Commit
•
91b112b
1
Parent(s):
541ca87
Upload 12 files
Browse files- .gitattributes +1 -0
- adapter_config.json +16 -0
- adapter_model.bin +3 -0
- config.json +25 -0
- databricks-dolly-15k-ja.json +3 -0
- generation_config.json +6 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +6 -0
- tokenizer.json +0 -0
- tokenizer_config.json +11 -0
- trainer_state.json +1488 -0
- training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ 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 |
+
databricks-dolly-15k-ja.json filter=lfs diff=lfs merge=lfs -text
|
adapter_config.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"base_model_name_or_path": "cyberagent/open-calm-large",
|
| 3 |
+
"bias": "none",
|
| 4 |
+
"fan_in_fan_out": false,
|
| 5 |
+
"inference_mode": true,
|
| 6 |
+
"init_lora_weights": true,
|
| 7 |
+
"lora_alpha": 16,
|
| 8 |
+
"lora_dropout": 0.05,
|
| 9 |
+
"modules_to_save": null,
|
| 10 |
+
"peft_type": "LORA",
|
| 11 |
+
"r": 8,
|
| 12 |
+
"target_modules": [
|
| 13 |
+
"query_key_value"
|
| 14 |
+
],
|
| 15 |
+
"task_type": "CAUSAL_LM"
|
| 16 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb82a0fc3d211e8b3b7793f1980299836eca50223f0b3da2705743fcc596dc6f
|
| 3 |
+
size 4736129
|
config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "open-calm-large",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTNeoXForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"eos_token_id": 0,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_size": 1536,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 6144,
|
| 12 |
+
"layer_norm_eps": 1e-05,
|
| 13 |
+
"max_position_embeddings": 2048,
|
| 14 |
+
"model_type": "gpt_neox",
|
| 15 |
+
"num_attention_heads": 16,
|
| 16 |
+
"num_hidden_layers": 24,
|
| 17 |
+
"rotary_emb_base": 10000,
|
| 18 |
+
"rotary_pct": 1.0,
|
| 19 |
+
"tie_word_embeddings": false,
|
| 20 |
+
"torch_dtype": "float16",
|
| 21 |
+
"transformers_version": "4.27.0.dev0",
|
| 22 |
+
"use_cache": true,
|
| 23 |
+
"use_parallel_residual": false,
|
| 24 |
+
"vocab_size": 52096
|
| 25 |
+
}
|
databricks-dolly-15k-ja.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4f814b77074a864ed15859b354470b10b8d970647165dc4e2ed08a6ff139417
|
| 3 |
+
size 17062018
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.27.0.dev0"
|
| 6 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7247fd55acc231c4b992aa616304bafa4f7d274596dcb407ce238b744240ab9
|
| 3 |
+
size 14575
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2cc30b9b2a57cf369f740406388258a373867af712078ebe300971a5df0c2944
|
| 3 |
+
size 627
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<|endoftext|>",
|
| 3 |
+
"eos_token": "<eos>",
|
| 4 |
+
"pad_token": "<|endoftext|>",
|
| 5 |
+
"unk_token": "<|endoftext|>"
|
| 6 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_eos_token": true,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"bos_token": "<|endoftext|>",
|
| 5 |
+
"clean_up_tokenization_spaces": true,
|
| 6 |
+
"eos_token": "<|endoftext|>",
|
| 7 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 8 |
+
"pad_token": "<|padding|>",
|
| 9 |
+
"tokenizer_class": "GPTNeoXTokenizer",
|
| 10 |
+
"unk_token": "<|endoftext|>"
|
| 11 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1488 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 4881,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 0.00029877074370006146,
|
| 13 |
+
"loss": 2.9685,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 0.00029754148740012294,
|
| 19 |
+
"loss": 2.2769,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 0.00029631223110018436,
|
| 25 |
+
"loss": 2.1295,
|
| 26 |
+
"step": 60
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 0.00029508297480024584,
|
| 31 |
+
"loss": 2.0832,
|
| 32 |
+
"step": 80
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 0.00029385371850030727,
|
| 37 |
+
"loss": 2.0765,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 0.00029262446220036875,
|
| 43 |
+
"loss": 2.0831,
|
| 44 |
+
"step": 120
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09,
|
| 48 |
+
"learning_rate": 0.00029139520590043023,
|
| 49 |
+
"loss": 2.1462,
|
| 50 |
+
"step": 140
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.1,
|
| 54 |
+
"learning_rate": 0.00029016594960049166,
|
| 55 |
+
"loss": 2.1126,
|
| 56 |
+
"step": 160
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.11,
|
| 60 |
+
"learning_rate": 0.00028893669330055314,
|
| 61 |
+
"loss": 2.0809,
|
| 62 |
+
"step": 180
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 0.0002877074370006146,
|
| 67 |
+
"loss": 2.1573,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.14,
|
| 72 |
+
"learning_rate": 0.00028647818070067605,
|
| 73 |
+
"loss": 2.1391,
|
| 74 |
+
"step": 220
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.15,
|
| 78 |
+
"learning_rate": 0.00028524892440073753,
|
| 79 |
+
"loss": 2.1265,
|
| 80 |
+
"step": 240
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.16,
|
| 84 |
+
"learning_rate": 0.00028401966810079896,
|
| 85 |
+
"loss": 2.0266,
|
| 86 |
+
"step": 260
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17,
|
| 90 |
+
"learning_rate": 0.00028279041180086044,
|
| 91 |
+
"loss": 2.0723,
|
| 92 |
+
"step": 280
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.18,
|
| 96 |
+
"learning_rate": 0.0002815611555009219,
|
| 97 |
+
"loss": 2.1082,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.2,
|
| 102 |
