commit version 1
Browse files- checkpoint-130000/config.json +23 -0
- checkpoint-130000/optimizer.pt +3 -0
- checkpoint-130000/pytorch_model.bin +3 -0
- checkpoint-130000/scheduler.pt +3 -0
- checkpoint-130000/trainer_state.json +1576 -0
- checkpoint-130000/training_args.bin +3 -0
- checkpoint-140000/config.json +23 -0
- checkpoint-140000/optimizer.pt +3 -0
- checkpoint-140000/pytorch_model.bin +3 -0
- checkpoint-140000/scheduler.pt +3 -0
- checkpoint-140000/trainer_state.json +1696 -0
- checkpoint-140000/training_args.bin +3 -0
- config.json +23 -0
- hasil/config.json +23 -0
- hasil/pytorch_model.bin +3 -0
- hasil/training_args.bin +3 -0
- pytorch_model.bin +3 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
checkpoint-130000/config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"gradient_checkpointing": false,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.1,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 3072,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"num_hidden_layers": 12,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"transformers_version": "4.2.2",
|
| 20 |
+
"type_vocab_size": 2,
|
| 21 |
+
"use_cache": true,
|
| 22 |
+
"vocab_size": 32000
|
| 23 |
+
}
|
checkpoint-130000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:790ab85114465c10d00587ad44efedb16a5b5e0eefeaceb54d1ab7a86f114a84
|
| 3 |
+
size 885334243
|
checkpoint-130000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:926e956398f87c99591eef5bb7592f3cc2ec708cd26c48c60b769340e9052609
|
| 3 |
+
size 442693650
|
checkpoint-130000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9629f1988f03c5ed49df2cb928d1d06c0303430bfbe46abbfb8c4285e514b160
|
| 3 |
+
size 623
|
checkpoint-130000/trainer_state.json
ADDED
|
@@ -0,0 +1,1576 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.223783170143323,
|
| 5 |
+
"global_step": 130000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 4.9822619554420325e-05,
|
| 13 |
+
"loss": 7.0399,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.07,
|
| 18 |
+
"learning_rate": 4.964523910884065e-05,
|
| 19 |
+
"loss": 6.9038,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.11,
|
| 24 |
+
"learning_rate": 4.946785866326096e-05,
|
| 25 |
+
"loss": 6.8142,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.14,
|
| 30 |
+
"learning_rate": 4.9290478217681286e-05,
|
| 31 |
+
"loss": 6.7261,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.18,
|
| 36 |
+
"learning_rate": 4.911309777210161e-05,
|
| 37 |
+
"loss": 6.6601,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.21,
|
| 42 |
+
"learning_rate": 4.893571732652193e-05,
|
| 43 |
+
"loss": 6.607,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.25,
|
| 48 |
+
"learning_rate": 4.8758336880942247e-05,
|
| 49 |
+
"loss": 6.565,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.28,
|
| 54 |
+
"learning_rate": 4.858095643536257e-05,
|
| 55 |
+
"loss": 6.5083,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.32,
|
| 60 |
+
"learning_rate": 4.840357598978289e-05,
|
| 61 |
+
"loss": 6.4739,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.35,
|
| 66 |
+
"learning_rate": 4.8226195544203214e-05,
|
| 67 |
+
"loss": 6.4472,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.39,
|
| 72 |
+
"learning_rate": 4.804881509862353e-05,
|
| 73 |
+
"loss": 6.4133,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.43,
|
| 78 |
+
"learning_rate": 4.787143465304385e-05,
|
| 79 |
+
"loss": 6.3857,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.46,
|
| 84 |
+
"learning_rate": 4.7694054207464175e-05,
|
| 85 |
+
"loss": 6.3575,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.5,
|
| 90 |
+
"learning_rate": 4.75166737618845e-05,
|
| 91 |
+
"loss": 6.3337,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.53,
|
| 96 |
+
"learning_rate": 4.733929331630481e-05,
|
| 97 |
+
"loss": 6.3248,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.57,
|
| 102 |
+
"learning_rate": 4.7161912870725136e-05,
|
| 103 |
+
"loss": 6.2974,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.6,
|
| 108 |
+
"learning_rate": 4.698453242514546e-05,
|
| 109 |
+
"loss": 6.2786,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.64,
|
| 114 |
+
"learning_rate": 4.680715197956578e-05,
|
| 115 |
+
"loss": 6.2569,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.67,
|
| 120 |
+
"learning_rate": 4.6629771533986097e-05,
|
| 121 |
+
"loss": 6.2477,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.71,
|
| 126 |
+
"learning_rate": 4.645239108840642e-05,
|
| 127 |
+
"loss": 6.2252,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.74,
|
| 132 |
+
"learning_rate": 4.627501064282674e-05,
|
| 133 |
+
"loss": 6.212,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.78,
|
| 138 |
+
"learning_rate": 4.609763019724706e-05,
|
| 139 |
+
"loss": 6.1802,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.82,
|
| 144 |
+
"learning_rate": 4.592024975166738e-05,
|
| 145 |
+
"loss": 6.1081,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.85,
|
| 150 |
+
"learning_rate": 4.5742869306087696e-05,
|
| 151 |
+
"loss": 6.0248,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.89,
|
| 156 |
+
"learning_rate": 4.556548886050802e-05,
|
| 157 |
+
"loss": 5.8882,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.92,
|
| 162 |
+
"learning_rate": 4.538810841492834e-05,
|
| 163 |
+
"loss": 5.7505,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.96,
|
| 168 |
+
"learning_rate": 4.5210727969348656e-05,
|
| 169 |
+
"loss": 5.5657,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.99,
|
| 174 |
+
"learning_rate": 4.503334752376898e-05,
|
| 175 |
+
"loss": 5.4052,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.03,
|
| 180 |
+
"learning_rate": 4.48559670781893e-05,
|
| 181 |
+
"loss": 5.2834,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.06,
|
| 186 |
+
"learning_rate": 4.4678586632609624e-05,
|
| 187 |
+
"loss": 5.1588,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.1,
|
| 192 |
+
"learning_rate": 4.450120618702994e-05,
|
| 193 |
+
"loss": 5.0617,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.14,
|
| 198 |
+
"learning_rate": 4.432382574145026e-05,
|
| 199 |
+
"loss": 4.9254,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.17,
|
| 204 |
+
"learning_rate": 4.4146445295870585e-05,
|
| 205 |
+
"loss": 4.7733,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.21,
|
| 210 |
+
"learning_rate": 4.396906485029091e-05,
|
| 211 |
+
"loss": 4.6491,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.24,
|
| 216 |
+
"learning_rate": 4.379168440471122e-05,
|
| 217 |
+
"loss": 4.5447,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.28,
|
| 222 |
+
"learning_rate": 4.3614303959131546e-05,
|
| 223 |
+
"loss": 4.4222,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.31,
|
| 228 |
+
"learning_rate": 4.343692351355187e-05,
|
| 229 |
+
"loss": 4.3276,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.35,
|
| 234 |
+
"learning_rate": 4.325954306797219e-05,
|
| 235 |
+
"loss": 4.2297,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.38,
|
| 240 |
+
"learning_rate": 4.3082162622392507e-05,
|
| 241 |
+
"loss": 4.169,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.42,
|
| 246 |
+
"learning_rate": 4.290478217681283e-05,
|
| 247 |
+
"loss": 4.1032,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.45,
|
| 252 |
+
"learning_rate": 4.272740173123315e-05,
|
| 253 |
+
"loss": 4.0278,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.49,
|
| 258 |
+
"learning_rate": 4.2550021285653474e-05,
|
| 259 |
+
"loss": 3.9557,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.53,
|
| 264 |
+
"learning_rate": 4.237264084007379e-05,
|
| 265 |
+
"loss": 3.8852,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.56,
|
| 270 |
+
"learning_rate": 4.219526039449411e-05,
|
| 271 |
+
"loss": 3.8525,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.6,
|
| 276 |
+
"learning_rate": 4.2017879948914435e-05,
|
| 277 |
+
"loss": 3.8033,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.63,
|
| 282 |
+
"learning_rate": 4.184049950333476e-05,
|
| 283 |
+
"loss": 3.7525,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.67,
|
| 288 |
+
"learning_rate": 4.166311905775507e-05,
|
| 289 |
+
"loss": 3.7301,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.7,
|
| 294 |
+
"learning_rate": 4.1485738612175396e-05,
|
| 295 |
+
"loss": 3.6619,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.74,
|
| 300 |
+
"learning_rate": 4.130835816659572e-05,
|
| 301 |
+
"loss": 3.6457,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.77,
|
| 306 |
+
"learning_rate": 4.113097772101604e-05,
|
| 307 |
+
"loss": 3.6079,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.81,
|
| 312 |
+
"learning_rate": 4.0953597275436357e-05,
|
| 313 |
+
"loss": 3.569,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.84,
|
| 318 |
+
"learning_rate": 4.077621682985668e-05,
|
| 319 |
+
"loss": 3.5467,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.88,
|
| 324 |
+
"learning_rate": 4.0598836384277e-05,
|
| 325 |
+
"loss": 3.5302,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.92,
|
| 330 |
+
"learning_rate": 4.0421455938697324e-05,
|
| 331 |
+
"loss": 3.4952,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.95,
|
| 336 |
+
"learning_rate": 4.024407549311764e-05,
|
| 337 |
+
"loss": 3.4743,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.99,
|
| 342 |
+
"learning_rate": 4.006669504753796e-05,
|
| 343 |
+
"loss": 3.4558,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.02,
|
| 348 |
+
"learning_rate": 3.9889314601958285e-05,
|
| 349 |
+
"loss": 3.4161,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.06,
|
| 354 |
+
"learning_rate": 3.971193415637861e-05,
|
| 355 |
+
"loss": 3.3899,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.09,
|
| 360 |
+
"learning_rate": 3.953455371079892e-05,
|
| 361 |
+
"loss": 3.3836,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.13,
|
| 366 |
+
"learning_rate": 3.9357173265219246e-05,
|
| 367 |
+
"loss": 3.3521,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.16,
|
| 372 |
+
"learning_rate": 3.917979281963957e-05,
|
| 373 |
+
"loss": 3.3418,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.2,
|
| 378 |
+
"learning_rate": 3.900241237405989e-05,
|
| 379 |
+
"loss": 3.3161,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.23,
|
| 384 |
+
"learning_rate": 3.8825031928480207e-05,
|
| 385 |
+
"loss": 3.3091,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.27,
|
| 390 |
+
"learning_rate": 3.864765148290053e-05,
|
| 391 |
+
"loss": 3.2917,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.31,
|
| 396 |
+
"learning_rate": 3.847027103732085e-05,
|
| 397 |
+
"loss": 3.268,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.34,
|
| 402 |
+
"learning_rate": 3.829289059174117e-05,
|
| 403 |
+
"loss": 3.2515,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.38,
|
| 408 |
+
"learning_rate": 3.811551014616149e-05,
|
| 409 |
+
"loss": 3.2415,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.41,
|
| 414 |
+
