pszemraj commited on
Commit
fd0e378
1 Parent(s): 83cf6d0
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 768,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,142 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: BEE-spoke-data/bert-plus-L8-4096-v1.0
3
+ library_name: sentence-transformers
4
+ pipeline_tag: sentence-similarity
5
+ tags:
6
+ - sentence-transformers
7
+ - feature-extraction
8
+ - sentence-similarity
9
+ - transformers
10
+ license: apache-2.0
11
+ widget:
12
+ - source_sentence: "How to discreetly optimize operating expenses?"
13
+ sentences:
14
+ - "Strategies for quietly reducing overhead costs"
15
+ - "Subtle ways to cut down on operating expenses"
16
+ - "Implementing technology to save on operating costs without broad announcements"
17
+ - "Lowering daily business expenses through unnoticed efficiencies"
18
+ - "Minimizing operational expenditures in small businesses without drawing attention"
19
+
20
+
21
+ ---
22
+
23
+ # BEE-spoke-data/bert-plus-L8-v1.0-allNLI_matryoshka
24
+
25
+
26
+
27
+ This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
28
+
29
+ - this was finetuned at 512 ctx (allNLI is all short-ctx examples) but the base model supports 4096
30
+ - Matryoshka dims: [768, 512, 256, 128, 64]
31
+
32
+ <!--- Describe your model here -->
33
+
34
+ ## Usage (Sentence-Transformers)
35
+
36
+ Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
37
+
38
+ ```
39
+ pip install -U sentence-transformers
40
+ ```
41
+
42
+ Then you can use the model like this:
43
+
44
+ ```python
45
+ from sentence_transformers import SentenceTransformer
46
+ sentences = ["This is an example sentence", "Each sentence is converted"]
47
+
48
+ model = SentenceTransformer('BEE-spoke-data/bert-plus-L8-v1.0-allNLI_matryoshka')
49
+ embeddings = model.encode(sentences)
50
+ print(embeddings)
51
+ ```
52
+
53
+
54
+
55
+ ## Usage (HuggingFace Transformers)
56
+ Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
57
+
58
+ ```python
59
+ from transformers import AutoTokenizer, AutoModel
60
+ import torch
61
+
62
+
63
+ #Mean Pooling - Take attention mask into account for correct averaging
64
+ def mean_pooling(model_output, attention_mask):
65
+ token_embeddings = model_output[0] #First element of model_output contains all token embeddings
66
+ input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
67
+ return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
68
+
69
+
70
+ # Sentences we want sentence embeddings for
71
+ sentences = ['This is an example sentence', 'Each sentence is converted']
72
+
73
+ # Load model from HuggingFace Hub
74
+ tokenizer = AutoTokenizer.from_pretrained('BEE-spoke-data/bert-plus-L8-v1.0-allNLI_matryoshka')
75
+ model = AutoModel.from_pretrained('BEE-spoke-data/bert-plus-L8-v1.0-allNLI_matryoshka')
76
+
77
+ # Tokenize sentences
78
+ encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
79
+
80
+ # Compute token embeddings
81
+ with torch.no_grad():
82
+ model_output = model(**encoded_input)
83
+
84
+ # Perform pooling. In this case, mean pooling.
85
+ sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
86
+
87
+ print("Sentence embeddings:")
88
+ print(sentence_embeddings)
89
+ ```
90
+
91
+
92
+
93
+ ## Evaluation Results
94
+
95
+ <!--- Describe how your model was evaluated -->
96
+
97
+ For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name=BEE-spoke-data/bert-plus-L8-v1.0-allNLI_matryoshka)
98
+
99
+
100
+ ## Training
101
+ The model was trained with the parameters:
102
+
103
+ **DataLoader**:
104
+
105
+ `sentence_transformers.datasets.NoDuplicatesDataLoader.NoDuplicatesDataLoader` of length 8663 with parameters:
106
+ ```
107
+ {'batch_size': 32}
108
+ ```
109
+
110
+ **Loss**:
111
+
112
+ `sentence_transformers.losses.MatryoshkaLoss.MatryoshkaLoss` with parameters:
113
+ ```
114
+ {'loss': 'MultipleNegativesRankingLoss', 'matryoshka_dims': [768, 512, 256, 128, 64], 'matryoshka_weights': [1, 1, 1, 1, 1], 'n_dims_per_step': -1}
115
+ ```
116
+
117
+ Parameters of the fit()-Method:
118
+ ```
119
+ {
120
+ "epochs": 1,
121
+ "evaluation_steps": 216,
122
+ "evaluator": "sentence_transformers.evaluation.EmbeddingSimilarityEvaluator.EmbeddingSimilarityEvaluator",
123
+ "max_grad_norm": 1,
124
+ "optimizer_class": "<class 'torch.optim.adamw.AdamW'>",
125
+ "optimizer_params": {
126
+ "lr": 2e-05
127
+ },
128
+ "scheduler": "WarmupLinear",
129
+ "steps_per_epoch": null,
130
+ "warmup_steps": 867,
131
+ "weight_decay": 0.01
132
+ }
133
+ ```
134
+
135
+
136
+ ## Full Model Architecture
137
+ ```
138
+ SentenceTransformer(
139
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
140
+ (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
141
+ )
142
+ ```
config_sentence_transformers.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "2.5.1",
4
+ "transformers": "4.38.2",
5
+ "pytorch": "2.1.0+cu121"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null
9
+ }
modules.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ }
14
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 4096,
3
+ "do_lower_case": false
4
+ }
trainer_state.json ADDED
@@ -0,0 +1,3279 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9652652915549957,
3
+ "best_model_checkpoint": "checkpoints/BEE-spoke-data-bert-plus-L8-v1.0-allNLI_matryoshka-synthetic-text-similarity-Mar-21_10-51/checkpoint-1800",
4
+ "epoch": 0.2894472361809045,
5
+ "eval_steps": 300,
6
+ "global_step": 1800,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "grad_norm": 3.4826388359069824,
14
+ "learning_rate": 1.2861736334405146e-07,
15
+ "loss": 0.0905,
16
+ "step": 4
17
+ },
18
+ {
19
+ "epoch": 0.0,
20
+ "grad_norm": 2.3191146850585938,
21
+ "learning_rate": 2.572347266881029e-07,
22
+ "loss": 0.1138,
23
+ "step": 8
24
+ },
25
+ {
26
+ "epoch": 0.0,
27
+ "grad_norm": 2.398422956466675,
28
+ "learning_rate": 3.858520900321544e-07,
29
+ "loss": 0.1055,
30
+ "step": 12
31
+ },
32
+ {
33
+ "epoch": 0.0,
34
+ "grad_norm": 3.034379720687866,
35
+ "learning_rate": 5.144694533762058e-07,
36
+ "loss": 0.1113,
37
+ "step": 16
38
+ },
39
+ {
40
+ "epoch": 0.0,
41
+ "grad_norm": 5.541696071624756,
42
+ "learning_rate": 6.430868167202573e-07,
43
+ "loss": 0.0875,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.0,
48
+ "grad_norm": 3.2232377529144287,
49
+ "learning_rate": 7.717041800643088e-07,
50
+ "loss": 0.1163,
51
+ "step": 24
52
+ },
53
+ {
54
+ "epoch": 0.0,
55
+ "grad_norm": 1.692831039428711,
56
+ "learning_rate": 9.003215434083602e-07,
57
+ "loss": 0.0746,
58
+ "step": 28
59
+ },
60
+ {
61
+ "epoch": 0.01,
62
+ "grad_norm": 0.9550053477287292,
63
+ "learning_rate": 1.0289389067524116e-06,
64
+ "loss": 0.0697,
65
+ "step": 32
66
+ },
67
+ {
68
+ "epoch": 0.01,
69
+ "grad_norm": 2.5329389572143555,
70
+ "learning_rate": 1.157556270096463e-06,
71
+ "loss": 0.0716,
72
+ "step": 36
73
+ },
74
+ {
75
+ "epoch": 0.01,
76
+ "grad_norm": 2.0605404376983643,
77
+ "learning_rate": 1.2861736334405146e-06,
78
+ "loss": 0.0478,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.01,
83
+ "grad_norm": 3.4411487579345703,
84
+ "learning_rate": 1.4147909967845661e-06,
85
+ "loss": 0.0971,
86
+ "step": 44
87
+ },
88
+ {
89
+ "epoch": 0.01,
90
+ "grad_norm": 1.9422132968902588,
91
+ "learning_rate": 1.5434083601286177e-06,
92
+ "loss": 0.0567,
93
+ "step": 48
94
+ },
95
+ {
96
+ "epoch": 0.01,
97
+ "grad_norm": 2.2038466930389404,
98
+ "learning_rate": 1.6720257234726688e-06,
99
+ "loss": 0.0477,
100
+ "step": 52
101
+ },
102
+ {
103
+ "epoch": 0.01,
104
+ "grad_norm": 1.7133512496948242,
105
+ "learning_rate": 1.8006430868167204e-06,
106
+ "loss": 0.0528,
107
+ "step": 56
108
+ },
109
+ {
110
+ "epoch": 0.01,
111
+ "grad_norm": 1.754583716392517,
112
+ "learning_rate": 1.9292604501607717e-06,
113
+ "loss": 0.048,
114
+ "step": 60
115
+ },
116
+ {
117
+ "epoch": 0.01,
118
+ "grad_norm": 0.8963572382926941,
119
+ "learning_rate": 2.0578778135048233e-06,
120
+ "loss": 0.0508,
121
+ "step": 64
122
+ },
123
+ {
124
+ "epoch": 0.01,
125
+ "grad_norm": 1.0214054584503174,
126
+ "learning_rate": 2.186495176848875e-06,
127
+ "loss": 0.0485,
128
+ "step": 68
129
+ },
130
+ {
131
+ "epoch": 0.01,
132
+ "grad_norm": 1.9729621410369873,
133
+ "learning_rate": 2.315112540192926e-06,
134
+ "loss": 0.051,
135
+ "step": 72
136
+ },
137
+ {
138
+ "epoch": 0.01,
139
+ "grad_norm": 1.5615208148956299,
140
+ "learning_rate": 2.4437299035369775e-06,
141
+ "loss": 0.0604,
142
+ "step": 76
143
+ },
144
+ {
145
+ "epoch": 0.01,
146
+ "grad_norm": 1.6368948221206665,
147
+ "learning_rate": 2.572347266881029e-06,
148
+ "loss": 0.0442,
149
+ "step": 80
150
+ },
151
+ {
152
+ "epoch": 0.01,
153
+ "grad_norm": 2.133140802383423,
154
+ "learning_rate": 2.7009646302250807e-06,
155
+ "loss": 0.0302,
156
+ "step": 84
157
+ },
158
+ {
159
+ "epoch": 0.01,
160
+ "grad_norm": 1.2055103778839111,
161
+ "learning_rate": 2.8295819935691322e-06,
162
+ "loss": 0.0432,
163
+ "step": 88
164
+ },
165
+ {
166
+ "epoch": 0.01,
167
+ "grad_norm": 1.5411338806152344,
168
+ "learning_rate": 2.9581993569131834e-06,
169
+ "loss": 0.0319,
170
+ "step": 92
171
+ },
172
+ {
173
+ "epoch": 0.02,
174
+ "grad_norm": 0.75223708152771,
175
+ "learning_rate": 3.0868167202572353e-06,
176
+ "loss": 0.0364,
177
+ "step": 96
178
+ },
179
+ {
180
+ "epoch": 0.02,
181
+ "grad_norm": 1.4433358907699585,
182
+ "learning_rate": 3.2154340836012865e-06,
183
+ "loss": 0.0661,
184
+ "step": 100
185
+ },
186
+ {
187
+ "epoch": 0.02,
188
+ "grad_norm": 1.5212054252624512,
189
+ "learning_rate": 3.3440514469453376e-06,
190
+ "loss": 0.0609,
191
+ "step": 104
192
+ },
193
+ {
194
+ "epoch": 0.02,
195
+ "grad_norm": 2.2315409183502197,
196
+ "learning_rate": 3.4726688102893896e-06,
197
+ "loss": 0.0546,
198
+ "step": 108
199
+ },
200
+ {
201
+ "epoch": 0.02,
202
+ "grad_norm": 1.073073148727417,
203
+ "learning_rate": 3.6012861736334407e-06,
204
+ "loss": 0.0686,
205
+ "step": 112
206
+ },
207
+ {
208
+ "epoch": 0.02,
209
+ "grad_norm": 1.4473097324371338,
210
+ "learning_rate": 3.7299035369774923e-06,
211
+ "loss": 0.0363,
212
+ "step": 116
213
+ },
214
+ {
215
+ "epoch": 0.02,
216
+ "grad_norm": 2.4474904537200928,
217
+ "learning_rate": 3.8585209003215434e-06,
218
+ "loss": 0.0425,
219
+ "step": 120
220
+ },
221
+ {
222
+ "epoch": 0.02,
223
+ "grad_norm": 2.410780429840088,
224
+ "learning_rate": 3.987138263665595e-06,
225
+ "loss": 0.057,
226
+ "step": 124
227
+ },
228
+ {
229
+ "epoch": 0.02,
230
+ "grad_norm": 1.380175232887268,
231
+ "learning_rate": 4.1157556270096466e-06,
232
+ "loss": 0.045,
233
+ "step": 128
234
+ },
235
+ {
236
+ "epoch": 0.02,
237
+ "grad_norm": 1.0906972885131836,
238
+ "learning_rate": 4.244372990353698e-06,
239
+ "loss": 0.0338,
240
+ "step": 132
241