+
"learning_rate": 0.0002803318992009834,
|
| 103 |
+
"loss": 2.061,
|
| 104 |
+
"step": 320
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.21,
|
| 108 |
+
"learning_rate": 0.0002791026429010449,
|
| 109 |
+
"loss": 2.0605,
|
| 110 |
+
"step": 340
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.22,
|
| 114 |
+
"learning_rate": 0.0002778733866011063,
|
| 115 |
+
"loss": 2.0401,
|
| 116 |
+
"step": 360
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.23,
|
| 120 |
+
"learning_rate": 0.00027664413030116774,
|
| 121 |
+
"loss": 2.0441,
|
| 122 |
+
"step": 380
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.25,
|
| 126 |
+
"learning_rate": 0.0002754148740012292,
|
| 127 |
+
"loss": 2.0578,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.26,
|
| 132 |
+
"learning_rate": 0.0002741856177012907,
|
| 133 |
+
"loss": 2.0634,
|
| 134 |
+
"step": 420
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.27,
|
| 138 |
+
"learning_rate": 0.00027295636140135213,
|
| 139 |
+
"loss": 1.9869,
|
| 140 |
+
"step": 440
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.28,
|
| 144 |
+
"learning_rate": 0.0002717271051014136,
|
| 145 |
+
"loss": 2.0628,
|
| 146 |
+
"step": 460
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.3,
|
| 150 |
+
"learning_rate": 0.0002704978488014751,
|
| 151 |
+
"loss": 2.0732,
|
| 152 |
+
"step": 480
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.31,
|
| 156 |
+
"learning_rate": 0.0002692685925015366,
|
| 157 |
+
"loss": 2.038,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.32,
|
| 162 |
+
"learning_rate": 0.000268039336201598,
|
| 163 |
+
"loss": 2.0333,
|
| 164 |
+
"step": 520
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.33,
|
| 168 |
+
"learning_rate": 0.0002668100799016595,
|
| 169 |
+
"loss": 1.97,
|
| 170 |
+
"step": 540
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.34,
|
| 174 |
+
"learning_rate": 0.0002655808236017209,
|
| 175 |
+
"loss": 2.0097,
|
| 176 |
+
"step": 560
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.36,
|
| 180 |
+
"learning_rate": 0.0002643515673017824,
|
| 181 |
+
"loss": 1.9377,
|
| 182 |
+
"step": 580
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.37,
|
| 186 |
+
"learning_rate": 0.00026312231100184387,
|
| 187 |
+
"loss": 2.059,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.38,
|
| 192 |
+
"learning_rate": 0.00026189305470190535,
|
| 193 |
+
"loss": 1.9423,
|
| 194 |
+
"step": 620
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.39,
|
| 198 |
+
"learning_rate": 0.0002606637984019668,
|
| 199 |
+
"loss": 2.0869,
|
| 200 |
+
"step": 640
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.41,
|
| 204 |
+
"learning_rate": 0.00025943454210202826,
|
| 205 |
+
"loss": 1.9231,
|
| 206 |
+
"step": 660
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.42,
|
| 210 |
+
"learning_rate": 0.0002582052858020897,
|
| 211 |
+
"loss": 2.0086,
|
| 212 |
+
"step": 680
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.43,
|
| 216 |
+
"learning_rate": 0.00025697602950215117,
|
| 217 |
+
"loss": 1.9623,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.44,
|
| 222 |
+
"learning_rate": 0.00025574677320221265,
|
| 223 |
+
"loss": 1.9348,
|
| 224 |
+
"step": 720
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.45,
|
| 228 |
+
"learning_rate": 0.0002545175169022741,
|
| 229 |
+
"loss": 2.0155,
|
| 230 |
+
"step": 740
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.47,
|
| 234 |
+
"learning_rate": 0.00025328826060233556,
|
| 235 |
+
"loss": 1.9043,
|
| 236 |
+
"step": 760
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.48,
|
| 240 |
+
"learning_rate": 0.00025205900430239704,
|
| 241 |
+
"loss": 2.0102,
|
| 242 |
+
"step": 780
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.49,
|
| 246 |
+
"learning_rate": 0.0002508297480024585,
|
| 247 |
+
"loss": 1.9552,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5,
|
| 252 |
+
"learning_rate": 0.00024960049170251995,
|
| 253 |
+
"loss": 1.9779,
|
| 254 |
+
"step": 820
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.52,
|
| 258 |
+
"learning_rate": 0.0002483712354025814,
|
| 259 |
+
"loss": 1.981,
|
| 260 |
+
"step": 840
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.53,
|
| 264 |
+
"learning_rate": 0.00024714197910264286,
|
| 265 |
+
"loss": 1.9206,
|
| 266 |
+
"step": 860
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.54,
|
| 270 |
+
"learning_rate": 0.00024591272280270434,
|
| 271 |
+
"loss": 1.9573,
|
| 272 |
+
"step": 880
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.55,
|
| 276 |
+
"learning_rate": 0.0002446834665027658,
|
| 277 |
+
"loss": 2.0444,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.57,
|
| 282 |
+
"learning_rate": 0.00024345421020282728,
|
| 283 |
+
"loss": 1.8819,
|
| 284 |
+
"step": 920
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.58,
|
| 288 |
+
"learning_rate": 0.00024222495390288873,
|
| 289 |
+
"loss": 1.9977,
|
| 290 |
+
"step": 940
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.59,
|
| 294 |
+
"learning_rate": 0.00024099569760295018,
|
| 295 |
+
"loss": 1.946,
|
| 296 |
+
"step": 960
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6,
|
| 300 |
+
"learning_rate": 0.00023976644130301167,
|
| 301 |
+
"loss": 1.953,
|
| 302 |
+
"step": 980
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.61,
|
| 306 |
+
"learning_rate": 0.00023853718500307312,
|
| 307 |
+
