Hugging Face's logo

cs182project
/
llama-2-7b-chat-lyre-lm

Text Generation
Transformers
PyTorch
llama
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
evanfrick commited on
Commit
802a40f
·
1 Parent(s): 09ff306

added model

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. added_tokens.json +5 -0
  3. config.json +27 -0
  4. generation_config.json +10 -0
  5. output-llama-2-7b-chat-finetuned_lyre/.gitattributes +0 -3
  6. pytorch_model.bin +3 -0
  7. special_tokens_map.json +6 -0
  8. tokenizer.model +3 -0
  9. tokenizer_config.json +45 -0
  10. trainer_state.json +3058 -0
  11. training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
37
+ training_args.bin filter=lfs diff=lfs merge=lfs -text
38
+ tokenizer.model filter=lfs diff=lfs merge=lfs -text
added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "</s>": 2,
3
+ "<s>": 1,
4
+ "<unk>": 0
5
+ }
config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "meta-llama/Llama-2-7b-chat-hf",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 4096,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 11008,
13
+ "max_position_embeddings": 4096,
14
+ "model_type": "llama",
15
+ "num_attention_heads": 32,
16
+ "num_hidden_layers": 32,
17
+ "num_key_value_heads": 32,
18
+ "pretraining_tp": 1,
19
+ "rms_norm_eps": 1e-05,
20
+ "rope_scaling": null,
21
+ "rope_theta": 10000.0,
22
+ "tie_word_embeddings": false,
23
+ "torch_dtype": "float32",
24
+ "transformers_version": "4.34.0",
25
+ "use_cache": true,
26
+ "vocab_size": 32000
27
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 1,
3
+ "do_sample": true,
4
+ "eos_token_id": 2,
5
+ "max_length": 4096,
6
+ "pad_token_id": 0,
7
+ "temperature": 0.6,
8
+ "top_p": 0.9,
9
+ "transformers_version": "4.34.0"
10
+ }
output-llama-2-7b-chat-finetuned_lyre/.gitattributes DELETED
@@ -1,3 +0,0 @@
1
- pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
2
- training_args.bin filter=lfs diff=lfs merge=lfs -text
3
- tokenizer.model filter=lfs diff=lfs merge=lfs -text
 
 
 
 
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca64e07fdda0b0cc45903bffdb7fb9a7a6b6d03cd849c60067326520d6c78ce6
3
+ size 26953810686
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "<unk>",
5
+ "unk_token": "<unk>"
6
+ }
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
tokenizer_config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": false
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": false
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": false
28
+ }
29
+ },
30
+ "additional_special_tokens": [],
31
+ "bos_token": "<s>",
32
+ "chat_template": "{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% set system_message = false %}{% endif %}{% for message in loop_messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if loop.index0 == 0 and system_message != false %}{% set content = '<<SYS>>\\n' + system_message + '\\n<</SYS>>\\n\\n' + message['content'] %}{% else %}{% set content = message['content'] %}{% endif %}{% if message['role'] == 'user' %}{{ bos_token + '[INST] ' + content.strip() + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ ' ' + content.strip() + ' ' + eos_token }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": false,
36
+ "model_max_length": 2048,
37
+ "pad_token": "<unk>",
38
+ "padding_side": "right",
39
+ "sp_model_kwargs": {},
40
+ "spaces_between_special_tokens": false,
41
+ "tokenizer_class": "LlamaTokenizer",
42
+ "tokenizer_file": "/data/evan/.cache/huggingface/hub/models--meta-llama--Llama-2-7b-chat-hf/snapshots/c1b0db933684edbfe29a06fa47eb19cc48025e93/tokenizer.json",
43
+ "unk_token": "<unk>",
44
+ "use_default_system_prompt": true
45
+ }
trainer_state.json ADDED
@@ -0,0 +1,3058 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 4.996907854050711,
5
+ "eval_steps": 500,
6
+ "global_step": 505,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "learning_rate": 1.25e-06,
14
+ "loss": 1.9805,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.02,
19
+ "learning_rate": 2.5e-06,
20
+ "loss": 1.9742,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.03,
25
+ "learning_rate": 3.7500000000000005e-06,
26
+ "loss": 1.9398,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.04,
31
+ "learning_rate": 5e-06,
32
+ "loss": 1.8268,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.05,
37
+ "learning_rate": 6.25e-06,
38
+ "loss": 1.9396,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.06,
43
+ "learning_rate": 7.500000000000001e-06,
44
+ "loss": 1.8366,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.07,
49
+ "learning_rate": 8.750000000000001e-06,
50
+ "loss": 1.7488,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.08,
55
+ "learning_rate": 1e-05,
56
+ "loss": 1.8458,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.09,
61
+ "learning_rate": 1.125e-05,
62
+ "loss": 1.818,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.1,
67
+ "learning_rate": 1.25e-05,
68
+ "loss": 1.7825,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.11,
73
+ "learning_rate": 1.375e-05,
74
+ "loss": 1.8167,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.12,
79
+ "learning_rate": 1.5000000000000002e-05,
80
+ "loss": 1.6685,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.13,
85
+ "learning_rate": 1.6250000000000002e-05,
86
+ "loss": 1.6598,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.14,
91
+ "learning_rate": 1.7500000000000002e-05,
92
+ "loss": 1.6377,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.15,
97
+ "learning_rate": 1.8750000000000002e-05,
98
+ "loss": 1.7032,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.16,
103
+ "learning_rate": 2e-05,
104
+ "loss": 1.7164,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.17,
109
+ "learning_rate": 1.9999793628111833e-05,
110
+ "loss": 1.6622,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.18,
115
+ "learning_rate": 1.9999174520965194e-05,
116
+ "loss": 1.6692,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.19,
121
+ "learning_rate": 1.999814270411335e-05,
122
+ "loss": 1.6883,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.2,
127
+ "learning_rate": 1.99966982201439e-05,
128
+ "loss": 1.698,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.21,
133
+ "learning_rate": 1.999484112867702e-05,
134
+ "loss": 1.7371,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 0.22,
139
+ "learning_rate": 1.9992571506362997e-05,
140
+ "loss": 1.6785,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 0.23,
145
+ "learning_rate": 1.9989889446879092e-05,
146
+ "loss": 1.6484,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 0.24,
151
+ "learning_rate": 1.9986795060925636e-05,
152
+ "loss": 1.6336,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 0.25,
157
+ "learning_rate": 1.9983288476221482e-05,
158
+ "loss": 1.5657,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 0.26,
163
+ "learning_rate": 1.997936983749873e-05,
164
+ "loss": 1.6443,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 0.27,
169
+ "learning_rate": 1.997503930649676e-05,
170
+ "loss": 1.6247,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 0.28,
175
+ "learning_rate": 1.9970297061955533e-05,
176
+ "loss": 1.5329,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 0.29,
181
+ "learning_rate": 1.9965143299608253e-05,
182
+ "loss": 1.6595,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 0.3,
187
+ "learning_rate": 1.995957823217325e-05,
188
+ "loss": 1.4952,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 0.31,
193
+ "learning_rate": 1.9953602089345215e-05,
194
+ "loss": 1.5722,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 0.32,
199
+ "learning_rate": 1.9947215117785727e-05,
200
+ "loss": 1.5964,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 0.33,
205
+ "learning_rate": 1.9940417581113062e-05,
206
+ "loss": 1.6671,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 0.34,
211
+ "learning_rate": 1.9933209759891318e-05,
212
+ "loss": 1.6731,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 0.35,
217
+ "learning_rate": 1.9925591951618822e-05,
218
+ "loss": 1.5952,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 0.36,
223
+ "learning_rate": 1.9917564470715876e-05,
224
+ "loss": 1.6384,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 0.37,
229
+ "learning_rate": 1.9909127648511758e-05,
230
+ "loss": 1.5503,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 0.38,
235
+ "learning_rate": 1.990028183323105e-05,
236
+ "loss": 1.731,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 0.39,
241
+ "learning_rate": 1.989102738997928e-05,
242
+ "loss": 1.68,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 0.4,
247
+ "learning_rate": 1.9881364700727827e-05,
