Commit
•
4ecd0e8
1
Parent(s):
a6eff7d
upload all files
Browse files- config.json +28 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2716 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"AlbertForMaskedLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0,
|
| 6 |
+
"bos_token_id": 2,
|
| 7 |
+
"classifier_dropout_prob": 0.1,
|
| 8 |
+
"embedding_size": 128,
|
| 9 |
+
"eos_token_id": 3,
|
| 10 |
+
"hidden_act": "gelu_new",
|
| 11 |
+
"hidden_dropout_prob": 0,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"inner_group_num": 1,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"layer_norm_eps": 1e-12,
|
| 17 |
+
"max_position_embeddings": 512,
|
| 18 |
+
"model_type": "albert",
|
| 19 |
+
"num_attention_heads": 12,
|
| 20 |
+
"num_hidden_groups": 1,
|
| 21 |
+
"num_hidden_layers": 12,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"position_embedding_type": "absolute",
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.10.2",
|
| 26 |
+
"type_vocab_size": 2,
|
| 27 |
+
"vocab_size": 50000
|
| 28 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:320ecb6c370355f4132d1aee6ae901335d5ed594aef50639729e0b80fe65b271
|
| 3 |
+
size 110451753
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2cd4e7b52c6e7bcd940fe7956f27a1e3cf04f3ccb048fbfeaa598f2df081df8a
|
| 3 |
+
size 23620
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0281fdaf239747dc383c1a97cc01b12944a91412f6ff7adc23490947460f6ab
|
| 3 |
+
size 559
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2716 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.079355858472596,
|
| 5 |
+
"global_step": 900000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 13 |
+
"loss": 10.6946,
|
| 14 |
+
"step": 2000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 19 |
+
"loss": 10.5675,
|
| 20 |
+
"step": 4000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 6.666666666666667e-05,
|
| 25 |
+
"loss": 10.4211,
|
| 26 |
+
"step": 6000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 8.888888888888889e-05,
|
| 31 |
+
"loss": 10.2041,
|
| 32 |
+
"step": 8000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.07,
|
| 36 |
+
"learning_rate": 9.988776655443322e-05,
|
| 37 |
+
"loss": 9.8916,
|
| 38 |
+
"step": 10000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08,
|
| 42 |
+
"learning_rate": 9.966329966329967e-05,
|
| 43 |
+
"loss": 9.4948,
|
| 44 |
+
"step": 12000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09,
|
| 48 |
+
"learning_rate": 9.943883277216612e-05,
|
| 49 |
+
"loss": 9.0658,
|
| 50 |
+
"step": 14000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.11,
|
| 54 |
+
"learning_rate": 9.921436588103256e-05,
|
| 55 |
+
"loss": 8.6284,
|
| 56 |
+
"step": 16000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.12,
|
| 60 |
+
"learning_rate": 9.8989898989899e-05,
|
| 61 |
+
"loss": 8.1946,
|
| 62 |
+
"step": 18000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.14,
|
| 66 |
+
"learning_rate": 9.876543209876543e-05,
|
| 67 |
+
"loss": 7.801,
|
| 68 |
+
"step": 20000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.15,
|
| 72 |
+
"learning_rate": 9.854096520763188e-05,
|
| 73 |
+
"loss": 7.4541,
|
| 74 |
+
"step": 22000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.16,
|
| 78 |
+
"learning_rate": 9.831649831649832e-05,
|
| 79 |
+
"loss": 7.283,
|
| 80 |
+
"step": 24000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.18,
|
| 84 |
+
"learning_rate": 9.809203142536476e-05,
|
| 85 |
+
"loss": 7.2067,
|
| 86 |
+
"step": 26000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.19,
|
| 90 |
+
"learning_rate": 9.78675645342312e-05,
|
| 91 |
+
"loss": 7.1146,
|
| 92 |
+
"step": 28000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.2,
|
| 96 |
+
"learning_rate": 9.764309764309764e-05,
|
| 97 |
+
"loss": 7.0541,
|
| 98 |
+
"step": 30000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.22,
|
| 102 |
+
"learning_rate": 9.74186307519641e-05,
|
| 103 |
+
"loss": 6.9996,
|
| 104 |
+
"step": 32000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.23,
|
| 108 |
+
"learning_rate": 9.719416386083054e-05,
|
| 109 |
+
"loss": 6.967,
|
| 110 |
+
"step": 34000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.24,
|
| 114 |
+
"learning_rate": 9.696969696969698e-05,
|
| 115 |
+
"loss": 6.9272,
|
| 116 |
+
"step": 36000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.26,
|
| 120 |
+
"learning_rate": 9.674523007856342e-05,
|
| 121 |
+
"loss": 6.9136,
|
| 122 |
+
"step": 38000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.27,
|
| 126 |
+
"learning_rate": 9.652076318742987e-05,
|
| 127 |
+
"loss": 6.8708,
|
| 128 |
+
"step": 40000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.28,
|
| 132 |
+
"learning_rate": 9.62962962962963e-05,
|
| 133 |
+
"loss": 6.8544,
|
| 134 |
+
"step": 42000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.3,
|
| 138 |
+
"learning_rate": 9.607182940516274e-05,
|
| 139 |
+
"loss": 6.8332,
|
| 140 |
+
"step": 44000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.31,
|
| 144 |
+
"learning_rate": 9.584736251402918e-05,
|
| 145 |
+
"loss": 6.8299,
|
| 146 |
+
"step": 46000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.32,
|
| 150 |
+
"learning_rate": 9.562289562289563e-05,
|
| 151 |
+
"loss": 6.806,
|
| 152 |
+
"step": 48000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.34,
|
| 156 |
+
"learning_rate": 9.539842873176207e-05,
|
| 157 |
+
"loss": 6.7812,
|
| 158 |
+
"step": 50000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.35,
|
| 162 |
+
"learning_rate": 9.517396184062851e-05,
|
| 163 |
+
"loss": 6.7821,
|
| 164 |
+
"step": 52000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.36,
|
| 168 |
+
"learning_rate": 9.494949494949495e-05,
|
| 169 |
+
"loss": 6.756,
|
| 170 |
+
"step": 54000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.38,
|
| 174 |
+
"learning_rate": 9.472502805836139e-05,
|
| 175 |
+
"loss": 6.7547,
|
| 176 |
+
"step": 56000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.39,
|
| 180 |
+
"learning_rate": 9.450056116722785e-05,
|
| 181 |
+
"loss": 6.7199,
|
| 182 |
+
"step": 58000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.41,
|
| 186 |
+
"learning_rate": 9.427609427609429e-05,
|
| 187 |
+
"loss": 6.6967,
|
| 188 |
+
"step": 60000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.42,
|
| 192 |
+
"learning_rate": 9.405162738496073e-05,
|
| 193 |
+
"loss": 6.6726,
|
| 194 |
+
"step": 62000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.43,
|
| 198 |
+
"learning_rate": 9.382716049382717e-05,
|
| 199 |
+
"loss": 6.6519,
|
| 200 |
+
"step": 64000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.45,
|
| 204 |
+
"learning_rate": 9.36026936026936e-05,
|
| 205 |
+
"loss": 6.6244,
|
| 206 |
+
"step": 66000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.46,
|
| 210 |
+
"learning_rate": 9.337822671156005e-05,
|
| 211 |
+
"loss": 6.5836,
|
| 212 |
+
"step": 68000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.47,
|
| 216 |
+
"learning_rate": 9.315375982042649e-05,
|
| 217 |
+
"loss": 6.5503,
|
| 218 |
+
"step": 70000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.49,
|
| 222 |
+
"learning_rate": 9.292929292929293e-05,
|
| 223 |
+
"loss": 6.5117,
|
| 224 |
+
"step": 72000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.5,
|
| 228 |
+
"learning_rate": 9.270482603815938e-05,
|
| 229 |
+
"loss": 6.4852,
|
| 230 |
+
"step": 74000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.51,
|
| 234 |
+
"learning_rate": 9.248035914702582e-05,
|
| 235 |
+
"loss": 6.4517,
|
| 236 |
+
"step": 76000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.53,
|
| 240 |
+
"learning_rate": 9.225589225589226e-05,
|
| 241 |
+
"loss": 6.405,
|
| 242 |
+
"step": 78000
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.54,
|
| 246 |
+
"learning_rate": 9.20314253647587e-05,
|
| 247 |
+
"loss": 6.3787,
|
| 248 |
+
"step": 80000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.55,
|
| 252 |
+
"learning_rate": 9.180695847362514e-05,
|
| 253 |
+
"loss": 6.3171,
|
| 254 |
+
"step": 82000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.57,
|
| 258 |
+
"learning_rate": 9.15824915824916e-05,
|
| 259 |
+
"loss": 6.2806,
|
| 260 |
+
"step": 84000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.58,
|
| 264 |
+
"learning_rate": 9.135802469135802e-05,
|
| 265 |
+
"loss": 6.3758,
|
| 266 |
+
"step": 86000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.59,
|
| 270 |
+
"learning_rate": 9.113355780022446e-05,
|
| 271 |
+
"loss": 6.3753,
|
| 272 |
+
"step": 88000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.61,
|
| 276 |
+
"learning_rate": 9.090909090909092e-05,
|
| 277 |
+
"loss": 6.2428,
|
| 278 |
+
"step": 90000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.62,
|
| 282 |
+
"learning_rate": 9.068462401795736e-05,
|
| 283 |
+
"loss": 5.9605,
|
| 284 |
+
"step": 92000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.63,
|
| 288 |
+
"learning_rate": 9.04601571268238e-05,
|
| 289 |
+
"loss": 5.774,
|
| 290 |
+
"step": 94000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.65,
|
| 294 |
+
"learning_rate": 9.023569023569024e-05,
|
| 295 |
+
"loss": 5.8005,
|
| 296 |
+
"step": 96000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.66,
|
| 300 |
+
"learning_rate": 9.001122334455668e-05,
|
| 301 |
+
"loss": 5.8735,
|
| 302 |
+
"step": 98000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.68,
|
| 306 |
+
"learning_rate": 8.978675645342313e-05,
|
| 307 |
+
"loss": 5.8452,
|
| 308 |
+
"step": 100000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.69,
|
| 312 |
+
"learning_rate": 8.956228956228957e-05,
|
