Training checkpoint - Epoch 3, Step 61224
Browse files- checkpoint-61224/1_Pooling/config.json +10 -0
- checkpoint-61224/config.json +25 -0
- checkpoint-61224/config_sentence_transformers.json +14 -0
- checkpoint-61224/modules.json +20 -0
- checkpoint-61224/sentence_bert_config.json +4 -0
- checkpoint-61224/special_tokens_map.json +37 -0
- checkpoint-61224/tokenizer.json +0 -0
- checkpoint-61224/tokenizer_config.json +65 -0
- checkpoint-61224/trainer_state.json +731 -0
- checkpoint-61224/training_metrics.json +396 -0
- checkpoint-61224/vocab.txt +0 -0
checkpoint-61224/1_Pooling/config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"word_embedding_dimension": 384,
|
| 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 |
+
}
|
checkpoint-61224/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 384,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 1536,
|
| 13 |
+
"layer_norm_eps": 1e-12,
|
| 14 |
+
"max_position_embeddings": 512,
|
| 15 |
+
"model_type": "bert",
|
| 16 |
+
"num_attention_heads": 12,
|
| 17 |
+
"num_hidden_layers": 6,
|
| 18 |
+
"pad_token_id": 0,
|
| 19 |
+
"position_embedding_type": "absolute",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.53.3",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
checkpoint-61224/config_sentence_transformers.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"__version__": {
|
| 3 |
+
"sentence_transformers": "5.2.0",
|
| 4 |
+
"transformers": "4.53.3",
|
| 5 |
+
"pytorch": "2.6.0+cu124"
|
| 6 |
+
},
|
| 7 |
+
"model_type": "SentenceTransformer",
|
| 8 |
+
"prompts": {
|
| 9 |
+
"query": "",
|
| 10 |
+
"document": ""
|
| 11 |
+
},
|
| 12 |
+
"default_prompt_name": null,
|
| 13 |
+
"similarity_fn_name": "cosine"
|
| 14 |
+
}
|
checkpoint-61224/modules.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
{
|
| 15 |
+
"idx": 2,
|
| 16 |
+
"name": "2",
|
| 17 |
+
"path": "2_Normalize",
|
| 18 |
+
"type": "sentence_transformers.models.Normalize"
|
| 19 |
+
}
|
| 20 |
+
]
|
checkpoint-61224/sentence_bert_config.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"max_seq_length": 256,
|
| 3 |
+
"do_lower_case": false
|
| 4 |
+
}
|
checkpoint-61224/special_tokens_map.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cls_token": {
|
| 3 |
+
"content": "[CLS]",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"mask_token": {
|
| 10 |
+
"content": "[MASK]",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "[PAD]",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"sep_token": {
|
| 24 |
+
"content": "[SEP]",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"unk_token": {
|
| 31 |
+
"content": "[UNK]",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
}
|
| 37 |
+
}
|
checkpoint-61224/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-61224/tokenizer_config.json
ADDED
|
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": false,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_basic_tokenize": true,
|
| 47 |
+
"do_lower_case": true,
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "[MASK]",
|
| 50 |
+
"max_length": 128,
|
| 51 |
+
"model_max_length": 256,
|
| 52 |
+
"never_split": null,
|
| 53 |
+
"pad_to_multiple_of": null,
|
| 54 |
+
"pad_token": "[PAD]",
|
| 55 |
+
"pad_token_type_id": 0,
|
| 56 |
+
"padding_side": "right",
|
| 57 |
+
"sep_token": "[SEP]",
|
| 58 |
+
"stride": 0,
|
| 59 |
+
"strip_accents": null,
|
| 60 |
+
"tokenize_chinese_chars": true,
|
| 61 |
+
"tokenizer_class": "BertTokenizer",
|
| 62 |
+
"truncation_side": "right",
|
| 63 |
+
"truncation_strategy": "longest_first",
|
| 64 |
+
"unk_token": "[UNK]"
|
| 65 |
+
}
|
checkpoint-61224/trainer_state.json
ADDED
|
@@ -0,0 +1,731 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 2000,
|
| 7 |
+
"global_step": 61224,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 4.9000392003136025e-05,
|
| 14 |
+
"grad_norm": NaN,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 146.2524,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.09800078400627205,
|
| 21 |
+
"grad_norm": 0.11796475946903229,
|
| 22 |
+
"learning_rate": 2.4424301812836847e-06,
|
| 23 |
+
"loss": 32.229,
|
| 24 |
+
"step": 2000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.09800078400627205,
|
| 28 |
+
"eval_cosine_accuracy": 0.6100063465199915,
|
| 29 |
+
"eval_cosine_accuracy_threshold": 0.9997795820236206,
|
| 30 |
+
"eval_cosine_ap": 0.6786677205526788,
|
| 31 |
+
"eval_cosine_f1": 0.6667136862150438,
|
| 32 |
+
"eval_cosine_f1_threshold": 0.991624653339386,
|
| 33 |
+
"eval_cosine_mcc": 0.00975822629781094,
|
| 34 |
+
"eval_cosine_precision": 0.5000793524837327,
|
| 35 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 36 |
+
"eval_loss": 26.897747039794922,
|
| 37 |
+
"eval_runtime": 22.9309,
|
| 38 |
+
"eval_samples_per_second": 824.566,
|
| 39 |
+
"eval_steps_per_second": 1.614,
|
| 40 |
+
"step": 2000
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.1960015680125441,
|
| 44 |
+
"grad_norm": 0.013558157719671726,
|
| 45 |
+
"learning_rate": 4.892209701126899e-06,
|
| 46 |
+
"loss": 0.0109,
|
| 47 |
+
"step": 4000
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.1960015680125441,
|
| 51 |
+
"eval_cosine_accuracy": 0.6240215781679712,
|
| 52 |
+
"eval_cosine_accuracy_threshold": 0.9999226331710815,
|
| 53 |
+
"eval_cosine_ap": 0.6923415808404929,
|
| 54 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 55 |
+
"eval_cosine_f1_threshold": 0.9939442873001099,
|
| 56 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 57 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 58 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 59 |
+
"eval_loss": 27.12621307373047,
|
| 60 |
+
"eval_runtime": 23.1359,
|
| 61 |
+
"eval_samples_per_second": 817.257,
|
| 62 |
+
"eval_steps_per_second": 1.599,
|
| 63 |
+
"step": 4000
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.29400235201881614,
|
| 67 |
+
"grad_norm": 0.0,
|
| 68 |
+
"learning_rate": 4.87678916827853e-06,
|
| 69 |
+
"loss": 0.5788,
|
| 70 |
+
"step": 6000
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.29400235201881614,
|
| 74 |
+
"eval_cosine_accuracy": 0.5835096255553205,
|
| 75 |
+
"eval_cosine_accuracy_threshold": 0.9478387236595154,
|
| 76 |
+
"eval_cosine_ap": 0.6313955266297183,
|
