Model save
Browse files- README.md +58 -0
- added_tokens.json +25 -0
- config.json +94 -0
- generation_config.json +6 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +32 -0
- tokenizer_config.json +216 -0
- trainer_state.json +3714 -0
- training_args.bin +3 -0
- vocab.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Nhanvi282/LLaVA_Qwen2.5_0.5b_SigLIPStage2
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
model-index:
|
| 6 |
+
- name: MoE_LlaVa_Qwen2.5_0.5bx4Top2_SigLIP
|
| 7 |
+
results: []
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 11 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 12 |
+
|
| 13 |
+
# MoE_LlaVa_Qwen2.5_0.5bx4Top2_SigLIP
|
| 14 |
+
|
| 15 |
+
This model is a fine-tuned version of [Nhanvi282/LLaVA_Qwen2.5_0.5b_SigLIPStage2](https://huggingface.co/Nhanvi282/LLaVA_Qwen2.5_0.5b_SigLIPStage2) on an unknown dataset.
|
| 16 |
+
|
| 17 |
+
## Model description
|
| 18 |
+
|
| 19 |
+
More information needed
|
| 20 |
+
|
| 21 |
+
## Intended uses & limitations
|
| 22 |
+
|
| 23 |
+
More information needed
|
| 24 |
+
|
| 25 |
+
## Training and evaluation data
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Training procedure
|
| 30 |
+
|
| 31 |
+
### Training hyperparameters
|
| 32 |
+
|
| 33 |
+
The following hyperparameters were used during training:
|
| 34 |
+
- learning_rate: 2e-06
|
| 35 |
+
- train_batch_size: 2
|
| 36 |
+
- eval_batch_size: 2
|
| 37 |
+
- seed: 42
|
| 38 |
+
- distributed_type: multi-GPU
|
| 39 |
+
- num_devices: 2
|
| 40 |
+
- gradient_accumulation_steps: 16
|
| 41 |
+
- total_train_batch_size: 64
|
| 42 |
+
- total_eval_batch_size: 4
|
| 43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
+
- lr_scheduler_type: cosine
|
| 45 |
+
- lr_scheduler_warmup_ratio: 0.03
|
| 46 |
+
- num_epochs: 4.0
|
| 47 |
+
- mixed_precision_training: Native AMP
|
| 48 |
+
|
| 49 |
+
### Training results
|
| 50 |
+
|
| 51 |
+
|
| 52 |
+
|
| 53 |
+
### Framework versions
|
| 54 |
+
|
| 55 |
+
- Transformers 4.37.0
|
| 56 |
+
- Pytorch 2.0.1+cu117
|
| 57 |
+
- Datasets 3.0.1
|
| 58 |
+
- Tokenizers 0.15.1
|
added_tokens.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</tool_call>": 151658,
|
| 3 |
+
"<tool_call>": 151657,
|
| 4 |
+
"<|box_end|>": 151649,
|
| 5 |
+
"<|box_start|>": 151648,
|
| 6 |
+
"<|endoftext|>": 151643,
|
| 7 |
+
"<|extra_0|>": 151665,
|
| 8 |
+
"<|file_sep|>": 151664,
|
| 9 |
+
"<|fim_middle|>": 151660,
|
| 10 |
+
"<|fim_pad|>": 151662,
|
| 11 |
+
"<|fim_prefix|>": 151659,
|
| 12 |
+
"<|fim_suffix|>": 151661,
|
| 13 |
+
"<|im_end|>": 151645,
|
| 14 |
+
"<|im_start|>": 151644,
|
| 15 |
+
"<|image_pad|>": 151655,
|
| 16 |
+
"<|object_ref_end|>": 151647,
|
| 17 |
+
"<|object_ref_start|>": 151646,
|
| 18 |
+
"<|quad_end|>": 151651,
|
| 19 |
+
"<|quad_start|>": 151650,
|
| 20 |
+
"<|repo_name|>": 151663,
|
| 21 |
+
"<|video_pad|>": 151656,
|
| 22 |
+
"<|vision_end|>": 151653,
|
| 23 |
+
"<|vision_pad|>": 151654,
|
| 24 |
+
"<|vision_start|>": 151652
|
| 25 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,94 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Nhanvi282/LLaVA_Qwen2.5_0.5b_SigLIPStage2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MoELLaVAQwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"freeze_mm_mlp_adapter": false,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 896,
|
| 12 |
+
"image_aspect_ratio": "pad",
|
| 13 |
+
"image_projector_type": "mlp2x_gelu",
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 4864,
|
| 16 |
+
"lora": {},
|
| 17 |
+
"max_position_embeddings": 32768,
|
| 18 |
+
"max_window_layers": 24,
|
| 19 |
+
"mm_hidden_size": 768,
|
| 20 |
+
"mm_image_tower": "google/siglip-base-patch16-256-multilingual",
|
| 21 |
+
"mm_projector_lr": null,
|
| 22 |
+
"mm_use_im_patch_token": false,
|
| 23 |
+
"mm_use_im_start_end": false,
|
| 24 |
+
"mm_video_tower": null,
|
| 25 |
+
"mm_vision_select_feature": "patch",
|
| 26 |
+
"mm_vision_select_layer": -2,
|
| 27 |
+
"model_type": "moe_llava_qwen2",
|
| 28 |
+
"moe": {
|
| 29 |
+
"capacity_factor": 1.5,
|
| 30 |
+
"ep_size": 1,
|
| 31 |
+
"eval_capacity_factor": 2.0,
|
| 32 |
+
"min_capacity": 0,
|
| 33 |
+
"moe_enable": true,
|
| 34 |
+
"moe_layers_idx": [
|
| 35 |
+
0,
|
| 36 |
+
2,
|
| 37 |
+
4,
|
| 38 |
+
6,
|
| 39 |
+
8,
|
| 40 |
+
10,
|
| 41 |
+
12,
|
| 42 |
+
14,
|
| 43 |
+
16,
|
| 44 |
+
18,
|
| 45 |
+
20,
|
| 46 |
+
22
|
| 47 |
+
],
|
| 48 |
+
"moe_mode": "sparse",
|
| 49 |
+
"num_experts": [
|
| 50 |
+
4,
|
| 51 |
+
4,
|
| 52 |
+
4,
|
| 53 |
+
4,
|
| 54 |
+
4,
|
| 55 |
+
4,
|
| 56 |
+
4,
|
| 57 |
+
4,
|
| 58 |
+
4,
|
| 59 |
+
4,
|
| 60 |
+
4,
|
| 61 |
+
4
|
| 62 |
+
],
|
| 63 |
+
"router_aux_loss_coef": 0.01,
|
| 64 |
+
"top_k_experts": 2,
|
| 65 |
+
"train_modules": [
|
| 66 |
+
"mlp.gate_proj",
|
| 67 |
+
"mlp.up_proj",
|
| 68 |
+
"mlp.down_proj",
|
| 69 |
+
"wg"
|
| 70 |
+
],
|
| 71 |
+
"use_residual": false
|
| 72 |
+
},
|
| 73 |
+
"num_attention_heads": 14,
|
| 74 |
+
"num_hidden_layers": 24,
|
| 75 |
+
"num_key_value_heads": 2,
|
| 76 |
+
"pad_token_id": 151665,
|
| 77 |
+
"rms_norm_eps": 1e-06,
|
| 78 |
+
"rope_theta": 1000000.0,
|
| 79 |
+
"sliding_window": 32768,
|
| 80 |
+
"tie_word_embeddings": true,
|
| 81 |
+
"tokenizer_padding_side": "right",
|
| 82 |
+
"torch_dtype": "float16",
|
| 83 |
+
"transformers_version": "4.37.0",
|
| 84 |
+
"tune_mm_mlp_adapter": false,
|
| 85 |
+
"use_cache": true,
|
| 86 |
+
"use_mm_proj": true,
|
| 87 |
+
"use_mrope": false,
|
| 88 |
+
"use_sliding_window": false,
|
| 89 |
+
"video_global_proj": false,
|
| 90 |
+
"video_projector_type": "linear",
|
| 91 |
+
"video_spatial_proj": false,
|
| 92 |
+
"video_temproal_proj": false,
|
| 93 |
+
"vocab_size": 151936
|
| 94 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.37.0"
|
| 6 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e465dfec31274be77a9d4baa978e1bd8727f19595065f505578ae62dc2e0d6c
|
| 3 |
+
size 4781336392
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a12ca5b9bcb4f74cf29e5a05920b4d37faa4efef3ca60aea8338ce01c6e95d71
|
| 3 |
+
size 2118541325
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>",
|
| 5 |
+
"<|object_ref_start|>",
|
| 6 |
+
"<|object_ref_end|>",
|
| 7 |
+
"<|box_start|>",
|
| 8 |
+
"<|box_end|>",
|
| 9 |
+
"<|quad_start|>",
|
| 10 |
+
"<|quad_end|>",
|
| 11 |
+
"<|vision_start|>",
|
| 12 |
+
"<|vision_end|>",
|
| 13 |
+
"<|vision_pad|>",
|
| 14 |
+
"<|image_pad|>",
|
| 15 |
+
"<|video_pad|>"
|
| 16 |
+
],
|
| 17 |
+
"eos_token": {
|
| 18 |
+
"content": "<|endoftext|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
},
|
| 24 |
+
"pad_token": "<|extra_0|>",
|
| 25 |
+
"unk_token": {
|
| 26 |
+
"content": "<|extra_0|>",
|
| 27 |
+
"lstrip": false,
|
| 28 |
+
"normalized": false,
|
| 29 |
+
"rstrip": false,
|
| 30 |
+
"single_word": false
|
| 31 |
+
}
|
| 32 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
},
|
| 181 |
+
"151665": {
|
| 182 |
+
"content": "<|extra_0|>",
|
| 183 |
+
"lstrip": false,
|
| 184 |
+
"normalized": false,
|
| 185 |
+
"rstrip": false,
|
| 186 |
+
"single_word": false,
|
| 187 |
+
"special": true
|
| 188 |
+
}
|
| 189 |
+
},
|
| 190 |
+
"additional_special_tokens": [
|
| 191 |
+
"<|im_start|>",
|
| 192 |
+
"<|im_end|>",
|
| 193 |
+
"<|object_ref_start|>",
|
| 194 |
+
"<|object_ref_end|>",
|
| 195 |
+
"<|box_start|>",
|
| 196 |
+
"<|box_end|>",
|
| 197 |
+
"<|quad_start|>",
|
| 198 |
+
"<|quad_end|>",
|
| 199 |
+
"<|vision_start|>",
|
| 200 |
+
"<|vision_end|>",
|
| 201 |
+
"<|vision_pad|>",
|
| 202 |
+
"<|image_pad|>",
|
| 203 |
+
"<|video_pad|>"
|
| 204 |
+
],
|
| 205 |
+
"bos_token": null,
|
| 206 |
+
"chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] }}\n {%- else %}\n {{- 'You are a helpful assistant.' }}\n {%- endif %}\n {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0]['role'] == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n {%- else %}\n {{- '<|im_start|>system\\nYou are a helpful assistant.<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role }}\n {%- if message.content %}\n {{- '\\n' + message.content }}\n {%- endif %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '\\n<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
|
| 207 |
+
"clean_up_tokenization_spaces": false,
