Training in progress, epoch 29, checkpoint
Browse files- last-checkpoint/config.json +46 -0
- last-checkpoint/model.safetensors +3 -0
- last-checkpoint/optimizer.pt +3 -0
- last-checkpoint/rng_state.pth +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/special_tokens_map.json +37 -0
- last-checkpoint/tokenizer.json +0 -0
- last-checkpoint/tokenizer_config.json +59 -0
- last-checkpoint/trainer_state.json +2730 -0
- last-checkpoint/training_args.bin +3 -0
- last-checkpoint/vocab.txt +0 -0
last-checkpoint/config.json
ADDED
|
@@ -0,0 +1,46 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "pt_bobai9192tok_freeze_emb_after_final_f",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"directionality": "bidi",
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.2,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "LABEL_0",
|
| 14 |
+
"1": "LABEL_1",
|
| 15 |
+
"2": "LABEL_2",
|
| 16 |
+
"3": "LABEL_3",
|
| 17 |
+
"4": "LABEL_4"
|
| 18 |
+
},
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 3072,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"LABEL_0": 0,
|
| 23 |
+
"LABEL_1": 1,
|
| 24 |
+
"LABEL_2": 2,
|
| 25 |
+
"LABEL_3": 3,
|
| 26 |
+
"LABEL_4": 4
|
| 27 |
+
},
|
| 28 |
+
"layer_norm_eps": 1e-12,
|
| 29 |
+
"max_position_embeddings": 512,
|
| 30 |
+
"model_type": "bert",
|
| 31 |
+
"num_attention_heads": 12,
|
| 32 |
+
"num_hidden_layers": 12,
|
| 33 |
+
"pad_token_id": 0,
|
| 34 |
+
"pooler_fc_size": 768,
|
| 35 |
+
"pooler_num_attention_heads": 12,
|
| 36 |
+
"pooler_num_fc_layers": 3,
|
| 37 |
+
"pooler_size_per_head": 128,
|
| 38 |
+
"pooler_type": "first_token_transform",
|
| 39 |
+
"position_embedding_type": "absolute",
|
| 40 |
+
"problem_type": "single_label_classification",
|
| 41 |
+
"torch_dtype": "float32",
|
| 42 |
+
"transformers_version": "4.42.4",
|
| 43 |
+
"type_vocab_size": 2,
|
| 44 |
+
"use_cache": true,
|
| 45 |
+
"vocab_size": 105879
|
| 46 |
+
}
|
last-checkpoint/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14abdf83fd59c4b56bbb236f9dc331b562128e11ff932f31f7ca1ae72425e9de
|
| 3 |
+
size 669464588
|
last-checkpoint/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c7513dfa09b26b044bca20dce2cfcf44f6b816275a2fd6e80987c703c3074e0
|
| 3 |
+
size 1339050234
|
last-checkpoint/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0619fcc05420edf096942dd88e18e541d3f50ad3e05fe1a8554228f39d2ccd79
|
| 3 |
+
size 14244
|
last-checkpoint/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fcbbdb17194da83c69961c26179188ec0752fbbff5804b4fa0a11bd799e5507f
|
| 3 |
+
size 1064
|
last-checkpoint/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 |
+
}
|
last-checkpoint/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
last-checkpoint/tokenizer_config.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"1": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"4": {
|
| 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": true,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"mask_token": "[MASK]",
|
| 48 |
+
"max_length": 512,
|
| 49 |
+
"model_max_length": 512,
|
| 50 |
+
"pad_token": "[PAD]",
|
| 51 |
+
"sep_token": "[SEP]",
|
| 52 |
+
"stride": 0,
|
| 53 |
+
"strip_accents": null,
|
| 54 |
+
"tokenize_chinese_chars": true,
|
| 55 |
+
"tokenizer_class": "BertTokenizer",
|
| 56 |
+
"truncation_side": "right",
|
| 57 |
+
"truncation_strategy": "longest_first",
|
| 58 |
+
"unk_token": "[UNK]"
|
| 59 |
+
}
|
last-checkpoint/trainer_state.json
ADDED
|
@@ -0,0 +1,2730 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8734395170821287,
|
| 3 |
+
"best_model_checkpoint": "Evier/9192_final_model/checkpoint-168",
|
| 4 |
+
"epoch": 29.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 348,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.08333333333333333,
|
| 13 |
+
"grad_norm": 1.508326530456543,
|
| 14 |
+
"learning_rate": 1.9999619230641714e-05,
|
| 15 |
+
"loss": 1.6372,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.16666666666666666,
|
| 20 |
+
"grad_norm": 1.9398980140686035,
|
| 21 |
+
"learning_rate": 1.9998476951563914e-05,
|
| 22 |
+
"loss": 1.6195,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.25,
|
| 27 |
+
"grad_norm": 1.470036506652832,
|
| 28 |
+
"learning_rate": 1.9996573249755573e-05,
|
| 29 |
+
"loss": 1.5765,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.3333333333333333,
|
| 34 |
+
"grad_norm": 1.6955764293670654,
|
| 35 |
+
"learning_rate": 1.999390827019096e-05,
|
| 36 |
+
"loss": 1.5741,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.4166666666666667,
|
| 41 |
+
"grad_norm": 1.6512634754180908,
|
| 42 |
+
"learning_rate": 1.999048221581858e-05,
|
| 43 |
+
"loss": 1.5339,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.5,
|
| 48 |
+
"grad_norm": 1.7195454835891724,
|
| 49 |
+
"learning_rate": 1.9986295347545738e-05,
|
| 50 |
+
"loss": 1.5036,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.5833333333333334,
|
| 55 |
+
"grad_norm": 1.891516089439392,
|
| 56 |
+
"learning_rate": 1.998134798421867e-05,
|
| 57 |
+
"loss": 1.4909,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.6666666666666666,
|
| 62 |
+
"grad_norm": 1.6854487657546997,
|
| 63 |
+
"learning_rate": 1.9975640502598243e-05,
|
| 64 |
+
"loss": 1.4888,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.75,
|
| 69 |
+
"grad_norm": 1.8361274003982544,
|
| 70 |
+
"learning_rate": 1.9969173337331283e-05,
|
| 71 |
+
"loss": 1.436,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.8333333333333334,
|
| 76 |
+
"grad_norm": 2.2142255306243896,
|
| 77 |
+
"learning_rate": 1.9961946980917457e-05,
|
| 78 |
+
"loss": 1.3977,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.9166666666666666,
|
| 83 |
+
"grad_norm": 1.7415549755096436,
|
| 84 |
+
"learning_rate": 1.9953961983671792e-05,
|
| 85 |
+
"loss": 1.4708,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.0,
|
| 90 |
+
"grad_norm": 1.9386205673217773,
|
| 91 |
+
"learning_rate": 1.9945218953682736e-05,
|
| 92 |
+
"loss": 1.4157,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.0,
|
| 97 |
+
"eval_f1": 0.45943301076734366,
|
| 98 |
+
"eval_loss": 1.333272099494934,
|
| 99 |
+
"eval_runtime": 0.1106,
|
| 100 |
+
"eval_samples_per_second": 1970.477,
|
| 101 |
+
"eval_steps_per_second": 36.156,
|
| 102 |
+
"step": 12
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.0833333333333333,
|
| 106 |
+
"grad_norm": 2.1051557064056396,
|
| 107 |
+
"learning_rate": 1.9935718556765878e-05,
|
| 108 |
+
"loss": 1.4316,
|
| 109 |
+
"step": 13
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.1666666666666667,
|
| 113 |
+
"grad_norm": 2.2947099208831787,
|
| 114 |
+
"learning_rate": 1.9925461516413224e-05,
|
| 115 |
+
"loss": 1.3958,
|
| 116 |
+
"step": 14
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.25,
|
| 120 |
+
"grad_norm": 1.881037950515747,
|
| 121 |
+
"learning_rate": 1.9914448613738107e-05,
|
| 122 |
+
"loss": 1.3299,
|
| 123 |
+
"step": 15
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.3333333333333333,
|
| 127 |
+
"grad_norm": 2.136141300201416,
|
| 128 |
+
"learning_rate": 1.9902680687415704e-05,
|
| 129 |
+
"loss": 1.2434,
|
| 130 |
+
"step": 16
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.4166666666666667,
|
| 134 |
+
"grad_norm": 2.0621819496154785,
|
| 135 |
+
"learning_rate": 1.989015863361917e-05,
|
| 136 |
+
"loss": 1.2908,
|
| 137 |
+
"step": 17
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.5,
|
| 141 |
+
"grad_norm": 2.067289113998413,
|
| 142 |
+
"learning_rate": 1.9876883405951378e-05,
|
| 143 |
+
"loss": 1.2058,
|
| 144 |
+
"step": 18
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.5833333333333335,
|
| 148 |
+
"grad_norm": 2.2645394802093506,
|
| 149 |
+
"learning_rate": 1.9862856015372315e-05,
|
| 150 |
+
"loss": 1.1692,
|
| 151 |
+
"step": 19
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.6666666666666665,
|
| 155 |
+
"grad_norm": 2.180142641067505,
|
| 156 |
+
"learning_rate": 1.9848077530122083e-05,
|
| 157 |
+
"loss": 1.1266,
|
| 158 |
+
"step": 20
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.75,
|
| 162 |
+
"grad_norm": 2.1327672004699707,
|
| 163 |
+
"learning_rate": 1.983254907563955e-05,
|
| 164 |
+
"loss": 1.1299,
|
| 165 |
+
"step": 21
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.8333333333333335,
|
| 169 |
+
"grad_norm": 2.3337109088897705,
|
| 170 |
+
"learning_rate": 1.9816271834476642e-05,
|
| 171 |
+
"loss": 1.167,
|
| 172 |
+
"step": 22
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.9166666666666665,
|
| 176 |
+
"grad_norm": 2.25063419342041,
|
| 177 |
+
"learning_rate": 1.9799247046208297e-05,
|
| 178 |
+
"loss": 1.0507,
|
| 179 |
+
"step": 23
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.0,
|
| 183 |
+
"grad_norm": 2.100271701812744,
|
| 184 |
+
"learning_rate": 1.9781476007338058e-05,
|
| 185 |
+
"loss": 1.094,
|
| 186 |
+
"step": 24
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.0,
|
| 190 |
+
"eval_f1": 0.5919379998124995,
|
| 191 |
+
"eval_loss": 0.9504638314247131,
|
| 192 |
+
"eval_runtime": 0.1142,
|
| 193 |
+
"eval_samples_per_second": 1908.934,
|
| 194 |
+
"eval_steps_per_second": 35.026,
|
| 195 |
+
"step": 24
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.0833333333333335,
|
| 199 |
+
"grad_norm": 2.3545162677764893,
|
| 200 |
+
"learning_rate": 1.9762960071199334e-05,
|
| 201 |
+
"loss": 0.9873,
|
| 202 |
+
"step": 25
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.1666666666666665,
|
| 206 |
+
"grad_norm": 1.8374146223068237,
|
| 207 |
+
"learning_rate": 1.9743700647852356e-05,
|
| 208 |
+
"loss": 1.0324,
|
| 209 |
+
"step": 26
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.25,
|
| 213 |
+
"grad_norm": 2.068077325820923,
|
| 214 |
+
"learning_rate": 1.9723699203976768e-05,
|
| 215 |
+
"loss": 1.0065,
|
| 216 |
+
"step": 27
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.3333333333333335,
|
| 220 |
+
"grad_norm": 2.6003549098968506,
|
| 221 |
+
"learning_rate": 1.9702957262759964e-05,
|
| 222 |
+
"loss": 1.0042,
|
| 223 |
+
"step": 28
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 2.4166666666666665,
|
| 227 |
+
"grad_norm": 2.3196046352386475,
|
| 228 |
+
"learning_rate": 1.968147640378108e-05,
|
| 229 |
+
"loss": 0.9483,
|
| 230 |
+
"step": 29
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.5,
|
| 234 |
+
"grad_norm": 1.9846868515014648,
|
| 235 |
+
"learning_rate": 1.9659258262890683e-05,
|
| 236 |
+
"loss": 0.9306,
|
| 237 |
+
"step": 30
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.5833333333333335,
|
| 241 |
+
"grad_norm": 2.3393170833587646,
|
| 242 |
+
"learning_rate": 1.963630453208623e-05,
|
| 243 |
+
"loss": 0.8026,
|
| 244 |
+
"step": 31
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.6666666666666665,
|
| 248 |
+
"grad_norm": 1.9397917985916138,
|
| 249 |
+
"learning_rate": 1.961261695938319e-05,
|
| 250 |
+
"loss": 0.8817,
|
| 251 |
+
"step": 32
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.75,
|
| 255 |
+
"grad_norm": 1.827669382095337,
|
| 256 |
+
"learning_rate": 1.958819734868193e-05,
|
| 257 |
+
"loss": 0.8372,
|
| 258 |
+
"step": 33
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.8333333333333335,
|
| 262 |
+
"grad_norm": 2.262103796005249,
|
| 263 |
+
