Hugging Face's logo

j-arkes
/
mt5-small-gos-nld-nonpretrained

Text2Text Generation
Transformers
PyTorch
t5
text-generation-inference
Inference Endpoints
Model card Files Community
1
j-arkes commited on
Commit
c7b4702
·
1 Parent(s): 59e0ede

first version

Browse files
Files changed (8) hide show
  1. config.json +31 -0
  2. generation_config.json +6 -0
  3. optimizer.pt +3 -0
  4. pytorch_model.bin +3 -0
  5. rng_state.pth +3 -0
  6. scheduler.pt +3 -0
  7. trainer_state.json +4216 -0
  8. training_args.bin +3 -0
config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/mt5-small",
3
+ "architectures": [
4
+ "T5ForConditionalGeneration"
5
+ ],
6
+ "d_ff": 1024,
7
+ "d_kv": 64,
8
+ "d_model": 512,
9
+ "decoder_start_token_id": 250099,
10
+ "dense_act_fn": "gelu_new",
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 1,
13
+ "feed_forward_proj": "gated-gelu",
14
+ "initializer_factor": 1.0,
15
+ "is_encoder_decoder": true,
16
+ "is_gated_act": true,
17
+ "layer_norm_epsilon": 1e-06,
18
+ "model_type": "t5",
19
+ "num_decoder_layers": 8,
20
+ "num_heads": 6,
21
+ "num_layers": 8,
22
+ "pad_token_id": 0,
23
+ "relative_attention_max_distance": 128,
24
+ "relative_attention_num_buckets": 32,
25
+ "tie_word_embeddings": false,
26
+ "tokenizer_class": "T5Tokenizer",
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.28.1",
29
+ "use_cache": true,
30
+ "vocab_size": 250112
31
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "decoder_start_token_id": 250099,
3
+ "eos_token_id": 1,
4
+ "pad_token_id": 0,
5
+ "transformers_version": "4.28.1"
6
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a094ba83ddf8e341a6943cc4881850aa7cc24f5c6a51c36b70f5cc23ad55726
3
+ size 2879301
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f85af26ff3cd3ab87d84660ecd52034885d37bf1caf3d6f1fad83936a5595b67
3
+ size 1200772613
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89723939282eaacbbab8bfcd721da961506e3197d070d6d0170873abaa990dda
3
+ size 14575
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:555734bc02f28323abc029b5c8032a4b42f3d054aa3a800c91f5e7ddbc24a9ba
3
+ size 563
trainer_state.json ADDED
@@ -0,0 +1,4216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 1.4148167371749878,
3
+ "best_model_checkpoint": "/scratch/s3545881/dumped/translation/mt5/3734190/checkpoint-4600",
4
+ "epoch": 18.231144872490503,
5
+ "global_step": 5600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 2.5e-06,
13
+ "loss": 16.635,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.07,
18
+ "learning_rate": 5e-06,
19
+ "loss": 16.8135,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.1,
24
+ "learning_rate": 7.5e-06,
25
+ "loss": 16.2077,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.13,
30
+ "learning_rate": 1e-05,
31
+ "loss": 15.7712,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.16,
36
+ "learning_rate": 1.25e-05,
37
+ "loss": 14.7744,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.2,
42
+ "learning_rate": 1.5e-05,
43
+ "loss": 14.0924,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.23,
48
+ "learning_rate": 1.7500000000000002e-05,
49
+ "loss": 13.2153,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.26,
54
+ "learning_rate": 2e-05,
55
+ "loss": 12.5009,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.29,
60
+ "learning_rate": 2.2499999999999998e-05,
61
+ "loss": 11.6579,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.33,
66
+ "learning_rate": 2.5e-05,
67
+ "loss": 10.6585,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.36,
72
+ "learning_rate": 2.75e-05,
73
+ "loss": 9.6627,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.39,
78
+ "learning_rate": 3e-05,
79
+ "loss": 9.1927,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.42,
84
+ "learning_rate": 3.2500000000000004e-05,
85
+ "loss": 8.4296,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.46,
90
+ "learning_rate": 3.5000000000000004e-05,
91
+ "loss": 7.9397,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.49,
96
+ "learning_rate": 3.75e-05,
97
+ "loss": 7.5002,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.52,
102
+ "learning_rate": 4e-05,
103
+ "loss": 7.0869,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.55,
108
+ "learning_rate": 4.25e-05,
109
+ "loss": 6.8123,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.59,
114
+ "learning_rate": 4.4999999999999996e-05,
115
+ "loss": 6.4624,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.62,
120
+ "learning_rate": 4.75e-05,
121
+ "loss": 6.2002,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.65,
126
+ "learning_rate": 5e-05,
127
+ "loss": 5.9634,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.65,
132
+ "eval_bp": 1.0,
133
+ "eval_counts": [
134
+ 8158,
135
+ 3455,
136
+ 1823,
137
+ 1029
138
+ ],
139
+ "eval_loss": 3.1029841899871826,
140
+ "eval_precisions": [
141
+ 2.863641788529988,
142
+ 1.2277067291121069,
143
+ 0.6539183086365283,
144
+ 0.3724954297815345
145
+ ],
146
+ "eval_ref_len": 41379,
147
+ "eval_runtime": 262.8531,
148
+ "eval_samples_per_second": 13.175,
149
+ "eval_score": 0.9619764567851472,
150
+ "eval_steps_per_second": 13.175,
151
+ "eval_sys_len": 284882,
152
+ "eval_totals": [
153
+ 284882,
154
+ 281419,
155
+ 278781,
156
+ 276245
157
+ ],
158
+ "step": 200
159
+ },
160
+ {
161
+ "epoch": 0.68,
162
+ "learning_rate": 5.25e-05,
163
+ "loss": 5.8077,
164
+ "step": 210
165
+ },
166
+ {
167
+ "epoch": 0.72,
168
+ "learning_rate": 5.5e-05,
169
+ "loss": 5.693,
170
+ "step": 220
171
+ },
172
+ {
173
+ "epoch": 0.75,
174
+ "learning_rate": 5.75e-05,
175
+ "loss": 5.5207,
176
+ "step": 230
177
+ },
178
+ {
179
+ "epoch": 0.78,
180
+ "learning_rate": 6e-05,
181
+ "loss": 5.4243,
182
+ "step": 240
183
+ },
184
+ {
185
+ "epoch": 0.81,
186
+ "learning_rate": 6.25e-05,
187
+ "loss": 5.3032,
188
+ "step": 250
189
+ },
190
+ {
191
+ "epoch": 0.85,
192
+ "learning_rate": 6.500000000000001e-05,
193
+ "loss": 5.2149,
194
+ "step": 260
195
+ },
196
+ {
197
+ "epoch": 0.88,
198
+ "learning_rate": 6.75e-05,
199
+ "loss": 5.1234,
200
+ "step": 270
201
+ },
202
+ {
203
+ "epoch": 0.91,
204
+ "learning_rate": 7.000000000000001e-05,
205
+ "loss": 5.0828,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.94,
210
+ "learning_rate": 7.25e-05,
211
+ "loss": 4.9444,
212
+ "step": 290
213
+ },
214
+ {
215
+ "epoch": 0.98,
216
+ "learning_rate": 7.5e-05,
217
+ "loss": 4.8644,
218
+ "step": 300
219
+ },
220
+ {
221
+ "epoch": 1.01,
222
+ "learning_rate": 7.75e-05,
223
+ "loss": 4.8276,
224
+ "step": 310
225
+ },
226
+ {
227
+ "epoch": 1.04,
228
+ "learning_rate": 8e-05,
229
+ "loss": 4.8093,
230
+ "step": 320
231
+ },
232
+ {
233
+ "epoch": 1.07,
234
+ "learning_rate": 8.25e-05,
235
+ "loss": 4.7373,
236
+ "step": 330
237
+ },
238
+ {
239
+ "epoch": 1.11,
240
+ "learning_rate": 8.5e-05,
241
+ "loss": 4.6358,
242
+ "step": 340
243
+ },
244
+ {
245
+ "epoch": 1.14,
246
+ "learning_rate": 8.75e-05,
247
+ "loss": 4.6567,
248
+ "step": 350
249
+ },
250
+ {
251
+ "epoch": 1.17,
252
+ "learning_rate": 8.999999999999999e-05,
253
+ "loss": 4.4909,
254
+ "step": 360
255
+ },
256
+ {
257
+ "epoch": 1.2,
258
+ "learning_rate": 9.25e-05,
259
+ "loss": 4.5973,
260
+ "step": 370
261
+ },
262
+ {
263
+ "epoch": 1.24,
264
+ "learning_rate": 9.5e-05,
265
+ "loss": 4.4554,
266
+ "step": 380
267
+ },
268
+ {
269
+ "epoch": 1.27,
270
+ "learning_rate": 9.750000000000001e-05,
271
+ "loss": 4.5076,
272
+ "step": 390
273
+ },
274
+ {
275
+ "epoch": 1.3,
276
+ "learning_rate": 0.0001,
277
+ "loss": 4.4274,
278
+ "step": 400
279
+ },
280
+ {
281
+ "epoch": 1.3,
282
+ "eval_bp": 1.0,
283
+ "eval_counts": [
284
+ 20993,
285
+ 12057,
286
+ 7745,
287
+ 5231
288
+ ],
289
+ "eval_loss": 2.268582582473755,
290
+ "eval_precisions": [
291
+ 27.83626816590644,
292
+ 16.75677178157964,
293
+ 11.30788997255154,
294
+ 8.037923139568832
295
+ ],
296
+ "eval_ref_len": 41379,
297
+ "eval_runtime": 232.9413,
298
+ "eval_samples_per_second": 14.866,
299
+ "eval_score": 14.349332444590503,
300
+ "eval_steps_per_second": 14.866,
301
+ "eval_sys_len": 75416,
302
+ "eval_totals": [
303
+ 75416,
304
+ 71953,
305
+ 68492,
306
+ 65079
307
+ ],
308
+ "step": 400
309
+ },
310
+ {
311
+ "epoch": 1.33,
312
+ "learning_rate": 0.0001025,
313
+ "loss": 4.4082,
314
+ "step": 410
315
+ },
316
+ {
317
+ "epoch": 1.37,
318
+ "learning_rate": 0.000105,
319
+ "loss": 4.3035,
320
+ "step": 420
321
+ },
322
+ {
323
+ "epoch": 1.4,
324
+ "learning_rate": 0.0001075,
325
+ "loss": 4.3256,
326
+ "step": 430
327
+ },
328
+ {
329
+ "epoch": 1.43,
330
+ "learning_rate": 0.00011,
331
+ "loss": 4.3323,
332
+ "step": 440
333
+ },
334
+ {
335
+ "epoch": 1.47,
336
+ "learning_rate": 0.00011250000000000001,
337
+ "loss": 4.2534,
338
+ "step": 450
339
+ },
340
+ {
341
+ "epoch": 1.5,
342
+ "learning_rate": 0.000115,
343
+ "loss": 4.291,
344
+ "step": 460
345
+ },
346
+ {
347
+ "epoch": 1.53,
348
+ "learning_rate": 0.0001175,
349
+ "loss": 4.2201,
350
+ "step": 470
351
+ },
352
+ {
353
+ "epoch": 1.56,
354
+ "learning_rate": 0.00012,
355
+ "loss": 4.3005,
356
+ "step": 480
357
+ },
358
+ {
359
+ "epoch": 1.6,
360
+ "learning_rate": 0.0001225,
361
+ "loss": 4.2619,
362
+ "step": 490
363
+ },
364
+ {
365
+ "epoch": 1.63,
366
+ "learning_rate": 0.000125,
367
+ "loss": 4.1464,
368
+ "step": 500
369
+ },
370
+ {
371
+ "epoch": 1.66,
372
+ "learning_rate": 0.0001275,
373
+ "loss": 4.104,
374
+ "step": 510
375
+ },
376
+ {
377
+ "epoch": 1.69,
378
+ "learning_rate": 0.00013000000000000002,
379
+ "loss": 4.1574,
380
+ "step": 520
381
+ },
382
+ {
383
+ "epoch": 1.73,
384
+ "learning_rate": 0.00013250000000000002,
385
+ "loss": 4.1,
386
+ "step": 530
387
+ },
388
+ {
389
+ "epoch": 1.76,
390
+ "learning_rate": 0.000135,
391
+ "loss": 4.078,
392
+ "step": 540
393
+ },
394
+ {
395
+ "epoch": 1.79,
396
+ "learning_rate": 0.0001375,
397
+ "loss": 4.1266,
398
+ "step": 550
399
+ },
400
+ {
401
+ "epoch": 1.82,
402
+ "learning_rate": 0.00014000000000000001,
403
+ "loss": 3.9953,
404
+ "step": 560
405
+ },
406
+ {
407
+ "epoch": 1.86,
408
+ "learning_rate": 0.0001425,
409
+ "loss": 4.0587,
410
+ "step": 570
411
+ },
412
+ {
413
+ "epoch": 1.89,
414
+ "learning_rate": 0.000145,
415
+ "loss": 4.0022,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 1.92,
420
+ "learning_rate": 0.0001475,
421
+ "loss": 3.963,
422
+ "step": 590
423
+ },
424
+ {
425
+ "epoch": 1.95,
426
+ "learning_rate": 0.00015,
427
+ "loss": 4.0209,
428
+ "step": 600
429
+ },
430
+ {
431
+ "epoch": 1.95,
432
+ "eval_bp": 1.0,
433
+ "eval_counts": [
434
+ 22518,
435
+ 13520,
436
+ 9025,
437
+ 6338
438
+ ],
439
+ "eval_loss": 2.0687694549560547,
440
+ "eval_precisions": [
441
+ 44.010554089709764,
442
+ 28.342627143515994,
443
+ 20.398246089865292,
444
+ 15.523280021553308
445
+ ],
446
+ "eval_ref_len": 41379,
447
+ "eval_runtime": 179.1653,
448
+ "eval_samples_per_second": 19.329,
449
+ "eval_score": 25.06936390823504,
450