|
| 208 |
+
"eos_token": "<|endoftext|>",
|
| 209 |
+
"errors": "replace",
|
| 210 |
+
"model_max_length": 2048,
|
| 211 |
+
"pad_token": "<|extra_0|>",
|
| 212 |
+
"padding_side": "right",
|
| 213 |
+
"split_special_tokens": false,
|
| 214 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 215 |
+
"unk_token": "<|extra_0|>"
|
| 216 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3714 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.990251827782291,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1228,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 0.0,
|
| 14 |
+
"loss": 1.8061,
|
| 15 |
+
"step": 2
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 5.4054054054054056e-08,
|
| 20 |
+
"loss": 1.8466,
|
| 21 |
+
"step": 4
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.02,
|
| 25 |
+
"learning_rate": 1.0810810810810811e-07,
|
| 26 |
+
"loss": 1.9168,
|
| 27 |
+
"step": 6
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.03,
|
| 31 |
+
"learning_rate": 2.1621621621621622e-07,
|
| 32 |
+
"loss": 1.7304,
|
| 33 |
+
"step": 8
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.03,
|
| 37 |
+
"learning_rate": 3.243243243243243e-07,
|
| 38 |
+
"loss": 1.8955,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.04,
|
| 43 |
+
"learning_rate": 4.3243243243243244e-07,
|
| 44 |
+
"loss": 1.9283,
|
| 45 |
+
"step": 12
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05,
|
| 49 |
+
"learning_rate": 5.405405405405406e-07,
|
| 50 |
+
"loss": 1.9517,
|
| 51 |
+
"step": 14
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05,
|
| 55 |
+
"learning_rate": 5.945945945945947e-07,
|
| 56 |
+
"loss": 1.8203,
|
| 57 |
+
"step": 16
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.06,
|
| 61 |
+
"learning_rate": 7.027027027027027e-07,
|
| 62 |
+
"loss": 1.7982,
|
| 63 |
+
"step": 18
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.06,
|
| 67 |
+
"learning_rate": 8.108108108108108e-07,
|
| 68 |
+
"loss": 1.8773,
|
| 69 |
+
"step": 20
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.07,
|
| 73 |
+
"learning_rate": 9.18918918918919e-07,
|
| 74 |
+
"loss": 1.891,
|
| 75 |
+
"step": 22
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.08,
|
| 79 |
+
"learning_rate": 1.0270270270270269e-06,
|
| 80 |
+
"loss": 1.801,
|
| 81 |
+
"step": 24
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.08,
|
| 85 |
+
"learning_rate": 1.135135135135135e-06,
|
| 86 |
+
"loss": 1.8555,
|
| 87 |
+
"step": 26
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09,
|
| 91 |
+
"learning_rate": 1.2432432432432432e-06,
|
| 92 |
+
"loss": 1.708,
|
| 93 |
+
"step": 28
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1,
|
| 97 |
+
"learning_rate": 1.3513513513513513e-06,
|
| 98 |
+
"loss": 1.7204,
|
| 99 |
+
"step": 30
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.1,
|
| 103 |
+
"learning_rate": 1.4594594594594594e-06,
|
| 104 |
+
"loss": 1.7953,
|
| 105 |
+
"step": 32
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.11,
|
| 109 |
+
"learning_rate": 1.5675675675675676e-06,
|
| 110 |
+
"loss": 1.7796,
|
| 111 |
+
"step": 34
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.12,
|
| 115 |
+
"learning_rate": 1.6756756756756755e-06,
|
| 116 |
+
"loss": 1.7226,
|
| 117 |
+
"step": 36
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.12,
|
| 121 |
+
"learning_rate": 1.7837837837837836e-06,
|
| 122 |
+
"loss": 1.7392,
|
| 123 |
+
"step": 38
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.13,
|
| 127 |
+
"learning_rate": 1.8918918918918918e-06,
|
| 128 |
+
"loss": 1.8025,
|
| 129 |
+
"step": 40
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14,
|
| 133 |
+
"learning_rate": 2e-06,
|
| 134 |
+
"loss": 1.6815,
|
| 135 |
+
"step": 42
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.14,
|
| 139 |
+
"learning_rate": 1.9999860842950866e-06,
|
| 140 |
+
"loss": 1.7512,
|
| 141 |
+
"step": 44
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.15,
|
| 145 |
+
"learning_rate": 1.99994433756764e-06,
|
| 146 |
+
"loss": 1.7699,
|
| 147 |
+
"step": 46
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.16,
|
| 151 |
+
"learning_rate": 1.9998747609795305e-06,
|
| 152 |
+
"loss": 1.7307,
|
| 153 |
+
"step": 48
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.16,
|
| 157 |
+
"learning_rate": 1.999777356467173e-06,
|
| 158 |
+
"loss": 1.6833,
|
| 159 |
+
"step": 50
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.17,
|
| 163 |
+
"learning_rate": 1.9996521267414718e-06,
|
| 164 |
+
"loss": 1.6805,
|
| 165 |
+
"step": 52
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.18,
|
| 169 |
+
"learning_rate": 1.999499075287747e-06,
|
| 170 |
+
"loss": 1.6625,
|
| 171 |
+
"step": 54
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.18,
|
| 175 |
+
"learning_rate": 1.999318206365636e-06,
|
| 176 |
+
"loss": 1.6638,
|
| 177 |
+
"step": 56
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.19,
|
| 181 |
+
"learning_rate": 1.999109525008977e-06,
|
| 182 |
+
"loss": 1.7162,
|
| 183 |
+
"step": 58
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.19,
|
| 187 |
+
"learning_rate": 1.998873037025665e-06,
|
| 188 |
+
"loss": 1.5856,
|
| 189 |
+
"step": 60
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.2,
|
| 193 |
+
"learning_rate": 1.998608748997495e-06,
|
| 194 |
+
"loss": 1.6406,
|
| 195 |
+
"step": 62
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.21,
|
| 199 |
+
"learning_rate": 1.9983166682799748e-06,
|
| 200 |
+
"loss": 1.7659,
|
| 201 |
+
"step": 64
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.21,
|
| 205 |
+
"learning_rate": 1.997996803002123e-06,
|
| 206 |
+
"loss": 1.6523,
|
| 207 |
+
"step": 66
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.22,
|
| 211 |
+
"learning_rate": 1.9976491620662405e-06,
|
| 212 |
+
"loss": 1.7458,
|
| 213 |
+
"step": 68
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.23,
|
| 217 |
+
"learning_rate": 1.9972737551476655e-06,
|
| 218 |
+
"loss": 1.6734,
|
| 219 |
+
"step": 70
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.23,
|
| 223 |
+
"learning_rate": 1.9968705926945013e-06,
|
| 224 |
+
"loss": 1.5952,
|
| 225 |
+
"step": 72
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.24,
|
| 229 |
+
"learning_rate": 1.9964396859273274e-06,
|
| 230 |
+
"loss": 1.6496,
|
| 231 |
+
"step": 74
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.25,
|
| 235 |
+
"learning_rate": 1.9959810468388867e-06,
|
| 236 |
+
"loss": 1.6758,
|
| 237 |
+
"step": 76
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.25,
|
| 241 |
+
"learning_rate": 1.9954946881937524e-06,
|
| 242 |
+
"loss": 1.721,
|
| 243 |
+
"step": 78
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.26,
|
| 247 |
+
"learning_rate": 1.9949806235279697e-06,
|
| 248 |
+
"loss": 1.6712,
|
| 249 |
+
"step": 80
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.27,
|
| 253 |
+
"learning_rate": 1.9944388671486843e-06,
|
| 254 |
+
"loss": 1.6497,
|
| 255 |
+
"step": 82
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.27,
|
| 259 |
+
"learning_rate": 1.9938694341337393e-06,
|
| 260 |
+
"loss": 1.644,
|
| 261 |
+
"step": 84
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.28,
|
| 265 |
+
"learning_rate": 1.993272340331259e-06,
|
| 266 |
+
"loss": 1.6608,
|
| 267 |
+
"step": 86
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.29,
|
| 271 |
+
"learning_rate": 1.992647602359205e-06,
|
| 272 |
+
"loss": 1.6927,
|
| 273 |
+
"step": 88
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.29,
|
| 277 |
+
"learning_rate": 1.991995237604916e-06,
|
| 278 |
+
"loss": 1.6459,
|
| 279 |
+
"step": 90
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.3,
|
| 283 |
+
"learning_rate": 1.991315264224623e-06,
|
| 284 |
+
"loss": 1.7367,
|
| 285 |
+
"step": 92
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.31,
|
| 289 |
+
"learning_rate": 1.990607701142944e-06,
|
| 290 |
+
"loss": 1.6752,
|
| 291 |
+
"step": 94
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.31,
|
| 295 |
+
"learning_rate": 1.9898725680523566e-06,
|
| 296 |
+
"loss": 1.615,
|
| 297 |
+
"step": 96
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.32,
|
| 301 |
+
"learning_rate": 1.9891098854126515e-06,