"learning_rate": 3.793812970058181e-05,
|
| 415 |
+
"loss": 3.2177,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.45,
|
| 420 |
+
"learning_rate": 3.776074925500213e-05,
|
| 421 |
+
"loss": 3.2102,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.48,
|
| 426 |
+
"learning_rate": 3.758336880942245e-05,
|
| 427 |
+
"loss": 3.2186,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.52,
|
| 432 |
+
"learning_rate": 3.740598836384277e-05,
|
| 433 |
+
"loss": 3.1951,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.55,
|
| 438 |
+
"learning_rate": 3.722860791826309e-05,
|
| 439 |
+
"loss": 3.1766,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.59,
|
| 444 |
+
"learning_rate": 3.705122747268341e-05,
|
| 445 |
+
"loss": 3.1603,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.63,
|
| 450 |
+
"learning_rate": 3.6873847027103734e-05,
|
| 451 |
+
"loss": 3.1269,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.66,
|
| 456 |
+
"learning_rate": 3.669646658152405e-05,
|
| 457 |
+
"loss": 3.1431,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.7,
|
| 462 |
+
"learning_rate": 3.651908613594437e-05,
|
| 463 |
+
"loss": 3.112,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.73,
|
| 468 |
+
"learning_rate": 3.6341705690364695e-05,
|
| 469 |
+
"loss": 3.1211,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.77,
|
| 474 |
+
"learning_rate": 3.616432524478502e-05,
|
| 475 |
+
"loss": 3.0896,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.8,
|
| 480 |
+
"learning_rate": 3.598694479920533e-05,
|
| 481 |
+
"loss": 3.0972,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.84,
|
| 486 |
+
"learning_rate": 3.5809564353625656e-05,
|
| 487 |
+
"loss": 3.0821,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.87,
|
| 492 |
+
"learning_rate": 3.563218390804598e-05,
|
| 493 |
+
"loss": 3.0626,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.91,
|
| 498 |
+
"learning_rate": 3.54548034624663e-05,
|
| 499 |
+
"loss": 3.0586,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.94,
|
| 504 |
+
"learning_rate": 3.5277423016886617e-05,
|
| 505 |
+
"loss": 3.0464,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.98,
|
| 510 |
+
"learning_rate": 3.510004257130694e-05,
|
| 511 |
+
"loss": 3.0368,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.02,
|
| 516 |
+
"learning_rate": 3.492266212572726e-05,
|
| 517 |
+
"loss": 3.0282,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.05,
|
| 522 |
+
"learning_rate": 3.4745281680147584e-05,
|
| 523 |
+
"loss": 2.9903,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 3.09,
|
| 528 |
+
"learning_rate": 3.45679012345679e-05,
|
| 529 |
+
"loss": 2.9912,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.12,
|
| 534 |
+
"learning_rate": 3.439052078898822e-05,
|
| 535 |
+
"loss": 2.9902,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.16,
|
| 540 |
+
"learning_rate": 3.4213140343408545e-05,
|
| 541 |
+
"loss": 2.995,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.19,
|
| 546 |
+
"learning_rate": 3.403575989782887e-05,
|
| 547 |
+
"loss": 2.9736,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 3.23,
|
| 552 |
+
"learning_rate": 3.385837945224918e-05,
|
| 553 |
+
"loss": 2.9724,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.26,
|
| 558 |
+
"learning_rate": 3.3680999006669506e-05,
|
| 559 |
+
"loss": 2.9636,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 3.3,
|
| 564 |
+
"learning_rate": 3.350361856108983e-05,
|
| 565 |
+
"loss": 2.957,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 3.33,
|
| 570 |
+
"learning_rate": 3.332623811551015e-05,
|
| 571 |
+
"loss": 2.9563,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.37,
|
| 576 |
+
"learning_rate": 3.3148857669930467e-05,
|
| 577 |
+
"loss": 2.9381,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 3.41,
|
| 582 |
+
"learning_rate": 3.297147722435079e-05,
|
| 583 |
+
"loss": 2.926,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.44,
|
| 588 |
+
"learning_rate": 3.279409677877111e-05,
|
| 589 |
+
"loss": 2.933,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.48,
|
| 594 |
+
"learning_rate": 3.2616716333191434e-05,
|
| 595 |
+
"loss": 2.9144,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.51,
|
| 600 |
+
"learning_rate": 3.243933588761175e-05,
|
| 601 |
+
"loss": 2.9084,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.55,
|
| 606 |
+
"learning_rate": 3.226195544203207e-05,
|
| 607 |
+
"loss": 2.9116,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.58,
|
| 612 |
+
"learning_rate": 3.2084574996452395e-05,
|
| 613 |
+
"loss": 2.8892,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.62,
|
| 618 |
+
"learning_rate": 3.190719455087272e-05,
|
| 619 |
+
"loss": 2.8828,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 3.65,
|
| 624 |
+
"learning_rate": 3.172981410529303e-05,
|
| 625 |
+
"loss": 2.8843,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.69,
|
| 630 |
+
"learning_rate": 3.1552433659713356e-05,
|
| 631 |
+
"loss": 2.8799,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.72,
|
| 636 |
+
"learning_rate": 3.137505321413368e-05,
|
| 637 |
+
"loss": 2.8586,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 3.76,
|
| 642 |
+
"learning_rate": 3.1197672768554e-05,
|
| 643 |
+
"loss": 2.8601,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 3.8,
|
| 648 |
+
"learning_rate": 3.1020292322974317e-05,
|
| 649 |
+
"loss": 2.8523,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 3.83,
|
| 654 |
+
"learning_rate": 3.084291187739464e-05,
|
| 655 |
+
"loss": 2.8517,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 3.87,
|
| 660 |
+
"learning_rate": 3.066553143181496e-05,
|
| 661 |
+
"loss": 2.8347,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.9,
|
| 666 |
+
"learning_rate": 3.048815098623528e-05,
|
| 667 |
+
"loss": 2.8427,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.94,
|
| 672 |
+
"learning_rate": 3.0310770540655597e-05,
|
| 673 |
+
"loss": 2.8471,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.97,
|
| 678 |
+
"learning_rate": 3.013339009507592e-05,
|
| 679 |
+
"loss": 2.8201,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 4.01,
|
| 684 |
+
"learning_rate": 2.995600964949624e-05,
|
| 685 |
+
"loss": 2.8236,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 4.04,
|
| 690 |
+
"learning_rate": 2.9778629203916564e-05,
|
| 691 |
+
"loss": 2.8043,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 4.08,
|
| 696 |
+
"learning_rate": 2.960124875833688e-05,
|
| 697 |
+
"loss": 2.7957,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.12,
|
| 702 |
+
"learning_rate": 2.9423868312757202e-05,
|
| 703 |
+
"loss": 2.8042,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 4.15,
|
| 708 |
+
"learning_rate": 2.9246487867177525e-05,
|
| 709 |
+
"loss": 2.7836,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.19,
|
| 714 |
+
"learning_rate": 2.9069107421597847e-05,
|
| 715 |
+
"loss": 2.7875,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.22,
|
| 720 |
+
"learning_rate": 2.8891726976018163e-05,
|
| 721 |
+
"loss": 2.7737,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 4.26,
|
| 726 |
+
"learning_rate": 2.8714346530438486e-05,
|
| 727 |
+
"loss": 2.7783,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 4.29,
|
| 732 |
+
"learning_rate": 2.8536966084858808e-05,
|
| 733 |
+
"loss": 2.7689,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.33,
|
| 738 |
+
"learning_rate": 2.835958563927913e-05,
|
| 739 |
+
"loss": 2.7689,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 4.36,
|
| 744 |
+
"learning_rate": 2.8182205193699447e-05,
|
| 745 |
+
"loss": 2.7737,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 4.4,
|
| 750 |
+
"learning_rate": 2.800482474811977e-05,
|
| 751 |
+
"loss": 2.7595,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.43,
|
| 756 |
+
"learning_rate": 2.7827444302540088e-05,
|
| 757 |
+
"loss": 2.7657,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 4.47,
|
| 762 |
+
"learning_rate": 2.765006385696041e-05,
|
| 763 |
+
"loss": 2.7664,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 4.51,
|
| 768 |
+
"learning_rate": 2.747268341138073e-05,
|
| 769 |
+
"loss": 2.7394,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 4.54,
|
| 774 |
+
"learning_rate": 2.729530296580105e-05,
|
| 775 |
+
"loss": 2.7357,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 4.58,
|
| 780 |
+
"learning_rate": 2.711792252022137e-05,
|
| 781 |
+
"loss": 2.7346,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 4.61,
|
| 786 |
+
"learning_rate": 2.6940542074641694e-05,
|
| 787 |
+
"loss": 2.7351,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 4.65,
|
| 792 |
+
"learning_rate": 2.676316162906201e-05,
|
| 793 |
+
"loss": 2.7268,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.68,
|
| 798 |
+
"learning_rate": 2.6585781183482332e-05,
|
| 799 |
+
"loss": 2.7319,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.72,
|
| 804 |
+
"learning_rate": 2.6408400737902655e-05,
|
| 805 |
+
"loss": 2.7282,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 4.75,
|
| 810 |
+
"learning_rate": 2.6231020292322977e-05,
|
| 811 |
+
"loss": 2.7297,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 4.79,
|
| 816 |
+
"learning_rate": 2.6053639846743293e-05,
|
| 817 |
+
"loss": 2.7112,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 4.82,
|
| 822 |
+
"learning_rate": 2.5876259401163616e-05,
|
| 823 |
+
"loss": 2.7315,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 4.86,
|
| 828 |
+
"learning_rate": 2.5698878955583938e-05,
|
| 829 |
+
"loss": 2.704,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 4.9,
|
| 834 |
+
"learning_rate": 2.552149851000426e-05,
|
| 835 |
+
"loss": 2.7036,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.93,
|
| 840 |
+
"learning_rate": 2.5344118064424577e-05,
|
| 841 |
+
"loss": 2.7103,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.97,
|
| 846 |
+
"learning_rate": 2.51667376188449e-05,
|
| 847 |
+
"loss": 2.7022,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 5.0,
|
| 852 |
+
"learning_rate": 2.498935717326522e-05,
|
| 853 |
+
"loss": 2.6949,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 5.04,
|
| 858 |
+
"learning_rate": 2.481197672768554e-05,
|
| 859 |
+
"loss": 2.681,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 5.07,
|
| 864 |
+
"learning_rate": 2.4634596282105863e-05,
|
| 865 |
+
"loss": 2.6636,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 5.11,
|
| 870 |
+
"learning_rate": 2.4457215836526182e-05,
|
| 871 |
+
"loss": 2.6868,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 5.14,
|
| 876 |
+
"learning_rate": 2.4279835390946505e-05,
|
| 877 |
+
"loss": 2.6683,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 5.18,
|
| 882 |
+
"learning_rate": 2.4102454945366824e-05,
|
| 883 |
+
"loss": 2.6765,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 5.21,
|
| 888 |
+
"learning_rate": 2.3925074499787147e-05,
|
| 889 |
+
"loss": 2.6665,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 5.25,
|
| 894 |
+
