Training in progress, epoch 1
Browse files- model.safetensors +1 -1
- run-2/checkpoint-1359/config.json +36 -0
- run-2/checkpoint-1359/model.safetensors +3 -0
- run-2/checkpoint-1359/optimizer.pt +3 -0
- run-2/checkpoint-1359/rng_state.pth +3 -0
- run-2/checkpoint-1359/scaler.pt +3 -0
- run-2/checkpoint-1359/scheduler.pt +3 -0
- run-2/checkpoint-1359/special_tokens_map.json +37 -0
- run-2/checkpoint-1359/tokenizer.json +0 -0
- run-2/checkpoint-1359/tokenizer_config.json +63 -0
- run-2/checkpoint-1359/trainer_state.json +1013 -0
- run-2/checkpoint-1359/training_args.bin +3 -0
- run-2/checkpoint-1359/vocab.txt +0 -0
- training_args.bin +1 -1
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 437961724
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17a9d1a7c0af34b2662d912bca96b4323fc85be50bfcd854f08488170d35f605
|
| 3 |
size 437961724
|
run-2/checkpoint-1359/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "LABEL_0",
|
| 14 |
+
"1": "LABEL_1",
|
| 15 |
+
"2": "LABEL_2"
|
| 16 |
+
},
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 3072,
|
| 19 |
+
"label2id": {
|
| 20 |
+
"LABEL_0": 0,
|
| 21 |
+
"LABEL_1": 1,
|
| 22 |
+
"LABEL_2": 2
|
| 23 |
+
},
|
| 24 |
+
"layer_norm_eps": 1e-12,
|
| 25 |
+
"max_position_embeddings": 512,
|
| 26 |
+
"model_type": "bert",
|
| 27 |
+
"num_attention_heads": 12,
|
| 28 |
+
"num_hidden_layers": 12,
|
| 29 |
+
"pad_token_id": 0,
|
| 30 |
+
"position_embedding_type": "absolute",
|
| 31 |
+
"problem_type": "single_label_classification",
|
| 32 |
+
"transformers_version": "4.57.1",
|
| 33 |
+
"type_vocab_size": 2,
|
| 34 |
+
"use_cache": true,
|
| 35 |
+
"vocab_size": 30522
|
| 36 |
+
}
|
run-2/checkpoint-1359/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17a9d1a7c0af34b2662d912bca96b4323fc85be50bfcd854f08488170d35f605
|
| 3 |
+
size 437961724
|
run-2/checkpoint-1359/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d46c33389283e208e1decfa83b3515a6379f814c2ff81822e1e0b958a0c8ec9b
|
| 3 |
+
size 876047755
|
run-2/checkpoint-1359/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3bf99a97e55e37df1321c689e62a0643f71031f8b594e08a09e50b16de6a73e3
|
| 3 |
+
size 14709
|
run-2/checkpoint-1359/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be3415804e0baef86602afc2f77976eabebf5e078b8f5f63b855d510c163c36b
|
| 3 |
+
size 1383
|
run-2/checkpoint-1359/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8e0bbbf384b81bbdda3c52b04df0b53dfa70b7e88d22ea4d6fa3aa455f7a45d
|
| 3 |
+
size 1465
|
run-2/checkpoint-1359/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 |
+
}
|
run-2/checkpoint-1359/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
run-2/checkpoint-1359/tokenizer_config.json
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"100": {
|
| 12 |
+
"content": "[UNK]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"101": {
|
| 20 |
+
"content": "[CLS]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"102": {
|
| 28 |
+
"content": "[SEP]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"103": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"clean_up_tokenization_spaces": false,
|
| 45 |
+
"cls_token": "[CLS]",
|
| 46 |
+
"do_lower_case": true,
|
| 47 |
+
"extra_special_tokens": {},
|
| 48 |
+
"mask_token": "[MASK]",
|
| 49 |
+
"max_length": 256,
|
| 50 |
+
"model_max_length": 512,
|
| 51 |
+
"pad_to_multiple_of": null,
|
| 52 |
+
"pad_token": "[PAD]",
|
| 53 |
+
"pad_token_type_id": 0,
|
| 54 |
+
"padding_side": "right",
|
| 55 |
+
"sep_token": "[SEP]",
|
| 56 |
+
"stride": 0,
|
| 57 |
+
"strip_accents": null,
|
| 58 |
+
"tokenize_chinese_chars": true,
|
| 59 |
+
"tokenizer_class": "BertTokenizer",
|
| 60 |
+
"truncation_side": "right",
|
| 61 |
+
"truncation_strategy": "longest_first",
|
| 62 |
+
"unk_token": "[UNK]"
|
| 63 |
+
}
|
run-2/checkpoint-1359/trainer_state.json
ADDED
|
@@ -0,0 +1,1013 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1359,
|
| 3 |
+
"best_metric": 0.7880794701986755,
|
| 4 |
+
"best_model_checkpoint": "bert-finetuned-sentiment/run-2/checkpoint-1359",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1359,
