Upload 5 files
Browse files- config (3).json +43 -0
- rng_state (1).pth +3 -0
- scheduler (1).pt +3 -0
- trainer_state (1).json +916 -0
- training_args (1).bin +3 -0
config (3).json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "bert-base-cased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 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 |
+
"3": "LABEL_3",
|
| 17 |
+
"4": "LABEL_4",
|
| 18 |
+
"5": "LABEL_5"
|
| 19 |
+
},
|
| 20 |
+
"initializer_range": 0.02,
|
| 21 |
+
"intermediate_size": 3072,
|
| 22 |
+
"label2id": {
|
| 23 |
+
"LABEL_0": 0,
|
| 24 |
+
"LABEL_1": 1,
|
| 25 |
+
"LABEL_2": 2,
|
| 26 |
+
"LABEL_3": 3,
|
| 27 |
+
"LABEL_4": 4,
|
| 28 |
+
"LABEL_5": 5
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-12,
|
| 31 |
+
"max_position_embeddings": 512,
|
| 32 |
+
"model_type": "bert",
|
| 33 |
+
"num_attention_heads": 12,
|
| 34 |
+
"num_hidden_layers": 12,
|
| 35 |
+
"pad_token_id": 0,
|
| 36 |
+
"position_embedding_type": "absolute",
|
| 37 |
+
"problem_type": "single_label_classification",
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.28.1",
|
| 40 |
+
"type_vocab_size": 2,
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 28996
|
| 43 |
+
}
|
rng_state (1).pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f59a86174567743dbba554ef7f6638075a8093746de00056da4244f5adbf765
|
| 3 |
+
size 13553
|
scheduler (1).pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8347f1c7314f6c68f76795f6c7f833d38efe6feff47ff93669dbeda77645d2e3
|
| 3 |
+
size 627
|
trainer_state (1).json
ADDED
|
@@ -0,0 +1,916 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.2820078962210942,
|
| 5 |
+
"global_step": 1500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 13 |
+
"loss": 1.7773,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.0,
|
| 18 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 19 |
+
"loss": 1.7255,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 3e-06,
|
| 25 |
+
"loss": 1.5729,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4.000000000000001e-06,
|
| 31 |
+
"loss": 1.3541,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 5e-06,
|
| 37 |
+
"loss": 1.0505,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 6e-06,
|
| 43 |
+
"loss": 0.7144,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01,
|
| 48 |
+
"learning_rate": 7.000000000000001e-06,
|
| 49 |
+
"loss": 0.3478,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.02,
|
| 54 |
+
"learning_rate": 8.000000000000001e-06,
|
| 55 |
+
"loss": 0.1306,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
+
"learning_rate": 9e-06,
|
| 61 |
+
"loss": 0.0424,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
+
"learning_rate": 1e-05,
|
| 67 |
+
"loss": 0.0169,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.02,
|
| 72 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 73 |
+
"loss": 0.0355,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.02,
|
| 78 |
+
"learning_rate": 1.2e-05,
|
| 79 |
+
"loss": 0.0346,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02,
|
| 84 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 85 |
+
"loss": 0.0904,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03,
|
| 90 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 91 |
+
"loss": 0.0351,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.03,
|
| 96 |
+
"learning_rate": 1.5e-05,
|
| 97 |
+
"loss": 0.0337,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 103 |
+
"loss": 0.033,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.03,
|
| 108 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 109 |
+
"loss": 0.0901,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.03,
|
| 114 |
+
"learning_rate": 1.8e-05,
|
| 115 |
+
"loss": 0.0081,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.04,
|
| 120 |
+
"learning_rate": 1.9e-05,
|
| 121 |
+
"loss": 0.0335,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.04,
|
| 126 |
+
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 0.0036,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04,
|
| 132 |
+
"learning_rate": 2.1e-05,
|
| 133 |
+
"loss": 0.0652,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.04,
|
| 138 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 139 |
+
"loss": 0.0346,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.04,
|
| 144 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 145 |
+
