Upload 8 files
Browse files- README.md +37 -0
- adapter_config.json +21 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map (1).json +24 -0
- tokenizer_config (1).json +39 -0
- trainer_state.json +1456 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
library_name: peft
|
| 2 |
+
Training procedure
|
| 3 |
+
The following bitsandbytes quantization config was used during training:
|
| 4 |
+
|
| 5 |
+
load_in_8bit: False
|
| 6 |
+
load_in_4bit: True
|
| 7 |
+
llm_int8_threshold: 6.0
|
| 8 |
+
llm_int8_skip_modules: None
|
| 9 |
+
llm_int8_enable_fp32_cpu_offload: False
|
| 10 |
+
llm_int8_has_fp16_weight: False
|
| 11 |
+
bnb_4bit_quant_type: nf4
|
| 12 |
+
bnb_4bit_use_double_quant: False
|
| 13 |
+
bnb_4bit_compute_dtype: float16
|
| 14 |
+
The following bitsandbytes quantization config was used during training:
|
| 15 |
+
|
| 16 |
+
load_in_8bit: False
|
| 17 |
+
|
| 18 |
+
load_in_4bit: True
|
| 19 |
+
|
| 20 |
+
llm_int8_threshold: 6.0
|
| 21 |
+
|
| 22 |
+
llm_int8_skip_modules: None
|
| 23 |
+
|
| 24 |
+
llm_int8_enable_fp32_cpu_offload: False
|
| 25 |
+
|
| 26 |
+
llm_int8_has_fp16_weight: False
|
| 27 |
+
|
| 28 |
+
bnb_4bit_quant_type: nf4
|
| 29 |
+
|
| 30 |
+
bnb_4bit_use_double_quant: False
|
| 31 |
+
|
| 32 |
+
bnb_4bit_compute_dtype: float16
|
| 33 |
+
|
| 34 |
+
Framework versions
|
| 35 |
+
PEFT 0.4.0
|
| 36 |
+
|
| 37 |
+
PEFT 0.4.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "danwils/mala-alpaca-base",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 16,
|
| 11 |
+
"lora_dropout": 0.1,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 64,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"v_proj"
|
| 19 |
+
],
|
| 20 |
+
"task_type": "CAUSAL_LM"
|
| 21 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3846cb7b8d1f03a47e641d5224362483391e727754ebcd43bb54624ff38873c
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:962dc8045e8bd336ec4241502f6dc70632ca6671cb91393b23391b4cc0531f95
|
| 3 |
+
size 1064
|
special_tokens_map (1).json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer_config (1).json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"bos_token": "<s>",
|
| 29 |
+
"clean_up_tokenization_spaces": false,
|
| 30 |
+
"eos_token": "</s>",
|
| 31 |
+
"legacy": false,
|
| 32 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 33 |
+
"pad_token": null,
|
| 34 |
+
"padding_side": "right",
|
| 35 |
+
"sp_model_kwargs": {},
|
| 36 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 37 |
+
"unk_token": "<unk>",
|
| 38 |
+
"use_default_system_prompt": false
|
| 39 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1456 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.769124264451367,
|
| 5 |
+
"global_step": 6000,
|
| 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": 3.8461538461538456e-05,
|
| 13 |
+
"loss": 3.0784,
|
| 14 |
+
"step": 25
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.02,
|
| 18 |
+
"learning_rate": 7.692307692307691e-05,
|
| 19 |
+
"loss": 2.3977,
|
| 20 |
+
"step": 50
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.03,
|
| 24 |
+
"learning_rate": 0.00011538461538461538,
|
| 25 |
+
"loss": 1.443,
|
| 26 |
+
"step": 75
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 0.00015384615384615382,
|
| 31 |
+
"loss": 0.8912,
|
| 32 |
+
"step": 100
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.06,
|
| 36 |
+
"learning_rate": 0.0001923076923076923,
|
| 37 |
+
"loss": 1.1574,
|
| 38 |
+
"step": 125
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07,
|
| 42 |
+
"learning_rate": 0.00023076923076923076,
|
| 43 |
+
"loss": 0.8311,
|
| 44 |
+
"step": 150
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08,
|
| 48 |
+
"learning_rate": 0.0002692307692307692,
|
| 49 |
+
"loss": 1.1308,
|
| 50 |
+
"step": 175
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.09,
|
| 54 |
+
"learning_rate": 0.0002999995341928201,
|
| 55 |
+
"loss": 0.7901,
|
| 56 |
+
"step": 200
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.1,
|
| 60 |
+
"learning_rate": 0.0002999832312452881,
|
| 61 |
+
"loss": 1.1048,
|
| 62 |
+
"step": 225
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.12,
|
| 66 |
+
"learning_rate": 0.0002999436408316976,
|
| 67 |
+
"loss": 0.7632,
|
| 68 |
+
"step": 250
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.13,
|
| 72 |
+
"learning_rate": 0.0002998807690991386,
|
| 73 |
+
"loss": 1.1239,
|
| 74 |
+
"step": 275
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.14,
|
| 78 |
+
"learning_rate": 0.00029979462580952446,
|
| 79 |
+
"loss": 0.7622,
|
| 80 |
+
"step": 300
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.15,
|
| 84 |
+
"learning_rate": 0.00029968522433807687,
|
| 85 |
+
"loss": 1.105,
|
| 86 |
+
"step": 325
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.16,
|
| 90 |
+
"learning_rate": 0.00029955258167124835,
|
| 91 |
+
"loss": 0.7669,
|
| 92 |
+
"step": 350
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.17,
|
| 96 |
+
"learning_rate": 0.0002993967184040853,
|
| 97 |
+
"loss": 1.0854,
|
| 98 |
+
"step": 375
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.18,
|
| 102 |
+
"learning_rate": 0.00029921765873703004,
|
| 103 |
+
"loss": 0.7586,
|
| 104 |
+
"step": 400
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2,
|
| 108 |
+