|
| 8 |
+
"is_hyper_param_search": true,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.007358351729212656,
|
| 14 |
+
"grad_norm": 15.73861026763916,
|
| 15 |
+
"learning_rate": 7.661248088138319e-07,
|
| 16 |
+
"loss": 0.1318,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.014716703458425313,
|
| 21 |
+
"grad_norm": 0.6841095685958862,
|
| 22 |
+
"learning_rate": 1.6173745963847562e-06,
|
| 23 |
+
"loss": 0.2228,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.02207505518763797,
|
| 28 |
+
"grad_norm": 20.415630340576172,
|
| 29 |
+
"learning_rate": 2.46862438395568e-06,
|
| 30 |
+
"loss": 0.2499,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.029433406916850625,
|
| 35 |
+
"grad_norm": 2.7358345985412598,
|
| 36 |
+
"learning_rate": 3.319874171526605e-06,
|
| 37 |
+
"loss": 0.2024,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03679175864606328,
|
| 42 |
+
"grad_norm": 6.356612205505371,
|
| 43 |
+
"learning_rate": 4.171123959097529e-06,
|
| 44 |
+
"loss": 0.1746,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04415011037527594,
|
| 49 |
+
"grad_norm": 0.4042581021785736,
|
| 50 |
+
"learning_rate": 5.022373746668453e-06,
|
| 51 |
+
"loss": 0.1,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.051508462104488596,
|
| 56 |
+
"grad_norm": 1.1148451566696167,
|
| 57 |
+
"learning_rate": 5.8736235342393774e-06,
|
| 58 |
+
"loss": 0.1023,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.05886681383370125,
|
| 63 |
+
"grad_norm": 39.592811584472656,
|
| 64 |
+
"learning_rate": 6.724873321810302e-06,
|
| 65 |
+
"loss": 0.0813,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.06622516556291391,
|
| 70 |
+
"grad_norm": 19.681055068969727,
|
| 71 |
+
"learning_rate": 7.576123109381227e-06,
|
| 72 |
+
"loss": 0.0931,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07358351729212656,
|
| 77 |
+
"grad_norm": 10.252055168151855,
|
| 78 |
+
"learning_rate": 8.42737289695215e-06,
|
| 79 |
+
"loss": 0.2333,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.08094186902133922,
|
| 84 |
+
"grad_norm": 33.362030029296875,
|
| 85 |
+
"learning_rate": 8.49814021437465e-06,
|
| 86 |
+
"loss": 0.1284,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.08830022075055188,
|
| 91 |
+
"grad_norm": 20.02435874938965,
|
| 92 |
+
"learning_rate": 8.482187257336215e-06,
|
| 93 |
+
"loss": 0.1234,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.09565857247976453,
|
| 98 |
+
"grad_norm": 66.22127532958984,
|
| 99 |
+
"learning_rate": 8.46623430029778e-06,
|
| 100 |
+
"loss": 0.0565,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.10301692420897719,
|
| 105 |
+
"grad_norm": 0.9394611120223999,
|
| 106 |
+
"learning_rate": 8.450281343259344e-06,
|
| 107 |
+
"loss": 0.0571,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.11037527593818984,
|
| 112 |
+
"grad_norm": 32.92816162109375,
|
| 113 |
+
"learning_rate": 8.434328386220908e-06,
|
| 114 |
+
"loss": 0.202,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1177336276674025,
|
| 119 |
+
"grad_norm": 30.171003341674805,
|
| 120 |
+
"learning_rate": 8.418375429182472e-06,
|
| 121 |
+
"loss": 0.2109,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.12509197939661515,
|
| 126 |
+
"grad_norm": 0.9405515193939209,
|
| 127 |
+
"learning_rate": 8.402422472144037e-06,
|
| 128 |
+
"loss": 0.1541,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.13245033112582782,
|
| 133 |
+
"grad_norm": 13.217094421386719,
|
| 134 |
+
"learning_rate": 8.386469515105601e-06,
|
| 135 |
+
"loss": 0.206,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.13980868285504047,
|
| 140 |
+
"grad_norm": 0.7528594136238098,
|
| 141 |
+
"learning_rate": 8.370516558067165e-06,
|
| 142 |
+
"loss": 0.0698,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.14716703458425312,
|
| 147 |
+
"grad_norm": 6.840025424957275,
|
| 148 |
+
"learning_rate": 8.35456360102873e-06,
|
| 149 |
+
"loss": 0.1986,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1545253863134658,
|
| 154 |
+