"loss": 2.0032,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.63,
|
| 312 |
+
"learning_rate": 0.0002373079287031346,
|
| 313 |
+
"loss": 1.9271,
|
| 314 |
+
"step": 1020
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.64,
|
| 318 |
+
"learning_rate": 0.00023607867240319603,
|
| 319 |
+
"loss": 2.0297,
|
| 320 |
+
"step": 1040
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.65,
|
| 324 |
+
"learning_rate": 0.0002348494161032575,
|
| 325 |
+
"loss": 2.049,
|
| 326 |
+
"step": 1060
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.66,
|
| 330 |
+
"learning_rate": 0.00023362015980331896,
|
| 331 |
+
"loss": 1.8794,
|
| 332 |
+
"step": 1080
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.68,
|
| 336 |
+
"learning_rate": 0.00023239090350338044,
|
| 337 |
+
"loss": 1.9826,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.69,
|
| 342 |
+
"learning_rate": 0.00023116164720344193,
|
| 343 |
+
"loss": 2.0297,
|
| 344 |
+
"step": 1120
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.7,
|
| 348 |
+
"learning_rate": 0.00022993239090350335,
|
| 349 |
+
"loss": 1.9343,
|
| 350 |
+
"step": 1140
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.71,
|
| 354 |
+
"learning_rate": 0.0002287031346035648,
|
| 355 |
+
"loss": 1.956,
|
| 356 |
+
"step": 1160
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.73,
|
| 360 |
+
"learning_rate": 0.0002274738783036263,
|
| 361 |
+
"loss": 1.9653,
|
| 362 |
+
"step": 1180
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.74,
|
| 366 |
+
"learning_rate": 0.00022624462200368777,
|
| 367 |
+
"loss": 1.9609,
|
| 368 |
+
"step": 1200
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.75,
|
| 372 |
+
"learning_rate": 0.00022501536570374922,
|
| 373 |
+
"loss": 2.0,
|
| 374 |
+
"step": 1220
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.76,
|
| 378 |
+
"learning_rate": 0.00022378610940381065,
|
| 379 |
+
"loss": 2.0268,
|
| 380 |
+
"step": 1240
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.77,
|
| 384 |
+
"learning_rate": 0.00022255685310387213,
|
| 385 |
+
"loss": 2.0342,
|
| 386 |
+
"step": 1260
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.79,
|
| 390 |
+
"learning_rate": 0.00022132759680393361,
|
| 391 |
+
"loss": 1.9875,
|
| 392 |
+
"step": 1280
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.8,
|
| 396 |
+
"learning_rate": 0.00022009834050399507,
|
| 397 |
+
"loss": 1.9526,
|
| 398 |
+
"step": 1300
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.81,
|
| 402 |
+
"learning_rate": 0.00021886908420405652,
|
| 403 |
+
"loss": 1.9678,
|
| 404 |
+
"step": 1320
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.82,
|
| 408 |
+
"learning_rate": 0.00021763982790411798,
|
| 409 |
+
"loss": 2.1014,
|
| 410 |
+
"step": 1340
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.84,
|
| 414 |
+
"learning_rate": 0.00021641057160417946,
|
| 415 |
+
"loss": 1.9,
|
| 416 |
+
"step": 1360
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.85,
|
| 420 |
+
"learning_rate": 0.0002151813153042409,
|
| 421 |
+
"loss": 1.9648,
|
| 422 |
+
"step": 1380
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.86,
|
| 426 |
+
"learning_rate": 0.0002139520590043024,
|
| 427 |
+
"loss": 2.0172,
|
| 428 |
+
"step": 1400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.87,
|
| 432 |
+
"learning_rate": 0.00021272280270436382,
|
| 433 |
+
"loss": 1.9327,
|
| 434 |
+
"step": 1420
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.89,
|
| 438 |
+
"learning_rate": 0.0002114935464044253,
|
| 439 |
+
"loss": 1.9365,
|
| 440 |
+
"step": 1440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.9,
|
| 444 |
+
"learning_rate": 0.00021026429010448676,
|
| 445 |
+
"loss": 1.9678,
|
| 446 |
+
"step": 1460
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.91,
|
| 450 |
+
"learning_rate": 0.00020903503380454824,
|
| 451 |
+
"loss": 1.9248,
|
| 452 |
+
"step": 1480
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.92,
|
| 456 |
+
"learning_rate": 0.0002078057775046097,
|
| 457 |
+
"loss": 1.9312,
|
| 458 |
+
"step": 1500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.93,
|
| 462 |
+
"learning_rate": 0.00020657652120467115,
|
| 463 |
+
"loss": 1.9107,
|
| 464 |
+
"step": 1520
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.95,
|
| 468 |
+
"learning_rate": 0.0002053472649047326,
|
| 469 |
+
"loss": 1.9331,
|
| 470 |
+
"step": 1540
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.96,
|
| 474 |
+
"learning_rate": 0.00020411800860479408,
|
| 475 |
+
"loss": 1.9028,
|
| 476 |
+
"step": 1560
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.97,
|
| 480 |
+
"learning_rate": 0.00020288875230485556,
|
| 481 |
+
"loss": 1.9537,
|
| 482 |
+
"step": 1580
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.98,
|
| 486 |
+
"learning_rate": 0.00020165949600491702,
|
| 487 |
+
"loss": 1.9189,
|
| 488 |
+
"step": 1600
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.0,
|
| 492 |
+
"learning_rate": 0.00020043023970497845,
|
| 493 |
+
"loss": 1.9677,
|
| 494 |
+
"step": 1620
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.01,
|
| 498 |
+
"learning_rate": 0.00019920098340503993,
|
| 499 |
+
"loss": 1.8744,
|
| 500 |
+
"step": 1640
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.02,
|
| 504 |
+
"learning_rate": 0.0001979717271051014,
|
| 505 |
+
"loss": 1.9705,
|
| 506 |
+
"step": 1660
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.03,