+ },
242
+ {
243
+ "epoch": 0.02,
244
+ "grad_norm": 1.6216851472854614,
245
+ "learning_rate": 4.37299035369775e-06,
246
+ "loss": 0.036,
247
+ "step": 136
248
+ },
249
+ {
250
+ "epoch": 0.02,
251
+ "grad_norm": 0.806192934513092,
252
+ "learning_rate": 4.501607717041801e-06,
253
+ "loss": 0.037,
254
+ "step": 140
255
+ },
256
+ {
257
+ "epoch": 0.02,
258
+ "grad_norm": 0.959051251411438,
259
+ "learning_rate": 4.630225080385852e-06,
260
+ "loss": 0.0473,
261
+ "step": 144
262
+ },
263
+ {
264
+ "epoch": 0.02,
265
+ "grad_norm": 0.4807804226875305,
266
+ "learning_rate": 4.758842443729904e-06,
267
+ "loss": 0.022,
268
+ "step": 148
269
+ },
270
+ {
271
+ "epoch": 0.02,
272
+ "grad_norm": 1.982123613357544,
273
+ "learning_rate": 4.887459807073955e-06,
274
+ "loss": 0.04,
275
+ "step": 152
276
+ },
277
+ {
278
+ "epoch": 0.03,
279
+ "grad_norm": 1.4609640836715698,
280
+ "learning_rate": 5.016077170418007e-06,
281
+ "loss": 0.0266,
282
+ "step": 156
283
+ },
284
+ {
285
+ "epoch": 0.03,
286
+ "grad_norm": 1.4000838994979858,
287
+ "learning_rate": 5.144694533762058e-06,
288
+ "loss": 0.0211,
289
+ "step": 160
290
+ },
291
+ {
292
+ "epoch": 0.03,
293
+ "grad_norm": 0.828529953956604,
294
+ "learning_rate": 5.273311897106109e-06,
295
+ "loss": 0.0239,
296
+ "step": 164
297
+ },
298
+ {
299
+ "epoch": 0.03,
300
+ "grad_norm": 0.8723558187484741,
301
+ "learning_rate": 5.401929260450161e-06,
302
+ "loss": 0.0273,
303
+ "step": 168
304
+ },
305
+ {
306
+ "epoch": 0.03,
307
+ "grad_norm": 1.4310917854309082,
308
+ "learning_rate": 5.530546623794213e-06,
309
+ "loss": 0.0383,
310
+ "step": 172
311
+ },
312
+ {
313
+ "epoch": 0.03,
314
+ "grad_norm": 1.0429316759109497,
315
+ "learning_rate": 5.6591639871382644e-06,
316
+ "loss": 0.0348,
317
+ "step": 176
318
+ },
319
+ {
320
+ "epoch": 0.03,
321
+ "grad_norm": 0.7097664475440979,
322
+ "learning_rate": 5.787781350482315e-06,
323
+ "loss": 0.0199,
324
+ "step": 180
325
+ },
326
+ {
327
+ "epoch": 0.03,
328
+ "grad_norm": 1.269923448562622,
329
+ "learning_rate": 5.916398713826367e-06,
330
+ "loss": 0.0274,
331
+ "step": 184
332
+ },
333
+ {
334
+ "epoch": 0.03,
335
+ "grad_norm": 1.1587966680526733,
336
+ "learning_rate": 6.045016077170418e-06,
337
+ "loss": 0.0347,
338
+ "step": 188
339
+ },
340
+ {
341
+ "epoch": 0.03,
342
+ "grad_norm": 1.1199533939361572,
343
+ "learning_rate": 6.173633440514471e-06,
344
+ "loss": 0.0294,
345
+ "step": 192
346
+ },
347
+ {
348
+ "epoch": 0.03,
349
+ "grad_norm": 0.9978614449501038,
350
+ "learning_rate": 6.302250803858521e-06,
351
+ "loss": 0.0237,
352
+ "step": 196
353
+ },
354
+ {
355
+ "epoch": 0.03,
356
+ "grad_norm": 1.0913878679275513,
357
+ "learning_rate": 6.430868167202573e-06,
358
+ "loss": 0.039,
359
+ "step": 200
360
+ },
361
+ {
362
+ "epoch": 0.03,
363
+ "grad_norm": 0.7430211901664734,
364
+ "learning_rate": 6.5594855305466245e-06,
365
+ "loss": 0.0328,
366
+ "step": 204
367
+ },
368
+ {
369
+ "epoch": 0.03,
370
+ "grad_norm": 1.597890019416809,
371
+ "learning_rate": 6.688102893890675e-06,
372
+ "loss": 0.033,
373
+ "step": 208
374
+ },
375
+ {
376
+ "epoch": 0.03,
377
+ "grad_norm": 0.8224192261695862,
378
+ "learning_rate": 6.816720257234727e-06,
379
+ "loss": 0.0207,
380
+ "step": 212
381
+ },
382
+ {
383
+ "epoch": 0.03,
384
+ "grad_norm": 2.1937623023986816,
385
+ "learning_rate": 6.945337620578779e-06,
386
+ "loss": 0.0341,
387
+ "step": 216
388
+ },
389
+ {
390
+ "epoch": 0.04,
391
+ "grad_norm": 1.337793231010437,
392
+ "learning_rate": 7.073954983922831e-06,
393
+ "loss": 0.0359,
394
+ "step": 220
395
+ },
396
+ {
397
+ "epoch": 0.04,
398
+ "grad_norm": 1.3222864866256714,
399
+ "learning_rate": 7.2025723472668815e-06,
400
+ "loss": 0.0341,
401
+ "step": 224
402
+ },
403
+ {
404
+ "epoch": 0.04,
405
+ "grad_norm": 1.6585116386413574,
406
+ "learning_rate": 7.331189710610933e-06,
407
+ "loss": 0.0319,
408
+ "step": 228
409
+ },
410
+ {
411
+ "epoch": 0.04,
412
+ "grad_norm": 1.9420249462127686,
413
+ "learning_rate": 7.459807073954985e-06,
414
+ "loss": 0.0458,
415
+ "step": 232
416
+ },
417
+ {
418
+ "epoch": 0.04,
419
+ "grad_norm": 2.1558046340942383,
420
+ "learning_rate": 7.588424437299035e-06,
421
+ "loss": 0.0309,
422
+ "step": 236
423
+ },
424
+ {
425
+ "epoch": 0.04,
426
+ "grad_norm": 2.778165340423584,
427
+ "learning_rate": 7.717041800643087e-06,
428
+ "loss": 0.0532,
429
+ "step": 240
430
+ },
431
+ {
432
+ "epoch": 0.04,
433
+ "grad_norm": 1.3744961023330688,
434
+ "learning_rate": 7.84565916398714e-06,
435
+ "loss": 0.0306,
436
+ "step": 244
437
+ },
438
+ {
439
+ "epoch": 0.04,
440
+ "grad_norm": 2.4764065742492676,
441
+ "learning_rate": 7.97427652733119e-06,
442
+ "loss": 0.0398,
443
+ "step": 248
444
+ },
445
+ {
446
+ "epoch": 0.04,
447
+ "grad_norm": 2.0220906734466553,
448
+ "learning_rate": 8.102893890675242e-06,
449
+ "loss": 0.0552,
450
+ "step": 252
451
+ },
452
+ {
453
+ "epoch": 0.04,
454
+ "grad_norm": 3.4686410427093506,
455
+ "learning_rate": 8.231511254019293e-06,
456
+ "loss": 0.0599,
457
+ "step": 256
458
+ },
459
+ {
460
+ "epoch": 0.04,
461
+ "grad_norm": 2.0917580127716064,
462
+ "learning_rate": 8.360128617363345e-06,
463
+ "loss": 0.036,
464
+ "step": 260
465
+ },
466
+ {
467
+ "epoch": 0.04,
468
+ "grad_norm": 1.3227124214172363,
469
+ "learning_rate": 8.488745980707396e-06,
470
+ "loss": 0.0337,
471
+ "step": 264
472
+ },
473
+ {
474
+ "epoch": 0.04,
475
+ "grad_norm": 1.853574275970459,
476
+ "learning_rate": 8.617363344051448e-06,
477
+ "loss": 0.0356,
478
+ "step": 268
479
+ },
480
+ {
481
+ "epoch": 0.04,
482
+ "grad_norm": 1.250444769859314,
483
+ "learning_rate": 8.7459807073955e-06,
484
+ "loss": 0.0625,
485
+ "step": 272
486
+ },
487
+ {
488
+ "epoch": 0.04,
489
+ "grad_norm": 1.1477417945861816,
490
+ "learning_rate": 8.874598070739551e-06,
491
+ "loss": 0.0325,
492
+ "step": 276
493
+ },
494
+ {
495
+ "epoch": 0.05,
496
+ "grad_norm": 1.7810678482055664,
497
+ "learning_rate": 9.003215434083602e-06,
498
+ "loss": 0.028,
499
+ "step": 280
500
+ },
501
+ {
502
+ "epoch": 0.05,
503
+ "grad_norm": 1.5006675720214844,
504
+ "learning_rate": 9.131832797427654e-06,
505
+ "loss": 0.046,
506
+ "step": 284
507
+ },
508
+ {
509
+ "epoch": 0.05,
510
+ "grad_norm": 0.8945916295051575,
511
+ "learning_rate": 9.260450160771704e-06,
512
+ "loss": 0.0318,
513
+ "step": 288
514
+ },
515
+ {
516
+ "epoch": 0.05,
517
+ "grad_norm": 1.4942963123321533,
518
+ "learning_rate": 9.389067524115757e-06,
519
+ "loss": 0.0254,
520
+ "step": 292
521
+ },
522
+ {
523
+ "epoch": 0.05,
524
+ "grad_norm": 0.9653424620628357,
525
+ "learning_rate": 9.517684887459809e-06,
526
+ "loss": 0.0286,
527
+ "step": 296
528
+ },
529
+ {
530
+ "epoch": 0.05,
531
+ "grad_norm": 1.740561604499817,
532
+ "learning_rate": 9.64630225080386e-06,
533
+ "loss": 0.032,
534
+ "step": 300
535
+ },
536
+ {
537
+ "epoch": 0.05,
538
+ "eval_loss": 0.03372881934046745,
539
+ "eval_pearson_cosine": 0.9332293835600463,
540
+ "eval_pearson_dot": 0.9301330880973007,
541
+ "eval_pearson_euclidean": 0.9260017674080668,
542
+ "eval_pearson_manhattan": 0.9271288221546901,
543
+ "eval_pearson_max": 0.9332293835600463,
544
+ "eval_runtime": 249.5828,
545
+ "eval_samples_per_second": 2.003,
546
+ "eval_spearman_cosine": 0.947022185388043,
547
+ "eval_spearman_dot": 0.9435553217732889,
548
+ "eval_spearman_euclidean": 0.940413083620648,
549
+ "eval_spearman_manhattan": 0.9413909920251474,
550
+ "eval_spearman_max": 0.947022185388043,
551
+ "eval_steps_per_second": 2.003,
552
+ "step": 300
553
+ },
554
+ {
555
+ "epoch": 0.05,
556
+ "grad_norm": 0.6091077923774719,
557
+ "learning_rate": 9.77491961414791e-06,
558
+ "loss": 0.0197,
559
+ "step": 304
560
+ },
561
+ {
562
+ "epoch": 0.05,
563
+ "grad_norm": 1.2511804103851318,
564
+ "learning_rate": 9.903536977491962e-06,
565
+ "loss": 0.0306,
566
+ "step": 308
567
+ },
568
+ {
569
+ "epoch": 0.05,
570
+ "grad_norm": 1.3721988201141357,
571
+ "learning_rate": 1.0032154340836013e-05,
572
+ "loss": 0.0187,
573
+ "step": 312
574
+ },
575
+ {
576
+ "epoch": 0.05,
577
+ "grad_norm": 0.7939477562904358,
578
+ "learning_rate": 1.0160771704180067e-05,
579
+ "loss": 0.0318,
580
+ "step": 316
581
+ },
582
+ {
583
+ "epoch": 0.05,
584
+ "grad_norm": 0.6942979097366333,
585
+ "learning_rate": 1.0289389067524116e-05,
586
+ "loss": 0.0264,
587
+ "step": 320
588
+ },
589
+ {
590
+ "epoch": 0.05,
591
+ "grad_norm": 1.3466440439224243,
592
+ "learning_rate": 1.0418006430868168e-05,
593
+ "loss": 0.0316,
594
+ "step": 324
595
+ },
596
+ {
597
+ "epoch": 0.05,
598
+ "grad_norm": 0.7388295531272888,
599
+ "learning_rate": 1.0546623794212218e-05,
600
+ "loss": 0.0322,
601
+ "step": 328
602
+ },
603
+ {
604
+ "epoch": 0.05,
605
+ "grad_norm": 1.094037413597107,
606
+ "learning_rate": 1.0675241157556271e-05,
607
+ "loss": 0.029,
608
+ "step": 332
609
+ },
610
+ {
611
+ "epoch": 0.05,
612
+ "grad_norm": 0.9884235262870789,
613
+ "learning_rate": 1.0803858520900323e-05,
614
+ "loss": 0.0424,
615
+ "step": 336
616
+ },
617
+ {
618
+ "epoch": 0.05,
619
+ "grad_norm": 1.2799253463745117,
620
+ "learning_rate": 1.0932475884244374e-05,
621
+ "loss": 0.0221,
622
+ "step": 340
623
+ },
624
+ {
625
+ "epoch": 0.06,
626
+ "grad_norm": 0.830769419670105,
627
+ "learning_rate": 1.1061093247588426e-05,
628
+ "loss": 0.0255,
629
+ "step": 344
630
+ },
631
+ {
632
+ "epoch": 0.06,
633
+ "grad_norm": 1.4065282344818115,
634
+ "learning_rate": 1.1189710610932476e-05,
635
+ "loss": 0.0257,
636
+ "step": 348
637
+ },
638
+ {
639
+ "epoch": 0.06,
640
+ "grad_norm": 1.4823110103607178,
641
+ "learning_rate": 1.1318327974276529e-05,
642
+ "loss": 0.0277,
643
+ "step": 352
644
+ },
645
+ {
646
+ "epoch": 0.06,
647
+ "grad_norm": 0.5865018367767334,
648
+ "learning_rate": 1.144694533762058e-05,
649
+ "loss": 0.0272,
650
+ "step": 356
651
+ },
652
+ {
653
+ "epoch": 0.06,
654
+ "grad_norm": 1.3011387586593628,
655
+ "learning_rate": 1.157556270096463e-05,
656
+ "loss": 0.0249,
657
+ "step": 360
658
+ },
659
+ {
660
+ "epoch": 0.06,
661
+ "grad_norm": 1.0222949981689453,
662
+ "learning_rate": 1.1704180064308684e-05,
663
+ "loss": 0.025,
664
+ "step": 364
665
+ },
666
+ {
667
+ "epoch": 0.06,
668
+ "grad_norm": 1.650606393814087,
669
+ "learning_rate": 1.1832797427652733e-05,
670
+ "loss": 0.0427,
671
+ "step": 368
672
+ },
673
+ {
674
+ "epoch": 0.06,
675
+ "grad_norm": 7.919157981872559,
676
+ "learning_rate": 1.1961414790996787e-05,
677
+ "loss": 0.0432,
678
+ "step": 372
679
+ },
680
+ {
681
+ "epoch": 0.06,
682
+ "grad_norm": 1.2726035118103027,
683
+ "learning_rate": 1.2090032154340837e-05,
684
+ "loss": 0.0244,
685
+ "step": 376
686
+ },
687
+ {
688
+ "epoch": 0.06,
689
+ "grad_norm": 1.1458909511566162,
690