| 77 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 78 |
+
"eval_cosine_f1_threshold": -0.09354762732982635,
|
| 79 |
+
"eval_cosine_mcc": 0.007272583024577166,
|
| 80 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 81 |
+
"eval_cosine_recall": 1.0,
|
| 82 |
+
"eval_loss": 32.61564254760742,
|
| 83 |
+
"eval_runtime": 23.1255,
|
| 84 |
+
"eval_samples_per_second": 817.626,
|
| 85 |
+
"eval_steps_per_second": 1.6,
|
| 86 |
+
"step": 6000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3920031360250882,
|
| 90 |
+
"grad_norm": 0.0,
|
| 91 |
+
"learning_rate": 4.747840103159252e-06,
|
| 92 |
+
"loss": 0.0,
|
| 93 |
+
"step": 8000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3920031360250882,
|
| 97 |
+
"eval_cosine_accuracy": 0.5764226782314364,
|
| 98 |
+
"eval_cosine_accuracy_threshold": 0.953868567943573,
|
| 99 |
+
"eval_cosine_ap": 0.6184434211752825,
|
| 100 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 101 |
+
"eval_cosine_f1_threshold": -0.09449770301580429,
|
| 102 |
+
"eval_cosine_mcc": 0.0,
|
| 103 |
+
"eval_cosine_precision": 0.5,
|
| 104 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 105 |
+
"eval_loss": 33.463409423828125,
|
| 106 |
+
"eval_runtime": 24.245,
|
| 107 |
+
"eval_samples_per_second": 779.873,
|
| 108 |
+
"eval_steps_per_second": 1.526,
|
| 109 |
+
"step": 8000
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.49000392003136023,
|
| 113 |
+
"grad_norm": 0.0,
|
| 114 |
+
"learning_rate": 4.618891038039975e-06,
|
| 115 |
+
"loss": 0.0,
|
| 116 |
+
"step": 10000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.49000392003136023,
|
| 120 |
+
"eval_cosine_accuracy": 0.5845673788872435,
|
| 121 |
+
"eval_cosine_accuracy_threshold": 0.9387475848197937,
|
| 122 |
+
"eval_cosine_ap": 0.6273588632357191,
|
| 123 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 124 |
+
"eval_cosine_f1_threshold": -0.1685822755098343,
|
| 125 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 126 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 127 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 128 |
+
"eval_loss": 35.26945495605469,
|
| 129 |
+
"eval_runtime": 23.2073,
|
| 130 |
+
"eval_samples_per_second": 814.742,
|
| 131 |
+
"eval_steps_per_second": 1.594,
|
| 132 |
+
"step": 10000
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.5880047040376323,
|
| 136 |
+
"grad_norm": 0.0,
|
| 137 |
+
"learning_rate": 4.489941972920697e-06,
|
| 138 |
+
"loss": 0.0,
|
| 139 |
+
"step": 12000
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.5880047040376323,
|
| 143 |
+
"eval_cosine_accuracy": 0.5943515972075312,
|
| 144 |
+
"eval_cosine_accuracy_threshold": 0.9397846460342407,
|
| 145 |
+
"eval_cosine_ap": 0.6521863345575756,
|
| 146 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 147 |
+
"eval_cosine_f1_threshold": -0.17127826809883118,
|
| 148 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 149 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 150 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 151 |
+
"eval_loss": 30.115076065063477,
|
| 152 |
+
"eval_runtime": 23.3715,
|
| 153 |
+
"eval_samples_per_second": 809.021,
|
| 154 |
+
"eval_steps_per_second": 1.583,
|
| 155 |
+
"step": 12000
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.6860054880439044,
|
| 159 |
+
"grad_norm": 0.0,
|
| 160 |
+
"learning_rate": 4.360992907801419e-06,
|
| 161 |
+
"loss": 0.0,
|
| 162 |
+
"step": 14000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.6860054880439044,
|
| 166 |
+
"eval_cosine_accuracy": 0.5931351808758197,
|
| 167 |
+
"eval_cosine_accuracy_threshold": 0.9391266107559204,
|
| 168 |
+
"eval_cosine_ap": 0.6494508725044261,
|
| 169 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 170 |
+
"eval_cosine_f1_threshold": -0.1697666049003601,
|
| 171 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 172 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 173 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 174 |
+
"eval_loss": 30.446081161499023,
|
| 175 |
+
"eval_runtime": 24.005,
|
| 176 |
+
"eval_samples_per_second": 787.669,
|
| 177 |
+
"eval_steps_per_second": 1.541,
|
| 178 |
+
"step": 14000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.7840062720501764,
|
| 182 |
+
"grad_norm": 0.0,
|
| 183 |
+
"learning_rate": 4.232043842682141e-06,
|
| 184 |
+
"loss": 0.0,
|
| 185 |
+
"step": 16000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.7840062720501764,
|
| 189 |
+
"eval_cosine_accuracy": 0.5934525068753966,
|
| 190 |
+
"eval_cosine_accuracy_threshold": 0.9391290545463562,
|
| 191 |
+
"eval_cosine_ap": 0.6498809737033857,
|
| 192 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 193 |
+
"eval_cosine_f1_threshold": -0.17070035636425018,
|
| 194 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 195 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 196 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 197 |
+
"eval_loss": 30.447351455688477,
|
| 198 |
+
"eval_runtime": 23.3368,
|
| 199 |
+
"eval_samples_per_second": 810.221,
|
| 200 |
+
"eval_steps_per_second": 1.585,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.8820070560564485,
|
| 205 |
+
"grad_norm": 0.0,
|
| 206 |
+
"learning_rate": 4.103094777562863e-06,
|
| 207 |
+
"loss": 0.0,
|
| 208 |
+
"step": 18000
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.8820070560564485,
|
| 212 |
+
"eval_cosine_accuracy": 0.593558282208589,
|
| 213 |
+
"eval_cosine_accuracy_threshold": 0.9389414191246033,
|
| 214 |
+
"eval_cosine_ap": 0.6502482026720758,
|
| 215 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 216 |
+
"eval_cosine_f1_threshold": -0.17154842615127563,
|
| 217 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 218 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 219 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 220 |
+
"eval_loss": 30.453508377075195,
|
| 221 |
+
"eval_runtime": 23.676,
|
| 222 |
+
"eval_samples_per_second": 798.616,
|
| 223 |
+
"eval_steps_per_second": 1.563,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.9800078400627205,
|
| 228 |
+
"grad_norm": 0.0,
|
| 229 |
+
"learning_rate": 3.974145712443585e-06,
|
| 230 |
+
"loss": 0.0,
|
| 231 |
+
"step": 20000