"grad_norm": 1.212363362312317,
|
| 155 |
+
"learning_rate": 8.338610643990296e-06,
|
| 156 |
+
"loss": 0.1699,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.16188373804267844,
|
| 161 |
+
"grad_norm": 3.122361183166504,
|
| 162 |
+
"learning_rate": 8.322657686951858e-06,
|
| 163 |
+
"loss": 0.1288,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.1692420897718911,
|
| 168 |
+
"grad_norm": 0.28292983770370483,
|
| 169 |
+
"learning_rate": 8.306704729913423e-06,
|
| 170 |
+
"loss": 0.007,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.17660044150110377,
|
| 175 |
+
"grad_norm": 0.11904435604810715,
|
| 176 |
+
"learning_rate": 8.290751772874989e-06,
|
| 177 |
+
"loss": 0.1291,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.18395879323031641,
|
| 182 |
+
"grad_norm": 62.53517532348633,
|
| 183 |
+
"learning_rate": 8.274798815836553e-06,
|
| 184 |
+
"loss": 0.0704,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.19131714495952906,
|
| 189 |
+
"grad_norm": 34.67837142944336,
|
| 190 |
+
"learning_rate": 8.258845858798116e-06,
|
| 191 |
+
"loss": 0.1849,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1986754966887417,
|
| 196 |
+
"grad_norm": 0.05347510427236557,
|
| 197 |
+
"learning_rate": 8.242892901759682e-06,
|
| 198 |
+
"loss": 0.1584,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.20603384841795438,
|
| 203 |
+
"grad_norm": 15.957563400268555,
|
| 204 |
+
"learning_rate": 8.226939944721246e-06,
|
| 205 |
+
"loss": 0.1044,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.21339220014716703,
|
| 210 |
+
"grad_norm": 0.15385572612285614,
|
| 211 |
+
"learning_rate": 8.21098698768281e-06,
|
| 212 |
+
"loss": 0.0807,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.22075055187637968,
|
| 217 |
+
"grad_norm": 0.08834321796894073,
|
| 218 |
+
"learning_rate": 8.195034030644375e-06,
|
| 219 |
+
"loss": 0.1526,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.22810890360559236,
|
| 224 |
+
"grad_norm": 5.847803115844727,
|
| 225 |
+
"learning_rate": 8.179081073605939e-06,
|
| 226 |
+
"loss": 0.2518,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.235467255334805,
|
| 231 |
+
"grad_norm": 76.60889434814453,
|
| 232 |
+
"learning_rate": 8.163128116567503e-06,
|
| 233 |
+
"loss": 0.3841,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.24282560706401765,
|
| 238 |
+
"grad_norm": 5.798557758331299,
|
| 239 |
+
"learning_rate": 8.147175159529068e-06,
|
| 240 |
+
"loss": 0.2359,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2501839587932303,
|
| 245 |
+
"grad_norm": 0.20405276119709015,
|
| 246 |
+
"learning_rate": 8.131222202490632e-06,
|
| 247 |
+
"loss": 0.1507,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.257542310522443,
|
| 252 |
+
"grad_norm": 0.35858389735221863,
|
| 253 |
+
"learning_rate": 8.115269245452196e-06,
|
| 254 |
+
"loss": 0.286,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.26490066225165565,
|
| 259 |
+
"grad_norm": 0.10192416608333588,
|
| 260 |
+
"learning_rate": 8.09931628841376e-06,
|
| 261 |
+
"loss": 0.0885,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.27225901398086827,
|
| 266 |
+
"grad_norm": 3.3536927700042725,
|
| 267 |
+
"learning_rate": 8.083363331375325e-06,
|
| 268 |
+
"loss": 0.2359,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.27961736571008095,
|
| 273 |
+
"grad_norm": 0.36128467321395874,
|
| 274 |
+
"learning_rate": 8.06741037433689e-06,
|
| 275 |
+
"loss": 0.2289,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2869757174392936,
|
| 280 |
+
"grad_norm": 35.29405975341797,
|
| 281 |
+
"learning_rate": 8.051457417298454e-06,
|
| 282 |
+
"loss": 0.1705,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.29433406916850624,
|
| 287 |
+
"grad_norm": 13.290261268615723,
|
| 288 |
+
"learning_rate": 8.035504460260018e-06,
|
| 289 |
+
"loss": 0.1391,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.3016924208977189,
|
| 294 |
+
"grad_norm": 4.086803436279297,
|
| 295 |
+
"learning_rate": 8.019551503221582e-06,
|
| 296 |
+
"loss": 0.3115,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3090507726269316,