+ "eval_steps_per_second": 19.329,
451
+ "eval_sys_len": 51165,
452
+ "eval_totals": [
453
+ 51165,
454
+ 47702,
455
+ 44244,
456
+ 40829
457
+ ],
458
+ "step": 600
459
+ },
460
+ {
461
+ "epoch": 1.99,
462
+ "learning_rate": 0.0001525,
463
+ "loss": 3.922,
464
+ "step": 610
465
+ },
466
+ {
467
+ "epoch": 2.02,
468
+ "learning_rate": 0.000155,
469
+ "loss": 3.872,
470
+ "step": 620
471
+ },
472
+ {
473
+ "epoch": 2.05,
474
+ "learning_rate": 0.0001575,
475
+ "loss": 3.9436,
476
+ "step": 630
477
+ },
478
+ {
479
+ "epoch": 2.08,
480
+ "learning_rate": 0.00016,
481
+ "loss": 3.8825,
482
+ "step": 640
483
+ },
484
+ {
485
+ "epoch": 2.12,
486
+ "learning_rate": 0.00016250000000000002,
487
+ "loss": 3.8445,
488
+ "step": 650
489
+ },
490
+ {
491
+ "epoch": 2.15,
492
+ "learning_rate": 0.000165,
493
+ "loss": 3.8438,
494
+ "step": 660
495
+ },
496
+ {
497
+ "epoch": 2.18,
498
+ "learning_rate": 0.0001675,
499
+ "loss": 3.9445,
500
+ "step": 670
501
+ },
502
+ {
503
+ "epoch": 2.21,
504
+ "learning_rate": 0.00017,
505
+ "loss": 3.8441,
506
+ "step": 680
507
+ },
508
+ {
509
+ "epoch": 2.25,
510
+ "learning_rate": 0.0001725,
511
+ "loss": 3.8509,
512
+ "step": 690
513
+ },
514
+ {
515
+ "epoch": 2.28,
516
+ "learning_rate": 0.000175,
517
+ "loss": 3.8228,
518
+ "step": 700
519
+ },
520
+ {
521
+ "epoch": 2.31,
522
+ "learning_rate": 0.0001775,
523
+ "loss": 3.7646,
524
+ "step": 710
525
+ },
526
+ {
527
+ "epoch": 2.34,
528
+ "learning_rate": 0.00017999999999999998,
529
+ "loss": 3.782,
530
+ "step": 720
531
+ },
532
+ {
533
+ "epoch": 2.38,
534
+ "learning_rate": 0.0001825,
535
+ "loss": 3.8468,
536
+ "step": 730
537
+ },
538
+ {
539
+ "epoch": 2.41,
540
+ "learning_rate": 0.000185,
541
+ "loss": 3.7436,
542
+ "step": 740
543
+ },
544
+ {
545
+ "epoch": 2.44,
546
+ "learning_rate": 0.0001875,
547
+ "loss": 3.7842,
548
+ "step": 750
549
+ },
550
+ {
551
+ "epoch": 2.47,
552
+ "learning_rate": 0.00019,
553
+ "loss": 3.7974,
554
+ "step": 760
555
+ },
556
+ {
557
+ "epoch": 2.51,
558
+ "learning_rate": 0.00019250000000000002,
559
+ "loss": 3.7282,
560
+ "step": 770
561
+ },
562
+ {
563
+ "epoch": 2.54,
564
+ "learning_rate": 0.00019500000000000002,
565
+ "loss": 3.7219,
566
+ "step": 780
567
+ },
568
+ {
569
+ "epoch": 2.57,
570
+ "learning_rate": 0.0001975,
571
+ "loss": 3.6277,
572
+ "step": 790
573
+ },
574
+ {
575
+ "epoch": 2.6,
576
+ "learning_rate": 0.0002,
577
+ "loss": 3.7015,
578
+ "step": 800
579
+ },
580
+ {
581
+ "epoch": 2.6,
582
+ "eval_bp": 1.0,
583
+ "eval_counts": [
584
+ 24390,
585
+ 15185,
586
+ 10398,
587
+ 7459
588
+ ],
589
+ "eval_loss": 1.9690614938735962,
590
+ "eval_precisions": [
591
+ 54.65668698458229,
592
+ 36.89171788829231,
593
+ 27.56116309274525,
594
+ 21.719760060567236
595
+ ],
596
+ "eval_ref_len": 41379,
597
+ "eval_runtime": 156.3814,
598
+ "eval_samples_per_second": 22.145,
599
+ "eval_score": 33.14600462748512,
600
+ "eval_steps_per_second": 22.145,
601
+ "eval_sys_len": 44624,
602
+ "eval_totals": [
603
+ 44624,
604
+ 41161,
605
+ 37727,
606
+ 34342
607
+ ],
608
+ "step": 800
609
+ },
610
+ {
611
+ "epoch": 2.64,
612
+ "learning_rate": 0.00020250000000000002,
613
+ "loss": 3.794,
614
+ "step": 810
615
+ },
616
+ {
617
+ "epoch": 2.67,
618
+ "learning_rate": 0.000205,
619
+ "loss": 3.6069,
620
+ "step": 820
621
+ },
622
+ {
623
+ "epoch": 2.7,
624
+ "learning_rate": 0.0002075,
625
+ "loss": 3.7057,
626
+ "step": 830
627
+ },
628
+ {
629
+ "epoch": 2.73,
630
+ "learning_rate": 0.00021,
631
+ "loss": 3.6715,
632
+ "step": 840
633
+ },
634
+ {
635
+ "epoch": 2.77,
636
+ "learning_rate": 0.0002125,
637
+ "loss": 3.6437,
638
+ "step": 850
639
+ },
640
+ {
641
+ "epoch": 2.8,
642
+ "learning_rate": 0.000215,
643
+ "loss": 3.6129,
644
+ "step": 860
645
+ },
646
+ {
647
+ "epoch": 2.83,
648
+ "learning_rate": 0.0002175,
649
+ "loss": 3.6232,
650
+ "step": 870
651
+ },
652
+ {
653
+ "epoch": 2.86,
654
+ "learning_rate": 0.00022,
655
+ "loss": 3.5967,
656
+ "step": 880
657
+ },
658
+ {
659
+ "epoch": 2.9,
660
+ "learning_rate": 0.00022250000000000001,
661
+ "loss": 3.5749,
662
+ "step": 890
663
+ },
664
+ {
665
+ "epoch": 2.93,
666
+ "learning_rate": 0.00022500000000000002,
667
+ "loss": 3.6316,
668
+ "step": 900
669
+ },
670
+ {
671
+ "epoch": 2.96,
672
+ "learning_rate": 0.0002275,
673
+ "loss": 3.6169,
674
+ "step": 910
675
+ },
676
+ {
677
+ "epoch": 3.0,
678
+ "learning_rate": 0.00023,
679
+ "loss": 3.5349,
680
+ "step": 920
681
+ },
682
+ {
683
+ "epoch": 3.03,
684
+ "learning_rate": 0.0002325,
685
+ "loss": 3.4972,
686
+ "step": 930
687
+ },
688
+ {
689
+ "epoch": 3.06,
690
+ "learning_rate": 0.000235,
691
+ "loss": 3.534,
692
+ "step": 940
693
+ },
694
+ {
695
+ "epoch": 3.09,
696
+ "learning_rate": 0.0002375,
697
+ "loss": 3.5623,
698
+ "step": 950
699
+ },
700
+ {
701
+ "epoch": 3.13,
702
+ "learning_rate": 0.00024,
703
+ "loss": 3.4621,
704
+ "step": 960
705
+ },
706
+ {
707
+ "epoch": 3.16,
708
+ "learning_rate": 0.00024249999999999999,
709
+ "loss": 3.5827,
710
+ "step": 970
711
+ },
712
+ {
713
+ "epoch": 3.19,
714
+ "learning_rate": 0.000245,
715
+ "loss": 3.4537,
716
+ "step": 980
717
+ },
718
+ {
719
+ "epoch": 3.22,
720
+ "learning_rate": 0.0002475,
721
+ "loss": 3.504,
722
+ "step": 990
723
+ },
724
+ {
725
+ "epoch": 3.26,
726
+ "learning_rate": 0.00025,
727
+ "loss": 3.5469,
728
+ "step": 1000
729
+ },
730
+ {
731
+ "epoch": 3.26,
732
+ "eval_bp": 1.0,
733
+ "eval_counts": [
734
+ 24938,
735
+ 15828,
736
+ 10952,
737
+ 7917
738
+ ],
739
+ "eval_loss": 1.8684660196304321,
740
+ "eval_precisions": [
741
+ 52.90423861852433,
742
+ 36.240412135088725,
743
+ 27.213318424649025,
744
+ 21.47740220281048
745
+ ],
746
+ "eval_ref_len": 41379,
747
+ "eval_runtime": 163.2477,
748
+ "eval_samples_per_second": 21.213,
749
+ "eval_score": 32.53581668713074,
750
+ "eval_steps_per_second": 21.213,
751
+ "eval_sys_len": 47138,
752
+ "eval_totals": [
753
+ 47138,
754
+ 43675,
755
+ 40245,
756
+ 36862
757
+ ],
758
+ "step": 1000
759
+ },
760
+ {
761
+ "epoch": 3.29,
762
+ "learning_rate": 0.0002525,
763
+ "loss": 3.4755,
764
+ "step": 1010
765
+ },
766
+ {
767
+ "epoch": 3.32,
768
+ "learning_rate": 0.000255,
769
+ "loss": 3.5018,
770
+ "step": 1020
771
+ },
772
+ {
773
+ "epoch": 3.35,
774
+ "learning_rate": 0.0002575,
775
+ "loss": 3.4461,
776
+ "step": 1030
777
+ },
778
+ {
779
+ "epoch": 3.39,
780
+ "learning_rate": 0.00026000000000000003,
781
+ "loss": 3.4587,
782
+ "step": 1040
783
+ },
784
+ {
785
+ "epoch": 3.42,
786
+ "learning_rate": 0.00026250000000000004,
787
+ "loss": 3.391,
788
+ "step": 1050
789
+ },
790
+ {
791
+ "epoch": 3.45,
792
+ "learning_rate": 0.00026500000000000004,
793
+ "loss": 3.4249,
794
+ "step": 1060
795
+ },
796
+ {
797
+ "epoch": 3.48,
798
+ "learning_rate": 0.0002675,
799
+ "loss": 3.5206,
800
+ "step": 1070
801
+ },
802
+ {
803
+ "epoch": 3.52,
804
+ "learning_rate": 0.00027,
805
+ "loss": 3.5261,
806
+ "step": 1080
807
+ },
808
+ {
809
+ "epoch": 3.55,
810
+ "learning_rate": 0.0002725,
811
+ "loss": 3.4603,
812
+ "step": 1090
813
+ },
814
+ {
815
+ "epoch": 3.58,
816
+ "learning_rate": 0.000275,
817
+ "loss": 3.4359,
818
+ "step": 1100
819
+ },
820
+ {
821
+ "epoch": 3.61,
822
+ "learning_rate": 0.0002775,
823
+ "loss": 3.4632,
824
+ "step": 1110
825
+ },
826
+ {
827
+ "epoch": 3.65,
828
+ "learning_rate": 0.00028000000000000003,
829
+ "loss": 3.4216,
830
+ "step": 1120
831
+ },
832
+ {
833
+ "epoch": 3.68,
834
+ "learning_rate": 0.0002825,
835
+ "loss": 3.3616,
836
+ "step": 1130
837
+ },
838
+ {
839
+ "epoch": 3.71,
840
+ "learning_rate": 0.000285,
841
+ "loss": 3.3384,
842
+ "step": 1140
843
+ },
844
+ {
845
+ "epoch": 3.74,
846
+ "learning_rate": 0.0002875,
847
+ "loss": 3.2927,
848
+ "step": 1150
849
+ },
850
+ {
851
+ "epoch": 3.78,
852
+ "learning_rate": 0.00029,
853
+ "loss": 3.4258,
854
+ "step": 1160
855
+ },
856
+ {
857
+ "epoch": 3.81,
858
+ "learning_rate": 0.0002925,
859
+ "loss": 3.4243,
860
+ "step": 1170
861
+ },
862
+ {
863
+ "epoch": 3.84,
864
+ "learning_rate": 0.000295,
865
+ "loss": 3.3272,
866
+ "step": 1180
867
+ },
868
+ {
869
+ "epoch": 3.87,
870
+ "learning_rate": 0.00029749999999999997,
871
+ "loss": 3.4568,
872
+ "step": 1190
873
+ },
874
+ {
875
+ "epoch": 3.91,
876
+ "learning_rate": 0.0003,
877
+ "loss": 3.4532,
878
+ "step": 1200
879
+ },
880
+ {
881
+ "epoch": 3.91,
882
+ "eval_bp": 1.0,
883
+ "eval_counts": [
884
+ 25039,
885
+ 16021,
886
+ 11141,
887
+ 8061
888
+ ],
889
+ "eval_loss": 1.7631980180740356,
890
+ "eval_precisions": [
891
+ 49.4558454640621,
892
+ 33.96726455497604,
893
+ 25.47329431132248,
894
+ 19.976704996034893
895
+ ],
896
+ "eval_ref_len": 41379,
897
+ "eval_runtime": 193.6778,
898
+ "eval_samples_per_second": 17.88,
899
+ "eval_score": 30.406875331612675,
900
+ "eval_steps_per_second": 17.88,
901
+ "eval_sys_len": 50629,
902
+ "eval_totals": [
903
+ 50629,
904
+ 47166,
905
+ 43736,
906
+ 40352
907
+ ],
908
+ "step": 1200
909
+ },
910
+ {
911
+ "epoch": 3.94,
912
+ "learning_rate": 0.0003025,
913
+ "loss": 3.3871,
914
+ "step": 1210
915
+ },
916
+ {
917
+ "epoch": 3.97,
918
+ "learning_rate": 0.000305,
919
+ "loss": 3.3748,
920
+ "step": 1220
921
+ },
922
+ {
923
+ "epoch": 4.0,
924
+ "learning_rate": 0.0003075,
925
+ "loss": 3.3834,
926
+ "step": 1230
927
+ },
928
+ {
929
+ "epoch": 4.04,
930
+ "learning_rate": 0.00031,
931
+ "loss": 3.2802,
932
+ "step": 1240
933
+ },
934
+ {
935
+ "epoch": 4.07,
936
+ "learning_rate": 0.0003125,
937
+ "loss": 3.3111,
938
+ "step": 1250
939
+ },
940
+ {
941
+ "epoch": 4.1,
942
+ "learning_rate": 0.000315,
943
+ "loss": 3.2518,
944
+ "step": 1260
945
+ },
946
+ {
947
+ "epoch": 4.13,
948
+ "learning_rate": 0.0003175,
949
+ "loss": 3.3339,
950
+ "step": 1270
951
+ },
952
+ {
953
+ "epoch": 4.17,
954
+ "learning_rate": 0.00032,
955
+ "loss": 3.2785,
956
+ "step": 1280
957
+ },
958
+ {
959
+ "epoch": 4.2,
960
+ "learning_rate": 0.00032250000000000003,
961
+ "loss": 3.2316,
962
+ "step": 1290
963
+ },
964
+ {
965
+ "epoch": 4.23,
966
+ "learning_rate": 0.00032500000000000004,
967
+ "loss": 3.3004,
968
+ "step": 1300
969
+ },
970
+ {
971
+ "epoch": 4.26,
972
+ "learning_rate": 0.00032750000000000005,
973
+ "loss": 3.236,
974
+ "step": 1310
975
+ },
976
+ {
977
+ "epoch": 4.3,
978
+ "learning_rate": 0.00033,
979
+ "loss": 3.3492,
980
+ "step": 1320
981
+ },
982
+ {
983
+ "epoch": 4.33,
984
+ "learning_rate": 0.0003325,
985
+ "loss": 3.2794,
986
+ "step": 1330
987
+ },
988
+ {
989
+ "epoch": 4.36,
990
+ "learning_rate": 0.000335,
991
+ "loss": 3.1176,
992
+ "step": 1340
993
+ },
994
+ {
995
+ "epoch": 4.4,
996
+ "learning_rate": 0.0003375,
997
+ "loss": 3.2454,
998
+ "step": 1350
999
+ },
1000
+ {
1001
+ "epoch": 4.43,
1002
+ "learning_rate": 0.00034,
1003
+ "loss": 3.2039,
1004
+ "step": 1360
1005
+ },
1006
+ {
1007
+ "epoch": 4.46,
1008
+ "learning_rate": 0.00034250000000000003,
1009
+ "loss": 3.2624,
1010
+ "step": 1370
1011
+ },
1012
+ {
1013
+ "epoch": 4.49,
1014