| 313 |
+
"loss": 5.7882,
|
| 314 |
+
"step": 102000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.7,
|
| 318 |
+
"learning_rate": 8.933782267115601e-05,
|
| 319 |
+
"loss": 6.0639,
|
| 320 |
+
"step": 104000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.72,
|
| 324 |
+
"learning_rate": 8.911335578002245e-05,
|
| 325 |
+
"loss": 6.0042,
|
| 326 |
+
"step": 106000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.73,
|
| 330 |
+
"learning_rate": 8.888888888888889e-05,
|
| 331 |
+
"loss": 5.9511,
|
| 332 |
+
"step": 108000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.74,
|
| 336 |
+
"learning_rate": 8.866442199775533e-05,
|
| 337 |
+
"loss": 5.8909,
|
| 338 |
+
"step": 110000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.76,
|
| 342 |
+
"learning_rate": 8.843995510662177e-05,
|
| 343 |
+
"loss": 5.8925,
|
| 344 |
+
"step": 112000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.77,
|
| 348 |
+
"learning_rate": 8.821548821548821e-05,
|
| 349 |
+
"loss": 5.9042,
|
| 350 |
+
"step": 114000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.78,
|
| 354 |
+
"learning_rate": 8.799102132435467e-05,
|
| 355 |
+
"loss": 5.9797,
|
| 356 |
+
"step": 116000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.8,
|
| 360 |
+
"learning_rate": 8.776655443322111e-05,
|
| 361 |
+
"loss": 6.1174,
|
| 362 |
+
"step": 118000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.81,
|
| 366 |
+
"learning_rate": 8.754208754208755e-05,
|
| 367 |
+
"loss": 6.2284,
|
| 368 |
+
"step": 120000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.82,
|
| 372 |
+
"learning_rate": 8.731762065095399e-05,
|
| 373 |
+
"loss": 6.1204,
|
| 374 |
+
"step": 122000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.84,
|
| 378 |
+
"learning_rate": 8.709315375982043e-05,
|
| 379 |
+
"loss": 5.6589,
|
| 380 |
+
"step": 124000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.85,
|
| 384 |
+
"learning_rate": 8.686868686868688e-05,
|
| 385 |
+
"loss": 5.7295,
|
| 386 |
+
"step": 126000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.86,
|
| 390 |
+
"learning_rate": 8.664421997755332e-05,
|
| 391 |
+
"loss": 5.7535,
|
| 392 |
+
"step": 128000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.88,
|
| 396 |
+
"learning_rate": 8.641975308641975e-05,
|
| 397 |
+
"loss": 5.6045,
|
| 398 |
+
"step": 130000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.89,
|
| 402 |
+
"learning_rate": 8.619528619528619e-05,
|
| 403 |
+
"loss": 5.6091,
|
| 404 |
+
"step": 132000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.91,
|
| 408 |
+
"learning_rate": 8.597081930415264e-05,
|
| 409 |
+
"loss": 5.5933,
|
| 410 |
+
"step": 134000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.92,
|
| 414 |
+
"learning_rate": 8.574635241301908e-05,
|
| 415 |
+
"loss": 5.6344,
|
| 416 |
+
"step": 136000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.93,
|
| 420 |
+
"learning_rate": 8.552188552188552e-05,
|
| 421 |
+
"loss": 5.6063,
|
| 422 |
+
"step": 138000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.95,
|
| 426 |
+
"learning_rate": 8.529741863075196e-05,
|
| 427 |
+
"loss": 5.4933,
|
| 428 |
+
"step": 140000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.96,
|
| 432 |
+
"learning_rate": 8.507295173961842e-05,
|
| 433 |
+
"loss": 5.5386,
|
| 434 |
+
"step": 142000
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.97,
|
| 438 |
+
"learning_rate": 8.484848484848486e-05,
|
| 439 |
+
"loss": 5.5085,
|
| 440 |
+
"step": 144000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.99,
|
| 444 |
+
"learning_rate": 8.46240179573513e-05,
|
| 445 |
+
"loss": 5.4098,
|
| 446 |
+
"step": 146000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.0,
|
| 450 |
+
"learning_rate": 8.439955106621774e-05,
|
| 451 |
+
"loss": 5.0913,
|
| 452 |
+
"step": 148000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.01,
|
| 456 |
+
"learning_rate": 8.417508417508418e-05,
|
| 457 |
+
"loss": 5.6115,
|
| 458 |
+
"step": 150000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.03,
|
| 462 |
+
"learning_rate": 8.395061728395062e-05,
|
| 463 |
+
"loss": 5.5311,
|
| 464 |
+
"step": 152000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.04,
|
| 468 |
+
"learning_rate": 8.372615039281706e-05,
|
| 469 |
+
"loss": 5.4583,
|
| 470 |
+
"step": 154000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.05,
|
| 474 |
+
"learning_rate": 8.35016835016835e-05,
|
| 475 |
+
"loss": 5.4188,
|
| 476 |
+
"step": 156000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.07,
|
| 480 |
+
"learning_rate": 8.327721661054994e-05,
|
| 481 |
+
"loss": 5.3814,
|
| 482 |
+
"step": 158000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.08,
|
| 486 |
+
"learning_rate": 8.30527497194164e-05,
|
| 487 |
+
"loss": 5.3292,
|
| 488 |
+
"step": 160000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.09,
|
| 492 |
+
"learning_rate": 8.282828282828283e-05,
|
| 493 |
+
"loss": 5.291,
|
| 494 |
+
"step": 162000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.11,
|
| 498 |
+
"learning_rate": 8.260381593714927e-05,
|
| 499 |
+
"loss": 5.2601,
|
| 500 |
+
"step": 164000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.12,
|
| 504 |
+
"learning_rate": 8.237934904601572e-05,
|
| 505 |
+
"loss": 5.2253,
|
| 506 |
+
"step": 166000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.13,
|
| 510 |
+
"learning_rate": 8.215488215488217e-05,
|
| 511 |
+
"loss": 5.2076,
|
| 512 |
+
"step": 168000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.15,
|
| 516 |
+
"learning_rate": 8.193041526374861e-05,
|
| 517 |
+
"loss": 5.1529,
|
| 518 |
+
"step": 170000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.16,
|
| 522 |
+
"learning_rate": 8.170594837261504e-05,
|
| 523 |
+
"loss": 5.1493,
|
| 524 |
+
"step": 172000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.18,
|
| 528 |
+
"learning_rate": 8.148148148148148e-05,
|
| 529 |
+
"loss": 5.1449,
|
| 530 |
+
"step": 174000
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.19,
|
| 534 |
+
"learning_rate": 8.125701459034793e-05,
|
| 535 |
+
"loss": 5.0768,
|
| 536 |
+
"step": 176000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.2,
|
| 540 |
+
"learning_rate": 8.103254769921437e-05,
|
| 541 |
+
"loss": 5.0558,
|
| 542 |
+
"step": 178000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.22,
|
| 546 |
+
"learning_rate": 8.080808080808081e-05,
|
| 547 |
+
"loss": 5.0232,
|
| 548 |
+
"step": 180000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.23,
|
| 552 |
+
"learning_rate": 8.058361391694725e-05,
|
| 553 |
+
"loss": 5.0037,
|
| 554 |
+
"step": 182000
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.24,
|
| 558 |
+
"learning_rate": 8.035914702581369e-05,
|
| 559 |
+
"loss": 4.9835,
|
| 560 |
+
"step": 184000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.26,
|
| 564 |
+
"learning_rate": 8.013468013468015e-05,
|
| 565 |
+
"loss": 4.9832,
|
| 566 |
+
"step": 186000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.27,
|
| 570 |
+
"learning_rate": 7.991021324354659e-05,
|
| 571 |
+
"loss": 4.9235,
|
| 572 |
+
"step": 188000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.28,
|
| 576 |
+
"learning_rate": 7.968574635241303e-05,
|
| 577 |
+
"loss": 4.9121,
|
| 578 |
+
"step": 190000
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.3,
|
| 582 |
+
"learning_rate": 7.946127946127947e-05,
|
| 583 |
+
"loss": 4.8973,
|
| 584 |
+
"step": 192000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.31,
|
| 588 |
+
"learning_rate": 7.92368125701459e-05,
|
| 589 |
+
"loss": 4.8992,
|
| 590 |
+
"step": 194000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.32,
|
| 594 |
+
"learning_rate": 7.901234567901235e-05,
|
| 595 |
+
"loss": 4.8595,
|
| 596 |
+
"step": 196000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.34,
|
| 600 |
+
"learning_rate": 7.878787878787879e-05,
|
| 601 |
+
"loss": 4.84,
|
| 602 |
+
"step": 198000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.35,
|
| 606 |
+
"learning_rate": 7.856341189674523e-05,
|
| 607 |
+
"loss": 4.8424,
|
| 608 |
+
"step": 200000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.36,
|
| 612 |
+
"learning_rate": 7.833894500561168e-05,
|
| 613 |
+
"loss": 4.8017,
|
| 614 |
+
"step": 202000
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.38,
|
| 618 |
+
"learning_rate": 7.811447811447812e-05,
|
| 619 |
+
"loss": 4.8083,
|
| 620 |
+
"step": 204000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.39,
|
| 624 |
+
"learning_rate": 7.789001122334456e-05,
|
| 625 |
+
"loss": 4.7732,
|
| 626 |
+
"step": 206000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.41,
|
| 630 |
+
"learning_rate": 7.7665544332211e-05,
|
| 631 |
+
"loss": 4.7528,
|
| 632 |
+
"step": 208000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.42,
|
| 636 |
+
"learning_rate": 7.744107744107744e-05,
|
| 637 |
+
"loss": 4.7499,
|
| 638 |
+
"step": 210000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.43,
|
| 642 |
+
"learning_rate": 7.72166105499439e-05,
|
| 643 |
+
"loss": 4.7275,
|
| 644 |
+
"step": 212000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.45,
|
| 648 |
+
"learning_rate": 7.699214365881034e-05,
|
| 649 |
+
"loss": 4.7118,
|
| 650 |
+
"step": 214000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.46,
|
| 654 |
+
"learning_rate": 7.676767676767676e-05,
|
| 655 |
+
"loss": 4.6839,
|
| 656 |
+
"step": 216000
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.47,
|
| 660 |
+
"learning_rate": 7.65432098765432e-05,
|
| 661 |
+
"loss": 4.6654,
|
| 662 |
+
"step": 218000
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.49,
|
| 666 |
+