"learning_rate": 0.0002990154304721634,
|
| 109 |
+
"loss": 1.0849,
|
| 110 |
+
"step": 425
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.21,
|
| 114 |
+
"learning_rate": 0.000298790065008888,
|
| 115 |
+
"loss": 0.7403,
|
| 116 |
+
"step": 450
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.22,
|
| 120 |
+
"learning_rate": 0.0002985415973390528,
|
| 121 |
+
"loss": 1.0786,
|
| 122 |
+
"step": 475
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.23,
|
| 126 |
+
"learning_rate": 0.00029827006604152,
|
| 127 |
+
"loss": 0.7676,
|
| 128 |
+
"step": 500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.24,
|
| 132 |
+
"learning_rate": 0.00029797551327617543,
|
| 133 |
+
"loss": 1.087,
|
| 134 |
+
"step": 525
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.25,
|
| 138 |
+
"learning_rate": 0.0002976579847773818,
|
| 139 |
+
"loss": 0.7685,
|
| 140 |
+
"step": 550
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.27,
|
| 144 |
+
"learning_rate": 0.00029731752984687835,
|
| 145 |
+
"loss": 1.072,
|
| 146 |
+
"step": 575
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.28,
|
| 150 |
+
"learning_rate": 0.0002969542013461254,
|
| 151 |
+
"loss": 0.7611,
|
| 152 |
+
"step": 600
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.29,
|
| 156 |
+
"learning_rate": 0.000296568055688097,
|
| 157 |
+
"loss": 1.0691,
|
| 158 |
+
"step": 625
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.3,
|
| 162 |
+
"learning_rate": 0.00029615915282852175,
|
| 163 |
+
"loss": 0.7512,
|
| 164 |
+
"step": 650
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.31,
|
| 168 |
+
"learning_rate": 0.00029572755625657365,
|
| 169 |
+
"loss": 1.0789,
|
| 170 |
+
"step": 675
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.32,
|
| 174 |
+
"learning_rate": 0.00029527333298501427,
|
| 175 |
+
"loss": 0.739,
|
| 176 |
+
"step": 700
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.33,
|
| 180 |
+
"learning_rate": 0.0002947965535397879,
|
| 181 |
+
"loss": 1.0944,
|
| 182 |
+
"step": 725
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.35,
|
| 186 |
+
"learning_rate": 0.0002942972919490714,
|
| 187 |
+
"loss": 0.7681,
|
| 188 |
+
"step": 750
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.36,
|
| 192 |
+
"learning_rate": 0.0002937756257317797,
|
| 193 |
+
"loss": 1.0773,
|
| 194 |
+
"step": 775
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.37,
|
| 198 |
+
"learning_rate": 0.00029323163588552975,
|
| 199 |
+
"loss": 0.7541,
|
| 200 |
+
"step": 800
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.38,
|
| 204 |
+
"learning_rate": 0.0002926654068740645,
|
| 205 |
+
"loss": 1.0885,
|
| 206 |
+
"step": 825
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.39,
|
| 210 |
+
"learning_rate": 0.0002920770266141382,
|
| 211 |
+
"loss": 0.7505,
|
| 212 |
+
"step": 850
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.4,
|
| 216 |
+
"learning_rate": 0.000291466586461866,
|
| 217 |
+
"loss": 1.0623,
|
| 218 |
+
"step": 875
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.42,
|
| 222 |
+
"learning_rate": 0.00029083418119853904,
|
| 223 |
+
"loss": 0.7429,
|
| 224 |
+
"step": 900
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.43,
|
| 228 |
+
"learning_rate": 0.0002901799090159085,
|
| 229 |
+
"loss": 1.0523,
|
| 230 |
+
"step": 925
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.44,
|
| 234 |
+
"learning_rate": 0.00028950387150093916,
|
| 235 |
+
"loss": 0.7258,
|
| 236 |
+
"step": 950
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.45,
|
| 240 |
+
"learning_rate": 0.0002888061736200367,
|
| 241 |
+
"loss": 1.0618,
|
| 242 |
+
"step": 975
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.46,
|
| 246 |
+
"learning_rate": 0.0002880869237027496,
|
| 247 |
+
"loss": 0.739,
|
| 248 |
+
"step": 1000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.47,
|
| 252 |
+
"learning_rate": 0.0002873462334249492,
|
| 253 |
+
"loss": 1.063,
|
| 254 |
+
"step": 1025
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.48,
|
| 258 |
+
"learning_rate": 0.00028658421779149,
|
| 259 |
+
"loss": 0.7298,
|
| 260 |
+
"step": 1050
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.5,
|
| 264 |
+
"learning_rate": 0.0002858009951183536,
|
| 265 |
+
"loss": 1.0526,
|
| 266 |
+
"step": 1075
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.51,
|
| 270 |
+
"learning_rate": 0.0002849966870142775,
|
| 271 |
+
"loss": 0.7322,
|
| 272 |
+
"step": 1100
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.52,
|
| 276 |
+
"learning_rate": 0.00028417141836187374,
|
| 277 |
+
"loss": 1.0509,
|
| 278 |
+
"step": 1125
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.53,
|
| 282 |
+
"learning_rate": 0.0002833253172982385,
|
| 283 |
+
"loss": 0.7245,
|
| 284 |
+
"step": 1150
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.54,
|
| 288 |
+
"learning_rate": 0.0002824585151950568,
|
| 289 |
+
"loss": 1.0746,
|
| 290 |
+
"step": 1175
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.55,
|
| 294 |
+
"learning_rate": 0.00028157114663820445,
|
| 295 |
+
"loss": 0.7327,
|
| 296 |
+
"step": 1200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.57,
|
| 300 |
+
"learning_rate": 0.0002806633494068519,
|
| 301 |
+
"loss": 1.0685,
|
| 302 |
+
"step": 1225
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.58,