+ "learning_rate": 1.2218649517684888e-05,
691
+ "loss": 0.026,
692
+ "step": 380
693
+ },
694
+ {
695
+ "epoch": 0.06,
696
+ "grad_norm": 1.4866142272949219,
697
+ "learning_rate": 1.2347266881028941e-05,
698
+ "loss": 0.0315,
699
+ "step": 384
700
+ },
701
+ {
702
+ "epoch": 0.06,
703
+ "grad_norm": 0.7649275064468384,
704
+ "learning_rate": 1.2475884244372991e-05,
705
+ "loss": 0.0253,
706
+ "step": 388
707
+ },
708
+ {
709
+ "epoch": 0.06,
710
+ "grad_norm": 0.5921647548675537,
711
+ "learning_rate": 1.2604501607717043e-05,
712
+ "loss": 0.0229,
713
+ "step": 392
714
+ },
715
+ {
716
+ "epoch": 0.06,
717
+ "grad_norm": 0.977344810962677,
718
+ "learning_rate": 1.2733118971061094e-05,
719
+ "loss": 0.0225,
720
+ "step": 396
721
+ },
722
+ {
723
+ "epoch": 0.06,
724
+ "grad_norm": 1.8201347589492798,
725
+ "learning_rate": 1.2861736334405146e-05,
726
+ "loss": 0.0219,
727
+ "step": 400
728
+ },
729
+ {
730
+ "epoch": 0.06,
731
+ "grad_norm": 0.8397660255432129,
732
+ "learning_rate": 1.2990353697749196e-05,
733
+ "loss": 0.0281,
734
+ "step": 404
735
+ },
736
+ {
737
+ "epoch": 0.07,
738
+ "grad_norm": 0.5883716940879822,
739
+ "learning_rate": 1.3118971061093249e-05,
740
+ "loss": 0.0256,
741
+ "step": 408
742
+ },
743
+ {
744
+ "epoch": 0.07,
745
+ "grad_norm": 1.352879524230957,
746
+ "learning_rate": 1.32475884244373e-05,
747
+ "loss": 0.0323,
748
+ "step": 412
749
+ },
750
+ {
751
+ "epoch": 0.07,
752
+ "grad_norm": 2.8051841259002686,
753
+ "learning_rate": 1.337620578778135e-05,
754
+ "loss": 0.0389,
755
+ "step": 416
756
+ },
757
+ {
758
+ "epoch": 0.07,
759
+ "grad_norm": 0.8029029965400696,
760
+ "learning_rate": 1.3504823151125404e-05,
761
+ "loss": 0.0198,
762
+ "step": 420
763
+ },
764
+ {
765
+ "epoch": 0.07,
766
+ "grad_norm": 2.1253607273101807,
767
+ "learning_rate": 1.3633440514469454e-05,
768
+ "loss": 0.0303,
769
+ "step": 424
770
+ },
771
+ {
772
+ "epoch": 0.07,
773
+ "grad_norm": 1.313376784324646,
774
+ "learning_rate": 1.3762057877813507e-05,
775
+ "loss": 0.0415,
776
+ "step": 428
777
+ },
778
+ {
779
+ "epoch": 0.07,
780
+ "grad_norm": 0.5886867046356201,
781
+ "learning_rate": 1.3890675241157558e-05,
782
+ "loss": 0.028,
783
+ "step": 432
784
+ },
785
+ {
786
+ "epoch": 0.07,
787
+ "grad_norm": 1.150387167930603,
788
+ "learning_rate": 1.4019292604501608e-05,
789
+ "loss": 0.0317,
790
+ "step": 436
791
+ },
792
+ {
793
+ "epoch": 0.07,
794
+ "grad_norm": 0.9141702651977539,
795
+ "learning_rate": 1.4147909967845662e-05,
796
+ "loss": 0.0375,
797
+ "step": 440
798
+ },
799
+ {
800
+ "epoch": 0.07,
801
+ "grad_norm": 1.5728639364242554,
802
+ "learning_rate": 1.4276527331189711e-05,
803
+ "loss": 0.035,
804
+ "step": 444
805
+ },
806
+ {
807
+ "epoch": 0.07,
808
+ "grad_norm": 0.9280940294265747,
809
+ "learning_rate": 1.4405144694533763e-05,
810
+ "loss": 0.0254,
811
+ "step": 448
812
+ },
813
+ {
814
+ "epoch": 0.07,
815
+ "grad_norm": 1.4746482372283936,
816
+ "learning_rate": 1.4533762057877815e-05,
817
+ "loss": 0.0449,
818
+ "step": 452
819
+ },
820
+ {
821
+ "epoch": 0.07,
822
+ "grad_norm": 0.9735682606697083,
823
+ "learning_rate": 1.4662379421221866e-05,
824
+ "loss": 0.0305,
825
+ "step": 456
826
+ },
827
+ {
828
+ "epoch": 0.07,
829
+ "grad_norm": 1.2252638339996338,
830
+ "learning_rate": 1.479099678456592e-05,
831
+ "loss": 0.0286,
832
+ "step": 460
833
+ },
834
+ {
835
+ "epoch": 0.07,
836
+ "grad_norm": 2.308204412460327,
837
+ "learning_rate": 1.491961414790997e-05,
838
+ "loss": 0.0376,
839
+ "step": 464
840
+ },
841
+ {
842
+ "epoch": 0.08,
843
+ "grad_norm": 1.1166713237762451,
844
+ "learning_rate": 1.504823151125402e-05,
845
+ "loss": 0.0309,
846
+ "step": 468
847
+ },
848
+ {
849
+ "epoch": 0.08,
850
+ "grad_norm": 1.1420267820358276,
851
+ "learning_rate": 1.517684887459807e-05,
852
+ "loss": 0.0243,
853
+ "step": 472
854
+ },
855
+ {
856
+ "epoch": 0.08,
857
+ "grad_norm": 2.7457785606384277,
858
+ "learning_rate": 1.5305466237942124e-05,
859
+ "loss": 0.0358,
860
+ "step": 476
861
+ },
862
+ {
863
+ "epoch": 0.08,
864
+ "grad_norm": 1.1351250410079956,
865
+ "learning_rate": 1.5434083601286174e-05,
866
+ "loss": 0.0381,
867
+ "step": 480
868
+ },
869
+ {
870
+ "epoch": 0.08,
871
+ "grad_norm": 0.669029951095581,
872
+ "learning_rate": 1.5562700964630227e-05,
873
+ "loss": 0.0233,
874
+ "step": 484
875
+ },
876
+ {
877
+ "epoch": 0.08,
878
+ "grad_norm": 1.1537421941757202,
879
+ "learning_rate": 1.569131832797428e-05,
880
+ "loss": 0.0319,
881
+ "step": 488
882
+ },
883
+ {
884
+ "epoch": 0.08,
885
+ "grad_norm": 1.1718066930770874,
886
+ "learning_rate": 1.581993569131833e-05,
887
+ "loss": 0.0366,
888
+ "step": 492
889
+ },
890
+ {
891
+ "epoch": 0.08,
892
+ "grad_norm": 0.6619621515274048,
893
+ "learning_rate": 1.594855305466238e-05,
894
+ "loss": 0.0433,
895
+ "step": 496
896
+ },
897
+ {
898
+ "epoch": 0.08,
899
+ "grad_norm": 0.9330528378486633,
900
+ "learning_rate": 1.607717041800643e-05,
901
+ "loss": 0.0336,
902
+ "step": 500
903
+ },
904
+ {
905
+ "epoch": 0.08,
906
+ "grad_norm": 1.4244128465652466,
907
+ "learning_rate": 1.6205787781350483e-05,
908
+ "loss": 0.0279,
909
+ "step": 504
910
+ },
911
+ {
912
+ "epoch": 0.08,
913
+ "grad_norm": 0.719264566898346,
914
+ "learning_rate": 1.6334405144694536e-05,
915
+ "loss": 0.026,
916
+ "step": 508
917
+ },
918
+ {
919
+ "epoch": 0.08,
920
+ "grad_norm": 0.5247528553009033,
921
+ "learning_rate": 1.6463022508038586e-05,
922
+ "loss": 0.0204,
923
+ "step": 512
924
+ },
925
+ {
926
+ "epoch": 0.08,
927
+ "grad_norm": 1.9459376335144043,
928
+ "learning_rate": 1.659163987138264e-05,
929
+ "loss": 0.0457,
930
+ "step": 516
931
+ },
932
+ {
933
+ "epoch": 0.08,
934
+ "grad_norm": 0.6106892824172974,
935
+ "learning_rate": 1.672025723472669e-05,
936
+ "loss": 0.024,
937
+ "step": 520
938
+ },
939
+ {
940
+ "epoch": 0.08,
941
+ "grad_norm": 1.9702305793762207,
942
+ "learning_rate": 1.6848874598070743e-05,
943
+ "loss": 0.0424,
944
+ "step": 524
945
+ },
946
+ {
947
+ "epoch": 0.08,
948
+ "grad_norm": 0.8634403347969055,
949
+ "learning_rate": 1.6977491961414792e-05,
950
+ "loss": 0.0377,
951
+ "step": 528
952
+ },
953
+ {
954
+ "epoch": 0.09,
955
+ "grad_norm": 1.0177710056304932,
956
+ "learning_rate": 1.7106109324758842e-05,
957
+ "loss": 0.0299,
958
+ "step": 532
959
+ },
960
+ {
961
+ "epoch": 0.09,
962
+ "grad_norm": 0.7187432646751404,
963
+ "learning_rate": 1.7234726688102896e-05,
964
+ "loss": 0.03,
965
+ "step": 536
966
+ },
967
+ {
968
+ "epoch": 0.09,
969
+ "grad_norm": 1.3794456720352173,
970
+ "learning_rate": 1.7363344051446945e-05,
971
+ "loss": 0.0461,
972
+ "step": 540
973
+ },
974
+ {
975
+ "epoch": 0.09,
976
+ "grad_norm": 2.011230707168579,
977
+ "learning_rate": 1.7491961414791e-05,
978
+ "loss": 0.0423,
979
+ "step": 544
980
+ },
981
+ {
982
+ "epoch": 0.09,
983
+ "grad_norm": 0.9847524166107178,
984
+ "learning_rate": 1.762057877813505e-05,
985
+ "loss": 0.0379,
986
+ "step": 548
987
+ },
988
+ {
989
+ "epoch": 0.09,
990
+ "grad_norm": 1.4435635805130005,
991
+ "learning_rate": 1.7749196141479102e-05,
992
+ "loss": 0.0407,
993
+ "step": 552
994
+ },
995
+ {
996
+ "epoch": 0.09,
997
+ "grad_norm": 0.6206502318382263,
998
+ "learning_rate": 1.7877813504823152e-05,
999
+ "loss": 0.0293,
1000
+ "step": 556
1001
+ },
1002
+ {
1003
+ "epoch": 0.09,
1004
+ "grad_norm": 0.6152936816215515,
1005
+ "learning_rate": 1.8006430868167205e-05,
1006
+ "loss": 0.0239,
1007
+ "step": 560
1008
+ },
1009
+ {
1010
+ "epoch": 0.09,
1011
+ "grad_norm": 0.9495165944099426,
1012
+ "learning_rate": 1.8135048231511255e-05,
1013
+ "loss": 0.0279,
1014
+ "step": 564
1015
+ },
1016
+ {
1017
+ "epoch": 0.09,
1018
+ "grad_norm": 1.1117522716522217,
1019
+ "learning_rate": 1.8263665594855308e-05,
1020
+ "loss": 0.0282,
1021
+ "step": 568
1022
+ },
1023
+ {
1024
+ "epoch": 0.09,
1025
+ "grad_norm": 0.9599608182907104,
1026
+ "learning_rate": 1.8392282958199358e-05,
1027
+ "loss": 0.0241,
1028
+ "step": 572
1029
+ },
1030
+ {
1031
+ "epoch": 0.09,
1032
+ "grad_norm": 0.665824294090271,
1033
+ "learning_rate": 1.8520900321543408e-05,
1034
+ "loss": 0.032,
1035
+ "step": 576
1036
+ },
1037
+ {
1038
+ "epoch": 0.09,
1039
+ "grad_norm": 0.9927297830581665,
1040
+ "learning_rate": 1.864951768488746e-05,
1041
+ "loss": 0.0386,
1042
+ "step": 580
1043
+ },
1044
+ {
1045
+ "epoch": 0.09,
1046
+ "grad_norm": 0.8864579200744629,
1047
+ "learning_rate": 1.8778135048231514e-05,
1048
+ "loss": 0.0424,
1049
+ "step": 584
1050
+ },
1051
+ {
1052
+ "epoch": 0.09,
1053
+ "grad_norm": 0.8934502601623535,
1054
+ "learning_rate": 1.8906752411575564e-05,
1055
+ "loss": 0.0303,
1056
+ "step": 588
1057
+ },
1058
+ {
1059
+ "epoch": 0.1,
1060
+ "grad_norm": 0.7153275609016418,
1061
+ "learning_rate": 1.9035369774919617e-05,
1062
+ "loss": 0.0351,
1063
+ "step": 592
1064
+ },
1065
+ {
1066
+ "epoch": 0.1,
1067
+ "grad_norm": 0.8853140473365784,
1068
+ "learning_rate": 1.9163987138263667e-05,
1069
+ "loss": 0.0251,
1070
+ "step": 596
1071
+ },
1072
+ {
1073
+ "epoch": 0.1,
1074
+ "grad_norm": 1.535198450088501,
1075
+ "learning_rate": 1.929260450160772e-05,
1076
+ "loss": 0.039,
1077
+ "step": 600
1078
+ },
1079
+ {
1080
+ "epoch": 0.1,
1081
+ "eval_loss": 0.03215770050883293,
1082
+ "eval_pearson_cosine": 0.9396208600788933,
1083
+ "eval_pearson_dot": 0.9361396977412548,
1084
+ "eval_pearson_euclidean": 0.9366702713808518,
1085
+ "eval_pearson_manhattan": 0.935916312626658,
1086
+ "eval_pearson_max": 0.9396208600788933,
1087
+ "eval_runtime": 249.4161,
1088
+ "eval_samples_per_second": 2.005,
1089
+ "eval_spearman_cosine": 0.9531478289779426,
1090
+ "eval_spearman_dot": 0.9484099676386145,
1091
+ "eval_spearman_euclidean": 0.9506823938736226,
1092
+ "eval_spearman_manhattan": 0.9494599723678361,
1093
+ "eval_spearman_max": 0.9531478289779426,
1094
+ "eval_steps_per_second": 2.005,
1095
+ "step": 600
1096
+ },
1097
+ {
1098
+ "epoch": 0.1,
1099
+ "grad_norm": 0.6895681023597717,
1100
+ "learning_rate": 1.942122186495177e-05,
1101
+ "loss": 0.0243,
1102
+ "step": 604
1103
+ },
1104
+ {
1105
+ "epoch": 0.1,
1106
+ "grad_norm": 1.0323344469070435,
1107
+ "learning_rate": 1.954983922829582e-05,
1108
+ "loss": 0.0366,
1109
+ "step": 608
1110
+ },
1111
+ {
1112
+ "epoch": 0.1,
1113
+ "grad_norm": 0.6633996367454529,
1114
+ "learning_rate": 1.9678456591639874e-05,
1115
+ "loss": 0.0472,
1116
+ "step": 612
1117
+ },
1118
+ {
1119
+ "epoch": 0.1,
1120
+ "grad_norm": 1.3444079160690308,
1121
+ "learning_rate": 1.9807073954983923e-05,
1122
+ "loss": 0.0323,
1123
+ "step": 616
1124
+ },
1125
+ {
1126
+ "epoch": 0.1,
1127
+ "grad_norm": 0.9151845574378967,
1128
+ "learning_rate": 1.9935691318327977e-05,
1129
+ "loss": 0.0263,
1130
+ "step": 620
1131
+ },
1132
+ {
1133