"learning_rate": 7.631874298540966e-05,
|
| 667 |
+
"loss": 4.6509,
|
| 668 |
+
"step": 220000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.5,
|
| 672 |
+
"learning_rate": 7.60942760942761e-05,
|
| 673 |
+
"loss": 4.6436,
|
| 674 |
+
"step": 222000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.51,
|
| 678 |
+
"learning_rate": 7.586980920314254e-05,
|
| 679 |
+
"loss": 4.6321,
|
| 680 |
+
"step": 224000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.53,
|
| 684 |
+
"learning_rate": 7.564534231200898e-05,
|
| 685 |
+
"loss": 4.6257,
|
| 686 |
+
"step": 226000
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 1.54,
|
| 690 |
+
"learning_rate": 7.542087542087543e-05,
|
| 691 |
+
"loss": 4.6259,
|
| 692 |
+
"step": 228000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.55,
|
| 696 |
+
"learning_rate": 7.519640852974187e-05,
|
| 697 |
+
"loss": 4.5876,
|
| 698 |
+
"step": 230000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.57,
|
| 702 |
+
"learning_rate": 7.497194163860831e-05,
|
| 703 |
+
"loss": 4.5918,
|
| 704 |
+
"step": 232000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.58,
|
| 708 |
+
"learning_rate": 7.474747474747475e-05,
|
| 709 |
+
"loss": 4.7349,
|
| 710 |
+
"step": 234000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.59,
|
| 714 |
+
"learning_rate": 7.452300785634119e-05,
|
| 715 |
+
"loss": 4.7906,
|
| 716 |
+
"step": 236000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.61,
|
| 720 |
+
"learning_rate": 7.429854096520763e-05,
|
| 721 |
+
"loss": 4.5217,
|
| 722 |
+
"step": 238000
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 1.62,
|
| 726 |
+
"learning_rate": 7.407407407407407e-05,
|
| 727 |
+
"loss": 4.2357,
|
| 728 |
+
"step": 240000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.63,
|
| 732 |
+
"learning_rate": 7.384960718294051e-05,
|
| 733 |
+
"loss": 4.1207,
|
| 734 |
+
"step": 242000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.65,
|
| 738 |
+
"learning_rate": 7.362514029180695e-05,
|
| 739 |
+
"loss": 4.1877,
|
| 740 |
+
"step": 244000
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.66,
|
| 744 |
+
"learning_rate": 7.340067340067341e-05,
|
| 745 |
+
"loss": 4.4264,
|
| 746 |
+
"step": 246000
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.68,
|
| 750 |
+
"learning_rate": 7.317620650953985e-05,
|
| 751 |
+
"loss": 4.3763,
|
| 752 |
+
"step": 248000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.69,
|
| 756 |
+
"learning_rate": 7.295173961840629e-05,
|
| 757 |
+
"loss": 4.3764,
|
| 758 |
+
"step": 250000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.7,
|
| 762 |
+
"learning_rate": 7.272727272727273e-05,
|
| 763 |
+
"loss": 4.6481,
|
| 764 |
+
"step": 252000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.72,
|
| 768 |
+
"learning_rate": 7.250280583613918e-05,
|
| 769 |
+
"loss": 4.6296,
|
| 770 |
+
"step": 254000
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.73,
|
| 774 |
+
"learning_rate": 7.227833894500562e-05,
|
| 775 |
+
"loss": 4.6029,
|
| 776 |
+
"step": 256000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.74,
|
| 780 |
+
"learning_rate": 7.205387205387206e-05,
|
| 781 |
+
"loss": 4.5606,
|
| 782 |
+
"step": 258000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.76,
|
| 786 |
+
"learning_rate": 7.182940516273849e-05,
|
| 787 |
+
"loss": 4.6015,
|
| 788 |
+
"step": 260000
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.77,
|
| 792 |
+
"learning_rate": 7.160493827160494e-05,
|
| 793 |
+
"loss": 4.6519,
|
| 794 |
+
"step": 262000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.78,
|
| 798 |
+
"learning_rate": 7.138047138047138e-05,
|
| 799 |
+
"loss": 4.7609,
|
| 800 |
+
"step": 264000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.8,
|
| 804 |
+
"learning_rate": 7.115600448933782e-05,
|
| 805 |
+
"loss": 4.9615,
|
| 806 |
+
"step": 266000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.81,
|
| 810 |
+
"learning_rate": 7.093153759820426e-05,
|
| 811 |
+
"loss": 5.1392,
|
| 812 |
+
"step": 268000
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.82,
|
| 816 |
+
"learning_rate": 7.07070707070707e-05,
|
| 817 |
+
"loss": 5.0403,
|
| 818 |
+
"step": 270000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.84,
|
| 822 |
+
"learning_rate": 7.048260381593716e-05,
|
| 823 |
+
"loss": 4.3987,
|
| 824 |
+
"step": 272000
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.85,
|
| 828 |
+
"learning_rate": 7.02581369248036e-05,
|
| 829 |
+
"loss": 4.6271,
|
| 830 |
+
"step": 274000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.86,
|
| 834 |
+
"learning_rate": 7.003367003367004e-05,
|
| 835 |
+
"loss": 4.6795,
|
| 836 |
+
"step": 276000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.88,
|
| 840 |
+
"learning_rate": 6.980920314253648e-05,
|
| 841 |
+
"loss": 4.5634,
|
| 842 |
+
"step": 278000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.89,
|
| 846 |
+
"learning_rate": 6.958473625140293e-05,
|
| 847 |
+
"loss": 4.5767,
|
| 848 |
+
"step": 280000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.9,
|
| 852 |
+
"learning_rate": 6.936026936026936e-05,
|
| 853 |
+
"loss": 4.5943,
|
| 854 |
+
"step": 282000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.92,
|
| 858 |
+
"learning_rate": 6.91358024691358e-05,
|
| 859 |
+
"loss": 4.6575,
|
| 860 |
+
"step": 284000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.93,
|
| 864 |
+
"learning_rate": 6.891133557800224e-05,
|
| 865 |
+
"loss": 4.6512,
|
| 866 |
+
"step": 286000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.95,
|
| 870 |
+
"learning_rate": 6.86868686868687e-05,
|
| 871 |
+
"loss": 4.5433,
|
| 872 |
+
"step": 288000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.96,
|
| 876 |
+
"learning_rate": 6.846240179573513e-05,
|
| 877 |
+
"loss": 4.6108,
|
| 878 |
+
"step": 290000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.97,
|
| 882 |
+
"learning_rate": 6.823793490460157e-05,
|
| 883 |
+
"loss": 4.6175,
|
| 884 |
+
"step": 292000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.99,
|
| 888 |
+
"learning_rate": 6.801346801346801e-05,
|
| 889 |
+
"loss": 4.5076,
|
| 890 |
+
"step": 294000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.0,
|
| 894 |
+
"learning_rate": 6.778900112233445e-05,
|
| 895 |
+
"loss": 4.131,
|
| 896 |
+
"step": 296000
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.01,
|
| 900 |
+
"learning_rate": 6.756453423120091e-05,
|
| 901 |
+
"loss": 4.4705,
|
| 902 |
+
"step": 298000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.03,
|
| 906 |
+
"learning_rate": 6.734006734006735e-05,
|
| 907 |
+
"loss": 4.4578,
|
| 908 |
+
"step": 300000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.04,
|
| 912 |
+
"learning_rate": 6.711560044893379e-05,
|
| 913 |
+
"loss": 4.4142,
|
| 914 |
+
"step": 302000
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.05,
|
| 918 |
+
"learning_rate": 6.689113355780023e-05,
|
| 919 |
+
"loss": 4.3968,
|
| 920 |
+
"step": 304000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.07,
|
| 924 |
+
"learning_rate": 6.666666666666667e-05,
|
| 925 |
+
"loss": 4.3893,
|
| 926 |
+
"step": 306000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.08,
|
| 930 |
+
"learning_rate": 6.644219977553311e-05,
|
| 931 |
+
"loss": 4.3628,
|
| 932 |
+
"step": 308000
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.09,
|
| 936 |
+
"learning_rate": 6.621773288439955e-05,
|
| 937 |
+
"loss": 4.3452,
|
| 938 |
+
"step": 310000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.11,
|
| 942 |
+
"learning_rate": 6.599326599326599e-05,
|
| 943 |
+
"loss": 4.3391,
|
| 944 |
+
"step": 312000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.12,
|
| 948 |
+
"learning_rate": 6.576879910213244e-05,
|
| 949 |
+
"loss": 4.3232,
|
| 950 |
+
"step": 314000
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.13,
|
| 954 |
+
"learning_rate": 6.554433221099888e-05,
|
| 955 |
+
"loss": 4.3286,
|
| 956 |
+
"step": 316000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.15,
|
| 960 |
+
"learning_rate": 6.531986531986533e-05,
|
| 961 |
+
"loss": 4.2963,
|
| 962 |
+
"step": 318000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.16,
|
| 966 |
+
"learning_rate": 6.509539842873177e-05,
|
| 967 |
+
"loss": 4.3098,
|
| 968 |
+
"step": 320000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.18,
|
| 972 |
+
"learning_rate": 6.48709315375982e-05,
|
| 973 |
+
"loss": 4.3239,
|
| 974 |
+
"step": 322000
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.19,
|
| 978 |
+
"learning_rate": 6.464646464646466e-05,
|
| 979 |
+
"loss": 4.2606,
|
| 980 |
+
"step": 324000
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.2,
|
| 984 |
+
"learning_rate": 6.442199775533109e-05,
|
| 985 |
+
"loss": 4.2617,
|
| 986 |
+
"step": 326000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.22,
|
| 990 |
+
"learning_rate": 6.419753086419753e-05,
|
| 991 |
+
"loss": 4.2429,
|
| 992 |
+
"step": 328000
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.23,
|
| 996 |
+
"learning_rate": 6.397306397306398e-05,
|
| 997 |
+
"loss": 4.2412,
|
| 998 |
+
"step": 330000
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.24,
|
| 1002 |
+
"learning_rate": 6.374859708193042e-05,
|
| 1003 |
+
"loss": 4.2415,
|
| 1004 |
+
"step": 332000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.26,
|
| 1008 |
+
"learning_rate": 6.352413019079686e-05,
|
| 1009 |
+
"loss": 4.2517,
|
| 1010 |
+
"step": 334000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.27,
|
| 1014 |
+
"learning_rate": 6.32996632996633e-05,
|
| 1015 |
+
"loss": 4.2051,
|
| 1016 |
+