"learning_rate": 2.3747694054207466e-05,
|
| 895 |
+
"loss": 2.6672,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 5.29,
|
| 900 |
+
"learning_rate": 2.3570313608627785e-05,
|
| 901 |
+
"loss": 2.6604,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 5.32,
|
| 906 |
+
"learning_rate": 2.3392933163048107e-05,
|
| 907 |
+
"loss": 2.6552,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 5.36,
|
| 912 |
+
"learning_rate": 2.3215552717468427e-05,
|
| 913 |
+
"loss": 2.6561,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 5.39,
|
| 918 |
+
"learning_rate": 2.3038172271888746e-05,
|
| 919 |
+
"loss": 2.6418,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.43,
|
| 924 |
+
"learning_rate": 2.2860791826309068e-05,
|
| 925 |
+
"loss": 2.6552,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 5.46,
|
| 930 |
+
"learning_rate": 2.2683411380729387e-05,
|
| 931 |
+
"loss": 2.6404,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 5.5,
|
| 936 |
+
"learning_rate": 2.250603093514971e-05,
|
| 937 |
+
"loss": 2.6317,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 5.53,
|
| 942 |
+
"learning_rate": 2.232865048957003e-05,
|
| 943 |
+
"loss": 2.6343,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 5.57,
|
| 948 |
+
"learning_rate": 2.215127004399035e-05,
|
| 949 |
+
"loss": 2.6446,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 5.61,
|
| 954 |
+
"learning_rate": 2.197388959841067e-05,
|
| 955 |
+
"loss": 2.6201,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 5.64,
|
| 960 |
+
"learning_rate": 2.1796509152830993e-05,
|
| 961 |
+
"loss": 2.6266,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 5.68,
|
| 966 |
+
"learning_rate": 2.1619128707251312e-05,
|
| 967 |
+
"loss": 2.6229,
|
| 968 |
+
"step": 80000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 5.71,
|
| 972 |
+
"learning_rate": 2.1441748261671635e-05,
|
| 973 |
+
"loss": 2.6297,
|
| 974 |
+
"step": 80500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 5.75,
|
| 978 |
+
"learning_rate": 2.1264367816091954e-05,
|
| 979 |
+
"loss": 2.6183,
|
| 980 |
+
"step": 81000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 5.78,
|
| 984 |
+
"learning_rate": 2.1086987370512277e-05,
|
| 985 |
+
"loss": 2.6183,
|
| 986 |
+
"step": 81500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 5.82,
|
| 990 |
+
"learning_rate": 2.0909606924932596e-05,
|
| 991 |
+
"loss": 2.6164,
|
| 992 |
+
"step": 82000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 5.85,
|
| 996 |
+
"learning_rate": 2.073222647935292e-05,
|
| 997 |
+
"loss": 2.6069,
|
| 998 |
+
"step": 82500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 5.89,
|
| 1002 |
+
"learning_rate": 2.0554846033773237e-05,
|
| 1003 |
+
"loss": 2.6219,
|
| 1004 |
+
"step": 83000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 5.92,
|
| 1008 |
+
"learning_rate": 2.037746558819356e-05,
|
| 1009 |
+
"loss": 2.592,
|
| 1010 |
+
"step": 83500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.96,
|
| 1014 |
+
"learning_rate": 2.020008514261388e-05,
|
| 1015 |
+
"loss": 2.5975,
|
| 1016 |
+
"step": 84000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 6.0,
|
| 1020 |
+
"learning_rate": 2.00227046970342e-05,
|
| 1021 |
+
"loss": 2.6029,
|
| 1022 |
+
"step": 84500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 6.03,
|
| 1026 |
+
"learning_rate": 1.984532425145452e-05,
|
| 1027 |
+
"loss": 2.5913,
|
| 1028 |
+
"step": 85000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 6.07,
|
| 1032 |
+
"learning_rate": 1.9667943805874843e-05,
|
| 1033 |
+
"loss": 2.5867,
|
| 1034 |
+
"step": 85500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 6.1,
|
| 1038 |
+
"learning_rate": 1.9490563360295162e-05,
|
| 1039 |
+
"loss": 2.5899,
|
| 1040 |
+
"step": 86000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 6.14,
|
| 1044 |
+
"learning_rate": 1.931318291471548e-05,
|
| 1045 |
+
"loss": 2.5923,
|
| 1046 |
+
"step": 86500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 6.17,
|
| 1050 |
+
"learning_rate": 1.91358024691358e-05,
|
| 1051 |
+
"loss": 2.5742,
|
| 1052 |
+
"step": 87000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 6.21,
|
| 1056 |
+
"learning_rate": 1.8958422023556123e-05,
|
| 1057 |
+
"loss": 2.5859,
|
| 1058 |
+
"step": 87500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 6.24,
|
| 1062 |
+
"learning_rate": 1.8781041577976442e-05,
|
| 1063 |
+
"loss": 2.5683,
|
| 1064 |
+
"step": 88000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 6.28,
|
| 1068 |
+
"learning_rate": 1.8603661132396765e-05,
|
| 1069 |
+
"loss": 2.5808,
|
| 1070 |
+
"step": 88500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 6.31,
|
| 1074 |
+
"learning_rate": 1.8426280686817084e-05,
|
| 1075 |
+
"loss": 2.5663,
|
| 1076 |
+
"step": 89000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 6.35,
|
| 1080 |
+
"learning_rate": 1.8248900241237407e-05,
|
| 1081 |
+
"loss": 2.5677,
|
| 1082 |
+
"step": 89500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 6.39,
|
| 1086 |
+
"learning_rate": 1.8071519795657726e-05,
|
| 1087 |
+
"loss": 2.5647,
|
| 1088 |
+
"step": 90000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 6.42,
|
| 1092 |
+
"learning_rate": 1.7894139350078048e-05,
|
| 1093 |
+
"loss": 2.5656,
|
| 1094 |
+
"step": 90500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 6.46,
|
| 1098 |
+
"learning_rate": 1.7716758904498367e-05,
|
| 1099 |
+
"loss": 2.5682,
|
| 1100 |
+
"step": 91000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 6.49,
|
| 1104 |
+
"learning_rate": 1.753937845891869e-05,
|
| 1105 |
+
"loss": 2.5645,
|
| 1106 |
+
"step": 91500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 6.53,
|
| 1110 |
+
"learning_rate": 1.736199801333901e-05,
|
| 1111 |
+
"loss": 2.5586,
|
| 1112 |
+
"step": 92000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 6.56,
|
| 1116 |
+
"learning_rate": 1.718461756775933e-05,
|
| 1117 |
+
"loss": 2.5488,
|
| 1118 |
+
"step": 92500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 6.6,
|
| 1122 |
+
"learning_rate": 1.700723712217965e-05,
|
| 1123 |
+
"loss": 2.5547,
|
| 1124 |
+
"step": 93000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 6.63,
|
| 1128 |
+
"learning_rate": 1.6829856676599973e-05,
|
| 1129 |
+
"loss": 2.5485,
|
| 1130 |
+
"step": 93500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 6.67,
|
| 1134 |
+
"learning_rate": 1.6652476231020292e-05,
|
| 1135 |
+
"loss": 2.5503,
|
| 1136 |
+
"step": 94000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 6.7,
|
| 1140 |
+
"learning_rate": 1.6475095785440615e-05,
|
| 1141 |
+
"loss": 2.547,
|
| 1142 |
+
"step": 94500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 6.74,
|
| 1146 |
+
"learning_rate": 1.6297715339860934e-05,
|
| 1147 |
+
"loss": 2.5468,
|
| 1148 |
+
"step": 95000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 6.78,
|
| 1152 |
+
"learning_rate": 1.6120334894281257e-05,
|
| 1153 |
+
"loss": 2.5416,
|
| 1154 |
+
"step": 95500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 6.81,
|
| 1158 |
+
"learning_rate": 1.5942954448701576e-05,
|
| 1159 |
+
"loss": 2.5562,
|
| 1160 |
+
"step": 96000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 6.85,
|
| 1164 |
+
"learning_rate": 1.57655740031219e-05,
|
| 1165 |
+
"loss": 2.5311,
|
| 1166 |
+
"step": 96500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 6.88,
|
| 1170 |
+
"learning_rate": 1.5588193557542217e-05,
|
| 1171 |
+
"loss": 2.5481,
|
| 1172 |
+
"step": 97000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 6.92,
|
| 1176 |
+
"learning_rate": 1.5410813111962537e-05,
|
| 1177 |
+
"loss": 2.5362,
|
| 1178 |
+
"step": 97500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.95,
|
| 1182 |
+
"learning_rate": 1.5233432666382857e-05,
|
| 1183 |
+
"loss": 2.5161,
|
| 1184 |
+
"step": 98000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 6.99,
|
| 1188 |
+
"learning_rate": 1.505605222080318e-05,
|
| 1189 |
+
"loss": 2.5354,
|
| 1190 |
+
"step": 98500
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 7.02,
|
| 1194 |
+
"learning_rate": 1.4878671775223499e-05,
|
| 1195 |
+
"loss": 2.5214,
|
| 1196 |
+
"step": 99000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 7.06,
|
| 1200 |
+
"learning_rate": 1.4701291329643822e-05,
|
| 1201 |
+
"loss": 2.5271,
|
| 1202 |
+
"step": 99500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 7.1,
|
| 1206 |
+
"learning_rate": 1.452391088406414e-05,
|
| 1207 |
+
"loss": 2.5024,
|
| 1208 |
+
"step": 100000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 7.13,
|
| 1212 |
+
"learning_rate": 1.4346530438484462e-05,
|
| 1213 |
+
"loss": 2.5096,
|
| 1214 |
+
"step": 100500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 7.17,
|
| 1218 |
+
"learning_rate": 1.4169149992904782e-05,
|
| 1219 |
+
"loss": 2.5094,
|
| 1220 |
+
"step": 101000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 7.2,
|
| 1224 |
+
"learning_rate": 1.3991769547325103e-05,
|
| 1225 |
+
"loss": 2.5151,
|
| 1226 |
+
"step": 101500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 7.24,
|
| 1230 |
+
"learning_rate": 1.3814389101745422e-05,
|
| 1231 |
+
"loss": 2.5141,
|
| 1232 |
+
"step": 102000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 7.27,
|
| 1236 |
+
"learning_rate": 1.3637008656165745e-05,
|
| 1237 |
+
"loss": 2.5102,
|
| 1238 |
+
"step": 102500
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 7.31,
|
| 1242 |
+
"learning_rate": 1.3459628210586064e-05,
|
| 1243 |
+
"loss": 2.5091,
|
| 1244 |
+
"step": 103000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 7.34,
|
| 1248 |
+
"learning_rate": 1.3282247765006387e-05,
|
| 1249 |
+
"loss": 2.5025,
|
| 1250 |
+
"step": 103500
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 7.38,
|
| 1254 |
+
"learning_rate": 1.3104867319426706e-05,
|
| 1255 |
+
"loss": 2.5025,
|
| 1256 |
+
"step": 104000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 7.41,
|
| 1260 |
+
"learning_rate": 1.2927486873847028e-05,
|
| 1261 |
+
"loss": 2.5051,
|
| 1262 |
+
"step": 104500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.45,
|
| 1266 |
+
"learning_rate": 1.2750106428267347e-05,
|
| 1267 |
+
"loss": 2.5084,
|
| 1268 |
+
"step": 105000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 7.49,
|
| 1272 |
+
"learning_rate": 1.257272598268767e-05,
|
| 1273 |
+
"loss": 2.4955,
|
| 1274 |
+
"step": 105500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 7.52,
|
| 1278 |
+
"learning_rate": 1.2395345537107989e-05,
|
| 1279 |
+
"loss": 2.4915,
|
| 1280 |
+
"step": 106000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 7.56,
|
| 1284 |
+
"learning_rate": 1.221796509152831e-05,
|
| 1285 |
+
"loss": 2.5053,
|
| 1286 |
+
"step": 106500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 7.59,
|
| 1290 |
+
"learning_rate": 1.204058464594863e-05,
|
| 1291 |
+
"loss": 2.4989,
|
| 1292 |
+
"step": 107000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 7.63,
|
| 1296 |
+
"learning_rate": 1.1863204200368952e-05,
|
| 1297 |
+
"loss": 2.4957,
|
| 1298 |
+
"step": 107500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 7.66,