248
+ "loss": 1.6194,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 0.41,
253
+ "learning_rate": 1.9871294164298175e-05,
254
+ "loss": 1.6532,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 0.42,
259
+ "learning_rate": 1.986081619634545e-05,
260
+ "loss": 1.7531,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 0.43,
265
+ "learning_rate": 1.9849931229341258e-05,
266
+ "loss": 1.6211,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 0.44,
271
+ "learning_rate": 1.9838639712555842e-05,
272
+ "loss": 1.6807,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 0.45,
277
+ "learning_rate": 1.982694211203952e-05,
278
+ "loss": 1.7175,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 0.46,
283
+ "learning_rate": 1.981483891060348e-05,
284
+ "loss": 1.6584,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 0.47,
289
+ "learning_rate": 1.9802330607799832e-05,
290
+ "loss": 1.5696,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 0.47,
295
+ "learning_rate": 1.978941771990098e-05,
296
+ "loss": 1.6854,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 0.48,
301
+ "learning_rate": 1.9776100779878344e-05,
302
+ "loss": 1.5761,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 0.49,
307
+ "learning_rate": 1.976238033738033e-05,
308
+ "loss": 1.6526,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 0.5,
313
+ "learning_rate": 1.9748256958709666e-05,
314
+ "loss": 1.5919,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 0.51,
319
+ "learning_rate": 1.9733731226800016e-05,
320
+ "loss": 1.5532,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 0.52,
325
+ "learning_rate": 1.9718803741191918e-05,
326
+ "loss": 1.6093,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 0.53,
331
+ "learning_rate": 1.970347511800806e-05,
332
+ "loss": 1.6669,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 0.54,
337
+ "learning_rate": 1.9687745989927823e-05,
338
+ "loss": 1.594,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 0.55,
343
+ "learning_rate": 1.967161700616117e-05,
344
+ "loss": 1.5944,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 0.56,
349
+ "learning_rate": 1.965508883242188e-05,
350
+ "loss": 1.6027,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 0.57,
355
+ "learning_rate": 1.9638162150900028e-05,
356
+ "loss": 1.5561,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 0.58,
361
+ "learning_rate": 1.9620837660233866e-05,
362
+ "loss": 1.6239,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 0.59,
367
+ "learning_rate": 1.960311607548096e-05,
368
+ "loss": 1.6061,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 0.6,
373
+ "learning_rate": 1.9584998128088686e-05,
374
+ "loss": 1.5782,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 0.61,
379
+ "learning_rate": 1.9566484565864056e-05,
380
+ "loss": 1.7248,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 0.62,
385
+ "learning_rate": 1.954757615294283e-05,
386
+ "loss": 1.5613,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 0.63,
391
+ "learning_rate": 1.9528273669757974e-05,
392
+ "loss": 1.7137,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 0.64,
397
+ "learning_rate": 1.9508577913007475e-05,
398
+ "loss": 1.5516,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 0.65,
403
+ "learning_rate": 1.9488489695621432e-05,
404
+ "loss": 1.5342,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 0.66,
409
+ "learning_rate": 1.9468009846728515e-05,
410
+ "loss": 1.519,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 0.67,
415
+ "learning_rate": 1.944713921162174e-05,
416
+ "loss": 1.5515,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 0.68,
421
+ "learning_rate": 1.942587865172359e-05,
422
+ "loss": 1.5703,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 0.69,
427
+ "learning_rate": 1.9404229044550432e-05,
428
+ "loss": 1.5526,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 0.7,
433
+ "learning_rate": 1.9382191283676336e-05,
434
+ "loss": 1.5859,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 0.71,
439
+ "learning_rate": 1.9359766278696165e-05,
440
+ "loss": 1.5213,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 0.72,
445
+ "learning_rate": 1.9336954955188042e-05,
446
+ "loss": 1.6142,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 0.73,
451
+ "learning_rate": 1.9313758254675143e-05,
452
+ "loss": 1.5944,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 0.74,
457
+ "learning_rate": 1.929017713458685e-05,
458
+ "loss": 1.4959,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 0.75,
463
+ "learning_rate": 1.9266212568219223e-05,
464
+ "loss": 1.494,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 0.76,
469
+ "learning_rate": 1.9241865544694817e-05,
470
+ "loss": 1.5576,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 0.77,
475
+ "learning_rate": 1.9217137068921875e-05,
476
+ "loss": 1.576,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 0.78,
481
+ "learning_rate": 1.9192028161552848e-05,
482
+ "loss": 1.6083,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 0.79,
487
+ "learning_rate": 1.9166539858942258e-05,
488
+ "loss": 1.5894,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 0.8,
493
+ "learning_rate": 1.9140673213103932e-05,
494
+ "loss": 1.678,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 0.81,
499
+ "learning_rate": 1.9114429291667583e-05,
500
+ "loss": 1.5729,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 0.82,
505
+ "learning_rate": 1.908780917783473e-05,
506
+ "loss": 1.6561,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 0.83,
511
+ "learning_rate": 1.906081397033401e-05,
512
+ "loss": 1.618,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 0.84,
517
+ "learning_rate": 1.9033444783375806e-05,
518
+ "loss": 1.5488,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 0.85,
523
+ "learning_rate": 1.9005702746606274e-05,
524
+ "loss": 1.5514,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 0.86,
529
+ "learning_rate": 1.8977589005060723e-05,
530
+ "loss": 1.47,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 0.87,
535
+ "learning_rate": 1.8949104719116334e-05,
536
+ "loss": 1.5237,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 0.88,
541
+ "learning_rate": 1.8920251064444284e-05,
542
+ "loss": 1.5974,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 0.89,
547
+ "learning_rate": 1.8891029231961208e-05,
548
+ "loss": 1.5083,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 0.9,
553
+ "learning_rate": 1.886144042778006e-05,
554
+ "loss": 1.6104,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 0.91,
559
+ "learning_rate": 1.8831485873160312e-05,
560
+ "loss": 1.5195,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 0.92,
565
+ "learning_rate": 1.880116680445757e-05,
566
+ "loss": 1.619,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 0.93,
571
+ "learning_rate": 1.8770484473072518e-05,
572
+ "loss": 1.5151,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 0.94,
577
+ "learning_rate": 1.8739440145399295e-05,
578
+ "loss": 1.6652,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 0.95,
583
+ "learning_rate": 1.8708035102773198e-05,
584
+ "loss": 1.5574,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 0.96,
589
+ "learning_rate": 1.8676270641417824e-05,
590
+ "loss": 1.6027,
591
+ "step": 97
592
+ },
593
+ {
594
+ "epoch": 0.97,
595
+ "learning_rate": 1.864414807239154e-05,
596
+ "loss": 1.6076,
597
+ "step": 98
598
+ },
599
+ {
600
+ "epoch": 0.98,
601
+ "learning_rate": 1.861166872153339e-05,
602
+ "loss": 1.65,
603
+ "step": 99
604
+ },
605
+ {
606
+ "epoch": 0.99,
607
+ "learning_rate": 1.857883392940837e-05,
608
+ "loss": 1.5535,
609
+ "step": 100
610
+ },
611
+ {
612
+ "epoch": 1.0,
613
+ "learning_rate": 1.8545645051252094e-05,
614
+ "loss": 1.4654,
615
+ "step": 101
616
+ },
617
+ {
618
+ "epoch": 1.01,
619
+ "learning_rate": 1.851210345691484e-05,
620
+ "loss": 1.6053,
621
+ "step": 102
622
+ },
623
+ {
624
+ "epoch": 1.02,
625
+ "learning_rate": 1.847821053080505e-05,
626
+ "loss": 1.5897,
627
+ "step": 103
628
+ },
629
+ {
630
+ "epoch": 1.03,
631
+ "learning_rate": 1.844396767183215e-05,
632
+ "loss": 1.5679,
633
+ "step": 104
634
+ },
635
+ {
636
+ "epoch": 1.04,
637
+ "learning_rate": 1.8409376293348836e-05,
638
+ "loss": 1.4711,
639
+ "step": 105
640
+ },
641
+ {
642
+ "epoch": 1.05,
643
+ "learning_rate": 1.8374437823092726e-05,
644
+ "loss": 1.6241,
645
+ "step": 106
646
+ },
647
+ {
648
+ "epoch": 1.06,
649
+ "learning_rate": 1.833915370312743e-05,
650
+ "loss": 1.5392,
651
+ "step": 107
652
+ },
653
+ {
654
+ "epoch": 1.07,
655
+ "learning_rate": 1.8303525389783045e-05,
656
+ "loss": 1.456,