|
| 306 |
+
"learning_rate": 0.00027973526445207084,
|
| 307 |
+
"loss": 0.7457,
|
| 308 |
+
"step": 1250
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.59,
|
| 312 |
+
"learning_rate": 0.00027878703587495,
|
| 313 |
+
"loss": 1.0731,
|
| 314 |
+
"step": 1275
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.6,
|
| 318 |
+
"learning_rate": 0.00027781881090422017,
|
| 319 |
+
"loss": 0.728,
|
| 320 |
+
"step": 1300
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.61,
|
| 324 |
+
"learning_rate": 0.00027683073987339523,
|
| 325 |
+
"loss": 1.0638,
|
| 326 |
+
"step": 1325
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.62,
|
| 330 |
+
"learning_rate": 0.0002758229761974294,
|
| 331 |
+
"loss": 0.734,
|
| 332 |
+
"step": 1350
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.63,
|
| 336 |
+
"learning_rate": 0.0002747956763488979,
|
| 337 |
+
"loss": 1.0548,
|
| 338 |
+
"step": 1375
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.65,
|
| 342 |
+
"learning_rate": 0.0002737489998337011,
|
| 343 |
+
"loss": 0.7357,
|
| 344 |
+
"step": 1400
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.66,
|
| 348 |
+
"learning_rate": 0.00027268310916629885,
|
| 349 |
+
"loss": 1.0528,
|
| 350 |
+
"step": 1425
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.67,
|
| 354 |
+
"learning_rate": 0.0002715981698444775,
|
| 355 |
+
"loss": 0.7227,
|
| 356 |
+
"step": 1450
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.68,
|
| 360 |
+
"learning_rate": 0.0002704943503236531,
|
| 361 |
+
"loss": 1.0546,
|
| 362 |
+
"step": 1475
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.69,
|
| 366 |
+
"learning_rate": 0.00026937182199071613,
|
| 367 |
+
"loss": 0.7228,
|
| 368 |
+
"step": 1500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.7,
|
| 372 |
+
"learning_rate": 0.0002682307591374206,
|
| 373 |
+
"loss": 1.052,
|
| 374 |
+
"step": 1525
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.72,
|
| 378 |
+
"learning_rate": 0.0002670713389333225,
|
| 379 |
+
"loss": 0.729,
|
| 380 |
+
"step": 1550
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.73,
|
| 384 |
+
"learning_rate": 0.0002658937413982708,
|
| 385 |
+
"loss": 1.0472,
|
| 386 |
+
"step": 1575
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.74,
|
| 390 |
+
"learning_rate": 0.00026469814937445655,
|
| 391 |
+
"loss": 0.7294,
|
| 392 |
+
"step": 1600
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.75,
|
| 396 |
+
"learning_rate": 0.00026348474849802357,
|
| 397 |
+
"loss": 1.0383,
|
| 398 |
+
"step": 1625
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.76,
|
| 402 |
+
"learning_rate": 0.000262253727170245,
|
| 403 |
+
"loss": 0.723,
|
| 404 |
+
"step": 1650
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.77,
|
| 408 |
+
"learning_rate": 0.0002610052765282711,
|
| 409 |
+
"loss": 1.0467,
|
| 410 |
+
"step": 1675
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.78,
|
| 414 |
+
"learning_rate": 0.0002597395904154515,
|
| 415 |
+
"loss": 0.7191,
|
| 416 |
+
"step": 1700
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.8,
|
| 420 |
+
"learning_rate": 0.0002584568653512381,
|
| 421 |
+
"loss": 1.0481,
|
| 422 |
+
"step": 1725
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.81,
|
| 426 |
+
"learning_rate": 0.0002571573005006718,
|
| 427 |
+
"loss": 0.6965,
|
| 428 |
+
"step": 1750
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.82,
|
| 432 |
+
"learning_rate": 0.00025584109764345905,
|
| 433 |
+
"loss": 1.0579,
|
| 434 |
+
"step": 1775
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.83,
|
| 438 |
+
"learning_rate": 0.0002545084611426415,
|
| 439 |
+
"loss": 0.7132,
|
| 440 |
+
"step": 1800
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.84,
|
| 444 |
+
"learning_rate": 0.00025315959791286585,
|
| 445 |
+
"loss": 1.0537,
|
| 446 |
+
"step": 1825
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.85,
|
| 450 |
+
"learning_rate": 0.0002517947173882562,
|
| 451 |
+
"loss": 0.7253,
|
| 452 |
+
"step": 1850
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.87,
|
| 456 |
+
"learning_rate": 0.0002504140314898962,
|
| 457 |
+
"loss": 1.0335,
|
| 458 |
+
"step": 1875
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.88,
|
| 462 |
+
"learning_rate": 0.0002490177545929246,
|
| 463 |
+
"loss": 0.7112,
|
| 464 |
+
"step": 1900
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.89,
|
| 468 |
+
"learning_rate": 0.00024760610349324955,
|
| 469 |
+
"loss": 1.07,
|
| 470 |
+
"step": 1925
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.9,
|
| 474 |
+
"learning_rate": 0.0002461792973738878,
|
| 475 |
+
"loss": 0.7135,
|
| 476 |
+
"step": 1950
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.91,
|
| 480 |
+
"learning_rate": 0.00024473755777093244,
|
| 481 |
+
"loss": 1.0404,
|
| 482 |
+
"step": 1975
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.92,
|
| 486 |
+
"learning_rate": 0.00024328110853915556,
|
| 487 |
+
"loss": 0.714,
|
| 488 |
+
"step": 2000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.93,
|
| 492 |
+
"learning_rate": 0.00024181017581725143,
|
| 493 |
+
"loss": 1.051,
|
| 494 |
+
"step": 2025
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.95,
|
| 498 |
+
"learning_rate": 0.00024032498799272402,
|
| 499 |
+