|
| 1302 |
+
"learning_rate": 1.1685823754789272e-05,
|
| 1303 |
+
"loss": 2.4909,
|
| 1304 |
+
"step": 108000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 7.7,
|
| 1308 |
+
"learning_rate": 1.1508443309209593e-05,
|
| 1309 |
+
"loss": 2.492,
|
| 1310 |
+
"step": 108500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 7.73,
|
| 1314 |
+
"learning_rate": 1.1331062863629914e-05,
|
| 1315 |
+
"loss": 2.4929,
|
| 1316 |
+
"step": 109000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 7.77,
|
| 1320 |
+
"learning_rate": 1.1153682418050235e-05,
|
| 1321 |
+
"loss": 2.4871,
|
| 1322 |
+
"step": 109500
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 7.8,
|
| 1326 |
+
"learning_rate": 1.0976301972470556e-05,
|
| 1327 |
+
"loss": 2.4843,
|
| 1328 |
+
"step": 110000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 7.84,
|
| 1332 |
+
"learning_rate": 1.0798921526890877e-05,
|
| 1333 |
+
"loss": 2.4849,
|
| 1334 |
+
"step": 110500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 7.88,
|
| 1338 |
+
"learning_rate": 1.0621541081311197e-05,
|
| 1339 |
+
"loss": 2.4903,
|
| 1340 |
+
"step": 111000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 7.91,
|
| 1344 |
+
"learning_rate": 1.0444160635731518e-05,
|
| 1345 |
+
"loss": 2.4766,
|
| 1346 |
+
"step": 111500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 7.95,
|
| 1350 |
+
"learning_rate": 1.0266780190151837e-05,
|
| 1351 |
+
"loss": 2.4895,
|
| 1352 |
+
"step": 112000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 7.98,
|
| 1356 |
+
"learning_rate": 1.0089399744572158e-05,
|
| 1357 |
+
"loss": 2.4818,
|
| 1358 |
+
"step": 112500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 8.02,
|
| 1362 |
+
"learning_rate": 9.912019298992479e-06,
|
| 1363 |
+
"loss": 2.4782,
|
| 1364 |
+
"step": 113000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 8.05,
|
| 1368 |
+
"learning_rate": 9.7346388534128e-06,
|
| 1369 |
+
"loss": 2.4671,
|
| 1370 |
+
"step": 113500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 8.09,
|
| 1374 |
+
"learning_rate": 9.55725840783312e-06,
|
| 1375 |
+
"loss": 2.4665,
|
| 1376 |
+
"step": 114000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 8.12,
|
| 1380 |
+
"learning_rate": 9.379877962253442e-06,
|
| 1381 |
+
"loss": 2.4669,
|
| 1382 |
+
"step": 114500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 8.16,
|
| 1386 |
+
"learning_rate": 9.202497516673762e-06,
|
| 1387 |
+
"loss": 2.4552,
|
| 1388 |
+
"step": 115000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 8.19,
|
| 1392 |
+
"learning_rate": 9.025117071094083e-06,
|
| 1393 |
+
"loss": 2.459,
|
| 1394 |
+
"step": 115500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 8.23,
|
| 1398 |
+
"learning_rate": 8.847736625514404e-06,
|
| 1399 |
+
"loss": 2.455,
|
| 1400 |
+
"step": 116000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 8.27,
|
| 1404 |
+
"learning_rate": 8.670356179934725e-06,
|
| 1405 |
+
"loss": 2.472,
|
| 1406 |
+
"step": 116500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 8.3,
|
| 1410 |
+
"learning_rate": 8.492975734355046e-06,
|
| 1411 |
+
"loss": 2.4517,
|
| 1412 |
+
"step": 117000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 8.34,
|
| 1416 |
+
"learning_rate": 8.315595288775365e-06,
|
| 1417 |
+
"loss": 2.458,
|
| 1418 |
+
"step": 117500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 8.37,
|
| 1422 |
+
"learning_rate": 8.138214843195686e-06,
|
| 1423 |
+
"loss": 2.4597,
|
| 1424 |
+
"step": 118000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 8.41,
|
| 1428 |
+
"learning_rate": 7.960834397616007e-06,
|
| 1429 |
+
"loss": 2.4567,
|
| 1430 |
+
"step": 118500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 8.44,
|
| 1434 |
+
"learning_rate": 7.783453952036327e-06,
|
| 1435 |
+
"loss": 2.4561,
|
| 1436 |
+
"step": 119000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 8.48,
|
| 1440 |
+
"learning_rate": 7.606073506456648e-06,
|
| 1441 |
+
"loss": 2.4614,
|
| 1442 |
+
"step": 119500
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 8.51,
|
| 1446 |
+
"learning_rate": 7.428693060876969e-06,
|
| 1447 |
+
"loss": 2.4592,
|
| 1448 |
+
"step": 120000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 8.55,
|
| 1452 |
+
"learning_rate": 7.25131261529729e-06,
|
| 1453 |
+
"loss": 2.456,
|
| 1454 |
+
"step": 120500
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 8.59,
|
| 1458 |
+
"learning_rate": 7.073932169717611e-06,
|
| 1459 |
+
"loss": 2.4593,
|
| 1460 |
+
"step": 121000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 8.62,
|
| 1464 |
+
"learning_rate": 6.896551724137932e-06,
|
| 1465 |
+
"loss": 2.452,
|
| 1466 |
+
"step": 121500
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 8.66,
|
| 1470 |
+
"learning_rate": 6.719171278558252e-06,
|
| 1471 |
+
"loss": 2.4414,
|
| 1472 |
+
"step": 122000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 8.69,
|
| 1476 |
+
"learning_rate": 6.5417908329785725e-06,
|
| 1477 |
+
"loss": 2.4558,
|
| 1478 |
+
"step": 122500
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 8.73,
|
| 1482 |
+
"learning_rate": 6.364410387398893e-06,
|
| 1483 |
+
"loss": 2.4417,
|
| 1484 |
+
"step": 123000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 8.76,
|
| 1488 |
+
"learning_rate": 6.187029941819214e-06,
|
| 1489 |
+
"loss": 2.4364,
|
| 1490 |
+
"step": 123500
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 8.8,
|
| 1494 |
+
"learning_rate": 6.009649496239535e-06,
|
| 1495 |
+
"loss": 2.4543,
|
| 1496 |
+
"step": 124000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 8.83,
|
| 1500 |
+
"learning_rate": 5.832269050659856e-06,
|
| 1501 |
+
"loss": 2.4419,
|
| 1502 |
+
"step": 124500
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 8.87,
|
| 1506 |
+
"learning_rate": 5.654888605080176e-06,
|
| 1507 |
+
"loss": 2.446,
|
| 1508 |
+
"step": 125000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 8.9,
|
| 1512 |
+
"learning_rate": 5.477508159500497e-06,
|
| 1513 |
+
"loss": 2.4395,
|
| 1514 |
+
"step": 125500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 8.94,
|
| 1518 |
+
"learning_rate": 5.3001277139208175e-06,
|
| 1519 |
+
"loss": 2.4466,
|
| 1520 |
+
"step": 126000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 8.98,
|
| 1524 |
+
"learning_rate": 5.122747268341138e-06,
|
| 1525 |
+
"loss": 2.4301,
|
| 1526 |
+
"step": 126500
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 9.01,
|
| 1530 |
+
"learning_rate": 4.945366822761459e-06,
|
| 1531 |
+
"loss": 2.4382,
|
| 1532 |
+
"step": 127000
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 9.05,
|
| 1536 |
+
"learning_rate": 4.767986377181779e-06,
|
| 1537 |
+
"loss": 2.4406,
|
| 1538 |
+
"step": 127500
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 9.08,
|
| 1542 |
+
"learning_rate": 4.5906059316021e-06,
|
| 1543 |
+
"loss": 2.4312,
|
| 1544 |
+
"step": 128000
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 9.12,
|
| 1548 |
+
"learning_rate": 4.413225486022421e-06,
|
| 1549 |
+
"loss": 2.4409,
|
| 1550 |
+
"step": 128500
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 9.15,
|
| 1554 |
+
"learning_rate": 4.235845040442742e-06,
|
| 1555 |
+
"loss": 2.4256,
|
| 1556 |
+
"step": 129000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 9.19,
|
| 1560 |
+
"learning_rate": 4.0584645948630625e-06,
|
| 1561 |
+
"loss": 2.4301,
|
| 1562 |
+
"step": 129500
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 9.22,
|
| 1566 |
+
"learning_rate": 3.881084149283383e-06,
|
| 1567 |
+
"loss": 2.4338,
|
| 1568 |
+
"step": 130000
|
| 1569 |
+
}
|
| 1570 |
+
],
|
| 1571 |
+
"max_steps": 140940,
|
| 1572 |
+
"num_train_epochs": 10,
|
| 1573 |
+
"total_flos": 640720730353827840,
|
| 1574 |
+
"trial_name": null,
|
| 1575 |
+
"trial_params": null
|
| 1576 |
+
}
|
checkpoint-130000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d218969f864d8514dc3a66fde3f1475f190986fb24b2e4acf5d7a01940a6be8
|
| 3 |
+
size 2031
|
checkpoint-140000/config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"gradient_checkpointing": false,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.1,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 3072,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"num_hidden_layers": 12,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"transformers_version": "4.2.2",
|
| 20 |
+
"type_vocab_size": 2,
|
| 21 |
+
"use_cache": true,
|
| 22 |
+
"vocab_size": 32000
|
| 23 |
+
}
|
checkpoint-140000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:42150721ddd4cdcb28819bcfda608315d7e9c95b9111daa643984474ed2a57bd
|
| 3 |
+
size 885334243
|
checkpoint-140000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1a9cc2c9618a591200fc997f9bf6fef80c0ce41dfe02df79f9d29dd034193d2
|
| 3 |
+
size 442693650
|
checkpoint-140000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc161521de4d88c6eb0abe7ca7301a0a45edb947f0ca11de1f102e22a6776a9c
|
| 3 |
+
size 623
|
checkpoint-140000/trainer_state.json
ADDED
|
@@ -0,0 +1,1696 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.933304952462041,
|
| 5 |
+
"global_step": 140000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 4.9822619554420325e-05,
|
| 13 |
+
"loss": 7.0399,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.07,
|
| 18 |
+
"learning_rate": 4.964523910884065e-05,
|
| 19 |
+
"loss": 6.9038,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.11,
|
| 24 |
+
"learning_rate": 4.946785866326096e-05,
|
| 25 |
+
"loss": 6.8142,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.14,
|
| 30 |
+
"learning_rate": 4.9290478217681286e-05,
|
| 31 |
+
"loss": 6.7261,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.18,
|
| 36 |
+
"learning_rate": 4.911309777210161e-05,
|
| 37 |
+
"loss": 6.6601,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.21,
|
| 42 |
+
"learning_rate": 4.893571732652193e-05,
|
| 43 |
+
"loss": 6.607,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.25,
|
| 48 |
+
"learning_rate": 4.8758336880942247e-05,
|
| 49 |
+
"loss": 6.565,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.28,
|
| 54 |
+
"learning_rate": 4.858095643536257e-05,
|
| 55 |
+
"loss": 6.5083,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.32,
|
| 60 |
+
"learning_rate": 4.840357598978289e-05,
|
| 61 |
+
"loss": 6.4739,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.35,
|
| 66 |
+
"learning_rate": 4.8226195544203214e-05,
|
| 67 |
+
"loss": 6.4472,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.39,
|
| 72 |
+
"learning_rate": 4.804881509862353e-05,
|
| 73 |
+
"loss": 6.4133,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.43,
|
| 78 |
+
"learning_rate": 4.787143465304385e-05,
|
| 79 |
+
"loss": 6.3857,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.46,
|
| 84 |
+
"learning_rate": 4.7694054207464175e-05,
|
| 85 |
+
"loss": 6.3575,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.5,
|
| 90 |
+
"learning_rate": 4.75166737618845e-05,
|
| 91 |
+
"loss": 6.3337,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.53,
|
| 96 |
+
"learning_rate": 4.733929331630481e-05,