"step": 336000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.28,
|
| 1020 |
+
"learning_rate": 6.307519640852974e-05,
|
| 1021 |
+
"loss": 4.2003,
|
| 1022 |
+
"step": 338000
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.3,
|
| 1026 |
+
"learning_rate": 6.28507295173962e-05,
|
| 1027 |
+
"loss": 4.2039,
|
| 1028 |
+
"step": 340000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.31,
|
| 1032 |
+
"learning_rate": 6.262626262626264e-05,
|
| 1033 |
+
"loss": 4.2197,
|
| 1034 |
+
"step": 342000
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.32,
|
| 1038 |
+
"learning_rate": 6.240179573512908e-05,
|
| 1039 |
+
"loss": 4.1898,
|
| 1040 |
+
"step": 344000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.34,
|
| 1044 |
+
"learning_rate": 6.21773288439955e-05,
|
| 1045 |
+
"loss": 4.1859,
|
| 1046 |
+
"step": 346000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.35,
|
| 1050 |
+
"learning_rate": 6.195286195286196e-05,
|
| 1051 |
+
"loss": 4.2003,
|
| 1052 |
+
"step": 348000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.36,
|
| 1056 |
+
"learning_rate": 6.17283950617284e-05,
|
| 1057 |
+
"loss": 4.1667,
|
| 1058 |
+
"step": 350000
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.38,
|
| 1062 |
+
"learning_rate": 6.150392817059484e-05,
|
| 1063 |
+
"loss": 4.1896,
|
| 1064 |
+
"step": 352000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.39,
|
| 1068 |
+
"learning_rate": 6.127946127946128e-05,
|
| 1069 |
+
"loss": 4.1576,
|
| 1070 |
+
"step": 354000
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.4,
|
| 1074 |
+
"learning_rate": 6.105499438832773e-05,
|
| 1075 |
+
"loss": 4.1515,
|
| 1076 |
+
"step": 356000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.42,
|
| 1080 |
+
"learning_rate": 6.0830527497194164e-05,
|
| 1081 |
+
"loss": 4.1666,
|
| 1082 |
+
"step": 358000
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.43,
|
| 1086 |
+
"learning_rate": 6.060606060606061e-05,
|
| 1087 |
+
"loss": 4.1459,
|
| 1088 |
+
"step": 360000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.45,
|
| 1092 |
+
"learning_rate": 6.038159371492705e-05,
|
| 1093 |
+
"loss": 4.1438,
|
| 1094 |
+
"step": 362000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.46,
|
| 1098 |
+
"learning_rate": 6.01571268237935e-05,
|
| 1099 |
+
"loss": 4.1272,
|
| 1100 |
+
"step": 364000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.47,
|
| 1104 |
+
"learning_rate": 5.993265993265994e-05,
|
| 1105 |
+
"loss": 4.1105,
|
| 1106 |
+
"step": 366000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.49,
|
| 1110 |
+
"learning_rate": 5.970819304152637e-05,
|
| 1111 |
+
"loss": 4.1111,
|
| 1112 |
+
"step": 368000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.5,
|
| 1116 |
+
"learning_rate": 5.948372615039281e-05,
|
| 1117 |
+
"loss": 4.1062,
|
| 1118 |
+
"step": 370000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.51,
|
| 1122 |
+
"learning_rate": 5.925925925925926e-05,
|
| 1123 |
+
"loss": 4.1072,
|
| 1124 |
+
"step": 372000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.53,
|
| 1128 |
+
"learning_rate": 5.90347923681257e-05,
|
| 1129 |
+
"loss": 4.1112,
|
| 1130 |
+
"step": 374000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.54,
|
| 1134 |
+
"learning_rate": 5.881032547699215e-05,
|
| 1135 |
+
"loss": 4.1251,
|
| 1136 |
+
"step": 376000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.55,
|
| 1140 |
+
"learning_rate": 5.858585858585859e-05,
|
| 1141 |
+
"loss": 4.0813,
|
| 1142 |
+
"step": 378000
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 2.57,
|
| 1146 |
+
"learning_rate": 5.8361391694725034e-05,
|
| 1147 |
+
"loss": 4.0908,
|
| 1148 |
+
"step": 380000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.58,
|
| 1152 |
+
"learning_rate": 5.8136924803591475e-05,
|
| 1153 |
+
"loss": 4.2286,
|
| 1154 |
+
"step": 382000
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.59,
|
| 1158 |
+
"learning_rate": 5.7912457912457915e-05,
|
| 1159 |
+
"loss": 4.3118,
|
| 1160 |
+
"step": 384000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.61,
|
| 1164 |
+
"learning_rate": 5.768799102132436e-05,
|
| 1165 |
+
"loss": 4.0238,
|
| 1166 |
+
"step": 386000
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.62,
|
| 1170 |
+
"learning_rate": 5.74635241301908e-05,
|
| 1171 |
+
"loss": 3.7601,
|
| 1172 |
+
"step": 388000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.63,
|
| 1176 |
+
"learning_rate": 5.7239057239057236e-05,
|
| 1177 |
+
"loss": 3.6653,
|
| 1178 |
+
"step": 390000
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.65,
|
| 1182 |
+
"learning_rate": 5.701459034792368e-05,
|
| 1183 |
+
"loss": 3.7115,
|
| 1184 |
+
"step": 392000
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 2.66,
|
| 1188 |
+
"learning_rate": 5.679012345679012e-05,
|
| 1189 |
+
"loss": 4.0332,
|
| 1190 |
+
"step": 394000
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.67,
|
| 1194 |
+
"learning_rate": 5.6565656565656563e-05,
|
| 1195 |
+
"loss": 3.9273,
|
| 1196 |
+
"step": 396000
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.69,
|
| 1200 |
+
"learning_rate": 5.634118967452301e-05,
|
| 1201 |
+
"loss": 3.9359,
|
| 1202 |
+
"step": 398000
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.7,
|
| 1206 |
+
"learning_rate": 5.611672278338945e-05,
|
| 1207 |
+
"loss": 4.2338,
|
| 1208 |
+
"step": 400000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.72,
|
| 1212 |
+
"learning_rate": 5.58922558922559e-05,
|
| 1213 |
+
"loss": 4.2178,
|
| 1214 |
+
"step": 402000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.73,
|
| 1218 |
+
"learning_rate": 5.566778900112234e-05,
|
| 1219 |
+
"loss": 4.2029,
|
| 1220 |
+
"step": 404000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.74,
|
| 1224 |
+
"learning_rate": 5.5443322109988785e-05,
|
| 1225 |
+
"loss": 4.1545,
|
| 1226 |
+
"step": 406000
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.76,
|
| 1230 |
+
"learning_rate": 5.5218855218855225e-05,
|
| 1231 |
+
"loss": 4.2009,
|
| 1232 |
+
"step": 408000
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.77,
|
| 1236 |
+
"learning_rate": 5.4994388327721666e-05,
|
| 1237 |
+
"loss": 4.2519,
|
| 1238 |
+
"step": 410000
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.78,
|
| 1242 |
+
"learning_rate": 5.47699214365881e-05,
|
| 1243 |
+
"loss": 4.3561,
|
| 1244 |
+
"step": 412000
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.8,
|
| 1248 |
+
"learning_rate": 5.4545454545454546e-05,
|
| 1249 |
+
"loss": 4.5544,
|
| 1250 |
+
"step": 414000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.81,
|
| 1254 |
+
"learning_rate": 5.4320987654320986e-05,
|
| 1255 |
+
"loss": 4.7515,
|
| 1256 |
+
"step": 416000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.82,
|
| 1260 |
+
"learning_rate": 5.4096520763187434e-05,
|
| 1261 |
+
"loss": 4.6801,
|
| 1262 |
+
"step": 418000
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.84,
|
| 1266 |
+
"learning_rate": 5.3872053872053874e-05,
|
| 1267 |
+
"loss": 3.9784,
|
| 1268 |
+
"step": 420000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.85,
|
| 1272 |
+
"learning_rate": 5.3647586980920314e-05,
|
| 1273 |
+
"loss": 4.2571,
|
| 1274 |
+
"step": 422000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.86,
|
| 1278 |
+
"learning_rate": 5.342312008978676e-05,
|
| 1279 |
+
"loss": 4.3184,
|
| 1280 |
+
"step": 424000
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.88,
|
| 1284 |
+
"learning_rate": 5.31986531986532e-05,
|
| 1285 |
+
"loss": 4.2075,
|
| 1286 |
+
"step": 426000
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.89,
|
| 1290 |
+
"learning_rate": 5.297418630751965e-05,
|
| 1291 |
+
"loss": 4.2215,
|
| 1292 |
+
"step": 428000
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 2.9,
|
| 1296 |
+
"learning_rate": 5.274971941638609e-05,
|
| 1297 |
+
"loss": 4.2475,
|
| 1298 |
+
"step": 430000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.92,
|
| 1302 |
+
"learning_rate": 5.2525252525252536e-05,
|
| 1303 |
+
"loss": 4.3064,
|
| 1304 |
+
"step": 432000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.93,
|
| 1308 |
+
"learning_rate": 5.230078563411896e-05,
|
| 1309 |
+
"loss": 4.3034,
|
| 1310 |
+
"step": 434000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.95,
|
| 1314 |
+
"learning_rate": 5.207631874298541e-05,
|
| 1315 |
+
"loss": 4.2119,
|
| 1316 |
+
"step": 436000
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.96,
|
| 1320 |
+
"learning_rate": 5.185185185185185e-05,
|
| 1321 |
+
"loss": 4.2714,
|
| 1322 |
+
"step": 438000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.97,
|
| 1326 |
+
"learning_rate": 5.16273849607183e-05,
|
| 1327 |
+
"loss": 4.2878,
|
| 1328 |
+
"step": 440000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.99,
|
| 1332 |
+
"learning_rate": 5.140291806958474e-05,
|
| 1333 |
+
"loss": 4.1744,
|
| 1334 |
+
"step": 442000
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 3.0,
|
| 1338 |
+
"learning_rate": 5.1178451178451184e-05,
|
| 1339 |
+
"loss": 3.8192,
|
| 1340 |
+
"step": 444000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.01,
|
| 1344 |
+
"learning_rate": 5.0953984287317624e-05,
|
| 1345 |
+
"loss": 4.0513,
|
| 1346 |
+
"step": 446000
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.03,
|
| 1350 |
+
"learning_rate": 5.0729517396184065e-05,
|
| 1351 |
+
"loss": 4.0666,
|
| 1352 |
+
"step": 448000
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.04,
|
| 1356 |
+
"learning_rate": 5.050505050505051e-05,
|
| 1357 |
+
"loss": 4.0297,
|
| 1358 |
+
"step": 450000
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 3.05,
|
| 1362 |
+