657
+ "step": 108
658
+ },
659
+ {
660
+ "epoch": 1.08,
661
+ "learning_rate": 1.8267554353596027e-05,
662
+ "loss": 1.5794,
663
+ "step": 109
664
+ },
665
+ {
666
+ "epoch": 1.09,
667
+ "learning_rate": 1.8231242079248512e-05,
668
+ "loss": 1.5491,
669
+ "step": 110
670
+ },
671
+ {
672
+ "epoch": 1.1,
673
+ "learning_rate": 1.819459006550702e-05,
674
+ "loss": 1.5011,
675
+ "step": 111
676
+ },
677
+ {
678
+ "epoch": 1.11,
679
+ "learning_rate": 1.815759982516061e-05,
680
+ "loss": 1.5362,
681
+ "step": 112
682
+ },
683
+ {
684
+ "epoch": 1.12,
685
+ "learning_rate": 1.812027288495843e-05,
686
+ "loss": 1.363,
687
+ "step": 113
688
+ },
689
+ {
690
+ "epoch": 1.13,
691
+ "learning_rate": 1.808261078554671e-05,
692
+ "loss": 1.3702,
693
+ "step": 114
694
+ },
695
+ {
696
+ "epoch": 1.14,
697
+ "learning_rate": 1.8044615081405153e-05,
698
+ "loss": 1.3961,
699
+ "step": 115
700
+ },
701
+ {
702
+ "epoch": 1.15,
703
+ "learning_rate": 1.8006287340782807e-05,
704
+ "loss": 1.4011,
705
+ "step": 116
706
+ },
707
+ {
708
+ "epoch": 1.16,
709
+ "learning_rate": 1.7967629145633312e-05,
710
+ "loss": 1.5261,
711
+ "step": 117
712
+ },
713
+ {
714
+ "epoch": 1.17,
715
+ "learning_rate": 1.7928642091549616e-05,
716
+ "loss": 1.3632,
717
+ "step": 118
718
+ },
719
+ {
720
+ "epoch": 1.18,
721
+ "learning_rate": 1.7889327787698105e-05,
722
+ "loss": 1.4158,
723
+ "step": 119
724
+ },
725
+ {
726
+ "epoch": 1.19,
727
+ "learning_rate": 1.784968785675221e-05,
728
+ "loss": 1.4353,
729
+ "step": 120
730
+ },
731
+ {
732
+ "epoch": 1.2,
733
+ "learning_rate": 1.7809723934825405e-05,
734
+ "loss": 1.4338,
735
+ "step": 121
736
+ },
737
+ {
738
+ "epoch": 1.21,
739
+ "learning_rate": 1.77694376714037e-05,
740
+ "loss": 1.4254,
741
+ "step": 122
742
+ },
743
+ {
744
+ "epoch": 1.22,
745
+ "learning_rate": 1.772883072927754e-05,
746
+ "loss": 1.4691,
747
+ "step": 123
748
+ },
749
+ {
750
+ "epoch": 1.23,
751
+ "learning_rate": 1.768790478447319e-05,
752
+ "loss": 1.3882,
753
+ "step": 124
754
+ },
755
+ {
756
+ "epoch": 1.24,
757
+ "learning_rate": 1.764666152618355e-05,
758
+ "loss": 1.388,
759
+ "step": 125
760
+ },
761
+ {
762
+ "epoch": 1.25,
763
+ "learning_rate": 1.7605102656698444e-05,
764
+ "loss": 1.3245,
765
+ "step": 126
766
+ },
767
+ {
768
+ "epoch": 1.26,
769
+ "learning_rate": 1.756322989133434e-05,
770
+ "loss": 1.4139,
771
+ "step": 127
772
+ },
773
+ {
774
+ "epoch": 1.27,
775
+ "learning_rate": 1.7521044958363567e-05,
776
+ "loss": 1.3521,
777
+ "step": 128
778
+ },
779
+ {
780
+ "epoch": 1.28,
781
+ "learning_rate": 1.7478549598942983e-05,
782
+ "loss": 1.2939,
783
+ "step": 129
784
+ },
785
+ {
786
+ "epoch": 1.29,
787
+ "learning_rate": 1.7435745567042096e-05,
788
+ "loss": 1.3955,
789
+ "step": 130
790
+ },
791
+ {
792
+ "epoch": 1.3,
793
+ "learning_rate": 1.7392634629370684e-05,
794
+ "loss": 1.2642,
795
+ "step": 131
796
+ },
797
+ {
798
+ "epoch": 1.31,
799
+ "learning_rate": 1.734921856530587e-05,
800
+ "loss": 1.3465,
801
+ "step": 132
802
+ },
803
+ {
804
+ "epoch": 1.32,
805
+ "learning_rate": 1.730549916681868e-05,
806
+ "loss": 1.3436,
807
+ "step": 133
808
+ },
809
+ {
810
+ "epoch": 1.33,
811
+ "learning_rate": 1.726147823840007e-05,
812
+ "loss": 1.401,
813
+ "step": 134
814
+ },
815
+ {
816
+ "epoch": 1.34,
817
+ "learning_rate": 1.7217157596986474e-05,
818
+ "loss": 1.4414,
819
+ "step": 135
820
+ },
821
+ {
822
+ "epoch": 1.35,
823
+ "learning_rate": 1.717253907188477e-05,
824
+ "loss": 1.3111,
825
+ "step": 136
826
+ },
827
+ {
828
+ "epoch": 1.36,
829
+ "learning_rate": 1.7127624504696824e-05,
830
+ "loss": 1.4064,
831
+ "step": 137
832
+ },
833
+ {
834
+ "epoch": 1.37,
835
+ "learning_rate": 1.7082415749243436e-05,
836
+ "loss": 1.3095,
837
+ "step": 138
838
+ },
839
+ {
840
+ "epoch": 1.38,
841
+ "learning_rate": 1.7036914671487854e-05,
842
+ "loss": 1.4675,
843
+ "step": 139
844
+ },
845
+ {
846
+ "epoch": 1.39,
847
+ "learning_rate": 1.699112314945874e-05,
848
+ "loss": 1.4249,
849
+ "step": 140
850
+ },
851
+ {
852
+ "epoch": 1.4,
853
+ "learning_rate": 1.694504307317267e-05,
854
+ "loss": 1.3799,
855
+ "step": 141
856
+ },
857
+ {
858
+ "epoch": 1.41,
859
+ "learning_rate": 1.689867634455612e-05,
860
+ "loss": 1.4039,
861
+ "step": 142
862
+ },
863
+ {
864
+ "epoch": 1.41,
865
+ "learning_rate": 1.6852024877366945e-05,
866
+ "loss": 1.5129,
867
+ "step": 143
868
+ },
869
+ {
870
+ "epoch": 1.42,
871
+ "learning_rate": 1.6805090597115424e-05,
872
+ "loss": 1.4277,
873
+ "step": 144
874
+ },
875
+ {
876
+ "epoch": 1.43,
877
+ "learning_rate": 1.675787544098477e-05,
878
+ "loss": 1.4467,
879
+ "step": 145
880
+ },
881
+ {
882
+ "epoch": 1.44,
883
+ "learning_rate": 1.6710381357751155e-05,
884
+ "loss": 1.4491,
885
+ "step": 146
886
+ },
887
+ {
888
+ "epoch": 1.45,
889
+ "learning_rate": 1.6662610307703318e-05,
890
+ "loss": 1.4455,
891
+ "step": 147
892
+ },
893
+ {
894
+ "epoch": 1.46,
895
+ "learning_rate": 1.661456426256161e-05,
896
+ "loss": 1.3074,
897
+ "step": 148
898
+ },
899
+ {
900
+ "epoch": 1.47,
901
+ "learning_rate": 1.6566245205396647e-05,
902
+ "loss": 1.4863,
903
+ "step": 149
904
+ },
905
+ {
906
+ "epoch": 1.48,
907
+ "learning_rate": 1.6517655130547435e-05,
908
+ "loss": 1.3705,
909
+ "step": 150
910
+ },
911
+ {
912
+ "epoch": 1.49,
913
+ "learning_rate": 1.6468796043539082e-05,
914
+ "loss": 1.4302,
915
+ "step": 151
916
+ },
917
+ {
918
+ "epoch": 1.5,
919
+ "learning_rate": 1.641966996099999e-05,
920
+ "loss": 1.3508,
921
+ "step": 152
922
+ },
923
+ {
924
+ "epoch": 1.51,
925
+ "learning_rate": 1.6370278910578644e-05,
926
+ "loss": 1.3224,
927
+ "step": 153
928
+ },
929
+ {
930
+ "epoch": 1.52,
931
+ "learning_rate": 1.6320624930859905e-05,
932
+ "loss": 1.3957,
933
+ "step": 154
934
+ },
935
+ {
936
+ "epoch": 1.53,
937
+ "learning_rate": 1.627071007128089e-05,
938
+ "loss": 1.4123,
939
+ "step": 155
940
+ },
941
+ {
942
+ "epoch": 1.54,
943
+ "learning_rate": 1.6220536392046357e-05,
944
+ "loss": 1.3938,
945
+ "step": 156
946
+ },
947
+ {
948
+ "epoch": 1.55,
949
+ "learning_rate": 1.6170105964043698e-05,
950
+ "loss": 1.3844,
951
+ "step": 157
952
+ },
953
+ {
954
+ "epoch": 1.56,
955
+ "learning_rate": 1.6119420868757433e-05,
956
+ "loss": 1.3879,
957
+ "step": 158
958
+ },
959
+ {
960
+ "epoch": 1.57,
961
+ "learning_rate": 1.606848319818333e-05,
962
+ "loss": 1.3161,
963
+ "step": 159
964
+ },
965
+ {
966
+ "epoch": 1.58,
967
+ "learning_rate": 1.6017295054742045e-05,
968
+ "loss": 1.4198,
969
+ "step": 160
970
+ },
971
+ {
972
+ "epoch": 1.59,
973
+ "learning_rate": 1.596585855119233e-05,
974
+ "loss": 1.3897,
975
+ "step": 161
976
+ },
977
+ {
978
+ "epoch": 1.6,
979
+ "learning_rate": 1.5914175810543868e-05,
980
+ "loss": 1.3397,
981
+ "step": 162
982
+ },
983
+ {
984
+ "epoch": 1.61,
985
+ "learning_rate": 1.5862248965969604e-05,
986
+ "loss": 1.4773,
987
+ "step": 163
988
+ },
989
+ {
990
+ "epoch": 1.62,
991
+ "learning_rate": 1.5810080160717737e-05,
992
+ "loss": 1.3474,
993
+ "step": 164
994
+ },
995
+ {
996
+ "epoch": 1.63,
997
+ "learning_rate": 1.575767154802323e-05,
998
+ "loss": 1.5057,
999
+ "step": 165
1000
+ },
1001
+ {
1002
+ "epoch": 1.64,
1003
+ "learning_rate": 1.570502529101896e-05,
1004
+ "loss": 1.3343,
1005
+ "step": 166
1006
+ },
1007
+ {
1008
+ "epoch": 1.65,
1009
+ "learning_rate": 1.5652143562646416e-05,
1010
+ "loss": 1.3129,
1011
+ "step": 167
1012
+ },
1013
+ {
1014
+ "epoch": 1.66,
1015
+ "learning_rate": 1.5599028545566028e-05,
1016
+ "loss": 1.2845,
1017
+ "step": 168
1018
+ },
1019
+ {
1020
+ "epoch": 1.67,
1021
+ "learning_rate": 1.5545682432067068e-05,
1022
+ "loss": 1.337,
1023
+ "step": 169
1024
+ },
1025
+ {
1026
+ "epoch": 1.68,
1027
+ "learning_rate": 1.5492107423977167e-05,
1028
+ "loss": 1.3302,
1029
+ "step": 170
1030
+ },
1031
+ {
1032
+ "epoch": 1.69,
1033
+ "learning_rate": 1.5438305732571445e-05,
1034
+ "loss": 1.348,
1035
+ "step": 171
1036
+ },
1037
+ {
1038
+ "epoch": 1.7,
1039
+ "learning_rate": 1.5384279578481223e-05,
1040
+ "loss": 1.3215,
1041
+ "step": 172
1042
+ },
1043
+ {
1044
+ "epoch": 1.71,
1045
+ "learning_rate": 1.5330031191602395e-05,
1046
+ "loss": 1.3139,
1047
+ "step": 173
1048
+ },
1049
+ {
1050
+ "epoch": 1.72,
1051
+ "learning_rate": 1.5275562811003363e-05,
1052
+ "loss": 1.3987,
1053
+ "step": 174
1054
+ },
1055
+ {
1056
+ "epoch": 1.73,
1057
+ "learning_rate": 1.522087668483264e-05,
1058
+ "loss": 1.3642,
1059
+ "step": 175
1060
+ },
1061
+ {
1062
+ "epoch": 1.74,
1063
+ "learning_rate": 1.5165975070226045e-05,
1064
+ "loss": 1.3091,
1065
+ "step": 176
1066
+ },
1067
+ {
1068
+ "epoch": 1.75,
1069
+ "learning_rate": 1.5110860233213556e-05,
1070
+ "loss": 1.2608,
1071