"learning_rate": 1.9563047559630356e-05,
|
| 264 |
+
"loss": 0.8173,
|
| 265 |
+
"step": 34
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.9166666666666665,
|
| 269 |
+
"grad_norm": 1.634775996208191,
|
| 270 |
+
"learning_rate": 1.953716950748227e-05,
|
| 271 |
+
"loss": 0.786,
|
| 272 |
+
"step": 35
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 3.0,
|
| 276 |
+
"grad_norm": 2.8899052143096924,
|
| 277 |
+
"learning_rate": 1.9510565162951538e-05,
|
| 278 |
+
"loss": 0.916,
|
| 279 |
+
"step": 36
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 3.0,
|
| 283 |
+
"eval_f1": 0.6702283745087566,
|
| 284 |
+
"eval_loss": 0.6923440098762512,
|
| 285 |
+
"eval_runtime": 0.1108,
|
| 286 |
+
"eval_samples_per_second": 1966.836,
|
| 287 |
+
"eval_steps_per_second": 36.089,
|
| 288 |
+
"step": 36
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.0833333333333335,
|
| 292 |
+
"grad_norm": 2.1823081970214844,
|
| 293 |
+
"learning_rate": 1.9483236552061996e-05,
|
| 294 |
+
"loss": 0.6862,
|
| 295 |
+
"step": 37
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 3.1666666666666665,
|
| 299 |
+
"grad_norm": 1.7626707553863525,
|
| 300 |
+
"learning_rate": 1.945518575599317e-05,
|
| 301 |
+
"loss": 0.7917,
|
| 302 |
+
"step": 38
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.25,
|
| 306 |
+
"grad_norm": 1.655781865119934,
|
| 307 |
+
"learning_rate": 1.9426414910921785e-05,
|
| 308 |
+
"loss": 0.7134,
|
| 309 |
+
"step": 39
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 3.3333333333333335,
|
| 313 |
+
"grad_norm": 3.0367038249969482,
|
| 314 |
+
"learning_rate": 1.9396926207859085e-05,
|
| 315 |
+
"loss": 0.7741,
|
| 316 |
+
"step": 40
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.4166666666666665,
|
| 320 |
+
"grad_norm": 1.7714229822158813,
|
| 321 |
+
"learning_rate": 1.9366721892483976e-05,
|
| 322 |
+
"loss": 0.7582,
|
| 323 |
+
"step": 41
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.5,
|
| 327 |
+
"grad_norm": 2.1124234199523926,
|
| 328 |
+
"learning_rate": 1.9335804264972018e-05,
|
| 329 |
+
"loss": 0.6961,
|
| 330 |
+
"step": 42
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.5833333333333335,
|
| 334 |
+
"grad_norm": 2.0323431491851807,
|
| 335 |
+
"learning_rate": 1.9304175679820247e-05,
|
| 336 |
+
"loss": 0.6512,
|
| 337 |
+
"step": 43
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.6666666666666665,
|
| 341 |
+
"grad_norm": 3.1592702865600586,
|
| 342 |
+
"learning_rate": 1.9271838545667876e-05,
|
| 343 |
+
"loss": 0.6052,
|
| 344 |
+
"step": 44
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.75,
|
| 348 |
+
"grad_norm": 2.1813063621520996,
|
| 349 |
+
"learning_rate": 1.9238795325112867e-05,
|
| 350 |
+
"loss": 0.7806,
|
| 351 |
+
"step": 45
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.8333333333333335,
|
| 355 |
+
"grad_norm": 2.1008193492889404,
|
| 356 |
+
"learning_rate": 1.9205048534524405e-05,
|
| 357 |
+
"loss": 0.7297,
|
| 358 |
+
"step": 46
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.9166666666666665,
|
| 362 |
+
"grad_norm": 1.7093874216079712,
|
| 363 |
+
"learning_rate": 1.917060074385124e-05,
|
| 364 |
+
"loss": 0.5639,
|
| 365 |
+
"step": 47
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 4.0,
|
| 369 |
+
"grad_norm": 2.2645668983459473,
|
| 370 |
+
"learning_rate": 1.913545457642601e-05,
|
| 371 |
+
"loss": 0.6286,
|
| 372 |
+
"step": 48
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 4.0,
|
| 376 |
+
"eval_f1": 0.8153789924338085,
|
| 377 |
+
"eval_loss": 0.562224268913269,
|
| 378 |
+
"eval_runtime": 0.1137,
|
| 379 |
+
"eval_samples_per_second": 1916.853,
|
| 380 |
+
"eval_steps_per_second": 35.172,
|
| 381 |
+
"step": 48
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 4.083333333333333,
|
| 385 |
+
"grad_norm": 3.357717752456665,
|
| 386 |
+
"learning_rate": 1.9099612708765432e-05,
|
| 387 |
+
"loss": 0.5976,
|
| 388 |
+
"step": 49
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.166666666666667,
|
| 392 |
+
"grad_norm": 3.129816770553589,
|
| 393 |
+
"learning_rate": 1.9063077870366504e-05,
|
| 394 |
+
"loss": 0.6326,
|
| 395 |
+
"step": 50
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 4.25,
|
| 399 |
+
"grad_norm": 2.7401657104492188,
|
| 400 |
+
"learning_rate": 1.902585284349861e-05,
|
| 401 |
+
"loss": 0.5995,
|
| 402 |
+
"step": 51
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 4.333333333333333,
|
| 406 |
+
"grad_norm": 2.191913604736328,
|
| 407 |
+
"learning_rate": 1.8987940462991673e-05,
|
| 408 |
+
"loss": 0.5869,
|
| 409 |
+
"step": 52
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.416666666666667,
|
| 413 |
+
"grad_norm": 2.8935346603393555,
|
| 414 |
+
"learning_rate": 1.894934361602025e-05,
|
| 415 |
+
"loss": 0.6083,
|
| 416 |
+
"step": 53
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.5,
|
| 420 |
+
"grad_norm": 2.674243450164795,
|
| 421 |
+
"learning_rate": 1.891006524188368e-05,
|
| 422 |
+
"loss": 0.5927,
|
| 423 |
+
"step": 54
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 4.583333333333333,
|
| 427 |
+
"grad_norm": 2.0203378200531006,
|
| 428 |
+
"learning_rate": 1.887010833178222e-05,
|
| 429 |
+
"loss": 0.6112,
|
| 430 |
+
"step": 55
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.666666666666667,
|
| 434 |
+
"grad_norm": 1.9910380840301514,
|
| 435 |
+
"learning_rate": 1.8829475928589272e-05,
|
| 436 |
+
"loss": 0.4967,
|
| 437 |
+
"step": 56
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 4.75,
|
| 441 |
+
"grad_norm": 2.2031772136688232,
|
| 442 |
+
"learning_rate": 1.8788171126619653e-05,
|
| 443 |
+
"loss": 0.5578,
|
| 444 |
+
"step": 57
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 4.833333333333333,
|
| 448 |
+
"grad_norm": 3.2422826290130615,
|
| 449 |
+
"learning_rate": 1.874619707139396e-05,
|
| 450 |
+
"loss": 0.5584,
|
| 451 |
+
"step": 58
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 4.916666666666667,
|
| 455 |
+
"grad_norm": 1.6707191467285156,
|
| 456 |
+
"learning_rate": 1.8703556959398998e-05,
|
| 457 |
+
"loss": 0.4944,
|
| 458 |
+
"step": 59
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 5.0,
|
| 462 |
+
"grad_norm": 3.649735927581787,
|
| 463 |
+
"learning_rate": 1.866025403784439e-05,
|
| 464 |
+
"loss": 0.5872,
|
| 465 |
+
"step": 60
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 5.0,
|
| 469 |
+
"eval_f1": 0.8406014900132547,
|
| 470 |
+
"eval_loss": 0.5035473108291626,
|
| 471 |
+
"eval_runtime": 0.1115,
|
| 472 |
+
"eval_samples_per_second": 1955.32,
|
| 473 |
+
"eval_steps_per_second": 35.877,
|
| 474 |
+
"step": 60
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 5.083333333333333,
|
| 478 |
+
"grad_norm": 3.136577606201172,
|
| 479 |
+
"learning_rate": 1.861629160441526e-05,
|
| 480 |
+
"loss": 0.538,
|
| 481 |
+
"step": 61
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 5.166666666666667,
|
| 485 |
+
"grad_norm": 1.9009243249893188,
|
| 486 |
+
"learning_rate": 1.8571673007021124e-05,
|
| 487 |
+
"loss": 0.5754,
|
| 488 |
+
"step": 62
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 5.25,
|
| 492 |
+
"grad_norm": 1.6683857440948486,
|
| 493 |
+
"learning_rate": 1.8526401643540924e-05,
|
| 494 |
+
"loss": 0.5381,
|
| 495 |
+
"step": 63
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 5.333333333333333,
|
| 499 |
+
"grad_norm": 1.3652187585830688,
|
| 500 |
+
"learning_rate": 1.848048096156426e-05,
|
| 501 |
+
"loss": 0.3973,
|
| 502 |
+
"step": 64
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 5.416666666666667,
|
| 506 |
+
"grad_norm": 3.1748218536376953,
|
| 507 |
+
"learning_rate": 1.843391445812886e-05,
|
| 508 |
+
"loss": 0.4358,
|
| 509 |
+
"step": 65
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 5.5,
|
| 513 |
+
"grad_norm": 1.84850013256073,
|
| 514 |
+
"learning_rate": 1.8386705679454243e-05,
|
| 515 |
+
"loss": 0.4504,
|
| 516 |
+
"step": 66
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 5.583333333333333,
|
| 520 |
+
"grad_norm": 1.8327257633209229,
|
| 521 |
+
"learning_rate": 1.8338858220671683e-05,
|
| 522 |
+
"loss": 0.4763,
|
| 523 |
+
"step": 67
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 5.666666666666667,
|
| 527 |
+
"grad_norm": 2.9052014350891113,
|
| 528 |
+
"learning_rate": 1.8290375725550417e-05,
|
| 529 |
+
"loss": 0.4949,
|
| 530 |
+
"step": 68
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 5.75,
|
| 534 |
+
"grad_norm": 3.871500015258789,
|
| 535 |
+
"learning_rate": 1.8241261886220155e-05,
|
| 536 |
+
"loss": 0.5134,
|
| 537 |
+
"step": 69
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 5.833333333333333,
|
| 541 |
+
"grad_norm": 1.9742496013641357,
|
| 542 |
+
"learning_rate": 1.819152044288992e-05,
|
| 543 |
+
"loss": 0.446,
|
| 544 |
+
"step": 70
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 5.916666666666667,
|
| 548 |
+
"grad_norm": 2.0321502685546875,
|
| 549 |
+
"learning_rate": 1.8141155183563195e-05,
|
| 550 |
+
"loss": 0.4747,
|
| 551 |
+
"step": 71
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 6.0,
|
| 555 |
+
"grad_norm": 2.1369895935058594,
|
| 556 |
+
"learning_rate": 1.8090169943749477e-05,
|
| 557 |
+
"loss": 0.5148,
|
| 558 |
+
"step": 72
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 6.0,
|
| 562 |
+
"eval_f1": 0.8577743286523773,
|
| 563 |
+
"eval_loss": 0.4574647545814514,
|
| 564 |
+
"eval_runtime": 0.115,
|
| 565 |
+
"eval_samples_per_second": 1895.546,
|
| 566 |
+
"eval_steps_per_second": 34.781,
|
| 567 |
+
"step": 72
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 6.083333333333333,
|
| 571 |
+
"grad_norm": 1.95284903049469,
|
| 572 |
+
"learning_rate": 1.8038568606172172e-05,
|
| 573 |
+
"loss": 0.4302,
|
| 574 |
+
"step": 73
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 6.166666666666667,
|
| 578 |
+
"grad_norm": 2.4043610095977783,
|
| 579 |
+
"learning_rate": 1.798635510047293e-05,
|
| 580 |
+
"loss": 0.5319,
|
| 581 |
+
"step": 74
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 6.25,
|
| 585 |
+
"grad_norm": 2.9988913536071777,
|
| 586 |
+
"learning_rate": 1.7933533402912354e-05,
|
| 587 |
+
"loss": 0.3815,
|
| 588 |
+
"step": 75
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 6.333333333333333,
|
| 592 |
+
"grad_norm": 1.631911277770996,
|
| 593 |
+
"learning_rate": 1.788010753606722e-05,
|
| 594 |
+
"loss": 0.3405,
|
| 595 |
+
"step": 76
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 6.416666666666667,
|
| 599 |
+
"grad_norm": 2.1044607162475586,
|
| 600 |
+
"learning_rate": 1.782608156852414e-05,
|
| 601 |
+
"loss": 0.3256,
|
| 602 |
+
"step": 77
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 6.5,
|
| 606 |
+
"grad_norm": 2.995670795440674,
|
| 607 |
+
"learning_rate": 1.777145961456971e-05,
|
| 608 |
+
"loss": 0.4149,
|
| 609 |
+
"step": 78
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 6.583333333333333,
|
| 613 |
+
"grad_norm": 2.1682353019714355,
|
| 614 |
+
"learning_rate": 1.7716245833877202e-05,
|
| 615 |
+
"loss": 0.3488,
|
| 616 |
+
"step": 79
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 6.666666666666667,
|
| 620 |
+
"grad_norm": 2.1982967853546143,
|
| 621 |
+
"learning_rate": 1.766044443118978e-05,
|
| 622 |
+
"loss": 0.4167,
|
| 623 |
+