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.9800078400627205,
|
| 235 |
+
"eval_cosine_accuracy": 0.5937169452083774,
|
| 236 |
+
"eval_cosine_accuracy_threshold": 0.9388357400894165,
|
| 237 |
+
"eval_cosine_ap": 0.6506732897432272,
|
| 238 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 239 |
+
"eval_cosine_f1_threshold": -0.1724291741847992,
|
| 240 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 241 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 242 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 243 |
+
"eval_loss": 30.439617156982422,
|
| 244 |
+
"eval_runtime": 23.0663,
|
| 245 |
+
"eval_samples_per_second": 819.725,
|
| 246 |
+
"eval_steps_per_second": 1.604,
|
| 247 |
+
"step": 20000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0780086240689926,
|
| 251 |
+
"grad_norm": 0.0,
|
| 252 |
+
"learning_rate": 3.845196647324307e-06,
|
| 253 |
+
"loss": 0.0,
|
| 254 |
+
"step": 22000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0780086240689926,
|
| 258 |
+
"eval_cosine_accuracy": 0.5937698328749735,
|
| 259 |
+
"eval_cosine_accuracy_threshold": 0.9385948181152344,
|
| 260 |
+
"eval_cosine_ap": 0.6509728340267652,
|
| 261 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 262 |
+
"eval_cosine_f1_threshold": -0.17285144329071045,
|
| 263 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 264 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 265 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 266 |
+
"eval_loss": 30.44361686706543,
|
| 267 |
+
"eval_runtime": 25.4596,
|
| 268 |
+
"eval_samples_per_second": 742.666,
|
| 269 |
+
"eval_steps_per_second": 1.453,
|
| 270 |
+
"step": 22000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.1760094080752646,
|
| 274 |
+
"grad_norm": 0.0,
|
| 275 |
+
"learning_rate": 3.716247582205029e-06,
|
| 276 |
+
"loss": 0.0,
|
| 277 |
+
"step": 24000
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.1760094080752646,
|
| 281 |
+
"eval_cosine_accuracy": 0.593928495874762,
|
| 282 |
+
"eval_cosine_accuracy_threshold": 0.9375462532043457,
|
| 283 |
+
"eval_cosine_ap": 0.6513894091627628,
|
| 284 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 285 |
+
"eval_cosine_f1_threshold": -0.17358532547950745,
|
| 286 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 287 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 288 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 289 |
+
"eval_loss": 30.4697208404541,
|
| 290 |
+
"eval_runtime": 23.202,
|
| 291 |
+
"eval_samples_per_second": 814.93,
|
| 292 |
+
"eval_steps_per_second": 1.595,
|
| 293 |
+
"step": 24000
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 1.2740101920815365,
|
| 297 |
+
"grad_norm": 0.0,
|
| 298 |
+
"learning_rate": 3.587362991618311e-06,
|
| 299 |
+
"loss": 0.0,
|
| 300 |
+
"step": 26000
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 1.2740101920815365,
|
| 304 |
+
"eval_cosine_accuracy": 0.5941400465411466,
|
| 305 |
+
"eval_cosine_accuracy_threshold": 0.9373874664306641,
|
| 306 |
+
"eval_cosine_ap": 0.6517436899727478,
|
| 307 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 308 |
+
"eval_cosine_f1_threshold": -0.17465540766716003,
|
| 309 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 310 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 311 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 312 |
+
"eval_loss": 30.457592010498047,
|
| 313 |
+
"eval_runtime": 23.1905,
|
| 314 |
+
"eval_samples_per_second": 815.335,
|
| 315 |
+
"eval_steps_per_second": 1.595,
|
| 316 |
+
"step": 26000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.3720109760878088,
|
| 320 |
+
"grad_norm": 0.0,
|
| 321 |
+
"learning_rate": 3.458607350096712e-06,
|
| 322 |
+
"loss": 0.0,
|
| 323 |
+
"step": 28000
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 1.3720109760878088,
|
| 327 |
+
"eval_cosine_accuracy": 0.6150306748466258,
|
| 328 |
+
"eval_cosine_accuracy_threshold": 0.9150234460830688,
|
| 329 |
+
"eval_cosine_ap": 0.6887500197266508,
|
| 330 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 331 |
+
"eval_cosine_f1_threshold": 0.005598240531980991,
|
| 332 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 333 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 334 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 335 |
+
"eval_loss": 27.762596130371094,
|
| 336 |
+
"eval_runtime": 24.3242,
|
| 337 |
+
"eval_samples_per_second": 777.334,
|
| 338 |
+
"eval_steps_per_second": 1.521,
|
| 339 |
+
"step": 28000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.4700117600940807,
|
| 343 |
+
"grad_norm": 0.0,
|
| 344 |
+
"learning_rate": 3.329658284977434e-06,
|
| 345 |
+
"loss": 0.0,
|
| 346 |
+
"step": 30000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.4700117600940807,
|
| 350 |
+
"eval_cosine_accuracy": 0.5982123968690501,
|
| 351 |
+
"eval_cosine_accuracy_threshold": 0.9481082558631897,
|
| 352 |
+
"eval_cosine_ap": 0.6620788779311684,
|
| 353 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 354 |
+
"eval_cosine_f1_threshold": -0.026352988556027412,
|
| 355 |
+
"eval_cosine_mcc": 0.0,
|
| 356 |
+
"eval_cosine_precision": 0.5,
|
| 357 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 358 |
+
"eval_loss": 25.89101791381836,
|
| 359 |
+
"eval_runtime": 23.4116,
|
| 360 |
+
"eval_samples_per_second": 807.635,
|
| 361 |
+
"eval_steps_per_second": 1.58,
|
| 362 |
+
"step": 30000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.5680125441003527,
|
| 366 |
+
"grad_norm": 0.0,
|
| 367 |
+
"learning_rate": 3.2007092198581564e-06,
|
| 368 |
+
"loss": 0.0,
|
| 369 |
+
"step": 32000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.5680125441003527,
|
| 373 |
+
"eval_cosine_accuracy": 0.5983181722022425,
|
| 374 |
+
"eval_cosine_accuracy_threshold": 0.947978138923645,
|
| 375 |
+
"eval_cosine_ap": 0.6623524771391387,
|
| 376 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 377 |
+
"eval_cosine_f1_threshold": -0.02726462110877037,
|
| 378 |
+
"eval_cosine_mcc": 0.0,
|
| 379 |
+
"eval_cosine_precision": 0.5,
|
| 380 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 381 |
+
"eval_loss": 25.8829402923584,
|
| 382 |
+
"eval_runtime": 23.5325,
|
| 383 |
+
"eval_samples_per_second": 803.484,
|
| 384 |
+
"eval_steps_per_second": 1.572,