+ "step": 177
1072
+ },
1073
+ {
1074
+ "epoch": 1.76,
1075
+ "learning_rate": 1.5055534448625766e-05,
1076
+ "loss": 1.3476,
1077
+ "step": 178
1078
+ },
1079
+ {
1080
+ "epoch": 1.77,
1081
+ "learning_rate": 1.5000000000000002e-05,
1082
+ "loss": 1.3632,
1083
+ "step": 179
1084
+ },
1085
+ {
1086
+ "epoch": 1.78,
1087
+ "learning_rate": 1.4944259179486068e-05,
1088
+ "loss": 1.3465,
1089
+ "step": 180
1090
+ },
1091
+ {
1092
+ "epoch": 1.79,
1093
+ "learning_rate": 1.488831428775164e-05,
1094
+ "loss": 1.3829,
1095
+ "step": 181
1096
+ },
1097
+ {
1098
+ "epoch": 1.8,
1099
+ "learning_rate": 1.4832167633887306e-05,
1100
+ "loss": 1.3986,
1101
+ "step": 182
1102
+ },
1103
+ {
1104
+ "epoch": 1.81,
1105
+ "learning_rate": 1.477582153531126e-05,
1106
+ "loss": 1.3538,
1107
+ "step": 183
1108
+ },
1109
+ {
1110
+ "epoch": 1.82,
1111
+ "learning_rate": 1.4719278317673655e-05,
1112
+ "loss": 1.4173,
1113
+ "step": 184
1114
+ },
1115
+ {
1116
+ "epoch": 1.83,
1117
+ "learning_rate": 1.4662540314760608e-05,
1118
+ "loss": 1.3762,
1119
+ "step": 185
1120
+ },
1121
+ {
1122
+ "epoch": 1.84,
1123
+ "learning_rate": 1.4605609868397874e-05,
1124
+ "loss": 1.3599,
1125
+ "step": 186
1126
+ },
1127
+ {
1128
+ "epoch": 1.85,
1129
+ "learning_rate": 1.4548489328354197e-05,
1130
+ "loss": 1.324,
1131
+ "step": 187
1132
+ },
1133
+ {
1134
+ "epoch": 1.86,
1135
+ "learning_rate": 1.4491181052244317e-05,
1136
+ "loss": 1.2417,
1137
+ "step": 188
1138
+ },
1139
+ {
1140
+ "epoch": 1.87,
1141
+ "learning_rate": 1.4433687405431663e-05,
1142
+ "loss": 1.2887,
1143
+ "step": 189
1144
+ },
1145
+ {
1146
+ "epoch": 1.88,
1147
+ "learning_rate": 1.437601076093073e-05,
1148
+ "loss": 1.3587,
1149
+ "step": 190
1150
+ },
1151
+ {
1152
+ "epoch": 1.89,
1153
+ "learning_rate": 1.4318153499309118e-05,
1154
+ "loss": 1.2952,
1155
+ "step": 191
1156
+ },
1157
+ {
1158
+ "epoch": 1.9,
1159
+ "learning_rate": 1.4260118008589294e-05,
1160
+ "loss": 1.3474,
1161
+ "step": 192
1162
+ },
1163
+ {
1164
+ "epoch": 1.91,
1165
+ "learning_rate": 1.420190668415002e-05,
1166
+ "loss": 1.348,
1167
+ "step": 193
1168
+ },
1169
+ {
1170
+ "epoch": 1.92,
1171
+ "learning_rate": 1.4143521928627479e-05,
1172
+ "loss": 1.3807,
1173
+ "step": 194
1174
+ },
1175
+ {
1176
+ "epoch": 1.93,
1177
+ "learning_rate": 1.4084966151816124e-05,
1178
+ "loss": 1.3044,
1179
+ "step": 195
1180
+ },
1181
+ {
1182
+ "epoch": 1.94,
1183
+ "learning_rate": 1.4026241770569198e-05,
1184
+ "loss": 1.4231,
1185
+ "step": 196
1186
+ },
1187
+ {
1188
+ "epoch": 1.95,
1189
+ "learning_rate": 1.3967351208698985e-05,
1190
+ "loss": 1.3206,
1191
+ "step": 197
1192
+ },
1193
+ {
1194
+ "epoch": 1.96,
1195
+ "learning_rate": 1.3908296896876778e-05,
1196
+ "loss": 1.3744,
1197
+ "step": 198
1198
+ },
1199
+ {
1200
+ "epoch": 1.97,
1201
+ "learning_rate": 1.3849081272532545e-05,
1202
+ "loss": 1.4214,
1203
+ "step": 199
1204
+ },
1205
+ {
1206
+ "epoch": 1.98,
1207
+ "learning_rate": 1.3789706779754326e-05,
1208
+ "loss": 1.4233,
1209
+ "step": 200
1210
+ },
1211
+ {
1212
+ "epoch": 1.99,
1213
+ "learning_rate": 1.373017586918736e-05,
1214
+ "loss": 1.3545,
1215
+ "step": 201
1216
+ },
1217
+ {
1218
+ "epoch": 2.0,
1219
+ "learning_rate": 1.3670490997932922e-05,
1220
+ "loss": 1.2513,
1221
+ "step": 202
1222
+ },
1223
+ {
1224
+ "epoch": 2.01,
1225
+ "learning_rate": 1.3610654629446938e-05,
1226
+ "loss": 1.3838,
1227
+ "step": 203
1228
+ },
1229
+ {
1230
+ "epoch": 2.02,
1231
+ "learning_rate": 1.3550669233438271e-05,
1232
+ "loss": 1.3655,
1233
+ "step": 204
1234
+ },
1235
+ {
1236
+ "epoch": 2.03,
1237
+ "learning_rate": 1.3490537285766809e-05,
1238
+ "loss": 1.3503,
1239
+ "step": 205
1240
+ },
1241
+ {
1242
+ "epoch": 2.04,
1243
+ "learning_rate": 1.3430261268341272e-05,
1244
+ "loss": 1.2948,
1245
+ "step": 206
1246
+ },
1247
+ {
1248
+ "epoch": 2.05,
1249
+ "learning_rate": 1.3369843669016757e-05,
1250
+ "loss": 1.3722,
1251
+ "step": 207
1252
+ },
1253
+ {
1254
+ "epoch": 2.06,
1255
+ "learning_rate": 1.3309286981492084e-05,
1256
+ "loss": 1.3155,
1257
+ "step": 208
1258
+ },
1259
+ {
1260
+ "epoch": 2.07,
1261
+ "learning_rate": 1.3248593705206838e-05,
1262
+ "loss": 1.2492,
1263
+ "step": 209
1264
+ },
1265
+ {
1266
+ "epoch": 2.08,
1267
+ "learning_rate": 1.3187766345238222e-05,
1268
+ "loss": 1.3089,
1269
+ "step": 210
1270
+ },
1271
+ {
1272
+ "epoch": 2.09,
1273
+ "learning_rate": 1.3126807412197666e-05,
1274
+ "loss": 1.3134,
1275
+ "step": 211
1276
+ },
1277
+ {
1278
+ "epoch": 2.1,
1279
+ "learning_rate": 1.3065719422127188e-05,
1280
+ "loss": 1.2995,
1281
+ "step": 212
1282
+ },
1283
+ {
1284
+ "epoch": 2.11,
1285
+ "learning_rate": 1.3004504896395564e-05,
1286
+ "loss": 1.3436,
1287
+ "step": 213
1288
+ },
1289
+ {
1290
+ "epoch": 2.12,
1291
+ "learning_rate": 1.2943166361594242e-05,
1292
+ "loss": 1.1327,
1293
+ "step": 214
1294
+ },
1295
+ {
1296
+ "epoch": 2.13,
1297
+ "learning_rate": 1.288170634943307e-05,
1298
+ "loss": 1.1199,
1299
+ "step": 215
1300
+ },
1301
+ {
1302
+ "epoch": 2.14,
1303
+ "learning_rate": 1.2820127396635802e-05,
1304
+ "loss": 1.1588,
1305
+ "step": 216
1306
+ },
1307
+ {
1308
+ "epoch": 2.15,
1309
+ "learning_rate": 1.275843204483539e-05,
1310
+ "loss": 1.1224,
1311
+ "step": 217
1312
+ },
1313
+ {
1314
+ "epoch": 2.16,
1315
+ "learning_rate": 1.2696622840469084e-05,
1316
+ "loss": 1.3147,
1317
+ "step": 218
1318
+ },
1319
+ {
1320
+ "epoch": 2.17,
1321
+ "learning_rate": 1.263470233467332e-05,
1322
+ "loss": 1.1618,
1323
+ "step": 219
1324
+ },
1325
+ {
1326
+ "epoch": 2.18,
1327
+ "learning_rate": 1.2572673083178448e-05,
1328
+ "loss": 1.195,
1329
+ "step": 220
1330
+ },
1331
+ {
1332
+ "epoch": 2.19,
1333
+ "learning_rate": 1.2510537646203209e-05,
1334
+ "loss": 1.1712,
1335
+ "step": 221
1336
+ },
1337
+ {
1338
+ "epoch": 2.2,
1339
+ "learning_rate": 1.2448298588349097e-05,
1340
+ "loss": 1.1937,
1341
+ "step": 222
1342
+ },
1343
+ {
1344
+ "epoch": 2.21,
1345
+ "learning_rate": 1.2385958478494487e-05,
1346
+ "loss": 1.192,
1347
+ "step": 223
1348
+ },
1349
+ {
1350
+ "epoch": 2.22,
1351
+ "learning_rate": 1.2323519889688615e-05,
1352
+ "loss": 1.2389,
1353
+ "step": 224
1354
+ },
1355
+ {
1356
+ "epoch": 2.23,
1357
+ "learning_rate": 1.2260985399045379e-05,
1358
+ "loss": 1.1535,
1359
+ "step": 225
1360
+ },
1361
+ {
1362
+ "epoch": 2.24,
1363
+ "learning_rate": 1.2198357587636958e-05,
1364
+ "loss": 1.1672,
1365
+ "step": 226
1366
+ },
1367
+ {
1368
+ "epoch": 2.25,
1369
+ "learning_rate": 1.2135639040387291e-05,
1370
+ "loss": 1.1308,
1371
+ "step": 227
1372
+ },
1373
+ {
1374
+ "epoch": 2.26,
1375
+ "learning_rate": 1.2072832345965381e-05,
1376
+ "loss": 1.1642,
1377
+ "step": 228
1378
+ },
1379
+ {
1380
+ "epoch": 2.27,
1381
+ "learning_rate": 1.2009940096678451e-05,
1382
+ "loss": 1.0826,
1383
+ "step": 229
1384
+ },
1385
+ {
1386
+ "epoch": 2.28,
1387
+ "learning_rate": 1.1946964888364949e-05,
1388
+ "loss": 1.0454,
1389
+ "step": 230
1390
+ },
1391
+ {
1392
+ "epoch": 2.29,
1393
+ "learning_rate": 1.1883909320287406e-05,
1394
+ "loss": 1.1272,
1395
+ "step": 231
1396
+ },
1397
+ {
1398
+ "epoch": 2.3,
1399
+ "learning_rate": 1.1820775995025147e-05,
1400
+ "loss": 1.0187,
1401
+ "step": 232
1402
+ },
1403
+ {
1404
+ "epoch": 2.31,
1405
+ "learning_rate": 1.1757567518366883e-05,
1406
+ "loss": 1.1232,
1407
+ "step": 233
1408
+ },
1409
+ {
1410
+ "epoch": 2.32,
1411
+ "learning_rate": 1.169428649920315e-05,
1412
+ "loss": 1.0917,
1413
+ "step": 234
1414
+ },
1415
+ {
1416
+ "epoch": 2.33,
1417
+ "learning_rate": 1.1630935549418627e-05,
1418
+ "loss": 1.1499,
1419
+ "step": 235
1420
+ },
1421
+ {
1422
+ "epoch": 2.34,
1423
+ "learning_rate": 1.1567517283784344e-05,
1424
+ "loss": 1.2013,
1425
+ "step": 236
1426
+ },
1427
+ {
1428
+ "epoch": 2.35,
1429
+ "learning_rate": 1.1504034319849741e-05,
1430
+ "loss": 1.0021,
1431
+ "step": 237
1432
+ },
1433
+ {
1434
+ "epoch": 2.35,
1435
+ "learning_rate": 1.1440489277834645e-05,
1436
+ "loss": 1.145,
1437
+ "step": 238
1438
+ },
1439
+ {
1440
+ "epoch": 2.36,
1441
+ "learning_rate": 1.1376884780521117e-05,
1442
+ "loss": 1.0523,
1443
+ "step": 239
1444
+ },
1445
+ {
1446
+ "epoch": 2.37,
1447
+ "learning_rate": 1.1313223453145202e-05,
1448
+ "loss": 1.199,
1449
+ "step": 240
1450
+ },
1451
+ {
1452
+ "epoch": 2.38,
1453
+ "learning_rate": 1.1249507923288563e-05,
1454
+ "loss": 1.1381,
1455
+ "step": 241
1456
+ },
1457
+ {
1458
+ "epoch": 2.39,
1459
+ "learning_rate": 1.1185740820770042e-05,
1460
+ "loss": 1.1056,
1461
+ "step": 242
1462
+ },
1463
+ {
1464
+ "epoch": 2.4,
1465
+ "learning_rate": 1.1121924777537108e-05,
1466
+ "loss": 1.152,
1467
+ "step": 243
1468
+ },
1469
+ {
1470