"step": 80
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 6.75,
|
| 627 |
+
"grad_norm": 1.896740198135376,
|
| 628 |
+
"learning_rate": 1.7604059656000313e-05,
|
| 629 |
+
"loss": 0.4102,
|
| 630 |
+
"step": 81
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 6.833333333333333,
|
| 634 |
+
"grad_norm": 2.5532283782958984,
|
| 635 |
+
"learning_rate": 1.7547095802227723e-05,
|
| 636 |
+
"loss": 0.453,
|
| 637 |
+
"step": 82
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 6.916666666666667,
|
| 641 |
+
"grad_norm": 2.8607945442199707,
|
| 642 |
+
"learning_rate": 1.7489557207890025e-05,
|
| 643 |
+
"loss": 0.439,
|
| 644 |
+
"step": 83
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 7.0,
|
| 648 |
+
"grad_norm": 3.566798210144043,
|
| 649 |
+
"learning_rate": 1.7431448254773943e-05,
|
| 650 |
+
"loss": 0.4552,
|
| 651 |
+
"step": 84
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 7.0,
|
| 655 |
+
"eval_f1": 0.8548547310106056,
|
| 656 |
+
"eval_loss": 0.4446788430213928,
|
| 657 |
+
"eval_runtime": 0.1123,
|
| 658 |
+
"eval_samples_per_second": 1940.427,
|
| 659 |
+
"eval_steps_per_second": 35.604,
|
| 660 |
+
"step": 84
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 7.083333333333333,
|
| 664 |
+
"grad_norm": 1.913853645324707,
|
| 665 |
+
"learning_rate": 1.737277336810124e-05,
|
| 666 |
+
"loss": 0.3472,
|
| 667 |
+
"step": 85
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 7.166666666666667,
|
| 671 |
+
"grad_norm": 2.719738721847534,
|
| 672 |
+
"learning_rate": 1.7313537016191706e-05,
|
| 673 |
+
"loss": 0.3941,
|
| 674 |
+
"step": 86
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 7.25,
|
| 678 |
+
"grad_norm": 2.5046474933624268,
|
| 679 |
+
"learning_rate": 1.7253743710122877e-05,
|
| 680 |
+
"loss": 0.3395,
|
| 681 |
+
"step": 87
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 7.333333333333333,
|
| 685 |
+
"grad_norm": 2.638500928878784,
|
| 686 |
+
"learning_rate": 1.7193398003386514e-05,
|
| 687 |
+
"loss": 0.3731,
|
| 688 |
+
"step": 88
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 7.416666666666667,
|
| 692 |
+
"grad_norm": 2.0507426261901855,
|
| 693 |
+
"learning_rate": 1.713250449154182e-05,
|
| 694 |
+
"loss": 0.4042,
|
| 695 |
+
"step": 89
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 7.5,
|
| 699 |
+
"grad_norm": 1.704081654548645,
|
| 700 |
+
"learning_rate": 1.7071067811865477e-05,
|
| 701 |
+
"loss": 0.3434,
|
| 702 |
+
"step": 90
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 7.583333333333333,
|
| 706 |
+
"grad_norm": 2.994464874267578,
|
| 707 |
+
"learning_rate": 1.700909264299851e-05,
|
| 708 |
+
"loss": 0.3913,
|
| 709 |
+
"step": 91
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 7.666666666666667,
|
| 713 |
+
"grad_norm": 2.510322093963623,
|
| 714 |
+
"learning_rate": 1.6946583704589973e-05,
|
| 715 |
+
"loss": 0.369,
|
| 716 |
+
"step": 92
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 7.75,
|
| 720 |
+
"grad_norm": 2.235898017883301,
|
| 721 |
+
"learning_rate": 1.688354575693754e-05,
|
| 722 |
+
"loss": 0.3372,
|
| 723 |
+
"step": 93
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 7.833333333333333,
|
| 727 |
+
"grad_norm": 2.09539532661438,
|
| 728 |
+
"learning_rate": 1.6819983600624986e-05,
|
| 729 |
+
"loss": 0.3133,
|
| 730 |
+
"step": 94
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 7.916666666666667,
|
| 734 |
+
"grad_norm": 3.6932387351989746,
|
| 735 |
+
"learning_rate": 1.6755902076156606e-05,
|
| 736 |
+
"loss": 0.3728,
|
| 737 |
+
"step": 95
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 8.0,
|
| 741 |
+
"grad_norm": 2.308973789215088,
|
| 742 |
+
"learning_rate": 1.6691306063588583e-05,
|
| 743 |
+
"loss": 0.3223,
|
| 744 |
+
"step": 96
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 8.0,
|
| 748 |
+
"eval_f1": 0.8471376397229375,
|
| 749 |
+
"eval_loss": 0.4268903136253357,
|
| 750 |
+
"eval_runtime": 0.1106,
|
| 751 |
+
"eval_samples_per_second": 1971.449,
|
| 752 |
+
"eval_steps_per_second": 36.173,
|
| 753 |
+
"step": 96
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 8.083333333333334,
|
| 757 |
+
"grad_norm": 1.779574990272522,
|
| 758 |
+
"learning_rate": 1.6626200482157378e-05,
|
| 759 |
+
"loss": 0.3047,
|
| 760 |
+
"step": 97
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 8.166666666666666,
|
| 764 |
+
"grad_norm": 2.538353681564331,
|
| 765 |
+
"learning_rate": 1.6560590289905074e-05,
|
| 766 |
+
"loss": 0.2682,
|
| 767 |
+
"step": 98
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 8.25,
|
| 771 |
+
"grad_norm": 2.067002058029175,
|
| 772 |
+
"learning_rate": 1.6494480483301836e-05,
|
| 773 |
+
"loss": 0.3435,
|
| 774 |
+
"step": 99
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 8.333333333333334,
|
| 778 |
+
"grad_norm": 2.4853057861328125,
|
| 779 |
+
"learning_rate": 1.6427876096865394e-05,
|
| 780 |
+
"loss": 0.3525,
|
| 781 |
+
"step": 100
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 8.416666666666666,
|
| 785 |
+
"grad_norm": 1.9690351486206055,
|
| 786 |
+
"learning_rate": 1.636078220277764e-05,
|
| 787 |
+
"loss": 0.3033,
|
| 788 |
+
"step": 101
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 8.5,
|
| 792 |
+
"grad_norm": 2.2814316749572754,
|
| 793 |
+
"learning_rate": 1.6293203910498375e-05,
|
| 794 |
+
"loss": 0.2569,
|
| 795 |
+
"step": 102
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 8.583333333333334,
|
| 799 |
+
"grad_norm": 1.6664941310882568,
|
| 800 |
+
"learning_rate": 1.6225146366376198e-05,
|
| 801 |
+
"loss": 0.1951,
|
| 802 |
+
"step": 103
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 8.666666666666666,
|
| 806 |
+
"grad_norm": 1.9773038625717163,
|
| 807 |
+
"learning_rate": 1.6156614753256583e-05,
|
| 808 |
+
"loss": 0.3379,
|
| 809 |
+
"step": 104
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 8.75,
|
| 813 |
+
"grad_norm": 2.8037147521972656,
|
| 814 |
+
"learning_rate": 1.608761429008721e-05,
|
| 815 |
+
"loss": 0.3027,
|
| 816 |
+
"step": 105
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 8.833333333333334,
|
| 820 |
+
"grad_norm": 2.945286512374878,
|
| 821 |
+
"learning_rate": 1.6018150231520486e-05,
|
| 822 |
+
"loss": 0.3796,
|
| 823 |
+
"step": 106
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 8.916666666666666,
|
| 827 |
+
"grad_norm": 2.62178373336792,
|
| 828 |
+
"learning_rate": 1.5948227867513416e-05,
|
| 829 |
+
"loss": 0.3388,
|
| 830 |
+
"step": 107
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 9.0,
|
| 834 |
+
"grad_norm": 3.8131866455078125,
|
| 835 |
+
"learning_rate": 1.5877852522924733e-05,
|
| 836 |
+
"loss": 0.2701,
|
| 837 |
+
"step": 108
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 9.0,
|
| 841 |
+
"eval_f1": 0.8585253422202574,
|
| 842 |
+
"eval_loss": 0.42710253596305847,
|
| 843 |
+
"eval_runtime": 0.1116,
|
| 844 |
+
"eval_samples_per_second": 1953.941,
|
| 845 |
+
"eval_steps_per_second": 35.852,
|
| 846 |
+
"step": 108
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 9.083333333333334,
|
| 850 |
+
"grad_norm": 1.8985729217529297,
|
| 851 |
+
"learning_rate": 1.5807029557109398e-05,
|
| 852 |
+
"loss": 0.2686,
|
| 853 |
+
"step": 109
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 9.166666666666666,
|
| 857 |
+
"grad_norm": 1.9291400909423828,
|
| 858 |
+
"learning_rate": 1.573576436351046e-05,
|
| 859 |
+
"loss": 0.3093,
|
| 860 |
+
"step": 110
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 9.25,
|
| 864 |
+
"grad_norm": 2.466015338897705,
|
| 865 |
+
"learning_rate": 1.566406236924833e-05,
|
| 866 |
+
"loss": 0.2529,
|
| 867 |
+
"step": 111
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 9.333333333333334,
|
| 871 |
+
"grad_norm": 1.6614748239517212,
|
| 872 |
+
"learning_rate": 1.5591929034707468e-05,
|
| 873 |
+
"loss": 0.2563,
|
| 874 |
+
"step": 112
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 9.416666666666666,
|
| 878 |
+
"grad_norm": 1.927484393119812,
|
| 879 |
+
"learning_rate": 1.5519369853120584e-05,
|
| 880 |
+
"loss": 0.2728,
|
| 881 |
+
"step": 113
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 9.5,
|
| 885 |
+
"grad_norm": 4.029143333435059,
|
| 886 |
+
"learning_rate": 1.5446390350150272e-05,
|
| 887 |
+
"loss": 0.3043,
|
| 888 |
+
"step": 114
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 9.583333333333334,
|
| 892 |
+
"grad_norm": 3.3003406524658203,
|
| 893 |
+
"learning_rate": 1.5372996083468242e-05,
|
| 894 |
+
"loss": 0.3676,
|
| 895 |
+
"step": 115
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 9.666666666666666,
|
| 899 |
+
"grad_norm": 2.3668899536132812,
|
| 900 |
+
"learning_rate": 1.529919264233205e-05,
|
| 901 |
+
"loss": 0.1894,
|
| 902 |
+
"step": 116
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 9.75,
|
| 906 |
+
"grad_norm": 4.198014259338379,
|
| 907 |
+
"learning_rate": 1.5224985647159489e-05,
|
| 908 |
+
"loss": 0.3147,
|
| 909 |
+
"step": 117
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 9.833333333333334,
|
| 913 |
+
"grad_norm": 3.053900718688965,
|
| 914 |
+
"learning_rate": 1.5150380749100545e-05,
|
| 915 |
+
"loss": 0.3324,
|
| 916 |
+
"step": 118
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 9.916666666666666,
|
| 920 |
+
"grad_norm": 2.420565366744995,
|
| 921 |
+
"learning_rate": 1.5075383629607043e-05,
|
| 922 |
+
"loss": 0.2264,
|
| 923 |
+
"step": 119
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 10.0,
|
| 927 |
+
"grad_norm": 2.6047873497009277,
|
| 928 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 929 |
+
"loss": 0.2349,
|
| 930 |
+
"step": 120
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 10.0,
|
| 934 |
+
"eval_f1": 0.8583761667305971,
|
| 935 |
+
"eval_loss": 0.4111815392971039,
|
| 936 |
+
"eval_runtime": 0.1123,
|
| 937 |
+
"eval_samples_per_second": 1940.485,
|
| 938 |
+
"eval_steps_per_second": 35.605,
|
| 939 |
+
"step": 120
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 10.083333333333334,
|
| 943 |
+
"grad_norm": 1.6441205739974976,
|
| 944 |
+
"learning_rate": 1.4924235601034673e-05,
|
| 945 |
+
"loss": 0.2693,
|
| 946 |
+
"step": 121
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 10.166666666666666,
|
| 950 |
+
"grad_norm": 2.5293445587158203,
|
| 951 |
+
"learning_rate": 1.4848096202463373e-05,
|
| 952 |
+
"loss": 0.2733,
|
| 953 |
+
"step": 122
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 10.25,
|
| 957 |
+
"grad_norm": 1.9516547918319702,
|
| 958 |
+
"learning_rate": 1.4771587602596085e-05,
|
| 959 |
+
"loss": 0.2199,
|
| 960 |
+
"step": 123
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 10.333333333333334,
|
| 964 |
+
"grad_norm": 3.799968719482422,
|
| 965 |
+
"learning_rate": 1.469471562785891e-05,
|
| 966 |
+
"loss": 0.261,
|
| 967 |
+
"step": 124
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 10.416666666666666,
|
| 971 |
+
"grad_norm": 2.764522075653076,
|
| 972 |
+
"learning_rate": 1.4617486132350343e-05,
|
| 973 |
+
"loss": 0.2217,
|
| 974 |
+
"step": 125
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 10.5,
|
| 978 |
+
"grad_norm": 4.601480484008789,
|
| 979 |
+
"learning_rate": 1.4539904997395468e-05,
|
| 980 |
+
"loss": 0.216,
|
| 981 |
+
"step": 126