"learning_rate": 5.028058361391695e-05,
|
| 1363 |
+
"loss": 4.019,
|
| 1364 |
+
"step": 452000
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 3.07,
|
| 1368 |
+
"learning_rate": 5.00561167227834e-05,
|
| 1369 |
+
"loss": 4.0183,
|
| 1370 |
+
"step": 454000
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 3.08,
|
| 1374 |
+
"learning_rate": 4.983164983164983e-05,
|
| 1375 |
+
"loss": 3.9988,
|
| 1376 |
+
"step": 456000
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 3.09,
|
| 1380 |
+
"learning_rate": 4.960718294051628e-05,
|
| 1381 |
+
"loss": 3.9882,
|
| 1382 |
+
"step": 458000
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 3.11,
|
| 1386 |
+
"learning_rate": 4.938271604938271e-05,
|
| 1387 |
+
"loss": 3.9879,
|
| 1388 |
+
"step": 460000
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.12,
|
| 1392 |
+
"learning_rate": 4.915824915824916e-05,
|
| 1393 |
+
"loss": 3.9752,
|
| 1394 |
+
"step": 462000
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 3.13,
|
| 1398 |
+
"learning_rate": 4.89337822671156e-05,
|
| 1399 |
+
"loss": 3.9901,
|
| 1400 |
+
"step": 464000
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 3.15,
|
| 1404 |
+
"learning_rate": 4.870931537598205e-05,
|
| 1405 |
+
"loss": 3.9586,
|
| 1406 |
+
"step": 466000
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 3.16,
|
| 1410 |
+
"learning_rate": 4.848484848484849e-05,
|
| 1411 |
+
"loss": 3.9711,
|
| 1412 |
+
"step": 468000
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 3.17,
|
| 1416 |
+
"learning_rate": 4.8260381593714935e-05,
|
| 1417 |
+
"loss": 4.0005,
|
| 1418 |
+
"step": 470000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 3.19,
|
| 1422 |
+
"learning_rate": 4.803591470258137e-05,
|
| 1423 |
+
"loss": 3.9334,
|
| 1424 |
+
"step": 472000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 3.2,
|
| 1428 |
+
"learning_rate": 4.7811447811447815e-05,
|
| 1429 |
+
"loss": 3.9461,
|
| 1430 |
+
"step": 474000
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.22,
|
| 1434 |
+
"learning_rate": 4.7586980920314256e-05,
|
| 1435 |
+
"loss": 3.9312,
|
| 1436 |
+
"step": 476000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 3.23,
|
| 1440 |
+
"learning_rate": 4.7362514029180696e-05,
|
| 1441 |
+
"loss": 3.928,
|
| 1442 |
+
"step": 478000
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 3.24,
|
| 1446 |
+
"learning_rate": 4.713804713804714e-05,
|
| 1447 |
+
"loss": 3.9341,
|
| 1448 |
+
"step": 480000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 3.26,
|
| 1452 |
+
"learning_rate": 4.691358024691358e-05,
|
| 1453 |
+
"loss": 3.949,
|
| 1454 |
+
"step": 482000
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 3.27,
|
| 1458 |
+
"learning_rate": 4.6689113355780024e-05,
|
| 1459 |
+
"loss": 3.9045,
|
| 1460 |
+
"step": 484000
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 3.28,
|
| 1464 |
+
"learning_rate": 4.6464646464646464e-05,
|
| 1465 |
+
"loss": 3.9063,
|
| 1466 |
+
"step": 486000
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 3.3,
|
| 1470 |
+
"learning_rate": 4.624017957351291e-05,
|
| 1471 |
+
"loss": 3.9079,
|
| 1472 |
+
"step": 488000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.31,
|
| 1476 |
+
"learning_rate": 4.601571268237935e-05,
|
| 1477 |
+
"loss": 3.9365,
|
| 1478 |
+
"step": 490000
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 3.32,
|
| 1482 |
+
"learning_rate": 4.57912457912458e-05,
|
| 1483 |
+
"loss": 3.908,
|
| 1484 |
+
"step": 492000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 3.34,
|
| 1488 |
+
"learning_rate": 4.556677890011223e-05,
|
| 1489 |
+
"loss": 3.9043,
|
| 1490 |
+
"step": 494000
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 3.35,
|
| 1494 |
+
"learning_rate": 4.534231200897868e-05,
|
| 1495 |
+
"loss": 3.9217,
|
| 1496 |
+
"step": 496000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 3.36,
|
| 1500 |
+
"learning_rate": 4.511784511784512e-05,
|
| 1501 |
+
"loss": 3.8926,
|
| 1502 |
+
"step": 498000
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 3.38,
|
| 1506 |
+
"learning_rate": 4.4893378226711566e-05,
|
| 1507 |
+
"loss": 3.9196,
|
| 1508 |
+
"step": 500000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 3.39,
|
| 1512 |
+
"learning_rate": 4.4668911335578006e-05,
|
| 1513 |
+
"loss": 3.8883,
|
| 1514 |
+
"step": 502000
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.4,
|
| 1518 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 1519 |
+
"loss": 3.8888,
|
| 1520 |
+
"step": 504000
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 3.42,
|
| 1524 |
+
"learning_rate": 4.421997755331089e-05,
|
| 1525 |
+
"loss": 3.9072,
|
| 1526 |
+
"step": 506000
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 3.43,
|
| 1530 |
+
"learning_rate": 4.3995510662177334e-05,
|
| 1531 |
+
"loss": 3.8872,
|
| 1532 |
+
"step": 508000
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 3.44,
|
| 1536 |
+
"learning_rate": 4.3771043771043774e-05,
|
| 1537 |
+
"loss": 3.8941,
|
| 1538 |
+
"step": 510000
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 3.46,
|
| 1542 |
+
"learning_rate": 4.3546576879910214e-05,
|
| 1543 |
+
"loss": 3.8768,
|
| 1544 |
+
"step": 512000
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 3.47,
|
| 1548 |
+
"learning_rate": 4.332210998877666e-05,
|
| 1549 |
+
"loss": 3.8667,
|
| 1550 |
+
"step": 514000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 3.49,
|
| 1554 |
+
"learning_rate": 4.3097643097643095e-05,
|
| 1555 |
+
"loss": 3.867,
|
| 1556 |
+
"step": 516000
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 3.5,
|
| 1560 |
+
"learning_rate": 4.287317620650954e-05,
|
| 1561 |
+
"loss": 3.8665,
|
| 1562 |
+
"step": 518000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 3.51,
|
| 1566 |
+
"learning_rate": 4.264870931537598e-05,
|
| 1567 |
+
"loss": 3.8677,
|
| 1568 |
+
"step": 520000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 3.53,
|
| 1572 |
+
"learning_rate": 4.242424242424243e-05,
|
| 1573 |
+
"loss": 3.8752,
|
| 1574 |
+
"step": 522000
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 3.54,
|
| 1578 |
+
"learning_rate": 4.219977553310887e-05,
|
| 1579 |
+
"loss": 3.888,
|
| 1580 |
+
"step": 524000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 3.55,
|
| 1584 |
+
"learning_rate": 4.197530864197531e-05,
|
| 1585 |
+
"loss": 3.849,
|
| 1586 |
+
"step": 526000
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 3.57,
|
| 1590 |
+
"learning_rate": 4.175084175084175e-05,
|
| 1591 |
+
"loss": 3.8591,
|
| 1592 |
+
"step": 528000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 3.58,
|
| 1596 |
+
"learning_rate": 4.15263748597082e-05,
|
| 1597 |
+
"loss": 3.9935,
|
| 1598 |
+
"step": 530000
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 3.59,
|
| 1602 |
+
"learning_rate": 4.130190796857464e-05,
|
| 1603 |
+
"loss": 4.1179,
|
| 1604 |
+
"step": 532000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 3.61,
|
| 1608 |
+
"learning_rate": 4.1077441077441085e-05,
|
| 1609 |
+
"loss": 3.8085,
|
| 1610 |
+
"step": 534000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 3.62,
|
| 1614 |
+
"learning_rate": 4.085297418630752e-05,
|
| 1615 |
+
"loss": 3.5572,
|
| 1616 |
+
"step": 536000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 3.63,
|
| 1620 |
+
"learning_rate": 4.0628507295173965e-05,
|
| 1621 |
+
"loss": 3.4772,
|
| 1622 |
+
"step": 538000
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 3.65,
|
| 1626 |
+
"learning_rate": 4.0404040404040405e-05,
|
| 1627 |
+
"loss": 3.4955,
|
| 1628 |
+
"step": 540000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 3.66,
|
| 1632 |
+
"learning_rate": 4.0179573512906846e-05,
|
| 1633 |
+
"loss": 3.8805,
|
| 1634 |
+
"step": 542000
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 3.67,
|
| 1638 |
+
"learning_rate": 3.995510662177329e-05,
|
| 1639 |
+
"loss": 3.7327,
|
| 1640 |
+
"step": 544000
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 3.69,
|
| 1644 |
+
"learning_rate": 3.973063973063973e-05,
|
| 1645 |
+
"loss": 3.7417,
|
| 1646 |
+
"step": 546000
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 3.7,
|
| 1650 |
+
"learning_rate": 3.950617283950617e-05,
|
| 1651 |
+
"loss": 4.0379,
|
| 1652 |
+
"step": 548000
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 3.72,
|
| 1656 |
+
"learning_rate": 3.9281705948372613e-05,
|
| 1657 |
+
"loss": 4.0257,
|
| 1658 |
+
"step": 550000
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 3.73,
|
| 1662 |
+
"learning_rate": 3.905723905723906e-05,
|
| 1663 |
+
"loss": 4.0116,
|
| 1664 |
+
"step": 552000
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 3.74,
|
| 1668 |
+
"learning_rate": 3.88327721661055e-05,
|
| 1669 |
+
"loss": 3.9664,
|
| 1670 |
+
"step": 554000
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 3.76,
|
| 1674 |
+
"learning_rate": 3.860830527497195e-05,
|
| 1675 |
+
"loss": 4.0149,
|
| 1676 |
+
"step": 556000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 3.77,
|
| 1680 |
+
"learning_rate": 3.838383838383838e-05,
|
| 1681 |
+
"loss": 4.0601,
|
| 1682 |
+
"step": 558000
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 3.78,
|
| 1686 |
+
"learning_rate": 3.815937149270483e-05,
|
| 1687 |
+
"loss": 4.1639,
|
| 1688 |
+
"step": 560000
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 3.8,
|
| 1692 |
+
"learning_rate": 3.793490460157127e-05,
|
| 1693 |
+
"loss": 4.3604,
|
| 1694 |
+
"step": 562000
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 3.81,
|
| 1698 |
+
"learning_rate": 3.7710437710437716e-05,
|
| 1699 |
+
"loss": 4.5598,
|
| 1700 |
+
"step": 564000
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 3.82,