"loss": 0.725,
|
| 500 |
+
"step": 2050
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.96,
|
| 504 |
+
"learning_rate": 0.0002388257756664263,
|
| 505 |
+
"loss": 1.054,
|
| 506 |
+
"step": 2075
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.97,
|
| 510 |
+
"learning_rate": 0.00023731277161675532,
|
| 511 |
+
"loss": 0.7111,
|
| 512 |
+
"step": 2100
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.98,
|
| 516 |
+
"learning_rate": 0.00023578621076350965,
|
| 517 |
+
"loss": 1.0462,
|
| 518 |
+
"step": 2125
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.99,
|
| 522 |
+
"learning_rate": 0.00023424633013141363,
|
| 523 |
+
"loss": 0.7244,
|
| 524 |
+
"step": 2150
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.0,
|
| 528 |
+
"learning_rate": 0.00023269336881331565,
|
| 529 |
+
"loss": 0.9686,
|
| 530 |
+
"step": 2175
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.02,
|
| 534 |
+
"learning_rate": 0.00023112756793306445,
|
| 535 |
+
"loss": 0.8741,
|
| 536 |
+
"step": 2200
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.03,
|
| 540 |
+
"learning_rate": 0.000229549170608071,
|
| 541 |
+
"loss": 0.7772,
|
| 542 |
+
"step": 2225
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.04,
|
| 546 |
+
"learning_rate": 0.00022795842191156,
|
| 547 |
+
"loss": 0.8901,
|
| 548 |
+
"step": 2250
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.05,
|
| 552 |
+
"learning_rate": 0.00022635556883451806,
|
| 553 |
+
"loss": 0.8139,
|
| 554 |
+
"step": 2275
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.06,
|
| 558 |
+
"learning_rate": 0.00022474086024734443,
|
| 559 |
+
"loss": 0.8782,
|
| 560 |
+
"step": 2300
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.07,
|
| 564 |
+
"learning_rate": 0.0002231145468612094,
|
| 565 |
+
"loss": 0.8002,
|
| 566 |
+
"step": 2325
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.08,
|
| 570 |
+
"learning_rate": 0.00022147688118912695,
|
| 571 |
+
"loss": 0.9184,
|
| 572 |
+
"step": 2350
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.1,
|
| 576 |
+
"learning_rate": 0.000219828117506748,
|
| 577 |
+
"loss": 0.7961,
|
| 578 |
+
"step": 2375
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.11,
|
| 582 |
+
"learning_rate": 0.0002181685118128797,
|
| 583 |
+
"loss": 0.9099,
|
| 584 |
+
"step": 2400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.12,
|
| 588 |
+
"learning_rate": 0.00021649832178973727,
|
| 589 |
+
"loss": 0.8191,
|
| 590 |
+
"step": 2425
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.13,
|
| 594 |
+
"learning_rate": 0.00021481780676293428,
|
| 595 |
+
"loss": 0.925,
|
| 596 |
+
"step": 2450
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.14,
|
| 600 |
+
"learning_rate": 0.00021312722766121808,
|
| 601 |
+
"loss": 0.7863,
|
| 602 |
+
"step": 2475
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.15,
|
| 606 |
+
"learning_rate": 0.00021142684697595605,
|
| 607 |
+
"loss": 0.9065,
|
| 608 |
+
"step": 2500
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.17,
|
| 612 |
+
"learning_rate": 0.00020971692872037923,
|
| 613 |
+
"loss": 0.7908,
|
| 614 |
+
"step": 2525
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.18,
|
| 618 |
+
"learning_rate": 0.00020799773838859004,
|
| 619 |
+
"loss": 0.9115,
|
| 620 |
+
"step": 2550
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.19,
|
| 624 |
+
"learning_rate": 0.00020626954291433935,
|
| 625 |
+
"loss": 0.7844,
|
| 626 |
+
"step": 2575
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.2,
|
| 630 |
+
"learning_rate": 0.0002045326106295808,
|
| 631 |
+
"loss": 0.9132,
|
| 632 |
+
"step": 2600
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.21,
|
| 636 |
+
"learning_rate": 0.0002027872112228075,
|
| 637 |
+
"loss": 0.7805,
|
| 638 |
+
"step": 2625
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.22,
|
| 642 |
+
"learning_rate": 0.00020103361569717837,
|
| 643 |
+
"loss": 0.9106,
|
| 644 |
+
"step": 2650
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.23,
|
| 648 |
+
"learning_rate": 0.00019927209632844006,
|
| 649 |
+
"loss": 0.791,
|
| 650 |
+
"step": 2675
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.25,
|
| 654 |
+
"learning_rate": 0.00019750292662265174,
|
| 655 |
+
"loss": 0.8947,
|
| 656 |
+
"step": 2700
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.26,
|
| 660 |
+
"learning_rate": 0.00019572638127371837,
|
| 661 |
+
"loss": 0.7896,
|
| 662 |
+
"step": 2725
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.27,
|
| 666 |
+
"learning_rate": 0.00019394273612073992,
|
| 667 |
+
"loss": 0.8987,
|
| 668 |
+
"step": 2750
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.28,
|
| 672 |
+
"learning_rate": 0.00019215226810518254,
|
| 673 |
+
"loss": 0.8036,
|
| 674 |
+
"step": 2775
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.29,
|
| 678 |
+
"learning_rate": 0.0001903552552278787,
|
| 679 |
+
"loss": 0.9274,
|
| 680 |
+
"step": 2800
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.3,
|
| 684 |
+
"learning_rate": 0.0001885519765058627,
|
| 685 |
+
"loss": 0.7937,
|
| 686 |
+
"step": 2825
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 1.32,
|
| 690 |
+
"learning_rate": 0.0001867427119290488,
|
| 691 |
+
"loss": 0.8966,
|
| 692 |
+
"step": 2850
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.33,