|
| 385 |
+
"step": 32000
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.666013328106625,
|
| 389 |
+
"grad_norm": 0.0,
|
| 390 |
+
"learning_rate": 3.0717601547388783e-06,
|
| 391 |
+
"loss": 0.0,
|
| 392 |
+
"step": 34000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.666013328106625,
|
| 396 |
+
"eval_cosine_accuracy": 0.5985826105352232,
|
| 397 |
+
"eval_cosine_accuracy_threshold": 0.9477732181549072,
|
| 398 |
+
"eval_cosine_ap": 0.6626794204198572,
|
| 399 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 400 |
+
"eval_cosine_f1_threshold": -0.027758542448282242,
|
| 401 |
+
"eval_cosine_mcc": 0.0,
|
| 402 |
+
"eval_cosine_precision": 0.5,
|
| 403 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 404 |
+
"eval_loss": 25.918596267700195,
|
| 405 |
+
"eval_runtime": 23.9566,
|
| 406 |
+
"eval_samples_per_second": 789.259,
|
| 407 |
+
"eval_steps_per_second": 1.544,
|
| 408 |
+
"step": 34000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.764014112112897,
|
| 412 |
+
"grad_norm": 0.0,
|
| 413 |
+
"learning_rate": 2.9428110896196007e-06,
|
| 414 |
+
"loss": 0.0,
|
| 415 |
+
"step": 36000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.764014112112897,
|
| 419 |
+
"eval_cosine_accuracy": 0.598847048868204,
|
| 420 |
+
"eval_cosine_accuracy_threshold": 0.9478007555007935,
|
| 421 |
+
"eval_cosine_ap": 0.6629160839294745,
|
| 422 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 423 |
+
"eval_cosine_f1_threshold": -0.028077710419893265,
|
| 424 |
+
"eval_cosine_mcc": 0.0,
|
| 425 |
+
"eval_cosine_precision": 0.5,
|
| 426 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 427 |
+
"eval_loss": 25.906251907348633,
|
| 428 |
+
"eval_runtime": 22.9116,
|
| 429 |
+
"eval_samples_per_second": 825.26,
|
| 430 |
+
"eval_steps_per_second": 1.615,
|
| 431 |
+
"step": 36000
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.862014896119169,
|
| 435 |
+
"grad_norm": 0.00737870205193758,
|
| 436 |
+
"learning_rate": 2.8147646679561574e-06,
|
| 437 |
+
"loss": 1.5404,
|
| 438 |
+
"step": 38000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.862014896119169,
|
| 442 |
+
"eval_cosine_accuracy": 0.6216416331711445,
|
| 443 |
+
"eval_cosine_accuracy_threshold": 0.9999633431434631,
|
| 444 |
+
"eval_cosine_ap": 0.6896995703091671,
|
| 445 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 446 |
+
"eval_cosine_f1_threshold": 0.9949546456336975,
|
| 447 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 448 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 449 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 450 |
+
"eval_loss": 28.107051849365234,
|
| 451 |
+
"eval_runtime": 23.2921,
|
| 452 |
+
"eval_samples_per_second": 811.777,
|
| 453 |
+
"eval_steps_per_second": 1.589,
|
| 454 |
+
"step": 38000
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.9600156801254411,
|
| 458 |
+
"grad_norm": 0.004108751192688942,
|
| 459 |
+
"learning_rate": 2.68581560283688e-06,
|
| 460 |
+
"loss": 0.0017,
|
| 461 |
+
"step": 40000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.9600156801254411,
|
| 465 |
+
"eval_cosine_accuracy": 0.623598476835202,
|
| 466 |
+
"eval_cosine_accuracy_threshold": 0.9999710917472839,
|
| 467 |
+
"eval_cosine_ap": 0.696279151150395,
|
| 468 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 469 |
+
"eval_cosine_f1_threshold": 0.9948583245277405,
|
| 470 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 471 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 472 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 473 |
+
"eval_loss": 27.634450912475586,
|
| 474 |
+
"eval_runtime": 23.2848,
|
| 475 |
+
"eval_samples_per_second": 812.033,
|
| 476 |
+
"eval_steps_per_second": 1.589,
|
| 477 |
+
"step": 40000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 2.058016464131713,
|
| 481 |
+
"grad_norm": 0.002198939910158515,
|
| 482 |
+
"learning_rate": 2.5568665377176017e-06,
|
| 483 |
+
"loss": 0.0007,
|
| 484 |
+
"step": 42000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 2.058016464131713,
|
| 488 |
+
"eval_cosine_accuracy": 0.6238100275015867,
|
| 489 |
+
"eval_cosine_accuracy_threshold": 0.9999785423278809,
|
| 490 |
+
"eval_cosine_ap": 0.7001012970861294,
|
| 491 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 492 |
+
"eval_cosine_f1_threshold": 0.9943636655807495,
|
| 493 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 494 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 495 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 496 |
+
"eval_loss": 26.901527404785156,
|
| 497 |
+
"eval_runtime": 24.7872,
|
| 498 |
+
"eval_samples_per_second": 762.814,
|
| 499 |
+
"eval_steps_per_second": 1.493,
|
| 500 |
+
"step": 42000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.156017248137985,
|
| 504 |
+
"grad_norm": 0.0011129860067740083,
|
| 505 |
+
"learning_rate": 2.427917472598324e-06,
|
| 506 |
+
"loss": 0.0003,
|
| 507 |
+
"step": 44000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.156017248137985,
|
| 511 |
+
"eval_cosine_accuracy": 0.6339644594880474,
|
| 512 |
+
"eval_cosine_accuracy_threshold": 0.9999836683273315,
|
| 513 |
+
"eval_cosine_ap": 0.7180590205623526,
|
| 514 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 515 |
+
"eval_cosine_f1_threshold": 0.9934605360031128,
|
| 516 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 517 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 518 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 519 |
+
"eval_loss": 26.755794525146484,
|
| 520 |
+
"eval_runtime": 23.8512,
|
| 521 |
+
"eval_samples_per_second": 792.75,
|
| 522 |
+
"eval_steps_per_second": 1.551,
|
| 523 |
+
"step": 44000
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 2.2540180321442573,
|
| 527 |
+
"grad_norm": 0.0,
|
| 528 |
+
"learning_rate": 2.298968407479046e-06,
|
| 529 |
+
"loss": 0.9002,
|
| 530 |
+
"step": 46000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.2540180321442573,
|
| 534 |
+
"eval_cosine_accuracy": 0.5470171356039771,
|
| 535 |
+
"eval_cosine_accuracy_threshold": 0.9858599305152893,
|
| 536 |
+
"eval_cosine_ap": 0.5604867418651215,
|
| 537 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 538 |
+
"eval_cosine_f1_threshold": -0.4505715072154999,