|
| 97 |
+
"loss": 6.3248,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.57,
|
| 102 |
+
"learning_rate": 4.7161912870725136e-05,
|
| 103 |
+
"loss": 6.2974,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.6,
|
| 108 |
+
"learning_rate": 4.698453242514546e-05,
|
| 109 |
+
"loss": 6.2786,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.64,
|
| 114 |
+
"learning_rate": 4.680715197956578e-05,
|
| 115 |
+
"loss": 6.2569,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.67,
|
| 120 |
+
"learning_rate": 4.6629771533986097e-05,
|
| 121 |
+
"loss": 6.2477,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.71,
|
| 126 |
+
"learning_rate": 4.645239108840642e-05,
|
| 127 |
+
"loss": 6.2252,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.74,
|
| 132 |
+
"learning_rate": 4.627501064282674e-05,
|
| 133 |
+
"loss": 6.212,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.78,
|
| 138 |
+
"learning_rate": 4.609763019724706e-05,
|
| 139 |
+
"loss": 6.1802,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.82,
|
| 144 |
+
"learning_rate": 4.592024975166738e-05,
|
| 145 |
+
"loss": 6.1081,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.85,
|
| 150 |
+
"learning_rate": 4.5742869306087696e-05,
|
| 151 |
+
"loss": 6.0248,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.89,
|
| 156 |
+
"learning_rate": 4.556548886050802e-05,
|
| 157 |
+
"loss": 5.8882,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.92,
|
| 162 |
+
"learning_rate": 4.538810841492834e-05,
|
| 163 |
+
"loss": 5.7505,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.96,
|
| 168 |
+
"learning_rate": 4.5210727969348656e-05,
|
| 169 |
+
"loss": 5.5657,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.99,
|
| 174 |
+
"learning_rate": 4.503334752376898e-05,
|
| 175 |
+
"loss": 5.4052,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.03,
|
| 180 |
+
"learning_rate": 4.48559670781893e-05,
|
| 181 |
+
"loss": 5.2834,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.06,
|
| 186 |
+
"learning_rate": 4.4678586632609624e-05,
|
| 187 |
+
"loss": 5.1588,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 1.1,
|
| 192 |
+
"learning_rate": 4.450120618702994e-05,
|
| 193 |
+
"loss": 5.0617,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 1.14,
|
| 198 |
+
"learning_rate": 4.432382574145026e-05,
|
| 199 |
+
"loss": 4.9254,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.17,
|
| 204 |
+
"learning_rate": 4.4146445295870585e-05,
|
| 205 |
+
"loss": 4.7733,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.21,
|
| 210 |
+
"learning_rate": 4.396906485029091e-05,
|
| 211 |
+
"loss": 4.6491,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.24,
|
| 216 |
+
"learning_rate": 4.379168440471122e-05,
|
| 217 |
+
"loss": 4.5447,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.28,
|
| 222 |
+
"learning_rate": 4.3614303959131546e-05,
|
| 223 |
+
"loss": 4.4222,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.31,
|
| 228 |
+
"learning_rate": 4.343692351355187e-05,
|
| 229 |
+
"loss": 4.3276,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.35,
|
| 234 |
+
"learning_rate": 4.325954306797219e-05,
|
| 235 |
+
"loss": 4.2297,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.38,
|
| 240 |
+
"learning_rate": 4.3082162622392507e-05,
|
| 241 |
+
"loss": 4.169,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.42,
|
| 246 |
+
"learning_rate": 4.290478217681283e-05,
|
| 247 |
+
"loss": 4.1032,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.45,
|
| 252 |
+
"learning_rate": 4.272740173123315e-05,
|
| 253 |
+
"loss": 4.0278,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.49,
|
| 258 |
+
"learning_rate": 4.2550021285653474e-05,
|
| 259 |
+
"loss": 3.9557,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.53,
|
| 264 |
+
"learning_rate": 4.237264084007379e-05,
|
| 265 |
+
"loss": 3.8852,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.56,
|
| 270 |
+
"learning_rate": 4.219526039449411e-05,
|
| 271 |
+
"loss": 3.8525,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.6,
|
| 276 |
+
"learning_rate": 4.2017879948914435e-05,
|
| 277 |
+
"loss": 3.8033,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.63,
|
| 282 |
+
"learning_rate": 4.184049950333476e-05,
|
| 283 |
+
"loss": 3.7525,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.67,
|
| 288 |
+
"learning_rate": 4.166311905775507e-05,
|
| 289 |
+
"loss": 3.7301,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.7,
|
| 294 |
+
"learning_rate": 4.1485738612175396e-05,
|
| 295 |
+
"loss": 3.6619,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.74,
|
| 300 |
+
"learning_rate": 4.130835816659572e-05,
|
| 301 |
+
"loss": 3.6457,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.77,
|
| 306 |
+
"learning_rate": 4.113097772101604e-05,
|
| 307 |
+
"loss": 3.6079,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.81,
|
| 312 |
+
"learning_rate": 4.0953597275436357e-05,
|
| 313 |
+
"loss": 3.569,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.84,
|
| 318 |
+
"learning_rate": 4.077621682985668e-05,
|
| 319 |
+
"loss": 3.5467,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.88,
|
| 324 |
+
"learning_rate": 4.0598836384277e-05,
|
| 325 |
+
"loss": 3.5302,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.92,
|
| 330 |
+
"learning_rate": 4.0421455938697324e-05,
|
| 331 |
+
"loss": 3.4952,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.95,
|
| 336 |
+
"learning_rate": 4.024407549311764e-05,
|
| 337 |
+
"loss": 3.4743,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.99,
|
| 342 |
+
"learning_rate": 4.006669504753796e-05,
|
| 343 |
+
"loss": 3.4558,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.02,
|
| 348 |
+
"learning_rate": 3.9889314601958285e-05,
|
| 349 |
+
"loss": 3.4161,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.06,
|
| 354 |
+
"learning_rate": 3.971193415637861e-05,
|
| 355 |
+
"loss": 3.3899,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.09,
|
| 360 |
+
"learning_rate": 3.953455371079892e-05,
|
| 361 |
+
"loss": 3.3836,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.13,
|
| 366 |
+
"learning_rate": 3.9357173265219246e-05,
|
| 367 |
+
"loss": 3.3521,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.16,
|
| 372 |
+
"learning_rate": 3.917979281963957e-05,
|
| 373 |
+
"loss": 3.3418,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.2,
|
| 378 |
+
"learning_rate": 3.900241237405989e-05,
|
| 379 |
+
"loss": 3.3161,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.23,
|
| 384 |
+
"learning_rate": 3.8825031928480207e-05,
|
| 385 |
+
"loss": 3.3091,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.27,
|
| 390 |
+
"learning_rate": 3.864765148290053e-05,
|
| 391 |
+
"loss": 3.2917,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.31,
|
| 396 |
+
"learning_rate": 3.847027103732085e-05,
|
| 397 |
+
"loss": 3.268,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.34,
|
| 402 |
+
"learning_rate": 3.829289059174117e-05,
|
| 403 |
+
"loss": 3.2515,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.38,
|
| 408 |
+
"learning_rate": 3.811551014616149e-05,
|
| 409 |
+
"loss": 3.2415,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.41,
|
| 414 |
+
"learning_rate": 3.793812970058181e-05,
|
| 415 |
+
"loss": 3.2177,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.45,
|
| 420 |
+
"learning_rate": 3.776074925500213e-05,
|
| 421 |
+
"loss": 3.2102,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.48,
|
| 426 |
+
"learning_rate": 3.758336880942245e-05,
|
| 427 |
+
"loss": 3.2186,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.52,
|
| 432 |
+
"learning_rate": 3.740598836384277e-05,
|
| 433 |
+
"loss": 3.1951,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.55,
|
| 438 |
+
"learning_rate": 3.722860791826309e-05,
|
| 439 |
+
"loss": 3.1766,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.59,
|
| 444 |
+
"learning_rate": 3.705122747268341e-05,
|
| 445 |
+
"loss": 3.1603,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.63,
|
| 450 |
+
"learning_rate": 3.6873847027103734e-05,
|
| 451 |
+
"loss": 3.1269,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.66,
|
| 456 |
+
"learning_rate": 3.669646658152405e-05,
|
| 457 |
+
"loss": 3.1431,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.7,
|
| 462 |
+
"learning_rate": 3.651908613594437e-05,
|
| 463 |
+
"loss": 3.112,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.73,
|
| 468 |
+
"learning_rate": 3.6341705690364695e-05,
|
| 469 |
+
"loss": 3.1211,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.77,
|
| 474 |
+
"learning_rate": 3.616432524478502e-05,
|
| 475 |
+
"loss": 3.0896,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.8,
|
| 480 |
+
"learning_rate": 3.598694479920533e-05,
|
| 481 |
+
"loss": 3.0972,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.84,
|
| 486 |
+
"learning_rate": 3.5809564353625656e-05,
|
| 487 |
+
"loss": 3.0821,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.87,
|
| 492 |
+
"learning_rate": 3.563218390804598e-05,
|
| 493 |
+
"loss": 3.0626,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.91,
|
| 498 |
+
"learning_rate": 3.54548034624663e-05,
|
| 499 |
+
"loss": 3.0586,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.94,
|
| 504 |
+
"learning_rate": 3.5277423016886617e-05,
|
| 505 |
+
"loss": 3.0464,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.98,
|
| 510 |
+
"learning_rate": 3.510004257130694e-05,
|
| 511 |
+
"loss": 3.0368,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.02,
|
| 516 |
+
"learning_rate": 3.492266212572726e-05,
|
| 517 |
+
"loss": 3.0282,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.05,
|
| 522 |
+
"learning_rate": 3.4745281680147584e-05,
|
| 523 |
+
"loss": 2.9903,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 3.09,
|
| 528 |
+
"learning_rate": 3.45679012345679e-05,
|
| 529 |
+
"loss": 2.9912,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.12,
|
| 534 |
+
"learning_rate": 3.439052078898822e-05,
|
| 535 |
+
"loss": 2.9902,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 3.16,
|
| 540 |
+
"learning_rate": 3.4213140343408545e-05,
|
| 541 |
+
"loss": 2.995,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 3.19,
|
| 546 |
+
"learning_rate": 3.403575989782887e-05,
|
| 547 |
+
"loss": 2.9736,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 3.23,
|
| 552 |
+
"learning_rate": 3.385837945224918e-05,
|
| 553 |
+
"loss": 2.9724,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 3.26,
|
| 558 |
+
"learning_rate": 3.3680999006669506e-05,
|
| 559 |
+
"loss": 2.9636,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 3.3,
|
| 564 |
+
"learning_rate": 3.350361856108983e-05,
|
| 565 |
+
"loss": 2.957,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 3.33,
|
| 570 |
+
"learning_rate": 3.332623811551015e-05,
|
| 571 |
+
"loss": 2.9563,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.37,
|
| 576 |
+
"learning_rate": 3.3148857669930467e-05,
|
| 577 |
+
"loss": 2.9381,
|
| 578 |
+