|
| 510 |
+
"learning_rate": 0.00019674247080516286,
|
| 511 |
+
"loss": 1.9018,
|
| 512 |
+
"step": 1680
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.04,
|
| 516 |
+
"learning_rate": 0.00019551321450522434,
|
| 517 |
+
"loss": 1.9697,
|
| 518 |
+
"step": 1700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.06,
|
| 522 |
+
"learning_rate": 0.00019428395820528577,
|
| 523 |
+
"loss": 1.9494,
|
| 524 |
+
"step": 1720
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.07,
|
| 528 |
+
"learning_rate": 0.00019305470190534725,
|
| 529 |
+
"loss": 1.9556,
|
| 530 |
+
"step": 1740
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.08,
|
| 534 |
+
"learning_rate": 0.0001918254456054087,
|
| 535 |
+
"loss": 2.0141,
|
| 536 |
+
"step": 1760
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.09,
|
| 540 |
+
"learning_rate": 0.0001905961893054702,
|
| 541 |
+
"loss": 1.9397,
|
| 542 |
+
"step": 1780
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.11,
|
| 546 |
+
"learning_rate": 0.00018936693300553164,
|
| 547 |
+
"loss": 1.9258,
|
| 548 |
+
"step": 1800
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.12,
|
| 552 |
+
"learning_rate": 0.0001881376767055931,
|
| 553 |
+
"loss": 1.8207,
|
| 554 |
+
"step": 1820
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.13,
|
| 558 |
+
"learning_rate": 0.00018690842040565455,
|
| 559 |
+
"loss": 1.9226,
|
| 560 |
+
"step": 1840
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.14,
|
| 564 |
+
"learning_rate": 0.00018567916410571603,
|
| 565 |
+
"loss": 1.9205,
|
| 566 |
+
"step": 1860
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.16,
|
| 570 |
+
"learning_rate": 0.00018444990780577749,
|
| 571 |
+
"loss": 1.9983,
|
| 572 |
+
"step": 1880
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.17,
|
| 576 |
+
"learning_rate": 0.00018322065150583897,
|
| 577 |
+
"loss": 1.9727,
|
| 578 |
+
"step": 1900
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.18,
|
| 582 |
+
"learning_rate": 0.0001819913952059004,
|
| 583 |
+
"loss": 1.9329,
|
| 584 |
+
"step": 1920
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.19,
|
| 588 |
+
"learning_rate": 0.00018076213890596188,
|
| 589 |
+
"loss": 2.0107,
|
| 590 |
+
"step": 1940
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.2,
|
| 594 |
+
"learning_rate": 0.00017953288260602333,
|
| 595 |
+
"loss": 1.9328,
|
| 596 |
+
"step": 1960
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.22,
|
| 600 |
+
"learning_rate": 0.0001783036263060848,
|
| 601 |
+
"loss": 1.8733,
|
| 602 |
+
"step": 1980
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.23,
|
| 606 |
+
"learning_rate": 0.0001770743700061463,
|
| 607 |
+
"loss": 1.9073,
|
| 608 |
+
"step": 2000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.24,
|
| 612 |
+
"learning_rate": 0.00017584511370620772,
|
| 613 |
+
"loss": 1.9431,
|
| 614 |
+
"step": 2020
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.25,
|
| 618 |
+
"learning_rate": 0.0001746158574062692,
|
| 619 |
+
"loss": 1.8844,
|
| 620 |
+
"step": 2040
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.27,
|
| 624 |
+
"learning_rate": 0.00017338660110633066,
|
| 625 |
+
"loss": 1.8902,
|
| 626 |
+
"step": 2060
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.28,
|
| 630 |
+
"learning_rate": 0.00017215734480639214,
|
| 631 |
+
"loss": 1.8887,
|
| 632 |
+
"step": 2080
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.29,
|
| 636 |
+
"learning_rate": 0.00017092808850645356,
|
| 637 |
+
"loss": 1.8653,
|
| 638 |
+
"step": 2100
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.3,
|
| 642 |
+
"learning_rate": 0.00016969883220651505,
|
| 643 |
+
"loss": 1.9092,
|
| 644 |
+
"step": 2120
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.32,
|
| 648 |
+
"learning_rate": 0.0001684695759065765,
|
| 649 |
+
"loss": 1.9427,
|
| 650 |
+
"step": 2140
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.33,
|
| 654 |
+
"learning_rate": 0.00016724031960663798,
|
| 655 |
+
"loss": 2.0016,
|
| 656 |
+
"step": 2160
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.34,
|
| 660 |
+
"learning_rate": 0.00016601106330669943,
|
| 661 |
+
"loss": 1.9192,
|
| 662 |
+
"step": 2180
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.35,
|
| 666 |
+
"learning_rate": 0.0001647818070067609,
|
| 667 |
+
"loss": 1.8911,
|
| 668 |
+
"step": 2200
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.36,
|
| 672 |
+
"learning_rate": 0.00016355255070682234,
|
| 673 |
+
"loss": 1.9627,
|
| 674 |
+
"step": 2220
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.38,
|
| 678 |
+
"learning_rate": 0.00016232329440688382,
|
| 679 |
+
"loss": 2.022,
|
| 680 |
+
"step": 2240
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.39,
|
| 684 |
+
"learning_rate": 0.00016109403810694528,
|
| 685 |
+
"loss": 1.8829,
|
| 686 |
+
"step": 2260
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 1.4,
|
| 690 |
+
"learning_rate": 0.00015986478180700676,
|
| 691 |
+
"loss": 2.0244,
|
| 692 |
+
"step": 2280
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.41,
|
| 696 |
+
"learning_rate": 0.0001586355255070682,
|
| 697 |
+
"loss": 1.9201,
|
| 698 |
+
"step": 2300
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.43,
|
| 702 |
+
"learning_rate": 0.00015740626920712967,
|
| 703 |
+
"loss": 1.9803,
|
| 704 |
+
"step": 2320
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.44,
|
| 708 |
+