|
| 302 |
+
"loss": 1.6755,
|
| 303 |
+
"step": 98
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.32,
|
| 307 |
+
"learning_rate": 1.988319674450362e-06,
|
| 308 |
+
"loss": 1.6614,
|
| 309 |
+
"step": 100
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.33,
|
| 313 |
+
"learning_rate": 1.9875019571581726e-06,
|
| 314 |
+
"loss": 1.6506,
|
| 315 |
+
"step": 102
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.34,
|
| 319 |
+
"learning_rate": 1.986656756294309e-06,
|
| 320 |
+
"loss": 1.6457,
|
| 321 |
+
"step": 104
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.34,
|
| 325 |
+
"learning_rate": 1.9857840953819023e-06,
|
| 326 |
+
"loss": 1.697,
|
| 327 |
+
"step": 106
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.35,
|
| 331 |
+
"learning_rate": 1.9848839987083364e-06,
|
| 332 |
+
"loss": 1.6985,
|
| 333 |
+
"step": 108
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.36,
|
| 337 |
+
"learning_rate": 1.983956491324571e-06,
|
| 338 |
+
"loss": 1.6595,
|
| 339 |
+
"step": 110
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.36,
|
| 343 |
+
"learning_rate": 1.9830015990444433e-06,
|
| 344 |
+
"loss": 1.6556,
|
| 345 |
+
"step": 112
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.37,
|
| 349 |
+
"learning_rate": 1.982019348443952e-06,
|
| 350 |
+
"loss": 1.7157,
|
| 351 |
+
"step": 114
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.38,
|
| 355 |
+
"learning_rate": 1.9810097668605167e-06,
|
| 356 |
+
"loss": 1.6435,
|
| 357 |
+
"step": 116
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.38,
|
| 361 |
+
"learning_rate": 1.9799728823922154e-06,
|
| 362 |
+
"loss": 1.7301,
|
| 363 |
+
"step": 118
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.39,
|
| 367 |
+
"learning_rate": 1.978908723897005e-06,
|
| 368 |
+
"loss": 1.7317,
|
| 369 |
+
"step": 120
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.4,
|
| 373 |
+
"learning_rate": 1.977817320991917e-06,
|
| 374 |
+
"loss": 1.6565,
|
| 375 |
+
"step": 122
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.4,
|
| 379 |
+
"learning_rate": 1.976698704052233e-06,
|
| 380 |
+
"loss": 1.6721,
|
| 381 |
+
"step": 124
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.41,
|
| 385 |
+
"learning_rate": 1.9755529042106393e-06,
|
| 386 |
+
"loss": 1.6775,
|
| 387 |
+
"step": 126
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.42,
|
| 391 |
+
"learning_rate": 1.97437995335636e-06,
|
| 392 |
+
"loss": 1.6714,
|
| 393 |
+
"step": 128
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.42,
|
| 397 |
+
"learning_rate": 1.9731798841342724e-06,
|
| 398 |
+
"loss": 1.6332,
|
| 399 |
+
"step": 130
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.43,
|
| 403 |
+
"learning_rate": 1.971952729943994e-06,
|
| 404 |
+
"loss": 1.6737,
|
| 405 |
+
"step": 132
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.44,
|
| 409 |
+
"learning_rate": 1.970698524938956e-06,
|
| 410 |
+
"loss": 1.715,
|
| 411 |
+
"step": 134
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.44,
|
| 415 |
+
"learning_rate": 1.9694173040254527e-06,
|
| 416 |
+
"loss": 1.5784,
|
| 417 |
+
"step": 136
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.45,
|
| 421 |
+
"learning_rate": 1.9681091028616676e-06,
|
| 422 |
+
"loss": 1.6485,
|
| 423 |
+
"step": 138
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.45,
|
| 427 |
+
"learning_rate": 1.9667739578566834e-06,
|
| 428 |
+
"loss": 1.6339,
|
| 429 |
+
"step": 140
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.46,
|
| 433 |
+
"learning_rate": 1.965411906169469e-06,
|
| 434 |
+
"loss": 1.6256,
|
| 435 |
+
"step": 142
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.47,
|
| 439 |
+
"learning_rate": 1.9640229857078413e-06,
|
| 440 |
+
"loss": 1.6671,
|
| 441 |
+
"step": 144
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.47,
|
| 445 |
+
"learning_rate": 1.9626072351274164e-06,
|
| 446 |
+
"loss": 1.5817,
|
| 447 |
+
"step": 146
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.48,
|
| 451 |
+
"learning_rate": 1.9611646938305286e-06,
|
| 452 |
+
"loss": 1.6282,
|
| 453 |
+
"step": 148
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.49,
|
| 457 |
+
"learning_rate": 1.9596954019651354e-06,
|
| 458 |
+
"loss": 1.5707,
|
| 459 |
+
"step": 150
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.49,
|
| 463 |
+
"learning_rate": 1.9581994004237012e-06,
|
| 464 |
+
"loss": 1.6142,
|
| 465 |
+
"step": 152
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5,
|
| 469 |
+
"learning_rate": 1.956676730842058e-06,
|
| 470 |
+
"loss": 1.7049,
|
| 471 |
+
"step": 154
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.51,
|
| 475 |
+
"learning_rate": 1.955127435598247e-06,
|
| 476 |
+
"loss": 1.7158,
|
| 477 |
+
"step": 156
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.51,
|
| 481 |
+
"learning_rate": 1.9535515578113387e-06,
|
| 482 |
+
"loss": 1.5922,
|
| 483 |
+
"step": 158
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.52,
|
| 487 |
+
"learning_rate": 1.9519491413402347e-06,
|
| 488 |
+
"loss": 1.6621,
|
| 489 |
+
"step": 160
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.53,
|
| 493 |
+
"learning_rate": 1.950320230782443e-06,
|
| 494 |
+
"loss": 1.736,
|
| 495 |
+
"step": 162
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.53,
|
| 499 |
+
"learning_rate": 1.9486648714728424e-06,
|
| 500 |
+
"loss": 1.6524,
|
| 501 |
+
"step": 164
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.54,
|
| 505 |
+
"learning_rate": 1.946983109482415e-06,
|
| 506 |
+
"loss": 1.6097,
|
| 507 |
+
"step": 166
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.55,
|
| 511 |
+
"learning_rate": 1.9452749916169685e-06,
|
| 512 |
+
"loss": 1.5994,
|
| 513 |
+
"step": 168
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.55,
|
| 517 |
+
"learning_rate": 1.9435405654158315e-06,
|
| 518 |
+
"loss": 1.6715,
|
| 519 |
+
"step": 170
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.56,
|
| 523 |
+
"learning_rate": 1.94177987915053e-06,
|
| 524 |
+
"loss": 1.7165,
|
| 525 |
+
"step": 172
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.57,
|
| 529 |
+
"learning_rate": 1.939992981823445e-06,
|
| 530 |
+
"loss": 1.6326,
|
| 531 |
+
"step": 174
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.57,
|
| 535 |
+
"learning_rate": 1.9381799231664492e-06,
|
| 536 |
+
"loss": 1.6364,
|
| 537 |
+
"step": 176
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.58,
|
| 541 |
+
"learning_rate": 1.93634075363952e-06,
|
| 542 |
+
"loss": 1.6556,
|
| 543 |
+
"step": 178
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.58,
|
| 547 |
+
"learning_rate": 1.934475524429339e-06,
|
| 548 |
+
"loss": 1.6754,
|
| 549 |
+
"step": 180
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.59,
|
| 553 |
+
"learning_rate": 1.9325842874478642e-06,
|
| 554 |
+
"loss": 1.6204,
|
| 555 |
+
"step": 182
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6,
|
| 559 |
+
"learning_rate": 1.9306670953308877e-06,
|
| 560 |
+
"loss": 1.7316,
|
| 561 |
+
"step": 184
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.6,
|
| 565 |
+
"learning_rate": 1.928724001436568e-06,
|
| 566 |
+
"loss": 1.6321,
|
| 567 |
+
"step": 186
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.61,
|
| 571 |
+
"learning_rate": 1.926755059843949e-06,
|
| 572 |
+
"loss": 1.6595,
|
| 573 |
+
"step": 188
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.62,
|
| 577 |
+
"learning_rate": 1.9247603253514502e-06,
|
| 578 |
+
"loss": 1.669,
|
| 579 |
+
"step": 190
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.62,
|
| 583 |
+
"learning_rate": 1.922739853475345e-06,
|
| 584 |
+
"loss": 1.6579,
|
| 585 |
+
"step": 192
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.63,
|
| 589 |
+
"learning_rate": 1.920693700448214e-06,
|
| 590 |
+
"loss": 1.6517,
|
| 591 |
+
"step": 194
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.64,
|
| 595 |
+
"learning_rate": 1.918621923217381e-06,
|
| 596 |
+
"loss": 1.6283,
|
| 597 |
+
"step": 196
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.64,
|
| 601 |
+
"learning_rate": 1.916524579443327e-06,
|
| 602 |
+
"loss": 1.7147,
|
| 603 |
+
"step": 198