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 10.583333333333334,
|
| 985 |
+
"grad_norm": 2.331076145172119,
|
| 986 |
+
"learning_rate": 1.4461978131098089e-05,
|
| 987 |
+
"loss": 0.2068,
|
| 988 |
+
"step": 127
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 10.666666666666666,
|
| 992 |
+
"grad_norm": 1.9390913248062134,
|
| 993 |
+
"learning_rate": 1.4383711467890776e-05,
|
| 994 |
+
"loss": 0.2177,
|
| 995 |
+
"step": 128
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 10.75,
|
| 999 |
+
"grad_norm": 2.847414493560791,
|
| 1000 |
+
"learning_rate": 1.4305110968082953e-05,
|
| 1001 |
+
"loss": 0.2672,
|
| 1002 |
+
"step": 129
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 10.833333333333334,
|
| 1006 |
+
"grad_norm": 1.6508653163909912,
|
| 1007 |
+
"learning_rate": 1.4226182617406996e-05,
|
| 1008 |
+
"loss": 0.166,
|
| 1009 |
+
"step": 130
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 10.916666666666666,
|
| 1013 |
+
"grad_norm": 2.4722893238067627,
|
| 1014 |
+
"learning_rate": 1.4146932426562391e-05,
|
| 1015 |
+
"loss": 0.2215,
|
| 1016 |
+
"step": 131
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 11.0,
|
| 1020 |
+
"grad_norm": 2.3666489124298096,
|
| 1021 |
+
"learning_rate": 1.4067366430758004e-05,
|
| 1022 |
+
"loss": 0.247,
|
| 1023 |
+
"step": 132
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 11.0,
|
| 1027 |
+
"eval_f1": 0.8425511801798409,
|
| 1028 |
+
"eval_loss": 0.42715930938720703,
|
| 1029 |
+
"eval_runtime": 0.1123,
|
| 1030 |
+
"eval_samples_per_second": 1940.369,
|
| 1031 |
+
"eval_steps_per_second": 35.603,
|
| 1032 |
+
"step": 132
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 11.083333333333334,
|
| 1036 |
+
"grad_norm": 5.439218521118164,
|
| 1037 |
+
"learning_rate": 1.3987490689252463e-05,
|
| 1038 |
+
"loss": 0.3612,
|
| 1039 |
+
"step": 133
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 11.166666666666666,
|
| 1043 |
+
"grad_norm": 1.708125114440918,
|
| 1044 |
+
"learning_rate": 1.3907311284892737e-05,
|
| 1045 |
+
"loss": 0.1717,
|
| 1046 |
+
"step": 134
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 11.25,
|
| 1050 |
+
"grad_norm": 2.919048309326172,
|
| 1051 |
+
"learning_rate": 1.3826834323650899e-05,
|
| 1052 |
+
"loss": 0.214,
|
| 1053 |
+
"step": 135
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 11.333333333333334,
|
| 1057 |
+
"grad_norm": 2.025909185409546,
|
| 1058 |
+
"learning_rate": 1.3746065934159123e-05,
|
| 1059 |
+
"loss": 0.1894,
|
| 1060 |
+
"step": 136
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 11.416666666666666,
|
| 1064 |
+
"grad_norm": 2.9236347675323486,
|
| 1065 |
+
"learning_rate": 1.3665012267242974e-05,
|
| 1066 |
+
"loss": 0.1728,
|
| 1067 |
+
"step": 137
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 11.5,
|
| 1071 |
+
"grad_norm": 2.8298561573028564,
|
| 1072 |
+
"learning_rate": 1.3583679495453e-05,
|
| 1073 |
+
"loss": 0.2073,
|
| 1074 |
+
"step": 138
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 11.583333333333334,
|
| 1078 |
+
"grad_norm": 1.7483054399490356,
|
| 1079 |
+
"learning_rate": 1.3502073812594677e-05,
|
| 1080 |
+
"loss": 0.1899,
|
| 1081 |
+
"step": 139
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 11.666666666666666,
|
| 1085 |
+
"grad_norm": 2.066528558731079,
|
| 1086 |
+
"learning_rate": 1.342020143325669e-05,
|
| 1087 |
+
"loss": 0.2361,
|
| 1088 |
+
"step": 140
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 11.75,
|
| 1092 |
+
"grad_norm": 2.8144662380218506,
|
| 1093 |
+
"learning_rate": 1.333806859233771e-05,
|
| 1094 |
+
"loss": 0.158,
|
| 1095 |
+
"step": 141
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 11.833333333333334,
|
| 1099 |
+
"grad_norm": 2.6354258060455322,
|
| 1100 |
+
"learning_rate": 1.3255681544571568e-05,
|
| 1101 |
+
"loss": 0.2086,
|
| 1102 |
+
"step": 142
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 11.916666666666666,
|
| 1106 |
+
"grad_norm": 2.479193687438965,
|
| 1107 |
+
"learning_rate": 1.3173046564050923e-05,
|
| 1108 |
+
"loss": 0.1669,
|
| 1109 |
+
"step": 143
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 12.0,
|
| 1113 |
+
"grad_norm": 3.766613006591797,
|
| 1114 |
+
"learning_rate": 1.3090169943749475e-05,
|
| 1115 |
+
"loss": 0.2962,
|
| 1116 |
+
"step": 144
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 12.0,
|
| 1120 |
+
"eval_f1": 0.8664029304029304,
|
| 1121 |
+
"eval_loss": 0.41419580578804016,
|
| 1122 |
+
"eval_runtime": 0.111,
|
| 1123 |
+
"eval_samples_per_second": 1964.798,
|
| 1124 |
+
"eval_steps_per_second": 36.051,
|
| 1125 |
+
"step": 144
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 12.083333333333334,
|
| 1129 |
+
"grad_norm": 1.6231375932693481,
|
| 1130 |
+
"learning_rate": 1.300705799504273e-05,
|
| 1131 |
+
"loss": 0.1251,
|
| 1132 |
+
"step": 145
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 12.166666666666666,
|
| 1136 |
+
"grad_norm": 2.5651540756225586,
|
| 1137 |
+
"learning_rate": 1.2923717047227368e-05,
|
| 1138 |
+
"loss": 0.1349,
|
| 1139 |
+
"step": 146
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 12.25,
|
| 1143 |
+
"grad_norm": 1.8243824243545532,
|
| 1144 |
+
"learning_rate": 1.284015344703923e-05,
|
| 1145 |
+
"loss": 0.242,
|
| 1146 |
+
"step": 147
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 12.333333333333334,
|
| 1150 |
+
"grad_norm": 2.2925655841827393,
|
| 1151 |
+
"learning_rate": 1.2756373558169992e-05,
|
| 1152 |
+
"loss": 0.1762,
|
| 1153 |
+
"step": 148
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 12.416666666666666,
|
| 1157 |
+
"grad_norm": 1.430468201637268,
|
| 1158 |
+
"learning_rate": 1.267238376078257e-05,
|
| 1159 |
+
"loss": 0.1472,
|
| 1160 |
+
"step": 149
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 12.5,
|
| 1164 |
+
"grad_norm": 2.6467137336730957,
|
| 1165 |
+
"learning_rate": 1.2588190451025209e-05,
|
| 1166 |
+
"loss": 0.2575,
|
| 1167 |
+
"step": 150
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 12.583333333333334,
|
| 1171 |
+
"grad_norm": 1.9027683734893799,
|
| 1172 |
+
"learning_rate": 1.2503800040544417e-05,
|
| 1173 |
+
"loss": 0.1942,
|
| 1174 |
+
"step": 151
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 12.666666666666666,
|
| 1178 |
+
"grad_norm": 2.648810625076294,
|
| 1179 |
+
"learning_rate": 1.2419218955996677e-05,
|
| 1180 |
+
"loss": 0.1643,
|
| 1181 |
+
"step": 152
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 12.75,
|
| 1185 |
+
"grad_norm": 1.919045329093933,
|
| 1186 |
+
"learning_rate": 1.2334453638559057e-05,
|
| 1187 |
+
"loss": 0.1527,
|
| 1188 |
+
"step": 153
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 12.833333333333334,
|
| 1192 |
+
"grad_norm": 1.9980669021606445,
|
| 1193 |
+
"learning_rate": 1.2249510543438652e-05,
|
| 1194 |
+
"loss": 0.1707,
|
| 1195 |
+
"step": 154
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 12.916666666666666,
|
| 1199 |
+
"grad_norm": 1.593509316444397,
|
| 1200 |
+
"learning_rate": 1.2164396139381029e-05,
|
| 1201 |
+
"loss": 0.1395,
|
| 1202 |
+
"step": 155
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 13.0,
|
| 1206 |
+
"grad_norm": 2.3133833408355713,
|
| 1207 |
+
"learning_rate": 1.2079116908177592e-05,
|
| 1208 |
+
"loss": 0.1445,
|
| 1209 |
+
"step": 156
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 13.0,
|
| 1213 |
+
"eval_f1": 0.8678026234826873,
|
| 1214 |
+
"eval_loss": 0.4235089123249054,
|
| 1215 |
+
"eval_runtime": 0.1126,
|
| 1216 |
+
"eval_samples_per_second": 1936.416,
|
| 1217 |
+
"eval_steps_per_second": 35.531,
|
| 1218 |
+
"step": 156
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 13.083333333333334,
|
| 1222 |
+
"grad_norm": 2.027728796005249,
|
| 1223 |
+
"learning_rate": 1.1993679344171973e-05,
|
| 1224 |
+
"loss": 0.2044,
|
| 1225 |
+
"step": 157
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 13.166666666666666,
|
| 1229 |
+
"grad_norm": 2.105649471282959,
|
| 1230 |
+
"learning_rate": 1.190808995376545e-05,
|
| 1231 |
+
"loss": 0.1672,
|
| 1232 |
+
"step": 158
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 13.25,
|
| 1236 |
+
"grad_norm": 2.1034839153289795,
|
| 1237 |
+
"learning_rate": 1.1822355254921478e-05,
|
| 1238 |
+
"loss": 0.1588,
|
| 1239 |
+
"step": 159
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 13.333333333333334,
|
| 1243 |
+
"grad_norm": 2.3096885681152344,
|
| 1244 |
+
"learning_rate": 1.1736481776669307e-05,
|
| 1245 |
+
"loss": 0.1874,
|
| 1246 |
+
"step": 160
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 13.416666666666666,
|
| 1250 |
+
"grad_norm": 2.7364308834075928,
|
| 1251 |
+
"learning_rate": 1.1650476058606776e-05,
|
| 1252 |
+
"loss": 0.2179,
|
| 1253 |
+
"step": 161
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 13.5,
|
| 1257 |
+
"grad_norm": 1.7314239740371704,
|
| 1258 |
+
"learning_rate": 1.156434465040231e-05,
|
| 1259 |
+
"loss": 0.1008,
|
| 1260 |
+
"step": 162
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 13.583333333333334,
|
| 1264 |
+
"grad_norm": 2.2747116088867188,
|
| 1265 |
+
"learning_rate": 1.1478094111296109e-05,
|
| 1266 |
+
"loss": 0.1548,
|
| 1267 |
+
"step": 163
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 13.666666666666666,
|
| 1271 |
+
"grad_norm": 1.9884597063064575,
|
| 1272 |
+
"learning_rate": 1.1391731009600655e-05,
|
| 1273 |
+
"loss": 0.1127,
|
| 1274 |
+
"step": 164
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 13.75,
|
| 1278 |
+
"grad_norm": 1.939298152923584,
|
| 1279 |
+
"learning_rate": 1.130526192220052e-05,
|
| 1280 |
+
"loss": 0.2087,
|
| 1281 |
+
"step": 165
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 13.833333333333334,
|
| 1285 |
+
"grad_norm": 2.0005686283111572,
|
| 1286 |
+
"learning_rate": 1.1218693434051475e-05,
|
| 1287 |
+
"loss": 0.1235,
|
| 1288 |
+
"step": 166
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 13.916666666666666,
|
| 1292 |
+
"grad_norm": 4.452102184295654,
|
| 1293 |
+
"learning_rate": 1.113203213767907e-05,
|
| 1294 |
+
"loss": 0.2373,
|
| 1295 |
+
"step": 167
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 14.0,
|
| 1299 |
+
"grad_norm": 3.5608956813812256,
|
| 1300 |
+
"learning_rate": 1.1045284632676535e-05,
|
| 1301 |
+
"loss": 0.1245,
|
| 1302 |
+
"step": 168
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 14.0,
|
| 1306 |
+
"eval_f1": 0.8734395170821287,
|
| 1307 |
+
"eval_loss": 0.4194470942020416,
|
| 1308 |
+
"eval_runtime": 0.1134,
|
| 1309 |
+
"eval_samples_per_second": 1923.066,
|
| 1310 |
+
"eval_steps_per_second": 35.286,
|
| 1311 |
+
"step": 168
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 14.083333333333334,
|
| 1315 |
+
"grad_norm": 1.8794602155685425,
|
| 1316 |
+
"learning_rate": 1.0958457525202241e-05,
|
| 1317 |
+
"loss": 0.1591,
|
| 1318 |
+
"step": 169
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 14.166666666666666,
|
| 1322 |
+
"grad_norm": 2.176715850830078,
|
| 1323 |
+
"learning_rate": 1.0871557427476585e-05,
|
| 1324 |
+
"loss": 0.1785,
|
| 1325 |
+
"step": 170
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 14.25,
|
| 1329 |
+
"grad_norm": 2.284532070159912,
|
| 1330 |
+
"learning_rate": 1.0784590957278452e-05,
|
| 1331 |
+
"loss": 0.1399,
|
| 1332 |
+
"step": 171