|
| 539 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 540 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 541 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 542 |
+
"eval_loss": 41.99852752685547,
|
| 543 |
+
"eval_runtime": 23.9606,
|
| 544 |
+
"eval_samples_per_second": 789.128,
|
| 545 |
+
"eval_steps_per_second": 1.544,
|
| 546 |
+
"step": 46000
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 2.352018816150529,
|
| 550 |
+
"grad_norm": 0.0,
|
| 551 |
+
"learning_rate": 2.1700193423597684e-06,
|
| 552 |
+
"loss": 0.0,
|
| 553 |
+
"step": 48000
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.352018816150529,
|
| 557 |
+
"eval_cosine_accuracy": 0.5442140892743812,
|
| 558 |
+
"eval_cosine_accuracy_threshold": 0.9832857847213745,
|
| 559 |
+
"eval_cosine_ap": 0.55157056706159,
|
| 560 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 561 |
+
"eval_cosine_f1_threshold": -0.2545183300971985,
|
| 562 |
+
"eval_cosine_mcc": 0.007272583024577166,
|
| 563 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 564 |
+
"eval_cosine_recall": 1.0,
|
| 565 |
+
"eval_loss": 44.62057876586914,
|
| 566 |
+
"eval_runtime": 24.1334,
|
| 567 |
+
"eval_samples_per_second": 783.478,
|
| 568 |
+
"eval_steps_per_second": 1.533,
|
| 569 |
+
"step": 48000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4500196001568013,
|
| 573 |
+
"grad_norm": 0.0,
|
| 574 |
+
"learning_rate": 2.0410702772404903e-06,
|
| 575 |
+
"loss": 0.0,
|
| 576 |
+
"step": 50000
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.4500196001568013,
|
| 580 |
+
"eval_cosine_accuracy": 0.5505606092659192,
|
| 581 |
+
"eval_cosine_accuracy_threshold": 0.9857516288757324,
|
| 582 |
+
"eval_cosine_ap": 0.5683410397390614,
|
| 583 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 584 |
+
"eval_cosine_f1_threshold": -0.29352766275405884,
|
| 585 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 586 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 587 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 588 |
+
"eval_loss": 40.676856994628906,
|
| 589 |
+
"eval_runtime": 25.7266,
|
| 590 |
+
"eval_samples_per_second": 734.96,
|
| 591 |
+
"eval_steps_per_second": 1.438,
|
| 592 |
+
"step": 50000
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 2.548020384163073,
|
| 596 |
+
"grad_norm": 0.0,
|
| 597 |
+
"learning_rate": 1.9121212121212123e-06,
|
| 598 |
+
"loss": 0.0,
|
| 599 |
+
"step": 52000
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 2.548020384163073,
|
| 603 |
+
"eval_cosine_accuracy": 0.5527290035963613,
|
| 604 |
+
"eval_cosine_accuracy_threshold": 0.9838961362838745,
|
| 605 |
+
"eval_cosine_ap": 0.5783079089292276,
|
| 606 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 607 |
+
"eval_cosine_f1_threshold": -0.3105597496032715,
|
| 608 |
+
"eval_cosine_mcc": 0.007272583024577166,
|
| 609 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 610 |
+
"eval_cosine_recall": 1.0,
|
| 611 |
+
"eval_loss": 49.68011474609375,
|
| 612 |
+
"eval_runtime": 24.0384,
|
| 613 |
+
"eval_samples_per_second": 786.575,
|
| 614 |
+
"eval_steps_per_second": 1.539,
|
| 615 |
+
"step": 52000
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.6460211681693453,
|
| 619 |
+
"grad_norm": 0.0,
|
| 620 |
+
"learning_rate": 1.7831721470019344e-06,
|
| 621 |
+
"loss": 0.0,
|
| 622 |
+
"step": 54000
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.6460211681693453,
|
| 626 |
+
"eval_cosine_accuracy": 0.5522530145969959,
|
| 627 |
+
"eval_cosine_accuracy_threshold": 0.9908125400543213,
|
| 628 |
+
"eval_cosine_ap": 0.573218654996148,
|
| 629 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 630 |
+
"eval_cosine_f1_threshold": -0.12428700923919678,
|
| 631 |
+
"eval_cosine_mcc": -0.007272583024577166,
|
| 632 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 633 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 634 |
+
"eval_loss": 33.039730072021484,
|
| 635 |
+
"eval_runtime": 23.0962,
|
| 636 |
+
"eval_samples_per_second": 818.662,
|
| 637 |
+
"eval_steps_per_second": 1.602,
|
| 638 |
+
"step": 54000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.7440219521756175,
|
| 642 |
+
"grad_norm": 0.0,
|
| 643 |
+
"learning_rate": 1.6542230818826566e-06,
|
| 644 |
+
"loss": 0.0,
|
| 645 |
+
"step": 56000
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 2.7440219521756175,
|
| 649 |
+
"eval_cosine_accuracy": 0.5532049925957266,
|
| 650 |
+
"eval_cosine_accuracy_threshold": 0.9827109575271606,
|
| 651 |
+
"eval_cosine_ap": 0.5732697527690604,
|
| 652 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 653 |
+
"eval_cosine_f1_threshold": -0.32465463876724243,
|
| 654 |
+
"eval_cosine_mcc": 0.007272583024577166,
|
| 655 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 656 |
+
"eval_cosine_recall": 1.0,
|
| 657 |
+
"eval_loss": 48.252845764160156,
|
| 658 |
+
"eval_runtime": 24.2479,
|
| 659 |
+
"eval_samples_per_second": 779.779,
|
| 660 |
+
"eval_steps_per_second": 1.526,
|
| 661 |
+
"step": 56000
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 2.8420227361818897,
|
| 665 |
+
"grad_norm": 0.0,
|
| 666 |
+
"learning_rate": 1.5252740167633787e-06,
|
| 667 |
+
"loss": 0.0,
|
| 668 |
+
"step": 58000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.8420227361818897,
|
| 672 |
+
"eval_cosine_accuracy": 0.5468584726041887,
|
| 673 |
+
"eval_cosine_accuracy_threshold": 0.9873851537704468,
|
| 674 |
+
"eval_cosine_ap": 0.5562632644839256,
|
| 675 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 676 |
+
"eval_cosine_f1_threshold": -0.1655093878507614,
|
| 677 |
+
"eval_cosine_mcc": 0.0,
|
| 678 |
+
"eval_cosine_precision": 0.5,
|
| 679 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 680 |
+
"eval_loss": 39.66588592529297,
|
| 681 |
+
"eval_runtime": 23.254,
|
| 682 |
+
"eval_samples_per_second": 813.108,
|
| 683 |
+
"eval_steps_per_second": 1.591,
|
| 684 |
+
"step": 58000
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.9400235201881615,
|
| 688 |
+
"grad_norm": 0.0,
|
| 689 |
+
"learning_rate": 1.3963249516441009e-06,
|
| 690 |
+
"loss": 0.0,
|
| 691 |
+
"step": 60000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 2.9400235201881615,
|
| 695 |
+