+ "learning_rate": 0.000345,
1015
+ "loss": 3.2248,
1016
+ "step": 1380
1017
+ },
1018
+ {
1019
+ "epoch": 4.53,
1020
+ "learning_rate": 0.0003475,
1021
+ "loss": 3.26,
1022
+ "step": 1390
1023
+ },
1024
+ {
1025
+ "epoch": 4.56,
1026
+ "learning_rate": 0.00035,
1027
+ "loss": 3.2288,
1028
+ "step": 1400
1029
+ },
1030
+ {
1031
+ "epoch": 4.56,
1032
+ "eval_bp": 1.0,
1033
+ "eval_counts": [
1034
+ 26211,
1035
+ 17123,
1036
+ 12060,
1037
+ 8793
1038
+ ],
1039
+ "eval_loss": 1.7274117469787598,
1040
+ "eval_precisions": [
1041
+ 58.21691135641783,
1042
+ 41.200673724735324,
1043
+ 31.649389843852514,
1044
+ 25.342979017754207
1045
+ ],
1046
+ "eval_ref_len": 41379,
1047
+ "eval_runtime": 140.7961,
1048
+ "eval_samples_per_second": 24.596,
1049
+ "eval_score": 37.242954038765475,
1050
+ "eval_steps_per_second": 24.596,
1051
+ "eval_sys_len": 45023,
1052
+ "eval_totals": [
1053
+ 45023,
1054
+ 41560,
1055
+ 38105,
1056
+ 34696
1057
+ ],
1058
+ "step": 1400
1059
+ },
1060
+ {
1061
+ "epoch": 4.59,
1062
+ "learning_rate": 0.0003525,
1063
+ "loss": 3.2446,
1064
+ "step": 1410
1065
+ },
1066
+ {
1067
+ "epoch": 4.62,
1068
+ "learning_rate": 0.000355,
1069
+ "loss": 3.1755,
1070
+ "step": 1420
1071
+ },
1072
+ {
1073
+ "epoch": 4.66,
1074
+ "learning_rate": 0.0003575,
1075
+ "loss": 3.274,
1076
+ "step": 1430
1077
+ },
1078
+ {
1079
+ "epoch": 4.69,
1080
+ "learning_rate": 0.00035999999999999997,
1081
+ "loss": 3.1518,
1082
+ "step": 1440
1083
+ },
1084
+ {
1085
+ "epoch": 4.72,
1086
+ "learning_rate": 0.0003625,
1087
+ "loss": 3.2472,
1088
+ "step": 1450
1089
+ },
1090
+ {
1091
+ "epoch": 4.75,
1092
+ "learning_rate": 0.000365,
1093
+ "loss": 3.231,
1094
+ "step": 1460
1095
+ },
1096
+ {
1097
+ "epoch": 4.79,
1098
+ "learning_rate": 0.0003675,
1099
+ "loss": 3.2377,
1100
+ "step": 1470
1101
+ },
1102
+ {
1103
+ "epoch": 4.82,
1104
+ "learning_rate": 0.00037,
1105
+ "loss": 3.2576,
1106
+ "step": 1480
1107
+ },
1108
+ {
1109
+ "epoch": 4.85,
1110
+ "learning_rate": 0.0003725,
1111
+ "loss": 3.2433,
1112
+ "step": 1490
1113
+ },
1114
+ {
1115
+ "epoch": 4.88,
1116
+ "learning_rate": 0.000375,
1117
+ "loss": 3.2144,
1118
+ "step": 1500
1119
+ },
1120
+ {
1121
+ "epoch": 4.92,
1122
+ "learning_rate": 0.0003775,
1123
+ "loss": 3.2166,
1124
+ "step": 1510
1125
+ },
1126
+ {
1127
+ "epoch": 4.95,
1128
+ "learning_rate": 0.00038,
1129
+ "loss": 3.1721,
1130
+ "step": 1520
1131
+ },
1132
+ {
1133
+ "epoch": 4.98,
1134
+ "learning_rate": 0.00038250000000000003,
1135
+ "loss": 3.1729,
1136
+ "step": 1530
1137
+ },
1138
+ {
1139
+ "epoch": 5.01,
1140
+ "learning_rate": 0.00038500000000000003,
1141
+ "loss": 3.1772,
1142
+ "step": 1540
1143
+ },
1144
+ {
1145
+ "epoch": 5.05,
1146
+ "learning_rate": 0.00038750000000000004,
1147
+ "loss": 3.1715,
1148
+ "step": 1550
1149
+ },
1150
+ {
1151
+ "epoch": 5.08,
1152
+ "learning_rate": 0.00039000000000000005,
1153
+ "loss": 3.0745,
1154
+ "step": 1560
1155
+ },
1156
+ {
1157
+ "epoch": 5.11,
1158
+ "learning_rate": 0.0003925,
1159
+ "loss": 3.1181,
1160
+ "step": 1570
1161
+ },
1162
+ {
1163
+ "epoch": 5.14,
1164
+ "learning_rate": 0.000395,
1165
+ "loss": 3.0892,
1166
+ "step": 1580
1167
+ },
1168
+ {
1169
+ "epoch": 5.18,
1170
+ "learning_rate": 0.0003975,
1171
+ "loss": 3.1591,
1172
+ "step": 1590
1173
+ },
1174
+ {
1175
+ "epoch": 5.21,
1176
+ "learning_rate": 0.0004,
1177
+ "loss": 3.1232,
1178
+ "step": 1600
1179
+ },
1180
+ {
1181
+ "epoch": 5.21,
1182
+ "eval_bp": 1.0,
1183
+ "eval_counts": [
1184
+ 25985,
1185
+ 17105,
1186
+ 12124,
1187
+ 8931
1188
+ ],
1189
+ "eval_loss": 1.6549996137619019,
1190
+ "eval_precisions": [
1191
+ 61.648873072360615,
1192
+ 44.2138185953938,
1193
+ 34.410921579201315,
1194
+ 28.063725490196077
1195
+ ],
1196
+ "eval_ref_len": 41379,
1197
+ "eval_runtime": 118.8752,
1198
+ "eval_samples_per_second": 29.131,
1199
+ "eval_score": 40.27923361043648,
1200
+ "eval_steps_per_second": 29.131,
1201
+ "eval_sys_len": 42150,
1202
+ "eval_totals": [
1203
+ 42150,
1204
+ 38687,
1205
+ 35233,
1206
+ 31824
1207
+ ],
1208
+ "step": 1600
1209
+ },
1210
+ {
1211
+ "epoch": 5.24,
1212
+ "learning_rate": 0.0004025,
1213
+ "loss": 3.049,
1214
+ "step": 1610
1215
+ },
1216
+ {
1217
+ "epoch": 5.27,
1218
+ "learning_rate": 0.00040500000000000003,
1219
+ "loss": 3.16,
1220
+ "step": 1620
1221
+ },
1222
+ {
1223
+ "epoch": 5.31,
1224
+ "learning_rate": 0.0004075,
1225
+ "loss": 3.1724,
1226
+ "step": 1630
1227
+ },
1228
+ {
1229
+ "epoch": 5.34,
1230
+ "learning_rate": 0.00041,
1231
+ "loss": 3.114,
1232
+ "step": 1640
1233
+ },
1234
+ {
1235
+ "epoch": 5.37,
1236
+ "learning_rate": 0.0004125,
1237
+ "loss": 3.1586,
1238
+ "step": 1650
1239
+ },
1240
+ {
1241
+ "epoch": 5.4,
1242
+ "learning_rate": 0.000415,
1243
+ "loss": 3.0389,
1244
+ "step": 1660
1245
+ },
1246
+ {
1247
+ "epoch": 5.44,
1248
+ "learning_rate": 0.0004175,
1249
+ "loss": 3.0599,
1250
+ "step": 1670
1251
+ },
1252
+ {
1253
+ "epoch": 5.47,
1254
+ "learning_rate": 0.00042,
1255
+ "loss": 3.1169,
1256
+ "step": 1680
1257
+ },
1258
+ {
1259
+ "epoch": 5.5,
1260
+ "learning_rate": 0.00042249999999999997,
1261
+ "loss": 3.0748,
1262
+ "step": 1690
1263
+ },
1264
+ {
1265
+ "epoch": 5.53,
1266
+ "learning_rate": 0.000425,
1267
+ "loss": 3.0661,
1268
+ "step": 1700
1269
+ },
1270
+ {
1271
+ "epoch": 5.57,
1272
+ "learning_rate": 0.0004275,
1273
+ "loss": 3.125,
1274
+ "step": 1710
1275
+ },
1276
+ {
1277
+ "epoch": 5.6,
1278
+ "learning_rate": 0.00043,
1279
+ "loss": 3.153,
1280
+ "step": 1720
1281
+ },
1282
+ {
1283
+ "epoch": 5.63,
1284
+ "learning_rate": 0.0004325,
1285
+ "loss": 3.0269,
1286
+ "step": 1730
1287
+ },
1288
+ {
1289
+ "epoch": 5.66,
1290
+ "learning_rate": 0.000435,
1291
+ "loss": 3.0262,
1292
+ "step": 1740
1293
+ },
1294
+ {
1295
+ "epoch": 5.7,
1296
+ "learning_rate": 0.0004375,
1297
+ "loss": 2.993,
1298
+ "step": 1750
1299
+ },
1300
+ {
1301
+ "epoch": 5.73,
1302
+ "learning_rate": 0.00044,
1303
+ "loss": 3.0263,
1304
+ "step": 1760
1305
+ },
1306
+ {
1307
+ "epoch": 5.76,
1308
+ "learning_rate": 0.0004425,
1309
+ "loss": 2.959,
1310
+ "step": 1770
1311
+ },
1312
+ {
1313
+ "epoch": 5.79,
1314
+ "learning_rate": 0.00044500000000000003,
1315
+ "loss": 3.0096,
1316
+ "step": 1780
1317
+ },
1318
+ {
1319
+ "epoch": 5.83,
1320
+ "learning_rate": 0.00044750000000000004,
1321
+ "loss": 3.0877,
1322
+ "step": 1790
1323
+ },
1324
+ {
1325
+ "epoch": 5.86,
1326
+ "learning_rate": 0.00045000000000000004,
1327
+ "loss": 3.0788,
1328
+ "step": 1800
1329
+ },
1330
+ {
1331
+ "epoch": 5.86,
1332
+ "eval_bp": 1.0,
1333
+ "eval_counts": [
1334
+ 25630,
1335
+ 16905,
1336
+ 12013,
1337
+ 8850
1338
+ ],
1339
+ "eval_loss": 1.6167092323303223,
1340
+ "eval_precisions": [
1341
+ 60.873076192285765,
1342
+ 43.748867782924876,
1343
+ 34.141419882907975,
1344
+ 27.850332001132895
1345
+ ],
1346
+ "eval_ref_len": 41379,
1347
+ "eval_runtime": 122.6416,
1348
+ "eval_samples_per_second": 28.237,
1349
+ "eval_score": 39.89109940073679,
1350
+ "eval_steps_per_second": 28.237,
1351
+ "eval_sys_len": 42104,
1352
+ "eval_totals": [
1353
+ 42104,
1354
+ 38641,
1355
+ 35186,
1356
+ 31777
1357
+ ],
1358
+ "step": 1800
1359
+ },
1360
+ {
1361
+ "epoch": 5.89,
1362
+ "learning_rate": 0.00045250000000000005,
1363
+ "loss": 3.1263,
1364
+ "step": 1810
1365
+ },
1366
+ {
1367
+ "epoch": 5.93,
1368
+ "learning_rate": 0.000455,
1369
+ "loss": 3.0045,
1370
+ "step": 1820
1371
+ },
1372
+ {
1373
+ "epoch": 5.96,
1374
+ "learning_rate": 0.0004575,
1375
+ "loss": 3.0452,
1376
+ "step": 1830
1377
+ },
1378
+ {
1379
+ "epoch": 5.99,
1380
+ "learning_rate": 0.00046,
1381
+ "loss": 3.0684,
1382
+ "step": 1840
1383
+ },
1384
+ {
1385
+ "epoch": 6.02,
1386
+ "learning_rate": 0.0004625,
1387
+ "loss": 3.0385,
1388
+ "step": 1850
1389
+ },
1390
+ {
1391
+ "epoch": 6.06,
1392
+ "learning_rate": 0.000465,
1393
+ "loss": 2.9502,
1394
+ "step": 1860
1395
+ },
1396
+ {
1397
+ "epoch": 6.09,
1398
+ "learning_rate": 0.00046750000000000003,
1399
+ "loss": 3.019,
1400
+ "step": 1870
1401
+ },
1402
+ {
1403
+ "epoch": 6.12,
1404
+ "learning_rate": 0.00047,
1405
+ "loss": 2.9622,
1406
+ "step": 1880
1407
+ },
1408
+ {
1409
+ "epoch": 6.15,
1410
+ "learning_rate": 0.0004725,
1411
+ "loss": 2.9952,
1412
+ "step": 1890
1413
+ },
1414
+ {
1415
+ "epoch": 6.19,
1416
+ "learning_rate": 0.000475,
1417
+ "loss": 2.961,
1418
+ "step": 1900
1419
+ },
1420
+ {
1421
+ "epoch": 6.22,
1422
+ "learning_rate": 0.0004775,
1423
+ "loss": 2.9056,
1424
+ "step": 1910
1425
+ },
1426
+ {
1427
+ "epoch": 6.25,
1428
+ "learning_rate": 0.00048,
1429
+ "loss": 3.03,
1430
+ "step": 1920
1431
+ },
1432
+ {
1433
+ "epoch": 6.28,
1434
+ "learning_rate": 0.0004825,
1435
+ "loss": 2.9475,
1436
+ "step": 1930
1437
+ },
1438
+ {
1439
+ "epoch": 6.32,
1440
+ "learning_rate": 0.00048499999999999997,
1441
+ "loss": 2.9383,
1442
+ "step": 1940
1443
+ },
1444
+ {
1445
+ "epoch": 6.35,
1446
+ "learning_rate": 0.0004875,
1447
+ "loss": 2.9497,
1448
+ "step": 1950
1449
+ },
1450
+ {
1451
+ "epoch": 6.38,
1452
+ "learning_rate": 0.00049,
1453
+ "loss": 2.9825,
1454
+ "step": 1960
1455
+ },
1456
+ {
1457
+ "epoch": 6.41,
1458
+ "learning_rate": 0.0004925,
1459
+ "loss": 2.8683,
1460
+ "step": 1970
1461
+ },
1462
+ {
1463
+ "epoch": 6.45,
1464
+ "learning_rate": 0.000495,
1465
+ "loss": 2.9585,
1466
+ "step": 1980
1467
+ },
1468
+ {
1469
+ "epoch": 6.48,
1470
+ "learning_rate": 0.0004975,
1471
+ "loss": 3.0039,
1472
+ "step": 1990
1473
+ },
1474
+ {
1475
+ "epoch": 6.51,
1476
+ "learning_rate": 0.0005,
1477
+ "loss": 2.8811,
1478
+ "step": 2000
1479
+ },
1480
+ {
1481
+ "epoch": 6.51,
1482
+ "eval_bp": 1.0,
1483
+ "eval_counts": [
1484
+ 27043,
1485
+ 18237,
1486
+ 13141,
1487
+ 9796
1488
+ ],
1489
+ "eval_loss": 1.6076596975326538,
1490
+ "eval_precisions": [
1491
+ 62.29957611500184,
1492
+ 45.6552760045062,
1493
+ 36.01359313765792,
1494
+ 29.613954472626137
1495
+ ],
1496
+ "eval_ref_len": 41379,
1497
+ "eval_runtime": 127.9727,
1498
+ "eval_samples_per_second": 27.06,
1499
+ "eval_score": 41.733492512975715,
1500
+ "eval_steps_per_second": 27.06,
1501
+ "eval_sys_len": 43408,
1502
+ "eval_totals": [
1503
+ 43408,
1504
+ 39945,
1505
+ 36489,
1506
+ 33079
1507
+ ],
1508
+ "step": 2000
1509
+ },
1510
+ {
1511
+ "epoch": 6.54,
1512
+ "learning_rate": 0.0005024999999999999,
1513
+ "loss": 2.9861,
1514
+ "step": 2010
1515
+ },
1516
+ {
1517
+ "epoch": 6.58,
1518
+ "learning_rate": 0.000505,
1519
+ "loss": 3.0033,
1520
+ "step": 2020
1521
+ },
1522
+ {
1523
+ "epoch": 6.61,
1524
+ "learning_rate": 0.0005074999999999999,
1525
+ "loss": 2.8877,
1526
+ "step": 2030
1527
+ },
1528
+ {
1529
+ "epoch": 6.64,
1530
+ "learning_rate": 0.00051,
1531
+ "loss": 2.9709,
1532
+ "step": 2040
1533
+ },
1534
+ {
1535
+ "epoch": 6.67,
1536
+ "learning_rate": 0.0005124999999999999,
1537
+ "loss": 2.8549,
1538
+ "step": 2050
1539
+ },
1540
+ {
1541
+ "epoch": 6.71,
1542
+ "learning_rate": 0.000515,
1543