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 14.333333333333334,
|
| 1336 |
+
"grad_norm": 1.7298234701156616,
|
| 1337 |
+
"learning_rate": 1.0697564737441254e-05,
|
| 1338 |
+
"loss": 0.1687,
|
| 1339 |
+
"step": 172
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 14.416666666666666,
|
| 1343 |
+
"grad_norm": 2.2304232120513916,
|
| 1344 |
+
"learning_rate": 1.0610485395348571e-05,
|
| 1345 |
+
"loss": 0.1129,
|
| 1346 |
+
"step": 173
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 14.5,
|
| 1350 |
+
"grad_norm": 2.125760793685913,
|
| 1351 |
+
"learning_rate": 1.0523359562429441e-05,
|
| 1352 |
+
"loss": 0.146,
|
| 1353 |
+
"step": 174
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 14.583333333333334,
|
| 1357 |
+
"grad_norm": 2.1295244693756104,
|
| 1358 |
+
"learning_rate": 1.0436193873653362e-05,
|
| 1359 |
+
"loss": 0.1257,
|
| 1360 |
+
"step": 175
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 14.666666666666666,
|
| 1364 |
+
"grad_norm": 2.1093266010284424,
|
| 1365 |
+
"learning_rate": 1.0348994967025012e-05,
|
| 1366 |
+
"loss": 0.1713,
|
| 1367 |
+
"step": 176
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 14.75,
|
| 1371 |
+
"grad_norm": 1.386696696281433,
|
| 1372 |
+
"learning_rate": 1.0261769483078734e-05,
|
| 1373 |
+
"loss": 0.1047,
|
| 1374 |
+
"step": 177
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 14.833333333333334,
|
| 1378 |
+
"grad_norm": 1.551113486289978,
|
| 1379 |
+
"learning_rate": 1.0174524064372837e-05,
|
| 1380 |
+
"loss": 0.1117,
|
| 1381 |
+
"step": 178
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 14.916666666666666,
|
| 1385 |
+
"grad_norm": 1.8801171779632568,
|
| 1386 |
+
"learning_rate": 1.008726535498374e-05,
|
| 1387 |
+
"loss": 0.1112,
|
| 1388 |
+
"step": 179
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 15.0,
|
| 1392 |
+
"grad_norm": 1.9673120975494385,
|
| 1393 |
+
"learning_rate": 1e-05,
|
| 1394 |
+
"loss": 0.1608,
|
| 1395 |
+
"step": 180
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 15.0,
|
| 1399 |
+
"eval_f1": 0.8613614978044017,
|
| 1400 |
+
"eval_loss": 0.4350447952747345,
|
| 1401 |
+
"eval_runtime": 0.1099,
|
| 1402 |
+
"eval_samples_per_second": 1983.243,
|
| 1403 |
+
"eval_steps_per_second": 36.39,
|
| 1404 |
+
"step": 180
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 15.083333333333334,
|
| 1408 |
+
"grad_norm": 2.1639723777770996,
|
| 1409 |
+
"learning_rate": 9.912734645016262e-06,
|
| 1410 |
+
"loss": 0.1573,
|
| 1411 |
+
"step": 181
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 15.166666666666666,
|
| 1415 |
+
"grad_norm": 2.2650837898254395,
|
| 1416 |
+
"learning_rate": 9.825475935627165e-06,
|
| 1417 |
+
"loss": 0.0938,
|
| 1418 |
+
"step": 182
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 15.25,
|
| 1422 |
+
"grad_norm": 3.4237184524536133,
|
| 1423 |
+
"learning_rate": 9.738230516921272e-06,
|
| 1424 |
+
"loss": 0.1833,
|
| 1425 |
+
"step": 183
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 15.333333333333334,
|
| 1429 |
+
"grad_norm": 2.6685876846313477,
|
| 1430 |
+
"learning_rate": 9.651005032974994e-06,
|
| 1431 |
+
"loss": 0.1026,
|
| 1432 |
+
"step": 184
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 15.416666666666666,
|
| 1436 |
+
"grad_norm": 2.295363187789917,
|
| 1437 |
+
"learning_rate": 9.563806126346643e-06,
|
| 1438 |
+
"loss": 0.1174,
|
| 1439 |
+
"step": 185
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 15.5,
|
| 1443 |
+
"grad_norm": 2.2559971809387207,
|
| 1444 |
+
"learning_rate": 9.476640437570562e-06,
|
| 1445 |
+
"loss": 0.1195,
|
| 1446 |
+
"step": 186
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 15.583333333333334,
|
| 1450 |
+
"grad_norm": 2.673501968383789,
|
| 1451 |
+
"learning_rate": 9.38951460465143e-06,
|
| 1452 |
+
"loss": 0.1345,
|
| 1453 |
+
"step": 187
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 15.666666666666666,
|
| 1457 |
+
"grad_norm": 2.226667642593384,
|
| 1458 |
+
"learning_rate": 9.302435262558748e-06,
|
| 1459 |
+
"loss": 0.0963,
|
| 1460 |
+
"step": 188
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 15.75,
|
| 1464 |
+
"grad_norm": 2.004056692123413,
|
| 1465 |
+
"learning_rate": 9.215409042721553e-06,
|
| 1466 |
+
"loss": 0.1008,
|
| 1467 |
+
"step": 189
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 15.833333333333334,
|
| 1471 |
+
"grad_norm": 2.363480806350708,
|
| 1472 |
+
"learning_rate": 9.128442572523418e-06,
|
| 1473 |
+
"loss": 0.1613,
|
| 1474 |
+
"step": 190
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 15.916666666666666,
|
| 1478 |
+
"grad_norm": 1.7934068441390991,
|
| 1479 |
+
"learning_rate": 9.04154247479776e-06,
|
| 1480 |
+
"loss": 0.0956,
|
| 1481 |
+
"step": 191
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 16.0,
|
| 1485 |
+
"grad_norm": 2.194617509841919,
|
| 1486 |
+
"learning_rate": 8.954715367323468e-06,
|
| 1487 |
+
"loss": 0.1073,
|
| 1488 |
+
"step": 192
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 16.0,
|
| 1492 |
+
"eval_f1": 0.8718701056124015,
|
| 1493 |
+
"eval_loss": 0.43139025568962097,
|
| 1494 |
+
"eval_runtime": 0.1114,
|
| 1495 |
+
"eval_samples_per_second": 1956.621,
|
| 1496 |
+
"eval_steps_per_second": 35.901,
|
| 1497 |
+
"step": 192
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 16.083333333333332,
|
| 1501 |
+
"grad_norm": 2.15915846824646,
|
| 1502 |
+
"learning_rate": 8.867967862320935e-06,
|
| 1503 |
+
"loss": 0.1073,
|
| 1504 |
+
"step": 193
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 16.166666666666668,
|
| 1508 |
+
"grad_norm": 1.6649281978607178,
|
| 1509 |
+
"learning_rate": 8.781306565948528e-06,
|
| 1510 |
+
"loss": 0.1163,
|
| 1511 |
+
"step": 194
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 16.25,
|
| 1515 |
+
"grad_norm": 1.5715163946151733,
|
| 1516 |
+
"learning_rate": 8.694738077799487e-06,
|
| 1517 |
+
"loss": 0.087,
|
| 1518 |
+
"step": 195
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 16.333333333333332,
|
| 1522 |
+
"grad_norm": 1.5008565187454224,
|
| 1523 |
+
"learning_rate": 8.60826899039935e-06,
|
| 1524 |
+
"loss": 0.0948,
|
| 1525 |
+
"step": 196
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 16.416666666666668,
|
| 1529 |
+
"grad_norm": 1.899385690689087,
|
| 1530 |
+
"learning_rate": 8.521905888703894e-06,
|
| 1531 |
+
"loss": 0.0903,
|
| 1532 |
+
"step": 197
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 16.5,
|
| 1536 |
+
"grad_norm": 3.4474494457244873,
|
| 1537 |
+
"learning_rate": 8.43565534959769e-06,
|
| 1538 |
+
"loss": 0.1356,
|
| 1539 |
+
"step": 198
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 16.583333333333332,
|
| 1543 |
+
"grad_norm": 0.969510555267334,
|
| 1544 |
+
"learning_rate": 8.349523941393224e-06,
|
| 1545 |
+
"loss": 0.0646,
|
| 1546 |
+
"step": 199
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 16.666666666666668,
|
| 1550 |
+
"grad_norm": 2.511319637298584,
|
| 1551 |
+
"learning_rate": 8.263518223330698e-06,
|
| 1552 |
+
"loss": 0.1066,
|
| 1553 |
+
"step": 200
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 16.75,
|
| 1557 |
+
"grad_norm": 3.4157521724700928,
|
| 1558 |
+
"learning_rate": 8.177644745078525e-06,
|
| 1559 |
+
"loss": 0.1279,
|
| 1560 |
+
"step": 201
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 16.833333333333332,
|
| 1564 |
+
"grad_norm": 1.7571994066238403,
|
| 1565 |
+
"learning_rate": 8.091910046234552e-06,
|
| 1566 |
+
"loss": 0.1098,
|
| 1567 |
+
"step": 202
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 16.916666666666668,
|
| 1571 |
+
"grad_norm": 1.884739875793457,
|
| 1572 |
+
"learning_rate": 8.00632065582803e-06,
|
| 1573 |
+
"loss": 0.1819,
|
| 1574 |
+
"step": 203
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 17.0,
|
| 1578 |
+
"grad_norm": 1.6282474994659424,
|
| 1579 |
+
"learning_rate": 7.92088309182241e-06,
|
| 1580 |
+
"loss": 0.0777,
|
| 1581 |
+
"step": 204
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 17.0,
|
| 1585 |
+
"eval_f1": 0.8576496308551251,
|
| 1586 |
+
"eval_loss": 0.4351810812950134,
|
| 1587 |
+
"eval_runtime": 0.1121,
|
| 1588 |
+
"eval_samples_per_second": 1944.281,
|
| 1589 |
+
"eval_steps_per_second": 35.675,
|
| 1590 |
+
"step": 204
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 17.083333333333332,
|
| 1594 |
+
"grad_norm": 1.4618171453475952,
|
| 1595 |
+
"learning_rate": 7.835603860618973e-06,
|
| 1596 |
+
"loss": 0.1195,
|
| 1597 |
+
"step": 205
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 17.166666666666668,
|
| 1601 |
+
"grad_norm": 1.6853649616241455,
|
| 1602 |
+
"learning_rate": 7.750489456561351e-06,
|
| 1603 |
+
"loss": 0.1022,
|
| 1604 |
+
"step": 206
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 17.25,
|
| 1608 |
+
"grad_norm": 1.8762054443359375,
|
| 1609 |
+
"learning_rate": 7.66554636144095e-06,
|
| 1610 |
+
"loss": 0.0968,
|
| 1611 |
+
"step": 207
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 17.333333333333332,
|
| 1615 |
+
"grad_norm": 1.7310266494750977,
|
| 1616 |
+
"learning_rate": 7.580781044003324e-06,
|
| 1617 |
+
"loss": 0.0824,
|
| 1618 |
+
"step": 208
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 17.416666666666668,
|
| 1622 |
+
"grad_norm": 1.7132197618484497,
|
| 1623 |
+
"learning_rate": 7.496199959455584e-06,
|
| 1624 |
+
"loss": 0.0883,
|
| 1625 |
+
"step": 209
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 17.5,
|
| 1629 |
+
"grad_norm": 1.2777665853500366,
|
| 1630 |
+
"learning_rate": 7.411809548974792e-06,
|
| 1631 |
+
"loss": 0.0782,
|
| 1632 |
+
"step": 210
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 17.583333333333332,
|
| 1636 |
+
"grad_norm": 2.170820951461792,
|
| 1637 |
+
"learning_rate": 7.327616239217432e-06,
|
| 1638 |
+
"loss": 0.0914,
|
| 1639 |
+
"step": 211
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 17.666666666666668,
|
| 1643 |
+
"grad_norm": 1.7623423337936401,
|
| 1644 |
+
"learning_rate": 7.243626441830009e-06,
|
| 1645 |
+
"loss": 0.1125,
|
| 1646 |
+
"step": 212
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 17.75,
|
| 1650 |
+
"grad_norm": 1.2972196340560913,
|
| 1651 |
+
"learning_rate": 7.159846552960774e-06,
|
| 1652 |
+
"loss": 0.0768,
|
| 1653 |
+
"step": 213
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 17.833333333333332,
|
| 1657 |
+
"grad_norm": 1.98750901222229,
|
| 1658 |
+
"learning_rate": 7.076282952772634e-06,
|
| 1659 |
+
"loss": 0.1047,
|
| 1660 |
+
"step": 214
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 17.916666666666668,
|
| 1664 |
+
"grad_norm": 1.9874991178512573,
|
| 1665 |
+
"learning_rate": 6.992942004957271e-06,
|
| 1666 |
+
"loss": 0.079,
|
| 1667 |
+
"step": 215
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 18.0,
|
| 1671 |
+
"grad_norm": 2.6068506240844727,
|
| 1672 |
+
"learning_rate": 6.909830056250527e-06,
|
| 1673 |
+
"loss": 0.122,
|
| 1674 |
+
"step": 216
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 18.0,
|
| 1678 |
+
"eval_f1": 0.8677297848792973,
|
| 1679 |
+
"eval_loss": 0.43646883964538574,
|
| 1680 |
+
"eval_runtime": 0.1109,
|
| 1681 |
+
"eval_samples_per_second": 1966.556,
|
| 1682 |
+
"eval_steps_per_second": 36.084,
|
| 1683 |
+
"step": 216