+ "epoch": 2.41,
1471
+ "learning_rate": 1.105806242755723e-05,
1472
+ "loss": 1.2463,
1473
+ "step": 244
1474
+ },
1475
+ {
1476
+ "epoch": 2.42,
1477
+ "learning_rate": 1.0994156406709155e-05,
1478
+ "loss": 1.1991,
1479
+ "step": 245
1480
+ },
1481
+ {
1482
+ "epoch": 2.43,
1483
+ "learning_rate": 1.0930209352674123e-05,
1484
+ "loss": 1.2074,
1485
+ "step": 246
1486
+ },
1487
+ {
1488
+ "epoch": 2.44,
1489
+ "learning_rate": 1.0866223904826992e-05,
1490
+ "loss": 1.2137,
1491
+ "step": 247
1492
+ },
1493
+ {
1494
+ "epoch": 2.45,
1495
+ "learning_rate": 1.0802202704127293e-05,
1496
+ "loss": 1.1762,
1497
+ "step": 248
1498
+ },
1499
+ {
1500
+ "epoch": 2.46,
1501
+ "learning_rate": 1.0738148393010251e-05,
1502
+ "loss": 1.1095,
1503
+ "step": 249
1504
+ },
1505
+ {
1506
+ "epoch": 2.47,
1507
+ "learning_rate": 1.0674063615277681e-05,
1508
+ "loss": 1.2061,
1509
+ "step": 250
1510
+ },
1511
+ {
1512
+ "epoch": 2.48,
1513
+ "learning_rate": 1.0609951015988907e-05,
1514
+ "loss": 1.1453,
1515
+ "step": 251
1516
+ },
1517
+ {
1518
+ "epoch": 2.49,
1519
+ "learning_rate": 1.054581324135156e-05,
1520
+ "loss": 1.1776,
1521
+ "step": 252
1522
+ },
1523
+ {
1524
+ "epoch": 2.5,
1525
+ "learning_rate": 1.0481652938612374e-05,
1526
+ "loss": 1.1084,
1527
+ "step": 253
1528
+ },
1529
+ {
1530
+ "epoch": 2.51,
1531
+ "learning_rate": 1.0417472755947908e-05,
1532
+ "loss": 1.1028,
1533
+ "step": 254
1534
+ },
1535
+ {
1536
+ "epoch": 2.52,
1537
+ "learning_rate": 1.0353275342355262e-05,
1538
+ "loss": 1.1285,
1539
+ "step": 255
1540
+ },
1541
+ {
1542
+ "epoch": 2.53,
1543
+ "learning_rate": 1.0289063347542727e-05,
1544
+ "loss": 1.1332,
1545
+ "step": 256
1546
+ },
1547
+ {
1548
+ "epoch": 2.54,
1549
+ "learning_rate": 1.0224839421820426e-05,
1550
+ "loss": 1.182,
1551
+ "step": 257
1552
+ },
1553
+ {
1554
+ "epoch": 2.55,
1555
+ "learning_rate": 1.0160606215990922e-05,
1556
+ "loss": 1.1823,
1557
+ "step": 258
1558
+ },
1559
+ {
1560
+ "epoch": 2.56,
1561
+ "learning_rate": 1.0096366381239808e-05,
1562
+ "loss": 1.14,
1563
+ "step": 259
1564
+ },
1565
+ {
1566
+ "epoch": 2.57,
1567
+ "learning_rate": 1.0032122569026284e-05,
1568
+ "loss": 1.0837,
1569
+ "step": 260
1570
+ },
1571
+ {
1572
+ "epoch": 2.58,
1573
+ "learning_rate": 9.967877430973716e-06,
1574
+ "loss": 1.1733,
1575
+ "step": 261
1576
+ },
1577
+ {
1578
+ "epoch": 2.59,
1579
+ "learning_rate": 9.903633618760195e-06,
1580
+ "loss": 1.128,
1581
+ "step": 262
1582
+ },
1583
+ {
1584
+ "epoch": 2.6,
1585
+ "learning_rate": 9.839393784009078e-06,
1586
+ "loss": 1.1153,
1587
+ "step": 263
1588
+ },
1589
+ {
1590
+ "epoch": 2.61,
1591
+ "learning_rate": 9.775160578179575e-06,
1592
+ "loss": 1.2042,
1593
+ "step": 264
1594
+ },
1595
+ {
1596
+ "epoch": 2.62,
1597
+ "learning_rate": 9.710936652457276e-06,
1598
+ "loss": 1.1457,
1599
+ "step": 265
1600
+ },
1601
+ {
1602
+ "epoch": 2.63,
1603
+ "learning_rate": 9.64672465764474e-06,
1604
+ "loss": 1.2454,
1605
+ "step": 266
1606
+ },
1607
+ {
1608
+ "epoch": 2.64,
1609
+ "learning_rate": 9.582527244052095e-06,
1610
+ "loss": 1.078,
1611
+ "step": 267
1612
+ },
1613
+ {
1614
+ "epoch": 2.65,
1615
+ "learning_rate": 9.518347061387629e-06,
1616
+ "loss": 1.0835,
1617
+ "step": 268
1618
+ },
1619
+ {
1620
+ "epoch": 2.66,
1621
+ "learning_rate": 9.454186758648444e-06,
1622
+ "loss": 1.0497,
1623
+ "step": 269
1624
+ },
1625
+ {
1626
+ "epoch": 2.67,
1627
+ "learning_rate": 9.390048984011095e-06,
1628
+ "loss": 1.0864,
1629
+ "step": 270
1630
+ },
1631
+ {
1632
+ "epoch": 2.68,
1633
+ "learning_rate": 9.325936384722322e-06,
1634
+ "loss": 1.0472,
1635
+ "step": 271
1636
+ },
1637
+ {
1638
+ "epoch": 2.69,
1639
+ "learning_rate": 9.261851606989754e-06,
1640
+ "loss": 1.1284,
1641
+ "step": 272
1642
+ },
1643
+ {
1644
+ "epoch": 2.7,
1645
+ "learning_rate": 9.197797295872709e-06,
1646
+ "loss": 1.0829,
1647
+ "step": 273
1648
+ },
1649
+ {
1650
+ "epoch": 2.71,
1651
+ "learning_rate": 9.133776095173015e-06,
1652
+ "loss": 1.0973,
1653
+ "step": 274
1654
+ },
1655
+ {
1656
+ "epoch": 2.72,
1657
+ "learning_rate": 9.069790647325879e-06,
1658
+ "loss": 1.1343,
1659
+ "step": 275
1660
+ },
1661
+ {
1662
+ "epoch": 2.73,
1663
+ "learning_rate": 9.005843593290849e-06,
1664
+ "loss": 1.1389,
1665
+ "step": 276
1666
+ },
1667
+ {
1668
+ "epoch": 2.74,
1669
+ "learning_rate": 8.941937572442773e-06,
1670
+ "loss": 1.0836,
1671
+ "step": 277
1672
+ },
1673
+ {
1674
+ "epoch": 2.75,
1675
+ "learning_rate": 8.878075222462896e-06,
1676
+ "loss": 1.0652,
1677
+ "step": 278
1678
+ },
1679
+ {
1680
+ "epoch": 2.76,
1681
+ "learning_rate": 8.81425917922996e-06,
1682
+ "loss": 1.0596,
1683
+ "step": 279
1684
+ },
1685
+ {
1686
+ "epoch": 2.77,
1687
+ "learning_rate": 8.750492076711439e-06,
1688
+ "loss": 1.1175,
1689
+ "step": 280
1690
+ },
1691
+ {
1692
+ "epoch": 2.78,
1693
+ "learning_rate": 8.6867765468548e-06,
1694
+ "loss": 1.0994,
1695
+ "step": 281
1696
+ },
1697
+ {
1698
+ "epoch": 2.79,
1699
+ "learning_rate": 8.623115219478884e-06,
1700
+ "loss": 1.1334,
1701
+ "step": 282
1702
+ },
1703
+ {
1704
+ "epoch": 2.8,
1705
+ "learning_rate": 8.55951072216536e-06,
1706
+ "loss": 1.1725,
1707
+ "step": 283
1708
+ },
1709
+ {
1710
+ "epoch": 2.81,
1711
+ "learning_rate": 8.49596568015026e-06,
1712
+ "loss": 1.0953,
1713
+ "step": 284
1714
+ },
1715
+ {
1716
+ "epoch": 2.82,
1717
+ "learning_rate": 8.432482716215663e-06,
1718
+ "loss": 1.1551,
1719
+ "step": 285
1720
+ },
1721
+ {
1722
+ "epoch": 2.83,
1723
+ "learning_rate": 8.369064450581374e-06,
1724
+ "loss": 1.0854,
1725
+ "step": 286
1726
+ },
1727
+ {
1728
+ "epoch": 2.84,
1729
+ "learning_rate": 8.305713500796852e-06,
1730
+ "loss": 1.1419,
1731
+ "step": 287
1732
+ },
1733
+ {
1734
+ "epoch": 2.85,
1735
+ "learning_rate": 8.242432481633119e-06,
1736
+ "loss": 1.0436,
1737
+ "step": 288
1738
+ },
1739
+ {
1740
+ "epoch": 2.86,
1741
+ "learning_rate": 8.179224004974857e-06,
1742
+ "loss": 1.0331,
1743
+ "step": 289
1744
+ },
1745
+ {
1746
+ "epoch": 2.87,
1747
+ "learning_rate": 8.116090679712601e-06,
1748
+ "loss": 1.0594,
1749
+ "step": 290
1750
+ },
1751
+ {
1752
+ "epoch": 2.88,
1753
+ "learning_rate": 8.053035111635054e-06,
1754
+ "loss": 1.0856,
1755
+ "step": 291
1756
+ },
1757
+ {
1758
+ "epoch": 2.89,
1759
+ "learning_rate": 7.990059903321554e-06,
1760
+ "loss": 1.0206,
1761
+ "step": 292
1762
+ },
1763
+ {
1764
+ "epoch": 2.9,
1765
+ "learning_rate": 7.927167654034622e-06,
1766
+ "loss": 1.1111,
1767
+ "step": 293
1768
+ },
1769
+ {
1770
+ "epoch": 2.91,
1771
+ "learning_rate": 7.864360959612714e-06,
1772
+ "loss": 1.1125,
1773
+ "step": 294
1774
+ },
1775
+ {
1776
+ "epoch": 2.92,
1777
+ "learning_rate": 7.801642412363042e-06,
1778
+ "loss": 1.1747,
1779
+ "step": 295
1780
+ },
1781
+ {
1782
+ "epoch": 2.93,
1783
+ "learning_rate": 7.739014600954623e-06,
1784
+ "loss": 1.0466,
1785
+ "step": 296
1786
+ },
1787
+ {
1788
+ "epoch": 2.94,
1789
+ "learning_rate": 7.676480110311385e-06,
1790
+ "loss": 1.1687,
1791
+ "step": 297
1792
+ },
1793
+ {
1794
+ "epoch": 2.95,
1795
+ "learning_rate": 7.614041521505517e-06,
1796
+ "loss": 1.0808,
1797
+ "step": 298
1798
+ },
1799
+ {
1800
+ "epoch": 2.96,
1801
+ "learning_rate": 7.5517014116509094e-06,
1802
+ "loss": 1.1393,
1803
+ "step": 299
1804
+ },
1805
+ {
1806
+ "epoch": 2.97,
1807
+ "learning_rate": 7.489462353796792e-06,
1808
+ "loss": 1.1399,
1809
+ "step": 300
1810
+ },
1811
+ {
1812
+ "epoch": 2.98,
1813
+ "learning_rate": 7.427326916821557e-06,
1814
+ "loss": 1.1977,
1815
+ "step": 301
1816
+ },
1817
+ {
1818
+ "epoch": 2.99,
1819
+ "learning_rate": 7.3652976653266785e-06,
1820
+ "loss": 1.1136,
1821
+ "step": 302
1822
+ },
1823
+ {
1824
+ "epoch": 3.0,
1825
+ "learning_rate": 7.303377159530919e-06,
1826
+ "loss": 1.0574,
1827
+ "step": 303
1828
+ },
1829
+ {
1830
+ "epoch": 3.01,
1831
+ "learning_rate": 7.24156795516461e-06,
1832
+ "loss": 1.1441,
1833
+ "step": 304
1834
+ },
1835
+ {
1836
+ "epoch": 3.02,
1837
+ "learning_rate": 7.1798726033642e-06,
1838
+ "loss": 1.1542,
1839
+ "step": 305
1840
+ },
1841
+ {
1842
+ "epoch": 3.03,
1843
+ "learning_rate": 7.118293650566931e-06,
1844
+ "loss": 1.1092,
1845
+ "step": 306
1846
+ },
1847
+ {
1848
+ "epoch": 3.04,
1849
+ "learning_rate": 7.056833638405762e-06,
1850
+ "loss": 1.064,
1851
+ "step": 307
1852
+ },
1853
+ {
1854
+ "epoch": 3.05,
1855
+ "learning_rate": 6.995495103604442e-06,
1856
+ "loss": 1.1304,
1857
+ "step": 308
1858
+ },
1859
+ {
1860
+ "epoch": 3.06,
1861
+ "learning_rate": 6.934280577872814e-06,
1862
+ "loss": 1.0682,
1863
+ "step": 309
1864
+ },
1865
+ {
1866
+ "epoch": 3.07,
1867
+ "learning_rate": 6.87319258780234e-06,
1868
+ "loss": 1.0223,