"eval_cosine_accuracy": 0.5468584726041887,
|
| 696 |
+
"eval_cosine_accuracy_threshold": 0.9873851537704468,
|
| 697 |
+
"eval_cosine_ap": 0.5562632644839256,
|
| 698 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 699 |
+
"eval_cosine_f1_threshold": -0.1655093878507614,
|
| 700 |
+
"eval_cosine_mcc": 0.0,
|
| 701 |
+
"eval_cosine_precision": 0.5,
|
| 702 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 703 |
+
"eval_loss": 39.66588592529297,
|
| 704 |
+
"eval_runtime": 23.1741,
|
| 705 |
+
"eval_samples_per_second": 815.91,
|
| 706 |
+
"eval_steps_per_second": 1.597,
|
| 707 |
+
"step": 60000
|
| 708 |
+
}
|
| 709 |
+
],
|
| 710 |
+
"logging_steps": 2000,
|
| 711 |
+
"max_steps": 81632,
|
| 712 |
+
"num_input_tokens_seen": 0,
|
| 713 |
+
"num_train_epochs": 4,
|
| 714 |
+
"save_steps": 500,
|
| 715 |
+
"stateful_callbacks": {
|
| 716 |
+
"TrainerControl": {
|
| 717 |
+
"args": {
|
| 718 |
+
"should_epoch_stop": false,
|
| 719 |
+
"should_evaluate": false,
|
| 720 |
+
"should_log": false,
|
| 721 |
+
"should_save": true,
|
| 722 |
+
"should_training_stop": false
|
| 723 |
+
},
|
| 724 |
+
"attributes": {}
|
| 725 |
+
}
|
| 726 |
+
},
|
| 727 |
+
"total_flos": 0.0,
|
| 728 |
+
"train_batch_size": 512,
|
| 729 |
+
"trial_name": null,
|
| 730 |
+
"trial_params": null
|
| 731 |
+
}
|
checkpoint-61224/training_metrics.json
ADDED
|
@@ -0,0 +1,396 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metrics_history": [
|
| 3 |
+
{
|
| 4 |
+
"epoch": 0.09800078400627205,
|
| 5 |
+
"global_step": 2000,
|
| 6 |
+
"eval_loss": 26.897747039794922,
|
| 7 |
+
"eval_cosine_accuracy": 0.6100063465199915,
|
| 8 |
+
"eval_cosine_accuracy_threshold": 0.9997795820236206,
|
| 9 |
+
"eval_cosine_f1": 0.6667136862150438,
|
| 10 |
+
"eval_cosine_f1_threshold": 0.991624653339386,
|
| 11 |
+
"eval_cosine_precision": 0.5000793524837327,
|
| 12 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 13 |
+
"eval_cosine_ap": 0.6786677205526788,
|
| 14 |
+
"eval_cosine_mcc": 0.00975822629781094
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.1960015680125441,
|
| 18 |
+
"global_step": 4000,
|
| 19 |
+
"eval_loss": 27.12621307373047,
|
| 20 |
+
"eval_cosine_accuracy": 0.6240215781679712,
|
| 21 |
+
"eval_cosine_accuracy_threshold": 0.9999226331710815,
|
| 22 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 23 |
+
"eval_cosine_f1_threshold": 0.9939442873001099,
|
| 24 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 25 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 26 |
+
"eval_cosine_ap": 0.6923415808404929,
|
| 27 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.29400235201881614,
|
| 31 |
+
"global_step": 6000,
|
| 32 |
+
"eval_loss": 32.61564254760742,
|
| 33 |
+
"eval_cosine_accuracy": 0.5835096255553205,
|
| 34 |
+
"eval_cosine_accuracy_threshold": 0.9478387236595154,
|
| 35 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 36 |
+
"eval_cosine_f1_threshold": -0.09354762732982635,
|
| 37 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 38 |
+
"eval_cosine_recall": 1.0,
|
| 39 |
+
"eval_cosine_ap": 0.6313955266297183,
|
| 40 |
+
"eval_cosine_mcc": 0.007272583024577166
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.3920031360250882,
|
| 44 |
+
"global_step": 8000,
|
| 45 |
+
"eval_loss": 33.463409423828125,
|
| 46 |
+
"eval_cosine_accuracy": 0.5764226782314364,
|
| 47 |
+
"eval_cosine_accuracy_threshold": 0.953868567943573,
|
| 48 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 49 |
+
"eval_cosine_f1_threshold": -0.09449770301580429,
|
| 50 |
+
"eval_cosine_precision": 0.5,
|
| 51 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 52 |
+
"eval_cosine_ap": 0.6184434211752825,
|
| 53 |
+
"eval_cosine_mcc": 0.0
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.49000392003136023,
|
| 57 |
+
"global_step": 10000,
|
| 58 |
+
"eval_loss": 35.26945495605469,
|
| 59 |
+
"eval_cosine_accuracy": 0.5845673788872435,
|
| 60 |
+
"eval_cosine_accuracy_threshold": 0.9387475848197937,
|
| 61 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 62 |
+
"eval_cosine_f1_threshold": -0.1685822755098343,
|
| 63 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 64 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 65 |
+
"eval_cosine_ap": 0.6273588632357191,
|
| 66 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.5880047040376323,
|
| 70 |
+
"global_step": 12000,
|
| 71 |
+
"eval_loss": 30.115076065063477,
|
| 72 |
+
"eval_cosine_accuracy": 0.5943515972075312,
|
| 73 |
+
"eval_cosine_accuracy_threshold": 0.9397846460342407,
|
| 74 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 75 |
+
"eval_cosine_f1_threshold": -0.17127826809883118,
|
| 76 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 77 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 78 |
+
"eval_cosine_ap": 0.6521863345575756,
|
| 79 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.6860054880439044,
|
| 83 |
+
"global_step": 14000,
|
| 84 |
+
"eval_loss": 30.446081161499023,
|
| 85 |
+
"eval_cosine_accuracy": 0.5931351808758197,
|
| 86 |
+
"eval_cosine_accuracy_threshold": 0.9391266107559204,
|
| 87 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 88 |
+
"eval_cosine_f1_threshold": -0.1697666049003601,
|
| 89 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 90 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 91 |
+
"eval_cosine_ap": 0.6494508725044261,
|
| 92 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.7840062720501764,
|
| 96 |
+
"global_step": 16000,
|
| 97 |
+
"eval_loss": 30.447351455688477,
|
| 98 |
+
"eval_cosine_accuracy": 0.5934525068753966,
|
| 99 |
+
"eval_cosine_accuracy_threshold": 0.9391290545463562,
|
| 100 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 101 |
+
"eval_cosine_f1_threshold": -0.17070035636425018,
|
| 102 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 103 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 104 |
+
"eval_cosine_ap": 0.6498809737033857,
|
| 105 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.8820070560564485,
|
| 109 |
+
"global_step": 18000,
|
| 110 |
+
"eval_loss": 30.453508377075195,
|
| 111 |
+
"eval_cosine_accuracy": 0.593558282208589,
|
| 112 |
+
"eval_cosine_accuracy_threshold": 0.9389414191246033,
|
| 113 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 114 |