"loss": 0.0891,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.05,
|
| 150 |
+
"learning_rate": 2.4e-05,
|
| 151 |
+
"loss": 0.0925,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.05,
|
| 156 |
+
"learning_rate": 2.5e-05,
|
| 157 |
+
"loss": 0.1139,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.05,
|
| 162 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 163 |
+
"loss": 0.0577,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.05,
|
| 168 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 169 |
+
"loss": 0.0073,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05,
|
| 174 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 175 |
+
"loss": 0.0298,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.05,
|
| 180 |
+
"learning_rate": 2.9e-05,
|
| 181 |
+
"loss": 0.0357,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.06,
|
| 186 |
+
"learning_rate": 3e-05,
|
| 187 |
+
"loss": 0.1385,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.06,
|
| 192 |
+
"learning_rate": 3.1e-05,
|
| 193 |
+
"loss": 0.0063,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.06,
|
| 198 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 199 |
+
"loss": 0.002,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.06,
|
| 204 |
+
"learning_rate": 3.3e-05,
|
| 205 |
+
"loss": 0.0694,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.06,
|
| 210 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 211 |
+
"loss": 0.0315,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.07,
|
| 216 |
+
"learning_rate": 3.5e-05,
|
| 217 |
+
"loss": 0.0035,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.07,
|
| 222 |
+
"learning_rate": 3.6e-05,
|
| 223 |
+
"loss": 0.0327,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.07,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 0.0324,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.07,
|
| 234 |
+
"learning_rate": 3.8e-05,
|
| 235 |
+
"loss": 0.0025,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.07,
|
| 240 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 241 |
+
"loss": 0.0014,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.08,
|
| 246 |
+
"learning_rate": 4e-05,
|
| 247 |
+
"loss": 0.074,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.08,
|
| 252 |
+
"learning_rate": 4.1e-05,
|
| 253 |
+
"loss": 0.0029,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.08,
|
| 258 |
+
"learning_rate": 4.2e-05,
|
| 259 |
+
"loss": 0.0335,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.08,
|
| 264 |
+
"learning_rate": 4.3e-05,
|
| 265 |
+
"loss": 0.0599,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.08,
|
| 270 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 271 |
+
"loss": 0.0321,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.08,
|
| 276 |
+
"learning_rate": 4.5e-05,
|
| 277 |
+
"loss": 0.0862,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.09,
|
| 282 |
+
"learning_rate": 4.600000000000001e-05,
|
| 283 |
+
"loss": 0.1408,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.09,
|
| 288 |
+
"learning_rate": 4.7e-05,
|
| 289 |
+
"loss": 0.0924,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.09,
|
| 294 |
+
"learning_rate": 4.8e-05,
|
| 295 |
+
"loss": 0.0052,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.09,
|
| 300 |
+
"learning_rate": 4.9e-05,
|
| 301 |
+
"loss": 0.0319,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.09,
|
| 306 |
+
"learning_rate": 5e-05,
|
| 307 |
+
"loss": 0.0654,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.1,
|
| 312 |
+
"learning_rate": 4.995068060761492e-05,
|
| 313 |
+
"loss": 0.0847,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.1,
|
| 318 |
+
"learning_rate": 4.990136121522983e-05,
|
| 319 |
+
"loss": 0.0073,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.1,
|
| 324 |
+
"learning_rate": 4.9852041822844744e-05,
|
| 325 |
+
"loss": 0.0351,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.1,
|
| 330 |
+
"learning_rate": 4.980272243045966e-05,
|
| 331 |
+
"loss": 0.123,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.1,
|
| 336 |
+
"learning_rate": 4.975340303807458e-05,
|
| 337 |
+
"loss": 0.0059,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.11,
|
| 342 |
+
"learning_rate": 4.9704083645689486e-05,
|
| 343 |
+
"loss": 0.0299,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.11,
|
| 348 |
+
"learning_rate": 4.96547642533044e-05,
|
| 349 |