|
| 696 |
+
"learning_rate": 0.00018492774241675742,
|
| 697 |
+
"loss": 0.7985,
|
| 698 |
+
"step": 2875
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.34,
|
| 702 |
+
"learning_rate": 0.00018310734977409826,
|
| 703 |
+
"loss": 0.8964,
|
| 704 |
+
"step": 2900
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.35,
|
| 708 |
+
"learning_rate": 0.00018128181664821475,
|
| 709 |
+
"loss": 0.7905,
|
| 710 |
+
"step": 2925
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.36,
|
| 714 |
+
"learning_rate": 0.00017945142648439843,
|
| 715 |
+
"loss": 0.8919,
|
| 716 |
+
"step": 2950
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.37,
|
| 720 |
+
"learning_rate": 0.00017761646348207917,
|
| 721 |
+
"loss": 0.8017,
|
| 722 |
+
"step": 2975
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 1.38,
|
| 726 |
+
"learning_rate": 0.0001757772125506984,
|
| 727 |
+
"loss": 0.8873,
|
| 728 |
+
"step": 3000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.4,
|
| 732 |
+
"learning_rate": 0.00017393395926547202,
|
| 733 |
+
"loss": 0.7904,
|
| 734 |
+
"step": 3025
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.41,
|
| 738 |
+
"learning_rate": 0.00017208698982304984,
|
| 739 |
+
"loss": 0.9017,
|
| 740 |
+
"step": 3050
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.42,
|
| 744 |
+
"learning_rate": 0.00017023659099707876,
|
| 745 |
+
"loss": 0.7917,
|
| 746 |
+
"step": 3075
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.43,
|
| 750 |
+
"learning_rate": 0.00016838305009367628,
|
| 751 |
+
"loss": 0.9097,
|
| 752 |
+
"step": 3100
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.44,
|
| 756 |
+
"learning_rate": 0.00016652665490682128,
|
| 757 |
+
"loss": 0.8175,
|
| 758 |
+
"step": 3125
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.45,
|
| 762 |
+
"learning_rate": 0.00016466769367366887,
|
| 763 |
+
"loss": 0.8899,
|
| 764 |
+
"step": 3150
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.47,
|
| 768 |
+
"learning_rate": 0.00016280645502979695,
|
| 769 |
+
"loss": 0.7847,
|
| 770 |
+
"step": 3175
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.48,
|
| 774 |
+
"learning_rate": 0.00016094322796439043,
|
| 775 |
+
"loss": 0.8796,
|
| 776 |
+
"step": 3200
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.49,
|
| 780 |
+
"learning_rate": 0.00015907830177537078,
|
| 781 |
+
"loss": 0.7959,
|
| 782 |
+
"step": 3225
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.5,
|
| 786 |
+
"learning_rate": 0.00015721196602447735,
|
| 787 |
+
"loss": 0.888,
|
| 788 |
+
"step": 3250
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.51,
|
| 792 |
+
"learning_rate": 0.0001553445104923083,
|
| 793 |
+
"loss": 0.7741,
|
| 794 |
+
"step": 3275
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.52,
|
| 798 |
+
"learning_rate": 0.00015347622513332685,
|
| 799 |
+
"loss": 0.9118,
|
| 800 |
+
"step": 3300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.53,
|
| 804 |
+
"learning_rate": 0.00015160740003084127,
|
| 805 |
+
"loss": 0.7928,
|
| 806 |
+
"step": 3325
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.55,
|
| 810 |
+
"learning_rate": 0.00014973832535196402,
|
| 811 |
+
"loss": 0.8985,
|
| 812 |
+
"step": 3350
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.56,
|
| 816 |
+
"learning_rate": 0.00014786929130255872,
|
| 817 |
+
"loss": 0.7885,
|
| 818 |
+
"step": 3375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.57,
|
| 822 |
+
"learning_rate": 0.00014600058808218047,
|
| 823 |
+
"loss": 0.8915,
|
| 824 |
+
"step": 3400
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.58,
|
| 828 |
+
"learning_rate": 0.0001441325058390176,
|
| 829 |
+
"loss": 0.775,
|
| 830 |
+
"step": 3425
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.59,
|
| 834 |
+
"learning_rate": 0.00014226533462484097,
|
| 835 |
+
"loss": 0.9084,
|
| 836 |
+
"step": 3450
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.6,
|
| 840 |
+
"learning_rate": 0.0001403993643499688,
|
| 841 |
+
"loss": 0.783,
|
| 842 |
+
"step": 3475
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.62,
|
| 846 |
+
"learning_rate": 0.00013853488473825263,
|
| 847 |
+
"loss": 0.897,
|
| 848 |
+
"step": 3500
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 1.63,
|
| 852 |
+
"learning_rate": 0.00013667218528209314,
|
| 853 |
+
"loss": 0.7799,
|
| 854 |
+
"step": 3525
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.64,
|
| 858 |
+
"learning_rate": 0.00013481155519749142,
|
| 859 |
+
"loss": 0.9029,
|
| 860 |
+
"step": 3550
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.65,
|
| 864 |
+
"learning_rate": 0.0001329532833791431,
|
| 865 |
+
"loss": 0.7867,
|
| 866 |
+
"step": 3575
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.66,
|
| 870 |
+
"learning_rate": 0.00013109765835558256,
|
| 871 |
+
"loss": 0.8932,
|
| 872 |
+
"step": 3600
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.67,
|
| 876 |
+
"learning_rate": 0.00012924496824438401,
|
| 877 |
+
"loss": 0.7868,
|
| 878 |
+
"step": 3625
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.68,
|
| 882 |
+
"learning_rate": 0.00012739550070742616,
|
| 883 |
+
"loss": 0.8851,
|
| 884 |
+
"step": 3650
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.7,
|
| 888 |
+
"learning_rate": 0.00012554954290622798,