"learning_rate": 0.00015617701290719112,
|
| 709 |
+
"loss": 1.9369,
|
| 710 |
+
"step": 2340
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.45,
|
| 714 |
+
"learning_rate": 0.0001549477566072526,
|
| 715 |
+
"loss": 1.9151,
|
| 716 |
+
"step": 2360
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.46,
|
| 720 |
+
"learning_rate": 0.00015371850030731409,
|
| 721 |
+
"loss": 1.8436,
|
| 722 |
+
"step": 2380
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 1.48,
|
| 726 |
+
"learning_rate": 0.0001524892440073755,
|
| 727 |
+
"loss": 1.9259,
|
| 728 |
+
"step": 2400
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.49,
|
| 732 |
+
"learning_rate": 0.00015125998770743697,
|
| 733 |
+
"loss": 1.8486,
|
| 734 |
+
"step": 2420
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.5,
|
| 738 |
+
"learning_rate": 0.00015003073140749845,
|
| 739 |
+
"loss": 1.8839,
|
| 740 |
+
"step": 2440
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.51,
|
| 744 |
+
"learning_rate": 0.00014880147510755993,
|
| 745 |
+
"loss": 1.917,
|
| 746 |
+
"step": 2460
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.52,
|
| 750 |
+
"learning_rate": 0.00014757221880762138,
|
| 751 |
+
"loss": 1.9474,
|
| 752 |
+
"step": 2480
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.54,
|
| 756 |
+
"learning_rate": 0.00014634296250768284,
|
| 757 |
+
"loss": 1.9225,
|
| 758 |
+
"step": 2500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.55,
|
| 762 |
+
"learning_rate": 0.0001451137062077443,
|
| 763 |
+
"loss": 1.9257,
|
| 764 |
+
"step": 2520
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.56,
|
| 768 |
+
"learning_rate": 0.00014388444990780577,
|
| 769 |
+
"loss": 1.9105,
|
| 770 |
+
"step": 2540
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.57,
|
| 774 |
+
"learning_rate": 0.00014265519360786723,
|
| 775 |
+
"loss": 1.9799,
|
| 776 |
+
"step": 2560
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.59,
|
| 780 |
+
"learning_rate": 0.00014142593730792868,
|
| 781 |
+
"loss": 1.9632,
|
| 782 |
+
"step": 2580
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.6,
|
| 786 |
+
"learning_rate": 0.00014019668100799016,
|
| 787 |
+
"loss": 1.888,
|
| 788 |
+
"step": 2600
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.61,
|
| 792 |
+
"learning_rate": 0.00013896742470805162,
|
| 793 |
+
"loss": 1.9576,
|
| 794 |
+
"step": 2620
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.62,
|
| 798 |
+
"learning_rate": 0.00013773816840811307,
|
| 799 |
+
"loss": 2.0318,
|
| 800 |
+
"step": 2640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.63,
|
| 804 |
+
"learning_rate": 0.00013650891210817455,
|
| 805 |
+
"loss": 1.9347,
|
| 806 |
+
"step": 2660
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.65,
|
| 810 |
+
"learning_rate": 0.000135279655808236,
|
| 811 |
+
"loss": 1.8473,
|
| 812 |
+
"step": 2680
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.66,
|
| 816 |
+
"learning_rate": 0.00013405039950829746,
|
| 817 |
+
"loss": 1.9583,
|
| 818 |
+
"step": 2700
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.67,
|
| 822 |
+
"learning_rate": 0.00013282114320835892,
|
| 823 |
+
"loss": 1.8072,
|
| 824 |
+
"step": 2720
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.68,
|
| 828 |
+
"learning_rate": 0.0001315918869084204,
|
| 829 |
+
"loss": 1.8624,
|
| 830 |
+
"step": 2740
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.7,
|
| 834 |
+
"learning_rate": 0.00013036263060848185,
|
| 835 |
+
"loss": 1.9646,
|
| 836 |
+
"step": 2760
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.71,
|
| 840 |
+
"learning_rate": 0.0001291333743085433,
|
| 841 |
+
"loss": 1.9538,
|
| 842 |
+
"step": 2780
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.72,
|
| 846 |
+
"learning_rate": 0.0001279041180086048,
|
| 847 |
+
"loss": 1.8872,
|
| 848 |
+
"step": 2800
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.73,
|
| 852 |
+
"learning_rate": 0.00012667486170866624,
|
| 853 |
+
"loss": 1.9137,
|
| 854 |
+
"step": 2820
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.75,
|
| 858 |
+
"learning_rate": 0.00012544560540872772,
|
| 859 |
+
"loss": 1.9049,
|
| 860 |
+
"step": 2840
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.76,
|
| 864 |
+
"learning_rate": 0.00012421634910878918,
|
| 865 |
+
"loss": 1.9567,
|
| 866 |
+
"step": 2860
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.77,
|
| 870 |
+
"learning_rate": 0.00012298709280885063,
|
| 871 |
+
"loss": 1.9053,
|
| 872 |
+
"step": 2880
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.78,
|
| 876 |
+
"learning_rate": 0.0001217578365089121,
|
| 877 |
+
"loss": 1.8927,
|
| 878 |
+
"step": 2900
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.79,
|
| 882 |
+
"learning_rate": 0.00012052858020897355,
|
| 883 |
+
"loss": 1.9475,
|
| 884 |
+
"step": 2920
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.81,
|
| 888 |
+
"learning_rate": 0.00011929932390903502,
|
| 889 |
+
"loss": 2.0315,
|
| 890 |
+
"step": 2940
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.82,
|
| 894 |
+
"learning_rate": 0.00011807006760909648,
|
| 895 |
+
"loss": 1.8144,
|
| 896 |
+
"step": 2960
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.83,
|
| 900 |
+
"learning_rate": 0.00011684081130915794,
|
| 901 |
+
"loss": 1.8562,
|
| 902 |
+
"step": 2980
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.84,
|
| 906 |
+
"learning_rate": 0.00011561155500921942,