+
"eval_cosine_f1_threshold": -0.17154842615127563,
|
| 115 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 116 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 117 |
+
"eval_cosine_ap": 0.6502482026720758,
|
| 118 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.9800078400627205,
|
| 122 |
+
"global_step": 20000,
|
| 123 |
+
"eval_loss": 30.439617156982422,
|
| 124 |
+
"eval_cosine_accuracy": 0.5937169452083774,
|
| 125 |
+
"eval_cosine_accuracy_threshold": 0.9388357400894165,
|
| 126 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 127 |
+
"eval_cosine_f1_threshold": -0.1724291741847992,
|
| 128 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 129 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 130 |
+
"eval_cosine_ap": 0.6506732897432272,
|
| 131 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.0780086240689926,
|
| 135 |
+
"global_step": 22000,
|
| 136 |
+
"eval_loss": 30.44361686706543,
|
| 137 |
+
"eval_cosine_accuracy": 0.5937698328749735,
|
| 138 |
+
"eval_cosine_accuracy_threshold": 0.9385948181152344,
|
| 139 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 140 |
+
"eval_cosine_f1_threshold": -0.17285144329071045,
|
| 141 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 142 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 143 |
+
"eval_cosine_ap": 0.6509728340267652,
|
| 144 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.1760094080752646,
|
| 148 |
+
"global_step": 24000,
|
| 149 |
+
"eval_loss": 30.4697208404541,
|
| 150 |
+
"eval_cosine_accuracy": 0.593928495874762,
|
| 151 |
+
"eval_cosine_accuracy_threshold": 0.9375462532043457,
|
| 152 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 153 |
+
"eval_cosine_f1_threshold": -0.17358532547950745,
|
| 154 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 155 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 156 |
+
"eval_cosine_ap": 0.6513894091627628,
|
| 157 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.2740101920815365,
|
| 161 |
+
"global_step": 26000,
|
| 162 |
+
"eval_loss": 30.457592010498047,
|
| 163 |
+
"eval_cosine_accuracy": 0.5941400465411466,
|
| 164 |
+
"eval_cosine_accuracy_threshold": 0.9373874664306641,
|
| 165 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 166 |
+
"eval_cosine_f1_threshold": -0.17465540766716003,
|
| 167 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 168 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 169 |
+
"eval_cosine_ap": 0.6517436899727478,
|
| 170 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.3720109760878088,
|
| 174 |
+
"global_step": 28000,
|
| 175 |
+
"eval_loss": 27.762596130371094,
|
| 176 |
+
"eval_cosine_accuracy": 0.6150306748466258,
|
| 177 |
+
"eval_cosine_accuracy_threshold": 0.9150234460830688,
|
| 178 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 179 |
+
"eval_cosine_f1_threshold": 0.005598240531980991,
|
| 180 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 181 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 182 |
+
"eval_cosine_ap": 0.6887500197266508,
|
| 183 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 1.4700117600940807,
|
| 187 |
+
"global_step": 30000,
|
| 188 |
+
"eval_loss": 25.89101791381836,
|
| 189 |
+
"eval_cosine_accuracy": 0.5982123968690501,
|
| 190 |
+
"eval_cosine_accuracy_threshold": 0.9481082558631897,
|
| 191 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 192 |
+
"eval_cosine_f1_threshold": -0.026352988556027412,
|
| 193 |
+
"eval_cosine_precision": 0.5,
|
| 194 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 195 |
+
"eval_cosine_ap": 0.6620788779311684,
|
| 196 |
+
"eval_cosine_mcc": 0.0
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.5680125441003527,
|
| 200 |
+
"global_step": 32000,
|
| 201 |
+
"eval_loss": 25.8829402923584,
|
| 202 |
+
"eval_cosine_accuracy": 0.5983181722022425,
|
| 203 |
+
"eval_cosine_accuracy_threshold": 0.947978138923645,
|
| 204 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 205 |
+
"eval_cosine_f1_threshold": -0.02726462110877037,
|
| 206 |
+
"eval_cosine_precision": 0.5,
|
| 207 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 208 |
+
"eval_cosine_ap": 0.6623524771391387,
|
| 209 |
+
"eval_cosine_mcc": 0.0
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.666013328106625,
|
| 213 |
+
"global_step": 34000,
|
| 214 |
+
"eval_loss": 25.918596267700195,
|
| 215 |
+
"eval_cosine_accuracy": 0.5985826105352232,
|
| 216 |
+
"eval_cosine_accuracy_threshold": 0.9477732181549072,
|
| 217 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 218 |
+
"eval_cosine_f1_threshold": -0.027758542448282242,
|
| 219 |
+
"eval_cosine_precision": 0.5,
|
| 220 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 221 |
+
"eval_cosine_ap": 0.6626794204198572,
|
| 222 |
+
"eval_cosine_mcc": 0.0
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.764014112112897,
|
| 226 |
+
"global_step": 36000,
|
| 227 |
+
"eval_loss": 25.906251907348633,
|
| 228 |
+
"eval_cosine_accuracy": 0.598847048868204,
|
| 229 |
+
"eval_cosine_accuracy_threshold": 0.9478007555007935,
|
| 230 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 231 |
+
"eval_cosine_f1_threshold": -0.028077710419893265,
|
| 232 |
+
"eval_cosine_precision": 0.5,
|
| 233 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 234 |
+
"eval_cosine_ap": 0.6629160839294745,
|
| 235 |
+
"eval_cosine_mcc": 0.0
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.862014896119169,
|
| 239 |
+
"global_step": 38000,
|
| 240 |
+
"eval_loss": 28.107051849365234,
|
| 241 |
+
"eval_cosine_accuracy": 0.6216416331711445,
|
| 242 |
+
"eval_cosine_accuracy_threshold": 0.9999633431434631,
|
| 243 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 244 |
+
"eval_cosine_f1_threshold": 0.9949546456336975,
|
| 245 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 246 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 247 |
+
"eval_cosine_ap": 0.6896995703091671,
|
| 248 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.9600156801254411,
|
| 252 |
+
"global_step": 40000,
|
| 253 |
+
"eval_loss": 27.634450912475586,
|
| 254 |
+
"eval_cosine_accuracy": 0.623598476835202,
|
| 255 |
+
"eval_cosine_accuracy_threshold": 0.9999710917472839,
|