|
| 889 |
+
"loss": 0.7783,
|
| 890 |
+
"step": 3675
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.71,
|
| 894 |
+
"learning_rate": 0.0001237073814573617,
|
| 895 |
+
"loss": 0.8929,
|
| 896 |
+
"step": 3700
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.72,
|
| 900 |
+
"learning_rate": 0.00012186930238795088,
|
| 901 |
+
"loss": 0.8003,
|
| 902 |
+
"step": 3725
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.73,
|
| 906 |
+
"learning_rate": 0.00012003559109125972,
|
| 907 |
+
"loss": 0.8833,
|
| 908 |
+
"step": 3750
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 1.74,
|
| 912 |
+
"learning_rate": 0.00011820653228238095,
|
| 913 |
+
"loss": 0.7831,
|
| 914 |
+
"step": 3775
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.75,
|
| 918 |
+
"learning_rate": 0.00011638240995402873,
|
| 919 |
+
"loss": 0.8936,
|
| 920 |
+
"step": 3800
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.77,
|
| 924 |
+
"learning_rate": 0.00011456350733244412,
|
| 925 |
+
"loss": 0.7678,
|
| 926 |
+
"step": 3825
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.78,
|
| 930 |
+
"learning_rate": 0.00011275010683341939,
|
| 931 |
+
"loss": 0.9017,
|
| 932 |
+
"step": 3850
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 1.79,
|
| 936 |
+
"learning_rate": 0.00011094249001844783,
|
| 937 |
+
"loss": 0.7693,
|
| 938 |
+
"step": 3875
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 1.8,
|
| 942 |
+
"learning_rate": 0.0001091409375510068,
|
| 943 |
+
"loss": 0.8821,
|
| 944 |
+
"step": 3900
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.81,
|
| 948 |
+
"learning_rate": 0.00010734572915297985,
|
| 949 |
+
"loss": 0.7796,
|
| 950 |
+
"step": 3925
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1.82,
|
| 954 |
+
"learning_rate": 0.000105557143561225,
|
| 955 |
+
"loss": 0.8948,
|
| 956 |
+
"step": 3950
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.83,
|
| 960 |
+
"learning_rate": 0.00010377545848429614,
|
| 961 |
+
"loss": 0.7797,
|
| 962 |
+
"step": 3975
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.85,
|
| 966 |
+
"learning_rate": 0.00010200095055932406,
|
| 967 |
+
"loss": 0.8789,
|
| 968 |
+
"step": 4000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.86,
|
| 972 |
+
"learning_rate": 0.00010023389530906374,
|
| 973 |
+
"loss": 0.791,
|
| 974 |
+
"step": 4025
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.87,
|
| 978 |
+
"learning_rate": 9.847456709911478e-05,
|
| 979 |
+
"loss": 0.8838,
|
| 980 |
+
"step": 4050
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.88,
|
| 984 |
+
"learning_rate": 9.672323909532129e-05,
|
| 985 |
+
"loss": 0.7831,
|
| 986 |
+
"step": 4075
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.89,
|
| 990 |
+
"learning_rate": 9.498018322135838e-05,
|
| 991 |
+
"loss": 0.8763,
|
| 992 |
+
"step": 4100
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 1.9,
|
| 996 |
+
"learning_rate": 9.324567011651136e-05,
|
| 997 |
+
"loss": 0.8028,
|
| 998 |
+
"step": 4125
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.92,
|
| 1002 |
+
"learning_rate": 9.151996909365419e-05,
|
| 1003 |
+
"loss": 0.9126,
|
| 1004 |
+
"step": 4150
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.93,
|
| 1008 |
+
"learning_rate": 8.980334809743414e-05,
|
| 1009 |
+
"loss": 0.7953,
|
| 1010 |
+
"step": 4175
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.94,
|
| 1014 |
+
"learning_rate": 8.809607366266878e-05,
|
| 1015 |
+
"loss": 0.8934,
|
| 1016 |
+
"step": 4200
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.95,
|
| 1020 |
+
"learning_rate": 8.639841087296207e-05,
|
| 1021 |
+
"loss": 0.7763,
|
| 1022 |
+
"step": 4225
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 1.96,
|
| 1026 |
+
"learning_rate": 8.471062331954525e-05,
|
| 1027 |
+
"loss": 0.8913,
|
| 1028 |
+
"step": 4250
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.97,
|
| 1032 |
+
"learning_rate": 8.303297306034988e-05,
|
| 1033 |
+
"loss": 0.7765,
|
| 1034 |
+
"step": 4275
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.98,
|
| 1038 |
+
"learning_rate": 8.136572057931912e-05,
|
| 1039 |
+
"loss": 0.8763,
|
| 1040 |
+
"step": 4300
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 2.0,
|
| 1044 |
+
"learning_rate": 7.970912474596287e-05,
|
| 1045 |
+
"loss": 0.7483,
|
| 1046 |
+
"step": 4325
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.01,
|
| 1050 |
+
"learning_rate": 7.806344277516381e-05,
|
| 1051 |
+
"loss": 0.9051,
|
| 1052 |
+
"step": 4350
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.02,
|
| 1056 |
+
"learning_rate": 7.642893018724062e-05,
|
| 1057 |
+
"loss": 0.703,
|
| 1058 |
+
"step": 4375
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 2.03,
|
| 1062 |
+
"learning_rate": 7.480584076827394e-05,
|
| 1063 |
+
"loss": 0.8799,
|
| 1064 |
+
"step": 4400
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 2.04,
|
| 1068 |
+
"learning_rate": 7.319442653070165e-05,
|
| 1069 |
+
"loss": 0.7338,
|
| 1070 |
+
"step": 4425
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.05,
|
| 1074 |
+
"learning_rate": 7.159493767418978e-05,
|
| 1075 |
+
"loss": 0.8673,
|
| 1076 |
+
"step": 4450
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.07,
|
| 1080 |
+
"learning_rate": 7.000762254678456e-05,
|
| 1081 |