+ "epoch": 0.1,
1134
+ "grad_norm": 1.1424955129623413,
1135
+ "learning_rate": 1.9992852037169407e-05,
1136
+ "loss": 0.0333,
1137
+ "step": 624
1138
+ },
1139
+ {
1140
+ "epoch": 0.1,
1141
+ "grad_norm": 1.0085678100585938,
1142
+ "learning_rate": 1.9978556111508223e-05,
1143
+ "loss": 0.0375,
1144
+ "step": 628
1145
+ },
1146
+ {
1147
+ "epoch": 0.1,
1148
+ "grad_norm": 1.071134328842163,
1149
+ "learning_rate": 1.9964260185847037e-05,
1150
+ "loss": 0.0299,
1151
+ "step": 632
1152
+ },
1153
+ {
1154
+ "epoch": 0.1,
1155
+ "grad_norm": 1.2681633234024048,
1156
+ "learning_rate": 1.9949964260185847e-05,
1157
+ "loss": 0.0317,
1158
+ "step": 636
1159
+ },
1160
+ {
1161
+ "epoch": 0.1,
1162
+ "grad_norm": 0.9661902189254761,
1163
+ "learning_rate": 1.9935668334524663e-05,
1164
+ "loss": 0.0358,
1165
+ "step": 640
1166
+ },
1167
+ {
1168
+ "epoch": 0.1,
1169
+ "grad_norm": 1.77792489528656,
1170
+ "learning_rate": 1.9921372408863477e-05,
1171
+ "loss": 0.0318,
1172
+ "step": 644
1173
+ },
1174
+ {
1175
+ "epoch": 0.1,
1176
+ "grad_norm": 0.4316706955432892,
1177
+ "learning_rate": 1.990707648320229e-05,
1178
+ "loss": 0.0288,
1179
+ "step": 648
1180
+ },
1181
+ {
1182
+ "epoch": 0.1,
1183
+ "grad_norm": 0.73268061876297,
1184
+ "learning_rate": 1.9892780557541103e-05,
1185
+ "loss": 0.0373,
1186
+ "step": 652
1187
+ },
1188
+ {
1189
+ "epoch": 0.11,
1190
+ "grad_norm": 1.6720423698425293,
1191
+ "learning_rate": 1.9878484631879917e-05,
1192
+ "loss": 0.0478,
1193
+ "step": 656
1194
+ },
1195
+ {
1196
+ "epoch": 0.11,
1197
+ "grad_norm": 0.5472896695137024,
1198
+ "learning_rate": 1.986418870621873e-05,
1199
+ "loss": 0.0169,
1200
+ "step": 660
1201
+ },
1202
+ {
1203
+ "epoch": 0.11,
1204
+ "grad_norm": 1.016210675239563,
1205
+ "learning_rate": 1.9849892780557543e-05,
1206
+ "loss": 0.0279,
1207
+ "step": 664
1208
+ },
1209
+ {
1210
+ "epoch": 0.11,
1211
+ "grad_norm": 0.6803948879241943,
1212
+ "learning_rate": 1.9835596854896357e-05,
1213
+ "loss": 0.0323,
1214
+ "step": 668
1215
+ },
1216
+ {
1217
+ "epoch": 0.11,
1218
+ "grad_norm": 0.570452868938446,
1219
+ "learning_rate": 1.982130092923517e-05,
1220
+ "loss": 0.0189,
1221
+ "step": 672
1222
+ },
1223
+ {
1224
+ "epoch": 0.11,
1225
+ "grad_norm": 1.1386935710906982,
1226
+ "learning_rate": 1.9807005003573983e-05,
1227
+ "loss": 0.0253,
1228
+ "step": 676
1229
+ },
1230
+ {
1231
+ "epoch": 0.11,
1232
+ "grad_norm": 1.8023974895477295,
1233
+ "learning_rate": 1.9792709077912797e-05,
1234
+ "loss": 0.0251,
1235
+ "step": 680
1236
+ },
1237
+ {
1238
+ "epoch": 0.11,
1239
+ "grad_norm": 1.1469191312789917,
1240
+ "learning_rate": 1.977841315225161e-05,
1241
+ "loss": 0.0247,
1242
+ "step": 684
1243
+ },
1244
+ {
1245
+ "epoch": 0.11,
1246
+ "grad_norm": 1.133340835571289,
1247
+ "learning_rate": 1.9764117226590423e-05,
1248
+ "loss": 0.0363,
1249
+ "step": 688
1250
+ },
1251
+ {
1252
+ "epoch": 0.11,
1253
+ "grad_norm": 1.3125261068344116,
1254
+ "learning_rate": 1.9749821300929237e-05,
1255
+ "loss": 0.022,
1256
+ "step": 692
1257
+ },
1258
+ {
1259
+ "epoch": 0.11,
1260
+ "grad_norm": 0.689136266708374,
1261
+ "learning_rate": 1.973552537526805e-05,
1262
+ "loss": 0.0332,
1263
+ "step": 696
1264
+ },
1265
+ {
1266
+ "epoch": 0.11,
1267
+ "grad_norm": 0.9812389016151428,
1268
+ "learning_rate": 1.9721229449606863e-05,
1269
+ "loss": 0.0309,
1270
+ "step": 700
1271
+ },
1272
+ {
1273
+ "epoch": 0.11,
1274
+ "grad_norm": 1.1285403966903687,
1275
+ "learning_rate": 1.9706933523945677e-05,
1276
+ "loss": 0.0295,
1277
+ "step": 704
1278
+ },
1279
+ {
1280
+ "epoch": 0.11,
1281
+ "grad_norm": 0.7089455723762512,
1282
+ "learning_rate": 1.969263759828449e-05,
1283
+ "loss": 0.0253,
1284
+ "step": 708
1285
+ },
1286
+ {
1287
+ "epoch": 0.11,
1288
+ "grad_norm": 0.8068435788154602,
1289
+ "learning_rate": 1.9678341672623303e-05,
1290
+ "loss": 0.0239,
1291
+ "step": 712
1292
+ },
1293
+ {
1294
+ "epoch": 0.12,
1295
+ "grad_norm": 1.2158197164535522,
1296
+ "learning_rate": 1.9664045746962117e-05,
1297
+ "loss": 0.0348,
1298
+ "step": 716
1299
+ },
1300
+ {
1301
+ "epoch": 0.12,
1302
+ "grad_norm": 1.1878087520599365,
1303
+ "learning_rate": 1.9649749821300933e-05,
1304
+ "loss": 0.0352,
1305
+ "step": 720
1306
+ },
1307
+ {
1308
+ "epoch": 0.12,
1309
+ "grad_norm": 1.019784927368164,
1310
+ "learning_rate": 1.9635453895639743e-05,
1311
+ "loss": 0.0468,
1312
+ "step": 724
1313
+ },
1314
+ {
1315
+ "epoch": 0.12,
1316
+ "grad_norm": 1.391402244567871,
1317
+ "learning_rate": 1.9621157969978557e-05,
1318
+ "loss": 0.0243,
1319
+ "step": 728
1320
+ },
1321
+ {
1322
+ "epoch": 0.12,
1323
+ "grad_norm": 1.5212937593460083,
1324
+ "learning_rate": 1.9606862044317373e-05,
1325
+ "loss": 0.0454,
1326
+ "step": 732
1327
+ },
1328
+ {
1329
+ "epoch": 0.12,
1330
+ "grad_norm": 0.7201647758483887,
1331
+ "learning_rate": 1.9592566118656183e-05,
1332
+ "loss": 0.0436,
1333
+ "step": 736
1334
+ },
1335
+ {
1336
+ "epoch": 0.12,
1337
+ "grad_norm": 1.0111092329025269,
1338
+ "learning_rate": 1.9578270192994997e-05,
1339
+ "loss": 0.0245,
1340
+ "step": 740
1341
+ },
1342
+ {
1343
+ "epoch": 0.12,
1344
+ "grad_norm": 0.8695981502532959,
1345
+ "learning_rate": 1.9563974267333813e-05,
1346
+ "loss": 0.0369,
1347
+ "step": 744
1348
+ },
1349
+ {
1350
+ "epoch": 0.12,
1351
+ "grad_norm": 0.9623269438743591,
1352
+ "learning_rate": 1.9549678341672623e-05,
1353
+ "loss": 0.0222,
1354
+ "step": 748
1355
+ },
1356
+ {
1357
+ "epoch": 0.12,
1358
+ "grad_norm": 0.6350664496421814,
1359
+ "learning_rate": 1.953538241601144e-05,
1360
+ "loss": 0.0211,
1361
+ "step": 752
1362
+ },
1363
+ {
1364
+ "epoch": 0.12,
1365
+ "grad_norm": 0.9464645981788635,
1366
+ "learning_rate": 1.9521086490350253e-05,
1367
+ "loss": 0.0416,
1368
+ "step": 756
1369
+ },
1370
+ {
1371
+ "epoch": 0.12,
1372
+ "grad_norm": 0.5733113288879395,
1373
+ "learning_rate": 1.9506790564689063e-05,
1374
+ "loss": 0.0217,
1375
+ "step": 760
1376
+ },
1377
+ {
1378
+ "epoch": 0.12,
1379
+ "grad_norm": 0.8460751175880432,
1380
+ "learning_rate": 1.949249463902788e-05,
1381
+ "loss": 0.0318,
1382
+ "step": 764
1383
+ },
1384
+ {
1385
+ "epoch": 0.12,
1386
+ "grad_norm": 1.7062273025512695,
1387
+ "learning_rate": 1.9478198713366693e-05,
1388
+ "loss": 0.0428,
1389
+ "step": 768
1390
+ },
1391
+ {
1392
+ "epoch": 0.12,
1393
+ "grad_norm": 0.8697042465209961,
1394
+ "learning_rate": 1.9463902787705507e-05,
1395
+ "loss": 0.0414,
1396
+ "step": 772
1397
+ },
1398
+ {
1399
+ "epoch": 0.12,
1400
+ "grad_norm": 1.0291727781295776,
1401
+ "learning_rate": 1.944960686204432e-05,
1402
+ "loss": 0.0288,
1403
+ "step": 776
1404
+ },
1405
+ {
1406
+ "epoch": 0.13,
1407
+ "grad_norm": 0.8506454229354858,
1408
+ "learning_rate": 1.9435310936383133e-05,
1409
+ "loss": 0.047,
1410
+ "step": 780
1411
+ },
1412
+ {
1413
+ "epoch": 0.13,
1414
+ "grad_norm": 0.9701406955718994,
1415
+ "learning_rate": 1.9421015010721947e-05,
1416
+ "loss": 0.0276,
1417
+ "step": 784
1418
+ },
1419
+ {
1420
+ "epoch": 0.13,
1421
+ "grad_norm": 2.1087899208068848,
1422
+ "learning_rate": 1.940671908506076e-05,
1423
+ "loss": 0.0448,
1424
+ "step": 788
1425
+ },
1426
+ {
1427
+ "epoch": 0.13,
1428
+ "grad_norm": 0.8718247413635254,
1429
+ "learning_rate": 1.9392423159399573e-05,
1430
+ "loss": 0.0241,
1431
+ "step": 792
1432
+ },
1433
+ {
1434
+ "epoch": 0.13,
1435
+ "grad_norm": 1.3904443979263306,
1436
+ "learning_rate": 1.9378127233738387e-05,
1437
+ "loss": 0.0451,
1438
+ "step": 796
1439
+ },
1440
+ {
1441
+ "epoch": 0.13,
1442
+ "grad_norm": 0.7410137057304382,
1443
+ "learning_rate": 1.93638313080772e-05,
1444
+ "loss": 0.0244,
1445
+ "step": 800
1446
+ },
1447
+ {
1448
+ "epoch": 0.13,
1449
+ "grad_norm": 0.9150621294975281,
1450
+ "learning_rate": 1.9349535382416013e-05,
1451
+ "loss": 0.0272,
1452
+ "step": 804
1453
+ },
1454
+ {
1455
+ "epoch": 0.13,
1456
+ "grad_norm": 0.7707653045654297,
1457
+ "learning_rate": 1.9335239456754827e-05,
1458
+ "loss": 0.036,
1459
+ "step": 808
1460
+ },
1461
+ {
1462
+ "epoch": 0.13,
1463
+ "grad_norm": 0.9467148184776306,
1464
+ "learning_rate": 1.932094353109364e-05,
1465
+ "loss": 0.0363,
1466
+ "step": 812
1467
+ },
1468
+ {
1469
+ "epoch": 0.13,
1470
+ "grad_norm": 1.2180854082107544,
1471
+ "learning_rate": 1.9306647605432453e-05,
1472
+ "loss": 0.039,
1473
+ "step": 816
1474
+ },
1475
+ {
1476
+ "epoch": 0.13,
1477
+ "grad_norm": 0.9589481949806213,
1478
+ "learning_rate": 1.9292351679771267e-05,
1479
+ "loss": 0.0346,
1480
+ "step": 820
1481
+ },
1482
+ {
1483
+ "epoch": 0.13,
1484
+ "grad_norm": 1.086827039718628,
1485
+ "learning_rate": 1.927805575411008e-05,
1486
+ "loss": 0.0352,
1487
+ "step": 824
1488
+ },
1489
+ {
1490
+ "epoch": 0.13,
1491
+ "grad_norm": 1.0733484029769897,
1492
+ "learning_rate": 1.9263759828448893e-05,
1493
+ "loss": 0.0308,
1494
+ "step": 828
1495
+ },
1496
+ {
1497
+ "epoch": 0.13,
1498
+ "grad_norm": 0.7703049182891846,
1499
+ "learning_rate": 1.9249463902787707e-05,
1500
+ "loss": 0.0271,
1501
+ "step": 832
1502
+ },
1503
+ {
1504
+ "epoch": 0.13,
1505
+ "grad_norm": 1.1060117483139038,
1506
+ "learning_rate": 1.923516797712652e-05,
1507
+ "loss": 0.0281,
1508
+ "step": 836
1509
+ },
1510
+ {
1511
+ "epoch": 0.14,
1512
+ "grad_norm": 0.7313349843025208,
1513
+ "learning_rate": 1.9220872051465333e-05,
1514
+ "loss": 0.0245,
1515
+ "step": 840
1516
+ },
1517
+ {
1518
+ "epoch": 0.14,
1519
+ "grad_norm": 0.7229084372520447,
1520
+ "learning_rate": 1.920657612580415e-05,
1521
+ "loss": 0.0249,
1522
+ "step": 844
1523
+ },
1524
+ {
1525
+ "epoch": 0.14,
1526
+ "grad_norm": 1.4665247201919556,
1527
+ "learning_rate": 1.919228020014296e-05,
1528
+ "loss": 0.0412,
1529
+ "step": 848
1530
+ },
1531
+ {
1532
+ "epoch": 0.14,
1533
+ "grad_norm": 0.8117924928665161,
1534
+ "learning_rate": 1.9177984274481773e-05,
1535
+ "loss": 0.0391,
1536
+ "step": 852
1537
+ },
1538
+ {
1539
+ "epoch": 0.14,
1540
+ "grad_norm": 0.9442553520202637,
1541
+ "learning_rate": 1.916368834882059e-05,
1542
+ "loss": 0.0328,
1543
+ "step": 856
1544
+ },
1545
+ {
1546
+ "epoch": 0.14,
1547
+ "grad_norm": 0.7885982394218445,
1548
+ "learning_rate": 1.91493924231594e-05,
1549
+ "loss": 0.0182,
1550
+ "step": 860
1551
+ },
1552
+ {
1553
+ "epoch": 0.14,
1554
+ "grad_norm": 1.2472141981124878,
1555
+ "learning_rate": 1.9135096497498217e-05,
1556
+ "loss": 0.0342,
1557
+ "step": 864
1558
+ },
1559
+ {
1560
+ "epoch": 0.14,
1561
+ "grad_norm": 0.8535823225975037,
1562
+ "learning_rate": 1.9120800571837027e-05,
1563
+ "loss": 0.0223,
1564
+ "step": 868
1565
+ },
1566
+ {
1567
+ "epoch": 0.14,
1568
+ "grad_norm": 1.84065842628479,
1569
+ "learning_rate": 1.910650464617584e-05,
1570
+ "loss": 0.0359,
1571
+ "step": 872
1572
+ },