| 256 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 257 |
+
"eval_cosine_f1_threshold": 0.9948583245277405,
|
| 258 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 259 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 260 |
+
"eval_cosine_ap": 0.696279151150395,
|
| 261 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 2.058016464131713,
|
| 265 |
+
"global_step": 42000,
|
| 266 |
+
"eval_loss": 26.901527404785156,
|
| 267 |
+
"eval_cosine_accuracy": 0.6238100275015867,
|
| 268 |
+
"eval_cosine_accuracy_threshold": 0.9999785423278809,
|
| 269 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 270 |
+
"eval_cosine_f1_threshold": 0.9943636655807495,
|
| 271 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 272 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 273 |
+
"eval_cosine_ap": 0.7001012970861294,
|
| 274 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.156017248137985,
|
| 278 |
+
"global_step": 44000,
|
| 279 |
+
"eval_loss": 26.755794525146484,
|
| 280 |
+
"eval_cosine_accuracy": 0.6339644594880474,
|
| 281 |
+
"eval_cosine_accuracy_threshold": 0.9999836683273315,
|
| 282 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 283 |
+
"eval_cosine_f1_threshold": 0.9934605360031128,
|
| 284 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 285 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 286 |
+
"eval_cosine_ap": 0.7180590205623526,
|
| 287 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 2.2540180321442573,
|
| 291 |
+
"global_step": 46000,
|
| 292 |
+
"eval_loss": 41.99852752685547,
|
| 293 |
+
"eval_cosine_accuracy": 0.5470171356039771,
|
| 294 |
+
"eval_cosine_accuracy_threshold": 0.9858599305152893,
|
| 295 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 296 |
+
"eval_cosine_f1_threshold": -0.4505715072154999,
|
| 297 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 298 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 299 |
+
"eval_cosine_ap": 0.5604867418651215,
|
| 300 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.352018816150529,
|
| 304 |
+
"global_step": 48000,
|
| 305 |
+
"eval_loss": 44.62057876586914,
|
| 306 |
+
"eval_cosine_accuracy": 0.5442140892743812,
|
| 307 |
+
"eval_cosine_accuracy_threshold": 0.9832857847213745,
|
| 308 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 309 |
+
"eval_cosine_f1_threshold": -0.2545183300971985,
|
| 310 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 311 |
+
"eval_cosine_recall": 1.0,
|
| 312 |
+
"eval_cosine_ap": 0.55157056706159,
|
| 313 |
+
"eval_cosine_mcc": 0.007272583024577166
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.4500196001568013,
|
| 317 |
+
"global_step": 50000,
|
| 318 |
+
"eval_loss": 40.676856994628906,
|
| 319 |
+
"eval_cosine_accuracy": 0.5505606092659192,
|
| 320 |
+
"eval_cosine_accuracy_threshold": 0.9857516288757324,
|
| 321 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 322 |
+
"eval_cosine_f1_threshold": -0.29352766275405884,
|
| 323 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 324 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 325 |
+
"eval_cosine_ap": 0.5683410397390614,
|
| 326 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.548020384163073,
|
| 330 |
+
"global_step": 52000,
|
| 331 |
+
"eval_loss": 49.68011474609375,
|
| 332 |
+
"eval_cosine_accuracy": 0.5527290035963613,
|
| 333 |
+
"eval_cosine_accuracy_threshold": 0.9838961362838745,
|
| 334 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 335 |
+
"eval_cosine_f1_threshold": -0.3105597496032715,
|
| 336 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 337 |
+
"eval_cosine_recall": 1.0,
|
| 338 |
+
"eval_cosine_ap": 0.5783079089292276,
|
| 339 |
+
"eval_cosine_mcc": 0.007272583024577166
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.6460211681693453,
|
| 343 |
+
"global_step": 54000,
|
| 344 |
+
"eval_loss": 33.039730072021484,
|
| 345 |
+
"eval_cosine_accuracy": 0.5522530145969959,
|
| 346 |
+
"eval_cosine_accuracy_threshold": 0.9908125400543213,
|
| 347 |
+
"eval_cosine_f1": 0.6666196537498678,
|
| 348 |
+
"eval_cosine_f1_threshold": -0.12428700923919678,
|
| 349 |
+
"eval_cosine_precision": 0.4999735547680753,
|
| 350 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 351 |
+
"eval_cosine_ap": 0.573218654996148,
|
| 352 |
+
"eval_cosine_mcc": -0.007272583024577166
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.7440219521756175,
|
| 356 |
+
"global_step": 56000,
|
| 357 |
+
"eval_loss": 48.252845764160156,
|
| 358 |
+
"eval_cosine_accuracy": 0.5532049925957266,
|
| 359 |
+
"eval_cosine_accuracy_threshold": 0.9827109575271606,
|
| 360 |
+
"eval_cosine_f1": 0.6666901731250661,
|
| 361 |
+
"eval_cosine_f1_threshold": -0.32465463876724243,
|
| 362 |
+
"eval_cosine_precision": 0.5000264452319246,
|
| 363 |
+
"eval_cosine_recall": 1.0,
|
| 364 |
+
"eval_cosine_ap": 0.5732697527690604,
|
| 365 |
+
"eval_cosine_mcc": 0.007272583024577166
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 2.8420227361818897,
|
| 369 |
+
"global_step": 58000,
|
| 370 |
+
"eval_loss": 39.66588592529297,
|
| 371 |
+
"eval_cosine_accuracy": 0.5468584726041887,
|
| 372 |
+
"eval_cosine_accuracy_threshold": 0.9873851537704468,
|
| 373 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 374 |
+
"eval_cosine_f1_threshold": -0.1655093878507614,
|
| 375 |
+
"eval_cosine_precision": 0.5,
|
| 376 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 377 |
+
"eval_cosine_ap": 0.5562632644839256,
|
| 378 |
+
"eval_cosine_mcc": 0.0
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 2.9400235201881615,
|
| 382 |
+
"global_step": 60000,
|
| 383 |
+
"eval_loss": 39.66588592529297,
|
| 384 |
+
"eval_cosine_accuracy": 0.5468584726041887,
|
| 385 |
+
"eval_cosine_accuracy_threshold": 0.9873851537704468,
|
| 386 |
+
"eval_cosine_f1": 0.6666431593794077,
|
| 387 |
+
"eval_cosine_f1_threshold": -0.1655093878507614,
|
| 388 |
+
"eval_cosine_precision": 0.5,
|
| 389 |
+
"eval_cosine_recall": 0.9998942246668077,
|
| 390 |
+
"eval_cosine_ap": 0.5562632644839256,
|
| 391 |
+
"eval_cosine_mcc": 0.0
|
| 392 |
+
}
|
| 393 |
+
],
|
| 394 |
+
"current_epoch": 3,
|
| 395 |
+
"global_step": 61224
|
| 396 |
+
}
|
checkpoint-61224/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|