|
| 907 |
+
"loss": 1.925,
|
| 908 |
+
"step": 3000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.86,
|
| 912 |
+
"learning_rate": 0.00011438229870928088,
|
| 913 |
+
"loss": 1.9984,
|
| 914 |
+
"step": 3020
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.87,
|
| 918 |
+
"learning_rate": 0.00011315304240934235,
|
| 919 |
+
"loss": 1.9431,
|
| 920 |
+
"step": 3040
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.88,
|
| 924 |
+
"learning_rate": 0.0001119237861094038,
|
| 925 |
+
"loss": 1.9025,
|
| 926 |
+
"step": 3060
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.89,
|
| 930 |
+
"learning_rate": 0.00011069452980946527,
|
| 931 |
+
"loss": 1.9075,
|
| 932 |
+
"step": 3080
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 1.91,
|
| 936 |
+
"learning_rate": 0.00010946527350952672,
|
| 937 |
+
"loss": 1.9387,
|
| 938 |
+
"step": 3100
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 1.92,
|
| 942 |
+
"learning_rate": 0.00010823601720958819,
|
| 943 |
+
"loss": 1.8919,
|
| 944 |
+
"step": 3120
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.93,
|
| 948 |
+
"learning_rate": 0.00010700676090964966,
|
| 949 |
+
"loss": 1.9138,
|
| 950 |
+
"step": 3140
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.94,
|
| 954 |
+
"learning_rate": 0.00010577750460971111,
|
| 955 |
+
"loss": 1.9591,
|
| 956 |
+
"step": 3160
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.95,
|
| 960 |
+
"learning_rate": 0.00010454824830977258,
|
| 961 |
+
"loss": 1.9794,
|
| 962 |
+
"step": 3180
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.97,
|
| 966 |
+
"learning_rate": 0.00010331899200983403,
|
| 967 |
+
"loss": 1.9314,
|
| 968 |
+
"step": 3200
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.98,
|
| 972 |
+
"learning_rate": 0.0001020897357098955,
|
| 973 |
+
"loss": 1.9484,
|
| 974 |
+
"step": 3220
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.99,
|
| 978 |
+
"learning_rate": 0.00010086047940995697,
|
| 979 |
+
"loss": 1.839,
|
| 980 |
+
"step": 3240
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.0,
|
| 984 |
+
"learning_rate": 9.963122311001842e-05,
|
| 985 |
+
"loss": 1.8986,
|
| 986 |
+
"step": 3260
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.02,
|
| 990 |
+
"learning_rate": 9.840196681007989e-05,
|
| 991 |
+
"loss": 1.9466,
|
| 992 |
+
"step": 3280
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.03,
|
| 996 |
+
"learning_rate": 9.717271051014135e-05,
|
| 997 |
+
"loss": 1.9425,
|
| 998 |
+
"step": 3300
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.04,
|
| 1002 |
+
"learning_rate": 9.594345421020281e-05,
|
| 1003 |
+
"loss": 1.8959,
|
| 1004 |
+
"step": 3320
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.05,
|
| 1008 |
+
"learning_rate": 9.471419791026428e-05,
|
| 1009 |
+
"loss": 1.8946,
|
| 1010 |
+
"step": 3340
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.07,
|
| 1014 |
+
"learning_rate": 9.348494161032574e-05,
|
| 1015 |
+
"loss": 1.974,
|
| 1016 |
+
"step": 3360
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.08,
|
| 1020 |
+
"learning_rate": 9.225568531038722e-05,
|
| 1021 |
+
"loss": 1.8717,
|
| 1022 |
+
"step": 3380
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 2.09,
|
| 1026 |
+
"learning_rate": 9.102642901044866e-05,
|
| 1027 |
+
"loss": 1.9481,
|
| 1028 |
+
"step": 3400
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.1,
|
| 1032 |
+
"learning_rate": 8.979717271051014e-05,
|
| 1033 |
+
"loss": 1.931,
|
| 1034 |
+
"step": 3420
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.11,
|
| 1038 |
+
"learning_rate": 8.856791641057161e-05,
|
| 1039 |
+
"loss": 1.9382,
|
| 1040 |
+
"step": 3440
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.13,
|
| 1044 |
+
"learning_rate": 8.733866011063306e-05,
|
| 1045 |
+
"loss": 1.9577,
|
| 1046 |
+
"step": 3460
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.14,
|
| 1050 |
+
"learning_rate": 8.610940381069453e-05,
|
| 1051 |
+
"loss": 1.8826,
|
| 1052 |
+
"step": 3480
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.15,
|
| 1056 |
+
"learning_rate": 8.488014751075598e-05,
|
| 1057 |
+
"loss": 1.902,
|
| 1058 |
+
"step": 3500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.16,
|
| 1062 |
+
"learning_rate": 8.365089121081745e-05,
|
| 1063 |
+
"loss": 1.9147,
|
| 1064 |
+
"step": 3520
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.18,
|
| 1068 |
+
"learning_rate": 8.24216349108789e-05,
|
| 1069 |
+
"loss": 1.8675,
|
| 1070 |
+
"step": 3540
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.19,
|
| 1074 |
+
"learning_rate": 8.119237861094037e-05,
|
| 1075 |
+
"loss": 1.877,
|
| 1076 |
+
"step": 3560
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.2,
|
| 1080 |
+
"learning_rate": 7.996312231100184e-05,
|
| 1081 |
+
"loss": 2.0258,
|
| 1082 |
+
"step": 3580
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.21,
|
| 1086 |
+
"learning_rate": 7.87338660110633e-05,
|
| 1087 |
+
"loss": 1.9204,
|
| 1088 |
+
"step": 3600
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.22,
|
| 1092 |
+
"learning_rate": 7.750460971112476e-05,
|
| 1093 |
+
"loss": 1.9073,
|
| 1094 |
+
"step": 3620
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.24,
|
| 1098 |
+
"learning_rate": 7.627535341118622e-05,
|
| 1099 |
+
"loss": 1.8481,
|
| 1100 |
+
"step": 3640
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.25,
|
| 1104 |
+
"learning_rate": 7.504609711124769e-05,