1573
+ {
1574
+ "epoch": 0.14,
1575
+ "grad_norm": 1.0987244844436646,
1576
+ "learning_rate": 1.9092208720514657e-05,
1577
+ "loss": 0.0343,
1578
+ "step": 876
1579
+ },
1580
+ {
1581
+ "epoch": 0.14,
1582
+ "grad_norm": 0.7632778286933899,
1583
+ "learning_rate": 1.9077912794853467e-05,
1584
+ "loss": 0.0323,
1585
+ "step": 880
1586
+ },
1587
+ {
1588
+ "epoch": 0.14,
1589
+ "grad_norm": 1.1485552787780762,
1590
+ "learning_rate": 1.906361686919228e-05,
1591
+ "loss": 0.0337,
1592
+ "step": 884
1593
+ },
1594
+ {
1595
+ "epoch": 0.14,
1596
+ "grad_norm": 2.6315150260925293,
1597
+ "learning_rate": 1.9049320943531097e-05,
1598
+ "loss": 0.0515,
1599
+ "step": 888
1600
+ },
1601
+ {
1602
+ "epoch": 0.14,
1603
+ "grad_norm": 1.0435465574264526,
1604
+ "learning_rate": 1.9035025017869907e-05,
1605
+ "loss": 0.0458,
1606
+ "step": 892
1607
+ },
1608
+ {
1609
+ "epoch": 0.14,
1610
+ "grad_norm": 1.001620888710022,
1611
+ "learning_rate": 1.9020729092208723e-05,
1612
+ "loss": 0.0357,
1613
+ "step": 896
1614
+ },
1615
+ {
1616
+ "epoch": 0.14,
1617
+ "grad_norm": 0.8570252060890198,
1618
+ "learning_rate": 1.9006433166547537e-05,
1619
+ "loss": 0.037,
1620
+ "step": 900
1621
+ },
1622
+ {
1623
+ "epoch": 0.14,
1624
+ "eval_loss": 0.04232440143823624,
1625
+ "eval_pearson_cosine": 0.9451831743059702,
1626
+ "eval_pearson_dot": 0.9441977144080326,
1627
+ "eval_pearson_euclidean": 0.9401266241811703,
1628
+ "eval_pearson_manhattan": 0.940026238812949,
1629
+ "eval_pearson_max": 0.9451831743059702,
1630
+ "eval_runtime": 249.3889,
1631
+ "eval_samples_per_second": 2.005,
1632
+ "eval_spearman_cosine": 0.9564561958788276,
1633
+ "eval_spearman_dot": 0.9527831233353083,
1634
+ "eval_spearman_euclidean": 0.9533796700221547,
1635
+ "eval_spearman_manhattan": 0.9531501329883199,
1636
+ "eval_spearman_max": 0.9564561958788276,
1637
+ "eval_steps_per_second": 2.005,
1638
+ "step": 900
1639
+ },
1640
+ {
1641
+ "epoch": 0.15,
1642
+ "grad_norm": 1.118642807006836,
1643
+ "learning_rate": 1.8992137240886347e-05,
1644
+ "loss": 0.0424,
1645
+ "step": 904
1646
+ },
1647
+ {
1648
+ "epoch": 0.15,
1649
+ "grad_norm": 1.344506025314331,
1650
+ "learning_rate": 1.8977841315225163e-05,
1651
+ "loss": 0.03,
1652
+ "step": 908
1653
+ },
1654
+ {
1655
+ "epoch": 0.15,
1656
+ "grad_norm": 0.7447184920310974,
1657
+ "learning_rate": 1.8963545389563977e-05,
1658
+ "loss": 0.0308,
1659
+ "step": 912
1660
+ },
1661
+ {
1662
+ "epoch": 0.15,
1663
+ "grad_norm": 0.6138767004013062,
1664
+ "learning_rate": 1.894924946390279e-05,
1665
+ "loss": 0.0314,
1666
+ "step": 916
1667
+ },
1668
+ {
1669
+ "epoch": 0.15,
1670
+ "grad_norm": 0.7356119751930237,
1671
+ "learning_rate": 1.8934953538241603e-05,
1672
+ "loss": 0.0283,
1673
+ "step": 920
1674
+ },
1675
+ {
1676
+ "epoch": 0.15,
1677
+ "grad_norm": 1.263514518737793,
1678
+ "learning_rate": 1.8920657612580417e-05,
1679
+ "loss": 0.0293,
1680
+ "step": 924
1681
+ },
1682
+ {
1683
+ "epoch": 0.15,
1684
+ "grad_norm": 0.6265131235122681,
1685
+ "learning_rate": 1.890636168691923e-05,
1686
+ "loss": 0.0183,
1687
+ "step": 928
1688
+ },
1689
+ {
1690
+ "epoch": 0.15,
1691
+ "grad_norm": 1.7762928009033203,
1692
+ "learning_rate": 1.8892065761258043e-05,
1693
+ "loss": 0.0449,
1694
+ "step": 932
1695
+ },
1696
+ {
1697
+ "epoch": 0.15,
1698
+ "grad_norm": 1.0329370498657227,
1699
+ "learning_rate": 1.8877769835596857e-05,
1700
+ "loss": 0.0397,
1701
+ "step": 936
1702
+ },
1703
+ {
1704
+ "epoch": 0.15,
1705
+ "grad_norm": 1.2647181749343872,
1706
+ "learning_rate": 1.886347390993567e-05,
1707
+ "loss": 0.0329,
1708
+ "step": 940
1709
+ },
1710
+ {
1711
+ "epoch": 0.15,
1712
+ "grad_norm": 0.7577453851699829,
1713
+ "learning_rate": 1.8849177984274483e-05,
1714
+ "loss": 0.043,
1715
+ "step": 944
1716
+ },
1717
+ {
1718
+ "epoch": 0.15,
1719
+ "grad_norm": 1.1184148788452148,
1720
+ "learning_rate": 1.8834882058613297e-05,
1721
+ "loss": 0.0419,
1722
+ "step": 948
1723
+ },
1724
+ {
1725
+ "epoch": 0.15,
1726
+ "grad_norm": 0.7198161482810974,
1727
+ "learning_rate": 1.882058613295211e-05,
1728
+ "loss": 0.0292,
1729
+ "step": 952
1730
+ },
1731
+ {
1732
+ "epoch": 0.15,
1733
+ "grad_norm": 1.0045865774154663,
1734
+ "learning_rate": 1.8806290207290923e-05,
1735
+ "loss": 0.0333,
1736
+ "step": 956
1737
+ },
1738
+ {
1739
+ "epoch": 0.15,
1740
+ "grad_norm": 0.5983513593673706,
1741
+ "learning_rate": 1.8791994281629737e-05,
1742
+ "loss": 0.022,
1743
+ "step": 960
1744
+ },
1745
+ {
1746
+ "epoch": 0.16,
1747
+ "grad_norm": 1.0982924699783325,
1748
+ "learning_rate": 1.877769835596855e-05,
1749
+ "loss": 0.0264,
1750
+ "step": 964
1751
+ },
1752
+ {
1753
+ "epoch": 0.16,
1754
+ "grad_norm": 0.6120471358299255,
1755
+ "learning_rate": 1.8763402430307363e-05,
1756
+ "loss": 0.0308,
1757
+ "step": 968
1758
+ },
1759
+ {
1760
+ "epoch": 0.16,
1761
+ "grad_norm": 1.5346020460128784,
1762
+ "learning_rate": 1.8749106504646177e-05,
1763
+ "loss": 0.0355,
1764
+ "step": 972
1765
+ },
1766
+ {
1767
+ "epoch": 0.16,
1768
+ "grad_norm": 0.42815306782722473,
1769
+ "learning_rate": 1.873481057898499e-05,
1770
+ "loss": 0.0236,
1771
+ "step": 976
1772
+ },
1773
+ {
1774
+ "epoch": 0.16,
1775
+ "grad_norm": 0.7016882300376892,
1776
+ "learning_rate": 1.8720514653323803e-05,
1777
+ "loss": 0.0278,
1778
+ "step": 980
1779
+ },
1780
+ {
1781
+ "epoch": 0.16,
1782
+ "grad_norm": 0.6955506801605225,
1783
+ "learning_rate": 1.8706218727662617e-05,
1784
+ "loss": 0.0208,
1785
+ "step": 984
1786
+ },
1787
+ {
1788
+ "epoch": 0.16,
1789
+ "grad_norm": 0.7579104900360107,
1790
+ "learning_rate": 1.8691922802001433e-05,
1791
+ "loss": 0.0303,
1792
+ "step": 988
1793
+ },
1794
+ {
1795
+ "epoch": 0.16,
1796
+ "grad_norm": 0.8700461983680725,
1797
+ "learning_rate": 1.8677626876340243e-05,
1798
+ "loss": 0.0212,
1799
+ "step": 992
1800
+ },
1801
+ {
1802
+ "epoch": 0.16,
1803
+ "grad_norm": 1.532637357711792,
1804
+ "learning_rate": 1.8663330950679057e-05,
1805
+ "loss": 0.0499,
1806
+ "step": 996
1807
+ },
1808
+ {
1809
+ "epoch": 0.16,
1810
+ "grad_norm": 1.3528228998184204,
1811
+ "learning_rate": 1.8649035025017873e-05,
1812
+ "loss": 0.0349,
1813
+ "step": 1000
1814
+ },
1815
+ {
1816
+ "epoch": 0.16,
1817
+ "grad_norm": 0.7273157238960266,
1818
+ "learning_rate": 1.8634739099356683e-05,
1819
+ "loss": 0.0202,
1820
+ "step": 1004
1821
+ },
1822
+ {
1823
+ "epoch": 0.16,
1824
+ "grad_norm": 0.952356219291687,
1825
+ "learning_rate": 1.86204431736955e-05,
1826
+ "loss": 0.0315,
1827
+ "step": 1008
1828
+ },
1829
+ {
1830
+ "epoch": 0.16,
1831
+ "grad_norm": 0.6654635071754456,
1832
+ "learning_rate": 1.8606147248034313e-05,
1833
+ "loss": 0.0248,
1834
+ "step": 1012
1835
+ },
1836
+ {
1837
+ "epoch": 0.16,
1838
+ "grad_norm": 0.578262984752655,
1839
+ "learning_rate": 1.8591851322373123e-05,
1840
+ "loss": 0.037,
1841
+ "step": 1016
1842
+ },
1843
+ {
1844
+ "epoch": 0.16,
1845
+ "grad_norm": 0.6199663877487183,
1846
+ "learning_rate": 1.857755539671194e-05,
1847
+ "loss": 0.0245,
1848
+ "step": 1020
1849
+ },
1850
+ {
1851
+ "epoch": 0.16,
1852
+ "grad_norm": 1.0458414554595947,
1853
+ "learning_rate": 1.8563259471050753e-05,
1854
+ "loss": 0.0264,
1855
+ "step": 1024
1856
+ },
1857
+ {
1858
+ "epoch": 0.17,
1859
+ "grad_norm": 0.5218886733055115,
1860
+ "learning_rate": 1.8548963545389563e-05,
1861
+ "loss": 0.0254,
1862
+ "step": 1028
1863
+ },
1864
+ {
1865
+ "epoch": 0.17,
1866
+ "grad_norm": 0.4931807518005371,
1867
+ "learning_rate": 1.853466761972838e-05,
1868
+ "loss": 0.0163,
1869
+ "step": 1032
1870
+ },
1871
+ {
1872
+ "epoch": 0.17,
1873
+ "grad_norm": 0.7079238891601562,
1874
+ "learning_rate": 1.8520371694067193e-05,
1875
+ "loss": 0.027,
1876
+ "step": 1036
1877
+ },
1878
+ {
1879
+ "epoch": 0.17,
1880
+ "grad_norm": 0.6107800006866455,
1881
+ "learning_rate": 1.8506075768406007e-05,
1882
+ "loss": 0.0353,
1883
+ "step": 1040
1884
+ },
1885
+ {
1886
+ "epoch": 0.17,
1887
+ "grad_norm": 1.5169730186462402,
1888
+ "learning_rate": 1.849177984274482e-05,
1889
+ "loss": 0.0353,
1890
+ "step": 1044
1891
+ },
1892
+ {
1893
+ "epoch": 0.17,
1894
+ "grad_norm": 0.6394426822662354,
1895
+ "learning_rate": 1.8477483917083633e-05,
1896
+ "loss": 0.0175,
1897
+ "step": 1048
1898
+ },
1899
+ {
1900
+ "epoch": 0.17,
1901
+ "grad_norm": 0.5369437336921692,
1902
+ "learning_rate": 1.8463187991422447e-05,
1903
+ "loss": 0.0456,
1904
+ "step": 1052
1905
+ },
1906
+ {
1907
+ "epoch": 0.17,
1908
+ "grad_norm": 2.259932041168213,
1909
+ "learning_rate": 1.844889206576126e-05,
1910
+ "loss": 0.0573,
1911
+ "step": 1056
1912
+ },
1913
+ {
1914
+ "epoch": 0.17,
1915
+ "grad_norm": 0.5754424333572388,
1916
+ "learning_rate": 1.8434596140100073e-05,
1917
+ "loss": 0.0208,
1918
+ "step": 1060
1919
+ },
1920
+ {
1921
+ "epoch": 0.17,
1922
+ "grad_norm": 0.48612886667251587,
1923
+ "learning_rate": 1.8420300214438887e-05,
1924
+ "loss": 0.0181,
1925
+ "step": 1064
1926
+ },
1927
+ {
1928
+ "epoch": 0.17,
1929
+ "grad_norm": 0.8483503460884094,
1930
+ "learning_rate": 1.84060042887777e-05,
1931
+ "loss": 0.0212,
1932
+ "step": 1068
1933
+ },
1934
+ {
1935
+ "epoch": 0.17,
1936
+ "grad_norm": 0.9669589996337891,
1937
+ "learning_rate": 1.8391708363116513e-05,
1938
+ "loss": 0.0355,
1939
+ "step": 1072
1940
+ },
1941
+ {
1942
+ "epoch": 0.17,
1943
+ "grad_norm": 0.6289377212524414,
1944
+ "learning_rate": 1.8377412437455326e-05,
1945
+ "loss": 0.0288,
1946
+ "step": 1076
1947
+ },
1948
+ {
1949
+ "epoch": 0.17,
1950
+ "grad_norm": 1.055770993232727,
1951
+ "learning_rate": 1.836311651179414e-05,
1952
+ "loss": 0.0321,
1953
+ "step": 1080
1954
+ },
1955
+ {
1956
+ "epoch": 0.17,
1957
+ "grad_norm": 0.5941286683082581,
1958
+ "learning_rate": 1.8348820586132953e-05,
1959
+ "loss": 0.0215,
1960
+ "step": 1084
1961
+ },
1962
+ {
1963
+ "epoch": 0.17,
1964
+ "grad_norm": 0.7000011801719666,
1965
+ "learning_rate": 1.8334524660471766e-05,
1966
+ "loss": 0.0295,
1967
+ "step": 1088
1968
+ },
1969
+ {
1970
+ "epoch": 0.18,
1971
+ "grad_norm": 0.5845941305160522,
1972
+ "learning_rate": 1.832022873481058e-05,
1973
+ "loss": 0.0257,
1974
+ "step": 1092
1975
+ },
1976
+ {
1977
+ "epoch": 0.18,
1978
+ "grad_norm": 1.1016316413879395,
1979
+ "learning_rate": 1.8305932809149393e-05,
1980
+ "loss": 0.0276,
1981
+ "step": 1096
1982
+ },
1983
+ {
1984
+ "epoch": 0.18,
1985
+ "grad_norm": 0.8285301327705383,
1986
+ "learning_rate": 1.8291636883488206e-05,
1987
+ "loss": 0.0241,
1988
+ "step": 1100
1989
+ },
1990
+ {
1991
+ "epoch": 0.18,
1992
+ "grad_norm": 0.8674764037132263,
1993
+ "learning_rate": 1.827734095782702e-05,
1994
+ "loss": 0.0274,
1995
+ "step": 1104
1996
+ },
1997
+ {
1998
+ "epoch": 0.18,
1999
+ "grad_norm": 0.654329240322113,
2000