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.65,
|
| 607 |
+
"learning_rate": 1.914401727498086e-06,
|
| 608 |
+
"loss": 1.6493,
|
| 609 |
+
"step": 200
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.66,
|
| 613 |
+
"learning_rate": 1.912253426463621e-06,
|
| 614 |
+
"loss": 1.5777,
|
| 615 |
+
"step": 202
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.66,
|
| 619 |
+
"learning_rate": 1.910079736130178e-06,
|
| 620 |
+
"loss": 1.5881,
|
| 621 |
+
"step": 204
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.67,
|
| 625 |
+
"learning_rate": 1.9078807169946233e-06,
|
| 626 |
+
"loss": 1.6019,
|
| 627 |
+
"step": 206
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.68,
|
| 631 |
+
"learning_rate": 1.9056564302587605e-06,
|
| 632 |
+
"loss": 1.7024,
|
| 633 |
+
"step": 208
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.68,
|
| 637 |
+
"learning_rate": 1.9034069378276248e-06,
|
| 638 |
+
"loss": 1.6262,
|
| 639 |
+
"step": 210
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.69,
|
| 643 |
+
"learning_rate": 1.901132302307762e-06,
|
| 644 |
+
"loss": 1.6722,
|
| 645 |
+
"step": 212
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.7,
|
| 649 |
+
"learning_rate": 1.8988325870054855e-06,
|
| 650 |
+
"loss": 1.6399,
|
| 651 |
+
"step": 214
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.7,
|
| 655 |
+
"learning_rate": 1.8965078559251141e-06,
|
| 656 |
+
"loss": 1.676,
|
| 657 |
+
"step": 216
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.71,
|
| 661 |
+
"learning_rate": 1.8941581737671916e-06,
|
| 662 |
+
"loss": 1.6346,
|
| 663 |
+
"step": 218
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.71,
|
| 667 |
+
"learning_rate": 1.8917836059266853e-06,
|
| 668 |
+
"loss": 1.6168,
|
| 669 |
+
"step": 220
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.72,
|
| 673 |
+
"learning_rate": 1.8893842184911652e-06,
|
| 674 |
+
"loss": 1.6366,
|
| 675 |
+
"step": 222
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.73,
|
| 679 |
+
"learning_rate": 1.8869600782389672e-06,
|
| 680 |
+
"loss": 1.6633,
|
| 681 |
+
"step": 224
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.73,
|
| 685 |
+
"learning_rate": 1.8845112526373314e-06,
|
| 686 |
+
"loss": 1.6987,
|
| 687 |
+
"step": 226
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.74,
|
| 691 |
+
"learning_rate": 1.8820378098405269e-06,
|
| 692 |
+
"loss": 1.6381,
|
| 693 |
+
"step": 228
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.75,
|
| 697 |
+
"learning_rate": 1.8795398186879542e-06,
|
| 698 |
+
"loss": 1.6275,
|
| 699 |
+
"step": 230
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.75,
|
| 703 |
+
"learning_rate": 1.8770173487022287e-06,
|
| 704 |
+
"loss": 1.5933,
|
| 705 |
+
"step": 232
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.76,
|
| 709 |
+
"learning_rate": 1.874470470087246e-06,
|
| 710 |
+
"loss": 1.5811,
|
| 711 |
+
"step": 234
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.77,
|
| 715 |
+
"learning_rate": 1.8718992537262289e-06,
|
| 716 |
+
"loss": 1.6296,
|
| 717 |
+
"step": 236
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.77,
|
| 721 |
+
"learning_rate": 1.8693037711797536e-06,
|
| 722 |
+
"loss": 1.6455,
|
| 723 |
+
"step": 238
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.78,
|
| 727 |
+
"learning_rate": 1.8666840946837588e-06,
|
| 728 |
+
"loss": 1.6686,
|
| 729 |
+
"step": 240
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.79,
|
| 733 |
+
"learning_rate": 1.865365206462597e-06,
|
| 734 |
+
"loss": 1.6371,
|
| 735 |
+
"step": 242
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.79,
|
| 739 |
+
"learning_rate": 1.8627093759571097e-06,
|
| 740 |
+
"loss": 1.6794,
|
| 741 |
+
"step": 244
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.8,
|
| 745 |
+
"learning_rate": 1.8600295350334187e-06,
|
| 746 |
+
"loss": 1.6423,
|
| 747 |
+
"step": 246
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.81,
|
| 751 |
+
"learning_rate": 1.8573257582752747e-06,
|
| 752 |
+
"loss": 1.6921,
|
| 753 |
+
"step": 248
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.81,
|
| 757 |
+
"learning_rate": 1.8545981209325974e-06,
|
| 758 |
+
"loss": 1.5967,
|
| 759 |
+
"step": 250
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.82,
|
| 763 |
+
"learning_rate": 1.8518466989193793e-06,
|
| 764 |
+
"loss": 1.6361,
|
| 765 |
+
"step": 252
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.83,
|
| 769 |
+
"learning_rate": 1.8490715688115736e-06,
|
| 770 |
+
"loss": 1.6242,
|
| 771 |
+
"step": 254
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.83,
|
| 775 |
+
"learning_rate": 1.846272807844964e-06,
|
| 776 |
+
"loss": 1.7019,
|
| 777 |
+
"step": 256
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.84,
|
| 781 |
+
"learning_rate": 1.8434504939130142e-06,
|
| 782 |
+
"loss": 1.6015,
|
| 783 |
+
"step": 258
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.84,
|
| 787 |
+
"learning_rate": 1.8406047055646997e-06,
|
| 788 |
+
"loss": 1.6254,
|
| 789 |
+
"step": 260
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.85,
|
| 793 |
+
"learning_rate": 1.837735522002322e-06,
|
| 794 |
+
"loss": 1.6742,
|
| 795 |
+
"step": 262
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.86,
|
| 799 |
+
"learning_rate": 1.8348430230793053e-06,
|
| 800 |
+
"loss": 1.5701,
|
| 801 |
+
"step": 264
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.86,
|
| 805 |
+
"learning_rate": 1.831927289297972e-06,
|
| 806 |
+
"loss": 1.6758,
|
| 807 |
+
"step": 266
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.87,
|
| 811 |
+
"learning_rate": 1.828988401807304e-06,
|
| 812 |
+
"loss": 1.5573,
|
| 813 |
+
"step": 268
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.88,
|
| 817 |
+
"learning_rate": 1.8260264424006842e-06,
|
| 818 |
+
"loss": 1.573,
|
| 819 |
+
"step": 270
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.88,
|
| 823 |
+
"learning_rate": 1.8230414935136173e-06,
|
| 824 |
+
"loss": 1.6287,
|
| 825 |
+
"step": 272
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.89,
|
| 829 |
+
"learning_rate": 1.8200336382214404e-06,
|
| 830 |
+
"loss": 1.6358,
|
| 831 |
+
"step": 274
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.9,
|
| 835 |
+
"learning_rate": 1.8170029602370065e-06,
|
| 836 |
+
"loss": 1.6163,
|
| 837 |
+
"step": 276
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.9,
|
| 841 |
+
"learning_rate": 1.8139495439083566e-06,
|
| 842 |
+
"loss": 1.6167,
|
| 843 |
+
"step": 278
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.91,
|
| 847 |
+
"learning_rate": 1.8108734742163714e-06,
|
| 848 |
+
"loss": 1.5929,
|
| 849 |
+
"step": 280
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.92,
|
| 853 |
+
"learning_rate": 1.807774836772408e-06,
|
| 854 |
+
"loss": 1.6121,
|
| 855 |
+
"step": 282
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.92,
|
| 859 |
+
"learning_rate": 1.8046537178159142e-06,
|
| 860 |
+
"loss": 1.643,
|
| 861 |
+
"step": 284
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.93,
|
| 865 |
+
"learning_rate": 1.8015102042120314e-06,
|
| 866 |
+
"loss": 1.6437,
|
| 867 |
+
"step": 286
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.94,
|
| 871 |
+
"learning_rate": 1.7983443834491748e-06,
|
| 872 |
+
"loss": 1.582,
|
| 873 |
+
"step": 288
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.94,
|
| 877 |
+
"learning_rate": 1.7951563436365994e-06,
|
| 878 |
+
"loss": 1.6458,
|
| 879 |
+
"step": 290
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.95,
|
| 883 |
+
"learning_rate": 1.791946173501948e-06,
|
| 884 |
+
"loss": 1.7392,
|
| 885 |
+
"step": 292
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.96,
|
| 889 |
+
"learning_rate": 1.7887139623887809e-06,
|
| 890 |
+
"loss": 1.5767,
|
| 891 |
+
"step": 294
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.96,
|
| 895 |
+
"learning_rate": 1.7854598002540904e-06,
|
| 896 |
+
"loss": 1.6668,
|
| 897 |
+
"step": 296
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.97,
|
| 901 |
+
"learning_rate": 1.7821837776657967e-06,
|
| 902 |
+
"loss": 1.5234,
|
| 903 |
+
"step": 298
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.97,
|
| 907 |
+