"step": 47500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 3.41,
|
| 582 |
+
"learning_rate": 3.297147722435079e-05,
|
| 583 |
+
"loss": 2.926,
|
| 584 |
+
"step": 48000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.44,
|
| 588 |
+
"learning_rate": 3.279409677877111e-05,
|
| 589 |
+
"loss": 2.933,
|
| 590 |
+
"step": 48500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.48,
|
| 594 |
+
"learning_rate": 3.2616716333191434e-05,
|
| 595 |
+
"loss": 2.9144,
|
| 596 |
+
"step": 49000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 3.51,
|
| 600 |
+
"learning_rate": 3.243933588761175e-05,
|
| 601 |
+
"loss": 2.9084,
|
| 602 |
+
"step": 49500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.55,
|
| 606 |
+
"learning_rate": 3.226195544203207e-05,
|
| 607 |
+
"loss": 2.9116,
|
| 608 |
+
"step": 50000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.58,
|
| 612 |
+
"learning_rate": 3.2084574996452395e-05,
|
| 613 |
+
"loss": 2.8892,
|
| 614 |
+
"step": 50500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.62,
|
| 618 |
+
"learning_rate": 3.190719455087272e-05,
|
| 619 |
+
"loss": 2.8828,
|
| 620 |
+
"step": 51000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 3.65,
|
| 624 |
+
"learning_rate": 3.172981410529303e-05,
|
| 625 |
+
"loss": 2.8843,
|
| 626 |
+
"step": 51500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.69,
|
| 630 |
+
"learning_rate": 3.1552433659713356e-05,
|
| 631 |
+
"loss": 2.8799,
|
| 632 |
+
"step": 52000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.72,
|
| 636 |
+
"learning_rate": 3.137505321413368e-05,
|
| 637 |
+
"loss": 2.8586,
|
| 638 |
+
"step": 52500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 3.76,
|
| 642 |
+
"learning_rate": 3.1197672768554e-05,
|
| 643 |
+
"loss": 2.8601,
|
| 644 |
+
"step": 53000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 3.8,
|
| 648 |
+
"learning_rate": 3.1020292322974317e-05,
|
| 649 |
+
"loss": 2.8523,
|
| 650 |
+
"step": 53500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 3.83,
|
| 654 |
+
"learning_rate": 3.084291187739464e-05,
|
| 655 |
+
"loss": 2.8517,
|
| 656 |
+
"step": 54000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 3.87,
|
| 660 |
+
"learning_rate": 3.066553143181496e-05,
|
| 661 |
+
"loss": 2.8347,
|
| 662 |
+
"step": 54500
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 3.9,
|
| 666 |
+
"learning_rate": 3.048815098623528e-05,
|
| 667 |
+
"loss": 2.8427,
|
| 668 |
+
"step": 55000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.94,
|
| 672 |
+
"learning_rate": 3.0310770540655597e-05,
|
| 673 |
+
"loss": 2.8471,
|
| 674 |
+
"step": 55500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.97,
|
| 678 |
+
"learning_rate": 3.013339009507592e-05,
|
| 679 |
+
"loss": 2.8201,
|
| 680 |
+
"step": 56000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 4.01,
|
| 684 |
+
"learning_rate": 2.995600964949624e-05,
|
| 685 |
+
"loss": 2.8236,
|
| 686 |
+
"step": 56500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 4.04,
|
| 690 |
+
"learning_rate": 2.9778629203916564e-05,
|
| 691 |
+
"loss": 2.8043,
|
| 692 |
+
"step": 57000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 4.08,
|
| 696 |
+
"learning_rate": 2.960124875833688e-05,
|
| 697 |
+
"loss": 2.7957,
|
| 698 |
+
"step": 57500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 4.12,
|
| 702 |
+
"learning_rate": 2.9423868312757202e-05,
|
| 703 |
+
"loss": 2.8042,
|
| 704 |
+
"step": 58000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 4.15,
|
| 708 |
+
"learning_rate": 2.9246487867177525e-05,
|
| 709 |
+
"loss": 2.7836,
|
| 710 |
+
"step": 58500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 4.19,
|
| 714 |
+
"learning_rate": 2.9069107421597847e-05,
|
| 715 |
+
"loss": 2.7875,
|
| 716 |
+
"step": 59000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.22,
|
| 720 |
+
"learning_rate": 2.8891726976018163e-05,
|
| 721 |
+
"loss": 2.7737,
|
| 722 |
+
"step": 59500
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 4.26,
|
| 726 |
+
"learning_rate": 2.8714346530438486e-05,
|
| 727 |
+
"loss": 2.7783,
|
| 728 |
+
"step": 60000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 4.29,
|
| 732 |
+
"learning_rate": 2.8536966084858808e-05,
|
| 733 |
+
"loss": 2.7689,
|
| 734 |
+
"step": 60500
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.33,
|
| 738 |
+
"learning_rate": 2.835958563927913e-05,
|
| 739 |
+
"loss": 2.7689,
|
| 740 |
+
"step": 61000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 4.36,
|
| 744 |
+
"learning_rate": 2.8182205193699447e-05,
|
| 745 |
+
"loss": 2.7737,
|
| 746 |
+
"step": 61500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 4.4,
|
| 750 |
+
"learning_rate": 2.800482474811977e-05,
|
| 751 |
+
"loss": 2.7595,
|
| 752 |
+
"step": 62000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.43,
|
| 756 |
+
"learning_rate": 2.7827444302540088e-05,
|
| 757 |
+
"loss": 2.7657,
|
| 758 |
+
"step": 62500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 4.47,
|
| 762 |
+
"learning_rate": 2.765006385696041e-05,
|
| 763 |
+
"loss": 2.7664,
|
| 764 |
+
"step": 63000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 4.51,
|
| 768 |
+
"learning_rate": 2.747268341138073e-05,
|
| 769 |
+
"loss": 2.7394,
|
| 770 |
+
"step": 63500
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 4.54,
|
| 774 |
+
"learning_rate": 2.729530296580105e-05,
|
| 775 |
+
"loss": 2.7357,
|
| 776 |
+
"step": 64000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 4.58,
|
| 780 |
+
"learning_rate": 2.711792252022137e-05,
|
| 781 |
+
"loss": 2.7346,
|
| 782 |
+
"step": 64500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 4.61,
|
| 786 |
+
"learning_rate": 2.6940542074641694e-05,
|
| 787 |
+
"loss": 2.7351,
|
| 788 |
+
"step": 65000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 4.65,
|
| 792 |
+
"learning_rate": 2.676316162906201e-05,
|
| 793 |
+
"loss": 2.7268,
|
| 794 |
+
"step": 65500
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.68,
|
| 798 |
+
"learning_rate": 2.6585781183482332e-05,
|
| 799 |
+
"loss": 2.7319,
|
| 800 |
+
"step": 66000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.72,
|
| 804 |
+
"learning_rate": 2.6408400737902655e-05,
|
| 805 |
+
"loss": 2.7282,
|
| 806 |
+
"step": 66500
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 4.75,
|
| 810 |
+
"learning_rate": 2.6231020292322977e-05,
|
| 811 |
+
"loss": 2.7297,
|
| 812 |
+
"step": 67000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 4.79,
|
| 816 |
+
"learning_rate": 2.6053639846743293e-05,
|
| 817 |
+
"loss": 2.7112,
|
| 818 |
+
"step": 67500
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 4.82,
|
| 822 |
+
"learning_rate": 2.5876259401163616e-05,
|
| 823 |
+
"loss": 2.7315,
|
| 824 |
+
"step": 68000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 4.86,
|
| 828 |
+
"learning_rate": 2.5698878955583938e-05,
|
| 829 |
+
"loss": 2.704,
|
| 830 |
+
"step": 68500
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 4.9,
|
| 834 |
+
"learning_rate": 2.552149851000426e-05,
|
| 835 |
+
"loss": 2.7036,
|
| 836 |
+
"step": 69000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.93,
|
| 840 |
+
"learning_rate": 2.5344118064424577e-05,
|
| 841 |
+
"loss": 2.7103,
|
| 842 |
+
"step": 69500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.97,
|
| 846 |
+
"learning_rate": 2.51667376188449e-05,
|
| 847 |
+
"loss": 2.7022,
|
| 848 |
+
"step": 70000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 5.0,
|
| 852 |
+
"learning_rate": 2.498935717326522e-05,
|
| 853 |
+
"loss": 2.6949,
|
| 854 |
+
"step": 70500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 5.04,
|
| 858 |
+
"learning_rate": 2.481197672768554e-05,
|
| 859 |
+
"loss": 2.681,
|
| 860 |
+
"step": 71000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 5.07,
|
| 864 |
+
"learning_rate": 2.4634596282105863e-05,
|
| 865 |
+
"loss": 2.6636,
|
| 866 |
+
"step": 71500
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 5.11,
|
| 870 |
+
"learning_rate": 2.4457215836526182e-05,
|
| 871 |
+
"loss": 2.6868,
|
| 872 |
+
"step": 72000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 5.14,
|
| 876 |
+
"learning_rate": 2.4279835390946505e-05,
|
| 877 |
+
"loss": 2.6683,
|
| 878 |
+
"step": 72500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 5.18,
|
| 882 |
+
"learning_rate": 2.4102454945366824e-05,
|
| 883 |
+
"loss": 2.6765,
|
| 884 |
+
"step": 73000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 5.21,
|
| 888 |
+
"learning_rate": 2.3925074499787147e-05,
|
| 889 |
+
"loss": 2.6665,
|
| 890 |
+
"step": 73500
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 5.25,
|
| 894 |
+
"learning_rate": 2.3747694054207466e-05,
|
| 895 |
+
"loss": 2.6672,
|
| 896 |
+
"step": 74000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 5.29,
|
| 900 |
+
"learning_rate": 2.3570313608627785e-05,
|
| 901 |
+
"loss": 2.6604,
|
| 902 |
+
"step": 74500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 5.32,
|
| 906 |
+
"learning_rate": 2.3392933163048107e-05,
|
| 907 |
+
"loss": 2.6552,
|
| 908 |
+
"step": 75000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 5.36,
|
| 912 |
+
"learning_rate": 2.3215552717468427e-05,
|
| 913 |
+
"loss": 2.6561,
|
| 914 |
+
"step": 75500
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 5.39,
|
| 918 |
+
"learning_rate": 2.3038172271888746e-05,
|
| 919 |
+
"loss": 2.6418,
|
| 920 |
+
"step": 76000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.43,
|
| 924 |
+
"learning_rate": 2.2860791826309068e-05,
|
| 925 |
+
"loss": 2.6552,
|
| 926 |
+
"step": 76500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 5.46,
|
| 930 |
+
"learning_rate": 2.2683411380729387e-05,
|
| 931 |
+
"loss": 2.6404,
|
| 932 |
+
"step": 77000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 5.5,
|
| 936 |
+
"learning_rate": 2.250603093514971e-05,
|
| 937 |
+
"loss": 2.6317,
|
| 938 |
+
"step": 77500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 5.53,
|
| 942 |
+
"learning_rate": 2.232865048957003e-05,
|
| 943 |
+
"loss": 2.6343,
|
| 944 |
+
"step": 78000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 5.57,
|
| 948 |
+
"learning_rate": 2.215127004399035e-05,
|
| 949 |
+
"loss": 2.6446,
|
| 950 |
+
"step": 78500
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 5.61,
|
| 954 |
+
"learning_rate": 2.197388959841067e-05,
|
| 955 |
+
"loss": 2.6201,
|
| 956 |
+
"step": 79000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 5.64,
|
| 960 |
+
"learning_rate": 2.1796509152830993e-05,
|
| 961 |
+
"loss": 2.6266,
|
| 962 |
+
"step": 79500
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 5.68,
|
| 966 |
+
"learning_rate": 2.1619128707251312e-05,
|
| 967 |
+
"loss": 2.6229,
|
| 968 |
+
"step": 80000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 5.71,
|
| 972 |
+
"learning_rate": 2.1441748261671635e-05,
|
| 973 |
+
"loss": 2.6297,
|
| 974 |
+
"step": 80500
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 5.75,
|
| 978 |
+
"learning_rate": 2.1264367816091954e-05,
|
| 979 |
+
"loss": 2.6183,
|
| 980 |
+