+ "loss": 2.9543,
1544
+ "step": 2060
1545
+ },
1546
+ {
1547
+ "epoch": 6.74,
1548
+ "learning_rate": 0.0005175,
1549
+ "loss": 2.9383,
1550
+ "step": 2070
1551
+ },
1552
+ {
1553
+ "epoch": 6.77,
1554
+ "learning_rate": 0.0005200000000000001,
1555
+ "loss": 2.8779,
1556
+ "step": 2080
1557
+ },
1558
+ {
1559
+ "epoch": 6.8,
1560
+ "learning_rate": 0.0005225,
1561
+ "loss": 2.969,
1562
+ "step": 2090
1563
+ },
1564
+ {
1565
+ "epoch": 6.84,
1566
+ "learning_rate": 0.0005250000000000001,
1567
+ "loss": 2.968,
1568
+ "step": 2100
1569
+ },
1570
+ {
1571
+ "epoch": 6.87,
1572
+ "learning_rate": 0.0005275,
1573
+ "loss": 2.8835,
1574
+ "step": 2110
1575
+ },
1576
+ {
1577
+ "epoch": 6.9,
1578
+ "learning_rate": 0.0005300000000000001,
1579
+ "loss": 3.0172,
1580
+ "step": 2120
1581
+ },
1582
+ {
1583
+ "epoch": 6.93,
1584
+ "learning_rate": 0.0005325,
1585
+ "loss": 2.9274,
1586
+ "step": 2130
1587
+ },
1588
+ {
1589
+ "epoch": 6.97,
1590
+ "learning_rate": 0.000535,
1591
+ "loss": 2.9879,
1592
+ "step": 2140
1593
+ },
1594
+ {
1595
+ "epoch": 7.0,
1596
+ "learning_rate": 0.0005375,
1597
+ "loss": 2.9297,
1598
+ "step": 2150
1599
+ },
1600
+ {
1601
+ "epoch": 7.03,
1602
+ "learning_rate": 0.00054,
1603
+ "loss": 2.8153,
1604
+ "step": 2160
1605
+ },
1606
+ {
1607
+ "epoch": 7.06,
1608
+ "learning_rate": 0.0005425,
1609
+ "loss": 2.8727,
1610
+ "step": 2170
1611
+ },
1612
+ {
1613
+ "epoch": 7.1,
1614
+ "learning_rate": 0.000545,
1615
+ "loss": 2.8239,
1616
+ "step": 2180
1617
+ },
1618
+ {
1619
+ "epoch": 7.13,
1620
+ "learning_rate": 0.0005475,
1621
+ "loss": 2.9188,
1622
+ "step": 2190
1623
+ },
1624
+ {
1625
+ "epoch": 7.16,
1626
+ "learning_rate": 0.00055,
1627
+ "loss": 2.8291,
1628
+ "step": 2200
1629
+ },
1630
+ {
1631
+ "epoch": 7.16,
1632
+ "eval_bp": 1.0,
1633
+ "eval_counts": [
1634
+ 26097,
1635
+ 17579,
1636
+ 12717,
1637
+ 9479
1638
+ ],
1639
+ "eval_loss": 1.545904278755188,
1640
+ "eval_precisions": [
1641
+ 55.22005924672027,
1642
+ 40.13745233691805,
1643
+ 31.5221971593585,
1644
+ 25.666089028484784
1645
+ ],
1646
+ "eval_ref_len": 41379,
1647
+ "eval_runtime": 165.394,
1648
+ "eval_samples_per_second": 20.938,
1649
+ "eval_score": 36.59364849906923,
1650
+ "eval_steps_per_second": 20.938,
1651
+ "eval_sys_len": 47260,
1652
+ "eval_totals": [
1653
+ 47260,
1654
+ 43797,
1655
+ 40343,
1656
+ 36932
1657
+ ],
1658
+ "step": 2200
1659
+ },
1660
+ {
1661
+ "epoch": 7.19,
1662
+ "learning_rate": 0.0005525,
1663
+ "loss": 2.8052,
1664
+ "step": 2210
1665
+ },
1666
+ {
1667
+ "epoch": 7.23,
1668
+ "learning_rate": 0.000555,
1669
+ "loss": 2.8976,
1670
+ "step": 2220
1671
+ },
1672
+ {
1673
+ "epoch": 7.26,
1674
+ "learning_rate": 0.0005575,
1675
+ "loss": 2.8193,
1676
+ "step": 2230
1677
+ },
1678
+ {
1679
+ "epoch": 7.29,
1680
+ "learning_rate": 0.0005600000000000001,
1681
+ "loss": 2.9084,
1682
+ "step": 2240
1683
+ },
1684
+ {
1685
+ "epoch": 7.33,
1686
+ "learning_rate": 0.0005625000000000001,
1687
+ "loss": 2.8175,
1688
+ "step": 2250
1689
+ },
1690
+ {
1691
+ "epoch": 7.36,
1692
+ "learning_rate": 0.000565,
1693
+ "loss": 2.8522,
1694
+ "step": 2260
1695
+ },
1696
+ {
1697
+ "epoch": 7.39,
1698
+ "learning_rate": 0.0005675,
1699
+ "loss": 2.8453,
1700
+ "step": 2270
1701
+ },
1702
+ {
1703
+ "epoch": 7.42,
1704
+ "learning_rate": 0.00057,
1705
+ "loss": 2.9098,
1706
+ "step": 2280
1707
+ },
1708
+ {
1709
+ "epoch": 7.46,
1710
+ "learning_rate": 0.0005725,
1711
+ "loss": 2.7735,
1712
+ "step": 2290
1713
+ },
1714
+ {
1715
+ "epoch": 7.49,
1716
+ "learning_rate": 0.000575,
1717
+ "loss": 2.8275,
1718
+ "step": 2300
1719
+ },
1720
+ {
1721
+ "epoch": 7.52,
1722
+ "learning_rate": 0.0005775,
1723
+ "loss": 2.8569,
1724
+ "step": 2310
1725
+ },
1726
+ {
1727
+ "epoch": 7.55,
1728
+ "learning_rate": 0.00058,
1729
+ "loss": 2.7792,
1730
+ "step": 2320
1731
+ },
1732
+ {
1733
+ "epoch": 7.59,
1734
+ "learning_rate": 0.0005825,
1735
+ "loss": 2.8173,
1736
+ "step": 2330
1737
+ },
1738
+ {
1739
+ "epoch": 7.62,
1740
+ "learning_rate": 0.000585,
1741
+ "loss": 2.8759,
1742
+ "step": 2340
1743
+ },
1744
+ {
1745
+ "epoch": 7.65,
1746
+ "learning_rate": 0.0005875,
1747
+ "loss": 2.8538,
1748
+ "step": 2350
1749
+ },
1750
+ {
1751
+ "epoch": 7.68,
1752
+ "learning_rate": 0.00059,
1753
+ "loss": 2.902,
1754
+ "step": 2360
1755
+ },
1756
+ {
1757
+ "epoch": 7.72,
1758
+ "learning_rate": 0.0005925,
1759
+ "loss": 2.7693,
1760
+ "step": 2370
1761
+ },
1762
+ {
1763
+ "epoch": 7.75,
1764
+ "learning_rate": 0.0005949999999999999,
1765
+ "loss": 2.8489,
1766
+ "step": 2380
1767
+ },
1768
+ {
1769
+ "epoch": 7.78,
1770
+ "learning_rate": 0.0005975,
1771
+ "loss": 2.8515,
1772
+ "step": 2390
1773
+ },
1774
+ {
1775
+ "epoch": 7.81,
1776
+ "learning_rate": 0.0006,
1777
+ "loss": 2.8358,
1778
+ "step": 2400
1779
+ },
1780
+ {
1781
+ "epoch": 7.81,
1782
+ "eval_bp": 0.951493698303458,
1783
+ "eval_counts": [
1784
+ 27661,
1785
+ 18999,
1786
+ 13858,
1787
+ 10389
1788
+ ],
1789
+ "eval_loss": 1.5499262809753418,
1790
+ "eval_precisions": [
1791
+ 70.17174459017225,
1792
+ 52.839581710980084,
1793
+ 42.60722521137586,
1794
+ 35.65324822402965
1795
+ ],
1796
+ "eval_ref_len": 41379,
1797
+ "eval_runtime": 72.495,
1798
+ "eval_samples_per_second": 47.769,
1799
+ "eval_score": 46.35343428916555,
1800
+ "eval_steps_per_second": 47.769,
1801
+ "eval_sys_len": 39419,
1802
+ "eval_totals": [
1803
+ 39419,
1804
+ 35956,
1805
+ 32525,
1806
+ 29139
1807
+ ],
1808
+ "step": 2400
1809
+ },
1810
+ {
1811
+ "epoch": 7.85,
1812
+ "learning_rate": 0.0006025000000000001,
1813
+ "loss": 2.8804,
1814
+ "step": 2410
1815
+ },
1816
+ {
1817
+ "epoch": 7.88,
1818
+ "learning_rate": 0.000605,
1819
+ "loss": 2.7958,
1820
+ "step": 2420
1821
+ },
1822
+ {
1823
+ "epoch": 7.91,
1824
+ "learning_rate": 0.0006075000000000001,
1825
+ "loss": 2.8646,
1826
+ "step": 2430
1827
+ },
1828
+ {
1829
+ "epoch": 7.94,
1830
+ "learning_rate": 0.00061,
1831
+ "loss": 2.7276,
1832
+ "step": 2440
1833
+ },
1834
+ {
1835
+ "epoch": 7.98,
1836
+ "learning_rate": 0.0006125000000000001,
1837
+ "loss": 2.8928,
1838
+ "step": 2450
1839
+ },
1840
+ {
1841
+ "epoch": 8.01,
1842
+ "learning_rate": 0.000615,
1843
+ "loss": 2.8545,
1844
+ "step": 2460
1845
+ },
1846
+ {
1847
+ "epoch": 8.04,
1848
+ "learning_rate": 0.0006175000000000001,
1849
+ "loss": 2.7296,
1850
+ "step": 2470
1851
+ },
1852
+ {
1853
+ "epoch": 8.07,
1854
+ "learning_rate": 0.00062,
1855
+ "loss": 2.7535,
1856
+ "step": 2480
1857
+ },
1858
+ {
1859
+ "epoch": 8.11,
1860
+ "learning_rate": 0.0006225000000000001,
1861
+ "loss": 2.8058,
1862
+ "step": 2490
1863
+ },
1864
+ {
1865
+ "epoch": 8.14,
1866
+ "learning_rate": 0.000625,
1867
+ "loss": 2.7666,
1868
+ "step": 2500
1869
+ },
1870
+ {
1871
+ "epoch": 8.17,
1872
+ "learning_rate": 0.0006274999999999999,
1873
+ "loss": 2.7758,
1874
+ "step": 2510
1875
+ },
1876
+ {
1877
+ "epoch": 8.2,
1878
+ "learning_rate": 0.00063,
1879
+ "loss": 2.6975,
1880
+ "step": 2520
1881
+ },
1882
+ {
1883
+ "epoch": 8.24,
1884
+ "learning_rate": 0.0006324999999999999,
1885
+ "loss": 2.7532,
1886
+ "step": 2530
1887
+ },
1888
+ {
1889
+ "epoch": 8.27,
1890
+ "learning_rate": 0.000635,
1891
+ "loss": 2.7222,
1892
+ "step": 2540
1893
+ },
1894
+ {
1895
+ "epoch": 8.3,
1896
+ "learning_rate": 0.0006374999999999999,
1897
+ "loss": 2.8252,
1898
+ "step": 2550
1899
+ },
1900
+ {
1901
+ "epoch": 8.33,
1902
+ "learning_rate": 0.00064,
1903
+ "loss": 2.8289,
1904
+ "step": 2560
1905
+ },
1906
+ {
1907
+ "epoch": 8.37,
1908
+ "learning_rate": 0.0006425,
1909
+ "loss": 2.7046,
1910
+ "step": 2570
1911
+ },
1912
+ {
1913
+ "epoch": 8.4,
1914
+ "learning_rate": 0.0006450000000000001,
1915
+ "loss": 2.7199,
1916
+ "step": 2580
1917
+ },
1918
+ {
1919
+ "epoch": 8.43,
1920
+ "learning_rate": 0.0006475,
1921
+ "loss": 2.6622,
1922
+ "step": 2590
1923
+ },
1924
+ {
1925
+ "epoch": 8.46,
1926
+ "learning_rate": 0.0006500000000000001,
1927
+ "loss": 2.764,
1928
+ "step": 2600
1929
+ },
1930
+ {
1931
+ "epoch": 8.46,
1932
+ "eval_bp": 1.0,
1933
+ "eval_counts": [
1934
+ 26024,
1935
+ 17669,
1936
+ 12817,
1937
+ 9561
1938
+ ],
1939
+ "eval_loss": 1.508206844329834,
1940
+ "eval_precisions": [
1941
+ 55.35019248356977,
1942
+ 40.56803049088488,
1943
+ 31.96418774003691,
1944
+ 26.063133791298657
1945
+ ],
1946
+ "eval_ref_len": 41379,
1947
+ "eval_runtime": 173.7893,
1948
+ "eval_samples_per_second": 19.926,
1949
+ "eval_score": 36.9826767402354,
1950
+ "eval_steps_per_second": 19.926,
1951
+ "eval_sys_len": 47017,
1952
+ "eval_totals": [
1953
+ 47017,
1954
+ 43554,
1955
+ 40098,
1956
+ 36684
1957
+ ],
1958
+ "step": 2600
1959
+ },
1960
+ {
1961
+ "epoch": 8.5,
1962
+ "learning_rate": 0.0006525,
1963
+ "loss": 2.7553,
1964
+ "step": 2610
1965
+ },
1966
+ {
1967
+ "epoch": 8.53,
1968
+ "learning_rate": 0.0006550000000000001,
1969
+ "loss": 2.7883,
1970
+ "step": 2620
1971
+ },
1972
+ {
1973
+ "epoch": 8.56,
1974
+ "learning_rate": 0.0006575,
1975
+ "loss": 2.7572,
1976
+ "step": 2630
1977
+ },
1978
+ {
1979
+ "epoch": 8.59,
1980
+ "learning_rate": 0.00066,
1981
+ "loss": 2.7026,
1982
+ "step": 2640
1983
+ },
1984
+ {
1985
+ "epoch": 8.63,
1986
+ "learning_rate": 0.0006625,
1987
+ "loss": 2.7265,
1988
+ "step": 2650
1989
+ },
1990
+ {
1991
+ "epoch": 8.66,
1992
+ "learning_rate": 0.000665,
1993
+ "loss": 2.5958,
1994
+ "step": 2660
1995
+ },
1996
+ {
1997
+ "epoch": 8.69,
1998
+ "learning_rate": 0.0006675,
1999
+ "loss": 2.69,
2000
+ "step": 2670
2001
+ },
2002
+ {
2003
+ "epoch": 8.72,
2004
+ "learning_rate": 0.00067,
2005
+ "loss": 2.7464,
2006
+ "step": 2680
2007
+ },
2008
+ {
2009
+ "epoch": 8.76,
2010
+ "learning_rate": 0.0006725,
2011
+ "loss": 2.7263,
2012
+ "step": 2690
2013
+ },
2014
+ {
2015
+ "epoch": 8.79,
2016
+ "learning_rate": 0.000675,
2017
+ "loss": 2.7336,
2018
+ "step": 2700
2019
+ },
2020
+ {
2021
+ "epoch": 8.82,
2022
+ "learning_rate": 0.0006775,
2023
+ "loss": 2.7562,
2024
+ "step": 2710
2025
+ },
2026
+ {
2027
+ "epoch": 8.86,
2028
+ "learning_rate": 0.00068,
2029
+ "loss": 2.7315,
2030
+ "step": 2720
2031
+ },
2032
+ {
2033
+ "epoch": 8.89,
2034
+ "learning_rate": 0.0006825000000000001,
2035
+ "loss": 2.7706,
2036
+ "step": 2730
2037
+ },
2038
+ {
2039
+ "epoch": 8.92,
2040
+ "learning_rate": 0.0006850000000000001,
2041
+ "loss": 2.6668,
2042
+ "step": 2740
2043
+ },
2044
+ {
2045
+ "epoch": 8.95,
2046
+ "learning_rate": 0.0006875,
2047
+ "loss": 2.7939,
2048
+ "step": 2750
2049
+ },
2050
+ {
2051
+ "epoch": 8.99,
2052
+ "learning_rate": 0.00069,
2053
+ "loss": 2.7894,
2054
+ "step": 2760
2055
+ },
2056
+ {
2057
+ "epoch": 9.02,
2058
+ "learning_rate": 0.0006925,
2059
+ "loss": 2.7196,
2060
+ "step": 2770
2061
+ },
2062
+ {
2063
+ "epoch": 9.05,
2064
+ "learning_rate": 0.000695,
2065
+ "loss": 2.6768,
2066
+ "step": 2780
2067
+ },
2068
+ {
2069
+ "epoch": 9.08,
2070
+ "learning_rate": 0.0006975,
2071
+ "loss": 2.5136,
2072
+ "step": 2790
2073
+ },
2074
+ {
2075