"learning_rate": 1.7788859858002267e-06,
|
| 908 |
+
"loss": 1.6067,
|
| 909 |
+
"step": 300
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.98,
|
| 913 |
+
"learning_rate": 1.7755665164395776e-06,
|
| 914 |
+
"loss": 1.7513,
|
| 915 |
+
"step": 302
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.99,
|
| 919 |
+
"learning_rate": 1.7722254619693617e-06,
|
| 920 |
+
"loss": 1.6095,
|
| 921 |
+
"step": 304
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.99,
|
| 925 |
+
"learning_rate": 1.768862915375835e-06,
|
| 926 |
+
"loss": 1.5938,
|
| 927 |
+
"step": 306
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.0,
|
| 931 |
+
"learning_rate": 1.7654789702434093e-06,
|
| 932 |
+
"loss": 1.6113,
|
| 933 |
+
"step": 308
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.01,
|
| 937 |
+
"learning_rate": 1.7620737207520498e-06,
|
| 938 |
+
"loss": 1.5611,
|
| 939 |
+
"step": 310
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 1.01,
|
| 943 |
+
"learning_rate": 1.7586472616746493e-06,
|
| 944 |
+
"loss": 1.524,
|
| 945 |
+
"step": 312
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 1.02,
|
| 949 |
+
"learning_rate": 1.7551996883743952e-06,
|
| 950 |
+
"loss": 1.6084,
|
| 951 |
+
"step": 314
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 1.03,
|
| 955 |
+
"learning_rate": 1.751731096802113e-06,
|
| 956 |
+
"loss": 1.5283,
|
| 957 |
+
"step": 316
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.03,
|
| 961 |
+
"learning_rate": 1.7482415834935958e-06,
|
| 962 |
+
"loss": 1.4722,
|
| 963 |
+
"step": 318
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.04,
|
| 967 |
+
"learning_rate": 1.7447312455669189e-06,
|
| 968 |
+
"loss": 1.5559,
|
| 969 |
+
"step": 320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.05,
|
| 973 |
+
"learning_rate": 1.7412001807197361e-06,
|
| 974 |
+
"loss": 1.5944,
|
| 975 |
+
"step": 322
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.05,
|
| 979 |
+
"learning_rate": 1.73764848722656e-06,
|
| 980 |
+
"loss": 1.4856,
|
| 981 |
+
"step": 324
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 1.06,
|
| 985 |
+
"learning_rate": 1.7340762639360275e-06,
|
| 986 |
+
"loss": 1.5987,
|
| 987 |
+
"step": 326
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 1.07,
|
| 991 |
+
"learning_rate": 1.7304836102681493e-06,
|
| 992 |
+
"loss": 1.6132,
|
| 993 |
+
"step": 328
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.07,
|
| 997 |
+
"learning_rate": 1.726870626211542e-06,
|
| 998 |
+
"loss": 1.6389,
|
| 999 |
+
"step": 330
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.08,
|
| 1003 |
+
"learning_rate": 1.7232374123206456e-06,
|
| 1004 |
+
"loss": 1.5557,
|
| 1005 |
+
"step": 332
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.09,
|
| 1009 |
+
"learning_rate": 1.719584069712925e-06,
|
| 1010 |
+
"loss": 1.6182,
|
| 1011 |
+
"step": 334
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.09,
|
| 1015 |
+
"learning_rate": 1.7159107000660548e-06,
|
| 1016 |
+
"loss": 1.5269,
|
| 1017 |
+
"step": 336
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.1,
|
| 1021 |
+
"learning_rate": 1.7122174056150923e-06,
|
| 1022 |
+
"loss": 1.575,
|
| 1023 |
+
"step": 338
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1.1,
|
| 1027 |
+
"learning_rate": 1.708504289149628e-06,
|
| 1028 |
+
"loss": 1.5117,
|
| 1029 |
+
"step": 340
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 1.11,
|
| 1033 |
+
"learning_rate": 1.7047714540109284e-06,
|
| 1034 |
+
"loss": 1.5364,
|
| 1035 |
+
"step": 342
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 1.12,
|
| 1039 |
+
"learning_rate": 1.7010190040890577e-06,
|
| 1040 |
+
"loss": 1.5653,
|
| 1041 |
+
"step": 344
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.12,
|
| 1045 |
+
"learning_rate": 1.697247043819988e-06,
|
| 1046 |
+
"loss": 1.5064,
|
| 1047 |
+
"step": 346
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.13,
|
| 1051 |
+
"learning_rate": 1.6934556781826914e-06,
|
| 1052 |
+
"loss": 1.5191,
|
| 1053 |
+
"step": 348
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.14,
|
| 1057 |
+
"learning_rate": 1.6896450126962182e-06,
|
| 1058 |
+
"loss": 1.5189,
|
| 1059 |
+
"step": 350
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.14,
|
| 1063 |
+
"learning_rate": 1.6858151534167616e-06,
|
| 1064 |
+
"loss": 1.5126,
|
| 1065 |
+
"step": 352
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 1.15,
|
| 1069 |
+
"learning_rate": 1.6819662069347053e-06,
|
| 1070 |
+
"loss": 1.6049,
|
| 1071 |
+
"step": 354
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 1.16,
|
| 1075 |
+
"learning_rate": 1.6780982803716554e-06,
|
| 1076 |
+
"loss": 1.5877,
|
| 1077 |
+
"step": 356
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.16,
|
| 1081 |
+
"learning_rate": 1.6742114813774618e-06,
|
| 1082 |
+
"loss": 1.5174,
|
| 1083 |
+
"step": 358
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.17,
|
| 1087 |
+
"learning_rate": 1.67030591812722e-06,
|
| 1088 |
+
"loss": 1.5522,
|
| 1089 |
+
"step": 360
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.18,
|
| 1093 |
+
"learning_rate": 1.6663816993182612e-06,
|
| 1094 |
+
"loss": 1.5089,
|
| 1095 |
+
"step": 362
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.18,
|
| 1099 |
+
"learning_rate": 1.6624389341671278e-06,
|
| 1100 |
+
"loss": 1.5059,
|
| 1101 |
+
"step": 364
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.19,
|
| 1105 |
+
"learning_rate": 1.6584777324065322e-06,
|
| 1106 |
+
"loss": 1.6044,
|
| 1107 |
+
"step": 366
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 1.2,
|
| 1111 |
+
"learning_rate": 1.6544982042823045e-06,
|
| 1112 |
+
"loss": 1.5062,
|
| 1113 |
+
"step": 368
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 1.2,
|
| 1117 |
+
"learning_rate": 1.6505004605503223e-06,
|
| 1118 |
+
"loss": 1.5777,
|
| 1119 |
+
"step": 370
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.21,
|
| 1123 |
+
"learning_rate": 1.6464846124734297e-06,
|
| 1124 |
+
"loss": 1.6269,
|
| 1125 |
+
"step": 372
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 1.22,
|
| 1129 |
+
"learning_rate": 1.642450771818341e-06,
|
| 1130 |
+
"loss": 1.4963,
|
| 1131 |
+
"step": 374
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.22,
|
| 1135 |
+
"learning_rate": 1.638399050852528e-06,
|
| 1136 |
+
"loss": 1.5251,
|
| 1137 |
+
"step": 376
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.23,
|
| 1141 |
+
"learning_rate": 1.6343295623410977e-06,
|
| 1142 |
+
"loss": 1.5058,
|
| 1143 |
+
"step": 378
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.23,
|
| 1147 |
+
"learning_rate": 1.630242419543653e-06,
|
| 1148 |
+
"loss": 1.5084,
|
| 1149 |
+
"step": 380
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 1.24,
|
| 1153 |
+
"learning_rate": 1.6261377362111396e-06,
|
| 1154 |
+
"loss": 1.5251,
|
| 1155 |
+
"step": 382
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 1.25,
|
| 1159 |
+
"learning_rate": 1.6220156265826814e-06,
|
| 1160 |
+
"loss": 1.5407,
|
| 1161 |
+
"step": 384
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.25,
|
| 1165 |
+
"learning_rate": 1.617876205382401e-06,
|
| 1166 |
+
"loss": 1.5053,
|
| 1167 |
+
"step": 386
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 1.26,
|
| 1171 |
+
"learning_rate": 1.6137195878162267e-06,
|
| 1172 |
+
"loss": 1.4987,
|
| 1173 |
+
"step": 388
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.27,
|
| 1177 |
+
"learning_rate": 1.6095458895686851e-06,
|
| 1178 |
+
"loss": 1.5446,
|
| 1179 |
+
"step": 390
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.27,
|
| 1183 |
+
"learning_rate": 1.6053552267996824e-06,
|
| 1184 |
+
"loss": 1.6929,
|
| 1185 |
+
"step": 392
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.28,
|
| 1189 |
+
"learning_rate": 1.601147716141272e-06,
|
| 1190 |
+
"loss": 1.6088,
|
| 1191 |
+
"step": 394
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 1.29,
|
| 1195 |
+
"learning_rate": 1.596923474694407e-06,
|
| 1196 |
+
"loss": 1.5453,
|
| 1197 |
+
"step": 396
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 1.29,
|
| 1201 |
+
"learning_rate": 1.5926826200256832e-06,
|
| 1202 |
+
"loss": 1.5575,
|
| 1203 |
+
"step": 398