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 18.083333333333332,
|
| 1687 |
+
"grad_norm": 1.915047287940979,
|
| 1688 |
+
"learning_rate": 6.826953435949081e-06,
|
| 1689 |
+
"loss": 0.0718,
|
| 1690 |
+
"step": 217
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 18.166666666666668,
|
| 1694 |
+
"grad_norm": 1.7660863399505615,
|
| 1695 |
+
"learning_rate": 6.744318455428436e-06,
|
| 1696 |
+
"loss": 0.1025,
|
| 1697 |
+
"step": 218
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 18.25,
|
| 1701 |
+
"grad_norm": 2.070688486099243,
|
| 1702 |
+
"learning_rate": 6.661931407662292e-06,
|
| 1703 |
+
"loss": 0.1481,
|
| 1704 |
+
"step": 219
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 18.333333333333332,
|
| 1708 |
+
"grad_norm": 1.786027193069458,
|
| 1709 |
+
"learning_rate": 6.579798566743314e-06,
|
| 1710 |
+
"loss": 0.0858,
|
| 1711 |
+
"step": 220
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 18.416666666666668,
|
| 1715 |
+
"grad_norm": 2.9605553150177,
|
| 1716 |
+
"learning_rate": 6.497926187405326e-06,
|
| 1717 |
+
"loss": 0.0774,
|
| 1718 |
+
"step": 221
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 18.5,
|
| 1722 |
+
"grad_norm": 2.3121283054351807,
|
| 1723 |
+
"learning_rate": 6.4163205045469975e-06,
|
| 1724 |
+
"loss": 0.074,
|
| 1725 |
+
"step": 222
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 18.583333333333332,
|
| 1729 |
+
"grad_norm": 3.045610189437866,
|
| 1730 |
+
"learning_rate": 6.334987732757028e-06,
|
| 1731 |
+
"loss": 0.1162,
|
| 1732 |
+
"step": 223
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 18.666666666666668,
|
| 1736 |
+
"grad_norm": 1.1786707639694214,
|
| 1737 |
+
"learning_rate": 6.25393406584088e-06,
|
| 1738 |
+
"loss": 0.0857,
|
| 1739 |
+
"step": 224
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 18.75,
|
| 1743 |
+
"grad_norm": 1.8060725927352905,
|
| 1744 |
+
"learning_rate": 6.173165676349103e-06,
|
| 1745 |
+
"loss": 0.0951,
|
| 1746 |
+
"step": 225
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 18.833333333333332,
|
| 1750 |
+
"grad_norm": 1.1193745136260986,
|
| 1751 |
+
"learning_rate": 6.092688715107265e-06,
|
| 1752 |
+
"loss": 0.076,
|
| 1753 |
+
"step": 226
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 18.916666666666668,
|
| 1757 |
+
"grad_norm": 1.3846014738082886,
|
| 1758 |
+
"learning_rate": 6.0125093107475385e-06,
|
| 1759 |
+
"loss": 0.0574,
|
| 1760 |
+
"step": 227
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 19.0,
|
| 1764 |
+
"grad_norm": 1.4988336563110352,
|
| 1765 |
+
"learning_rate": 5.932633569242e-06,
|
| 1766 |
+
"loss": 0.0678,
|
| 1767 |
+
"step": 228
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 19.0,
|
| 1771 |
+
"eval_f1": 0.8522723841933064,
|
| 1772 |
+
"eval_loss": 0.44795626401901245,
|
| 1773 |
+
"eval_runtime": 0.1106,
|
| 1774 |
+
"eval_samples_per_second": 1971.505,
|
| 1775 |
+
"eval_steps_per_second": 36.174,
|
| 1776 |
+
"step": 228
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 19.083333333333332,
|
| 1780 |
+
"grad_norm": 2.050723075866699,
|
| 1781 |
+
"learning_rate": 5.853067573437612e-06,
|
| 1782 |
+
"loss": 0.0937,
|
| 1783 |
+
"step": 229
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 19.166666666666668,
|
| 1787 |
+
"grad_norm": 2.385880708694458,
|
| 1788 |
+
"learning_rate": 5.773817382593008e-06,
|
| 1789 |
+
"loss": 0.0849,
|
| 1790 |
+
"step": 230
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 19.25,
|
| 1794 |
+
"grad_norm": 2.387587070465088,
|
| 1795 |
+
"learning_rate": 5.694889031917047e-06,
|
| 1796 |
+
"loss": 0.0689,
|
| 1797 |
+
"step": 231
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 19.333333333333332,
|
| 1801 |
+
"grad_norm": 1.3751537799835205,
|
| 1802 |
+
"learning_rate": 5.616288532109225e-06,
|
| 1803 |
+
"loss": 0.1126,
|
| 1804 |
+
"step": 232
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 19.416666666666668,
|
| 1808 |
+
"grad_norm": 2.130180835723877,
|
| 1809 |
+
"learning_rate": 5.5380218689019125e-06,
|
| 1810 |
+
"loss": 0.1119,
|
| 1811 |
+
"step": 233
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 19.5,
|
| 1815 |
+
"grad_norm": 1.6648838520050049,
|
| 1816 |
+
"learning_rate": 5.460095002604533e-06,
|
| 1817 |
+
"loss": 0.0599,
|
| 1818 |
+
"step": 234
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 19.583333333333332,
|
| 1822 |
+
"grad_norm": 1.598511815071106,
|
| 1823 |
+
"learning_rate": 5.382513867649663e-06,
|
| 1824 |
+
"loss": 0.0732,
|
| 1825 |
+
"step": 235
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 19.666666666666668,
|
| 1829 |
+
"grad_norm": 2.2660152912139893,
|
| 1830 |
+
"learning_rate": 5.305284372141095e-06,
|
| 1831 |
+
"loss": 0.0989,
|
| 1832 |
+
"step": 236
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 19.75,
|
| 1836 |
+
"grad_norm": 1.437440037727356,
|
| 1837 |
+
"learning_rate": 5.228412397403916e-06,
|
| 1838 |
+
"loss": 0.0731,
|
| 1839 |
+
"step": 237
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 19.833333333333332,
|
| 1843 |
+
"grad_norm": 2.2391016483306885,
|
| 1844 |
+
"learning_rate": 5.151903797536631e-06,
|
| 1845 |
+
"loss": 0.0906,
|
| 1846 |
+
"step": 238
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 19.916666666666668,
|
| 1850 |
+
"grad_norm": 1.3197194337844849,
|
| 1851 |
+
"learning_rate": 5.075764398965331e-06,
|
| 1852 |
+
"loss": 0.0646,
|
| 1853 |
+
"step": 239
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 20.0,
|
| 1857 |
+
"grad_norm": 1.2512582540512085,
|
| 1858 |
+
"learning_rate": 5.000000000000003e-06,
|
| 1859 |
+
"loss": 0.0509,
|
| 1860 |
+
"step": 240
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 20.0,
|
| 1864 |
+
"eval_f1": 0.8578401118067667,
|
| 1865 |
+
"eval_loss": 0.4534085690975189,
|
| 1866 |
+
"eval_runtime": 0.1114,
|
| 1867 |
+
"eval_samples_per_second": 1957.609,
|
| 1868 |
+
"eval_steps_per_second": 35.919,
|
| 1869 |
+
"step": 240
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 20.083333333333332,
|
| 1873 |
+
"grad_norm": 1.6873009204864502,
|
| 1874 |
+
"learning_rate": 4.924616370392962e-06,
|
| 1875 |
+
"loss": 0.0865,
|
| 1876 |
+
"step": 241
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 20.166666666666668,
|
| 1880 |
+
"grad_norm": 2.690499782562256,
|
| 1881 |
+
"learning_rate": 4.849619250899458e-06,
|
| 1882 |
+
"loss": 0.0616,
|
| 1883 |
+
"step": 242
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 20.25,
|
| 1887 |
+
"grad_norm": 2.2095584869384766,
|
| 1888 |
+
"learning_rate": 4.775014352840512e-06,
|
| 1889 |
+
"loss": 0.077,
|
| 1890 |
+
"step": 243
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 20.333333333333332,
|
| 1894 |
+
"grad_norm": 2.5876200199127197,
|
| 1895 |
+
"learning_rate": 4.700807357667953e-06,
|
| 1896 |
+
"loss": 0.0806,
|
| 1897 |
+
"step": 244
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 20.416666666666668,
|
| 1901 |
+
"grad_norm": 2.9219162464141846,
|
| 1902 |
+
"learning_rate": 4.627003916531761e-06,
|
| 1903 |
+
"loss": 0.0875,
|
| 1904 |
+
"step": 245
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 20.5,
|
| 1908 |
+
"grad_norm": 1.6222124099731445,
|
| 1909 |
+
"learning_rate": 4.5536096498497295e-06,
|
| 1910 |
+
"loss": 0.1145,
|
| 1911 |
+
"step": 246
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 20.583333333333332,
|
| 1915 |
+
"grad_norm": 2.0779643058776855,
|
| 1916 |
+
"learning_rate": 4.480630146879419e-06,
|
| 1917 |
+
"loss": 0.0757,
|
| 1918 |
+
"step": 247
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 20.666666666666668,
|
| 1922 |
+
"grad_norm": 2.8607051372528076,
|
| 1923 |
+
"learning_rate": 4.408070965292534e-06,
|
| 1924 |
+
"loss": 0.0813,
|
| 1925 |
+
"step": 248
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 20.75,
|
| 1929 |
+
"grad_norm": 1.2463396787643433,
|
| 1930 |
+
"learning_rate": 4.335937630751675e-06,
|
| 1931 |
+
"loss": 0.0543,
|
| 1932 |
+
"step": 249
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 20.833333333333332,
|
| 1936 |
+
"grad_norm": 2.674499273300171,
|
| 1937 |
+
"learning_rate": 4.264235636489542e-06,
|
| 1938 |
+
"loss": 0.0651,
|
| 1939 |
+
"step": 250
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 20.916666666666668,
|
| 1943 |
+
"grad_norm": 2.0927512645721436,
|
| 1944 |
+
"learning_rate": 4.192970442890602e-06,
|
| 1945 |
+
"loss": 0.0675,
|
| 1946 |
+
"step": 251
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 21.0,
|
| 1950 |
+
"grad_norm": 1.6905465126037598,
|
| 1951 |
+
"learning_rate": 4.12214747707527e-06,
|
| 1952 |
+
"loss": 0.0771,
|
| 1953 |
+
"step": 252
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 21.0,
|
| 1957 |
+
"eval_f1": 0.8624653833164471,
|
| 1958 |
+
"eval_loss": 0.4554041028022766,
|
| 1959 |
+
"eval_runtime": 0.1165,
|
| 1960 |
+
"eval_samples_per_second": 1870.546,
|
| 1961 |
+
"eval_steps_per_second": 34.322,
|
| 1962 |
+
"step": 252
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 21.083333333333332,
|
| 1966 |
+
"grad_norm": 1.1010918617248535,
|
| 1967 |
+
"learning_rate": 4.051772132486589e-06,
|
| 1968 |
+
"loss": 0.0535,
|
| 1969 |
+
"step": 253
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 21.166666666666668,
|
| 1973 |
+
"grad_norm": 2.0781238079071045,
|
| 1974 |
+
"learning_rate": 3.981849768479516e-06,
|
| 1975 |
+
"loss": 0.0801,
|
| 1976 |
+
"step": 254
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 21.25,
|
| 1980 |
+
"grad_norm": 1.435062050819397,
|
| 1981 |
+
"learning_rate": 3.912385709912794e-06,
|
| 1982 |
+
"loss": 0.0567,
|
| 1983 |
+
"step": 255
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 21.333333333333332,
|
| 1987 |
+
"grad_norm": 2.286985397338867,
|
| 1988 |
+
"learning_rate": 3.8433852467434175e-06,
|
| 1989 |
+
"loss": 0.1132,
|
| 1990 |
+
"step": 256
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 21.416666666666668,
|
| 1994 |
+
"grad_norm": 1.7003625631332397,
|
| 1995 |
+
"learning_rate": 3.774853633623806e-06,
|
| 1996 |
+
"loss": 0.0883,
|
| 1997 |
+
"step": 257
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 21.5,
|
| 2001 |
+
"grad_norm": 1.5241916179656982,
|
| 2002 |
+
"learning_rate": 3.7067960895016277e-06,
|
| 2003 |
+
"loss": 0.0561,
|
| 2004 |
+
"step": 258
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 21.583333333333332,
|
| 2008 |
+
"grad_norm": 1.5684936046600342,
|
| 2009 |
+
"learning_rate": 3.6392177972223596e-06,
|
| 2010 |
+
"loss": 0.0827,
|
| 2011 |
+
"step": 259
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 21.666666666666668,
|
| 2015 |
+
"grad_norm": 1.1295371055603027,
|
| 2016 |
+
"learning_rate": 3.5721239031346067e-06,
|
| 2017 |
+
"loss": 0.0503,
|
| 2018 |
+
"step": 260
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 21.75,
|
| 2022 |
+
"grad_norm": 1.6498470306396484,
|
| 2023 |
+
"learning_rate": 3.505519516698165e-06,
|
| 2024 |
+
"loss": 0.088,
|
| 2025 |
+
"step": 261
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 21.833333333333332,
|
| 2029 |
+
"grad_norm": 2.4177968502044678,
|
| 2030 |
+
"learning_rate": 3.4394097100949286e-06,
|
| 2031 |
+
"loss": 0.0516,
|
| 2032 |
+
"step": 262
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 21.916666666666668,
|
| 2036 |
+