+
"loss": 0.0552,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.11,
|
| 354 |
+
"learning_rate": 4.960544486091931e-05,
|
| 355 |
+
"loss": 0.0616,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.11,
|
| 360 |
+
"learning_rate": 4.955612546853423e-05,
|
| 361 |
+
"loss": 0.0049,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.11,
|
| 366 |
+
"learning_rate": 4.9506806076149144e-05,
|
| 367 |
+
"loss": 0.003,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.11,
|
| 372 |
+
"learning_rate": 4.945748668376406e-05,
|
| 373 |
+
"loss": 0.0688,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.12,
|
| 378 |
+
"learning_rate": 4.940816729137897e-05,
|
| 379 |
+
"loss": 0.1185,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.12,
|
| 384 |
+
"learning_rate": 4.9358847898993886e-05,
|
| 385 |
+
"loss": 0.007,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.12,
|
| 390 |
+
"learning_rate": 4.9309528506608796e-05,
|
| 391 |
+
"loss": 0.058,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.12,
|
| 396 |
+
"learning_rate": 4.926020911422372e-05,
|
| 397 |
+
"loss": 0.0864,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.12,
|
| 402 |
+
"learning_rate": 4.921088972183863e-05,
|
| 403 |
+
"loss": 0.0788,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.13,
|
| 408 |
+
"learning_rate": 4.9161570329453544e-05,
|
| 409 |
+
"loss": 0.0614,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.13,
|
| 414 |
+
"learning_rate": 4.911225093706846e-05,
|
| 415 |
+
"loss": 0.0632,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.13,
|
| 420 |
+
"learning_rate": 4.906293154468337e-05,
|
| 421 |
+
"loss": 0.0834,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.13,
|
| 426 |
+
"learning_rate": 4.9013612152298286e-05,
|
| 427 |
+
"loss": 0.0792,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.13,
|
| 432 |
+
"learning_rate": 4.89642927599132e-05,
|
| 433 |
+
"loss": 0.0096,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.14,
|
| 438 |
+
"learning_rate": 4.891497336752812e-05,
|
| 439 |
+
"loss": 0.0601,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.14,
|
| 444 |
+
"learning_rate": 4.886565397514303e-05,
|
| 445 |
+
"loss": 0.0041,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.14,
|
| 450 |
+
"learning_rate": 4.8816334582757945e-05,
|
| 451 |
+
"loss": 0.0025,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.14,
|
| 456 |
+
"learning_rate": 4.8767015190372854e-05,
|
| 457 |
+
"loss": 0.0326,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.14,
|
| 462 |
+
"learning_rate": 4.871769579798777e-05,
|
| 463 |
+
"loss": 0.0017,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.14,
|
| 468 |
+
"learning_rate": 4.8668376405602686e-05,
|
| 469 |
+
"loss": 0.0015,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.15,
|
| 474 |
+
"learning_rate": 4.86190570132176e-05,
|
| 475 |
+
"loss": 0.0913,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.15,
|
| 480 |
+
"learning_rate": 4.856973762083251e-05,
|
| 481 |
+
"loss": 0.1069,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.15,
|
| 486 |
+
"learning_rate": 4.852041822844743e-05,
|
| 487 |
+
"loss": 0.0625,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.15,
|
| 492 |
+
"learning_rate": 4.847109883606234e-05,
|
| 493 |
+
"loss": 0.0062,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.15,
|
| 498 |
+
"learning_rate": 4.8421779443677254e-05,
|
| 499 |
+
"loss": 0.061,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.16,
|
| 504 |
+
"learning_rate": 4.837246005129217e-05,
|
| 505 |
+
"loss": 0.0311,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.16,
|
| 510 |
+
"learning_rate": 4.8323140658907086e-05,
|
| 511 |
+
"loss": 0.0339,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.16,
|
| 516 |
+
"learning_rate": 4.8273821266522e-05,
|
| 517 |
+
"loss": 0.0028,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.16,
|
| 522 |
+
"learning_rate": 4.822450187413691e-05,
|
| 523 |
+
"loss": 0.0588,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.16,
|
| 528 |
+
"learning_rate": 4.817518248175183e-05,
|
| 529 |
+
"loss": 0.0638,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.17,
|
| 534 |
+
"learning_rate": 4.812586308936674e-05,
|
| 535 |
+
"loss": 0.0035,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.17,
|
| 540 |
+
"learning_rate": 4.8076543696981654e-05,
|
| 541 |
+