+ "learning_rate": 1.8263045032165833e-05,
2001
+ "loss": 0.0213,
2002
+ "step": 1108
2003
+ },
2004
+ {
2005
+ "epoch": 0.18,
2006
+ "grad_norm": 0.6802071928977966,
2007
+ "learning_rate": 1.824874910650465e-05,
2008
+ "loss": 0.0321,
2009
+ "step": 1112
2010
+ },
2011
+ {
2012
+ "epoch": 0.18,
2013
+ "grad_norm": 0.7723608016967773,
2014
+ "learning_rate": 1.823445318084346e-05,
2015
+ "loss": 0.0289,
2016
+ "step": 1116
2017
+ },
2018
+ {
2019
+ "epoch": 0.18,
2020
+ "grad_norm": 0.9586684703826904,
2021
+ "learning_rate": 1.8220157255182273e-05,
2022
+ "loss": 0.0264,
2023
+ "step": 1120
2024
+ },
2025
+ {
2026
+ "epoch": 0.18,
2027
+ "grad_norm": 0.758579432964325,
2028
+ "learning_rate": 1.820586132952109e-05,
2029
+ "loss": 0.0278,
2030
+ "step": 1124
2031
+ },
2032
+ {
2033
+ "epoch": 0.18,
2034
+ "grad_norm": 1.0153886079788208,
2035
+ "learning_rate": 1.81915654038599e-05,
2036
+ "loss": 0.0305,
2037
+ "step": 1128
2038
+ },
2039
+ {
2040
+ "epoch": 0.18,
2041
+ "grad_norm": 1.1771838665008545,
2042
+ "learning_rate": 1.8177269478198717e-05,
2043
+ "loss": 0.0512,
2044
+ "step": 1132
2045
+ },
2046
+ {
2047
+ "epoch": 0.18,
2048
+ "grad_norm": 1.2490832805633545,
2049
+ "learning_rate": 1.816297355253753e-05,
2050
+ "loss": 0.0295,
2051
+ "step": 1136
2052
+ },
2053
+ {
2054
+ "epoch": 0.18,
2055
+ "grad_norm": 0.5716216564178467,
2056
+ "learning_rate": 1.814867762687634e-05,
2057
+ "loss": 0.0246,
2058
+ "step": 1140
2059
+ },
2060
+ {
2061
+ "epoch": 0.18,
2062
+ "grad_norm": 0.7177873253822327,
2063
+ "learning_rate": 1.8134381701215157e-05,
2064
+ "loss": 0.0293,
2065
+ "step": 1144
2066
+ },
2067
+ {
2068
+ "epoch": 0.18,
2069
+ "grad_norm": 0.5648506879806519,
2070
+ "learning_rate": 1.812008577555397e-05,
2071
+ "loss": 0.0225,
2072
+ "step": 1148
2073
+ },
2074
+ {
2075
+ "epoch": 0.19,
2076
+ "grad_norm": 0.721409261226654,
2077
+ "learning_rate": 1.8105789849892783e-05,
2078
+ "loss": 0.0242,
2079
+ "step": 1152
2080
+ },
2081
+ {
2082
+ "epoch": 0.19,
2083
+ "grad_norm": 0.6926946640014648,
2084
+ "learning_rate": 1.8091493924231596e-05,
2085
+ "loss": 0.0276,
2086
+ "step": 1156
2087
+ },
2088
+ {
2089
+ "epoch": 0.19,
2090
+ "grad_norm": 0.6999531388282776,
2091
+ "learning_rate": 1.807719799857041e-05,
2092
+ "loss": 0.0187,
2093
+ "step": 1160
2094
+ },
2095
+ {
2096
+ "epoch": 0.19,
2097
+ "grad_norm": 0.7933658957481384,
2098
+ "learning_rate": 1.8062902072909223e-05,
2099
+ "loss": 0.0244,
2100
+ "step": 1164
2101
+ },
2102
+ {
2103
+ "epoch": 0.19,
2104
+ "grad_norm": 0.4972588121891022,
2105
+ "learning_rate": 1.8048606147248036e-05,
2106
+ "loss": 0.0418,
2107
+ "step": 1168
2108
+ },
2109
+ {
2110
+ "epoch": 0.19,
2111
+ "grad_norm": 1.0972784757614136,
2112
+ "learning_rate": 1.803431022158685e-05,
2113
+ "loss": 0.0319,
2114
+ "step": 1172
2115
+ },
2116
+ {
2117
+ "epoch": 0.19,
2118
+ "grad_norm": 1.2647324800491333,
2119
+ "learning_rate": 1.8020014295925663e-05,
2120
+ "loss": 0.0325,
2121
+ "step": 1176
2122
+ },
2123
+ {
2124
+ "epoch": 0.19,
2125
+ "grad_norm": 0.8409667015075684,
2126
+ "learning_rate": 1.8005718370264476e-05,
2127
+ "loss": 0.0236,
2128
+ "step": 1180
2129
+ },
2130
+ {
2131
+ "epoch": 0.19,
2132
+ "grad_norm": 0.6574044227600098,
2133
+ "learning_rate": 1.799142244460329e-05,
2134
+ "loss": 0.032,
2135
+ "step": 1184
2136
+ },
2137
+ {
2138
+ "epoch": 0.19,
2139
+ "grad_norm": 0.5404456257820129,
2140
+ "learning_rate": 1.7977126518942103e-05,
2141
+ "loss": 0.0313,
2142
+ "step": 1188
2143
+ },
2144
+ {
2145
+ "epoch": 0.19,
2146
+ "grad_norm": 0.7715393900871277,
2147
+ "learning_rate": 1.7962830593280916e-05,
2148
+ "loss": 0.0252,
2149
+ "step": 1192
2150
+ },
2151
+ {
2152
+ "epoch": 0.19,
2153
+ "grad_norm": 1.3372646570205688,
2154
+ "learning_rate": 1.794853466761973e-05,
2155
+ "loss": 0.0268,
2156
+ "step": 1196
2157
+ },
2158
+ {
2159
+ "epoch": 0.19,
2160
+ "grad_norm": 0.5368986129760742,
2161
+ "learning_rate": 1.7934238741958543e-05,
2162
+ "loss": 0.0234,
2163
+ "step": 1200
2164
+ },
2165
+ {
2166
+ "epoch": 0.19,
2167
+ "eval_loss": 0.026826824992895126,
2168
+ "eval_pearson_cosine": 0.9529420998918104,
2169
+ "eval_pearson_dot": 0.951139070288119,
2170
+ "eval_pearson_euclidean": 0.945750278447336,
2171
+ "eval_pearson_manhattan": 0.945116677167893,
2172
+ "eval_pearson_max": 0.9529420998918104,
2173
+ "eval_runtime": 249.7054,
2174
+ "eval_samples_per_second": 2.002,
2175
+ "eval_spearman_cosine": 0.9627865603907902,
2176
+ "eval_spearman_dot": 0.9601426124824484,
2177
+ "eval_spearman_euclidean": 0.9601416524781244,
2178
+ "eval_spearman_manhattan": 0.9589273910090907,
2179
+ "eval_spearman_max": 0.9627865603907902,
2180
+ "eval_steps_per_second": 2.002,
2181
+ "step": 1200
2182
+ },
2183
+ {
2184
+ "epoch": 0.19,
2185
+ "grad_norm": 0.749082624912262,
2186
+ "learning_rate": 1.7919942816297356e-05,
2187
+ "loss": 0.0222,
2188
+ "step": 1204
2189
+ },
2190
+ {
2191
+ "epoch": 0.19,
2192
+ "grad_norm": 0.8751634955406189,
2193
+ "learning_rate": 1.790564689063617e-05,
2194
+ "loss": 0.0267,
2195
+ "step": 1208
2196
+ },
2197
+ {
2198
+ "epoch": 0.19,
2199
+ "grad_norm": 0.622231125831604,
2200
+ "learning_rate": 1.7891350964974983e-05,
2201
+ "loss": 0.02,
2202
+ "step": 1212
2203
+ },
2204
+ {
2205
+ "epoch": 0.2,
2206
+ "grad_norm": 1.1356748342514038,
2207
+ "learning_rate": 1.7877055039313796e-05,
2208
+ "loss": 0.0286,
2209
+ "step": 1216
2210
+ },
2211
+ {
2212
+ "epoch": 0.2,
2213
+ "grad_norm": 0.6375044584274292,
2214
+ "learning_rate": 1.786275911365261e-05,
2215
+ "loss": 0.024,
2216
+ "step": 1220
2217
+ },
2218
+ {
2219
+ "epoch": 0.2,
2220
+ "grad_norm": 0.9166258573532104,
2221
+ "learning_rate": 1.7848463187991427e-05,
2222
+ "loss": 0.0285,
2223
+ "step": 1224
2224
+ },
2225
+ {
2226
+ "epoch": 0.2,
2227
+ "grad_norm": 0.7663798332214355,
2228
+ "learning_rate": 1.7834167262330236e-05,
2229
+ "loss": 0.0188,
2230
+ "step": 1228
2231
+ },
2232
+ {
2233
+ "epoch": 0.2,
2234
+ "grad_norm": 1.0997885465621948,
2235
+ "learning_rate": 1.781987133666905e-05,
2236
+ "loss": 0.0345,
2237
+ "step": 1232
2238
+ },
2239
+ {
2240
+ "epoch": 0.2,
2241
+ "grad_norm": 0.8155802488327026,
2242
+ "learning_rate": 1.7805575411007866e-05,
2243
+ "loss": 0.0291,
2244
+ "step": 1236
2245
+ },
2246
+ {
2247
+ "epoch": 0.2,
2248
+ "grad_norm": 0.5071162581443787,
2249
+ "learning_rate": 1.7791279485346676e-05,
2250
+ "loss": 0.0225,
2251
+ "step": 1240
2252
+ },
2253
+ {
2254
+ "epoch": 0.2,
2255
+ "grad_norm": 0.4950205981731415,
2256
+ "learning_rate": 1.777698355968549e-05,
2257
+ "loss": 0.0255,
2258
+ "step": 1244
2259
+ },
2260
+ {
2261
+ "epoch": 0.2,
2262
+ "grad_norm": 0.6503371000289917,
2263
+ "learning_rate": 1.7762687634024306e-05,
2264
+ "loss": 0.0353,
2265
+ "step": 1248
2266
+ },
2267
+ {
2268
+ "epoch": 0.2,
2269
+ "grad_norm": 0.4491406977176666,
2270
+ "learning_rate": 1.7748391708363116e-05,
2271
+ "loss": 0.0287,
2272
+ "step": 1252
2273
+ },
2274
+ {
2275
+ "epoch": 0.2,
2276
+ "grad_norm": 0.5856379270553589,
2277
+ "learning_rate": 1.7734095782701933e-05,
2278
+ "loss": 0.029,
2279
+ "step": 1256
2280
+ },
2281
+ {
2282
+ "epoch": 0.2,
2283
+ "grad_norm": 0.9508484601974487,
2284
+ "learning_rate": 1.7719799857040746e-05,
2285
+ "loss": 0.02,
2286
+ "step": 1260
2287
+ },
2288
+ {
2289
+ "epoch": 0.2,
2290
+ "grad_norm": 1.2820254564285278,
2291
+ "learning_rate": 1.7705503931379556e-05,
2292
+ "loss": 0.0323,
2293
+ "step": 1264
2294
+ },
2295
+ {
2296
+ "epoch": 0.2,
2297
+ "grad_norm": 0.885857105255127,
2298
+ "learning_rate": 1.7691208005718373e-05,
2299
+ "loss": 0.0241,
2300
+ "step": 1268
2301
+ },
2302
+ {
2303
+ "epoch": 0.2,
2304
+ "grad_norm": 0.5690828561782837,
2305
+ "learning_rate": 1.7676912080057183e-05,
2306
+ "loss": 0.0205,
2307
+ "step": 1272
2308
+ },
2309
+ {
2310
+ "epoch": 0.21,
2311
+ "grad_norm": 0.41895803809165955,
2312
+ "learning_rate": 1.7662616154396e-05,
2313
+ "loss": 0.0188,
2314
+ "step": 1276
2315
+ },
2316
+ {
2317
+ "epoch": 0.21,
2318
+ "grad_norm": 0.8826864361763,
2319
+ "learning_rate": 1.7648320228734813e-05,
2320
+ "loss": 0.026,
2321
+ "step": 1280
2322
+ },
2323
+ {
2324
+ "epoch": 0.21,
2325
+ "grad_norm": 0.9115862250328064,
2326
+ "learning_rate": 1.7634024303073623e-05,
2327
+ "loss": 0.0328,
2328
+ "step": 1284
2329
+ },
2330
+ {
2331
+ "epoch": 0.21,
2332
+ "grad_norm": 1.5499801635742188,
2333
+ "learning_rate": 1.761972837741244e-05,
2334
+ "loss": 0.0385,
2335
+ "step": 1288
2336
+ },
2337
+ {
2338
+ "epoch": 0.21,
2339
+ "grad_norm": 1.07961106300354,
2340
+ "learning_rate": 1.7605432451751253e-05,
2341
+ "loss": 0.0276,
2342
+ "step": 1292
2343
+ },
2344
+ {
2345
+ "epoch": 0.21,
2346
+ "grad_norm": 1.7965257167816162,
2347
+ "learning_rate": 1.7591136526090066e-05,
2348
+ "loss": 0.0305,
2349
+ "step": 1296
2350
+ },
2351
+ {
2352
+ "epoch": 0.21,
2353
+ "grad_norm": 0.47339317202568054,
2354
+ "learning_rate": 1.757684060042888e-05,
2355
+ "loss": 0.0181,
2356
+ "step": 1300
2357
+ },
2358
+ {
2359
+ "epoch": 0.21,
2360
+ "grad_norm": 0.6750252842903137,
2361
+ "learning_rate": 1.7562544674767693e-05,
2362
+ "loss": 0.0208,
2363
+ "step": 1304
2364
+ },
2365
+ {
2366
+ "epoch": 0.21,
2367
+ "grad_norm": 0.9396247267723083,
2368
+ "learning_rate": 1.7548248749106506e-05,
2369
+ "loss": 0.0361,
2370
+ "step": 1308
2371
+ },
2372
+ {
2373
+ "epoch": 0.21,
2374
+ "grad_norm": 1.0881524085998535,
2375
+ "learning_rate": 1.753395282344532e-05,
2376
+ "loss": 0.0331,
2377
+ "step": 1312
2378
+ },
2379
+ {
2380
+ "epoch": 0.21,
2381
+ "grad_norm": 0.7517051100730896,
2382
+ "learning_rate": 1.7519656897784133e-05,
2383
+ "loss": 0.0281,
2384
+ "step": 1316
2385
+ },
2386
+ {
2387
+ "epoch": 0.21,
2388
+ "grad_norm": 0.7083280682563782,
2389
+ "learning_rate": 1.7505360972122946e-05,
2390
+ "loss": 0.0209,
2391
+ "step": 1320
2392
+ },
2393
+ {
2394
+ "epoch": 0.21,
2395
+ "grad_norm": 0.727603018283844,
2396
+ "learning_rate": 1.749106504646176e-05,
2397
+ "loss": 0.0312,
2398
+ "step": 1324
2399
+ },
2400
+ {
2401
+ "epoch": 0.21,
2402
+ "grad_norm": 0.4598117470741272,
2403
+ "learning_rate": 1.7476769120800573e-05,
2404
+ "loss": 0.02,
2405
+ "step": 1328
2406
+ },
2407
+ {
2408
+ "epoch": 0.21,
2409
+ "grad_norm": 0.6653364896774292,
2410
+ "learning_rate": 1.7462473195139386e-05,
2411
+ "loss": 0.0263,
2412
+ "step": 1332
2413
+ },
2414
+ {
2415
+ "epoch": 0.21,
2416
+ "grad_norm": 0.527958869934082,
2417
+ "learning_rate": 1.74481772694782e-05,
2418
+ "loss": 0.0176,
2419
+ "step": 1336
2420
+ },
2421
+ {
2422
+ "epoch": 0.22,
2423
+ "grad_norm": 0.9711959362030029,
2424
+ "learning_rate": 1.7433881343817013e-05,
2425
+ "loss": 0.0336,
2426
+ "step": 1340
2427