|
| 301 |
+
"grad_norm": 85.40064239501953,
|
| 302 |
+
"learning_rate": 8.003598546183147e-06,
|
| 303 |
+
"loss": 0.1905,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3164091243561442,
|
| 308 |
+
"grad_norm": 3.496741533279419,
|
| 309 |
+
"learning_rate": 7.987645589144711e-06,
|
| 310 |
+
"loss": 0.1074,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3237674760853569,
|
| 315 |
+
"grad_norm": 32.27187728881836,
|
| 316 |
+
"learning_rate": 7.971692632106275e-06,
|
| 317 |
+
"loss": 0.1561,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.33112582781456956,
|
| 322 |
+
"grad_norm": 8.242842674255371,
|
| 323 |
+
"learning_rate": 7.95573967506784e-06,
|
| 324 |
+
"loss": 0.2492,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3384841795437822,
|
| 329 |
+
"grad_norm": 29.650043487548828,
|
| 330 |
+
"learning_rate": 7.939786718029404e-06,
|
| 331 |
+
"loss": 0.2973,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.34584253127299486,
|
| 336 |
+
"grad_norm": 79.97142791748047,
|
| 337 |
+
"learning_rate": 7.923833760990968e-06,
|
| 338 |
+
"loss": 0.1544,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.35320088300220753,
|
| 343 |
+
"grad_norm": 29.046892166137695,
|
| 344 |
+
"learning_rate": 7.907880803952533e-06,
|
| 345 |
+
"loss": 0.1099,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.36055923473142015,
|
| 350 |
+
"grad_norm": 0.17732320725917816,
|
| 351 |
+
"learning_rate": 7.891927846914099e-06,
|
| 352 |
+
"loss": 0.1003,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.36791758646063283,
|
| 357 |
+
"grad_norm": 0.7188624143600464,
|
| 358 |
+
"learning_rate": 7.875974889875661e-06,
|
| 359 |
+
"loss": 0.3034,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.37527593818984545,
|
| 364 |
+
"grad_norm": 21.04668617248535,
|
| 365 |
+
"learning_rate": 7.860021932837226e-06,
|
| 366 |
+
"loss": 0.2042,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3826342899190581,
|
| 371 |
+
"grad_norm": 5.541590690612793,
|
| 372 |
+
"learning_rate": 7.844068975798792e-06,
|
| 373 |
+
"loss": 0.2426,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.3899926416482708,
|
| 378 |
+
"grad_norm": 0.21457338333129883,
|
| 379 |
+
"learning_rate": 7.828116018760356e-06,
|
| 380 |
+
"loss": 0.2613,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.3973509933774834,
|
| 385 |
+
"grad_norm": 6.815304279327393,
|
| 386 |
+
"learning_rate": 7.812163061721919e-06,
|
| 387 |
+
"loss": 0.1203,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.4047093451066961,
|
| 392 |
+
"grad_norm": 1.4606707096099854,
|
| 393 |
+
"learning_rate": 7.796210104683485e-06,
|
| 394 |
+
"loss": 0.1307,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.41206769683590877,
|
| 399 |
+
"grad_norm": 170.10000610351562,
|
| 400 |
+
"learning_rate": 7.780257147645049e-06,
|
| 401 |
+
"loss": 0.2187,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4194260485651214,
|
| 406 |
+
"grad_norm": 52.976402282714844,
|
| 407 |
+
"learning_rate": 7.764304190606613e-06,
|
| 408 |
+
"loss": 0.1297,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.42678440029433407,
|
| 413 |
+
"grad_norm": 1.0618321895599365,
|
| 414 |
+
"learning_rate": 7.748351233568176e-06,
|
| 415 |
+
"loss": 0.1972,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.43414275202354674,
|
| 420 |
+
"grad_norm": 0.2524672746658325,
|
| 421 |
+
"learning_rate": 7.732398276529742e-06,
|
| 422 |
+
"loss": 0.2513,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.44150110375275936,
|
| 427 |
+
"grad_norm": 0.7126919627189636,
|
| 428 |
+
"learning_rate": 7.716445319491306e-06,
|
| 429 |
+
"loss": 0.2055,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.44885945548197204,
|
| 434 |
+
"grad_norm": 6.286591529846191,
|
| 435 |
+
"learning_rate": 7.70049236245287e-06,
|
| 436 |
+
"loss": 0.0236,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.4562178072111847,
|
| 441 |
+
"grad_norm": 0.11959370225667953,
|
| 442 |
+
"learning_rate": 7.684539405414435e-06,