"loss": 0.0297,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.17,
|
| 546 |
+
"learning_rate": 4.802722430459657e-05,
|
| 547 |
+
"loss": 0.0028,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.17,
|
| 552 |
+
"learning_rate": 4.7977904912211487e-05,
|
| 553 |
+
"loss": 0.0307,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.17,
|
| 558 |
+
"learning_rate": 4.7928585519826396e-05,
|
| 559 |
+
"loss": 0.0638,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.17,
|
| 564 |
+
"learning_rate": 4.787926612744131e-05,
|
| 565 |
+
"loss": 0.0292,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.18,
|
| 570 |
+
"learning_rate": 4.782994673505622e-05,
|
| 571 |
+
"loss": 0.0308,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.18,
|
| 576 |
+
"learning_rate": 4.7780627342671145e-05,
|
| 577 |
+
"loss": 0.0305,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.18,
|
| 582 |
+
"learning_rate": 4.7731307950286054e-05,
|
| 583 |
+
"loss": 0.0589,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.18,
|
| 588 |
+
"learning_rate": 4.768198855790097e-05,
|
| 589 |
+
"loss": 0.0821,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.18,
|
| 594 |
+
"learning_rate": 4.763266916551588e-05,
|
| 595 |
+
"loss": 0.032,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.19,
|
| 600 |
+
"learning_rate": 4.7583349773130796e-05,
|
| 601 |
+
"loss": 0.0051,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.19,
|
| 606 |
+
"learning_rate": 4.7534030380745706e-05,
|
| 607 |
+
"loss": 0.0464,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.19,
|
| 612 |
+
"learning_rate": 4.748471098836063e-05,
|
| 613 |
+
"loss": 0.1152,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.19,
|
| 618 |
+
"learning_rate": 4.7435391595975545e-05,
|
| 619 |
+
"loss": 0.0045,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.19,
|
| 624 |
+
"learning_rate": 4.7386072203590454e-05,
|
| 625 |
+
"loss": 0.0552,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.2,
|
| 630 |
+
"learning_rate": 4.733675281120537e-05,
|
| 631 |
+
"loss": 0.0593,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.2,
|
| 636 |
+
"learning_rate": 4.728743341882028e-05,
|
| 637 |
+
"loss": 0.0318,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.2,
|
| 642 |
+
"learning_rate": 4.7238114026435196e-05,
|
| 643 |
+
"loss": 0.0297,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.2,
|
| 648 |
+
"learning_rate": 4.718879463405011e-05,
|
| 649 |
+
"loss": 0.0424,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.2,
|
| 654 |
+
"learning_rate": 4.713947524166503e-05,
|
| 655 |
+
"loss": 0.0565,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.2,
|
| 660 |
+
"learning_rate": 4.709015584927994e-05,
|
| 661 |
+
"loss": 0.0029,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.21,
|
| 666 |
+
"learning_rate": 4.7040836456894854e-05,
|
| 667 |
+
"loss": 0.0593,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.21,
|
| 672 |
+
"learning_rate": 4.6991517064509764e-05,
|
| 673 |
+
"loss": 0.0327,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.21,
|
| 678 |
+
"learning_rate": 4.694219767212468e-05,
|
| 679 |
+
"loss": 0.03,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.21,
|
| 684 |
+
"learning_rate": 4.6892878279739596e-05,
|
| 685 |
+
"loss": 0.0304,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.21,
|
| 690 |
+
"learning_rate": 4.684355888735451e-05,
|
| 691 |
+
"loss": 0.0619,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.22,
|
| 696 |
+
"learning_rate": 4.679423949496942e-05,
|
| 697 |
+
"loss": 0.0385,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.22,
|
| 702 |
+
"learning_rate": 4.674492010258434e-05,
|
| 703 |
+
"loss": 0.0313,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.22,
|
| 708 |
+
"learning_rate": 4.669560071019925e-05,
|
| 709 |
+
"loss": 0.0281,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.22,
|
| 714 |
+
"learning_rate": 4.6646281317814164e-05,
|
| 715 |
+
"loss": 0.0304,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.22,
|
| 720 |
+
"learning_rate": 4.659696192542909e-05,
|
| 721 |
+
"loss": 0.0319,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.23,
|
| 726 |
+
"learning_rate": 4.6547642533043996e-05,
|
| 727 |
+
"loss": 0.0021,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.23,
|
| 732 |
+
"learning_rate": 4.649832314065891e-05,
|
| 733 |
+