|
| 1105 |
+
"loss": 1.896,
|
| 1106 |
+
"step": 3660
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.26,
|
| 1110 |
+
"learning_rate": 7.381684081130915e-05,
|
| 1111 |
+
"loss": 1.9285,
|
| 1112 |
+
"step": 3680
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.27,
|
| 1116 |
+
"learning_rate": 7.258758451137061e-05,
|
| 1117 |
+
"loss": 1.8239,
|
| 1118 |
+
"step": 3700
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.29,
|
| 1122 |
+
"learning_rate": 7.135832821143208e-05,
|
| 1123 |
+
"loss": 1.9289,
|
| 1124 |
+
"step": 3720
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.3,
|
| 1128 |
+
"learning_rate": 7.012907191149354e-05,
|
| 1129 |
+
"loss": 1.8431,
|
| 1130 |
+
"step": 3740
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.31,
|
| 1134 |
+
"learning_rate": 6.8899815611555e-05,
|
| 1135 |
+
"loss": 1.9239,
|
| 1136 |
+
"step": 3760
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.32,
|
| 1140 |
+
"learning_rate": 6.767055931161647e-05,
|
| 1141 |
+
"loss": 1.9132,
|
| 1142 |
+
"step": 3780
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 2.34,
|
| 1146 |
+
"learning_rate": 6.644130301167793e-05,
|
| 1147 |
+
"loss": 1.9019,
|
| 1148 |
+
"step": 3800
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.35,
|
| 1152 |
+
"learning_rate": 6.521204671173939e-05,
|
| 1153 |
+
"loss": 1.9947,
|
| 1154 |
+
"step": 3820
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.36,
|
| 1158 |
+
"learning_rate": 6.398279041180086e-05,
|
| 1159 |
+
"loss": 1.9146,
|
| 1160 |
+
"step": 3840
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.37,
|
| 1164 |
+
"learning_rate": 6.275353411186232e-05,
|
| 1165 |
+
"loss": 1.8672,
|
| 1166 |
+
"step": 3860
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.38,
|
| 1170 |
+
"learning_rate": 6.152427781192378e-05,
|
| 1171 |
+
"loss": 1.9189,
|
| 1172 |
+
"step": 3880
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.4,
|
| 1176 |
+
"learning_rate": 6.0295021511985245e-05,
|
| 1177 |
+
"loss": 1.8431,
|
| 1178 |
+
"step": 3900
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.41,
|
| 1182 |
+
"learning_rate": 5.9065765212046706e-05,
|
| 1183 |
+
"loss": 1.9629,
|
| 1184 |
+
"step": 3920
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 2.42,
|
| 1188 |
+
"learning_rate": 5.7836508912108174e-05,
|
| 1189 |
+
"loss": 1.8209,
|
| 1190 |
+
"step": 3940
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.43,
|
| 1194 |
+
"learning_rate": 5.6607252612169635e-05,
|
| 1195 |
+
"loss": 1.9104,
|
| 1196 |
+
"step": 3960
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.45,
|
| 1200 |
+
"learning_rate": 5.5377996312231096e-05,
|
| 1201 |
+
"loss": 1.8563,
|
| 1202 |
+
"step": 3980
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.46,
|
| 1206 |
+
"learning_rate": 5.414874001229256e-05,
|
| 1207 |
+
"loss": 1.9091,
|
| 1208 |
+
"step": 4000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.47,
|
| 1212 |
+
"learning_rate": 5.291948371235402e-05,
|
| 1213 |
+
"loss": 1.9038,
|
| 1214 |
+
"step": 4020
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.48,
|
| 1218 |
+
"learning_rate": 5.169022741241548e-05,
|
| 1219 |
+
"loss": 1.9083,
|
| 1220 |
+
"step": 4040
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.5,
|
| 1224 |
+
"learning_rate": 5.046097111247695e-05,
|
| 1225 |
+
"loss": 1.9037,
|
| 1226 |
+
"step": 4060
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.51,
|
| 1230 |
+
"learning_rate": 4.9231714812538415e-05,
|
| 1231 |
+
"loss": 1.8808,
|
| 1232 |
+
"step": 4080
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.52,
|
| 1236 |
+
"learning_rate": 4.8002458512599876e-05,
|
| 1237 |
+
"loss": 1.9638,
|
| 1238 |
+
"step": 4100
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.53,
|
| 1242 |
+
"learning_rate": 4.677320221266134e-05,
|
| 1243 |
+
"loss": 1.9124,
|
| 1244 |
+
"step": 4120
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.54,
|
| 1248 |
+
"learning_rate": 4.55439459127228e-05,
|
| 1249 |
+
"loss": 1.9119,
|
| 1250 |
+
"step": 4140
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.56,
|
| 1254 |
+
"learning_rate": 4.4314689612784266e-05,
|
| 1255 |
+
"loss": 1.8596,
|
| 1256 |
+
"step": 4160
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.57,
|
| 1260 |
+
"learning_rate": 4.3085433312845727e-05,
|
| 1261 |
+
"loss": 1.8779,
|
| 1262 |
+
"step": 4180
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.58,
|
| 1266 |
+
"learning_rate": 4.185617701290719e-05,
|
| 1267 |
+
"loss": 1.9442,
|
| 1268 |
+
"step": 4200
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.59,
|
| 1272 |
+
"learning_rate": 4.062692071296865e-05,
|
| 1273 |
+
"loss": 1.8899,
|
| 1274 |
+
"step": 4220
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.61,
|
| 1278 |
+
"learning_rate": 3.939766441303011e-05,
|
| 1279 |
+
"loss": 1.8925,
|
| 1280 |
+
"step": 4240
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.62,
|
| 1284 |
+
"learning_rate": 3.816840811309157e-05,
|
| 1285 |
+
"loss": 1.9306,
|
| 1286 |
+
"step": 4260
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.63,
|
| 1290 |
+
"learning_rate": 3.693915181315304e-05,
|
| 1291 |
+
"loss": 1.988,
|
| 1292 |
+
"step": 4280
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 2.64,
|
| 1296 |
+
"learning_rate": 3.5709895513214506e-05,
|
| 1297 |
+
"loss": 1.8516,
|
| 1298 |
+
"step": 4300
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.66,
