|
| 2566 |
+
"loss": 1.4427,
|
| 2567 |
+
"step": 208000
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 9.11,
|
| 2571 |
+
"learning_rate": 3.581072224406868e-07,
|
| 2572 |
+
"loss": 1.4428,
|
| 2573 |
+
"step": 208500
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 9.13,
|
| 2577 |
+
"learning_rate": 3.5177174815397383e-07,
|
| 2578 |
+
"loss": 1.4425,
|
| 2579 |
+
"step": 209000
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 9.15,
|
| 2583 |
+
"learning_rate": 3.4543627386726087e-07,
|
| 2584 |
+
"loss": 1.4428,
|
| 2585 |
+
"step": 209500
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 9.18,
|
| 2589 |
+
"learning_rate": 3.3910079958054786e-07,
|
| 2590 |
+
"loss": 1.4425,
|
| 2591 |
+
"step": 210000
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 9.2,
|
| 2595 |
+
"learning_rate": 3.327653252938349e-07,
|
| 2596 |
+
"loss": 1.4427,
|
| 2597 |
+
"step": 210500
|
| 2598 |
+
},
|
| 2599 |
+
{
|
| 2600 |
+
"epoch": 9.22,
|
| 2601 |
+
"learning_rate": 3.2642985100712194e-07,
|
| 2602 |
+
"loss": 1.4423,
|
| 2603 |
+
"step": 211000
|
| 2604 |
+
},
|
| 2605 |
+
{
|
| 2606 |
+
"epoch": 9.24,
|
| 2607 |
+
"learning_rate": 3.20094376720409e-07,
|
| 2608 |
+
"loss": 1.4427,
|
| 2609 |
+
"step": 211500
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 9.26,
|
| 2613 |
+
"learning_rate": 3.1375890243369597e-07,
|
| 2614 |
+
"loss": 1.4426,
|
| 2615 |
+
"step": 212000
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 9.28,
|
| 2619 |
+
"learning_rate": 3.07423428146983e-07,
|
| 2620 |
+
"loss": 1.4427,
|
| 2621 |
+
"step": 212500
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 9.31,
|
| 2625 |
+
"learning_rate": 3.0108795386027005e-07,
|
| 2626 |
+
"loss": 1.4427,
|
| 2627 |
+
"step": 213000
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 9.33,
|
| 2631 |
+
"learning_rate": 2.9475247957355704e-07,
|
| 2632 |
+
"loss": 1.4425,
|
| 2633 |
+
"step": 213500
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 9.35,
|
| 2637 |
+
"learning_rate": 2.884170052868441e-07,
|
| 2638 |
+
"loss": 1.4423,
|
| 2639 |
+
"step": 214000
|
| 2640 |
+
},
|
| 2641 |
+
{
|
| 2642 |
+
"epoch": 9.37,
|
| 2643 |
+
"learning_rate": 2.820815310001311e-07,
|
| 2644 |
+
"loss": 1.4427,
|
| 2645 |
+
"step": 214500
|
| 2646 |
+
},
|
| 2647 |
+
{
|
| 2648 |
+
"epoch": 9.39,
|
| 2649 |
+
"learning_rate": 2.757460567134181e-07,
|
| 2650 |
+
"loss": 1.4426,
|
| 2651 |
+
"step": 215000
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 9.42,
|
| 2655 |
+
"learning_rate": 2.694105824267052e-07,
|
| 2656 |
+
"loss": 1.4425,
|
| 2657 |
+
"step": 215500
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 9.44,
|
| 2661 |
+
"learning_rate": 2.630751081399922e-07,
|
| 2662 |
+
"loss": 1.4426,
|
| 2663 |
+
"step": 216000
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 9.46,
|
| 2667 |
+
"learning_rate": 2.5673963385327923e-07,
|
| 2668 |
+
"loss": 1.4427,
|
| 2669 |
+
"step": 216500
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 9.48,
|
| 2673 |
+
"learning_rate": 2.504041595665663e-07,
|
| 2674 |
+
"loss": 1.4426,
|
| 2675 |
+
"step": 217000
|
| 2676 |
+
},
|
| 2677 |
+
{
|
| 2678 |
+
"epoch": 9.5,
|
| 2679 |
+
"learning_rate": 2.4406868527985326e-07,
|
| 2680 |
+
"loss": 1.4424,
|
| 2681 |
+
"step": 217500
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 9.53,
|
| 2685 |
+
"learning_rate": 2.377332109931403e-07,
|
| 2686 |
+
"loss": 1.4426,
|
| 2687 |
+
"step": 218000
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"epoch": 9.55,
|
| 2691 |
+
"learning_rate": 2.3139773670642737e-07,
|
| 2692 |
+
"loss": 1.4425,
|
| 2693 |
+
"step": 218500
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 9.57,
|
| 2697 |
+
"learning_rate": 2.2506226241971439e-07,
|
| 2698 |
+
"loss": 1.4424,