|
| 443 |
+
"loss": 0.2016,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.46357615894039733,
|
| 448 |
+
"grad_norm": 6.085186004638672,
|
| 449 |
+
"learning_rate": 7.668586448376e-06,
|
| 450 |
+
"loss": 0.3102,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.47093451066961,
|
| 455 |
+
"grad_norm": 96.2499008178711,
|
| 456 |
+
"learning_rate": 7.652633491337564e-06,
|
| 457 |
+
"loss": 0.2181,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.4782928623988227,
|
| 462 |
+
"grad_norm": 14.058534622192383,
|
| 463 |
+
"learning_rate": 7.636680534299128e-06,
|
| 464 |
+
"loss": 0.2536,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.4856512141280353,
|
| 469 |
+
"grad_norm": 33.89338684082031,
|
| 470 |
+
"learning_rate": 7.620727577260692e-06,
|
| 471 |
+
"loss": 0.2979,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.493009565857248,
|
| 476 |
+
"grad_norm": 0.733513355255127,
|
| 477 |
+
"learning_rate": 7.604774620222257e-06,
|
| 478 |
+
"loss": 0.1619,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5003679175864606,
|
| 483 |
+
"grad_norm": 0.18061041831970215,
|
| 484 |
+
"learning_rate": 7.588821663183821e-06,
|
| 485 |
+
"loss": 0.1896,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5077262693156733,
|
| 490 |
+
"grad_norm": 0.1485278606414795,
|
| 491 |
+
"learning_rate": 7.572868706145385e-06,
|
| 492 |
+
"loss": 0.0868,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.515084621044886,
|
| 497 |
+
"grad_norm": 65.20437622070312,
|
| 498 |
+
"learning_rate": 7.55691574910695e-06,
|
| 499 |
+
"loss": 0.1969,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5224429727740986,
|
| 504 |
+
"grad_norm": 33.507415771484375,
|
| 505 |
+
"learning_rate": 7.540962792068515e-06,
|
| 506 |
+
"loss": 0.2309,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5298013245033113,
|
| 511 |
+
"grad_norm": 7.292962551116943,
|
| 512 |
+
"learning_rate": 7.525009835030078e-06,
|
| 513 |
+
"loss": 0.3215,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.5371596762325239,
|
| 518 |
+
"grad_norm": 0.141510471701622,
|
| 519 |
+
"learning_rate": 7.509056877991643e-06,
|
| 520 |
+
"loss": 0.1861,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.5445180279617365,
|
| 525 |
+
"grad_norm": 10.76659107208252,
|
| 526 |
+
"learning_rate": 7.493103920953208e-06,
|
| 527 |
+
"loss": 0.1563,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.5518763796909493,
|
| 532 |
+
"grad_norm": 0.535017192363739,
|
| 533 |
+
"learning_rate": 7.477150963914772e-06,
|
| 534 |
+
"loss": 0.1037,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.5592347314201619,
|
| 539 |
+
"grad_norm": 0.12771788239479065,
|
| 540 |
+
"learning_rate": 7.461198006876336e-06,
|
| 541 |
+
"loss": 0.0714,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.5665930831493745,
|
| 546 |
+
"grad_norm": 0.13390083611011505,
|
| 547 |
+
"learning_rate": 7.4452450498379e-06,
|
| 548 |
+
"loss": 0.2959,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.5739514348785872,
|
| 553 |
+
"grad_norm": 56.0723991394043,
|
| 554 |
+
"learning_rate": 7.429292092799465e-06,
|
| 555 |
+
"loss": 0.1132,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.5813097866077999,
|
| 560 |
+
"grad_norm": 0.137408047914505,
|
| 561 |
+
"learning_rate": 7.4133391357610295e-06,
|
| 562 |
+
"loss": 0.2051,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.5886681383370125,
|
| 567 |
+
"grad_norm": 12.899972915649414,
|
| 568 |
+
"learning_rate": 7.397386178722593e-06,
|
| 569 |
+
"loss": 0.3972,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.5960264900662252,
|
| 574 |
+
"grad_norm": 11.240960121154785,
|
| 575 |
+
"learning_rate": 7.381433221684158e-06,
|
| 576 |
+
"loss": 0.3619,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6033848417954378,
|
| 581 |
+
"grad_norm": 8.278285026550293,
|
| 582 |
+
"learning_rate": 7.3654802646457225e-06,
|
| 583 |
+
"loss": 0.1447,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6107431935246505,
|
| 588 |
+
"grad_norm": 2.9699766635894775,
|
| 589 |
+