|
| 1704 |
+
"learning_rate": 3.7485970819304156e-05,
|
| 1705 |
+
"loss": 4.5141,
|
| 1706 |
+
"step": 566000
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 3.84,
|
| 1710 |
+
"learning_rate": 3.7261503928170596e-05,
|
| 1711 |
+
"loss": 3.8061,
|
| 1712 |
+
"step": 568000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 3.85,
|
| 1716 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 1717 |
+
"loss": 4.0652,
|
| 1718 |
+
"step": 570000
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 3.86,
|
| 1722 |
+
"learning_rate": 3.681257014590348e-05,
|
| 1723 |
+
"loss": 4.1378,
|
| 1724 |
+
"step": 572000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 3.88,
|
| 1728 |
+
"learning_rate": 3.6588103254769924e-05,
|
| 1729 |
+
"loss": 4.0305,
|
| 1730 |
+
"step": 574000
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 3.89,
|
| 1734 |
+
"learning_rate": 3.6363636363636364e-05,
|
| 1735 |
+
"loss": 4.0567,
|
| 1736 |
+
"step": 576000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 3.9,
|
| 1740 |
+
"learning_rate": 3.613916947250281e-05,
|
| 1741 |
+
"loss": 4.0735,
|
| 1742 |
+
"step": 578000
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 3.92,
|
| 1746 |
+
"learning_rate": 3.5914702581369245e-05,
|
| 1747 |
+
"loss": 4.1457,
|
| 1748 |
+
"step": 580000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 3.93,
|
| 1752 |
+
"learning_rate": 3.569023569023569e-05,
|
| 1753 |
+
"loss": 4.1228,
|
| 1754 |
+
"step": 582000
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 3.94,
|
| 1758 |
+
"learning_rate": 3.546576879910213e-05,
|
| 1759 |
+
"loss": 4.061,
|
| 1760 |
+
"step": 584000
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 3.96,
|
| 1764 |
+
"learning_rate": 3.524130190796858e-05,
|
| 1765 |
+
"loss": 4.1077,
|
| 1766 |
+
"step": 586000
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 3.97,
|
| 1770 |
+
"learning_rate": 3.501683501683502e-05,
|
| 1771 |
+
"loss": 4.1329,
|
| 1772 |
+
"step": 588000
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 3.99,
|
| 1776 |
+
"learning_rate": 3.4792368125701466e-05,
|
| 1777 |
+
"loss": 4.0207,
|
| 1778 |
+
"step": 590000
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 4.0,
|
| 1782 |
+
"learning_rate": 3.45679012345679e-05,
|
| 1783 |
+
"loss": 3.6761,
|
| 1784 |
+
"step": 592000
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 4.01,
|
| 1788 |
+
"learning_rate": 3.434343434343435e-05,
|
| 1789 |
+
"loss": 3.8304,
|
| 1790 |
+
"step": 594000
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 4.03,
|
| 1794 |
+
"learning_rate": 3.411896745230079e-05,
|
| 1795 |
+
"loss": 3.8686,
|
| 1796 |
+
"step": 596000
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 4.04,
|
| 1800 |
+
"learning_rate": 3.389450056116723e-05,
|
| 1801 |
+
"loss": 3.8363,
|
| 1802 |
+
"step": 598000
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.05,
|
| 1806 |
+
"learning_rate": 3.3670033670033675e-05,
|
| 1807 |
+
"loss": 3.8302,
|
| 1808 |
+
"step": 600000
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.07,
|
| 1812 |
+
"learning_rate": 3.3445566778900115e-05,
|
| 1813 |
+
"loss": 3.8316,
|
| 1814 |
+
"step": 602000
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 4.08,
|
| 1818 |
+
"learning_rate": 3.3221099887766555e-05,
|
| 1819 |
+
"loss": 3.8158,
|
| 1820 |
+
"step": 604000
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 4.09,
|
| 1824 |
+
"learning_rate": 3.2996632996632995e-05,
|
| 1825 |
+
"loss": 3.8058,
|
| 1826 |
+
"step": 606000
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 4.11,
|
| 1830 |
+
"learning_rate": 3.277216610549944e-05,
|
| 1831 |
+
"loss": 3.8102,
|
| 1832 |
+
"step": 608000
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 4.12,
|
| 1836 |
+
"learning_rate": 3.254769921436588e-05,
|
| 1837 |
+
"loss": 3.8001,
|
| 1838 |
+
"step": 610000
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 4.13,
|
| 1842 |
+
"learning_rate": 3.232323232323233e-05,
|
| 1843 |
+
"loss": 3.8153,
|
| 1844 |
+
"step": 612000
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.15,
|
| 1848 |
+
"learning_rate": 3.209876543209876e-05,
|
| 1849 |
+
"loss": 3.7888,
|
| 1850 |
+
"step": 614000
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.16,
|
| 1854 |
+
"learning_rate": 3.187429854096521e-05,
|
| 1855 |
+
"loss": 3.8019,
|
| 1856 |
+
"step": 616000
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 4.17,
|
| 1860 |
+
"learning_rate": 3.164983164983165e-05,
|
| 1861 |
+
"loss": 3.8379,
|
| 1862 |
+
"step": 618000
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 4.19,
|
| 1866 |
+
"learning_rate": 3.14253647586981e-05,
|
| 1867 |
+
"loss": 3.7686,
|
| 1868 |
+
"step": 620000
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 4.2,
|
| 1872 |
+
"learning_rate": 3.120089786756454e-05,
|
| 1873 |
+
"loss": 3.7825,
|
| 1874 |
+
"step": 622000
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 4.22,
|
| 1878 |
+
"learning_rate": 3.097643097643098e-05,
|
| 1879 |
+
"loss": 3.7695,
|
| 1880 |
+
"step": 624000
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 4.23,
|
| 1884 |
+
"learning_rate": 3.075196408529742e-05,
|
| 1885 |
+
"loss": 3.7713,
|
| 1886 |
+
"step": 626000
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.24,
|
| 1890 |
+
"learning_rate": 3.0527497194163865e-05,
|
| 1891 |
+
"loss": 3.7789,
|
| 1892 |
+
"step": 628000
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.26,
|
| 1896 |
+
"learning_rate": 3.0303030303030306e-05,
|
| 1897 |
+
"loss": 3.7927,
|
| 1898 |
+
"step": 630000
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 4.27,
|
| 1902 |
+
"learning_rate": 3.007856341189675e-05,
|
| 1903 |
+
"loss": 3.7568,
|
| 1904 |
+
"step": 632000
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 4.28,
|
| 1908 |
+
"learning_rate": 2.9854096520763186e-05,
|
| 1909 |
+
"loss": 3.7555,
|
| 1910 |
+
"step": 634000
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 4.3,
|
| 1914 |
+
"learning_rate": 2.962962962962963e-05,
|
| 1915 |
+
"loss": 3.761,
|
| 1916 |
+
"step": 636000
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 4.31,
|
| 1920 |
+
"learning_rate": 2.9405162738496074e-05,
|
| 1921 |
+
"loss": 3.7894,
|
| 1922 |
+
"step": 638000
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 4.32,
|
| 1926 |
+
"learning_rate": 2.9180695847362517e-05,
|
| 1927 |
+
"loss": 3.7595,
|
| 1928 |
+
"step": 640000
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.34,
|
| 1932 |
+
"learning_rate": 2.8956228956228958e-05,
|
| 1933 |
+
"loss": 3.763,
|
| 1934 |
+
"step": 642000
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.35,
|
| 1938 |
+
"learning_rate": 2.87317620650954e-05,
|
| 1939 |
+
"loss": 3.7774,
|
| 1940 |
+
"step": 644000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 4.36,
|
| 1944 |
+
"learning_rate": 2.850729517396184e-05,
|
| 1945 |
+
"loss": 3.7551,
|
| 1946 |
+
"step": 646000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 4.38,
|
| 1950 |
+
"learning_rate": 2.8282828282828282e-05,
|
| 1951 |
+
"loss": 3.7829,
|
| 1952 |
+
"step": 648000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 4.39,
|
| 1956 |
+
"learning_rate": 2.8058361391694725e-05,
|
| 1957 |
+
"loss": 3.7527,
|
| 1958 |
+
"step": 650000
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 4.4,
|
| 1962 |
+
"learning_rate": 2.783389450056117e-05,
|
| 1963 |
+
"loss": 3.753,
|
| 1964 |
+
"step": 652000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 4.42,
|
| 1968 |
+
"learning_rate": 2.7609427609427613e-05,
|
| 1969 |
+
"loss": 3.7746,
|
| 1970 |
+
"step": 654000
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.43,
|
| 1974 |
+
"learning_rate": 2.738496071829405e-05,
|
| 1975 |
+
"loss": 3.7543,
|
| 1976 |
+
"step": 656000
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 4.44,
|
| 1980 |
+
"learning_rate": 2.7160493827160493e-05,
|
| 1981 |
+
"loss": 3.7604,
|
| 1982 |
+
"step": 658000
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 4.46,
|
| 1986 |
+
"learning_rate": 2.6936026936026937e-05,
|
| 1987 |
+
"loss": 3.7473,
|
| 1988 |
+
"step": 660000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 4.47,
|
| 1992 |
+
"learning_rate": 2.671156004489338e-05,
|
| 1993 |
+
"loss": 3.7377,
|
| 1994 |
+
"step": 662000
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 4.49,
|
| 1998 |
+
"learning_rate": 2.6487093153759824e-05,
|
| 1999 |
+
"loss": 3.7398,
|
| 2000 |
+
"step": 664000
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 4.5,
|
| 2004 |
+
"learning_rate": 2.6262626262626268e-05,
|
| 2005 |
+
"loss": 3.7432,
|
| 2006 |
+
"step": 666000
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 4.51,
|
| 2010 |
+
"learning_rate": 2.6038159371492705e-05,
|
| 2011 |
+
"loss": 3.7484,
|
| 2012 |
+
"step": 668000
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 4.53,
|
| 2016 |
+
"learning_rate": 2.581369248035915e-05,
|
| 2017 |
+
"loss": 3.7542,
|
| 2018 |
+
"step": 670000
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 4.54,
|
| 2022 |
+
"learning_rate": 2.5589225589225592e-05,
|
| 2023 |
+
"loss": 3.7721,
|
| 2024 |
+
"step": 672000
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 4.55,
|
| 2028 |
+
"learning_rate": 2.5364758698092032e-05,
|
| 2029 |
+
"loss": 3.7291,
|
| 2030 |
+
"step": 674000
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 4.57,
|
| 2034 |
+
"learning_rate": 2.5140291806958476e-05,
|
| 2035 |
+
"loss": 3.7455,
|
| 2036 |
+
"step": 676000
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 4.58,
|
| 2040 |
+
"learning_rate": 2.4915824915824916e-05,
|
| 2041 |
+
"loss": 3.8808,
|
| 2042 |
+
"step": 678000