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.3,
|
| 1207 |
+
"learning_rate": 1.5884252701640634e-06,
|
| 1208 |
+
"loss": 1.5664,
|
| 1209 |
+
"step": 400
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 1.31,
|
| 1213 |
+
"learning_rate": 1.584151543597598e-06,
|
| 1214 |
+
"loss": 1.5568,
|
| 1215 |
+
"step": 402
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.31,
|
| 1219 |
+
"learning_rate": 1.5798615592701212e-06,
|
| 1220 |
+
"loss": 1.5608,
|
| 1221 |
+
"step": 404
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.32,
|
| 1225 |
+
"learning_rate": 1.5755554365779455e-06,
|
| 1226 |
+
"loss": 1.592,
|
| 1227 |
+
"step": 406
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.33,
|
| 1231 |
+
"learning_rate": 1.5712332953665366e-06,
|
| 1232 |
+
"loss": 1.4785,
|
| 1233 |
+
"step": 408
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 1.33,
|
| 1237 |
+
"learning_rate": 1.5668952559271775e-06,
|
| 1238 |
+
"loss": 1.5541,
|
| 1239 |
+
"step": 410
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 1.34,
|
| 1243 |
+
"learning_rate": 1.5625414389936218e-06,
|
| 1244 |
+
"loss": 1.5963,
|
| 1245 |
+
"step": 412
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.35,
|
| 1249 |
+
"learning_rate": 1.558171965738733e-06,
|
| 1250 |
+
"loss": 1.5598,
|
| 1251 |
+
"step": 414
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 1.35,
|
| 1255 |
+
"learning_rate": 1.5537869577711118e-06,
|
| 1256 |
+
"loss": 1.5968,
|
| 1257 |
+
"step": 416
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.36,
|
| 1261 |
+
"learning_rate": 1.5515886663910286e-06,
|
| 1262 |
+
"loss": 1.6047,
|
| 1263 |
+
"step": 418
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.36,
|
| 1267 |
+
"learning_rate": 1.5471805853152792e-06,
|
| 1268 |
+
"loss": 1.5337,
|
| 1269 |
+
"step": 420
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.37,
|
| 1273 |
+
"learning_rate": 1.542757275432411e-06,
|
| 1274 |
+
"loss": 1.5303,
|
| 1275 |
+
"step": 422
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 1.38,
|
| 1279 |
+
"learning_rate": 1.5383188598493732e-06,
|
| 1280 |
+
"loss": 1.5346,
|
| 1281 |
+
"step": 424
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 1.38,
|
| 1285 |
+
"learning_rate": 1.5338654620935297e-06,
|
| 1286 |
+
"loss": 1.5407,
|
| 1287 |
+
"step": 426
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.39,
|
| 1291 |
+
"learning_rate": 1.5293972061092185e-06,
|
| 1292 |
+
"loss": 1.5472,
|
| 1293 |
+
"step": 428
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 1.4,
|
| 1297 |
+
"learning_rate": 1.5249142162543025e-06,
|
| 1298 |
+
"loss": 1.5588,
|
| 1299 |
+
"step": 430
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.4,
|
| 1303 |
+
"learning_rate": 1.52041661729671e-06,
|
| 1304 |
+
"loss": 1.6215,
|
| 1305 |
+
"step": 432
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.41,
|
| 1309 |
+
"learning_rate": 1.515904534410961e-06,
|
| 1310 |
+
"loss": 1.424,
|
| 1311 |
+
"step": 434
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.42,
|
| 1315 |
+
"learning_rate": 1.5113780931746835e-06,
|
| 1316 |
+
"loss": 1.6234,
|
| 1317 |
+
"step": 436
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 1.42,
|
| 1321 |
+
"learning_rate": 1.5068374195651182e-06,
|
| 1322 |
+
"loss": 1.4829,
|
| 1323 |
+
"step": 438
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 1.43,
|
| 1327 |
+
"learning_rate": 1.5022826399556133e-06,
|
| 1328 |
+
"loss": 1.473,
|
| 1329 |
+
"step": 440
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 1.44,
|
| 1333 |
+
"learning_rate": 1.497713881112107e-06,
|
| 1334 |
+
"loss": 1.5315,
|
| 1335 |
+
"step": 442
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 1.44,
|
| 1339 |
+
"learning_rate": 1.4931312701895989e-06,
|
| 1340 |
+
"loss": 1.5482,
|
| 1341 |
+
"step": 444
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.45,
|
| 1345 |
+
"learning_rate": 1.4885349347286115e-06,
|
| 1346 |
+
"loss": 1.5099,
|
| 1347 |
+
"step": 446
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.46,
|
| 1351 |
+
"learning_rate": 1.4839250026516415e-06,
|
| 1352 |
+
"loss": 1.5285,
|
| 1353 |
+
"step": 448
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.46,
|
| 1357 |
+
"learning_rate": 1.479301602259597e-06,
|
| 1358 |
+
"loss": 1.5137,
|
| 1359 |
+
"step": 450
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 1.47,
|
| 1363 |
+
"learning_rate": 1.474664862228229e-06,
|
| 1364 |
+
"loss": 1.6417,
|
| 1365 |
+
"step": 452
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 1.48,
|
| 1369 |
+
"learning_rate": 1.4700149116045504e-06,
|
| 1370 |
+
"loss": 1.4725,
|
| 1371 |
+
"step": 454
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 1.48,
|
| 1375 |
+
"learning_rate": 1.4653518798032422e-06,
|
| 1376 |
+
"loss": 1.5678,
|
| 1377 |
+
"step": 456
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.49,
|
| 1381 |
+
"learning_rate": 1.4606758966030534e-06,
|
| 1382 |
+
"loss": 1.4743,
|
| 1383 |
+
"step": 458
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.49,
|
| 1387 |
+
"learning_rate": 1.455987092143189e-06,
|
| 1388 |
+
"loss": 1.5511,
|
| 1389 |
+
"step": 460
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.5,
|
| 1393 |
+
"learning_rate": 1.4512855969196872e-06,
|
| 1394 |
+
"loss": 1.5214,
|
| 1395 |
+
"step": 462
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.51,
|
| 1399 |
+
"learning_rate": 1.4465715417817888e-06,
|
| 1400 |
+
"loss": 1.5138,
|
| 1401 |
+
"step": 464
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 1.51,
|
| 1405 |
+
"learning_rate": 1.4418450579282939e-06,
|
| 1406 |
+
"loss": 1.5529,
|
| 1407 |
+
"step": 466
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 1.52,
|
| 1411 |
+
"learning_rate": 1.4371062769039117e-06,
|
| 1412 |
+
"loss": 1.5379,
|
| 1413 |
+
"step": 468
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.53,
|
| 1417 |
+
"learning_rate": 1.4323553305955997e-06,
|
| 1418 |
+
"loss": 1.5192,
|
| 1419 |
+
"step": 470
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 1.53,
|
| 1423 |
+
"learning_rate": 1.4275923512288905e-06,
|
| 1424 |
+
"loss": 1.6183,
|
| 1425 |
+
"step": 472
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.54,
|
| 1429 |
+
"learning_rate": 1.4228174713642158e-06,
|
| 1430 |
+
"loss": 1.5813,
|
| 1431 |
+
"step": 474
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.55,
|
| 1435 |
+
"learning_rate": 1.4180308238932135e-06,
|
| 1436 |
+
"loss": 1.6012,
|
| 1437 |
+
"step": 476
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.55,
|
| 1441 |
+
"learning_rate": 1.4132325420350313e-06,
|
| 1442 |
+
"loss": 1.5235,
|
| 1443 |
+
"step": 478
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 1.56,
|
| 1447 |
+
"learning_rate": 1.4084227593326176e-06,
|
| 1448 |
+
"loss": 1.4738,
|
| 1449 |
+
"step": 480
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 1.57,
|
| 1453 |
+
"learning_rate": 1.4036016096490064e-06,
|
| 1454 |
+
"loss": 1.5581,
|
| 1455 |
+
"step": 482
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.57,
|
| 1459 |
+
"learning_rate": 1.3987692271635908e-06,
|
| 1460 |
+
"loss": 1.5567,
|
| 1461 |
+
"step": 484
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 1.58,
|
| 1465 |
+
"learning_rate": 1.393925746368387e-06,
|
| 1466 |
+
"loss": 1.555,
|
| 1467 |
+
"step": 486
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.59,
|
| 1471 |
+
"learning_rate": 1.389071302064295e-06,
|
| 1472 |
+
"loss": 1.5704,
|
| 1473 |
+
"step": 488
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.59,
|
| 1477 |
+
"learning_rate": 1.3842060293573435e-06,
|
| 1478 |
+
"loss": 1.4905,
|
| 1479 |
+
"step": 490
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.6,
|
| 1483 |
+
"learning_rate": 1.379330063654931e-06,
|
| 1484 |
+
"loss": 1.5386,
|
| 1485 |
+
"step": 492
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 1.61,
|
| 1489 |
+
"learning_rate": 1.374443540662057e-06,
|
| 1490 |
+
"loss": 1.4526,
|
| 1491 |
+
"step": 494
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1.61,
|
| 1495 |
+
"learning_rate": 1.369546596377546e-06,
|
| 1496 |
+
"loss": 1.6076,
|
| 1497 |
+
"step": 496
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.62,