+ "epoch": 9.12,
2076
+ "learning_rate": 0.0007,
2077
+ "loss": 2.6378,
2078
+ "step": 2800
2079
+ },
2080
+ {
2081
+ "epoch": 9.12,
2082
+ "eval_bp": 0.982521383577959,
2083
+ "eval_counts": [
2084
+ 27410,
2085
+ 18929,
2086
+ 13880,
2087
+ 10452
2088
+ ],
2089
+ "eval_loss": 1.4984954595565796,
2090
+ "eval_precisions": [
2091
+ 67.40937484629384,
2092
+ 50.88577649936826,
2093
+ 41.13445751711466,
2094
+ 34.460929772502475
2095
+ ],
2096
+ "eval_ref_len": 41379,
2097
+ "eval_runtime": 88.8496,
2098
+ "eval_samples_per_second": 38.976,
2099
+ "eval_score": 46.13753405463289,
2100
+ "eval_steps_per_second": 38.976,
2101
+ "eval_sys_len": 40662,
2102
+ "eval_totals": [
2103
+ 40662,
2104
+ 37199,
2105
+ 33743,
2106
+ 30330
2107
+ ],
2108
+ "step": 2800
2109
+ },
2110
+ {
2111
+ "epoch": 9.15,
2112
+ "learning_rate": 0.0007025,
2113
+ "loss": 2.6048,
2114
+ "step": 2810
2115
+ },
2116
+ {
2117
+ "epoch": 9.18,
2118
+ "learning_rate": 0.000705,
2119
+ "loss": 2.6136,
2120
+ "step": 2820
2121
+ },
2122
+ {
2123
+ "epoch": 9.21,
2124
+ "learning_rate": 0.0007075,
2125
+ "loss": 2.5976,
2126
+ "step": 2830
2127
+ },
2128
+ {
2129
+ "epoch": 9.25,
2130
+ "learning_rate": 0.00071,
2131
+ "loss": 2.6645,
2132
+ "step": 2840
2133
+ },
2134
+ {
2135
+ "epoch": 9.28,
2136
+ "learning_rate": 0.0007125,
2137
+ "loss": 2.7466,
2138
+ "step": 2850
2139
+ },
2140
+ {
2141
+ "epoch": 9.31,
2142
+ "learning_rate": 0.000715,
2143
+ "loss": 2.6573,
2144
+ "step": 2860
2145
+ },
2146
+ {
2147
+ "epoch": 9.34,
2148
+ "learning_rate": 0.0007175,
2149
+ "loss": 2.6746,
2150
+ "step": 2870
2151
+ },
2152
+ {
2153
+ "epoch": 9.38,
2154
+ "learning_rate": 0.0007199999999999999,
2155
+ "loss": 2.6083,
2156
+ "step": 2880
2157
+ },
2158
+ {
2159
+ "epoch": 9.41,
2160
+ "learning_rate": 0.0007225,
2161
+ "loss": 2.718,
2162
+ "step": 2890
2163
+ },
2164
+ {
2165
+ "epoch": 9.44,
2166
+ "learning_rate": 0.000725,
2167
+ "loss": 2.6234,
2168
+ "step": 2900
2169
+ },
2170
+ {
2171
+ "epoch": 9.47,
2172
+ "learning_rate": 0.0007275000000000001,
2173
+ "loss": 2.6455,
2174
+ "step": 2910
2175
+ },
2176
+ {
2177
+ "epoch": 9.51,
2178
+ "learning_rate": 0.00073,
2179
+ "loss": 2.6674,
2180
+ "step": 2920
2181
+ },
2182
+ {
2183
+ "epoch": 9.54,
2184
+ "learning_rate": 0.0007325000000000001,
2185
+ "loss": 2.6671,
2186
+ "step": 2930
2187
+ },
2188
+ {
2189
+ "epoch": 9.57,
2190
+ "learning_rate": 0.000735,
2191
+ "loss": 2.6839,
2192
+ "step": 2940
2193
+ },
2194
+ {
2195
+ "epoch": 9.6,
2196
+ "learning_rate": 0.0007375000000000001,
2197
+ "loss": 2.7052,
2198
+ "step": 2950
2199
+ },
2200
+ {
2201
+ "epoch": 9.64,
2202
+ "learning_rate": 0.00074,
2203
+ "loss": 2.7401,
2204
+ "step": 2960
2205
+ },
2206
+ {
2207
+ "epoch": 9.67,
2208
+ "learning_rate": 0.0007425000000000001,
2209
+ "loss": 2.6638,
2210
+ "step": 2970
2211
+ },
2212
+ {
2213
+ "epoch": 9.7,
2214
+ "learning_rate": 0.000745,
2215
+ "loss": 2.5944,
2216
+ "step": 2980
2217
+ },
2218
+ {
2219
+ "epoch": 9.73,
2220
+ "learning_rate": 0.0007475000000000001,
2221
+ "loss": 2.6262,
2222
+ "step": 2990
2223
+ },
2224
+ {
2225
+ "epoch": 9.77,
2226
+ "learning_rate": 0.00075,
2227
+ "loss": 2.701,
2228
+ "step": 3000
2229
+ },
2230
+ {
2231
+ "epoch": 9.77,
2232
+ "eval_bp": 1.0,
2233
+ "eval_counts": [
2234
+ 26685,
2235
+ 18421,
2236
+ 13517,
2237
+ 10181
2238
+ ],
2239
+ "eval_loss": 1.4741408824920654,
2240
+ "eval_precisions": [
2241
+ 63.57355568790947,
2242
+ 47.83184461985874,
2243
+ 38.558306709265175,
2244
+ 32.17457257529311
2245
+ ],
2246
+ "eval_ref_len": 41379,
2247
+ "eval_runtime": 129.3598,
2248
+ "eval_samples_per_second": 26.77,
2249
+ "eval_score": 44.071323498930454,
2250
+ "eval_steps_per_second": 26.77,
2251
+ "eval_sys_len": 41975,
2252
+ "eval_totals": [
2253
+ 41975,
2254
+ 38512,
2255
+ 35056,
2256
+ 31643
2257
+ ],
2258
+ "step": 3000
2259
+ },
2260
+ {
2261
+ "epoch": 9.8,
2262
+ "learning_rate": 0.0007524999999999999,
2263
+ "loss": 2.6454,
2264
+ "step": 3010
2265
+ },
2266
+ {
2267
+ "epoch": 9.83,
2268
+ "learning_rate": 0.000755,
2269
+ "loss": 2.7318,
2270
+ "step": 3020
2271
+ },
2272
+ {
2273
+ "epoch": 9.86,
2274
+ "learning_rate": 0.0007574999999999999,
2275
+ "loss": 2.6678,
2276
+ "step": 3030
2277
+ },
2278
+ {
2279
+ "epoch": 9.9,
2280
+ "learning_rate": 0.00076,
2281
+ "loss": 2.6132,
2282
+ "step": 3040
2283
+ },
2284
+ {
2285
+ "epoch": 9.93,
2286
+ "learning_rate": 0.0007624999999999999,
2287
+ "loss": 2.6956,
2288
+ "step": 3050
2289
+ },
2290
+ {
2291
+ "epoch": 9.96,
2292
+ "learning_rate": 0.0007650000000000001,
2293
+ "loss": 2.5574,
2294
+ "step": 3060
2295
+ },
2296
+ {
2297
+ "epoch": 9.99,
2298
+ "learning_rate": 0.0007675,
2299
+ "loss": 2.6114,
2300
+ "step": 3070
2301
+ },
2302
+ {
2303
+ "epoch": 10.03,
2304
+ "learning_rate": 0.0007700000000000001,
2305
+ "loss": 2.6667,
2306
+ "step": 3080
2307
+ },
2308
+ {
2309
+ "epoch": 10.06,
2310
+ "learning_rate": 0.0007725,
2311
+ "loss": 2.5687,
2312
+ "step": 3090
2313
+ },
2314
+ {
2315
+ "epoch": 10.09,
2316
+ "learning_rate": 0.0007750000000000001,
2317
+ "loss": 2.4745,
2318
+ "step": 3100
2319
+ },
2320
+ {
2321
+ "epoch": 10.12,
2322
+ "learning_rate": 0.0007775,
2323
+ "loss": 2.6286,
2324
+ "step": 3110
2325
+ },
2326
+ {
2327
+ "epoch": 10.16,
2328
+ "learning_rate": 0.0007800000000000001,
2329
+ "loss": 2.6025,
2330
+ "step": 3120
2331
+ },
2332
+ {
2333
+ "epoch": 10.19,
2334
+ "learning_rate": 0.0007825,
2335
+ "loss": 2.5623,
2336
+ "step": 3130
2337
+ },
2338
+ {
2339
+ "epoch": 10.22,
2340
+ "learning_rate": 0.000785,
2341
+ "loss": 2.6493,
2342
+ "step": 3140
2343
+ },
2344
+ {
2345
+ "epoch": 10.26,
2346
+ "learning_rate": 0.0007875,
2347
+ "loss": 2.5866,
2348
+ "step": 3150
2349
+ },
2350
+ {
2351
+ "epoch": 10.29,
2352
+ "learning_rate": 0.00079,
2353
+ "loss": 2.5541,
2354
+ "step": 3160
2355
+ },
2356
+ {
2357
+ "epoch": 10.32,
2358
+ "learning_rate": 0.0007925,
2359
+ "loss": 2.6685,
2360
+ "step": 3170
2361
+ },
2362
+ {
2363
+ "epoch": 10.35,
2364
+ "learning_rate": 0.000795,
2365
+ "loss": 2.5534,
2366
+ "step": 3180
2367
+ },
2368
+ {
2369
+ "epoch": 10.39,
2370
+ "learning_rate": 0.0007975,
2371
+ "loss": 2.6261,
2372
+ "step": 3190
2373
+ },
2374
+ {
2375
+ "epoch": 10.42,
2376
+ "learning_rate": 0.0008,
2377
+ "loss": 2.6547,
2378
+ "step": 3200
2379
+ },
2380
+ {
2381
+ "epoch": 10.42,
2382
+ "eval_bp": 0.9986941383168589,
2383
+ "eval_counts": [
2384
+ 27057,
2385
+ 18731,
2386
+ 13763,
2387
+ 10375
2388
+ ],
2389
+ "eval_loss": 1.4648829698562622,
2390
+ "eval_precisions": [
2391
+ 65.47368421052632,
2392
+ 49.47176588664096,
2393
+ 40.00174388188107,
2394
+ 33.47422081693231
2395
+ ],
2396
+ "eval_ref_len": 41379,
2397
+ "eval_runtime": 105.6078,
2398
+ "eval_samples_per_second": 32.791,
2399
+ "eval_score": 45.57596776445642,
2400
+ "eval_steps_per_second": 32.791,
2401
+ "eval_sys_len": 41325,
2402
+ "eval_totals": [
2403
+ 41325,
2404
+ 37862,
2405
+ 34406,
2406
+ 30994
2407
+ ],
2408
+ "step": 3200
2409
+ },
2410
+ {
2411
+ "epoch": 10.45,
2412
+ "learning_rate": 0.0008025,
2413
+ "loss": 2.4607,
2414
+ "step": 3210
2415
+ },
2416
+ {
2417
+ "epoch": 10.48,
2418
+ "learning_rate": 0.000805,
2419
+ "loss": 2.596,
2420
+ "step": 3220
2421
+ },
2422
+ {
2423
+ "epoch": 10.52,
2424
+ "learning_rate": 0.0008075000000000001,
2425
+ "loss": 2.5666,
2426
+ "step": 3230
2427
+ },
2428
+ {
2429
+ "epoch": 10.55,
2430
+ "learning_rate": 0.0008100000000000001,
2431
+ "loss": 2.5642,
2432
+ "step": 3240
2433
+ },
2434
+ {
2435
+ "epoch": 10.58,
2436
+ "learning_rate": 0.0008125000000000001,
2437
+ "loss": 2.5876,
2438
+ "step": 3250
2439
+ },
2440
+ {
2441
+ "epoch": 10.61,
2442
+ "learning_rate": 0.000815,
2443
+ "loss": 2.5194,
2444
+ "step": 3260
2445
+ },
2446
+ {
2447
+ "epoch": 10.65,
2448
+ "learning_rate": 0.0008175,
2449
+ "loss": 2.71,
2450
+ "step": 3270
2451
+ },
2452
+ {
2453
+ "epoch": 10.68,
2454
+ "learning_rate": 0.00082,
2455
+ "loss": 2.5289,
2456
+ "step": 3280
2457
+ },
2458
+ {
2459
+ "epoch": 10.71,
2460
+ "learning_rate": 0.0008225,
2461
+ "loss": 2.5532,
2462
+ "step": 3290
2463
+ },
2464
+ {
2465
+ "epoch": 10.74,
2466
+ "learning_rate": 0.000825,
2467
+ "loss": 2.5693,
2468
+ "step": 3300
2469
+ },
2470
+ {
2471
+ "epoch": 10.78,
2472
+ "learning_rate": 0.0008275,
2473
+ "loss": 2.5581,
2474
+ "step": 3310
2475
+ },
2476
+ {
2477
+ "epoch": 10.81,
2478
+ "learning_rate": 0.00083,
2479
+ "loss": 2.5051,
2480
+ "step": 3320
2481
+ },
2482
+ {
2483
+ "epoch": 10.84,
2484
+ "learning_rate": 0.0008325,
2485
+ "loss": 2.5048,
2486
+ "step": 3330
2487
+ },
2488
+ {
2489
+ "epoch": 10.87,
2490
+ "learning_rate": 0.000835,
2491
+ "loss": 2.5013,
2492
+ "step": 3340
2493
+ },
2494
+ {
2495
+ "epoch": 10.91,
2496
+ "learning_rate": 0.0008375,
2497
+ "loss": 2.5961,
2498
+ "step": 3350
2499
+ },
2500
+ {
2501
+ "epoch": 10.94,
2502
+ "learning_rate": 0.00084,
2503
+ "loss": 2.5747,
2504
+ "step": 3360
2505
+ },
2506
+ {
2507
+ "epoch": 10.97,
2508
+ "learning_rate": 0.0008425,
2509
+ "loss": 2.513,
2510
+ "step": 3370
2511
+ },
2512
+ {
2513
+ "epoch": 11.0,
2514
+ "learning_rate": 0.0008449999999999999,
2515
+ "loss": 2.5362,
2516
+ "step": 3380
2517
+ },
2518
+ {
2519
+ "epoch": 11.04,
2520
+ "learning_rate": 0.0008475000000000001,
2521
+ "loss": 2.4747,
2522
+ "step": 3390
2523
+ },
2524
+ {
2525
+ "epoch": 11.07,
2526
+ "learning_rate": 0.00085,
2527
+ "loss": 2.4868,
2528
+ "step": 3400
2529
+ },
2530
+ {
2531
+ "epoch": 11.07,
2532
+ "eval_bp": 1.0,
2533
+ "eval_counts": [
2534
+ 27887,
2535
+ 19470,
2536
+ 14351,
2537
+ 10835
2538
+ ],
2539
+ "eval_loss": 1.463901162147522,
2540
+ "eval_precisions": [
2541
+ 65.5994918962151,
2542
+ 49.861708666256916,
2543
+ 40.29255692506388,
2544
+ 33.617747440273035
2545
+ ],
2546
+ "eval_ref_len": 41379,
2547
+ "eval_runtime": 111.3022,
2548
+ "eval_samples_per_second": 31.113,
2549
+ "eval_score": 45.87913991332588,
2550
+ "eval_steps_per_second": 31.113,
2551
+ "eval_sys_len": 42511,
2552
+ "eval_totals": [
2553
+ 42511,
2554
+ 39048,
2555
+ 35617,
2556
+ 32230
2557
+ ],
2558
+ "step": 3400
2559
+ },
2560
+ {
2561
+ "epoch": 11.1,
2562
+ "learning_rate": 0.0008525000000000001,
2563
+ "loss": 2.4984,
2564
+ "step": 3410
2565
+ },
2566
+ {
2567
+ "epoch": 11.13,
2568
+ "learning_rate": 0.000855,
2569
+ "loss": 2.4708,
2570
+ "step": 3420
2571
+ },
2572
+ {
2573
+ "epoch": 11.17,
2574
+ "learning_rate": 0.0008575000000000001,
2575
+ "loss": 2.4994,
2576
+ "step": 3430
2577
+ },
2578
+ {
2579
+ "epoch": 11.2,
2580
+ "learning_rate": 0.00086,
2581
+ "loss": 2.3728,
2582
+ "step": 3440
2583
+ },
2584
+ {
2585
+ "epoch": 11.23,
2586
+ "learning_rate": 0.0008625000000000001,
2587
+ "loss": 2.4748,
2588
+ "step": 3450
2589
+ },
2590
+ {
2591
+ "epoch": 11.26,
2592
+ "learning_rate": 0.000865,
2593
+ "loss": 2.4959,
2594
+ "step": 3460
2595
+ },
2596
+ {
2597
+ "epoch": 11.3,
2598