+ },
2428
+ {
2429
+ "epoch": 0.22,
2430
+ "grad_norm": 0.8734799027442932,
2431
+ "learning_rate": 1.7419585418155826e-05,
2432
+ "loss": 0.0384,
2433
+ "step": 1344
2434
+ },
2435
+ {
2436
+ "epoch": 0.22,
2437
+ "grad_norm": 0.599764883518219,
2438
+ "learning_rate": 1.740528949249464e-05,
2439
+ "loss": 0.0317,
2440
+ "step": 1348
2441
+ },
2442
+ {
2443
+ "epoch": 0.22,
2444
+ "grad_norm": 0.6115812659263611,
2445
+ "learning_rate": 1.7390993566833453e-05,
2446
+ "loss": 0.0246,
2447
+ "step": 1352
2448
+ },
2449
+ {
2450
+ "epoch": 0.22,
2451
+ "grad_norm": 0.660057008266449,
2452
+ "learning_rate": 1.7376697641172266e-05,
2453
+ "loss": 0.027,
2454
+ "step": 1356
2455
+ },
2456
+ {
2457
+ "epoch": 0.22,
2458
+ "grad_norm": 1.2455826997756958,
2459
+ "learning_rate": 1.736240171551108e-05,
2460
+ "loss": 0.0434,
2461
+ "step": 1360
2462
+ },
2463
+ {
2464
+ "epoch": 0.22,
2465
+ "grad_norm": 1.07332181930542,
2466
+ "learning_rate": 1.7348105789849893e-05,
2467
+ "loss": 0.038,
2468
+ "step": 1364
2469
+ },
2470
+ {
2471
+ "epoch": 0.22,
2472
+ "grad_norm": 0.6912384629249573,
2473
+ "learning_rate": 1.733380986418871e-05,
2474
+ "loss": 0.0252,
2475
+ "step": 1368
2476
+ },
2477
+ {
2478
+ "epoch": 0.22,
2479
+ "grad_norm": 0.7599236369132996,
2480
+ "learning_rate": 1.731951393852752e-05,
2481
+ "loss": 0.025,
2482
+ "step": 1372
2483
+ },
2484
+ {
2485
+ "epoch": 0.22,
2486
+ "grad_norm": 0.7361788153648376,
2487
+ "learning_rate": 1.7305218012866333e-05,
2488
+ "loss": 0.0413,
2489
+ "step": 1376
2490
+ },
2491
+ {
2492
+ "epoch": 0.22,
2493
+ "grad_norm": 0.6756102442741394,
2494
+ "learning_rate": 1.729092208720515e-05,
2495
+ "loss": 0.0374,
2496
+ "step": 1380
2497
+ },
2498
+ {
2499
+ "epoch": 0.22,
2500
+ "grad_norm": 0.5338143706321716,
2501
+ "learning_rate": 1.727662616154396e-05,
2502
+ "loss": 0.0226,
2503
+ "step": 1384
2504
+ },
2505
+ {
2506
+ "epoch": 0.22,
2507
+ "grad_norm": 0.5998290777206421,
2508
+ "learning_rate": 1.7262330235882773e-05,
2509
+ "loss": 0.0244,
2510
+ "step": 1388
2511
+ },
2512
+ {
2513
+ "epoch": 0.22,
2514
+ "grad_norm": 0.5274918079376221,
2515
+ "learning_rate": 1.724803431022159e-05,
2516
+ "loss": 0.0267,
2517
+ "step": 1392
2518
+ },
2519
+ {
2520
+ "epoch": 0.22,
2521
+ "grad_norm": 0.6132778525352478,
2522
+ "learning_rate": 1.72337383845604e-05,
2523
+ "loss": 0.0351,
2524
+ "step": 1396
2525
+ },
2526
+ {
2527
+ "epoch": 0.23,
2528
+ "grad_norm": 0.7461394667625427,
2529
+ "learning_rate": 1.7219442458899216e-05,
2530
+ "loss": 0.0314,
2531
+ "step": 1400
2532
+ },
2533
+ {
2534
+ "epoch": 0.23,
2535
+ "grad_norm": 0.7438216209411621,
2536
+ "learning_rate": 1.720514653323803e-05,
2537
+ "loss": 0.0261,
2538
+ "step": 1404
2539
+ },
2540
+ {
2541
+ "epoch": 0.23,
2542
+ "grad_norm": 0.730912983417511,
2543
+ "learning_rate": 1.719085060757684e-05,
2544
+ "loss": 0.0229,
2545
+ "step": 1408
2546
+ },
2547
+ {
2548
+ "epoch": 0.23,
2549
+ "grad_norm": 1.2940622568130493,
2550
+ "learning_rate": 1.7176554681915656e-05,
2551
+ "loss": 0.0273,
2552
+ "step": 1412
2553
+ },
2554
+ {
2555
+ "epoch": 0.23,
2556
+ "grad_norm": 0.7200002670288086,
2557
+ "learning_rate": 1.716225875625447e-05,
2558
+ "loss": 0.0267,
2559
+ "step": 1416
2560
+ },
2561
+ {
2562
+ "epoch": 0.23,
2563
+ "grad_norm": 1.3229494094848633,
2564
+ "learning_rate": 1.7147962830593283e-05,
2565
+ "loss": 0.0334,
2566
+ "step": 1420
2567
+ },
2568
+ {
2569
+ "epoch": 0.23,
2570
+ "grad_norm": 0.7718178033828735,
2571
+ "learning_rate": 1.7133666904932096e-05,
2572
+ "loss": 0.0312,
2573
+ "step": 1424
2574
+ },
2575
+ {
2576
+ "epoch": 0.23,
2577
+ "grad_norm": 0.6765419244766235,
2578
+ "learning_rate": 1.711937097927091e-05,
2579
+ "loss": 0.0236,
2580
+ "step": 1428
2581
+ },
2582
+ {
2583
+ "epoch": 0.23,
2584
+ "grad_norm": 0.753416121006012,
2585
+ "learning_rate": 1.7105075053609723e-05,
2586
+ "loss": 0.0271,
2587
+ "step": 1432
2588
+ },
2589
+ {
2590
+ "epoch": 0.23,
2591
+ "grad_norm": 0.6442211866378784,
2592
+ "learning_rate": 1.7090779127948536e-05,
2593
+ "loss": 0.0151,
2594
+ "step": 1436
2595
+ },
2596
+ {
2597
+ "epoch": 0.23,
2598
+ "grad_norm": 0.5224249362945557,
2599
+ "learning_rate": 1.707648320228735e-05,
2600
+ "loss": 0.0355,
2601
+ "step": 1440
2602
+ },
2603
+ {
2604
+ "epoch": 0.23,
2605
+ "grad_norm": 0.7751701474189758,
2606
+ "learning_rate": 1.7062187276626163e-05,
2607
+ "loss": 0.0251,
2608
+ "step": 1444
2609
+ },
2610
+ {
2611
+ "epoch": 0.23,
2612
+ "grad_norm": 0.7004714012145996,
2613
+ "learning_rate": 1.7047891350964976e-05,
2614
+ "loss": 0.0171,
2615
+ "step": 1448
2616
+ },
2617
+ {
2618
+ "epoch": 0.23,
2619
+ "grad_norm": 0.7819869518280029,
2620
+ "learning_rate": 1.703359542530379e-05,
2621
+ "loss": 0.0309,
2622
+ "step": 1452
2623
+ },
2624
+ {
2625
+ "epoch": 0.23,
2626
+ "grad_norm": 0.5115138292312622,
2627
+ "learning_rate": 1.7019299499642603e-05,
2628
+ "loss": 0.0321,
2629
+ "step": 1456
2630
+ },
2631
+ {
2632
+ "epoch": 0.23,
2633
+ "grad_norm": 0.3391636610031128,
2634
+ "learning_rate": 1.7005003573981416e-05,
2635
+ "loss": 0.0221,
2636
+ "step": 1460
2637
+ },
2638
+ {
2639
+ "epoch": 0.24,
2640
+ "grad_norm": 1.2895567417144775,
2641
+ "learning_rate": 1.699070764832023e-05,
2642
+ "loss": 0.0317,
2643
+ "step": 1464
2644
+ },
2645
+ {
2646
+ "epoch": 0.24,
2647
+ "grad_norm": 0.7947621941566467,
2648
+ "learning_rate": 1.6976411722659043e-05,
2649
+ "loss": 0.0266,
2650
+ "step": 1468
2651
+ },
2652
+ {
2653
+ "epoch": 0.24,
2654
+ "grad_norm": 0.7865754961967468,
2655
+ "learning_rate": 1.6962115796997856e-05,
2656
+ "loss": 0.0421,
2657
+ "step": 1472
2658
+ },
2659
+ {
2660
+ "epoch": 0.24,
2661
+ "grad_norm": 1.002314567565918,
2662
+ "learning_rate": 1.694781987133667e-05,
2663
+ "loss": 0.0308,
2664
+ "step": 1476
2665
+ },
2666
+ {
2667
+ "epoch": 0.24,
2668
+ "grad_norm": 0.6675818562507629,
2669
+ "learning_rate": 1.6933523945675483e-05,
2670
+ "loss": 0.0202,
2671
+ "step": 1480
2672
+ },
2673
+ {
2674
+ "epoch": 0.24,
2675
+ "grad_norm": 1.2045787572860718,
2676
+ "learning_rate": 1.6919228020014296e-05,
2677
+ "loss": 0.0387,
2678
+ "step": 1484
2679
+ },
2680
+ {
2681
+ "epoch": 0.24,
2682
+ "grad_norm": 0.5665823817253113,
2683
+ "learning_rate": 1.690493209435311e-05,
2684
+ "loss": 0.0149,
2685
+ "step": 1488
2686
+ },
2687
+ {
2688
+ "epoch": 0.24,
2689
+ "grad_norm": 0.5733370780944824,
2690
+ "learning_rate": 1.6890636168691926e-05,
2691
+ "loss": 0.024,
2692
+ "step": 1492
2693
+ },
2694
+ {
2695
+ "epoch": 0.24,
2696
+ "grad_norm": 0.32628941535949707,
2697
+ "learning_rate": 1.6876340243030736e-05,
2698
+ "loss": 0.0181,
2699
+ "step": 1496
2700
+ },
2701
+ {
2702
+ "epoch": 0.24,
2703
+ "grad_norm": 0.5624873638153076,
2704
+ "learning_rate": 1.686204431736955e-05,
2705
+ "loss": 0.0308,
2706
+ "step": 1500
2707
+ },
2708
+ {
2709
+ "epoch": 0.24,
2710
+ "eval_loss": 0.02633051760494709,
2711
+ "eval_pearson_cosine": 0.9510150119057015,
2712
+ "eval_pearson_dot": 0.9497414425975934,
2713
+ "eval_pearson_euclidean": 0.9338662221774623,
2714
+ "eval_pearson_manhattan": 0.9334364652708103,
2715
+ "eval_pearson_max": 0.9510150119057015,
2716
+ "eval_runtime": 249.0704,
2717
+ "eval_samples_per_second": 2.007,
2718
+ "eval_spearman_cosine": 0.9551964302048428,
2719
+ "eval_spearman_dot": 0.9547767163144516,
2720
+ "eval_spearman_euclidean": 0.9485047200653796,
2721
+ "eval_spearman_manhattan": 0.9472191302750829,
2722
+ "eval_spearman_max": 0.9551964302048428,
2723
+ "eval_steps_per_second": 2.007,
2724
+ "step": 1500
2725
+ },
2726
+ {
2727
+ "epoch": 0.24,
2728
+ "grad_norm": 0.4453175663948059,
2729
+ "learning_rate": 1.6847748391708366e-05,
2730
+ "loss": 0.0251,
2731
+ "step": 1504
2732
+ },
2733
+ {
2734
+ "epoch": 0.24,
2735
+ "grad_norm": 0.8831977844238281,
2736
+ "learning_rate": 1.6833452466047176e-05,
2737
+ "loss": 0.0253,
2738
+ "step": 1508
2739
+ },
2740
+ {
2741
+ "epoch": 0.24,
2742
+ "grad_norm": 0.9551718831062317,
2743
+ "learning_rate": 1.6819156540385993e-05,
2744
+ "loss": 0.0451,
2745
+ "step": 1512
2746
+ },
2747
+ {
2748
+ "epoch": 0.24,
2749
+ "grad_norm": 0.9147098064422607,
2750
+ "learning_rate": 1.6804860614724806e-05,
2751
+ "loss": 0.0233,
2752
+ "step": 1516
2753
+ },
2754
+ {
2755
+ "epoch": 0.24,
2756
+ "grad_norm": 0.8430941104888916,
2757
+ "learning_rate": 1.6790564689063616e-05,
2758
+ "loss": 0.0224,
2759
+ "step": 1520
2760
+ },
2761
+ {
2762
+ "epoch": 0.25,
2763
+ "grad_norm": 0.6627079844474792,
2764
+ "learning_rate": 1.6776268763402433e-05,
2765
+ "loss": 0.0221,
2766
+ "step": 1524
2767
+ },
2768
+ {
2769
+ "epoch": 0.25,
2770
+ "grad_norm": 0.7855361104011536,
2771
+ "learning_rate": 1.6761972837741246e-05,
2772
+ "loss": 0.0307,
2773
+ "step": 1528
2774
+ },
2775
+ {
2776
+ "epoch": 0.25,
2777
+ "grad_norm": 0.45924311876296997,
2778
+ "learning_rate": 1.6747676912080056e-05,
2779
+ "loss": 0.0257,
2780
+ "step": 1532
2781
+ },
2782
+ {
2783
+ "epoch": 0.25,
2784
+ "grad_norm": 0.7786813378334045,
2785
+ "learning_rate": 1.6733380986418873e-05,
2786
+ "loss": 0.0345,
2787
+ "step": 1536
2788
+ },
2789
+ {
2790
+ "epoch": 0.25,
2791
+ "grad_norm": 0.5647817254066467,
2792
+ "learning_rate": 1.6719085060757686e-05,
2793
+ "loss": 0.0181,
2794
+ "step": 1540
2795
+ },
2796
+ {
2797
+ "epoch": 0.25,
2798
+ "grad_norm": 1.0387967824935913,
2799
+ "learning_rate": 1.67047891350965e-05,
2800
+ "loss": 0.0297,
2801
+ "step": 1544
2802
+ },
2803
+ {
2804
+ "epoch": 0.25,
2805
+ "grad_norm": 0.5947245359420776,
2806
+ "learning_rate": 1.6690493209435313e-05,
2807
+ "loss": 0.0239,
2808
+ "step": 1548
2809
+ },
2810
+ {
2811
+ "epoch": 0.25,
2812
+ "grad_norm": 0.5582525134086609,
2813
+ "learning_rate": 1.6676197283774126e-05,
2814
+ "loss": 0.0305,
2815
+ "step": 1552
2816
+ },
2817
+ {
2818
+ "epoch": 0.25,
2819
+ "grad_norm": 0.664482057094574,
2820
+ "learning_rate": 1.666190135811294e-05,
2821
+ "loss": 0.0249,
2822
+ "step": 1556
2823
+ },
2824
+ {
2825
+ "epoch": 0.25,
2826
+ "grad_norm": 0.9069839715957642,
2827
+ "learning_rate": 1.6647605432451753e-05,
2828
+ "loss": 0.0298,
2829
+ "step": 1560
2830
+ },
2831
+ {
2832
+ "epoch": 0.25,
2833
+ "grad_norm": 0.8026562333106995,
2834
+ "learning_rate": 1.6633309506790566e-05,
2835
+ "loss": 0.0203,
2836
+ "step": 1564
2837
+ },
2838
+ {
2839
+ "epoch": 0.25,
2840
+ "grad_norm": 0.6699361801147461,
2841
+ "learning_rate": 1.661901358112938e-05,
2842
+ "loss": 0.0312,
2843
+ "step": 1568
2844
+ },
2845
+ {
2846
+ "epoch": 0.25,
2847
+ "grad_norm": 0.6312674283981323,
2848
+ "learning_rate": 1.6604717655468193e-05,
2849
+ "loss": 0.0204,
2850
+ "step": 1572
2851
+ },
2852
+ {
2853
+ "epoch": 0.25,
2854