"learning_rate": 7.349527307607287e-06,
|
| 590 |
+
"loss": 0.1212,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6181015452538632,
|
| 595 |
+
"grad_norm": 12.325569152832031,
|
| 596 |
+
"learning_rate": 7.333574350568851e-06,
|
| 597 |
+
"loss": 0.2392,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.6254598969830758,
|
| 602 |
+
"grad_norm": 1.178353190422058,
|
| 603 |
+
"learning_rate": 7.3176213935304155e-06,
|
| 604 |
+
"loss": 0.193,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.6328182487122884,
|
| 609 |
+
"grad_norm": 42.010711669921875,
|
| 610 |
+
"learning_rate": 7.30166843649198e-06,
|
| 611 |
+
"loss": 0.178,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.6401766004415012,
|
| 616 |
+
"grad_norm": 0.831745445728302,
|
| 617 |
+
"learning_rate": 7.285715479453545e-06,
|
| 618 |
+
"loss": 0.2348,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.6475349521707138,
|
| 623 |
+
"grad_norm": 68.13945007324219,
|
| 624 |
+
"learning_rate": 7.2697625224151084e-06,
|
| 625 |
+
"loss": 0.0788,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.6548933038999264,
|
| 630 |
+
"grad_norm": 33.832366943359375,
|
| 631 |
+
"learning_rate": 7.253809565376673e-06,
|
| 632 |
+
"loss": 0.1783,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.6622516556291391,
|
| 637 |
+
"grad_norm": 1.357936978340149,
|
| 638 |
+
"learning_rate": 7.237856608338238e-06,
|
| 639 |
+
"loss": 0.1938,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.6696100073583517,
|
| 644 |
+
"grad_norm": 53.46693420410156,
|
| 645 |
+
"learning_rate": 7.221903651299802e-06,
|
| 646 |
+
"loss": 0.4283,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.6769683590875644,
|
| 651 |
+
"grad_norm": 4.254789352416992,
|
| 652 |
+
"learning_rate": 7.205950694261366e-06,
|
| 653 |
+
"loss": 0.2616,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.6843267108167771,
|
| 658 |
+
"grad_norm": 75.10453796386719,
|
| 659 |
+
"learning_rate": 7.189997737222931e-06,
|
| 660 |
+
"loss": 0.1924,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.6916850625459897,
|
| 665 |
+
"grad_norm": 0.7422951459884644,
|
| 666 |
+
"learning_rate": 7.174044780184495e-06,
|
| 667 |
+
"loss": 0.13,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.6990434142752023,
|
| 672 |
+
"grad_norm": 18.9062557220459,
|
| 673 |
+
"learning_rate": 7.15809182314606e-06,
|
| 674 |
+
"loss": 0.1283,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.7064017660044151,
|
| 679 |
+
"grad_norm": 6.750290393829346,
|
| 680 |
+
"learning_rate": 7.142138866107623e-06,
|
| 681 |
+
"loss": 0.3134,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.7137601177336277,
|
| 686 |
+
"grad_norm": 16.308557510375977,
|
| 687 |
+
"learning_rate": 7.126185909069188e-06,
|
| 688 |
+
"loss": 0.2575,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.7211184694628403,
|
| 693 |
+
"grad_norm": 67.54432678222656,
|
| 694 |
+
"learning_rate": 7.110232952030753e-06,
|
| 695 |
+
"loss": 0.3646,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.7284768211920529,
|
| 700 |
+
"grad_norm": 20.565406799316406,
|
| 701 |
+
"learning_rate": 7.094279994992317e-06,
|
| 702 |
+
"loss": 0.1607,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.7358351729212657,
|
| 707 |
+
"grad_norm": 9.457584381103516,
|
| 708 |
+
"learning_rate": 7.078327037953881e-06,
|
| 709 |
+
"loss": 0.4157,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.7431935246504783,
|
| 714 |
+
"grad_norm": 1.9208470582962036,
|
| 715 |
+
"learning_rate": 7.062374080915446e-06,
|
| 716 |
+
"loss": 0.288,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.7505518763796909,
|
| 721 |
+
"grad_norm": 0.18186113238334656,
|
| 722 |
+
"learning_rate": 7.04642112387701e-06,
|
| 723 |
+
"loss": 0.1683,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.7579102281089036,
|
| 728 |
+
"grad_norm": 56.17685317993164,
|
| 729 |
+
"learning_rate": 7.030468166838575e-06,
|
| 730 |
+
"loss": 0.1789,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.7652685798381162,
|
| 735 |
+