+ "learning_rate": 0.0008675000000000001,
2599
+ "loss": 2.513,
2600
+ "step": 3470
2601
+ },
2602
+ {
2603
+ "epoch": 11.33,
2604
+ "learning_rate": 0.00087,
2605
+ "loss": 2.5693,
2606
+ "step": 3480
2607
+ },
2608
+ {
2609
+ "epoch": 11.36,
2610
+ "learning_rate": 0.0008725000000000001,
2611
+ "loss": 2.5396,
2612
+ "step": 3490
2613
+ },
2614
+ {
2615
+ "epoch": 11.39,
2616
+ "learning_rate": 0.000875,
2617
+ "loss": 2.5051,
2618
+ "step": 3500
2619
+ },
2620
+ {
2621
+ "epoch": 11.43,
2622
+ "learning_rate": 0.0008774999999999999,
2623
+ "loss": 2.4765,
2624
+ "step": 3510
2625
+ },
2626
+ {
2627
+ "epoch": 11.46,
2628
+ "learning_rate": 0.00088,
2629
+ "loss": 2.4833,
2630
+ "step": 3520
2631
+ },
2632
+ {
2633
+ "epoch": 11.49,
2634
+ "learning_rate": 0.0008824999999999999,
2635
+ "loss": 2.509,
2636
+ "step": 3530
2637
+ },
2638
+ {
2639
+ "epoch": 11.52,
2640
+ "learning_rate": 0.000885,
2641
+ "loss": 2.5082,
2642
+ "step": 3540
2643
+ },
2644
+ {
2645
+ "epoch": 11.56,
2646
+ "learning_rate": 0.0008874999999999999,
2647
+ "loss": 2.5851,
2648
+ "step": 3550
2649
+ },
2650
+ {
2651
+ "epoch": 11.59,
2652
+ "learning_rate": 0.0008900000000000001,
2653
+ "loss": 2.4521,
2654
+ "step": 3560
2655
+ },
2656
+ {
2657
+ "epoch": 11.62,
2658
+ "learning_rate": 0.0008925,
2659
+ "loss": 2.4873,
2660
+ "step": 3570
2661
+ },
2662
+ {
2663
+ "epoch": 11.65,
2664
+ "learning_rate": 0.0008950000000000001,
2665
+ "loss": 2.4671,
2666
+ "step": 3580
2667
+ },
2668
+ {
2669
+ "epoch": 11.69,
2670
+ "learning_rate": 0.0008975,
2671
+ "loss": 2.5156,
2672
+ "step": 3590
2673
+ },
2674
+ {
2675
+ "epoch": 11.72,
2676
+ "learning_rate": 0.0009000000000000001,
2677
+ "loss": 2.4155,
2678
+ "step": 3600
2679
+ },
2680
+ {
2681
+ "epoch": 11.72,
2682
+ "eval_bp": 1.0,
2683
+ "eval_counts": [
2684
+ 26675,
2685
+ 18446,
2686
+ 13573,
2687
+ 10248
2688
+ ],
2689
+ "eval_loss": 1.4466536045074463,
2690
+ "eval_precisions": [
2691
+ 64.46350894151765,
2692
+ 48.64836353087006,
2693
+ 39.38655291488929,
2694
+ 33.009083295754685
2695
+ ],
2696
+ "eval_ref_len": 41379,
2697
+ "eval_runtime": 110.662,
2698
+ "eval_samples_per_second": 31.293,
2699
+ "eval_score": 44.93563710796045,
2700
+ "eval_steps_per_second": 31.293,
2701
+ "eval_sys_len": 41380,
2702
+ "eval_totals": [
2703
+ 41380,
2704
+ 37917,
2705
+ 34461,
2706
+ 31046
2707
+ ],
2708
+ "step": 3600
2709
+ },
2710
+ {
2711
+ "epoch": 11.75,
2712
+ "learning_rate": 0.0009025,
2713
+ "loss": 2.5093,
2714
+ "step": 3610
2715
+ },
2716
+ {
2717
+ "epoch": 11.79,
2718
+ "learning_rate": 0.0009050000000000001,
2719
+ "loss": 2.532,
2720
+ "step": 3620
2721
+ },
2722
+ {
2723
+ "epoch": 11.82,
2724
+ "learning_rate": 0.0009075,
2725
+ "loss": 2.5547,
2726
+ "step": 3630
2727
+ },
2728
+ {
2729
+ "epoch": 11.85,
2730
+ "learning_rate": 0.00091,
2731
+ "loss": 2.3751,
2732
+ "step": 3640
2733
+ },
2734
+ {
2735
+ "epoch": 11.88,
2736
+ "learning_rate": 0.0009125,
2737
+ "loss": 2.5222,
2738
+ "step": 3650
2739
+ },
2740
+ {
2741
+ "epoch": 11.92,
2742
+ "learning_rate": 0.000915,
2743
+ "loss": 2.4935,
2744
+ "step": 3660
2745
+ },
2746
+ {
2747
+ "epoch": 11.95,
2748
+ "learning_rate": 0.0009175,
2749
+ "loss": 2.5756,
2750
+ "step": 3670
2751
+ },
2752
+ {
2753
+ "epoch": 11.98,
2754
+ "learning_rate": 0.00092,
2755
+ "loss": 2.5278,
2756
+ "step": 3680
2757
+ },
2758
+ {
2759
+ "epoch": 12.01,
2760
+ "learning_rate": 0.0009225,
2761
+ "loss": 2.4911,
2762
+ "step": 3690
2763
+ },
2764
+ {
2765
+ "epoch": 12.05,
2766
+ "learning_rate": 0.000925,
2767
+ "loss": 2.4734,
2768
+ "step": 3700
2769
+ },
2770
+ {
2771
+ "epoch": 12.08,
2772
+ "learning_rate": 0.0009275,
2773
+ "loss": 2.3916,
2774
+ "step": 3710
2775
+ },
2776
+ {
2777
+ "epoch": 12.11,
2778
+ "learning_rate": 0.00093,
2779
+ "loss": 2.443,
2780
+ "step": 3720
2781
+ },
2782
+ {
2783
+ "epoch": 12.14,
2784
+ "learning_rate": 0.0009325000000000001,
2785
+ "loss": 2.3527,
2786
+ "step": 3730
2787
+ },
2788
+ {
2789
+ "epoch": 12.18,
2790
+ "learning_rate": 0.0009350000000000001,
2791
+ "loss": 2.3896,
2792
+ "step": 3740
2793
+ },
2794
+ {
2795
+ "epoch": 12.21,
2796
+ "learning_rate": 0.0009375,
2797
+ "loss": 2.4948,
2798
+ "step": 3750
2799
+ },
2800
+ {
2801
+ "epoch": 12.24,
2802
+ "learning_rate": 0.00094,
2803
+ "loss": 2.4026,
2804
+ "step": 3760
2805
+ },
2806
+ {
2807
+ "epoch": 12.27,
2808
+ "learning_rate": 0.0009425,
2809
+ "loss": 2.3684,
2810
+ "step": 3770
2811
+ },
2812
+ {
2813
+ "epoch": 12.31,
2814
+ "learning_rate": 0.000945,
2815
+ "loss": 2.434,
2816
+ "step": 3780
2817
+ },
2818
+ {
2819
+ "epoch": 12.34,
2820
+ "learning_rate": 0.0009475,
2821
+ "loss": 2.426,
2822
+ "step": 3790
2823
+ },
2824
+ {
2825
+ "epoch": 12.37,
2826
+ "learning_rate": 0.00095,
2827
+ "loss": 2.4975,
2828
+ "step": 3800
2829
+ },
2830
+ {
2831
+ "epoch": 12.37,
2832
+ "eval_bp": 1.0,
2833
+ "eval_counts": [
2834
+ 26470,
2835
+ 18231,
2836
+ 13333,
2837
+ 10001
2838
+ ],
2839
+ "eval_loss": 1.4460936784744263,
2840
+ "eval_precisions": [
2841
+ 56.14593276063209,
2842
+ 41.73572638615448,
2843
+ 33.11888320333847,
2844
+ 27.118414273705902
2845
+ ],
2846
+ "eval_ref_len": 41379,
2847
+ "eval_runtime": 158.852,
2848
+ "eval_samples_per_second": 21.8,
2849
+ "eval_score": 38.08829771113567,
2850
+ "eval_steps_per_second": 21.8,
2851
+ "eval_sys_len": 47145,
2852
+ "eval_totals": [
2853
+ 47145,
2854
+ 43682,
2855
+ 40258,
2856
+ 36879
2857
+ ],
2858
+ "step": 3800
2859
+ },
2860
+ {
2861
+ "epoch": 12.4,
2862
+ "learning_rate": 0.0009525,
2863
+ "loss": 2.4968,
2864
+ "step": 3810
2865
+ },
2866
+ {
2867
+ "epoch": 12.44,
2868
+ "learning_rate": 0.000955,
2869
+ "loss": 2.4392,
2870
+ "step": 3820
2871
+ },
2872
+ {
2873
+ "epoch": 12.47,
2874
+ "learning_rate": 0.0009575,
2875
+ "loss": 2.4044,
2876
+ "step": 3830
2877
+ },
2878
+ {
2879
+ "epoch": 12.5,
2880
+ "learning_rate": 0.00096,
2881
+ "loss": 2.4352,
2882
+ "step": 3840
2883
+ },
2884
+ {
2885
+ "epoch": 12.53,
2886
+ "learning_rate": 0.0009625,
2887
+ "loss": 2.398,
2888
+ "step": 3850
2889
+ },
2890
+ {
2891
+ "epoch": 12.57,
2892
+ "learning_rate": 0.000965,
2893
+ "loss": 2.5434,
2894
+ "step": 3860
2895
+ },
2896
+ {
2897
+ "epoch": 12.6,
2898
+ "learning_rate": 0.0009675,
2899
+ "loss": 2.335,
2900
+ "step": 3870
2901
+ },
2902
+ {
2903
+ "epoch": 12.63,
2904
+ "learning_rate": 0.0009699999999999999,
2905
+ "loss": 2.41,
2906
+ "step": 3880
2907
+ },
2908
+ {
2909
+ "epoch": 12.66,
2910
+ "learning_rate": 0.0009725000000000001,
2911
+ "loss": 2.4202,
2912
+ "step": 3890
2913
+ },
2914
+ {
2915
+ "epoch": 12.7,
2916
+ "learning_rate": 0.000975,
2917
+ "loss": 2.4073,
2918
+ "step": 3900
2919
+ },
2920
+ {
2921
+ "epoch": 12.73,
2922
+ "learning_rate": 0.0009775,
2923
+ "loss": 2.3379,
2924
+ "step": 3910
2925
+ },
2926
+ {
2927
+ "epoch": 12.76,
2928
+ "learning_rate": 0.00098,
2929
+ "loss": 2.4531,
2930
+ "step": 3920
2931
+ },
2932
+ {
2933
+ "epoch": 12.79,
2934
+ "learning_rate": 0.0009825,
2935
+ "loss": 2.3775,
2936
+ "step": 3930
2937
+ },
2938
+ {
2939
+ "epoch": 12.83,
2940
+ "learning_rate": 0.000985,
2941
+ "loss": 2.4836,
2942
+ "step": 3940
2943
+ },
2944
+ {
2945
+ "epoch": 12.86,
2946
+ "learning_rate": 0.0009875,
2947
+ "loss": 2.5058,
2948
+ "step": 3950
2949
+ },
2950
+ {
2951
+ "epoch": 12.89,
2952
+ "learning_rate": 0.00099,
2953
+ "loss": 2.4745,
2954
+ "step": 3960
2955
+ },
2956
+ {
2957
+ "epoch": 12.92,
2958
+ "learning_rate": 0.0009925000000000001,
2959
+ "loss": 2.426,
2960
+ "step": 3970
2961
+ },
2962
+ {
2963
+ "epoch": 12.96,
2964
+ "learning_rate": 0.000995,
2965
+ "loss": 2.4218,
2966
+ "step": 3980
2967
+ },
2968
+ {
2969
+ "epoch": 12.99,
2970
+ "learning_rate": 0.0009975000000000001,
2971
+ "loss": 2.3397,
2972
+ "step": 3990
2973
+ },
2974
+ {
2975
+ "epoch": 13.02,
2976
+ "learning_rate": 0.001,
2977
+ "loss": 2.3011,
2978
+ "step": 4000
2979
+ },
2980
+ {
2981
+ "epoch": 13.02,
2982
+ "eval_bp": 0.9755140388093076,
2983
+ "eval_counts": [
2984
+ 28338,
2985
+ 19935,
2986
+ 14788,
2987
+ 11220
2988
+ ],
2989
+ "eval_loss": 1.4539415836334229,
2990
+ "eval_precisions": [
2991
+ 70.18178215860122,
2992
+ 54.00243803331979,
2993
+ 44.164377015888185,
2994
+ 37.280701754385966
2995
+ ],
2996
+ "eval_ref_len": 41379,
2997
+ "eval_runtime": 75.0864,
2998
+ "eval_samples_per_second": 46.12,
2999
+ "eval_score": 48.756442564740674,
3000
+ "eval_steps_per_second": 46.12,
3001
+ "eval_sys_len": 40378,
3002
+ "eval_totals": [
3003
+ 40378,
3004
+ 36915,
3005
+ 33484,
3006
+ 30096
3007
+ ],
3008
+ "step": 4000
3009
+ },
3010
+ {
3011
+ "epoch": 13.05,
3012
+ "learning_rate": 0.001,
3013
+ "loss": 2.4352,
3014
+ "step": 4010
3015
+ },
3016
+ {
3017
+ "epoch": 13.09,
3018
+ "learning_rate": 0.001,
3019
+ "loss": 2.3339,
3020
+ "step": 4020
3021
+ },
3022
+ {
3023
+ "epoch": 13.12,
3024
+ "learning_rate": 0.001,
3025
+ "loss": 2.3264,
3026
+ "step": 4030
3027
+ },
3028
+ {
3029
+ "epoch": 13.15,
3030
+ "learning_rate": 0.001,
3031
+ "loss": 2.3393,
3032
+ "step": 4040
3033
+ },
3034
+ {
3035
+ "epoch": 13.19,
3036
+ "learning_rate": 0.001,
3037
+ "loss": 2.4056,
3038
+ "step": 4050
3039
+ },
3040
+ {
3041
+ "epoch": 13.22,
3042
+ "learning_rate": 0.001,
3043
+ "loss": 2.3324,
3044
+ "step": 4060
3045
+ },
3046
+ {
3047
+ "epoch": 13.25,
3048
+ "learning_rate": 0.001,
3049
+ "loss": 2.3016,
3050
+ "step": 4070
3051
+ },
3052
+ {
3053
+ "epoch": 13.28,
3054
+ "learning_rate": 0.001,
3055
+ "loss": 2.3232,
3056
+ "step": 4080
3057
+ },
3058
+ {
3059
+ "epoch": 13.32,
3060
+ "learning_rate": 0.001,
3061
+ "loss": 2.3512,
3062
+ "step": 4090
3063
+ },
3064
+ {
3065
+ "epoch": 13.35,
3066
+ "learning_rate": 0.001,
3067
+ "loss": 2.2561,
3068
+ "step": 4100
3069
+ },
3070
+ {
3071
+ "epoch": 13.38,
3072
+ "learning_rate": 0.001,
3073
+ "loss": 2.3663,
3074
+ "step": 4110
3075
+ },
3076
+ {
3077
+ "epoch": 13.41,
3078
+ "learning_rate": 0.001,
3079
+ "loss": 2.414,
3080
+ "step": 4120
3081
+ },
3082
+ {
3083
+ "epoch": 13.45,
3084
+ "learning_rate": 0.001,
3085
+ "loss": 2.3334,
3086
+ "step": 4130
3087
+ },
3088
+ {
3089
+ "epoch": 13.48,
3090
+ "learning_rate": 0.001,
3091
+ "loss": 2.324,
3092
+ "step": 4140
3093
+ },
3094
+ {
3095
+ "epoch": 13.51,
3096
+ "learning_rate": 0.001,
3097
+ "loss": 2.4558,
3098
+ "step": 4150
3099
+ },
3100
+ {
3101
+ "epoch": 13.54,
3102
+ "learning_rate": 0.001,
3103
+ "loss": 2.2729,
3104
+ "step": 4160
3105
+ },
3106
+ {
3107
+ "epoch": 13.58,
3108
+ "learning_rate": 0.001,
3109
+ "loss": 2.3651,
3110
+ "step": 4170
3111
+ },
3112
+ {
3113
+ "epoch": 13.61,
3114
+ "learning_rate": 0.001,
3115
+ "loss": 2.4071,
3116
+ "step": 4180
3117
+ },
3118
+ {
3119
+ "epoch": 13.64,
3120
+ "learning_rate": 0.001,
3121
+ "loss": 2.3455,
3122
+ "step": 4190
3123
+ },
3124
+ {
3125
+ "epoch": 13.67,
3126
+ "learning_rate": 0.001,
3127
+ "loss": 2.383,
3128
+ "step": 4200
3129
+ },
3130
+ {
3131
+ "epoch": 13.67,
3132