1869
+ "step": 310
1870
+ },
1871
+ {
1872
+ "epoch": 3.08,
1873
+ "learning_rate": 6.812233654761779e-06,
1874
+ "loss": 1.0467,
1875
+ "step": 311
1876
+ },
1877
+ {
1878
+ "epoch": 3.09,
1879
+ "learning_rate": 6.7514062947931655e-06,
1880
+ "loss": 1.0722,
1881
+ "step": 312
1882
+ },
1883
+ {
1884
+ "epoch": 3.1,
1885
+ "learning_rate": 6.690713018507917e-06,
1886
+ "loss": 1.0569,
1887
+ "step": 313
1888
+ },
1889
+ {
1890
+ "epoch": 3.11,
1891
+ "learning_rate": 6.630156330983244e-06,
1892
+ "loss": 1.1766,
1893
+ "step": 314
1894
+ },
1895
+ {
1896
+ "epoch": 3.12,
1897
+ "learning_rate": 6.569738731658735e-06,
1898
+ "loss": 0.9828,
1899
+ "step": 315
1900
+ },
1901
+ {
1902
+ "epoch": 3.13,
1903
+ "learning_rate": 6.509462714233194e-06,
1904
+ "loss": 0.9987,
1905
+ "step": 316
1906
+ },
1907
+ {
1908
+ "epoch": 3.14,
1909
+ "learning_rate": 6.449330766561735e-06,
1910
+ "loss": 0.9584,
1911
+ "step": 317
1912
+ },
1913
+ {
1914
+ "epoch": 3.15,
1915
+ "learning_rate": 6.389345370553065e-06,
1916
+ "loss": 0.9931,
1917
+ "step": 318
1918
+ },
1919
+ {
1920
+ "epoch": 3.16,
1921
+ "learning_rate": 6.32950900206708e-06,
1922
+ "loss": 1.1554,
1923
+ "step": 319
1924
+ },
1925
+ {
1926
+ "epoch": 3.17,
1927
+ "learning_rate": 6.269824130812645e-06,
1928
+ "loss": 0.9724,
1929
+ "step": 320
1930
+ },
1931
+ {
1932
+ "epoch": 3.18,
1933
+ "learning_rate": 6.210293220245678e-06,
1934
+ "loss": 1.0328,
1935
+ "step": 321
1936
+ },
1937
+ {
1938
+ "epoch": 3.19,
1939
+ "learning_rate": 6.150918727467455e-06,
1940
+ "loss": 0.9994,
1941
+ "step": 322
1942
+ },
1943
+ {
1944
+ "epoch": 3.2,
1945
+ "learning_rate": 6.091703103123223e-06,
1946
+ "loss": 1.0123,
1947
+ "step": 323
1948
+ },
1949
+ {
1950
+ "epoch": 3.21,
1951
+ "learning_rate": 6.032648791301019e-06,
1952
+ "loss": 1.003,
1953
+ "step": 324
1954
+ },
1955
+ {
1956
+ "epoch": 3.22,
1957
+ "learning_rate": 5.973758229430806e-06,
1958
+ "loss": 1.0226,
1959
+ "step": 325
1960
+ },
1961
+ {
1962
+ "epoch": 3.23,
1963
+ "learning_rate": 5.91503384818388e-06,
1964
+ "loss": 0.9729,
1965
+ "step": 326
1966
+ },
1967
+ {
1968
+ "epoch": 3.24,
1969
+ "learning_rate": 5.856478071372521e-06,
1970
+ "loss": 0.9728,
1971
+ "step": 327
1972
+ },
1973
+ {
1974
+ "epoch": 3.25,
1975
+ "learning_rate": 5.798093315849984e-06,
1976
+ "loss": 0.9271,
1977
+ "step": 328
1978
+ },
1979
+ {
1980
+ "epoch": 3.26,
1981
+ "learning_rate": 5.739881991410707e-06,
1982
+ "loss": 0.962,
1983
+ "step": 329
1984
+ },
1985
+ {
1986
+ "epoch": 3.27,
1987
+ "learning_rate": 5.681846500690884e-06,
1988
+ "loss": 0.8697,
1989
+ "step": 330
1990
+ },
1991
+ {
1992
+ "epoch": 3.28,
1993
+ "learning_rate": 5.623989239069275e-06,
1994
+ "loss": 0.8456,
1995
+ "step": 331
1996
+ },
1997
+ {
1998
+ "epoch": 3.29,
1999
+ "learning_rate": 5.56631259456834e-06,
2000
+ "loss": 0.9068,
2001
+ "step": 332
2002
+ },
2003
+ {
2004
+ "epoch": 3.29,
2005
+ "learning_rate": 5.508818947755687e-06,
2006
+ "loss": 0.8098,
2007
+ "step": 333
2008
+ },
2009
+ {
2010
+ "epoch": 3.3,
2011
+ "learning_rate": 5.451510671645806e-06,
2012
+ "loss": 0.9104,
2013
+ "step": 334
2014
+ },
2015
+ {
2016
+ "epoch": 3.31,
2017
+ "learning_rate": 5.394390131602133e-06,
2018
+ "loss": 0.904,
2019
+ "step": 335
2020
+ },
2021
+ {
2022
+ "epoch": 3.32,
2023
+ "learning_rate": 5.337459685239395e-06,
2024
+ "loss": 0.9783,
2025
+ "step": 336
2026
+ },
2027
+ {
2028
+ "epoch": 3.33,
2029
+ "learning_rate": 5.280721682326349e-06,
2030
+ "loss": 1.0197,
2031
+ "step": 337
2032
+ },
2033
+ {
2034
+ "epoch": 3.34,
2035
+ "learning_rate": 5.224178464688742e-06,
2036
+ "loss": 0.8167,
2037
+ "step": 338
2038
+ },
2039
+ {
2040
+ "epoch": 3.35,
2041
+ "learning_rate": 5.167832366112695e-06,
2042
+ "loss": 0.9123,
2043
+ "step": 339
2044
+ },
2045
+ {
2046
+ "epoch": 3.36,
2047
+ "learning_rate": 5.111685712248364e-06,
2048
+ "loss": 0.8262,
2049
+ "step": 340
2050
+ },
2051
+ {
2052
+ "epoch": 3.37,
2053
+ "learning_rate": 5.055740820513932e-06,
2054
+ "loss": 1.0102,
2055
+ "step": 341
2056
+ },
2057
+ {
2058
+ "epoch": 3.38,
2059
+ "learning_rate": 5.000000000000003e-06,
2060
+ "loss": 0.9307,
2061
+ "step": 342
2062
+ },
2063
+ {
2064
+ "epoch": 3.39,
2065
+ "learning_rate": 4.944465551374238e-06,
2066
+ "loss": 0.8742,
2067
+ "step": 343
2068
+ },
2069
+ {
2070
+ "epoch": 3.4,
2071
+ "learning_rate": 4.889139766786447e-06,
2072
+ "loss": 0.9395,
2073
+ "step": 344
2074
+ },
2075
+ {
2076
+ "epoch": 3.41,
2077
+ "learning_rate": 4.834024929773956e-06,
2078
+ "loss": 1.0364,
2079
+ "step": 345
2080
+ },
2081
+ {
2082
+ "epoch": 3.42,
2083
+ "learning_rate": 4.779123315167362e-06,
2084
+ "loss": 1.0022,
2085
+ "step": 346
2086
+ },
2087
+ {
2088
+ "epoch": 3.43,
2089
+ "learning_rate": 4.7244371889966374e-06,
2090
+ "loss": 0.9957,
2091
+ "step": 347
2092
+ },
2093
+ {
2094
+ "epoch": 3.44,
2095
+ "learning_rate": 4.669968808397609e-06,
2096
+ "loss": 1.027,
2097
+ "step": 348
2098
+ },
2099
+ {
2100
+ "epoch": 3.45,
2101
+ "learning_rate": 4.61572042151878e-06,
2102
+ "loss": 0.9526,
2103
+ "step": 349
2104
+ },
2105
+ {
2106
+ "epoch": 3.46,
2107
+ "learning_rate": 4.56169426742856e-06,
2108
+ "loss": 0.9176,
2109
+ "step": 350
2110
+ },
2111
+ {
2112
+ "epoch": 3.47,
2113
+ "learning_rate": 4.507892576022838e-06,
2114
+ "loss": 1.0136,
2115
+ "step": 351
2116
+ },
2117
+ {
2118
+ "epoch": 3.48,
2119
+ "learning_rate": 4.4543175679329345e-06,
2120
+ "loss": 0.959,
2121
+ "step": 352
2122
+ },
2123
+ {
2124
+ "epoch": 3.49,
2125
+ "learning_rate": 4.4009714544339755e-06,
2126
+ "loss": 0.9847,
2127
+ "step": 353
2128
+ },
2129
+ {
2130
+ "epoch": 3.5,
2131
+ "learning_rate": 4.347856437353584e-06,
2132
+ "loss": 0.8823,
2133
+ "step": 354
2134
+ },
2135
+ {
2136
+ "epoch": 3.51,
2137
+ "learning_rate": 4.294974708981041e-06,
2138
+ "loss": 0.917,
2139
+ "step": 355
2140
+ },
2141
+ {
2142
+ "epoch": 3.52,
2143
+ "learning_rate": 4.242328451976774e-06,
2144
+ "loss": 0.9402,
2145
+ "step": 356
2146
+ },
2147
+ {
2148
+ "epoch": 3.53,
2149
+ "learning_rate": 4.189919839282265e-06,
2150
+ "loss": 0.9244,
2151
+ "step": 357
2152
+ },
2153
+ {
2154
+ "epoch": 3.54,
2155
+ "learning_rate": 4.1377510340304e-06,
2156
+ "loss": 1.0061,
2157
+ "step": 358
2158
+ },
2159
+ {
2160
+ "epoch": 3.55,
2161
+ "learning_rate": 4.085824189456136e-06,
2162
+ "loss": 0.9854,
2163
+ "step": 359
2164
+ },
2165
+ {
2166
+ "epoch": 3.56,
2167
+ "learning_rate": 4.03414144880767e-06,
2168
+ "loss": 0.9865,
2169
+ "step": 360
2170
+ },
2171
+ {
2172
+ "epoch": 3.57,
2173
+ "learning_rate": 3.982704945257957e-06,
2174
+ "loss": 0.9202,
2175
+ "step": 361
2176
+ },
2177
+ {
2178
+ "epoch": 3.58,
2179
+ "learning_rate": 3.931516801816668e-06,
2180
+ "loss": 0.971,
2181
+ "step": 362
2182
+ },
2183
+ {
2184
+ "epoch": 3.59,
2185
+ "learning_rate": 3.880579131242567e-06,
2186
+ "loss": 0.9601,
2187
+ "step": 363
2188
+ },
2189
+ {
2190
+ "epoch": 3.6,
2191
+ "learning_rate": 3.829894035956306e-06,
2192
+ "loss": 0.9382,
2193
+ "step": 364
2194
+ },
2195
+ {
2196
+ "epoch": 3.61,
2197
+ "learning_rate": 3.779463607953644e-06,
2198
+ "loss": 0.9866,
2199
+ "step": 365
2200
+ },
2201
+ {
2202
+ "epoch": 3.62,
2203
+ "learning_rate": 3.729289928719113e-06,
2204
+ "loss": 0.9524,
2205
+ "step": 366
2206
+ },
2207
+ {
2208
+ "epoch": 3.63,
2209
+ "learning_rate": 3.6793750691400996e-06,
2210
+ "loss": 1.0385,
2211
+ "step": 367
2212
+ },
2213
+ {
2214
+ "epoch": 3.64,
2215
+ "learning_rate": 3.62972108942136e-06,
2216
+ "loss": 0.8792,
2217
+ "step": 368
2218
+ },
2219
+ {
2220
+ "epoch": 3.65,
2221
+ "learning_rate": 3.580330039000014e-06,
2222
+ "loss": 0.9042,
2223
+ "step": 369
2224
+ },
2225
+ {
2226
+ "epoch": 3.66,
2227
+ "learning_rate": 3.5312039564609203e-06,
2228
+ "loss": 0.8561,
2229
+ "step": 370
2230
+ },
2231
+ {
2232
+ "epoch": 3.67,
2233
+ "learning_rate": 3.482344869452565e-06,
2234
+ "loss": 0.948,
2235
+ "step": 371
2236
+ },
2237
+ {
2238
+ "epoch": 3.68,
2239
+ "learning_rate": 3.4337547946033557e-06,
2240
+ "loss": 0.8488,
2241
+ "step": 372
2242
+ },
2243
+ {
2244
+ "epoch": 3.69,
2245
+ "learning_rate": 3.3854357374383905e-06,
2246
+ "loss": 0.9748,
2247
+ "step": 373
2248
+ },
2249
+ {
2250
+ "epoch": 3.7,
2251
+ "learning_rate": 3.3373896922966863e-06,
2252
+ "loss": 0.9182,
2253
+ "step": 374
2254
+ },
2255
+ {
2256
+ "epoch": 3.71,
2257
+ "learning_rate": 3.2896186422488463e-06,
2258
+ "loss": 0.9528,
2259
+ "step": 375
2260
+ },
2261
+ {
2262
+ "epoch": 3.72,
2263
+ "learning_rate": 3.242124559015234e-06,
2264
+ "loss": 0.9543,
2265
+ "step": 376
2266
+ },
2267
+ {
2268
+ "epoch": 3.73,
2269
+ "learning_rate": 3.194909402884576e-06,