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 4.59,
|
| 2046 |
+
"learning_rate": 2.4691358024691357e-05,
|
| 2047 |
+
"loss": 4.0555,
|
| 2048 |
+
"step": 680000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 4.61,
|
| 2052 |
+
"learning_rate": 2.44668911335578e-05,
|
| 2053 |
+
"loss": 3.7132,
|
| 2054 |
+
"step": 682000
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 4.62,
|
| 2058 |
+
"learning_rate": 2.4242424242424244e-05,
|
| 2059 |
+
"loss": 3.4602,
|
| 2060 |
+
"step": 684000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 4.63,
|
| 2064 |
+
"learning_rate": 2.4017957351290684e-05,
|
| 2065 |
+
"loss": 3.4057,
|
| 2066 |
+
"step": 686000
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 4.65,
|
| 2070 |
+
"learning_rate": 2.3793490460157128e-05,
|
| 2071 |
+
"loss": 3.397,
|
| 2072 |
+
"step": 688000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 4.66,
|
| 2076 |
+
"learning_rate": 2.356902356902357e-05,
|
| 2077 |
+
"loss": 3.8375,
|
| 2078 |
+
"step": 690000
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 4.67,
|
| 2082 |
+
"learning_rate": 2.3344556677890012e-05,
|
| 2083 |
+
"loss": 3.6615,
|
| 2084 |
+
"step": 692000
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 4.69,
|
| 2088 |
+
"learning_rate": 2.3120089786756455e-05,
|
| 2089 |
+
"loss": 3.6624,
|
| 2090 |
+
"step": 694000
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 4.7,
|
| 2094 |
+
"learning_rate": 2.28956228956229e-05,
|
| 2095 |
+
"loss": 3.9217,
|
| 2096 |
+
"step": 696000
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 4.71,
|
| 2100 |
+
"learning_rate": 2.267115600448934e-05,
|
| 2101 |
+
"loss": 3.9308,
|
| 2102 |
+
"step": 698000
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 4.73,
|
| 2106 |
+
"learning_rate": 2.2446689113355783e-05,
|
| 2107 |
+
"loss": 3.9061,
|
| 2108 |
+
"step": 700000
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 4.74,
|
| 2112 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 2113 |
+
"loss": 3.87,
|
| 2114 |
+
"step": 702000
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 4.76,
|
| 2118 |
+
"learning_rate": 2.1997755331088667e-05,
|
| 2119 |
+
"loss": 3.919,
|
| 2120 |
+
"step": 704000
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 4.77,
|
| 2124 |
+
"learning_rate": 2.1773288439955107e-05,
|
| 2125 |
+
"loss": 3.9621,
|
| 2126 |
+
"step": 706000
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 4.78,
|
| 2130 |
+
"learning_rate": 2.1548821548821547e-05,
|
| 2131 |
+
"loss": 4.0667,
|
| 2132 |
+
"step": 708000
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 4.8,
|
| 2136 |
+
"learning_rate": 2.132435465768799e-05,
|
| 2137 |
+
"loss": 4.2579,
|
| 2138 |
+
"step": 710000
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 4.81,
|
| 2142 |
+
"learning_rate": 2.1099887766554435e-05,
|
| 2143 |
+
"loss": 4.4613,
|
| 2144 |
+
"step": 712000
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 4.82,
|
| 2148 |
+
"learning_rate": 2.0875420875420875e-05,
|
| 2149 |
+
"loss": 4.4471,
|
| 2150 |
+
"step": 714000
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 4.84,
|
| 2154 |
+
"learning_rate": 2.065095398428732e-05,
|
| 2155 |
+
"loss": 3.7612,
|
| 2156 |
+
"step": 716000
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 4.85,
|
| 2160 |
+
"learning_rate": 2.042648709315376e-05,
|
| 2161 |
+
"loss": 3.9657,
|
| 2162 |
+
"step": 718000
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 4.86,
|
| 2166 |
+
"learning_rate": 2.0202020202020203e-05,
|
| 2167 |
+
"loss": 4.0575,
|
| 2168 |
+
"step": 720000
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 4.88,
|
| 2172 |
+
"learning_rate": 1.9977553310886646e-05,
|
| 2173 |
+
"loss": 3.9554,
|
| 2174 |
+
"step": 722000
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 4.89,
|
| 2178 |
+
"learning_rate": 1.9753086419753087e-05,
|
| 2179 |
+
"loss": 3.9786,
|
| 2180 |
+
"step": 724000
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 4.9,
|
| 2184 |
+
"learning_rate": 1.952861952861953e-05,
|
| 2185 |
+
"loss": 3.9809,
|
| 2186 |
+
"step": 726000
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 4.92,
|
| 2190 |
+
"learning_rate": 1.9304152637485974e-05,
|
| 2191 |
+
"loss": 4.0793,
|
| 2192 |
+
"step": 728000
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 4.93,
|
| 2196 |
+
"learning_rate": 1.9079685746352414e-05,
|
| 2197 |
+
"loss": 4.0373,
|
| 2198 |
+
"step": 730000
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 4.94,
|
| 2202 |
+
"learning_rate": 1.8855218855218858e-05,
|
| 2203 |
+
"loss": 4.0029,
|
| 2204 |
+
"step": 732000
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 4.96,
|
| 2208 |
+
"learning_rate": 1.8630751964085298e-05,
|
| 2209 |
+
"loss": 4.0193,
|
| 2210 |
+
"step": 734000
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 4.97,
|
| 2214 |
+
"learning_rate": 1.840628507295174e-05,
|
| 2215 |
+
"loss": 4.0719,
|
| 2216 |
+
"step": 736000
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 4.99,
|
| 2220 |
+
"learning_rate": 1.8181818181818182e-05,
|
| 2221 |
+
"loss": 3.9453,
|
| 2222 |
+
"step": 738000
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 5.0,
|
| 2226 |
+
"learning_rate": 1.7957351290684622e-05,
|
| 2227 |
+
"loss": 3.6204,
|
| 2228 |
+
"step": 740000
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 5.01,
|
| 2232 |
+
"learning_rate": 1.7732884399551066e-05,
|
| 2233 |
+
"loss": 3.6973,
|
| 2234 |
+
"step": 742000
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 5.03,
|
| 2238 |
+
"learning_rate": 1.750841750841751e-05,
|
| 2239 |
+
"loss": 3.7598,
|
| 2240 |
+
"step": 744000
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 5.04,
|
| 2244 |
+
"learning_rate": 1.728395061728395e-05,
|
| 2245 |
+
"loss": 3.7362,
|
| 2246 |
+
"step": 746000
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 5.05,
|
| 2250 |
+
"learning_rate": 1.7059483726150394e-05,
|
| 2251 |
+
"loss": 3.7313,
|
| 2252 |
+
"step": 748000
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 5.07,
|
| 2256 |
+
"learning_rate": 1.6835016835016837e-05,
|
| 2257 |
+
"loss": 3.7332,
|
| 2258 |
+
"step": 750000
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 5.08,
|
| 2262 |
+
"learning_rate": 1.6610549943883278e-05,
|
| 2263 |
+
"loss": 3.722,
|
| 2264 |
+
"step": 752000
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 5.09,
|
| 2268 |
+
"learning_rate": 1.638608305274972e-05,
|
| 2269 |
+
"loss": 3.713,
|
| 2270 |
+
"step": 754000
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 5.11,
|
| 2274 |
+
"learning_rate": 1.6161616161616165e-05,
|
| 2275 |
+
"loss": 3.72,
|
| 2276 |
+
"step": 756000
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 5.12,
|
| 2280 |
+
"learning_rate": 1.5937149270482605e-05,
|
| 2281 |
+
"loss": 3.7103,
|
| 2282 |
+
"step": 758000
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 5.13,
|
| 2286 |
+
"learning_rate": 1.571268237934905e-05,
|
| 2287 |
+
"loss": 3.7315,
|
| 2288 |
+
"step": 760000
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 5.15,
|
| 2292 |
+
"learning_rate": 1.548821548821549e-05,
|
| 2293 |
+
"loss": 3.7024,
|
| 2294 |
+
"step": 762000
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 5.16,
|
| 2298 |
+
"learning_rate": 1.5263748597081933e-05,
|
| 2299 |
+
"loss": 3.7156,
|
| 2300 |
+
"step": 764000
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 5.17,
|
| 2304 |
+
"learning_rate": 1.5039281705948375e-05,
|
| 2305 |
+
"loss": 3.7549,
|
| 2306 |
+
"step": 766000
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 5.19,
|
| 2310 |
+
"learning_rate": 1.4814814814814815e-05,
|
| 2311 |
+
"loss": 3.6887,
|
| 2312 |
+
"step": 768000
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 5.2,
|
| 2316 |
+
"learning_rate": 1.4590347923681259e-05,
|
| 2317 |
+
"loss": 3.7033,
|
| 2318 |
+
"step": 770000
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 5.21,
|
| 2322 |
+
"learning_rate": 1.43658810325477e-05,
|
| 2323 |
+
"loss": 3.6917,
|
| 2324 |
+
"step": 772000
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 5.23,
|
| 2328 |
+
"learning_rate": 1.4141414141414141e-05,
|
| 2329 |
+
"loss": 3.6936,
|
| 2330 |
+
"step": 774000
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 5.24,
|
| 2334 |
+
"learning_rate": 1.3916947250280585e-05,
|
| 2335 |
+
"loss": 3.7019,
|
| 2336 |
+
"step": 776000
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 5.26,
|
| 2340 |
+
"learning_rate": 1.3692480359147025e-05,
|
| 2341 |
+
"loss": 3.7172,
|
| 2342 |
+
"step": 778000
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 5.27,
|
| 2346 |
+
"learning_rate": 1.3468013468013468e-05,
|
| 2347 |
+
"loss": 3.6812,
|
| 2348 |
+
"step": 780000
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 5.28,
|
| 2352 |
+
"learning_rate": 1.3243546576879912e-05,
|
| 2353 |
+
"loss": 3.6841,
|
| 2354 |
+
"step": 782000
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 5.3,
|
| 2358 |
+
"learning_rate": 1.3019079685746352e-05,
|
| 2359 |
+
"loss": 3.6837,
|
| 2360 |
+
"step": 784000
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 5.31,
|
| 2364 |
+
"learning_rate": 1.2794612794612796e-05,
|
| 2365 |
+
"loss": 3.7225,
|
| 2366 |
+
"step": 786000
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 5.32,
|
| 2370 |
+
"learning_rate": 1.2570145903479238e-05,
|
| 2371 |
+
"loss": 3.6897,
|
| 2372 |
+
"step": 788000
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 5.34,
|
| 2376 |
+
"learning_rate": 1.2345679012345678e-05,
|
| 2377 |
+
"loss": 3.6941,
|
| 2378 |
+
"step": 790000
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 5.35,
|
| 2382 |
+