+ "eval_bp": 1.0,
3133
+ "eval_counts": [
3134
+ 26856,
3135
+ 18659,
3136
+ 13769,
3137
+ 10407
3138
+ ],
3139
+ "eval_loss": 1.422828197479248,
3140
+ "eval_precisions": [
3141
+ 63.941334730125476,
3142
+ 48.41714671233588,
3143
+ 39.242454470316645,
3144
+ 32.85037878787879
3145
+ ],
3146
+ "eval_ref_len": 41379,
3147
+ "eval_runtime": 112.127,
3148
+ "eval_samples_per_second": 30.885,
3149
+ "eval_score": 44.6960720730811,
3150
+ "eval_steps_per_second": 30.885,
3151
+ "eval_sys_len": 42001,
3152
+ "eval_totals": [
3153
+ 42001,
3154
+ 38538,
3155
+ 35087,
3156
+ 31680
3157
+ ],
3158
+ "step": 4200
3159
+ },
3160
+ {
3161
+ "epoch": 13.71,
3162
+ "learning_rate": 0.001,
3163
+ "loss": 2.2856,
3164
+ "step": 4210
3165
+ },
3166
+ {
3167
+ "epoch": 13.74,
3168
+ "learning_rate": 0.001,
3169
+ "loss": 2.33,
3170
+ "step": 4220
3171
+ },
3172
+ {
3173
+ "epoch": 13.77,
3174
+ "learning_rate": 0.001,
3175
+ "loss": 2.3685,
3176
+ "step": 4230
3177
+ },
3178
+ {
3179
+ "epoch": 13.8,
3180
+ "learning_rate": 0.001,
3181
+ "loss": 2.3132,
3182
+ "step": 4240
3183
+ },
3184
+ {
3185
+ "epoch": 13.84,
3186
+ "learning_rate": 0.001,
3187
+ "loss": 2.3598,
3188
+ "step": 4250
3189
+ },
3190
+ {
3191
+ "epoch": 13.87,
3192
+ "learning_rate": 0.001,
3193
+ "loss": 2.3929,
3194
+ "step": 4260
3195
+ },
3196
+ {
3197
+ "epoch": 13.9,
3198
+ "learning_rate": 0.001,
3199
+ "loss": 2.2815,
3200
+ "step": 4270
3201
+ },
3202
+ {
3203
+ "epoch": 13.93,
3204
+ "learning_rate": 0.001,
3205
+ "loss": 2.3766,
3206
+ "step": 4280
3207
+ },
3208
+ {
3209
+ "epoch": 13.97,
3210
+ "learning_rate": 0.001,
3211
+ "loss": 2.4054,
3212
+ "step": 4290
3213
+ },
3214
+ {
3215
+ "epoch": 14.0,
3216
+ "learning_rate": 0.001,
3217
+ "loss": 2.3955,
3218
+ "step": 4300
3219
+ },
3220
+ {
3221
+ "epoch": 14.03,
3222
+ "learning_rate": 0.001,
3223
+ "loss": 2.3044,
3224
+ "step": 4310
3225
+ },
3226
+ {
3227
+ "epoch": 14.06,
3228
+ "learning_rate": 0.001,
3229
+ "loss": 2.2661,
3230
+ "step": 4320
3231
+ },
3232
+ {
3233
+ "epoch": 14.1,
3234
+ "learning_rate": 0.001,
3235
+ "loss": 2.2011,
3236
+ "step": 4330
3237
+ },
3238
+ {
3239
+ "epoch": 14.13,
3240
+ "learning_rate": 0.001,
3241
+ "loss": 2.2519,
3242
+ "step": 4340
3243
+ },
3244
+ {
3245
+ "epoch": 14.16,
3246
+ "learning_rate": 0.001,
3247
+ "loss": 2.2913,
3248
+ "step": 4350
3249
+ },
3250
+ {
3251
+ "epoch": 14.19,
3252
+ "learning_rate": 0.001,
3253
+ "loss": 2.284,
3254
+ "step": 4360
3255
+ },
3256
+ {
3257
+ "epoch": 14.23,
3258
+ "learning_rate": 0.001,
3259
+ "loss": 2.2612,
3260
+ "step": 4370
3261
+ },
3262
+ {
3263
+ "epoch": 14.26,
3264
+ "learning_rate": 0.001,
3265
+ "loss": 2.2675,
3266
+ "step": 4380
3267
+ },
3268
+ {
3269
+ "epoch": 14.29,
3270
+ "learning_rate": 0.001,
3271
+ "loss": 2.2262,
3272
+ "step": 4390
3273
+ },
3274
+ {
3275
+ "epoch": 14.32,
3276
+ "learning_rate": 0.001,
3277
+ "loss": 2.2696,
3278
+ "step": 4400
3279
+ },
3280
+ {
3281
+ "epoch": 14.32,
3282
+ "eval_bp": 0.9583631202356716,
3283
+ "eval_counts": [
3284
+ 27315,
3285
+ 19078,
3286
+ 14092,
3287
+ 10677
3288
+ ],
3289
+ "eval_loss": 1.4258805513381958,
3290
+ "eval_precisions": [
3291
+ 68.81912776196116,
3292
+ 52.660925251186924,
3293
+ 42.99880999603332,
3294
+ 36.36208834247182
3295
+ ],
3296
+ "eval_ref_len": 41379,
3297
+ "eval_runtime": 79.9145,
3298
+ "eval_samples_per_second": 43.334,
3299
+ "eval_score": 46.757988379346315,
3300
+ "eval_steps_per_second": 43.334,
3301
+ "eval_sys_len": 39691,
3302
+ "eval_totals": [
3303
+ 39691,
3304
+ 36228,
3305
+ 32773,
3306
+ 29363
3307
+ ],
3308
+ "step": 4400
3309
+ },
3310
+ {
3311
+ "epoch": 14.36,
3312
+ "learning_rate": 0.001,
3313
+ "loss": 2.3408,
3314
+ "step": 4410
3315
+ },
3316
+ {
3317
+ "epoch": 14.39,
3318
+ "learning_rate": 0.001,
3319
+ "loss": 2.32,
3320
+ "step": 4420
3321
+ },
3322
+ {
3323
+ "epoch": 14.42,
3324
+ "learning_rate": 0.001,
3325
+ "loss": 2.2473,
3326
+ "step": 4430
3327
+ },
3328
+ {
3329
+ "epoch": 14.45,
3330
+ "learning_rate": 0.001,
3331
+ "loss": 2.243,
3332
+ "step": 4440
3333
+ },
3334
+ {
3335
+ "epoch": 14.49,
3336
+ "learning_rate": 0.001,
3337
+ "loss": 2.294,
3338
+ "step": 4450
3339
+ },
3340
+ {
3341
+ "epoch": 14.52,
3342
+ "learning_rate": 0.001,
3343
+ "loss": 2.2595,
3344
+ "step": 4460
3345
+ },
3346
+ {
3347
+ "epoch": 14.55,
3348
+ "learning_rate": 0.001,
3349
+ "loss": 2.2146,
3350
+ "step": 4470
3351
+ },
3352
+ {
3353
+ "epoch": 14.58,
3354
+ "learning_rate": 0.001,
3355
+ "loss": 2.2468,
3356
+ "step": 4480
3357
+ },
3358
+ {
3359
+ "epoch": 14.62,
3360
+ "learning_rate": 0.001,
3361
+ "loss": 2.3313,
3362
+ "step": 4490
3363
+ },
3364
+ {
3365
+ "epoch": 14.65,
3366
+ "learning_rate": 0.001,
3367
+ "loss": 2.2623,
3368
+ "step": 4500
3369
+ },
3370
+ {
3371
+ "epoch": 14.68,
3372
+ "learning_rate": 0.001,
3373
+ "loss": 2.2167,
3374
+ "step": 4510
3375
+ },
3376
+ {
3377
+ "epoch": 14.72,
3378
+ "learning_rate": 0.001,
3379
+ "loss": 2.3084,
3380
+ "step": 4520
3381
+ },
3382
+ {
3383
+ "epoch": 14.75,
3384
+ "learning_rate": 0.001,
3385
+ "loss": 2.2722,
3386
+ "step": 4530
3387
+ },
3388
+ {
3389
+ "epoch": 14.78,
3390
+ "learning_rate": 0.001,
3391
+ "loss": 2.2878,
3392
+ "step": 4540
3393
+ },
3394
+ {
3395
+ "epoch": 14.81,
3396
+ "learning_rate": 0.001,
3397
+ "loss": 2.2647,
3398
+ "step": 4550
3399
+ },
3400
+ {
3401
+ "epoch": 14.85,
3402
+ "learning_rate": 0.001,
3403
+ "loss": 2.35,
3404
+ "step": 4560
3405
+ },
3406
+ {
3407
+ "epoch": 14.88,
3408
+ "learning_rate": 0.001,
3409
+ "loss": 2.312,
3410
+ "step": 4570
3411
+ },
3412
+ {
3413
+ "epoch": 14.91,
3414
+ "learning_rate": 0.001,
3415
+ "loss": 2.3462,
3416
+ "step": 4580
3417
+ },
3418
+ {
3419
+ "epoch": 14.94,
3420
+ "learning_rate": 0.001,
3421
+ "loss": 2.2526,
3422
+ "step": 4590
3423
+ },
3424
+ {
3425
+ "epoch": 14.98,
3426
+ "learning_rate": 0.001,
3427
+ "loss": 2.2648,
3428
+ "step": 4600
3429
+ },
3430
+ {
3431
+ "epoch": 14.98,
3432
+ "eval_bp": 0.96840860582464,
3433
+ "eval_counts": [
3434
+ 27875,
3435
+ 19603,
3436
+ 14568,
3437
+ 11071
3438
+ ],
3439
+ "eval_loss": 1.4148167371749878,
3440
+ "eval_precisions": [
3441
+ 69.52758655093285,
3442
+ 53.5177045510388,
3443
+ 43.91258477769404,
3444
+ 37.19094329481322
3445
+ ],
3446
+ "eval_ref_len": 41379,
3447
+ "eval_runtime": 80.5243,
3448
+ "eval_samples_per_second": 43.006,
3449
+ "eval_score": 48.08159289033059,
3450
+ "eval_steps_per_second": 43.006,
3451
+ "eval_sys_len": 40092,
3452
+ "eval_totals": [
3453
+ 40092,
3454
+ 36629,
3455
+ 33175,
3456
+ 29768
3457
+ ],
3458
+ "step": 4600
3459
+ },
3460
+ {
3461
+ "epoch": 15.01,
3462
+ "learning_rate": 0.001,
3463
+ "loss": 2.2865,
3464
+ "step": 4610
3465
+ },
3466
+ {
3467
+ "epoch": 15.04,
3468
+ "learning_rate": 0.001,
3469
+ "loss": 2.1537,
3470
+ "step": 4620
3471
+ },
3472
+ {
3473
+ "epoch": 15.07,
3474
+ "learning_rate": 0.001,
3475
+ "loss": 2.1461,
3476
+ "step": 4630
3477
+ },
3478
+ {
3479
+ "epoch": 15.11,
3480
+ "learning_rate": 0.001,
3481
+ "loss": 2.2309,
3482
+ "step": 4640
3483
+ },
3484
+ {
3485
+ "epoch": 15.14,
3486
+ "learning_rate": 0.001,
3487
+ "loss": 2.1519,
3488
+ "step": 4650
3489
+ },
3490
+ {
3491
+ "epoch": 15.17,
3492
+ "learning_rate": 0.001,
3493
+ "loss": 2.2306,
3494
+ "step": 4660
3495
+ },
3496
+ {
3497
+ "epoch": 15.2,
3498
+ "learning_rate": 0.001,
3499
+ "loss": 2.1776,
3500
+ "step": 4670
3501
+ },
3502
+ {
3503
+ "epoch": 15.24,
3504
+ "learning_rate": 0.001,
3505
+ "loss": 2.1811,
3506
+ "step": 4680
3507
+ },
3508
+ {
3509
+ "epoch": 15.27,
3510
+ "learning_rate": 0.001,
3511
+ "loss": 2.2001,
3512
+ "step": 4690
3513
+ },
3514
+ {
3515
+ "epoch": 15.3,
3516
+ "learning_rate": 0.001,
3517
+ "loss": 2.1924,
3518
+ "step": 4700
3519
+ },
3520
+ {
3521
+ "epoch": 15.33,
3522
+ "learning_rate": 0.001,
3523
+ "loss": 2.1418,
3524
+ "step": 4710
3525
+ },
3526
+ {
3527
+ "epoch": 15.37,
3528
+ "learning_rate": 0.001,
3529
+ "loss": 2.2936,
3530
+ "step": 4720
3531
+ },
3532
+ {
3533
+ "epoch": 15.4,
3534
+ "learning_rate": 0.001,
3535
+ "loss": 2.2203,
3536
+ "step": 4730
3537
+ },
3538
+ {
3539
+ "epoch": 15.43,
3540
+ "learning_rate": 0.001,
3541
+ "loss": 2.2136,
3542
+ "step": 4740
3543
+ },
3544
+ {
3545
+ "epoch": 15.46,
3546
+ "learning_rate": 0.001,
3547
+ "loss": 2.1555,
3548
+ "step": 4750
3549
+ },
3550
+ {
3551
+ "epoch": 15.5,
3552
+ "learning_rate": 0.001,
3553
+ "loss": 2.3135,
3554
+ "step": 4760
3555
+ },
3556
+ {
3557
+ "epoch": 15.53,
3558
+ "learning_rate": 0.001,
3559
+ "loss": 2.2464,
3560
+ "step": 4770
3561
+ },
3562
+ {
3563
+ "epoch": 15.56,
3564
+ "learning_rate": 0.001,
3565
+ "loss": 2.2005,
3566
+ "step": 4780
3567
+ },
3568
+ {
3569
+ "epoch": 15.59,
3570
+ "learning_rate": 0.001,
3571
+ "loss": 2.1989,
3572
+ "step": 4790
3573
+ },
3574
+ {
3575
+ "epoch": 15.63,
3576
+ "learning_rate": 0.001,
3577
+ "loss": 2.2123,
3578
+ "step": 4800
3579
+ },
3580
+ {
3581
+ "epoch": 15.63,
3582
+ "eval_bp": 1.0,
3583
+ "eval_counts": [
3584
+ 28009,
3585
+ 19746,
3586
+ 14685,
3587
+ 11165
3588
+ ],
3589
+ "eval_loss": 1.432946801185608,
3590
+ "eval_precisions": [
3591
+ 67.3374203630244,
3592
+ 51.78327913563411,
3593
+ 42.3491752220556,
3594
+ 35.71314333237373
3595
+ ],
3596
+ "eval_ref_len": 41379,
3597
+ "eval_runtime": 95.3413,
3598
+ "eval_samples_per_second": 36.322,
3599
+ "eval_score": 47.92142944979905,
3600
+ "eval_steps_per_second": 36.322,
3601
+ "eval_sys_len": 41595,
3602
+ "eval_totals": [
3603
+ 41595,
3604
+ 38132,
3605
+ 34676,
3606
+ 31263
3607
+ ],
3608
+ "step": 4800
3609
+ },
3610
+ {
3611
+ "epoch": 15.66,
3612
+ "learning_rate": 0.001,
3613
+ "loss": 2.2782,
3614
+ "step": 4810
3615
+ },
3616
+ {
3617
+ "epoch": 15.69,
3618
+ "learning_rate": 0.001,
3619
+ "loss": 2.2205,
3620
+ "step": 4820
3621
+ },
3622
+ {
3623
+ "epoch": 15.72,
3624
+ "learning_rate": 0.001,
3625
+ "loss": 2.212,
3626
+ "step": 4830
3627
+ },
3628
+ {
3629
+ "epoch": 15.76,
3630
+ "learning_rate": 0.001,
3631
+ "loss": 2.2153,
3632
+ "step": 4840
3633
+ },
3634
+ {
3635
+ "epoch": 15.79,
3636
+ "learning_rate": 0.001,
3637
+ "loss": 2.2444,
3638
+ "step": 4850
3639
+ },
3640
+ {
3641
+ "epoch": 15.82,
3642
+ "learning_rate": 0.001,
3643
+ "loss": 2.285,
3644
+ "step": 4860
3645
+ },
3646
+ {
3647
+ "epoch": 15.85,
3648
+ "learning_rate": 0.001,
3649
+ "loss": 2.2001,
3650
+ "step": 4870
3651
+ },
3652
+ {
3653
+ "epoch": 15.89,
3654
+ "learning_rate": 0.001,
3655
+ "loss": 2.3122,
3656
+ "step": 4880
3657
+ },
3658
+ {
3659
+ "epoch": 15.92,
3660
+ "learning_rate": 0.001,
3661
+ "loss": 2.2141,
3662
+ "step": 4890
3663
+ },
3664
+ {
3665
+ "epoch": 15.95,
3666
+ "learning_rate": 0.001,
3667
+ "loss": 2.1724,
3668
+ "step": 4900
3669
+ },
3670
+ {
3671
+ "epoch": 15.98,
3672
+ "learning_rate": 0.001,
3673
+ "loss": 2.2171,
3674
+ "step": 4910
3675
+ },
3676
+ {
3677
+ "epoch": 16.02,
3678
+ "learning_rate": 0.001,