+
"loss": 0.7347,
|
| 1082 |
+
"step": 4475
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 2.08,
|
| 1086 |
+
"learning_rate": 6.843272760635235e-05,
|
| 1087 |
+
"loss": 0.8641,
|
| 1088 |
+
"step": 4500
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.09,
|
| 1092 |
+
"learning_rate": 6.687049738231268e-05,
|
| 1093 |
+
"loss": 0.7346,
|
| 1094 |
+
"step": 4525
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.1,
|
| 1098 |
+
"learning_rate": 6.532117443767072e-05,
|
| 1099 |
+
"loss": 0.8766,
|
| 1100 |
+
"step": 4550
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 2.11,
|
| 1104 |
+
"learning_rate": 6.378499933135577e-05,
|
| 1105 |
+
"loss": 0.7404,
|
| 1106 |
+
"step": 4575
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.12,
|
| 1110 |
+
"learning_rate": 6.226221058086958e-05,
|
| 1111 |
+
"loss": 0.8855,
|
| 1112 |
+
"step": 4600
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.13,
|
| 1116 |
+
"learning_rate": 6.0753044625253063e-05,
|
| 1117 |
+
"loss": 0.741,
|
| 1118 |
+
"step": 4625
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.15,
|
| 1122 |
+
"learning_rate": 5.9257735788374854e-05,
|
| 1123 |
+
"loss": 0.8679,
|
| 1124 |
+
"step": 4650
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 2.16,
|
| 1128 |
+
"learning_rate": 5.7776516242548395e-05,
|
| 1129 |
+
"loss": 0.7291,
|
| 1130 |
+
"step": 4675
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.17,
|
| 1134 |
+
"learning_rate": 5.6309615972483455e-05,
|
| 1135 |
+
"loss": 0.8408,
|
| 1136 |
+
"step": 4700
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.18,
|
| 1140 |
+
"learning_rate": 5.485726273957677e-05,
|
| 1141 |
+
"loss": 0.7349,
|
| 1142 |
+
"step": 4725
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 2.19,
|
| 1146 |
+
"learning_rate": 5.341968204654852e-05,
|
| 1147 |
+
"loss": 0.8473,
|
| 1148 |
+
"step": 4750
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 2.2,
|
| 1152 |
+
"learning_rate": 5.199709710242896e-05,
|
| 1153 |
+
"loss": 0.7233,
|
| 1154 |
+
"step": 4775
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.22,
|
| 1158 |
+
"learning_rate": 5.058972878790129e-05,
|
| 1159 |
+
"loss": 0.8649,
|
| 1160 |
+
"step": 4800
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.23,
|
| 1164 |
+
"learning_rate": 4.919779562100643e-05,
|
| 1165 |
+
"loss": 0.7435,
|
| 1166 |
+
"step": 4825
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.24,
|
| 1170 |
+
"learning_rate": 4.782151372321425e-05,
|
| 1171 |
+
"loss": 0.8768,
|
| 1172 |
+
"step": 4850
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.25,
|
| 1176 |
+
"learning_rate": 4.646109678586675e-05,
|
| 1177 |
+
"loss": 0.7419,
|
| 1178 |
+
"step": 4875
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.26,
|
| 1182 |
+
"learning_rate": 4.511675603699927e-05,
|
| 1183 |
+
"loss": 0.8745,
|
| 1184 |
+
"step": 4900
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 2.27,
|
| 1188 |
+
"learning_rate": 4.3788700208543235e-05,
|
| 1189 |
+
"loss": 0.7294,
|
| 1190 |
+
"step": 4925
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 2.28,
|
| 1194 |
+
"learning_rate": 4.2477135503917515e-05,
|
| 1195 |
+
"loss": 0.8768,
|
| 1196 |
+
"step": 4950
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.3,
|
| 1200 |
+
"learning_rate": 4.1182265566011184e-05,
|
| 1201 |
+
"loss": 0.74,
|
| 1202 |
+
"step": 4975
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 2.31,
|
| 1206 |
+
"learning_rate": 3.9904291445564937e-05,
|
| 1207 |
+
"loss": 0.8695,
|
| 1208 |
+
"step": 5000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 2.32,
|
| 1212 |
+
"learning_rate": 3.864341156995439e-05,
|
| 1213 |
+
"loss": 0.7214,
|
| 1214 |
+
"step": 5025
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.33,
|
| 1218 |
+
"learning_rate": 3.739982171238066e-05,
|
| 1219 |
+
"loss": 0.8579,
|
| 1220 |
+
"step": 5050
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.34,
|
| 1224 |
+
"learning_rate": 3.617371496147347e-05,
|
| 1225 |
+
"loss": 0.7466,
|
| 1226 |
+
"step": 5075
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.35,
|
| 1230 |
+
"learning_rate": 3.4965281691310755e-05,
|
| 1231 |
+
"loss": 0.8625,
|
| 1232 |
+
"step": 5100
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 2.37,
|
| 1236 |
+
"learning_rate": 3.3774709531859636e-05,
|
| 1237 |
+
"loss": 0.7496,
|
| 1238 |
+
"step": 5125
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.38,
|
| 1242 |
+
"learning_rate": 3.260218333984402e-05,
|
| 1243 |
+
"loss": 0.8741,
|
| 1244 |
+
"step": 5150
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 2.39,
|
| 1248 |
+
"learning_rate": 3.144788517004204e-05,
|
| 1249 |
+
"loss": 0.7245,
|
| 1250 |
+
"step": 5175
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 2.4,
|
| 1254 |
+
"learning_rate": 3.0311994247019213e-05,
|
| 1255 |
+
"loss": 0.8712,
|
| 1256 |
+
"step": 5200
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.41,
|
| 1260 |
+
"learning_rate": 2.9194686937300664e-05,
|
| 1261 |
+
"loss": 0.7163,
|
| 1262 |
+
"step": 5225
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.42,
|
| 1266 |
+
"learning_rate": 2.8096136721987088e-05,
|
| 1267 |
+
"loss": 0.8638,
|
| 1268 |
+
"step": 5250
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 2.43,
|
| 1272 |
+