"step": 81000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 5.78,
|
| 984 |
+
"learning_rate": 2.1086987370512277e-05,
|
| 985 |
+
"loss": 2.6183,
|
| 986 |
+
"step": 81500
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 5.82,
|
| 990 |
+
"learning_rate": 2.0909606924932596e-05,
|
| 991 |
+
"loss": 2.6164,
|
| 992 |
+
"step": 82000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 5.85,
|
| 996 |
+
"learning_rate": 2.073222647935292e-05,
|
| 997 |
+
"loss": 2.6069,
|
| 998 |
+
"step": 82500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 5.89,
|
| 1002 |
+
"learning_rate": 2.0554846033773237e-05,
|
| 1003 |
+
"loss": 2.6219,
|
| 1004 |
+
"step": 83000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 5.92,
|
| 1008 |
+
"learning_rate": 2.037746558819356e-05,
|
| 1009 |
+
"loss": 2.592,
|
| 1010 |
+
"step": 83500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.96,
|
| 1014 |
+
"learning_rate": 2.020008514261388e-05,
|
| 1015 |
+
"loss": 2.5975,
|
| 1016 |
+
"step": 84000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 6.0,
|
| 1020 |
+
"learning_rate": 2.00227046970342e-05,
|
| 1021 |
+
"loss": 2.6029,
|
| 1022 |
+
"step": 84500
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 6.03,
|
| 1026 |
+
"learning_rate": 1.984532425145452e-05,
|
| 1027 |
+
"loss": 2.5913,
|
| 1028 |
+
"step": 85000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 6.07,
|
| 1032 |
+
"learning_rate": 1.9667943805874843e-05,
|
| 1033 |
+
"loss": 2.5867,
|
| 1034 |
+
"step": 85500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 6.1,
|
| 1038 |
+
"learning_rate": 1.9490563360295162e-05,
|
| 1039 |
+
"loss": 2.5899,
|
| 1040 |
+
"step": 86000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 6.14,
|
| 1044 |
+
"learning_rate": 1.931318291471548e-05,
|
| 1045 |
+
"loss": 2.5923,
|
| 1046 |
+
"step": 86500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 6.17,
|
| 1050 |
+
"learning_rate": 1.91358024691358e-05,
|
| 1051 |
+
"loss": 2.5742,
|
| 1052 |
+
"step": 87000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 6.21,
|
| 1056 |
+
"learning_rate": 1.8958422023556123e-05,
|
| 1057 |
+
"loss": 2.5859,
|
| 1058 |
+
"step": 87500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 6.24,
|
| 1062 |
+
"learning_rate": 1.8781041577976442e-05,
|
| 1063 |
+
"loss": 2.5683,
|
| 1064 |
+
"step": 88000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 6.28,
|
| 1068 |
+
"learning_rate": 1.8603661132396765e-05,
|
| 1069 |
+
"loss": 2.5808,
|
| 1070 |
+
"step": 88500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 6.31,
|
| 1074 |
+
"learning_rate": 1.8426280686817084e-05,
|
| 1075 |
+
"loss": 2.5663,
|
| 1076 |
+
"step": 89000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 6.35,
|
| 1080 |
+
"learning_rate": 1.8248900241237407e-05,
|
| 1081 |
+
"loss": 2.5677,
|
| 1082 |
+
"step": 89500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 6.39,
|
| 1086 |
+
"learning_rate": 1.8071519795657726e-05,
|
| 1087 |
+
"loss": 2.5647,
|
| 1088 |
+
"step": 90000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 6.42,
|
| 1092 |
+
"learning_rate": 1.7894139350078048e-05,
|
| 1093 |
+
"loss": 2.5656,
|
| 1094 |
+
"step": 90500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 6.46,
|
| 1098 |
+
"learning_rate": 1.7716758904498367e-05,
|
| 1099 |
+
"loss": 2.5682,
|
| 1100 |
+
"step": 91000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 6.49,
|
| 1104 |
+
"learning_rate": 1.753937845891869e-05,
|
| 1105 |
+
"loss": 2.5645,
|
| 1106 |
+
"step": 91500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 6.53,
|
| 1110 |
+
"learning_rate": 1.736199801333901e-05,
|
| 1111 |
+
"loss": 2.5586,
|
| 1112 |
+
"step": 92000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 6.56,
|
| 1116 |
+
"learning_rate": 1.718461756775933e-05,
|
| 1117 |
+
"loss": 2.5488,
|
| 1118 |
+
"step": 92500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 6.6,
|
| 1122 |
+
"learning_rate": 1.700723712217965e-05,
|
| 1123 |
+
"loss": 2.5547,
|
| 1124 |
+
"step": 93000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 6.63,
|
| 1128 |
+
"learning_rate": 1.6829856676599973e-05,
|
| 1129 |
+
"loss": 2.5485,
|
| 1130 |
+
"step": 93500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 6.67,
|
| 1134 |
+
"learning_rate": 1.6652476231020292e-05,
|
| 1135 |
+
"loss": 2.5503,
|
| 1136 |
+
"step": 94000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 6.7,
|
| 1140 |
+
"learning_rate": 1.6475095785440615e-05,
|
| 1141 |
+
"loss": 2.547,
|
| 1142 |
+
"step": 94500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 6.74,
|
| 1146 |
+
"learning_rate": 1.6297715339860934e-05,
|
| 1147 |
+
"loss": 2.5468,
|
| 1148 |
+
"step": 95000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 6.78,
|
| 1152 |
+
"learning_rate": 1.6120334894281257e-05,
|
| 1153 |
+
"loss": 2.5416,
|
| 1154 |
+
"step": 95500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 6.81,
|
| 1158 |
+
"learning_rate": 1.5942954448701576e-05,
|
| 1159 |
+
"loss": 2.5562,
|
| 1160 |
+
"step": 96000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 6.85,
|
| 1164 |
+
"learning_rate": 1.57655740031219e-05,
|
| 1165 |
+
"loss": 2.5311,
|
| 1166 |
+
"step": 96500
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 6.88,
|
| 1170 |
+
"learning_rate": 1.5588193557542217e-05,
|
| 1171 |
+
"loss": 2.5481,
|
| 1172 |
+
"step": 97000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 6.92,
|
| 1176 |
+
"learning_rate": 1.5410813111962537e-05,
|
| 1177 |
+
"loss": 2.5362,
|
| 1178 |
+
"step": 97500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.95,
|
| 1182 |
+
"learning_rate": 1.5233432666382857e-05,
|
| 1183 |
+
"loss": 2.5161,
|
| 1184 |
+
"step": 98000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 6.99,
|
| 1188 |
+
"learning_rate": 1.505605222080318e-05,
|
| 1189 |
+
"loss": 2.5354,
|
| 1190 |
+
"step": 98500
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 7.02,
|
| 1194 |
+
"learning_rate": 1.4878671775223499e-05,
|
| 1195 |
+
"loss": 2.5214,
|
| 1196 |
+
"step": 99000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 7.06,
|
| 1200 |
+
"learning_rate": 1.4701291329643822e-05,
|
| 1201 |
+
"loss": 2.5271,
|
| 1202 |
+
"step": 99500
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 7.1,
|
| 1206 |
+
"learning_rate": 1.452391088406414e-05,
|
| 1207 |
+
"loss": 2.5024,
|
| 1208 |
+
"step": 100000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 7.13,
|
| 1212 |
+
"learning_rate": 1.4346530438484462e-05,
|
| 1213 |
+
"loss": 2.5096,
|
| 1214 |
+
"step": 100500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 7.17,
|
| 1218 |
+
"learning_rate": 1.4169149992904782e-05,
|
| 1219 |
+
"loss": 2.5094,
|
| 1220 |
+
"step": 101000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 7.2,
|
| 1224 |
+
"learning_rate": 1.3991769547325103e-05,
|
| 1225 |
+
"loss": 2.5151,
|
| 1226 |
+
"step": 101500
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 7.24,
|
| 1230 |
+
"learning_rate": 1.3814389101745422e-05,
|
| 1231 |
+
"loss": 2.5141,
|
| 1232 |
+
"step": 102000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 7.27,
|
| 1236 |
+
"learning_rate": 1.3637008656165745e-05,
|
| 1237 |
+
"loss": 2.5102,
|
| 1238 |
+
"step": 102500
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 7.31,
|
| 1242 |
+
"learning_rate": 1.3459628210586064e-05,
|
| 1243 |
+
"loss": 2.5091,
|
| 1244 |
+
"step": 103000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 7.34,
|
| 1248 |
+
"learning_rate": 1.3282247765006387e-05,
|
| 1249 |
+
"loss": 2.5025,
|
| 1250 |
+
"step": 103500
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 7.38,
|
| 1254 |
+
"learning_rate": 1.3104867319426706e-05,
|
| 1255 |
+
"loss": 2.5025,
|
| 1256 |
+
"step": 104000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 7.41,
|
| 1260 |
+
"learning_rate": 1.2927486873847028e-05,
|
| 1261 |
+
"loss": 2.5051,
|
| 1262 |
+
"step": 104500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 7.45,
|
| 1266 |
+
"learning_rate": 1.2750106428267347e-05,
|
| 1267 |
+
"loss": 2.5084,
|
| 1268 |
+
"step": 105000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 7.49,
|
| 1272 |
+
"learning_rate": 1.257272598268767e-05,
|
| 1273 |
+
"loss": 2.4955,
|
| 1274 |
+
"step": 105500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 7.52,
|
| 1278 |
+
"learning_rate": 1.2395345537107989e-05,
|
| 1279 |
+
"loss": 2.4915,
|
| 1280 |
+
"step": 106000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 7.56,
|
| 1284 |
+
"learning_rate": 1.221796509152831e-05,
|
| 1285 |
+
"loss": 2.5053,
|
| 1286 |
+
"step": 106500
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 7.59,
|
| 1290 |
+
"learning_rate": 1.204058464594863e-05,
|
| 1291 |
+
"loss": 2.4989,
|
| 1292 |
+
"step": 107000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 7.63,
|
| 1296 |
+
"learning_rate": 1.1863204200368952e-05,
|
| 1297 |
+
"loss": 2.4957,
|
| 1298 |
+
"step": 107500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 7.66,
|
| 1302 |
+
"learning_rate": 1.1685823754789272e-05,
|
| 1303 |
+
"loss": 2.4909,
|
| 1304 |
+
"step": 108000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 7.7,
|
| 1308 |
+
"learning_rate": 1.1508443309209593e-05,
|
| 1309 |
+
"loss": 2.492,
|
| 1310 |
+
"step": 108500
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 7.73,
|
| 1314 |
+
"learning_rate": 1.1331062863629914e-05,
|
| 1315 |
+
"loss": 2.4929,
|
| 1316 |
+
"step": 109000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 7.77,
|
| 1320 |
+
"learning_rate": 1.1153682418050235e-05,
|
| 1321 |
+
"loss": 2.4871,
|
| 1322 |
+
"step": 109500
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 7.8,
|
| 1326 |
+
"learning_rate": 1.0976301972470556e-05,
|
| 1327 |
+
"loss": 2.4843,
|
| 1328 |
+
"step": 110000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 7.84,
|
| 1332 |
+
"learning_rate": 1.0798921526890877e-05,
|
| 1333 |
+
"loss": 2.4849,
|
| 1334 |
+
"step": 110500
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 7.88,
|
| 1338 |
+
"learning_rate": 1.0621541081311197e-05,
|
| 1339 |
+
"loss": 2.4903,
|
| 1340 |
+
"step": 111000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 7.91,
|
| 1344 |
+
"learning_rate": 1.0444160635731518e-05,
|
| 1345 |
+
"loss": 2.4766,
|
| 1346 |
+
"step": 111500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 7.95,
|
| 1350 |
+
"learning_rate": 1.0266780190151837e-05,
|
| 1351 |
+
"loss": 2.4895,
|
| 1352 |
+
"step": 112000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 7.98,
|
| 1356 |
+
"learning_rate": 1.0089399744572158e-05,
|
| 1357 |
+
"loss": 2.4818,
|
| 1358 |
+
"step": 112500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 8.02,
|
| 1362 |
+
"learning_rate": 9.912019298992479e-06,
|
| 1363 |
+
"loss": 2.4782,
|
| 1364 |
+
"step": 113000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 8.05,
|
| 1368 |
+
"learning_rate": 9.7346388534128e-06,
|
| 1369 |
+
"loss": 2.4671,
|
| 1370 |
+
"step": 113500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 8.09,
|
| 1374 |
+