3679
+ "loss": 2.1489,
3680
+ "step": 4920
3681
+ },
3682
+ {
3683
+ "epoch": 16.05,
3684
+ "learning_rate": 0.001,
3685
+ "loss": 2.1308,
3686
+ "step": 4930
3687
+ },
3688
+ {
3689
+ "epoch": 16.08,
3690
+ "learning_rate": 0.001,
3691
+ "loss": 2.1169,
3692
+ "step": 4940
3693
+ },
3694
+ {
3695
+ "epoch": 16.12,
3696
+ "learning_rate": 0.001,
3697
+ "loss": 2.0744,
3698
+ "step": 4950
3699
+ },
3700
+ {
3701
+ "epoch": 16.15,
3702
+ "learning_rate": 0.001,
3703
+ "loss": 2.1527,
3704
+ "step": 4960
3705
+ },
3706
+ {
3707
+ "epoch": 16.18,
3708
+ "learning_rate": 0.001,
3709
+ "loss": 2.1514,
3710
+ "step": 4970
3711
+ },
3712
+ {
3713
+ "epoch": 16.21,
3714
+ "learning_rate": 0.001,
3715
+ "loss": 2.2027,
3716
+ "step": 4980
3717
+ },
3718
+ {
3719
+ "epoch": 16.25,
3720
+ "learning_rate": 0.001,
3721
+ "loss": 2.1814,
3722
+ "step": 4990
3723
+ },
3724
+ {
3725
+ "epoch": 16.28,
3726
+ "learning_rate": 0.001,
3727
+ "loss": 2.1292,
3728
+ "step": 5000
3729
+ },
3730
+ {
3731
+ "epoch": 16.28,
3732
+ "eval_bp": 0.9822754617987439,
3733
+ "eval_counts": [
3734
+ 27961,
3735
+ 19662,
3736
+ 14591,
3737
+ 11071
3738
+ ],
3739
+ "eval_loss": 1.4380427598953247,
3740
+ "eval_precisions": [
3741
+ 68.7813637705402,
3742
+ 52.87047245153137,
3743
+ 43.253097764866304,
3744
+ 36.513852242744065
3745
+ ],
3746
+ "eval_ref_len": 41379,
3747
+ "eval_runtime": 88.9695,
3748
+ "eval_samples_per_second": 38.923,
3749
+ "eval_score": 48.08648210020106,
3750
+ "eval_steps_per_second": 38.923,
3751
+ "eval_sys_len": 40652,
3752
+ "eval_totals": [
3753
+ 40652,
3754
+ 37189,
3755
+ 33734,
3756
+ 30320
3757
+ ],
3758
+ "step": 5000
3759
+ },
3760
+ {
3761
+ "epoch": 16.31,
3762
+ "learning_rate": 0.001,
3763
+ "loss": 2.185,
3764
+ "step": 5010
3765
+ },
3766
+ {
3767
+ "epoch": 16.34,
3768
+ "learning_rate": 0.001,
3769
+ "loss": 2.1878,
3770
+ "step": 5020
3771
+ },
3772
+ {
3773
+ "epoch": 16.38,
3774
+ "learning_rate": 0.001,
3775
+ "loss": 2.1767,
3776
+ "step": 5030
3777
+ },
3778
+ {
3779
+ "epoch": 16.41,
3780
+ "learning_rate": 0.001,
3781
+ "loss": 2.011,
3782
+ "step": 5040
3783
+ },
3784
+ {
3785
+ "epoch": 16.44,
3786
+ "learning_rate": 0.001,
3787
+ "loss": 2.0795,
3788
+ "step": 5050
3789
+ },
3790
+ {
3791
+ "epoch": 16.47,
3792
+ "learning_rate": 0.001,
3793
+ "loss": 2.2293,
3794
+ "step": 5060
3795
+ },
3796
+ {
3797
+ "epoch": 16.51,
3798
+ "learning_rate": 0.001,
3799
+ "loss": 2.1918,
3800
+ "step": 5070
3801
+ },
3802
+ {
3803
+ "epoch": 16.54,
3804
+ "learning_rate": 0.001,
3805
+ "loss": 2.1273,
3806
+ "step": 5080
3807
+ },
3808
+ {
3809
+ "epoch": 16.57,
3810
+ "learning_rate": 0.001,
3811
+ "loss": 2.1646,
3812
+ "step": 5090
3813
+ },
3814
+ {
3815
+ "epoch": 16.6,
3816
+ "learning_rate": 0.001,
3817
+ "loss": 2.1413,
3818
+ "step": 5100
3819
+ },
3820
+ {
3821
+ "epoch": 16.64,
3822
+ "learning_rate": 0.001,
3823
+ "loss": 2.0892,
3824
+ "step": 5110
3825
+ },
3826
+ {
3827
+ "epoch": 16.67,
3828
+ "learning_rate": 0.001,
3829
+ "loss": 2.1659,
3830
+ "step": 5120
3831
+ },
3832
+ {
3833
+ "epoch": 16.7,
3834
+ "learning_rate": 0.001,
3835
+ "loss": 2.1922,
3836
+ "step": 5130
3837
+ },
3838
+ {
3839
+ "epoch": 16.73,
3840
+ "learning_rate": 0.001,
3841
+ "loss": 2.2153,
3842
+ "step": 5140
3843
+ },
3844
+ {
3845
+ "epoch": 16.77,
3846
+ "learning_rate": 0.001,
3847
+ "loss": 2.1297,
3848
+ "step": 5150
3849
+ },
3850
+ {
3851
+ "epoch": 16.8,
3852
+ "learning_rate": 0.001,
3853
+ "loss": 2.2311,
3854
+ "step": 5160
3855
+ },
3856
+ {
3857
+ "epoch": 16.83,
3858
+ "learning_rate": 0.001,
3859
+ "loss": 2.1413,
3860
+ "step": 5170
3861
+ },
3862
+ {
3863
+ "epoch": 16.86,
3864
+ "learning_rate": 0.001,
3865
+ "loss": 2.1689,
3866
+ "step": 5180
3867
+ },
3868
+ {
3869
+ "epoch": 16.9,
3870
+ "learning_rate": 0.001,
3871
+ "loss": 2.1698,
3872
+ "step": 5190
3873
+ },
3874
+ {
3875
+ "epoch": 16.93,
3876
+ "learning_rate": 0.001,
3877
+ "loss": 2.0922,
3878
+ "step": 5200
3879
+ },
3880
+ {
3881
+ "epoch": 16.93,
3882
+ "eval_bp": 0.9634606403593156,
3883
+ "eval_counts": [
3884
+ 27055,
3885
+ 18947,
3886
+ 14032,
3887
+ 10647
3888
+ ],
3889
+ "eval_loss": 1.420142650604248,
3890
+ "eval_precisions": [
3891
+ 67.8172156213967,
3892
+ 52.00790535532925,
3893
+ 42.54699818071558,
3894
+ 36.002434653230985
3895
+ ],
3896
+ "eval_ref_len": 41379,
3897
+ "eval_runtime": 88.5384,
3898
+ "eval_samples_per_second": 39.113,
3899
+ "eval_score": 46.450084211438956,
3900
+ "eval_steps_per_second": 39.113,
3901
+ "eval_sys_len": 39894,
3902
+ "eval_totals": [
3903
+ 39894,
3904
+ 36431,
3905
+ 32980,
3906
+ 29573
3907
+ ],
3908
+ "step": 5200
3909
+ },
3910
+ {
3911
+ "epoch": 16.96,
3912
+ "learning_rate": 0.001,
3913
+ "loss": 2.212,
3914
+ "step": 5210
3915
+ },
3916
+ {
3917
+ "epoch": 16.99,
3918
+ "learning_rate": 0.001,
3919
+ "loss": 2.1306,
3920
+ "step": 5220
3921
+ },
3922
+ {
3923
+ "epoch": 17.03,
3924
+ "learning_rate": 0.001,
3925
+ "loss": 2.0919,
3926
+ "step": 5230
3927
+ },
3928
+ {
3929
+ "epoch": 17.06,
3930
+ "learning_rate": 0.001,
3931
+ "loss": 1.977,
3932
+ "step": 5240
3933
+ },
3934
+ {
3935
+ "epoch": 17.09,
3936
+ "learning_rate": 0.001,
3937
+ "loss": 2.0461,
3938
+ "step": 5250
3939
+ },
3940
+ {
3941
+ "epoch": 17.12,
3942
+ "learning_rate": 0.001,
3943
+ "loss": 2.1332,
3944
+ "step": 5260
3945
+ },
3946
+ {
3947
+ "epoch": 17.16,
3948
+ "learning_rate": 0.001,
3949
+ "loss": 2.116,
3950
+ "step": 5270
3951
+ },
3952
+ {
3953
+ "epoch": 17.19,
3954
+ "learning_rate": 0.001,
3955
+ "loss": 2.0639,
3956
+ "step": 5280
3957
+ },
3958
+ {
3959
+ "epoch": 17.22,
3960
+ "learning_rate": 0.001,
3961
+ "loss": 2.0844,
3962
+ "step": 5290
3963
+ },
3964
+ {
3965
+ "epoch": 17.25,
3966
+ "learning_rate": 0.001,
3967
+ "loss": 2.0423,
3968
+ "step": 5300
3969
+ },
3970
+ {
3971
+ "epoch": 17.29,
3972
+ "learning_rate": 0.001,
3973
+ "loss": 2.0831,
3974
+ "step": 5310
3975
+ },
3976
+ {
3977
+ "epoch": 17.32,
3978
+ "learning_rate": 0.001,
3979
+ "loss": 2.0964,
3980
+ "step": 5320
3981
+ },
3982
+ {
3983
+ "epoch": 17.35,
3984
+ "learning_rate": 0.001,
3985
+ "loss": 2.0526,
3986
+ "step": 5330
3987
+ },
3988
+ {
3989
+ "epoch": 17.38,
3990
+ "learning_rate": 0.001,
3991
+ "loss": 2.0584,
3992
+ "step": 5340
3993
+ },
3994
+ {
3995
+ "epoch": 17.42,
3996
+ "learning_rate": 0.001,
3997
+ "loss": 2.1096,
3998
+ "step": 5350
3999
+ },
4000
+ {
4001
+ "epoch": 17.45,
4002
+ "learning_rate": 0.001,
4003
+ "loss": 2.0576,
4004
+ "step": 5360
4005
+ },
4006
+ {
4007
+ "epoch": 17.48,
4008
+ "learning_rate": 0.001,
4009
+ "loss": 2.0952,
4010
+ "step": 5370
4011
+ },
4012
+ {
4013
+ "epoch": 17.51,
4014
+ "learning_rate": 0.001,
4015
+ "loss": 2.0561,
4016
+ "step": 5380
4017
+ },
4018
+ {
4019
+ "epoch": 17.55,
4020
+ "learning_rate": 0.001,
4021
+ "loss": 2.1358,
4022
+ "step": 5390
4023
+ },
4024
+ {
4025
+ "epoch": 17.58,
4026
+ "learning_rate": 0.001,
4027
+ "loss": 2.066,
4028
+ "step": 5400
4029
+ },
4030
+ {
4031
+ "epoch": 17.58,
4032
+ "eval_bp": 1.0,
4033
+ "eval_counts": [
4034
+ 27996,
4035
+ 19728,
4036
+ 14711,
4037
+ 11223
4038
+ ],
4039
+ "eval_loss": 1.4336364269256592,
4040
+ "eval_precisions": [
4041
+ 66.97607655502392,
4042
+ 51.45942562015807,
4043
+ 42.17240489636786,
4044
+ 35.660269445856635
4045
+ ],
4046
+ "eval_ref_len": 41379,
4047
+ "eval_runtime": 98.8235,
4048
+ "eval_samples_per_second": 35.042,
4049
+ "eval_score": 47.71439333194693,
4050
+ "eval_steps_per_second": 35.042,
4051
+ "eval_sys_len": 41800,
4052
+ "eval_totals": [
4053
+ 41800,
4054
+ 38337,
4055
+ 34883,
4056
+ 31472
4057
+ ],
4058
+ "step": 5400
4059
+ },
4060
+ {
4061
+ "epoch": 17.61,
4062
+ "learning_rate": 0.001,
4063
+ "loss": 2.1324,
4064
+ "step": 5410
4065
+ },
4066
+ {
4067
+ "epoch": 17.65,
4068
+ "learning_rate": 0.001,
4069
+ "loss": 2.1562,
4070
+ "step": 5420
4071
+ },
4072
+ {
4073
+ "epoch": 17.68,
4074
+ "learning_rate": 0.001,
4075
+ "loss": 2.064,
4076
+ "step": 5430
4077
+ },
4078
+ {
4079
+ "epoch": 17.71,
4080
+ "learning_rate": 0.001,
4081
+ "loss": 2.1753,
4082
+ "step": 5440
4083
+ },
4084
+ {
4085
+ "epoch": 17.74,
4086
+ "learning_rate": 0.001,
4087
+ "loss": 2.1545,
4088
+ "step": 5450
4089
+ },
4090
+ {
4091
+ "epoch": 17.78,
4092
+ "learning_rate": 0.001,
4093
+ "loss": 2.1808,
4094
+ "step": 5460
4095
+ },
4096
+ {
4097
+ "epoch": 17.81,
4098
+ "learning_rate": 0.001,
4099
+ "loss": 1.9854,
4100
+ "step": 5470
4101
+ },
4102
+ {
4103
+ "epoch": 17.84,
4104
+ "learning_rate": 0.001,
4105
+ "loss": 2.1477,
4106
+ "step": 5480
4107
+ },
4108
+ {
4109
+ "epoch": 17.87,
4110
+ "learning_rate": 0.001,
4111
+ "loss": 2.147,
4112
+ "step": 5490
4113
+ },
4114
+ {
4115
+ "epoch": 17.91,
4116
+ "learning_rate": 0.001,
4117
+ "loss": 2.1071,
4118
+ "step": 5500
4119
+ },
4120
+ {
4121
+ "epoch": 17.94,
4122
+ "learning_rate": 0.001,
4123
+ "loss": 2.1168,
4124
+ "step": 5510
4125
+ },
4126
+ {
4127
+ "epoch": 17.97,
4128
+ "learning_rate": 0.001,
4129
+ "loss": 2.0857,
4130
+ "step": 5520
4131
+ },
4132
+ {
4133
+ "epoch": 18.0,
4134
+ "learning_rate": 0.001,
4135
+ "loss": 2.1181,
4136
+ "step": 5530
4137
+ },
4138
+ {
4139
+ "epoch": 18.04,
4140
+ "learning_rate": 0.001,
4141
+ "loss": 2.0222,
4142
+ "step": 5540
4143
+ },
4144
+ {
4145
+ "epoch": 18.07,
4146
+ "learning_rate": 0.001,
4147
+ "loss": 1.9394,
4148
+ "step": 5550
4149
+ },
4150
+ {
4151
+ "epoch": 18.1,
4152
+ "learning_rate": 0.001,
4153
+ "loss": 2.0617,
4154
+ "step": 5560
4155
+ },
4156
+ {
4157
+ "epoch": 18.13,
4158
+ "learning_rate": 0.001,
4159
+ "loss": 1.9781,
4160
+ "step": 5570
4161
+ },
4162
+ {
4163
+ "epoch": 18.17,
4164
+ "learning_rate": 0.001,
4165
+ "loss": 1.9584,
4166
+ "step": 5580
4167
+ },
4168
+ {
4169
+ "epoch": 18.2,
4170
+ "learning_rate": 0.001,
4171
+ "loss": 1.9962,
4172
+ "step": 5590
4173
+ },
4174
+ {
4175
+ "epoch": 18.23,
4176
+ "learning_rate": 0.001,
4177
+ "loss": 1.9179,
4178
+ "step": 5600
4179
+ },
4180
+ {
4181
+ "epoch": 18.23,
4182
+ "eval_bp": 0.9862765874121399,
4183
+ "eval_counts": [
4184
+ 28447,
4185
+ 20181,
4186
+ 15105,
4187
+ 11544
4188
+ ],
4189
+ "eval_loss": 1.4454234838485718,
4190
+ "eval_precisions": [
4191
+ 69.6974151659929,
4192
+ 54.02923538230885,
4193
+ 44.560151041359376,
4194
+ 37.866561700452664
4195
+ ],
4196
+ "eval_ref_len": 41379,
4197
+ "eval_runtime": 80.4217,
4198
+ "eval_samples_per_second": 43.06,
4199
+ "eval_score": 49.51773016565914,
4200
+ "eval_steps_per_second": 43.06,
4201
+ "eval_sys_len": 40815,
4202
+ "eval_totals": [
4203
+ 40815,
4204
+ 37352,
4205
+ 33898,
4206
+ 30486
4207
+ ],
4208
+ "step": 5600
4209
+ }
4210
+ ],
4211
+ "max_steps": 76750,
4212
+ "num_train_epochs": 250,
4213
+ "total_flos": 5.39878684251648e+16,
4214
+ "trial_name": null,
4215
+ "trial_params": null
4216
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db455e9de0e5656e19b335317e6684f522ce7643f4689ab5bb2bae7f281326b4
3
+ size 3771