"grad_norm": 1.5983963012695312,
|
| 2037 |
+
"learning_rate": 3.3737995178426276e-06,
|
| 2038 |
+
"loss": 0.0635,
|
| 2039 |
+
"step": 263
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 22.0,
|
| 2043 |
+
"grad_norm": 3.1225271224975586,
|
| 2044 |
+
"learning_rate": 3.308693936411421e-06,
|
| 2045 |
+
"loss": 0.0955,
|
| 2046 |
+
"step": 264
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 22.0,
|
| 2050 |
+
"eval_f1": 0.8582176116005,
|
| 2051 |
+
"eval_loss": 0.4626469016075134,
|
| 2052 |
+
"eval_runtime": 0.1141,
|
| 2053 |
+
"eval_samples_per_second": 1910.773,
|
| 2054 |
+
"eval_steps_per_second": 35.06,
|
| 2055 |
+
"step": 264
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 22.083333333333332,
|
| 2059 |
+
"grad_norm": 1.6301372051239014,
|
| 2060 |
+
"learning_rate": 3.2440979238433977e-06,
|
| 2061 |
+
"loss": 0.0518,
|
| 2062 |
+
"step": 265
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 22.166666666666668,
|
| 2066 |
+
"grad_norm": 2.003352165222168,
|
| 2067 |
+
"learning_rate": 3.1800163993750166e-06,
|
| 2068 |
+
"loss": 0.0823,
|
| 2069 |
+
"step": 266
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 22.25,
|
| 2073 |
+
"grad_norm": 1.1521472930908203,
|
| 2074 |
+
"learning_rate": 3.116454243062459e-06,
|
| 2075 |
+
"loss": 0.0641,
|
| 2076 |
+
"step": 267
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 22.333333333333332,
|
| 2080 |
+
"grad_norm": 0.8404669761657715,
|
| 2081 |
+
"learning_rate": 3.0534162954100264e-06,
|
| 2082 |
+
"loss": 0.0452,
|
| 2083 |
+
"step": 268
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 22.416666666666668,
|
| 2087 |
+
"grad_norm": 0.963030993938446,
|
| 2088 |
+
"learning_rate": 2.990907357001491e-06,
|
| 2089 |
+
"loss": 0.0427,
|
| 2090 |
+
"step": 269
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 22.5,
|
| 2094 |
+
"grad_norm": 2.651822566986084,
|
| 2095 |
+
"learning_rate": 2.9289321881345257e-06,
|
| 2096 |
+
"loss": 0.0927,
|
| 2097 |
+
"step": 270
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 22.583333333333332,
|
| 2101 |
+
"grad_norm": 1.8716309070587158,
|
| 2102 |
+
"learning_rate": 2.867495508458186e-06,
|
| 2103 |
+
"loss": 0.0706,
|
| 2104 |
+
"step": 271
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 22.666666666666668,
|
| 2108 |
+
"grad_norm": 1.8748180866241455,
|
| 2109 |
+
"learning_rate": 2.8066019966134907e-06,
|
| 2110 |
+
"loss": 0.0834,
|
| 2111 |
+
"step": 272
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 22.75,
|
| 2115 |
+
"grad_norm": 1.859476089477539,
|
| 2116 |
+
"learning_rate": 2.746256289877126e-06,
|
| 2117 |
+
"loss": 0.0669,
|
| 2118 |
+
"step": 273
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 22.833333333333332,
|
| 2122 |
+
"grad_norm": 0.9171540141105652,
|
| 2123 |
+
"learning_rate": 2.6864629838082957e-06,
|
| 2124 |
+
"loss": 0.045,
|
| 2125 |
+
"step": 274
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 22.916666666666668,
|
| 2129 |
+
"grad_norm": 1.7010291814804077,
|
| 2130 |
+
"learning_rate": 2.6272266318987606e-06,
|
| 2131 |
+
"loss": 0.0753,
|
| 2132 |
+
"step": 275
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 23.0,
|
| 2136 |
+
"grad_norm": 2.0408334732055664,
|
| 2137 |
+
"learning_rate": 2.5685517452260566e-06,
|
| 2138 |
+
"loss": 0.0911,
|
| 2139 |
+
"step": 276
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 23.0,
|
| 2143 |
+
"eval_f1": 0.8520156075449599,
|
| 2144 |
+
"eval_loss": 0.4653885066509247,
|
| 2145 |
+
"eval_runtime": 0.1119,
|
| 2146 |
+
"eval_samples_per_second": 1948.254,
|
| 2147 |
+
"eval_steps_per_second": 35.748,
|
| 2148 |
+
"step": 276
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 23.083333333333332,
|
| 2152 |
+
"grad_norm": 2.0291988849639893,
|
| 2153 |
+
"learning_rate": 2.5104427921099783e-06,
|
| 2154 |
+
"loss": 0.0645,
|
| 2155 |
+
"step": 277
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 23.166666666666668,
|
| 2159 |
+
"grad_norm": 0.9454895257949829,
|
| 2160 |
+
"learning_rate": 2.45290419777228e-06,
|
| 2161 |
+
"loss": 0.0473,
|
| 2162 |
+
"step": 278
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 23.25,
|
| 2166 |
+
"grad_norm": 1.0240933895111084,
|
| 2167 |
+
"learning_rate": 2.395940343999691e-06,
|
| 2168 |
+
"loss": 0.0504,
|
| 2169 |
+
"step": 279
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 23.333333333333332,
|
| 2173 |
+
"grad_norm": 1.8231364488601685,
|
| 2174 |
+
"learning_rate": 2.339555568810221e-06,
|
| 2175 |
+
"loss": 0.0578,
|
| 2176 |
+
"step": 280
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 23.416666666666668,
|
| 2180 |
+
"grad_norm": 2.723802089691162,
|
| 2181 |
+
"learning_rate": 2.2837541661228024e-06,
|
| 2182 |
+
"loss": 0.111,
|
| 2183 |
+
"step": 281
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 23.5,
|
| 2187 |
+
"grad_norm": 1.4741343259811401,
|
| 2188 |
+
"learning_rate": 2.2285403854302912e-06,
|
| 2189 |
+
"loss": 0.0536,
|
| 2190 |
+
"step": 282
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 23.583333333333332,
|
| 2194 |
+
"grad_norm": 3.269688844680786,
|
| 2195 |
+
"learning_rate": 2.173918431475861e-06,
|
| 2196 |
+
"loss": 0.0904,
|
| 2197 |
+
"step": 283
|
| 2198 |
+
},
|
| 2199 |
+
{
|
| 2200 |
+
"epoch": 23.666666666666668,
|
| 2201 |
+
"grad_norm": 2.1853411197662354,
|
| 2202 |
+
"learning_rate": 2.119892463932781e-06,
|
| 2203 |
+
"loss": 0.1204,
|
| 2204 |
+
"step": 284
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 23.75,
|
| 2208 |
+
"grad_norm": 4.364923000335693,
|
| 2209 |
+
"learning_rate": 2.0664665970876496e-06,
|
| 2210 |
+
"loss": 0.1136,
|
| 2211 |
+
"step": 285
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 23.833333333333332,
|
| 2215 |
+
"grad_norm": 1.406599760055542,
|
| 2216 |
+
"learning_rate": 2.013644899527074e-06,
|
| 2217 |
+
"loss": 0.0516,
|
| 2218 |
+
"step": 286
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 23.916666666666668,
|
| 2222 |
+
"grad_norm": 1.8309155702590942,
|
| 2223 |
+
"learning_rate": 1.961431393827827e-06,
|
| 2224 |
+
"loss": 0.0553,
|
| 2225 |
+
"step": 287
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 24.0,
|
| 2229 |
+
"grad_norm": 1.9190561771392822,
|
| 2230 |
+
"learning_rate": 1.9098300562505266e-06,
|
| 2231 |
+
"loss": 0.0753,
|
| 2232 |
+
"step": 288
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 24.0,
|
| 2236 |
+
"eval_f1": 0.8567961564859299,
|
| 2237 |
+
"eval_loss": 0.4646962285041809,
|
| 2238 |
+
"eval_runtime": 0.1111,
|
| 2239 |
+
"eval_samples_per_second": 1962.146,
|
| 2240 |
+
"eval_steps_per_second": 36.003,
|
| 2241 |
+
"step": 288
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 24.083333333333332,
|
| 2245 |
+
"grad_norm": 2.4621970653533936,
|
| 2246 |
+
"learning_rate": 1.858844816436809e-06,
|
| 2247 |
+
"loss": 0.1027,
|
| 2248 |
+
"step": 289
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 24.166666666666668,
|
| 2252 |
+
"grad_norm": 2.6477534770965576,
|
| 2253 |
+
"learning_rate": 1.808479557110081e-06,
|
| 2254 |
+
"loss": 0.0844,
|
| 2255 |
+
"step": 290
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 24.25,
|
| 2259 |
+
"grad_norm": 1.848976492881775,
|
| 2260 |
+
"learning_rate": 1.7587381137798432e-06,
|
| 2261 |
+
"loss": 0.0592,
|
| 2262 |
+
"step": 291
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 24.333333333333332,
|
| 2266 |
+
"grad_norm": 1.919098973274231,
|
| 2267 |
+
"learning_rate": 1.709624274449584e-06,
|
| 2268 |
+
"loss": 0.058,
|
| 2269 |
+
"step": 292
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 24.416666666666668,
|
| 2273 |
+
"grad_norm": 1.8423515558242798,
|
| 2274 |
+
"learning_rate": 1.6611417793283192e-06,
|
| 2275 |
+
"loss": 0.0767,
|
| 2276 |
+
"step": 293
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 24.5,
|
| 2280 |
+
"grad_norm": 1.4706356525421143,
|
| 2281 |
+
"learning_rate": 1.6132943205457607e-06,
|
| 2282 |
+
"loss": 0.057,
|
| 2283 |
+
"step": 294
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 24.583333333333332,
|
| 2287 |
+
"grad_norm": 0.7103891968727112,
|
| 2288 |
+
"learning_rate": 1.566085541871145e-06,
|
| 2289 |
+
"loss": 0.0378,
|
| 2290 |
+
"step": 295
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 24.666666666666668,
|
| 2294 |
+
"grad_norm": 1.5366411209106445,
|
| 2295 |
+
"learning_rate": 1.5195190384357405e-06,
|
| 2296 |
+
"loss": 0.0459,
|
| 2297 |
+
"step": 296
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 24.75,
|
| 2301 |
+
"grad_norm": 1.5824693441390991,
|
| 2302 |
+
"learning_rate": 1.4735983564590784e-06,
|
| 2303 |
+
"loss": 0.0755,
|
| 2304 |
+
"step": 297
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 24.833333333333332,
|
| 2308 |
+
"grad_norm": 2.309704065322876,
|
| 2309 |
+
"learning_rate": 1.4283269929788779e-06,
|
| 2310 |
+
"loss": 0.0797,
|
| 2311 |
+
"step": 298
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 24.916666666666668,
|
| 2315 |
+
"grad_norm": 1.1554335355758667,
|
| 2316 |
+
"learning_rate": 1.3837083955847418e-06,
|
| 2317 |
+
"loss": 0.0497,
|
| 2318 |
+
"step": 299
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 25.0,
|
| 2322 |
+
"grad_norm": 1.337747573852539,
|
| 2323 |
+
"learning_rate": 1.339745962155613e-06,
|
| 2324 |
+
"loss": 0.0503,
|
| 2325 |
+
"step": 300
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 25.0,
|
| 2329 |
+
"eval_f1": 0.8678584635765896,
|
| 2330 |
+
"eval_loss": 0.4654601514339447,
|
| 2331 |
+
"eval_runtime": 0.1165,
|
| 2332 |
+
"eval_samples_per_second": 1872.043,
|
| 2333 |
+
"eval_steps_per_second": 34.349,
|
| 2334 |
+
"step": 300
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 25.083333333333332,
|
| 2338 |
+
"grad_norm": 2.1485424041748047,
|
| 2339 |
+
"learning_rate": 1.2964430406010032e-06,
|
| 2340 |
+
"loss": 0.0631,
|
| 2341 |
+
"step": 301
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 25.166666666666668,
|
| 2345 |
+
"grad_norm": 2.0165176391601562,
|
| 2346 |
+
"learning_rate": 1.2538029286060428e-06,
|
| 2347 |
+
"loss": 0.0675,
|
| 2348 |
+
"step": 302
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 25.25,
|
| 2352 |
+
"grad_norm": 1.0340204238891602,
|
| 2353 |
+
"learning_rate": 1.2118288733803474e-06,
|
| 2354 |
+
"loss": 0.0447,
|
| 2355 |
+
"step": 303
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 25.333333333333332,
|
| 2359 |
+
"grad_norm": 1.1221197843551636,
|
| 2360 |
+
"learning_rate": 1.1705240714107301e-06,
|
| 2361 |
+
"loss": 0.0449,
|
| 2362 |
+
"step": 304
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 25.416666666666668,
|
| 2366 |
+
"grad_norm": 1.3471612930297852,
|
| 2367 |
+
"learning_rate": 1.129891668217783e-06,
|
| 2368 |
+
"loss": 0.0554,
|
| 2369 |
+
"step": 305
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 25.5,
|
| 2373 |
+
"grad_norm": 1.3579171895980835,
|
| 2374 |
+
"learning_rate": 1.0899347581163222e-06,
|
| 2375 |
+
"loss": 0.0507,
|
| 2376 |
+
"step": 306
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 25.583333333333332,
|
| 2380 |
+
"grad_norm": 1.9497525691986084,
|
| 2381 |
+
"learning_rate": 1.0506563839797501e-06,
|
| 2382 |
+
"loss": 0.0779,
|
| 2383 |
+
"step": 307
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 25.666666666666668,