+ "grad_norm": 0.4308748245239258,
2855
+ "learning_rate": 1.6590421729807006e-05,
2856
+ "loss": 0.0209,
2857
+ "step": 1576
2858
+ },
2859
+ {
2860
+ "epoch": 0.25,
2861
+ "grad_norm": 1.3553575277328491,
2862
+ "learning_rate": 1.657612580414582e-05,
2863
+ "loss": 0.0257,
2864
+ "step": 1580
2865
+ },
2866
+ {
2867
+ "epoch": 0.25,
2868
+ "grad_norm": 1.4249012470245361,
2869
+ "learning_rate": 1.6561829878484633e-05,
2870
+ "loss": 0.0306,
2871
+ "step": 1584
2872
+ },
2873
+ {
2874
+ "epoch": 0.26,
2875
+ "grad_norm": 0.4437320828437805,
2876
+ "learning_rate": 1.6547533952823446e-05,
2877
+ "loss": 0.017,
2878
+ "step": 1588
2879
+ },
2880
+ {
2881
+ "epoch": 0.26,
2882
+ "grad_norm": 0.6957021951675415,
2883
+ "learning_rate": 1.653323802716226e-05,
2884
+ "loss": 0.0227,
2885
+ "step": 1592
2886
+ },
2887
+ {
2888
+ "epoch": 0.26,
2889
+ "grad_norm": 0.5041042566299438,
2890
+ "learning_rate": 1.6518942101501073e-05,
2891
+ "loss": 0.0228,
2892
+ "step": 1596
2893
+ },
2894
+ {
2895
+ "epoch": 0.26,
2896
+ "grad_norm": 0.6849528551101685,
2897
+ "learning_rate": 1.6504646175839886e-05,
2898
+ "loss": 0.0205,
2899
+ "step": 1600
2900
+ },
2901
+ {
2902
+ "epoch": 0.26,
2903
+ "grad_norm": 0.8483012318611145,
2904
+ "learning_rate": 1.64903502501787e-05,
2905
+ "loss": 0.0336,
2906
+ "step": 1604
2907
+ },
2908
+ {
2909
+ "epoch": 0.26,
2910
+ "grad_norm": 0.9397591352462769,
2911
+ "learning_rate": 1.6476054324517513e-05,
2912
+ "loss": 0.029,
2913
+ "step": 1608
2914
+ },
2915
+ {
2916
+ "epoch": 0.26,
2917
+ "grad_norm": 0.9127416014671326,
2918
+ "learning_rate": 1.6461758398856326e-05,
2919
+ "loss": 0.0225,
2920
+ "step": 1612
2921
+ },
2922
+ {
2923
+ "epoch": 0.26,
2924
+ "grad_norm": 0.8871793150901794,
2925
+ "learning_rate": 1.6447462473195143e-05,
2926
+ "loss": 0.0185,
2927
+ "step": 1616
2928
+ },
2929
+ {
2930
+ "epoch": 0.26,
2931
+ "grad_norm": 0.7608364224433899,
2932
+ "learning_rate": 1.6433166547533953e-05,
2933
+ "loss": 0.0273,
2934
+ "step": 1620
2935
+ },
2936
+ {
2937
+ "epoch": 0.26,
2938
+ "grad_norm": 0.778032660484314,
2939
+ "learning_rate": 1.6418870621872766e-05,
2940
+ "loss": 0.02,
2941
+ "step": 1624
2942
+ },
2943
+ {
2944
+ "epoch": 0.26,
2945
+ "grad_norm": 0.6790982484817505,
2946
+ "learning_rate": 1.6404574696211583e-05,
2947
+ "loss": 0.025,
2948
+ "step": 1628
2949
+ },
2950
+ {
2951
+ "epoch": 0.26,
2952
+ "grad_norm": 0.4867960810661316,
2953
+ "learning_rate": 1.6390278770550393e-05,
2954
+ "loss": 0.0175,
2955
+ "step": 1632
2956
+ },
2957
+ {
2958
+ "epoch": 0.26,
2959
+ "grad_norm": 0.5005691647529602,
2960
+ "learning_rate": 1.637598284488921e-05,
2961
+ "loss": 0.0329,
2962
+ "step": 1636
2963
+ },
2964
+ {
2965
+ "epoch": 0.26,
2966
+ "grad_norm": 0.7641982436180115,
2967
+ "learning_rate": 1.6361686919228023e-05,
2968
+ "loss": 0.0312,
2969
+ "step": 1640
2970
+ },
2971
+ {
2972
+ "epoch": 0.26,
2973
+ "grad_norm": 1.112260103225708,
2974
+ "learning_rate": 1.6347390993566833e-05,
2975
+ "loss": 0.0289,
2976
+ "step": 1644
2977
+ },
2978
+ {
2979
+ "epoch": 0.27,
2980
+ "grad_norm": 0.8543418049812317,
2981
+ "learning_rate": 1.633309506790565e-05,
2982
+ "loss": 0.0205,
2983
+ "step": 1648
2984
+ },
2985
+ {
2986
+ "epoch": 0.27,
2987
+ "grad_norm": 0.5461686849594116,
2988
+ "learning_rate": 1.6318799142244463e-05,
2989
+ "loss": 0.0155,
2990
+ "step": 1652
2991
+ },
2992
+ {
2993
+ "epoch": 0.27,
2994
+ "grad_norm": 0.5375934839248657,
2995
+ "learning_rate": 1.6304503216583276e-05,
2996
+ "loss": 0.0237,
2997
+ "step": 1656
2998
+ },
2999
+ {
3000
+ "epoch": 0.27,
3001
+ "grad_norm": 0.6225507855415344,
3002
+ "learning_rate": 1.629020729092209e-05,
3003
+ "loss": 0.031,
3004
+ "step": 1660
3005
+ },
3006
+ {
3007
+ "epoch": 0.27,
3008
+ "grad_norm": 1.2959158420562744,
3009
+ "learning_rate": 1.62759113652609e-05,
3010
+ "loss": 0.0315,
3011
+ "step": 1664
3012
+ },
3013
+ {
3014
+ "epoch": 0.27,
3015
+ "grad_norm": 0.8647311329841614,
3016
+ "learning_rate": 1.6261615439599716e-05,
3017
+ "loss": 0.0204,
3018
+ "step": 1668
3019
+ },
3020
+ {
3021
+ "epoch": 0.27,
3022
+ "grad_norm": 0.5671218633651733,
3023
+ "learning_rate": 1.624731951393853e-05,
3024
+ "loss": 0.0182,
3025
+ "step": 1672
3026
+ },
3027
+ {
3028
+ "epoch": 0.27,
3029
+ "grad_norm": 0.5024406313896179,
3030
+ "learning_rate": 1.623302358827734e-05,
3031
+ "loss": 0.0257,
3032
+ "step": 1676
3033
+ },
3034
+ {
3035
+ "epoch": 0.27,
3036
+ "grad_norm": 0.4015944004058838,
3037
+ "learning_rate": 1.6218727662616156e-05,
3038
+ "loss": 0.023,
3039
+ "step": 1680
3040
+ },
3041
+ {
3042
+ "epoch": 0.27,
3043
+ "grad_norm": 0.5640401840209961,
3044
+ "learning_rate": 1.620443173695497e-05,
3045
+ "loss": 0.0179,
3046
+ "step": 1684
3047
+ },
3048
+ {
3049
+ "epoch": 0.27,
3050
+ "grad_norm": 0.7147939205169678,
3051
+ "learning_rate": 1.6190135811293783e-05,
3052
+ "loss": 0.0176,
3053
+ "step": 1688
3054
+ },
3055
+ {
3056
+ "epoch": 0.27,
3057
+ "grad_norm": 0.8704376816749573,
3058
+ "learning_rate": 1.6175839885632596e-05,
3059
+ "loss": 0.02,
3060
+ "step": 1692
3061
+ },
3062
+ {
3063
+ "epoch": 0.27,
3064
+ "grad_norm": 0.7839465737342834,
3065
+ "learning_rate": 1.616154395997141e-05,
3066
+ "loss": 0.0209,
3067
+ "step": 1696
3068
+ },
3069
+ {
3070
+ "epoch": 0.27,
3071
+ "grad_norm": 0.4096917510032654,
3072
+ "learning_rate": 1.6147248034310223e-05,
3073
+ "loss": 0.0204,
3074
+ "step": 1700
3075
+ },
3076
+ {
3077
+ "epoch": 0.27,
3078
+ "grad_norm": 0.5869227051734924,
3079
+ "learning_rate": 1.6132952108649036e-05,
3080
+ "loss": 0.0198,
3081
+ "step": 1704
3082
+ },
3083
+ {
3084
+ "epoch": 0.27,
3085
+ "grad_norm": 0.6609891057014465,
3086
+ "learning_rate": 1.611865618298785e-05,
3087
+ "loss": 0.0251,
3088
+ "step": 1708
3089
+ },
3090
+ {
3091
+ "epoch": 0.28,
3092
+ "grad_norm": 0.5184406042098999,
3093
+ "learning_rate": 1.6104360257326663e-05,
3094
+ "loss": 0.0234,
3095
+ "step": 1712
3096
+ },
3097
+ {
3098
+ "epoch": 0.28,
3099
+ "grad_norm": 0.4317310154438019,
3100
+ "learning_rate": 1.6090064331665476e-05,
3101
+ "loss": 0.0263,
3102
+ "step": 1716
3103
+ },
3104
+ {
3105
+ "epoch": 0.28,
3106
+ "grad_norm": 0.5459049940109253,
3107
+ "learning_rate": 1.607576840600429e-05,
3108
+ "loss": 0.0243,
3109
+ "step": 1720
3110
+ },
3111
+ {
3112
+ "epoch": 0.28,
3113
+ "grad_norm": 0.7071524858474731,
3114
+ "learning_rate": 1.6061472480343103e-05,
3115
+ "loss": 0.0235,
3116
+ "step": 1724
3117
+ },
3118
+ {
3119
+ "epoch": 0.28,
3120
+ "grad_norm": 0.8446561694145203,
3121
+ "learning_rate": 1.604717655468192e-05,
3122
+ "loss": 0.0206,
3123
+ "step": 1728
3124
+ },
3125
+ {
3126
+ "epoch": 0.28,
3127
+ "grad_norm": 0.6042218804359436,
3128
+ "learning_rate": 1.603288062902073e-05,
3129
+ "loss": 0.0332,
3130
+ "step": 1732
3131
+ },
3132
+ {
3133
+ "epoch": 0.28,
3134
+ "grad_norm": 0.5404290556907654,
3135
+ "learning_rate": 1.6018584703359543e-05,
3136
+ "loss": 0.0198,
3137
+ "step": 1736
3138
+ },
3139
+ {
3140
+ "epoch": 0.28,
3141
+ "grad_norm": 0.3870123326778412,
3142
+ "learning_rate": 1.6004288777698356e-05,
3143
+ "loss": 0.0216,
3144
+ "step": 1740
3145
+ },
3146
+ {
3147
+ "epoch": 0.28,
3148
+ "grad_norm": 0.803578794002533,
3149
+ "learning_rate": 1.598999285203717e-05,
3150
+ "loss": 0.0248,
3151
+ "step": 1744
3152
+ },
3153
+ {
3154
+ "epoch": 0.28,
3155
+ "grad_norm": 0.7153476476669312,
3156
+ "learning_rate": 1.5975696926375983e-05,
3157
+ "loss": 0.0237,
3158
+ "step": 1748
3159
+ },
3160
+ {
3161
+ "epoch": 0.28,
3162
+ "grad_norm": 1.2977997064590454,
3163
+ "learning_rate": 1.5961401000714796e-05,
3164
+ "loss": 0.037,
3165
+ "step": 1752
3166
+ },
3167
+ {
3168
+ "epoch": 0.28,
3169
+ "grad_norm": 0.7146950364112854,
3170
+ "learning_rate": 1.594710507505361e-05,
3171
+ "loss": 0.02,
3172
+ "step": 1756
3173
+ },
3174
+ {
3175
+ "epoch": 0.28,
3176
+ "grad_norm": 1.2458608150482178,
3177
+ "learning_rate": 1.5932809149392426e-05,
3178
+ "loss": 0.0396,
3179
+ "step": 1760
3180
+ },
3181
+ {
3182
+ "epoch": 0.28,
3183
+ "grad_norm": 1.0099600553512573,
3184
+ "learning_rate": 1.5918513223731236e-05,
3185
+ "loss": 0.0244,
3186
+ "step": 1764
3187
+ },
3188
+ {
3189
+ "epoch": 0.28,
3190
+ "grad_norm": 0.8322357535362244,
3191
+ "learning_rate": 1.590421729807005e-05,
3192
+ "loss": 0.0224,
3193
+ "step": 1768
3194
+ },
3195
+ {
3196
+ "epoch": 0.28,
3197
+ "grad_norm": 0.7080321907997131,
3198
+ "learning_rate": 1.5889921372408866e-05,
3199
+ "loss": 0.0294,
3200
+ "step": 1772
3201
+ },
3202
+ {
3203
+ "epoch": 0.29,
3204
+ "grad_norm": 0.5599932670593262,
3205
+ "learning_rate": 1.5875625446747676e-05,
3206
+ "loss": 0.0194,
3207
+ "step": 1776
3208
+ },
3209
+ {
3210
+ "epoch": 0.29,
3211
+ "grad_norm": 0.4317460060119629,
3212
+ "learning_rate": 1.5861329521086493e-05,
3213
+ "loss": 0.0144,
3214
+ "step": 1780
3215
+ },
3216
+ {
3217
+ "epoch": 0.29,
3218
+ "grad_norm": 0.8583347797393799,
3219
+ "learning_rate": 1.5847033595425306e-05,
3220
+ "loss": 0.022,
3221
+ "step": 1784
3222
+ },
3223
+ {
3224
+ "epoch": 0.29,
3225
+ "grad_norm": 0.6212904453277588,
3226
+ "learning_rate": 1.5832737669764116e-05,
3227
+ "loss": 0.0235,
3228
+ "step": 1788
3229
+ },
3230
+ {
3231
+ "epoch": 0.29,
3232
+ "grad_norm": 0.6250016689300537,
3233
+ "learning_rate": 1.5818441744102933e-05,
3234
+ "loss": 0.024,
3235
+ "step": 1792
3236
+ },
3237
+ {
3238
+ "epoch": 0.29,
3239
+ "grad_norm": 0.7036072611808777,
3240
+ "learning_rate": 1.5804145818441746e-05,
3241
+ "loss": 0.0155,
3242
+ "step": 1796
3243
+ },
3244
+ {
3245
+ "epoch": 0.29,
3246
+ "grad_norm": 0.7857645750045776,
3247
+ "learning_rate": 1.578984989278056e-05,
3248
+ "loss": 0.0177,
3249
+ "step": 1800
3250
+ },
3251
+ {
3252
+ "epoch": 0.29,
3253
+ "eval_loss": 0.022910235449671745,
3254
+ "eval_pearson_cosine": 0.9571668454101061,
3255
+ "eval_pearson_dot": 0.9545933082183199,
3256
+ "eval_pearson_euclidean": 0.9493360141559007,
3257
+ "eval_pearson_manhattan": 0.9494330807780875,
3258
+ "eval_pearson_max": 0.9571668454101061,
3259
+ "eval_runtime": 248.8356,
3260
+ "eval_samples_per_second": 2.009,
3261
+ "eval_spearman_cosine": 0.9652652915549957,
3262
+ "eval_spearman_dot": 0.9626083355880656,
3263
+ "eval_spearman_euclidean": 0.9635709799238159,
3264
+ "eval_spearman_manhattan": 0.9633432188979801,
3265
+ "eval_spearman_max": 0.9652652915549957,
3266
+ "eval_steps_per_second": 2.009,
3267
+ "step": 1800
3268
+ }
3269
+ ],
3270
+ "logging_steps": 4,
3271
+ "max_steps": 6218,
3272
+ "num_input_tokens_seen": 0,
3273
+ "num_train_epochs": 1,
3274
+ "save_steps": 100,
3275
+ "total_flos": 0.0,
3276
+ "train_batch_size": 1,
3277
+ "trial_name": null,
3278
+ "trial_params": null
3279
+ }