"grad_norm": 22.048622131347656,
|
| 736 |
+
"learning_rate": 7.014515209800139e-06,
|
| 737 |
+
"loss": 0.2506,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.7726269315673289,
|
| 742 |
+
"grad_norm": 3.697582244873047,
|
| 743 |
+
"learning_rate": 6.998562252761703e-06,
|
| 744 |
+
"loss": 0.1667,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.7799852832965416,
|
| 749 |
+
"grad_norm": 6.335505962371826,
|
| 750 |
+
"learning_rate": 6.982609295723268e-06,
|
| 751 |
+
"loss": 0.1866,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.7873436350257542,
|
| 756 |
+
"grad_norm": 18.40208625793457,
|
| 757 |
+
"learning_rate": 6.9666563386848324e-06,
|
| 758 |
+
"loss": 0.2035,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.7947019867549668,
|
| 763 |
+
"grad_norm": 0.3842657506465912,
|
| 764 |
+
"learning_rate": 6.950703381646396e-06,
|
| 765 |
+
"loss": 0.3858,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.8020603384841796,
|
| 770 |
+
"grad_norm": 0.10957049578428268,
|
| 771 |
+
"learning_rate": 6.934750424607961e-06,
|
| 772 |
+
"loss": 0.1044,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.8094186902133922,
|
| 777 |
+
"grad_norm": 52.896461486816406,
|
| 778 |
+
"learning_rate": 6.9187974675695254e-06,
|
| 779 |
+
"loss": 0.0542,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.8167770419426048,
|
| 784 |
+
"grad_norm": 77.98719787597656,
|
| 785 |
+
"learning_rate": 6.90284451053109e-06,
|
| 786 |
+
"loss": 0.2078,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.8241353936718175,
|
| 791 |
+
"grad_norm": 103.59429931640625,
|
| 792 |
+
"learning_rate": 6.886891553492654e-06,
|
| 793 |
+
"loss": 0.1757,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.8314937454010302,
|
| 798 |
+
"grad_norm": 14.622530937194824,
|
| 799 |
+
"learning_rate": 6.8709385964542184e-06,
|
| 800 |
+
"loss": 0.1741,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.8388520971302428,
|
| 805 |
+
"grad_norm": 0.8220998048782349,
|
| 806 |
+
"learning_rate": 6.854985639415783e-06,
|
| 807 |
+
"loss": 0.1864,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.8462104488594555,
|
| 812 |
+
"grad_norm": 44.293025970458984,
|
| 813 |
+
"learning_rate": 6.839032682377347e-06,
|
| 814 |
+
"loss": 0.3249,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.8535688005886681,
|
| 819 |
+
"grad_norm": 45.35190200805664,
|
| 820 |
+
"learning_rate": 6.823079725338911e-06,
|
| 821 |
+
"loss": 0.2564,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.8609271523178808,
|
| 826 |
+
"grad_norm": 92.81620025634766,
|
| 827 |
+
"learning_rate": 6.807126768300476e-06,
|
| 828 |
+
"loss": 0.1886,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.8682855040470935,
|
| 833 |
+
"grad_norm": 110.62460327148438,
|
| 834 |
+
"learning_rate": 6.79117381126204e-06,
|
| 835 |
+
"loss": 0.2484,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.8756438557763061,
|
| 840 |
+
"grad_norm": 25.606109619140625,
|
| 841 |
+
"learning_rate": 6.775220854223605e-06,
|
| 842 |
+
"loss": 0.1544,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.8830022075055187,
|
| 847 |
+
"grad_norm": 0.10603518784046173,
|
| 848 |
+
"learning_rate": 6.759267897185169e-06,
|
| 849 |
+
"loss": 0.3421,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.8903605592347315,
|
| 854 |
+
"grad_norm": 0.8049039840698242,
|
| 855 |
+
"learning_rate": 6.743314940146733e-06,
|
| 856 |
+
"loss": 0.3171,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.8977189109639441,
|
| 861 |
+
"grad_norm": 0.1788051277399063,
|
| 862 |
+
"learning_rate": 6.727361983108298e-06,
|
| 863 |
+
"loss": 0.2324,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.9050772626931567,
|
| 868 |
+
"grad_norm": 4.411452293395996,
|
| 869 |
+
"learning_rate": 6.711409026069863e-06,
|
| 870 |
+
"loss": 0.0503,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.9124356144223694,
|
| 875 |
+
"grad_norm": 0.08730533719062805,
|
| 876 |
+
"learning_rate": 6.695456069031426e-06,
|
| 877 |
+