|
| 2387 |
+
"grad_norm": 1.707481026649475,
|
| 2388 |
+
"learning_rate": 1.012059537008332e-06,
|
| 2389 |
+
"loss": 0.0553,
|
| 2390 |
+
"step": 308
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 25.75,
|
| 2394 |
+
"grad_norm": 1.7610981464385986,
|
| 2395 |
+
"learning_rate": 9.74147156501396e-07,
|
| 2396 |
+
"loss": 0.0635,
|
| 2397 |
+
"step": 309
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 25.833333333333332,
|
| 2401 |
+
"grad_norm": 1.118330955505371,
|
| 2402 |
+
"learning_rate": 9.369221296335007e-07,
|
| 2403 |
+
"loss": 0.0684,
|
| 2404 |
+
"step": 310
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 25.916666666666668,
|
| 2408 |
+
"grad_norm": 1.4681358337402344,
|
| 2409 |
+
"learning_rate": 9.00387291234569e-07,
|
| 2410 |
+
"loss": 0.0506,
|
| 2411 |
+
"step": 311
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 26.0,
|
| 2415 |
+
"grad_norm": 1.6009918451309204,
|
| 2416 |
+
"learning_rate": 8.645454235739903e-07,
|
| 2417 |
+
"loss": 0.051,
|
| 2418 |
+
"step": 312
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 26.0,
|
| 2422 |
+
"eval_f1": 0.8680197132616488,
|
| 2423 |
+
"eval_loss": 0.4655022919178009,
|
| 2424 |
+
"eval_runtime": 0.1104,
|
| 2425 |
+
"eval_samples_per_second": 1974.911,
|
| 2426 |
+
"eval_steps_per_second": 36.237,
|
| 2427 |
+
"step": 312
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 26.083333333333332,
|
| 2431 |
+
"grad_norm": 2.7175793647766113,
|
| 2432 |
+
"learning_rate": 8.293992561487596e-07,
|
| 2433 |
+
"loss": 0.0789,
|
| 2434 |
+
"step": 313
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 26.166666666666668,
|
| 2438 |
+
"grad_norm": 1.6440610885620117,
|
| 2439 |
+
"learning_rate": 7.949514654755963e-07,
|
| 2440 |
+
"loss": 0.0877,
|
| 2441 |
+
"step": 314
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 26.25,
|
| 2445 |
+
"grad_norm": 0.9847065806388855,
|
| 2446 |
+
"learning_rate": 7.612046748871327e-07,
|
| 2447 |
+
"loss": 0.0394,
|
| 2448 |
+
"step": 315
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 26.333333333333332,
|
| 2452 |
+
"grad_norm": 1.6793824434280396,
|
| 2453 |
+
"learning_rate": 7.281614543321269e-07,
|
| 2454 |
+
"loss": 0.0859,
|
| 2455 |
+
"step": 316
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 26.416666666666668,
|
| 2459 |
+
"grad_norm": 1.883439302444458,
|
| 2460 |
+
"learning_rate": 6.958243201797554e-07,
|
| 2461 |
+
"loss": 0.0708,
|
| 2462 |
+
"step": 317
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 26.5,
|
| 2466 |
+
"grad_norm": 1.155173420906067,
|
| 2467 |
+
"learning_rate": 6.641957350279838e-07,
|
| 2468 |
+
"loss": 0.0429,
|
| 2469 |
+
"step": 318
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 26.583333333333332,
|
| 2473 |
+
"grad_norm": 1.1771842241287231,
|
| 2474 |
+
"learning_rate": 6.332781075160244e-07,
|
| 2475 |
+
"loss": 0.0466,
|
| 2476 |
+
"step": 319
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 26.666666666666668,
|
| 2480 |
+
"grad_norm": 2.2564806938171387,
|
| 2481 |
+
"learning_rate": 6.030737921409169e-07,
|
| 2482 |
+
"loss": 0.0721,
|
| 2483 |
+
"step": 320
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 26.75,
|
| 2487 |
+
"grad_norm": 1.109580159187317,
|
| 2488 |
+
"learning_rate": 5.735850890782158e-07,
|
| 2489 |
+
"loss": 0.044,
|
| 2490 |
+
"step": 321
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 26.833333333333332,
|
| 2494 |
+
"grad_norm": 1.1505159139633179,
|
| 2495 |
+
"learning_rate": 5.448142440068316e-07,
|
| 2496 |
+
"loss": 0.0514,
|
| 2497 |
+
"step": 322
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 26.916666666666668,
|
| 2501 |
+
"grad_norm": 1.2787203788757324,
|
| 2502 |
+
"learning_rate": 5.167634479380068e-07,
|
| 2503 |
+
"loss": 0.053,
|
| 2504 |
+
"step": 323
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 27.0,
|
| 2508 |
+
"grad_norm": 0.9756549000740051,
|
| 2509 |
+
"learning_rate": 4.894348370484648e-07,
|
| 2510 |
+
"loss": 0.0464,
|
| 2511 |
+
"step": 324
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 27.0,
|
| 2515 |
+
"eval_f1": 0.8680197132616488,
|
| 2516 |
+
"eval_loss": 0.4661083519458771,
|
| 2517 |
+
"eval_runtime": 0.1099,
|
| 2518 |
+
"eval_samples_per_second": 1983.161,
|
| 2519 |
+
"eval_steps_per_second": 36.388,
|
| 2520 |
+
"step": 324
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 27.083333333333332,
|
| 2524 |
+
"grad_norm": 1.9149914979934692,
|
| 2525 |
+
"learning_rate": 4.628304925177318e-07,
|
| 2526 |
+
"loss": 0.0856,
|
| 2527 |
+
"step": 325
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 27.166666666666668,
|
| 2531 |
+
"grad_norm": 0.5506188869476318,
|
| 2532 |
+
"learning_rate": 4.3695244036964567e-07,
|
| 2533 |
+
"loss": 0.0329,
|
| 2534 |
+
"step": 326
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 27.25,
|
| 2538 |
+
"grad_norm": 1.3258453607559204,
|
| 2539 |
+
"learning_rate": 4.118026513180695e-07,
|
| 2540 |
+
"loss": 0.0474,
|
| 2541 |
+
"step": 327
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 27.333333333333332,
|
| 2545 |
+
"grad_norm": 1.2896806001663208,
|
| 2546 |
+
"learning_rate": 3.8738304061681107e-07,
|
| 2547 |
+
"loss": 0.0562,
|
| 2548 |
+
"step": 328
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 27.416666666666668,
|
| 2552 |
+
"grad_norm": 1.6682846546173096,
|
| 2553 |
+
"learning_rate": 3.6369546791377054e-07,
|
| 2554 |
+
"loss": 0.0653,
|
| 2555 |
+
"step": 329
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 27.5,
|
| 2559 |
+
"grad_norm": 1.1842490434646606,
|
| 2560 |
+
"learning_rate": 3.4074173710931804e-07,
|
| 2561 |
+
"loss": 0.0395,
|
| 2562 |
+
"step": 330
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 27.583333333333332,
|
| 2566 |
+
"grad_norm": 1.6091792583465576,
|
| 2567 |
+
"learning_rate": 3.185235962189237e-07,
|
| 2568 |
+
"loss": 0.0644,
|
| 2569 |
+
"step": 331
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 27.666666666666668,
|
| 2573 |
+
"grad_norm": 1.7737246751785278,
|
| 2574 |
+
"learning_rate": 2.970427372400353e-07,
|
| 2575 |
+
"loss": 0.0534,
|
| 2576 |
+
"step": 332
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 27.75,
|
| 2580 |
+
"grad_norm": 1.5198438167572021,
|
| 2581 |
+
"learning_rate": 2.7630079602323447e-07,
|
| 2582 |
+
"loss": 0.0482,
|
| 2583 |
+
"step": 333
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 27.833333333333332,
|
| 2587 |
+
"grad_norm": 1.4363399744033813,
|
| 2588 |
+
"learning_rate": 2.5629935214764866e-07,
|
| 2589 |
+
"loss": 0.0571,
|
| 2590 |
+
"step": 334
|
| 2591 |
+
},
|
| 2592 |
+
{
|
| 2593 |
+
"epoch": 27.916666666666668,
|
| 2594 |
+
"grad_norm": 0.946235179901123,
|
| 2595 |
+
"learning_rate": 2.370399288006664e-07,
|
| 2596 |
+
"loss": 0.0489,
|
| 2597 |
+
"step": 335
|
| 2598 |
+
},
|
| 2599 |
+
{
|
| 2600 |
+
"epoch": 28.0,
|
| 2601 |
+
"grad_norm": 1.5237261056900024,
|
| 2602 |
+
"learning_rate": 2.1852399266194312e-07,
|
| 2603 |
+
"loss": 0.0697,
|
| 2604 |
+
"step": 336
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 28.0,
|
| 2608 |
+
"eval_f1": 0.8680197132616488,
|
| 2609 |
+
"eval_loss": 0.465838223695755,
|
| 2610 |
+
"eval_runtime": 0.1167,
|
| 2611 |
+
"eval_samples_per_second": 1868.401,
|
| 2612 |
+
"eval_steps_per_second": 34.283,
|
| 2613 |
+
"step": 336
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 28.083333333333332,
|
| 2617 |
+
"grad_norm": 1.9383271932601929,
|
| 2618 |
+
"learning_rate": 2.0075295379170413e-07,
|
| 2619 |
+
"loss": 0.0581,
|
| 2620 |
+
"step": 337
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 28.166666666666668,
|
| 2624 |
+
"grad_norm": 1.3384016752243042,
|
| 2625 |
+
"learning_rate": 1.8372816552336025e-07,
|
| 2626 |
+
"loss": 0.0613,
|
| 2627 |
+
"step": 338
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 28.25,
|
| 2631 |
+
"grad_norm": 2.645017147064209,
|
| 2632 |
+
"learning_rate": 1.6745092436045495e-07,
|
| 2633 |
+
"loss": 0.0683,
|
| 2634 |
+
"step": 339
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 28.333333333333332,
|
| 2638 |
+
"grad_norm": 1.1161837577819824,
|
| 2639 |
+
"learning_rate": 1.519224698779198e-07,
|
| 2640 |
+
"loss": 0.0493,
|
| 2641 |
+
"step": 340
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 28.416666666666668,
|
| 2645 |
+
"grad_norm": 0.978988528251648,
|
| 2646 |
+
"learning_rate": 1.3714398462768563e-07,
|
| 2647 |
+
"loss": 0.0422,
|
| 2648 |
+
"step": 341
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 28.5,
|
| 2652 |
+
"grad_norm": 2.342298746109009,
|
| 2653 |
+
"learning_rate": 1.231165940486234e-07,
|
| 2654 |
+
"loss": 0.0749,
|
| 2655 |
+
"step": 342
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 28.583333333333332,
|
| 2659 |
+
"grad_norm": 1.992661476135254,
|
| 2660 |
+
"learning_rate": 1.0984136638083176e-07,
|
| 2661 |
+
"loss": 0.0511,
|
| 2662 |
+
"step": 343
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 28.666666666666668,
|
| 2666 |
+
"grad_norm": 0.9711017608642578,
|
| 2667 |
+
"learning_rate": 9.731931258429638e-08,
|
| 2668 |
+
"loss": 0.0392,
|
| 2669 |
+
"step": 344
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 28.75,
|
| 2673 |
+
"grad_norm": 1.7103900909423828,
|
| 2674 |
+
"learning_rate": 8.555138626189619e-08,
|
| 2675 |
+
"loss": 0.0619,
|
| 2676 |
+
"step": 345
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 28.833333333333332,
|
| 2680 |
+
"grad_norm": 3.338014841079712,
|
| 2681 |
+
"learning_rate": 7.453848358678018e-08,
|
| 2682 |
+
"loss": 0.1132,
|
| 2683 |
+
"step": 346
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 28.916666666666668,
|
| 2687 |
+
"grad_norm": 2.25351619720459,
|
| 2688 |
+
"learning_rate": 6.428144323412544e-08,
|
| 2689 |
+
"loss": 0.0653,
|
| 2690 |
+
"step": 347
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 29.0,
|
| 2694 |
+
"grad_norm": 2.362285614013672,
|
| 2695 |
+
"learning_rate": 5.4781046317267103e-08,
|
| 2696 |
+
"loss": 0.082,
|
| 2697 |
+
"step": 348
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 29.0,
|
| 2701 |
+
"eval_f1": 0.8624653833164471,
|
| 2702 |
+
"eval_loss": 0.4657937288284302,
|
| 2703 |
+
"eval_runtime": 0.1112,
|
| 2704 |
+
"eval_samples_per_second": 1960.342,
|
| 2705 |
+
"eval_steps_per_second": 35.97,
|
| 2706 |
+
"step": 348
|
| 2707 |
+
}
|
| 2708 |
+
],
|
| 2709 |
+
"logging_steps": 1,
|
| 2710 |
+
"max_steps": 360,
|
| 2711 |
+
"num_input_tokens_seen": 0,
|
| 2712 |
+
"num_train_epochs": 30,
|
| 2713 |
+
"save_steps": 500,
|
| 2714 |
+
"stateful_callbacks": {
|
| 2715 |
+
"TrainerControl": {
|
| 2716 |
+
"args": {
|
| 2717 |
+
"should_epoch_stop": false,
|
| 2718 |
+
"should_evaluate": false,
|
| 2719 |
+
"should_log": false,
|
| 2720 |
+
"should_save": true,
|
| 2721 |
+
"should_training_stop": false
|
| 2722 |
+
},
|
| 2723 |
+
"attributes": {}
|
| 2724 |
+
}
|
| 2725 |
+
},
|
| 2726 |
+
"total_flos": 706091924575560.0,
|
| 2727 |
+
"train_batch_size": 128,
|
| 2728 |
+
"trial_name": null,
|
| 2729 |
+
"trial_params": null
|
| 2730 |
+
}
|
last-checkpoint/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1877de0b03c5a8d1c62291ce9d6583d58142e218d2cd950523ecb3ac34374b7c
|
| 3 |
+
size 5240
|
last-checkpoint/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|