"learning_rate": 2.7016514169819016e-05,
|
| 1273 |
+
"loss": 0.7206,
|
| 1274 |
+
"step": 5275
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 2.45,
|
| 1278 |
+
"learning_rate": 2.595598691069304e-05,
|
| 1279 |
+
"loss": 0.8693,
|
| 1280 |
+
"step": 5300
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.46,
|
| 1284 |
+
"learning_rate": 2.491471960963434e-05,
|
| 1285 |
+
"loss": 0.7308,
|
| 1286 |
+
"step": 5325
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.47,
|
| 1290 |
+
"learning_rate": 2.3892873941229695e-05,
|
| 1291 |
+
"loss": 0.8619,
|
| 1292 |
+
"step": 5350
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 2.48,
|
| 1296 |
+
"learning_rate": 2.2890608564524616e-05,
|
| 1297 |
+
"loss": 0.7062,
|
| 1298 |
+
"step": 5375
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.49,
|
| 1302 |
+
"learning_rate": 2.190807909838896e-05,
|
| 1303 |
+
"loss": 0.8553,
|
| 1304 |
+
"step": 5400
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.5,
|
| 1308 |
+
"learning_rate": 2.0945438097354384e-05,
|
| 1309 |
+
"loss": 0.7415,
|
| 1310 |
+
"step": 5425
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.52,
|
| 1314 |
+
"learning_rate": 2.0002835027927544e-05,
|
| 1315 |
+
"loss": 0.8702,
|
| 1316 |
+
"step": 5450
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.53,
|
| 1320 |
+
"learning_rate": 1.908041624538304e-05,
|
| 1321 |
+
"loss": 0.7552,
|
| 1322 |
+
"step": 5475
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.54,
|
| 1326 |
+
"learning_rate": 1.817832497103901e-05,
|
| 1327 |
+
"loss": 0.8496,
|
| 1328 |
+
"step": 5500
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 2.55,
|
| 1332 |
+
"learning_rate": 1.729670127001987e-05,
|
| 1333 |
+
"loss": 0.7167,
|
| 1334 |
+
"step": 5525
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 2.56,
|
| 1338 |
+
"learning_rate": 1.643568202950854e-05,
|
| 1339 |
+
"loss": 0.8556,
|
| 1340 |
+
"step": 5550
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.57,
|
| 1344 |
+
"learning_rate": 1.559540093749247e-05,
|
| 1345 |
+
"loss": 0.7245,
|
| 1346 |
+
"step": 5575
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.58,
|
| 1350 |
+
"learning_rate": 1.4775988462006316e-05,
|
| 1351 |
+
"loss": 0.8672,
|
| 1352 |
+
"step": 5600
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 2.6,
|
| 1356 |
+
"learning_rate": 1.3977571830874524e-05,
|
| 1357 |
+
"loss": 0.7386,
|
| 1358 |
+
"step": 5625
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 2.61,
|
| 1362 |
+
"learning_rate": 1.3200275011956901e-05,
|
| 1363 |
+
"loss": 0.862,
|
| 1364 |
+
"step": 5650
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.62,
|
| 1368 |
+
"learning_rate": 1.2444218693900826e-05,
|
| 1369 |
+
"loss": 0.7431,
|
| 1370 |
+
"step": 5675
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 2.63,
|
| 1374 |
+
"learning_rate": 1.170952026740178e-05,
|
| 1375 |
+
"loss": 0.8653,
|
| 1376 |
+
"step": 5700
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 2.64,
|
| 1380 |
+
"learning_rate": 1.0996293806976907e-05,
|
| 1381 |
+
"loss": 0.743,
|
| 1382 |
+
"step": 5725
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.65,
|
| 1386 |
+
"learning_rate": 1.0304650053252634e-05,
|
| 1387 |
+
"loss": 0.8539,
|
| 1388 |
+
"step": 5750
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.67,
|
| 1392 |
+
"learning_rate": 9.63469639577048e-06,
|
| 1393 |
+
"loss": 0.7485,
|
| 1394 |
+
"step": 5775
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 2.68,
|
| 1398 |
+
"learning_rate": 8.986536856312987e-06,
|
| 1399 |
+
"loss": 0.861,
|
| 1400 |
+
"step": 5800
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 2.69,
|
| 1404 |
+
"learning_rate": 8.360272072752406e-06,
|
| 1405 |
+
"loss": 0.7405,
|
| 1406 |
+
"step": 5825
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.7,
|
| 1410 |
+
"learning_rate": 7.755999283425168e-06,
|
| 1411 |
+
"loss": 0.8565,
|
| 1412 |
+
"step": 5850
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 2.71,
|
| 1416 |
+
"learning_rate": 7.1738123120338e-06,
|
| 1417 |
+
"loss": 0.7328,
|
| 1418 |
+
"step": 5875
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 2.72,
|
| 1422 |
+
"learning_rate": 6.613801553079184e-06,
|
| 1423 |
+
"loss": 0.8614,
|
| 1424 |
+
"step": 5900
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.73,
|
| 1428 |
+
"learning_rate": 6.076053957825411e-06,
|
| 1429 |
+
"loss": 0.7389,
|
| 1430 |
+
"step": 5925
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.75,
|
| 1434 |
+
"learning_rate": 5.560653020798906e-06,
|
| 1435 |
+
"loss": 0.8642,
|
| 1436 |
+
"step": 5950
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 2.76,
|
| 1440 |
+
"learning_rate": 5.067678766824651e-06,
|
| 1441 |
+
"loss": 0.7141,
|
| 1442 |
+
"step": 5975
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 2.77,
|
| 1446 |
+
"learning_rate": 4.597207738600911e-06,
|
| 1447 |
+
"loss": 0.8602,
|
| 1448 |
+
"step": 6000
|
| 1449 |
+
}
|
| 1450 |
+
],
|
| 1451 |
+
"max_steps": 6498,
|
| 1452 |
+
"num_train_epochs": 3,
|
| 1453 |
+
"total_flos": 4.8008476405304525e+17,
|
| 1454 |
+
"trial_name": null,
|
| 1455 |
+
"trial_params": null
|
| 1456 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53fdfcaf43ad16ff6ca49ce2445dbcc4efeb22a1be7a88fbf9579ae35f5cd0da
|
| 3 |
+
size 4408
|