"loss": 0.1698,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.919793966151582,
|
| 882 |
+
"grad_norm": 14.45964527130127,
|
| 883 |
+
"learning_rate": 6.679503111992991e-06,
|
| 884 |
+
"loss": 0.1356,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.9271523178807947,
|
| 889 |
+
"grad_norm": 99.87224578857422,
|
| 890 |
+
"learning_rate": 6.663550154954556e-06,
|
| 891 |
+
"loss": 0.0962,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.9345106696100074,
|
| 896 |
+
"grad_norm": 0.4108155369758606,
|
| 897 |
+
"learning_rate": 6.64759719791612e-06,
|
| 898 |
+
"loss": 0.1787,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.94186902133922,
|
| 903 |
+
"grad_norm": 108.73641204833984,
|
| 904 |
+
"learning_rate": 6.631644240877684e-06,
|
| 905 |
+
"loss": 0.279,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.9492273730684326,
|
| 910 |
+
"grad_norm": 131.88836669921875,
|
| 911 |
+
"learning_rate": 6.6156912838392486e-06,
|
| 912 |
+
"loss": 0.1746,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.9565857247976454,
|
| 917 |
+
"grad_norm": 0.547515869140625,
|
| 918 |
+
"learning_rate": 6.599738326800813e-06,
|
| 919 |
+
"loss": 0.0208,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.963944076526858,
|
| 924 |
+
"grad_norm": 19.8458251953125,
|
| 925 |
+
"learning_rate": 6.583785369762378e-06,
|
| 926 |
+
"loss": 0.1389,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.9713024282560706,
|
| 931 |
+
"grad_norm": 0.5687592625617981,
|
| 932 |
+
"learning_rate": 6.5678324127239416e-06,
|
| 933 |
+
"loss": 0.1856,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.9786607799852833,
|
| 938 |
+
"grad_norm": 0.06650309264659882,
|
| 939 |
+
"learning_rate": 6.551879455685506e-06,
|
| 940 |
+
"loss": 0.2146,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.986019131714496,
|
| 945 |
+
"grad_norm": 38.21836471557617,
|
| 946 |
+
"learning_rate": 6.53592649864707e-06,
|
| 947 |
+
"loss": 0.2825,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.9933774834437086,
|
| 952 |
+
"grad_norm": 2.0989203453063965,
|
| 953 |
+
"learning_rate": 6.519973541608635e-06,
|
| 954 |
+
"loss": 0.341,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.0,
|
| 959 |
+
"eval_accuracy": 0.7880794701986755,
|
| 960 |
+
"eval_confusion_matrix": [
|
| 961 |
+
[
|
| 962 |
+
676,
|
| 963 |
+
8,
|
| 964 |
+
78
|
| 965 |
+
],
|
| 966 |
+
[
|
| 967 |
+
7,
|
| 968 |
+
70,
|
| 969 |
+
34
|
| 970 |
+
],
|
| 971 |
+
[
|
| 972 |
+
110,
|
| 973 |
+
19,
|
| 974 |
+
206
|
| 975 |
+
]
|
| 976 |
+
],
|
| 977 |
+
"eval_f1": 0.7852632048062911,
|
| 978 |
+
"eval_loss": 1.1113489866256714,
|
| 979 |
+
"eval_precision": 0.7836833128242224,
|
| 980 |
+
"eval_recall": 0.7880794701986755,
|
| 981 |
+
"eval_runtime": 10.673,
|
| 982 |
+
"eval_samples_per_second": 113.183,
|
| 983 |
+
"eval_steps_per_second": 3.56,
|
| 984 |
+
"step": 1359
|
| 985 |
+
}
|
| 986 |
+
],
|
| 987 |
+
"logging_steps": 10,
|
| 988 |
+
"max_steps": 5436,
|
| 989 |
+
"num_input_tokens_seen": 0,
|
| 990 |
+
"num_train_epochs": 4,
|
| 991 |
+
"save_steps": 500,
|
| 992 |
+
"stateful_callbacks": {
|
| 993 |
+
"TrainerControl": {
|
| 994 |
+
"args": {
|
| 995 |
+
"should_epoch_stop": false,
|
| 996 |
+
"should_evaluate": false,
|
| 997 |
+
"should_log": false,
|
| 998 |
+
"should_save": true,
|
| 999 |
+
"should_training_stop": false
|
| 1000 |
+
},
|
| 1001 |
+
"attributes": {}
|
| 1002 |
+
}
|
| 1003 |
+
},
|
| 1004 |
+
"total_flos": 1158870714941520.0,
|
| 1005 |
+
"train_batch_size": 8,
|
| 1006 |
+
"trial_name": null,
|
| 1007 |
+
"trial_params": {
|
| 1008 |
+
"learning_rate": 8.512497875709243e-06,
|
| 1009 |
+
"num_train_epochs": 4,
|
| 1010 |
+
"per_device_train_batch_size": 4,
|
| 1011 |
+
"seed": 14
|
| 1012 |
+
}
|
| 1013 |
+
}
|
run-2/checkpoint-1359/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1ed213c84afc0c72cb8266b9163d38ffb171814d89789d606f80a887cc280d4
|
| 3 |
+
size 5777
|
run-2/checkpoint-1359/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1ed213c84afc0c72cb8266b9163d38ffb171814d89789d606f80a887cc280d4
|
| 3 |
size 5777
|