|
| 1501 |
+
"learning_rate": 1.364639367090261e-06,
|
| 1502 |
+
"loss": 1.4629,
|
| 1503 |
+
"step": 498
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.62,
|
| 1507 |
+
"learning_rate": 1.3597219893753117e-06,
|
| 1508 |
+
"loss": 1.4478,
|
| 1509 |
+
"step": 500
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.63,
|
| 1513 |
+
"learning_rate": 1.3547946000902522e-06,
|
| 1514 |
+
"loss": 1.5674,
|
| 1515 |
+
"step": 502
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.64,
|
| 1519 |
+
"learning_rate": 1.3498573363712736e-06,
|
| 1520 |
+
"loss": 1.4905,
|
| 1521 |
+
"step": 504
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.64,
|
| 1525 |
+
"learning_rate": 1.3449103356293852e-06,
|
| 1526 |
+
"loss": 1.565,
|
| 1527 |
+
"step": 506
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 1.65,
|
| 1531 |
+
"learning_rate": 1.3399537355465923e-06,
|
| 1532 |
+
"loss": 1.5596,
|
| 1533 |
+
"step": 508
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1.66,
|
| 1537 |
+
"learning_rate": 1.3349876740720633e-06,
|
| 1538 |
+
"loss": 1.5123,
|
| 1539 |
+
"step": 510
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 1.66,
|
| 1543 |
+
"learning_rate": 1.3300122894182909e-06,
|
| 1544 |
+
"loss": 1.4944,
|
| 1545 |
+
"step": 512
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 1.67,
|
| 1549 |
+
"learning_rate": 1.3250277200572431e-06,
|
| 1550 |
+
"loss": 1.5019,
|
| 1551 |
+
"step": 514
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.68,
|
| 1555 |
+
"learning_rate": 1.3200341047165133e-06,
|
| 1556 |
+
"loss": 1.4684,
|
| 1557 |
+
"step": 516
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.68,
|
| 1561 |
+
"learning_rate": 1.315031582375457e-06,
|
| 1562 |
+
"loss": 1.477,
|
| 1563 |
+
"step": 518
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.69,
|
| 1567 |
+
"learning_rate": 1.310020292261323e-06,
|
| 1568 |
+
"loss": 1.5424,
|
| 1569 |
+
"step": 520
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 1.7,
|
| 1573 |
+
"learning_rate": 1.3050003738453804e-06,
|
| 1574 |
+
"loss": 1.4738,
|
| 1575 |
+
"step": 522
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 1.7,
|
| 1579 |
+
"learning_rate": 1.299971966839036e-06,
|
| 1580 |
+
"loss": 1.6079,
|
| 1581 |
+
"step": 524
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.71,
|
| 1585 |
+
"learning_rate": 1.294935211189946e-06,
|
| 1586 |
+
"loss": 1.4931,
|
| 1587 |
+
"step": 526
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.72,
|
| 1591 |
+
"learning_rate": 1.2898902470781208e-06,
|
| 1592 |
+
"loss": 1.5408,
|
| 1593 |
+
"step": 528
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.72,
|
| 1597 |
+
"learning_rate": 1.2848372149120246e-06,
|
| 1598 |
+
"loss": 1.5499,
|
| 1599 |
+
"step": 530
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.73,
|
| 1603 |
+
"learning_rate": 1.2797762553246662e-06,
|
| 1604 |
+
"loss": 1.5375,
|
| 1605 |
+
"step": 532
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.74,
|
| 1609 |
+
"learning_rate": 1.2747075091696862e-06,
|
| 1610 |
+
"loss": 1.4326,
|
| 1611 |
+
"step": 534
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.74,
|
| 1615 |
+
"learning_rate": 1.2696311175174357e-06,
|
| 1616 |
+
"loss": 1.5611,
|
| 1617 |
+
"step": 536
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.75,
|
| 1621 |
+
"learning_rate": 1.2645472216510516e-06,
|
| 1622 |
+
"loss": 1.5441,
|
| 1623 |
+
"step": 538
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.75,
|
| 1627 |
+
"learning_rate": 1.2594559630625233e-06,
|
| 1628 |
+
"loss": 1.5056,
|
| 1629 |
+
"step": 540
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.76,
|
| 1633 |
+
"learning_rate": 1.254357483448755e-06,
|
| 1634 |
+
"loss": 1.5582,
|
| 1635 |
+
"step": 542
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.77,
|
| 1639 |
+
"learning_rate": 1.249251924707622e-06,
|
| 1640 |
+
"loss": 1.5138,
|
| 1641 |
+
"step": 544
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.77,
|
| 1645 |
+
"learning_rate": 1.2441394289340226e-06,
|
| 1646 |
+
"loss": 1.5066,
|
| 1647 |
+
"step": 546
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.78,
|
| 1651 |
+
"learning_rate": 1.2390201384159219e-06,
|
| 1652 |
+
"loss": 1.5965,
|
| 1653 |
+
"step": 548
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 1.79,
|
| 1657 |
+
"learning_rate": 1.2338941956303917e-06,
|
| 1658 |
+
"loss": 1.5904,
|
| 1659 |
+
"step": 550
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.79,
|
| 1663 |
+
"learning_rate": 1.2287617432396466e-06,
|
| 1664 |
+
"loss": 1.6234,
|
| 1665 |
+
"step": 552
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.8,
|
| 1669 |
+
"learning_rate": 1.223622924087073e-06,
|
| 1670 |
+
"loss": 1.4397,
|
| 1671 |
+
"step": 554
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.81,
|
| 1675 |
+
"learning_rate": 1.2184778811932527e-06,
|
| 1676 |
+
"loss": 1.528,
|
| 1677 |
+
"step": 556
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 1.81,
|
| 1681 |
+
"learning_rate": 1.2133267577519824e-06,
|
| 1682 |
+
"loss": 1.4858,
|
| 1683 |
+
"step": 558
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.82,
|
| 1687 |
+
"learning_rate": 1.2081696971262903e-06,
|
| 1688 |
+
"loss": 1.6103,
|
| 1689 |
+
"step": 560
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.83,
|
| 1693 |
+
"learning_rate": 1.203006842844444e-06,
|
| 1694 |
+
"loss": 1.5404,
|
| 1695 |
+
"step": 562
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 1.83,
|
| 1699 |
+
"learning_rate": 1.1978383385959566e-06,
|
| 1700 |
+
"loss": 1.433,
|
| 1701 |
+
"step": 564
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 1.84,
|
| 1705 |
+
"learning_rate": 1.1926643282275882e-06,
|
| 1706 |
+
"loss": 1.6102,
|
| 1707 |
+
"step": 566
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.85,
|
| 1711 |
+
"learning_rate": 1.1874849557393423e-06,
|
| 1712 |
+
"loss": 1.493,
|
| 1713 |
+
"step": 568
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 1.85,
|
| 1717 |
+
"learning_rate": 1.1823003652804569e-06,
|
| 1718 |
+
"loss": 1.5125,
|
| 1719 |
+
"step": 570
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 1.86,
|
| 1723 |
+
"learning_rate": 1.1771107011453933e-06,
|
| 1724 |
+
"loss": 1.5232,
|
| 1725 |
+
"step": 572
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.87,
|
| 1729 |
+
"learning_rate": 1.1719161077698217e-06,
|
| 1730 |
+
"loss": 1.4354,
|
| 1731 |
+
"step": 574
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.87,
|
| 1735 |
+
"learning_rate": 1.1667167297265993e-06,
|
| 1736 |
+
"loss": 1.5592,
|
| 1737 |
+
"step": 576
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 1.88,
|
| 1741 |
+
"learning_rate": 1.1615127117217463e-06,
|
| 1742 |
+
"loss": 1.5371,
|
| 1743 |
+
"step": 578
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 1.88,
|
| 1747 |
+
"learning_rate": 1.1563041985904212e-06,
|
| 1748 |
+
"loss": 1.5508,
|
| 1749 |
+
"step": 580
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.89,
|
| 1753 |
+
"learning_rate": 1.1510913352928878e-06,
|
| 1754 |
+
"loss": 1.5688,
|
| 1755 |
+
"step": 582
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 1.9,
|
| 1759 |
+
"learning_rate": 1.1458742669104803e-06,
|
| 1760 |
+
"loss": 1.4938,
|
| 1761 |
+
"step": 584
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 1.9,
|
| 1765 |
+
"learning_rate": 1.1406531386415673e-06,
|
| 1766 |
+
"loss": 1.5762,
|
| 1767 |
+
"step": 586
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.91,
|
| 1771 |
+
"learning_rate": 1.1354280957975092e-06,
|
| 1772 |
+
"loss": 1.5239,
|
| 1773 |
+
"step": 588
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.92,
|
| 1777 |
+
"learning_rate": 1.130199283798615e-06,
|
| 1778 |
+
"loss": 1.4697,
|
| 1779 |
+
"step": 590
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 1.92,
|
| 1783 |
+
"learning_rate": 1.1249668481700944e-06,
|
| 1784 |
+
"loss": 1.5472,
|
| 1785 |
+
"step": 592
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 1.93,
|
| 1789 |
+
"learning_rate": 1.1197309345380072e-06,
|
| 1790 |
+
"loss": 1.52,
|
| 1791 |
+
"step": 594
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 1.94,
|
| 1795 |
+