|
| 2699 |
+
"step": 219000
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 9.59,
|
| 2703 |
+
"learning_rate": 2.187267881330014e-07,
|
| 2704 |
+
"loss": 1.4423,
|
| 2705 |
+
"step": 219500
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 9.61,
|
| 2709 |
+
"learning_rate": 2.1239131384628841e-07,
|
| 2710 |
+
"loss": 1.4423,
|
| 2711 |
+
"step": 220000
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 9.63,
|
| 2715 |
+
"learning_rate": 2.0605583955957543e-07,
|
| 2716 |
+
"loss": 1.4428,
|
| 2717 |
+
"step": 220500
|
| 2718 |
+
},
|
| 2719 |
+
{
|
| 2720 |
+
"epoch": 9.66,
|
| 2721 |
+
"learning_rate": 1.997203652728625e-07,
|
| 2722 |
+
"loss": 1.4423,
|
| 2723 |
+
"step": 221000
|
| 2724 |
+
},
|
| 2725 |
+
{
|
| 2726 |
+
"epoch": 9.68,
|
| 2727 |
+
"learning_rate": 1.933848909861492e-07,
|
| 2728 |
+
"loss": 1.4425,
|
| 2729 |
+
"step": 221500
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 9.7,
|
| 2733 |
+
"learning_rate": 1.870494166994362e-07,
|
| 2734 |
+
"loss": 1.4425,
|
| 2735 |
+
"step": 222000
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 9.72,
|
| 2739 |
+
"learning_rate": 1.8071394241272325e-07,
|
| 2740 |
+
"loss": 1.4424,
|
| 2741 |
+
"step": 222500
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 9.74,
|
| 2745 |
+
"learning_rate": 1.7437846812601026e-07,
|
| 2746 |
+
"loss": 1.4425,
|
| 2747 |
+
"step": 223000
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 9.77,
|
| 2751 |
+
"learning_rate": 1.6804299383929728e-07,
|
| 2752 |
+
"loss": 1.4424,
|
| 2753 |
+
"step": 223500
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 9.79,
|
| 2757 |
+
"learning_rate": 1.6170751955258432e-07,
|
| 2758 |
+
"loss": 1.4424,
|
| 2759 |
+
"step": 224000
|
| 2760 |
+
},
|
| 2761 |
+
{
|
| 2762 |
+
"epoch": 9.81,
|
| 2763 |
+
"learning_rate": 1.5537204526587136e-07,
|
| 2764 |
+
"loss": 1.4424,
|
| 2765 |
+
"step": 224500
|
| 2766 |
+
},
|
| 2767 |
+
{
|
| 2768 |
+
"epoch": 9.83,
|
| 2769 |
+
"learning_rate": 1.4903657097915838e-07,
|
| 2770 |
+
"loss": 1.4423,
|
| 2771 |
+
"step": 225000
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 9.85,
|
| 2775 |
+
"learning_rate": 1.427010966924454e-07,
|
| 2776 |
+
"loss": 1.4426,
|
| 2777 |
+
"step": 225500
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 9.87,
|
| 2781 |
+
"learning_rate": 1.3636562240573243e-07,
|
| 2782 |
+
"loss": 1.4427,
|
| 2783 |
+
"step": 226000
|
| 2784 |
+
},
|
| 2785 |
+
{
|
| 2786 |
+
"epoch": 9.9,
|
| 2787 |
+
"learning_rate": 1.3003014811901947e-07,
|
| 2788 |
+
"loss": 1.4423,
|
| 2789 |
+
"step": 226500
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 9.92,
|
| 2793 |
+
"learning_rate": 1.236946738323065e-07,
|
| 2794 |
+
"loss": 1.4423,
|
| 2795 |
+
"step": 227000
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 9.94,
|
| 2799 |
+
"learning_rate": 1.1735919954559353e-07,
|
| 2800 |
+
"loss": 1.4421,
|
| 2801 |
+
"step": 227500
|
| 2802 |
+
},
|
| 2803 |
+
{
|
| 2804 |
+
"epoch": 9.96,
|
| 2805 |
+
"learning_rate": 1.1102372525888054e-07,
|
| 2806 |
+
"loss": 1.4423,
|
| 2807 |
+
"step": 228000
|
| 2808 |
+
},
|
| 2809 |
+
{
|
| 2810 |
+
"epoch": 9.98,
|
| 2811 |
+
"learning_rate": 1.0468825097216756e-07,
|
| 2812 |
+
"loss": 1.4427,
|
| 2813 |
+
"step": 228500
|
| 2814 |
+
},
|
| 2815 |
+
{
|
| 2816 |
+
"epoch": 10.0,
|
| 2817 |
+
"eval_loss": 1.4457148313522339,
|
| 2818 |
+
"eval_runtime": 13.0706,
|
| 2819 |
+
"eval_samples_per_second": 75.436,
|
| 2820 |
+
"step": 228870
|
| 2821 |
+
}
|
| 2822 |
+
],
|
| 2823 |
+
"max_steps": 228870,
|
| 2824 |
+
"num_train_epochs": 10,
|
| 2825 |
+
"total_flos": 1.288906420929577e+19,
|
| 2826 |
+
"trial_name": null,
|
| 2827 |
+
"trial_params": null
|
| 2828 |
+
}
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|