"learning_rate": 1.2121212121212122e-05,
|
| 2383 |
+
"loss": 3.7125,
|
| 2384 |
+
"step": 792000
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 5.36,
|
| 2388 |
+
"learning_rate": 1.1896745230078564e-05,
|
| 2389 |
+
"loss": 3.688,
|
| 2390 |
+
"step": 794000
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 5.38,
|
| 2394 |
+
"learning_rate": 1.1672278338945006e-05,
|
| 2395 |
+
"loss": 3.7184,
|
| 2396 |
+
"step": 796000
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 5.39,
|
| 2400 |
+
"learning_rate": 1.144781144781145e-05,
|
| 2401 |
+
"loss": 3.6901,
|
| 2402 |
+
"step": 798000
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 5.4,
|
| 2406 |
+
"learning_rate": 1.1223344556677892e-05,
|
| 2407 |
+
"loss": 3.6886,
|
| 2408 |
+
"step": 800000
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 5.42,
|
| 2412 |
+
"learning_rate": 1.0998877665544333e-05,
|
| 2413 |
+
"loss": 3.7129,
|
| 2414 |
+
"step": 802000
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 5.43,
|
| 2418 |
+
"learning_rate": 1.0774410774410774e-05,
|
| 2419 |
+
"loss": 3.6953,
|
| 2420 |
+
"step": 804000
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 5.44,
|
| 2424 |
+
"learning_rate": 1.0549943883277217e-05,
|
| 2425 |
+
"loss": 3.7016,
|
| 2426 |
+
"step": 806000
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 5.46,
|
| 2430 |
+
"learning_rate": 1.032547699214366e-05,
|
| 2431 |
+
"loss": 3.69,
|
| 2432 |
+
"step": 808000
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 5.47,
|
| 2436 |
+
"learning_rate": 1.0101010101010101e-05,
|
| 2437 |
+
"loss": 3.6834,
|
| 2438 |
+
"step": 810000
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 5.48,
|
| 2442 |
+
"learning_rate": 9.876543209876543e-06,
|
| 2443 |
+
"loss": 3.6809,
|
| 2444 |
+
"step": 812000
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 5.5,
|
| 2448 |
+
"learning_rate": 9.652076318742987e-06,
|
| 2449 |
+
"loss": 3.6861,
|
| 2450 |
+
"step": 814000
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 5.51,
|
| 2454 |
+
"learning_rate": 9.427609427609429e-06,
|
| 2455 |
+
"loss": 3.6977,
|
| 2456 |
+
"step": 816000
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 5.53,
|
| 2460 |
+
"learning_rate": 9.20314253647587e-06,
|
| 2461 |
+
"loss": 3.7002,
|
| 2462 |
+
"step": 818000
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 5.54,
|
| 2466 |
+
"learning_rate": 8.978675645342311e-06,
|
| 2467 |
+
"loss": 3.7243,
|
| 2468 |
+
"step": 820000
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 5.55,
|
| 2472 |
+
"learning_rate": 8.754208754208755e-06,
|
| 2473 |
+
"loss": 3.6768,
|
| 2474 |
+
"step": 822000
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 5.57,
|
| 2478 |
+
"learning_rate": 8.529741863075197e-06,
|
| 2479 |
+
"loss": 3.6947,
|
| 2480 |
+
"step": 824000
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 5.58,
|
| 2484 |
+
"learning_rate": 8.305274971941639e-06,
|
| 2485 |
+
"loss": 3.8538,
|
| 2486 |
+
"step": 826000
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 5.59,
|
| 2490 |
+
"learning_rate": 8.080808080808082e-06,
|
| 2491 |
+
"loss": 4.0943,
|
| 2492 |
+
"step": 828000
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 5.61,
|
| 2496 |
+
"learning_rate": 7.856341189674524e-06,
|
| 2497 |
+
"loss": 3.7048,
|
| 2498 |
+
"step": 830000
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 5.62,
|
| 2502 |
+
"learning_rate": 7.631874298540966e-06,
|
| 2503 |
+
"loss": 3.4471,
|
| 2504 |
+
"step": 832000
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 5.63,
|
| 2508 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 2509 |
+
"loss": 3.4131,
|
| 2510 |
+
"step": 834000
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 5.65,
|
| 2514 |
+
"learning_rate": 7.18294051627385e-06,
|
| 2515 |
+
"loss": 3.3965,
|
| 2516 |
+
"step": 836000
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 5.66,
|
| 2520 |
+
"learning_rate": 6.958473625140292e-06,
|
| 2521 |
+
"loss": 3.9014,
|
| 2522 |
+
"step": 838000
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 5.67,
|
| 2526 |
+
"learning_rate": 6.734006734006734e-06,
|
| 2527 |
+
"loss": 3.6888,
|
| 2528 |
+
"step": 840000
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 5.69,
|
| 2532 |
+
"learning_rate": 6.509539842873176e-06,
|
| 2533 |
+
"loss": 3.6862,
|
| 2534 |
+
"step": 842000
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 5.7,
|
| 2538 |
+
"learning_rate": 6.285072951739619e-06,
|
| 2539 |
+
"loss": 3.8646,
|
| 2540 |
+
"step": 844000
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"epoch": 5.71,
|
| 2544 |
+
"learning_rate": 6.060606060606061e-06,
|
| 2545 |
+
"loss": 3.8818,
|
| 2546 |
+
"step": 846000
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 5.73,
|
| 2550 |
+
"learning_rate": 5.836139169472503e-06,
|
| 2551 |
+
"loss": 3.8666,
|
| 2552 |
+
"step": 848000
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 5.74,
|
| 2556 |
+
"learning_rate": 5.611672278338946e-06,
|
| 2557 |
+
"loss": 3.8303,
|
| 2558 |
+
"step": 850000
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 5.76,
|
| 2562 |
+
"learning_rate": 5.387205387205387e-06,
|
| 2563 |
+
"loss": 3.8764,
|
| 2564 |
+
"step": 852000
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 5.77,
|
| 2568 |
+
"learning_rate": 5.16273849607183e-06,
|
| 2569 |
+
"loss": 3.9326,
|
| 2570 |
+
"step": 854000
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"epoch": 5.78,
|
| 2574 |
+
"learning_rate": 4.938271604938272e-06,
|
| 2575 |
+
"loss": 4.0327,
|
| 2576 |
+
"step": 856000
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 5.8,
|
| 2580 |
+
"learning_rate": 4.7138047138047145e-06,
|
| 2581 |
+
"loss": 4.2255,
|
| 2582 |
+
"step": 858000
|
| 2583 |
+
},
|
| 2584 |
+
{
|
| 2585 |
+
"epoch": 5.81,
|
| 2586 |
+
"learning_rate": 4.489337822671156e-06,
|
| 2587 |
+
"loss": 4.4367,
|
| 2588 |
+
"step": 860000
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 5.82,
|
| 2592 |
+
"learning_rate": 4.264870931537598e-06,
|
| 2593 |
+
"loss": 4.4619,
|
| 2594 |
+
"step": 862000
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 5.84,
|
| 2598 |
+
"learning_rate": 4.040404040404041e-06,
|
| 2599 |
+
"loss": 3.8376,
|
| 2600 |
+
"step": 864000
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 5.85,
|
| 2604 |
+
"learning_rate": 3.815937149270483e-06,
|
| 2605 |
+
"loss": 3.9389,
|
| 2606 |
+
"step": 866000
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 5.86,
|
| 2610 |
+
"learning_rate": 3.591470258136925e-06,
|
| 2611 |
+
"loss": 4.051,
|
| 2612 |
+
"step": 868000
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 5.88,
|
| 2616 |
+
"learning_rate": 3.367003367003367e-06,
|
| 2617 |
+
"loss": 3.9391,
|
| 2618 |
+
"step": 870000
|
| 2619 |
+
},
|
| 2620 |
+
{
|
| 2621 |
+
"epoch": 5.89,
|
| 2622 |
+
"learning_rate": 3.1425364758698095e-06,
|
| 2623 |
+
"loss": 3.969,
|
| 2624 |
+
"step": 872000
|
| 2625 |
+
},
|
| 2626 |
+
{
|
| 2627 |
+
"epoch": 5.9,
|
| 2628 |
+
"learning_rate": 2.9180695847362515e-06,
|
| 2629 |
+
"loss": 3.9726,
|
| 2630 |
+
"step": 874000
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 5.92,
|
| 2634 |
+
"learning_rate": 2.6936026936026934e-06,
|
| 2635 |
+
"loss": 4.081,
|
| 2636 |
+
"step": 876000
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 5.93,
|
| 2640 |
+
"learning_rate": 2.469135802469136e-06,
|
| 2641 |
+
"loss": 4.0299,
|
| 2642 |
+
"step": 878000
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 5.94,
|
| 2646 |
+
"learning_rate": 2.244668911335578e-06,
|
| 2647 |
+
"loss": 4.0083,
|
| 2648 |
+
"step": 880000
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 5.96,
|
| 2652 |
+
"learning_rate": 2.0202020202020206e-06,
|
| 2653 |
+
"loss": 4.0233,
|
| 2654 |
+
"step": 882000
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 5.97,
|
| 2658 |
+
"learning_rate": 1.7957351290684626e-06,
|
| 2659 |
+
"loss": 4.0916,
|
| 2660 |
+
"step": 884000
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 5.98,
|
| 2664 |
+
"learning_rate": 1.5712682379349048e-06,
|
| 2665 |
+
"loss": 3.9541,
|
| 2666 |
+
"step": 886000
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 6.0,
|
| 2670 |
+
"learning_rate": 1.3468013468013467e-06,
|
| 2671 |
+
"loss": 3.6427,
|
| 2672 |
+
"step": 888000
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 6.01,
|
| 2676 |
+
"learning_rate": 1.122334455667789e-06,
|
| 2677 |
+
"loss": 3.6296,
|
| 2678 |
+
"step": 890000
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 6.03,
|
| 2682 |
+
"learning_rate": 8.978675645342313e-07,
|
| 2683 |
+
"loss": 3.7085,
|
| 2684 |
+
"step": 892000
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 6.04,
|
| 2688 |
+
"learning_rate": 6.734006734006734e-07,
|
| 2689 |
+
"loss": 3.6918,
|
| 2690 |
+
"step": 894000
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 6.05,
|
| 2694 |
+
"learning_rate": 4.4893378226711564e-07,
|
| 2695 |
+
"loss": 3.6907,
|
| 2696 |
+
"step": 896000
|
| 2697 |
+
},
|
| 2698 |
+
{
|
| 2699 |
+
"epoch": 6.07,
|
| 2700 |
+
"learning_rate": 2.2446689113355782e-07,
|
| 2701 |
+
"loss": 3.6985,
|
| 2702 |
+
"step": 898000
|
| 2703 |
+
},
|
| 2704 |
+
{
|
| 2705 |
+
"epoch": 6.08,
|
| 2706 |
+
"learning_rate": 0.0,
|
| 2707 |
+
"loss": 3.6945,
|
| 2708 |
+
"step": 900000
|
| 2709 |
+
}
|
| 2710 |
+
],
|
| 2711 |
+
"max_steps": 900000,
|
| 2712 |
+
"num_train_epochs": 7,
|
| 2713 |
+
"total_flos": 1.894127697690476e+18,
|
| 2714 |
+
"trial_name": null,
|
| 2715 |
+
"trial_params": null
|
| 2716 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bfc27bb0a8680fdb510a73d211eb1e15775bca955d40be99592e20cf65f4fb6
|
| 3 |
+
size 2607
|