2270
+ "loss": 0.9576,
2271
+ "step": 377
2272
+ },
2273
+ {
2274
+ "epoch": 3.74,
2275
+ "learning_rate": 3.1479751226330567e-06,
2276
+ "loss": 0.8864,
2277
+ "step": 378
2278
+ },
2279
+ {
2280
+ "epoch": 3.75,
2281
+ "learning_rate": 3.101323655443882e-06,
2282
+ "loss": 0.9,
2283
+ "step": 379
2284
+ },
2285
+ {
2286
+ "epoch": 3.76,
2287
+ "learning_rate": 3.0549569268273316e-06,
2288
+ "loss": 0.9028,
2289
+ "step": 380
2290
+ },
2291
+ {
2292
+ "epoch": 3.77,
2293
+ "learning_rate": 3.0088768505412623e-06,
2294
+ "loss": 0.9111,
2295
+ "step": 381
2296
+ },
2297
+ {
2298
+ "epoch": 3.78,
2299
+ "learning_rate": 2.9630853285121506e-06,
2300
+ "loss": 0.9345,
2301
+ "step": 382
2302
+ },
2303
+ {
2304
+ "epoch": 3.79,
2305
+ "learning_rate": 2.91758425075657e-06,
2306
+ "loss": 0.9312,
2307
+ "step": 383
2308
+ },
2309
+ {
2310
+ "epoch": 3.8,
2311
+ "learning_rate": 2.872375495303178e-06,
2312
+ "loss": 0.9943,
2313
+ "step": 384
2314
+ },
2315
+ {
2316
+ "epoch": 3.81,
2317
+ "learning_rate": 2.8274609281152322e-06,
2318
+ "loss": 0.9443,
2319
+ "step": 385
2320
+ },
2321
+ {
2322
+ "epoch": 3.82,
2323
+ "learning_rate": 2.7828424030135305e-06,
2324
+ "loss": 0.9603,
2325
+ "step": 386
2326
+ },
2327
+ {
2328
+ "epoch": 3.83,
2329
+ "learning_rate": 2.7385217615999303e-06,
2330
+ "loss": 0.9195,
2331
+ "step": 387
2332
+ },
2333
+ {
2334
+ "epoch": 3.84,
2335
+ "learning_rate": 2.694500833181323e-06,
2336
+ "loss": 0.9257,
2337
+ "step": 388
2338
+ },
2339
+ {
2340
+ "epoch": 3.85,
2341
+ "learning_rate": 2.65078143469413e-06,
2342
+ "loss": 0.8675,
2343
+ "step": 389
2344
+ },
2345
+ {
2346
+ "epoch": 3.86,
2347
+ "learning_rate": 2.6073653706293202e-06,
2348
+ "loss": 0.8377,
2349
+ "step": 390
2350
+ },
2351
+ {
2352
+ "epoch": 3.87,
2353
+ "learning_rate": 2.5642544329579088e-06,
2354
+ "loss": 0.8685,
2355
+ "step": 391
2356
+ },
2357
+ {
2358
+ "epoch": 3.88,
2359
+ "learning_rate": 2.5214504010570217e-06,
2360
+ "loss": 0.8838,
2361
+ "step": 392
2362
+ },
2363
+ {
2364
+ "epoch": 3.89,
2365
+ "learning_rate": 2.478955041636435e-06,
2366
+ "loss": 0.8548,
2367
+ "step": 393
2368
+ },
2369
+ {
2370
+ "epoch": 3.9,
2371
+ "learning_rate": 2.4367701086656625e-06,
2372
+ "loss": 0.9218,
2373
+ "step": 394
2374
+ },
2375
+ {
2376
+ "epoch": 3.91,
2377
+ "learning_rate": 2.3948973433015564e-06,
2378
+ "loss": 0.9963,
2379
+ "step": 395
2380
+ },
2381
+ {
2382
+ "epoch": 3.92,
2383
+ "learning_rate": 2.353338473816451e-06,
2384
+ "loss": 0.9655,
2385
+ "step": 396
2386
+ },
2387
+ {
2388
+ "epoch": 3.93,
2389
+ "learning_rate": 2.312095215526814e-06,
2390
+ "loss": 0.8899,
2391
+ "step": 397
2392
+ },
2393
+ {
2394
+ "epoch": 3.94,
2395
+ "learning_rate": 2.271169270722464e-06,
2396
+ "loss": 0.9778,
2397
+ "step": 398
2398
+ },
2399
+ {
2400
+ "epoch": 3.95,
2401
+ "learning_rate": 2.230562328596306e-06,
2402
+ "loss": 0.946,
2403
+ "step": 399
2404
+ },
2405
+ {
2406
+ "epoch": 3.96,
2407
+ "learning_rate": 2.190276065174596e-06,
2408
+ "loss": 0.9608,
2409
+ "step": 400
2410
+ },
2411
+ {
2412
+ "epoch": 3.97,
2413
+ "learning_rate": 2.1503121432477936e-06,
2414
+ "loss": 0.9379,
2415
+ "step": 401
2416
+ },
2417
+ {
2418
+ "epoch": 3.98,
2419
+ "learning_rate": 2.1106722123018965e-06,
2420
+ "loss": 1.0413,
2421
+ "step": 402
2422
+ },
2423
+ {
2424
+ "epoch": 3.99,
2425
+ "learning_rate": 2.0713579084503877e-06,
2426
+ "loss": 0.9567,
2427
+ "step": 403
2428
+ },
2429
+ {
2430
+ "epoch": 4.0,
2431
+ "learning_rate": 2.0323708543666888e-06,
2432
+ "loss": 0.8854,
2433
+ "step": 404
2434
+ },
2435
+ {
2436
+ "epoch": 4.01,
2437
+ "learning_rate": 1.993712659217194e-06,
2438
+ "loss": 0.952,
2439
+ "step": 405
2440
+ },
2441
+ {
2442
+ "epoch": 4.02,
2443
+ "learning_rate": 1.9553849185948514e-06,
2444
+ "loss": 1.0019,
2445
+ "step": 406
2446
+ },
2447
+ {
2448
+ "epoch": 4.03,
2449
+ "learning_rate": 1.9173892144532957e-06,
2450
+ "loss": 0.965,
2451
+ "step": 407
2452
+ },
2453
+ {
2454
+ "epoch": 4.04,
2455
+ "learning_rate": 1.8797271150415709e-06,
2456
+ "loss": 0.8878,
2457
+ "step": 408
2458
+ },
2459
+ {
2460
+ "epoch": 4.05,
2461
+ "learning_rate": 1.8424001748393905e-06,
2462
+ "loss": 0.9816,
2463
+ "step": 409
2464
+ },
2465
+ {
2466
+ "epoch": 4.06,
2467
+ "learning_rate": 1.8054099344929833e-06,
2468
+ "loss": 0.9631,
2469
+ "step": 410
2470
+ },
2471
+ {
2472
+ "epoch": 4.07,
2473
+ "learning_rate": 1.7687579207514893e-06,
2474
+ "loss": 0.8626,
2475
+ "step": 411
2476
+ },
2477
+ {
2478
+ "epoch": 4.08,
2479
+ "learning_rate": 1.7324456464039751e-06,
2480
+ "loss": 0.8982,
2481
+ "step": 412
2482
+ },
2483
+ {
2484
+ "epoch": 4.09,
2485
+ "learning_rate": 1.6964746102169582e-06,
2486
+ "loss": 0.9542,
2487
+ "step": 413
2488
+ },
2489
+ {
2490
+ "epoch": 4.1,
2491
+ "learning_rate": 1.6608462968725736e-06,
2492
+ "loss": 0.8928,
2493
+ "step": 414
2494
+ },
2495
+ {
2496
+ "epoch": 4.11,
2497
+ "learning_rate": 1.6255621769072805e-06,
2498
+ "loss": 1.0105,
2499
+ "step": 415
2500
+ },
2501
+ {
2502
+ "epoch": 4.12,
2503
+ "learning_rate": 1.5906237066511643e-06,
2504
+ "loss": 0.8821,
2505
+ "step": 416
2506
+ },
2507
+ {
2508
+ "epoch": 4.13,
2509
+ "learning_rate": 1.5560323281678514e-06,
2510
+ "loss": 0.8703,
2511
+ "step": 417
2512
+ },
2513
+ {
2514
+ "epoch": 4.14,
2515
+ "learning_rate": 1.521789469194952e-06,
2516
+ "loss": 0.8437,
2517
+ "step": 418
2518
+ },
2519
+ {
2520
+ "epoch": 4.15,
2521
+ "learning_rate": 1.4878965430851612e-06,
2522
+ "loss": 0.8663,
2523
+ "step": 419
2524
+ },
2525
+ {
2526
+ "epoch": 4.16,
2527
+ "learning_rate": 1.4543549487479092e-06,
2528
+ "loss": 0.9994,
2529
+ "step": 420
2530
+ },
2531
+ {
2532
+ "epoch": 4.17,
2533
+ "learning_rate": 1.4211660705916286e-06,
2534
+ "loss": 0.8591,
2535
+ "step": 421
2536
+ },
2537
+ {
2538
+ "epoch": 4.18,
2539
+ "learning_rate": 1.3883312784666091e-06,
2540
+ "loss": 0.8815,
2541
+ "step": 422
2542
+ },
2543
+ {
2544
+ "epoch": 4.19,
2545
+ "learning_rate": 1.3558519276084636e-06,
2546
+ "loss": 0.8615,
2547
+ "step": 423
2548
+ },
2549
+ {
2550
+ "epoch": 4.2,
2551
+ "learning_rate": 1.3237293585821786e-06,
2552
+ "loss": 0.8511,
2553
+ "step": 424
2554
+ },
2555
+ {
2556
+ "epoch": 4.21,
2557
+ "learning_rate": 1.291964897226803e-06,
2558
+ "loss": 0.8524,
2559
+ "step": 425
2560
+ },
2561
+ {
2562
+ "epoch": 4.22,
2563
+ "learning_rate": 1.260559854600709e-06,
2564
+ "loss": 0.8683,
2565
+ "step": 426
2566
+ },
2567
+ {
2568
+ "epoch": 4.23,
2569
+ "learning_rate": 1.2295155269274827e-06,
2570
+ "loss": 0.8299,
2571
+ "step": 427
2572
+ },
2573
+ {
2574
+ "epoch": 4.24,
2575
+ "learning_rate": 1.198833195542435e-06,
2576
+ "loss": 0.8564,
2577
+ "step": 428
2578
+ },
2579
+ {
2580
+ "epoch": 4.24,
2581
+ "learning_rate": 1.1685141268396906e-06,
2582
+ "loss": 0.7924,
2583
+ "step": 429
2584
+ },
2585
+ {
2586
+ "epoch": 4.25,
2587
+ "learning_rate": 1.1385595722199438e-06,
2588
+ "loss": 0.8414,
2589
+ "step": 430
2590
+ },
2591
+ {
2592
+ "epoch": 4.26,
2593
+ "learning_rate": 1.1089707680387962e-06,
2594
+ "loss": 0.7434,
2595
+ "step": 431
2596
+ },
2597
+ {
2598
+ "epoch": 4.27,
2599
+ "learning_rate": 1.079748935555719e-06,
2600
+ "loss": 0.7299,
2601
+ "step": 432
2602
+ },
2603
+ {
2604
+ "epoch": 4.28,
2605
+ "learning_rate": 1.0508952808836682e-06,
2606
+ "loss": 0.7797,
2607
+ "step": 433
2608
+ },
2609
+ {
2610
+ "epoch": 4.29,
2611
+ "learning_rate": 1.022410994939279e-06,
2612
+ "loss": 0.7147,
2613
+ "step": 434
2614
+ },
2615
+ {
2616
+ "epoch": 4.3,
2617
+ "learning_rate": 9.942972533937268e-07,
2618
+ "loss": 0.8071,
2619
+ "step": 435
2620
+ },
2621
+ {
2622
+ "epoch": 4.31,
2623
+ "learning_rate": 9.665552166241965e-07,
2624
+ "loss": 0.7704,
2625
+ "step": 436
2626
+ },
2627
+ {
2628
+ "epoch": 4.32,
2629
+ "learning_rate": 9.391860296659916e-07,
2630
+ "loss": 0.8572,
2631
+ "step": 437
2632
+ },
2633
+ {
2634
+ "epoch": 4.33,
2635
+ "learning_rate": 9.121908221652675e-07,
2636
+ "loss": 0.8579,
2637
+ "step": 438
2638
+ },
2639
+ {
2640
+ "epoch": 4.34,
2641
+ "learning_rate": 8.855707083324183e-07,
2642
+ "loss": 0.7123,
2643
+ "step": 439
2644
+ },
2645
+ {
2646
+ "epoch": 4.35,
2647
+ "learning_rate": 8.593267868960675e-07,
2648
+ "loss": 0.7775,
2649
+ "step": 440
2650
+ },
2651
+ {
2652
+ "epoch": 4.36,
2653
+ "learning_rate": 8.334601410577436e-07,
2654
+ "loss": 0.7357,
2655
+ "step": 441
2656
+ },
2657
+ {
2658
+ "epoch": 4.37,
2659
+ "learning_rate": 8.079718384471557e-07,
2660
+ "loss": 0.867,
2661
+ "step": 442
2662
+ },
2663
+ {
2664
+ "epoch": 4.38,
2665
+ "learning_rate": 7.828629310781266e-07,
2666
+ "loss": 0.821,
2667
+ "step": 443
2668
+ },