|
| 1302 |
+
"learning_rate": 3.448063921327597e-05,
|
| 1303 |
+
"loss": 1.9874,
|
| 1304 |
+
"step": 4320
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.67,
|
| 1308 |
+
"learning_rate": 3.325138291333743e-05,
|
| 1309 |
+
"loss": 1.9059,
|
| 1310 |
+
"step": 4340
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.68,
|
| 1314 |
+
"learning_rate": 3.202212661339889e-05,
|
| 1315 |
+
"loss": 1.9219,
|
| 1316 |
+
"step": 4360
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.69,
|
| 1320 |
+
"learning_rate": 3.079287031346035e-05,
|
| 1321 |
+
"loss": 1.8807,
|
| 1322 |
+
"step": 4380
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.7,
|
| 1326 |
+
"learning_rate": 2.9563614013521815e-05,
|
| 1327 |
+
"loss": 1.842,
|
| 1328 |
+
"step": 4400
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.72,
|
| 1332 |
+
"learning_rate": 2.833435771358328e-05,
|
| 1333 |
+
"loss": 1.9281,
|
| 1334 |
+
"step": 4420
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 2.73,
|
| 1338 |
+
"learning_rate": 2.7105101413644744e-05,
|
| 1339 |
+
"loss": 1.8918,
|
| 1340 |
+
"step": 4440
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.74,
|
| 1344 |
+
"learning_rate": 2.5875845113706208e-05,
|
| 1345 |
+
"loss": 1.8154,
|
| 1346 |
+
"step": 4460
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.75,
|
| 1350 |
+
"learning_rate": 2.464658881376767e-05,
|
| 1351 |
+
"loss": 1.8775,
|
| 1352 |
+
"step": 4480
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 2.77,
|
| 1356 |
+
"learning_rate": 2.341733251382913e-05,
|
| 1357 |
+
"loss": 1.9259,
|
| 1358 |
+
"step": 4500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 2.78,
|
| 1362 |
+
"learning_rate": 2.2188076213890594e-05,
|
| 1363 |
+
"loss": 1.8731,
|
| 1364 |
+
"step": 4520
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.79,
|
| 1368 |
+
"learning_rate": 2.095881991395206e-05,
|
| 1369 |
+
"loss": 1.916,
|
| 1370 |
+
"step": 4540
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 2.8,
|
| 1374 |
+
"learning_rate": 1.972956361401352e-05,
|
| 1375 |
+
"loss": 1.8511,
|
| 1376 |
+
"step": 4560
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 2.81,
|
| 1380 |
+
"learning_rate": 1.850030731407498e-05,
|
| 1381 |
+
"loss": 1.8983,
|
| 1382 |
+
"step": 4580
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.83,
|
| 1386 |
+
"learning_rate": 1.7271051014136445e-05,
|
| 1387 |
+
"loss": 1.8668,
|
| 1388 |
+
"step": 4600
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.84,
|
| 1392 |
+
"learning_rate": 1.604179471419791e-05,
|
| 1393 |
+
"loss": 1.8947,
|
| 1394 |
+
"step": 4620
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 2.85,
|
| 1398 |
+
"learning_rate": 1.4812538414259372e-05,
|
| 1399 |
+
"loss": 1.9118,
|
| 1400 |
+
"step": 4640
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 2.86,
|
| 1404 |
+
"learning_rate": 1.3583282114320835e-05,
|
| 1405 |
+
"loss": 1.9495,
|
| 1406 |
+
"step": 4660
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.88,
|
| 1410 |
+
"learning_rate": 1.2354025814382298e-05,
|
| 1411 |
+
"loss": 1.9175,
|
| 1412 |
+
"step": 4680
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 2.89,
|
| 1416 |
+
"learning_rate": 1.112476951444376e-05,
|
| 1417 |
+
"loss": 1.9496,
|
| 1418 |
+
"step": 4700
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 2.9,
|
| 1422 |
+
"learning_rate": 9.895513214505223e-06,
|
| 1423 |
+
"loss": 1.9861,
|
| 1424 |
+
"step": 4720
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.91,
|
| 1428 |
+
"learning_rate": 8.666256914566686e-06,
|
| 1429 |
+
"loss": 1.8718,
|
| 1430 |
+
"step": 4740
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.93,
|
| 1434 |
+
"learning_rate": 7.4370006146281495e-06,
|
| 1435 |
+
"loss": 1.88,
|
| 1436 |
+
"step": 4760
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 2.94,
|
| 1440 |
+
"learning_rate": 6.207744314689612e-06,
|
| 1441 |
+
"loss": 1.8669,
|
| 1442 |
+
"step": 4780
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 2.95,
|
| 1446 |
+
"learning_rate": 4.978488014751076e-06,
|
| 1447 |
+
"loss": 1.8853,
|
| 1448 |
+
"step": 4800
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.96,
|
| 1452 |
+
"learning_rate": 3.749231714812538e-06,
|
| 1453 |
+
"loss": 1.958,
|
| 1454 |
+
"step": 4820
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 2.97,
|
| 1458 |
+
"learning_rate": 2.519975414874001e-06,
|
| 1459 |
+
"loss": 1.8813,
|
| 1460 |
+
"step": 4840
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 2.99,
|
| 1464 |
+
"learning_rate": 1.2907191149354641e-06,
|
| 1465 |
+
"loss": 1.8783,
|
| 1466 |
+
"step": 4860
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 3.0,
|
| 1470 |
+
"learning_rate": 6.146281499692685e-08,
|
| 1471 |
+
"loss": 1.9501,
|
| 1472 |
+
"step": 4880
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.0,
|
| 1476 |
+
"eval_loss": 1.9423085451126099,
|
| 1477 |
+
"eval_runtime": 128.9524,
|
| 1478 |
+
"eval_samples_per_second": 15.51,
|
| 1479 |
+
"eval_steps_per_second": 1.939,
|
| 1480 |
+
"step": 4881
|
| 1481 |
+
}
|
| 1482 |
+
],
|
| 1483 |
+
"max_steps": 4881,
|
| 1484 |
+
"num_train_epochs": 3,
|
| 1485 |
+
"total_flos": 4.564905941532672e+16,
|
| 1486 |
+
"trial_name": null,
|
| 1487 |
+
"trial_params": null
|
| 1488 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93bccf288a74034fa1d2abb6ca22fd64ee72d461467f59ed61e4d7f5068789a9
|
| 3 |
+
size 3899
|