"learning_rate": 9.55725840783312e-06,
|
| 1375 |
+
"loss": 2.4665,
|
| 1376 |
+
"step": 114000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 8.12,
|
| 1380 |
+
"learning_rate": 9.379877962253442e-06,
|
| 1381 |
+
"loss": 2.4669,
|
| 1382 |
+
"step": 114500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 8.16,
|
| 1386 |
+
"learning_rate": 9.202497516673762e-06,
|
| 1387 |
+
"loss": 2.4552,
|
| 1388 |
+
"step": 115000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 8.19,
|
| 1392 |
+
"learning_rate": 9.025117071094083e-06,
|
| 1393 |
+
"loss": 2.459,
|
| 1394 |
+
"step": 115500
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 8.23,
|
| 1398 |
+
"learning_rate": 8.847736625514404e-06,
|
| 1399 |
+
"loss": 2.455,
|
| 1400 |
+
"step": 116000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 8.27,
|
| 1404 |
+
"learning_rate": 8.670356179934725e-06,
|
| 1405 |
+
"loss": 2.472,
|
| 1406 |
+
"step": 116500
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 8.3,
|
| 1410 |
+
"learning_rate": 8.492975734355046e-06,
|
| 1411 |
+
"loss": 2.4517,
|
| 1412 |
+
"step": 117000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 8.34,
|
| 1416 |
+
"learning_rate": 8.315595288775365e-06,
|
| 1417 |
+
"loss": 2.458,
|
| 1418 |
+
"step": 117500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 8.37,
|
| 1422 |
+
"learning_rate": 8.138214843195686e-06,
|
| 1423 |
+
"loss": 2.4597,
|
| 1424 |
+
"step": 118000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 8.41,
|
| 1428 |
+
"learning_rate": 7.960834397616007e-06,
|
| 1429 |
+
"loss": 2.4567,
|
| 1430 |
+
"step": 118500
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 8.44,
|
| 1434 |
+
"learning_rate": 7.783453952036327e-06,
|
| 1435 |
+
"loss": 2.4561,
|
| 1436 |
+
"step": 119000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 8.48,
|
| 1440 |
+
"learning_rate": 7.606073506456648e-06,
|
| 1441 |
+
"loss": 2.4614,
|
| 1442 |
+
"step": 119500
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 8.51,
|
| 1446 |
+
"learning_rate": 7.428693060876969e-06,
|
| 1447 |
+
"loss": 2.4592,
|
| 1448 |
+
"step": 120000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 8.55,
|
| 1452 |
+
"learning_rate": 7.25131261529729e-06,
|
| 1453 |
+
"loss": 2.456,
|
| 1454 |
+
"step": 120500
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 8.59,
|
| 1458 |
+
"learning_rate": 7.073932169717611e-06,
|
| 1459 |
+
"loss": 2.4593,
|
| 1460 |
+
"step": 121000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 8.62,
|
| 1464 |
+
"learning_rate": 6.896551724137932e-06,
|
| 1465 |
+
"loss": 2.452,
|
| 1466 |
+
"step": 121500
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 8.66,
|
| 1470 |
+
"learning_rate": 6.719171278558252e-06,
|
| 1471 |
+
"loss": 2.4414,
|
| 1472 |
+
"step": 122000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 8.69,
|
| 1476 |
+
"learning_rate": 6.5417908329785725e-06,
|
| 1477 |
+
"loss": 2.4558,
|
| 1478 |
+
"step": 122500
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 8.73,
|
| 1482 |
+
"learning_rate": 6.364410387398893e-06,
|
| 1483 |
+
"loss": 2.4417,
|
| 1484 |
+
"step": 123000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 8.76,
|
| 1488 |
+
"learning_rate": 6.187029941819214e-06,
|
| 1489 |
+
"loss": 2.4364,
|
| 1490 |
+
"step": 123500
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 8.8,
|
| 1494 |
+
"learning_rate": 6.009649496239535e-06,
|
| 1495 |
+
"loss": 2.4543,
|
| 1496 |
+
"step": 124000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 8.83,
|
| 1500 |
+
"learning_rate": 5.832269050659856e-06,
|
| 1501 |
+
"loss": 2.4419,
|
| 1502 |
+
"step": 124500
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 8.87,
|
| 1506 |
+
"learning_rate": 5.654888605080176e-06,
|
| 1507 |
+
"loss": 2.446,
|
| 1508 |
+
"step": 125000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 8.9,
|
| 1512 |
+
"learning_rate": 5.477508159500497e-06,
|
| 1513 |
+
"loss": 2.4395,
|
| 1514 |
+
"step": 125500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 8.94,
|
| 1518 |
+
"learning_rate": 5.3001277139208175e-06,
|
| 1519 |
+
"loss": 2.4466,
|
| 1520 |
+
"step": 126000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 8.98,
|
| 1524 |
+
"learning_rate": 5.122747268341138e-06,
|
| 1525 |
+
"loss": 2.4301,
|
| 1526 |
+
"step": 126500
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 9.01,
|
| 1530 |
+
"learning_rate": 4.945366822761459e-06,
|
| 1531 |
+
"loss": 2.4382,
|
| 1532 |
+
"step": 127000
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 9.05,
|
| 1536 |
+
"learning_rate": 4.767986377181779e-06,
|
| 1537 |
+
"loss": 2.4406,
|
| 1538 |
+
"step": 127500
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 9.08,
|
| 1542 |
+
"learning_rate": 4.5906059316021e-06,
|
| 1543 |
+
"loss": 2.4312,
|
| 1544 |
+
"step": 128000
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 9.12,
|
| 1548 |
+
"learning_rate": 4.413225486022421e-06,
|
| 1549 |
+
"loss": 2.4409,
|
| 1550 |
+
"step": 128500
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 9.15,
|
| 1554 |
+
"learning_rate": 4.235845040442742e-06,
|
| 1555 |
+
"loss": 2.4256,
|
| 1556 |
+
"step": 129000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 9.19,
|
| 1560 |
+
"learning_rate": 4.0584645948630625e-06,
|
| 1561 |
+
"loss": 2.4301,
|
| 1562 |
+
"step": 129500
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 9.22,
|
| 1566 |
+
"learning_rate": 3.881084149283383e-06,
|
| 1567 |
+
"loss": 2.4338,
|
| 1568 |
+
"step": 130000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 9.26,
|
| 1572 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 1573 |
+
"loss": 2.4367,
|
| 1574 |
+
"step": 130500
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 9.29,
|
| 1578 |
+
"learning_rate": 3.5263232581240246e-06,
|
| 1579 |
+
"loss": 2.4198,
|
| 1580 |
+
"step": 131000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 9.33,
|
| 1584 |
+
"learning_rate": 3.348942812544345e-06,
|
| 1585 |
+
"loss": 2.4228,
|
| 1586 |
+
"step": 131500
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 9.37,
|
| 1590 |
+
"learning_rate": 3.171562366964666e-06,
|
| 1591 |
+
"loss": 2.4257,
|
| 1592 |
+
"step": 132000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 9.4,
|
| 1596 |
+
"learning_rate": 2.9941819213849867e-06,
|
| 1597 |
+
"loss": 2.4286,
|
| 1598 |
+
"step": 132500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 9.44,
|
| 1602 |
+
"learning_rate": 2.8168014758053075e-06,
|
| 1603 |
+
"loss": 2.4153,
|
| 1604 |
+
"step": 133000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 9.47,
|
| 1608 |
+
"learning_rate": 2.6394210302256283e-06,
|
| 1609 |
+
"loss": 2.4235,
|
| 1610 |
+
"step": 133500
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 9.51,
|
| 1614 |
+
"learning_rate": 2.4620405846459487e-06,
|
| 1615 |
+
"loss": 2.4243,
|
| 1616 |
+
"step": 134000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 9.54,
|
| 1620 |
+
"learning_rate": 2.2846601390662696e-06,
|
| 1621 |
+
"loss": 2.4235,
|
| 1622 |
+
"step": 134500
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 9.58,
|
| 1626 |
+
"learning_rate": 2.1072796934865904e-06,
|
| 1627 |
+
"loss": 2.4297,
|
| 1628 |
+
"step": 135000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 9.61,
|
| 1632 |
+
"learning_rate": 1.929899247906911e-06,
|
| 1633 |
+
"loss": 2.4356,
|
| 1634 |
+
"step": 135500
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 9.65,
|
| 1638 |
+
"learning_rate": 1.7525188023272317e-06,
|
| 1639 |
+
"loss": 2.4223,
|
| 1640 |
+
"step": 136000
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 9.68,
|
| 1644 |
+
"learning_rate": 1.5751383567475523e-06,
|
| 1645 |
+
"loss": 2.4218,
|
| 1646 |
+
"step": 136500
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 9.72,
|
| 1650 |
+
"learning_rate": 1.397757911167873e-06,
|
| 1651 |
+
"loss": 2.4211,
|
| 1652 |
+
"step": 137000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 9.76,
|
| 1656 |
+
"learning_rate": 1.2203774655881937e-06,
|
| 1657 |
+
"loss": 2.4233,
|
| 1658 |
+
"step": 137500
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 9.79,
|
| 1662 |
+
"learning_rate": 1.0429970200085144e-06,
|
| 1663 |
+
"loss": 2.4107,
|
| 1664 |
+
"step": 138000
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 9.83,
|
| 1668 |
+
"learning_rate": 8.65616574428835e-07,
|
| 1669 |
+
"loss": 2.4176,
|
| 1670 |
+
"step": 138500
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 9.86,
|
| 1674 |
+
"learning_rate": 6.882361288491557e-07,
|
| 1675 |
+
"loss": 2.4227,
|
| 1676 |
+
"step": 139000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 9.9,
|
| 1680 |
+
"learning_rate": 5.108556832694765e-07,
|
| 1681 |
+
"loss": 2.4229,
|
| 1682 |
+
"step": 139500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 9.93,
|
| 1686 |
+
"learning_rate": 3.334752376897971e-07,
|
| 1687 |
+
"loss": 2.4089,
|
| 1688 |
+
"step": 140000
|
| 1689 |
+
}
|
| 1690 |
+
],
|
| 1691 |
+
"max_steps": 140940,
|
| 1692 |
+
"num_train_epochs": 10,
|
| 1693 |
+
"total_flos": 690009058341027840,
|
| 1694 |
+
"trial_name": null,
|
| 1695 |
+
"trial_params": null
|
| 1696 |
+
}
|
checkpoint-140000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d218969f864d8514dc3a66fde3f1475f190986fb24b2e4acf5d7a01940a6be8
|
| 3 |
+
size 2031
|
config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"gradient_checkpointing": false,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.1,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 3072,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"num_hidden_layers": 12,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"transformers_version": "4.2.2",
|
| 20 |
+
"type_vocab_size": 2,
|
| 21 |
+
"use_cache": true,
|
| 22 |
+
"vocab_size": 32000
|
| 23 |
+
}
|
hasil/config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"gradient_checkpointing": false,
|
| 7 |
+
"hidden_act": "gelu",
|
| 8 |
+
"hidden_dropout_prob": 0.1,
|
| 9 |
+
"hidden_size": 768,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 3072,
|
| 12 |
+
"layer_norm_eps": 1e-12,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "bert",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"num_hidden_layers": 12,
|
| 17 |
+
"pad_token_id": 0,
|
| 18 |
+
"position_embedding_type": "absolute",
|
| 19 |
+
"transformers_version": "4.2.2",
|
| 20 |
+
"type_vocab_size": 2,
|
| 21 |
+
"use_cache": true,
|
| 22 |
+
"vocab_size": 32000
|
| 23 |
+
}
|
hasil/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74d1764324ec9ee3cf1fbbca6cf8521fe8fc2f74b82a215431e1123ca7067651
|
| 3 |
+
size 442693650
|
hasil/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d218969f864d8514dc3a66fde3f1475f190986fb24b2e4acf5d7a01940a6be8
|
| 3 |
+
size 2031
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74d1764324ec9ee3cf1fbbca6cf8521fe8fc2f74b82a215431e1123ca7067651
|
| 3 |
+
size 442693650
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d218969f864d8514dc3a66fde3f1475f190986fb24b2e4acf5d7a01940a6be8
|
| 3 |
+
size 2031
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|