"loss": 0.0018,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.23,
|
| 738 |
+
"learning_rate": 4.644900374827382e-05,
|
| 739 |
+
"loss": 0.0625,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.23,
|
| 744 |
+
"learning_rate": 4.639968435588874e-05,
|
| 745 |
+
"loss": 0.0023,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.23,
|
| 750 |
+
"learning_rate": 4.635036496350365e-05,
|
| 751 |
+
"loss": 0.0303,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.24,
|
| 756 |
+
"learning_rate": 4.630104557111857e-05,
|
| 757 |
+
"loss": 0.0886,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.24,
|
| 762 |
+
"learning_rate": 4.625172617873348e-05,
|
| 763 |
+
"loss": 0.0029,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.24,
|
| 768 |
+
"learning_rate": 4.6202406786348396e-05,
|
| 769 |
+
"loss": 0.0836,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.24,
|
| 774 |
+
"learning_rate": 4.6153087393963306e-05,
|
| 775 |
+
"loss": 0.0289,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.24,
|
| 780 |
+
"learning_rate": 4.610376800157822e-05,
|
| 781 |
+
"loss": 0.0665,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.24,
|
| 786 |
+
"learning_rate": 4.605444860919313e-05,
|
| 787 |
+
"loss": 0.0053,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.25,
|
| 792 |
+
"learning_rate": 4.6005129216808054e-05,
|
| 793 |
+
"loss": 0.0335,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.25,
|
| 798 |
+
"learning_rate": 4.5955809824422964e-05,
|
| 799 |
+
"loss": 0.0027,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.25,
|
| 804 |
+
"learning_rate": 4.590649043203788e-05,
|
| 805 |
+
"loss": 0.0021,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.25,
|
| 810 |
+
"learning_rate": 4.585717103965279e-05,
|
| 811 |
+
"loss": 0.0326,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.25,
|
| 816 |
+
"learning_rate": 4.5807851647267706e-05,
|
| 817 |
+
"loss": 0.03,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.26,
|
| 822 |
+
"learning_rate": 4.575853225488262e-05,
|
| 823 |
+
"loss": 0.1424,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.26,
|
| 828 |
+
"learning_rate": 4.570921286249754e-05,
|
| 829 |
+
"loss": 0.0091,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.26,
|
| 834 |
+
"learning_rate": 4.5659893470112455e-05,
|
| 835 |
+
"loss": 0.0032,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.26,
|
| 840 |
+
"learning_rate": 4.5610574077727364e-05,
|
| 841 |
+
"loss": 0.0319,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.26,
|
| 846 |
+
"learning_rate": 4.556125468534228e-05,
|
| 847 |
+
"loss": 0.0602,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.27,
|
| 852 |
+
"learning_rate": 4.551193529295719e-05,
|
| 853 |
+
"loss": 0.0325,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.27,
|
| 858 |
+
"learning_rate": 4.5462615900572106e-05,
|
| 859 |
+
"loss": 0.0871,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.27,
|
| 864 |
+
"learning_rate": 4.541329650818702e-05,
|
| 865 |
+
"loss": 0.0039,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.27,
|
| 870 |
+
"learning_rate": 4.536397711580194e-05,
|
| 871 |
+
"loss": 0.119,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.27,
|
| 876 |
+
"learning_rate": 4.531465772341685e-05,
|
| 877 |
+
"loss": 0.0058,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.27,
|
| 882 |
+
"learning_rate": 4.5265338331031764e-05,
|
| 883 |
+
"loss": 0.0586,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.28,
|
| 888 |
+
"learning_rate": 4.5216018938646674e-05,
|
| 889 |
+
"loss": 0.03,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.28,
|
| 894 |
+
"learning_rate": 4.516669954626159e-05,
|
| 895 |
+
"loss": 0.0594,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.28,
|
| 900 |
+
"learning_rate": 4.5117380153876506e-05,
|
| 901 |
+
"loss": 0.0663,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.28,
|
| 906 |
+
"learning_rate": 4.506806076149142e-05,
|
| 907 |
+
"loss": 0.032,
|
| 908 |
+
"step": 1500
|
| 909 |
+
}
|
| 910 |
+
],
|
| 911 |
+
"max_steps": 10638,
|
| 912 |
+
"num_train_epochs": 2,
|
| 913 |
+
"total_flos": 9472338173952000.0,
|
| 914 |
+
"trial_name": null,
|
| 915 |
+
"trial_params": null
|
| 916 |
+
}
|
training_args (1).bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2ec049d4430a0f4619a1bf6076acb5fdee3110102fdcc72c8da1eca7429ee2e
|
| 3 |
+
size 3515
|