2669
+ {
2670
+ "epoch": 4.39,
2671
+ "learning_rate": 7.581344553051873e-07,
2672
+ "loss": 0.7604,
2673
+ "step": 444
2674
+ },
2675
+ {
2676
+ "epoch": 4.4,
2677
+ "learning_rate": 7.337874317807803e-07,
2678
+ "loss": 0.8181,
2679
+ "step": 445
2680
+ },
2681
+ {
2682
+ "epoch": 4.41,
2683
+ "learning_rate": 7.098228654131489e-07,
2684
+ "loss": 0.8672,
2685
+ "step": 446
2686
+ },
2687
+ {
2688
+ "epoch": 4.42,
2689
+ "learning_rate": 6.862417453248593e-07,
2690
+ "loss": 0.871,
2691
+ "step": 447
2692
+ },
2693
+ {
2694
+ "epoch": 4.43,
2695
+ "learning_rate": 6.630450448119618e-07,
2696
+ "loss": 0.8699,
2697
+ "step": 448
2698
+ },
2699
+ {
2700
+ "epoch": 4.44,
2701
+ "learning_rate": 6.402337213038379e-07,
2702
+ "loss": 0.9146,
2703
+ "step": 449
2704
+ },
2705
+ {
2706
+ "epoch": 4.45,
2707
+ "learning_rate": 6.178087163236645e-07,
2708
+ "loss": 0.8129,
2709
+ "step": 450
2710
+ },
2711
+ {
2712
+ "epoch": 4.46,
2713
+ "learning_rate": 5.957709554495683e-07,
2714
+ "loss": 0.7845,
2715
+ "step": 451
2716
+ },
2717
+ {
2718
+ "epoch": 4.47,
2719
+ "learning_rate": 5.741213482764118e-07,
2720
+ "loss": 0.9181,
2721
+ "step": 452
2722
+ },
2723
+ {
2724
+ "epoch": 4.48,
2725
+ "learning_rate": 5.528607883782599e-07,
2726
+ "loss": 0.8529,
2727
+ "step": 453
2728
+ },
2729
+ {
2730
+ "epoch": 4.49,
2731
+ "learning_rate": 5.319901532714877e-07,
2732
+ "loss": 0.8863,
2733
+ "step": 454
2734
+ },
2735
+ {
2736
+ "epoch": 4.5,
2737
+ "learning_rate": 5.115103043785718e-07,
2738
+ "loss": 0.7747,
2739
+ "step": 455
2740
+ },
2741
+ {
2742
+ "epoch": 4.51,
2743
+ "learning_rate": 4.91422086992529e-07,
2744
+ "loss": 0.8225,
2745
+ "step": 456
2746
+ },
2747
+ {
2748
+ "epoch": 4.52,
2749
+ "learning_rate": 4.717263302420283e-07,
2750
+ "loss": 0.8055,
2751
+ "step": 457
2752
+ },
2753
+ {
2754
+ "epoch": 4.53,
2755
+ "learning_rate": 4.5242384705717404e-07,
2756
+ "loss": 0.8073,
2757
+ "step": 458
2758
+ },
2759
+ {
2760
+ "epoch": 4.54,
2761
+ "learning_rate": 4.3351543413594263e-07,
2762
+ "loss": 0.9058,
2763
+ "step": 459
2764
+ },
2765
+ {
2766
+ "epoch": 4.55,
2767
+ "learning_rate": 4.150018719113147e-07,
2768
+ "loss": 0.8856,
2769
+ "step": 460
2770
+ },
2771
+ {
2772
+ "epoch": 4.56,
2773
+ "learning_rate": 3.968839245190448e-07,
2774
+ "loss": 0.8794,
2775
+ "step": 461
2776
+ },
2777
+ {
2778
+ "epoch": 4.57,
2779
+ "learning_rate": 3.7916233976613704e-07,
2780
+ "loss": 0.8192,
2781
+ "step": 462
2782
+ },
2783
+ {
2784
+ "epoch": 4.58,
2785
+ "learning_rate": 3.618378490999719e-07,
2786
+ "loss": 0.8839,
2787
+ "step": 463
2788
+ },
2789
+ {
2790
+ "epoch": 4.59,
2791
+ "learning_rate": 3.4491116757812024e-07,
2792
+ "loss": 0.8596,
2793
+ "step": 464
2794
+ },
2795
+ {
2796
+ "epoch": 4.6,
2797
+ "learning_rate": 3.283829938388294e-07,
2798
+ "loss": 0.8171,
2799
+ "step": 465
2800
+ },
2801
+ {
2802
+ "epoch": 4.61,
2803
+ "learning_rate": 3.122540100721794e-07,
2804
+ "loss": 0.8854,
2805
+ "step": 466
2806
+ },
2807
+ {
2808
+ "epoch": 4.62,
2809
+ "learning_rate": 2.965248819919397e-07,
2810
+ "loss": 0.8724,
2811
+ "step": 467
2812
+ },
2813
+ {
2814
+ "epoch": 4.63,
2815
+ "learning_rate": 2.8119625880808185e-07,
2816
+ "loss": 0.9394,
2817
+ "step": 468
2818
+ },
2819
+ {
2820
+ "epoch": 4.64,
2821
+ "learning_rate": 2.66268773199988e-07,
2822
+ "loss": 0.7804,
2823
+ "step": 469
2824
+ },
2825
+ {
2826
+ "epoch": 4.65,
2827
+ "learning_rate": 2.5174304129033655e-07,
2828
+ "loss": 0.7995,
2829
+ "step": 470
2830
+ },
2831
+ {
2832
+ "epoch": 4.66,
2833
+ "learning_rate": 2.3761966261967252e-07,
2834
+ "loss": 0.7713,
2835
+ "step": 471
2836
+ },
2837
+ {
2838
+ "epoch": 4.67,
2839
+ "learning_rate": 2.2389922012165944e-07,
2840
+ "loss": 0.8227,
2841
+ "step": 472
2842
+ },
2843
+ {
2844
+ "epoch": 4.68,
2845
+ "learning_rate": 2.1058228009902094e-07,
2846
+ "loss": 0.763,
2847
+ "step": 473
2848
+ },
2849
+ {
2850
+ "epoch": 4.69,
2851
+ "learning_rate": 1.9766939220017155e-07,
2852
+ "loss": 0.8846,
2853
+ "step": 474
2854
+ },
2855
+ {
2856
+ "epoch": 4.7,
2857
+ "learning_rate": 1.8516108939651945e-07,
2858
+ "loss": 0.8191,
2859
+ "step": 475
2860
+ },
2861
+ {
2862
+ "epoch": 4.71,
2863
+ "learning_rate": 1.7305788796048274e-07,
2864
+ "loss": 0.8646,
2865
+ "step": 476
2866
+ },
2867
+ {
2868
+ "epoch": 4.72,
2869
+ "learning_rate": 1.613602874441622e-07,
2870
+ "loss": 0.8464,
2871
+ "step": 477
2872
+ },
2873
+ {
2874
+ "epoch": 4.73,
2875
+ "learning_rate": 1.5006877065874338e-07,
2876
+ "loss": 0.8767,
2877
+ "step": 478
2878
+ },
2879
+ {
2880
+ "epoch": 4.74,
2881
+ "learning_rate": 1.3918380365455232e-07,
2882
+ "loss": 0.7908,
2883
+ "step": 479
2884
+ },
2885
+ {
2886
+ "epoch": 4.75,
2887
+ "learning_rate": 1.287058357018278e-07,
2888
+ "loss": 0.8039,
2889
+ "step": 480
2890
+ },
2891
+ {
2892
+ "epoch": 4.76,
2893
+ "learning_rate": 1.1863529927217731e-07,
2894
+ "loss": 0.7987,
2895
+ "step": 481
2896
+ },
2897
+ {
2898
+ "epoch": 4.77,
2899
+ "learning_rate": 1.0897261002072223e-07,
2900
+ "loss": 0.8231,
2901
+ "step": 482
2902
+ },
2903
+ {
2904
+ "epoch": 4.78,
2905
+ "learning_rate": 9.971816676894952e-08,
2906
+ "loss": 0.8304,
2907
+ "step": 483
2908
+ },
2909
+ {
2910
+ "epoch": 4.79,
2911
+ "learning_rate": 9.08723514882437e-08,
2912
+ "loss": 0.846,
2913
+ "step": 484
2914
+ },
2915
+ {
2916
+ "epoch": 4.8,
2917
+ "learning_rate": 8.243552928412501e-08,
2918
+ "loss": 0.8961,
2919
+ "step": 485
2920
+ },
2921
+ {
2922
+ "epoch": 4.81,
2923
+ "learning_rate": 7.440804838117932e-08,
2924
+ "loss": 0.8834,
2925
+ "step": 486
2926
+ },
2927
+ {
2928
+ "epoch": 4.82,
2929
+ "learning_rate": 6.679024010868617e-08,
2930
+ "loss": 0.8625,
2931
+ "step": 487
2932
+ },
2933
+ {
2934
+ "epoch": 4.83,
2935
+ "learning_rate": 5.958241888693872e-08,
2936
+ "loss": 0.8292,
2937
+ "step": 488
2938
+ },
2939
+ {
2940
+ "epoch": 4.84,
2941
+ "learning_rate": 5.2784882214274025e-08,
2942
+ "loss": 0.8513,
2943
+ "step": 489
2944
+ },
2945
+ {
2946
+ "epoch": 4.85,
2947
+ "learning_rate": 4.639791065478738e-08,
2948
+ "loss": 0.7609,
2949
+ "step": 490
2950
+ },
2951
+ {
2952
+ "epoch": 4.86,
2953
+ "learning_rate": 4.042176782675267e-08,
2954
+ "loss": 0.7706,
2955
+ "step": 491
2956
+ },
2957
+ {
2958
+ "epoch": 4.87,
2959
+ "learning_rate": 3.485670039174882e-08,
2960
+ "loss": 0.7774,
2961
+ "step": 492
2962
+ },
2963
+ {
2964
+ "epoch": 4.88,
2965
+ "learning_rate": 2.9702938044468e-08,
2966
+ "loss": 0.8261,
2967
+ "step": 493
2968
+ },
2969
+ {
2970
+ "epoch": 4.89,
2971
+ "learning_rate": 2.496069350324537e-08,
2972
+ "loss": 0.7775,
2973
+ "step": 494
2974
+ },
2975
+ {
2976
+ "epoch": 4.9,
2977
+ "learning_rate": 2.063016250127281e-08,
2978
+ "loss": 0.8115,
2979
+ "step": 495
2980
+ },
2981
+ {
2982
+ "epoch": 4.91,
2983
+ "learning_rate": 1.671152377852092e-08,
2984
+ "loss": 0.9538,
2985
+ "step": 496
2986
+ },
2987
+ {
2988
+ "epoch": 4.92,
2989
+ "learning_rate": 1.320493907436604e-08,
2990
+ "loss": 0.8701,
2991
+ "step": 497
2992
+ },
2993
+ {
2994
+ "epoch": 4.93,
2995
+ "learning_rate": 1.0110553120908917e-08,
2996
+ "loss": 0.8319,
2997
+ "step": 498
2998
+ },
2999
+ {
3000
+ "epoch": 4.94,
3001
+ "learning_rate": 7.42849363700282e-09,
3002
+ "loss": 0.8792,
3003
+ "step": 499
3004
+ },
3005
+ {
3006
+ "epoch": 4.95,
3007
+ "learning_rate": 5.158871322984426e-09,
3008
+ "loss": 0.8577,
3009
+ "step": 500
3010
+ },
3011
+ {
3012
+ "epoch": 4.96,
3013
+ "learning_rate": 3.3017798561030268e-09,
3014
+ "loss": 0.9151,
3015
+ "step": 501
3016
+ },
3017
+ {
3018
+ "epoch": 4.97,
3019
+ "learning_rate": 1.8572958866514e-09,
3020
+ "loss": 0.8622,
3021
+ "step": 502
3022
+ },
3023
+ {
3024
+ "epoch": 4.98,
3025
+ "learning_rate": 8.254790348072306e-10,
3026
+ "loss": 0.9606,
3027
+ "step": 503
3028
+ },
3029
+ {
3030
+ "epoch": 4.99,
3031
+ "learning_rate": 2.063718881695209e-10,
3032
+ "loss": 0.8872,
3033
+ "step": 504
3034
+ },
3035
+ {
3036
+ "epoch": 5.0,
3037
+ "learning_rate": 0.0,
3038
+ "loss": 0.819,
3039
+ "step": 505
3040
+ },
3041
+ {
3042
+ "epoch": 5.0,
3043
+ "step": 505,
3044
+ "total_flos": 2.6761493632096666e+18,
3045
+ "train_loss": 1.198445222283354,
3046
+ "train_runtime": 29037.4355,
3047
+ "train_samples_per_second": 2.227,
3048
+ "train_steps_per_second": 0.017
3049
+ }
3050
+ ],
3051
+ "logging_steps": 1.0,
3052
+ "max_steps": 505,
3053
+ "num_train_epochs": 5,
3054
+ "save_steps": 101,
3055
+ "total_flos": 2.6761493632096666e+18,
3056
+ "trial_name": null,
3057
+ "trial_params": null
3058
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:975ed3fe5462a4ce9f8cc80fd157ff577d42152921974ac7407b77bd839f7b9b
3
+ size 4664