"learning_rate": 1.1144916886252124e-06,
|
| 1796 |
+
"loss": 1.5085,
|
| 1797 |
+
"step": 596
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 1.94,
|
| 1801 |
+
"learning_rate": 1.1092492562473098e-06,
|
| 1802 |
+
"loss": 1.5219,
|
| 1803 |
+
"step": 598
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 1.95,
|
| 1807 |
+
"learning_rate": 1.104003783308583e-06,
|
| 1808 |
+
"loss": 1.465,
|
| 1809 |
+
"step": 600
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.96,
|
| 1813 |
+
"learning_rate": 1.098755415797939e-06,
|
| 1814 |
+
"loss": 1.5284,
|
| 1815 |
+
"step": 602
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.96,
|
| 1819 |
+
"learning_rate": 1.0935042997848461e-06,
|
| 1820 |
+
"loss": 1.5163,
|
| 1821 |
+
"step": 604
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.97,
|
| 1825 |
+
"learning_rate": 1.088250581415265e-06,
|
| 1826 |
+
"loss": 1.5255,
|
| 1827 |
+
"step": 606
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.98,
|
| 1831 |
+
"learning_rate": 1.0829944069075847e-06,
|
| 1832 |
+
"loss": 1.4877,
|
| 1833 |
+
"step": 608
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.98,
|
| 1837 |
+
"learning_rate": 1.0777359225485529e-06,
|
| 1838 |
+
"loss": 1.5613,
|
| 1839 |
+
"step": 610
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.99,
|
| 1843 |
+
"learning_rate": 1.0724752746892023e-06,
|
| 1844 |
+
"loss": 1.5088,
|
| 1845 |
+
"step": 612
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 2.0,
|
| 1849 |
+
"learning_rate": 1.0672126097407795e-06,
|
| 1850 |
+
"loss": 1.5147,
|
| 1851 |
+
"step": 614
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.0,
|
| 1855 |
+
"learning_rate": 1.0619480741706691e-06,
|
| 1856 |
+
"loss": 1.5313,
|
| 1857 |
+
"step": 616
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.01,
|
| 1861 |
+
"learning_rate": 1.0566818144983189e-06,
|
| 1862 |
+
"loss": 1.4273,
|
| 1863 |
+
"step": 618
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 2.01,
|
| 1867 |
+
"learning_rate": 1.0514139772911597e-06,
|
| 1868 |
+
"loss": 1.4914,
|
| 1869 |
+
"step": 620
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 2.02,
|
| 1873 |
+
"learning_rate": 1.0461447091605275e-06,
|
| 1874 |
+
"loss": 1.5086,
|
| 1875 |
+
"step": 622
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 2.03,
|
| 1879 |
+
"learning_rate": 1.0408741567575834e-06,
|
| 1880 |
+
"loss": 1.5238,
|
| 1881 |
+
"step": 624
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 2.03,
|
| 1885 |
+
"learning_rate": 1.0356024667692314e-06,
|
| 1886 |
+
"loss": 1.4624,
|
| 1887 |
+
"step": 626
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 2.04,
|
| 1891 |
+
"learning_rate": 1.0303297859140356e-06,
|
| 1892 |
+
"loss": 1.496,
|
| 1893 |
+
"step": 628
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.05,
|
| 1897 |
+
"learning_rate": 1.025056260938138e-06,
|
| 1898 |
+
"loss": 1.4343,
|
| 1899 |
+
"step": 630
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.05,
|
| 1903 |
+
"learning_rate": 1.0197820386111737e-06,
|
| 1904 |
+
"loss": 1.413,
|
| 1905 |
+
"step": 632
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 2.06,
|
| 1909 |
+
"learning_rate": 1.0145072657221858e-06,
|
| 1910 |
+
"loss": 1.4242,
|
| 1911 |
+
"step": 634
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 2.07,
|
| 1915 |
+
"learning_rate": 1.0092320890755397e-06,
|
| 1916 |
+
"loss": 1.4592,
|
| 1917 |
+
"step": 636
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 2.07,
|
| 1921 |
+
"learning_rate": 1.0039566554868392e-06,
|
| 1922 |
+
"loss": 1.4259,
|
| 1923 |
+
"step": 638
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 2.08,
|
| 1927 |
+
"learning_rate": 9.986811117788386e-07,
|
| 1928 |
+
"loss": 1.4989,
|
| 1929 |
+
"step": 640
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 2.09,
|
| 1933 |
+
"learning_rate": 9.934056047773561e-07,
|
| 1934 |
+
"loss": 1.5084,
|
| 1935 |
+
"step": 642
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.09,
|
| 1939 |
+
"learning_rate": 9.881302813071896e-07,
|
| 1940 |
+
"loss": 1.4395,
|
| 1941 |
+
"step": 644
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.1,
|
| 1945 |
+
"learning_rate": 9.828552881880288e-07,
|
| 1946 |
+
"loss": 1.4138,
|
| 1947 |
+
"step": 646
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 2.11,
|
| 1951 |
+
"learning_rate": 9.775807722303687e-07,
|
| 1952 |
+
"loss": 1.4226,
|
| 1953 |
+
"step": 648
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 2.11,
|
| 1957 |
+
"learning_rate": 9.723068802314246e-07,
|
| 1958 |
+
"loss": 1.534,
|
| 1959 |
+
"step": 650
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 2.12,
|
| 1963 |
+
"learning_rate": 9.670337589710467e-07,
|
| 1964 |
+
"loss": 1.4433,
|
| 1965 |
+
"step": 652
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 2.13,
|
| 1969 |
+
"learning_rate": 9.61761555207633e-07,
|
| 1970 |
+
"loss": 1.4163,
|
| 1971 |
+
"step": 654
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 2.13,
|
| 1975 |
+
"learning_rate": 9.564904156740471e-07,
|
| 1976 |
+
"loss": 1.4728,
|
| 1977 |
+
"step": 656
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.14,
|
| 1981 |
+
"learning_rate": 9.512204870735342e-07,
|
| 1982 |
+
"loss": 1.4101,
|
| 1983 |
+
"step": 658
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.14,
|
| 1987 |
+
"learning_rate": 9.459519160756369e-07,
|
| 1988 |
+
"loss": 1.5015,
|
| 1989 |
+
"step": 660
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 2.15,
|
| 1993 |
+
"learning_rate": 9.406848493121134e-07,
|
| 1994 |
+
"loss": 1.4454,
|
| 1995 |
+
"step": 662
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 2.16,
|
| 1999 |
+
"learning_rate": 9.354194333728576e-07,
|
| 2000 |
+
"loss": 1.52,
|
| 2001 |
+
"step": 664
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 2.16,
|
| 2005 |
+
"learning_rate": 9.301558148018188e-07,
|
| 2006 |
+
"loss": 1.5274,
|
| 2007 |
+
"step": 666
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 2.17,
|
| 2011 |
+
"learning_rate": 9.248941400929222e-07,
|
| 2012 |
+
"loss": 1.5197,
|
| 2013 |
+
"step": 668
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 2.18,
|
| 2017 |
+
"learning_rate": 9.196345556859928e-07,
|
| 2018 |
+
"loss": 1.5062,
|
| 2019 |
+
"step": 670
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.18,
|
| 2023 |
+
"learning_rate": 9.1437720796268e-07,
|
| 2024 |
+
"loss": 1.5285,
|
| 2025 |
+
"step": 672
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.19,
|
| 2029 |
+
"learning_rate": 9.09122243242383e-07,
|
| 2030 |
+
"loss": 1.4948,
|
| 2031 |
+
"step": 674
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 2.2,
|
| 2035 |
+
"learning_rate": 9.038698077781786e-07,
|
| 2036 |
+
"loss": 1.4591,
|
| 2037 |
+
"step": 676
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 2.2,
|
| 2041 |
+
"learning_rate": 8.986200477527504e-07,
|
| 2042 |
+
"loss": 1.4638,
|
| 2043 |
+
"step": 678
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 2.21,
|
| 2047 |
+
"learning_rate": 8.933731092743219e-07,
|
| 2048 |
+
"loss": 1.4873,
|
| 2049 |
+
"step": 680
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 2.22,
|
| 2053 |
+
"learning_rate": 8.881291383725875e-07,
|
| 2054 |
+
"loss": 1.4668,
|
| 2055 |
+
"step": 682
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 2.22,
|
| 2059 |
+
"learning_rate": 8.828882809946506e-07,
|
| 2060 |
+
"loss": 1.432,
|
| 2061 |
+
"step": 684
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.23,
|
| 2065 |
+
"learning_rate": 8.776506830009607e-07,
|
| 2066 |
+
"loss": 1.3995,
|
| 2067 |
+
"step": 686
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.24,
|
| 2071 |
+
"learning_rate": 8.724164901612545e-07,
|
| 2072 |
+
"loss": 1.4243,
|
| 2073 |
+
"step": 688
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 2.24,
|
| 2077 |
+
"learning_rate": 8.671858481504977e-07,
|
| 2078 |
+
"loss": 1.4486,
|
| 2079 |
+
"step": 690
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 2.25,
|
| 2083 |
+
"learning_rate": 8.619589025448318e-07,
|
| 2084 |
+
"loss": 1.4809,
|
| 2085 |
+
"step": 692
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 2.26,
|
| 2089 |
+
"learning_rate": 8.567357988175224e-07,
|
| 2090 |
+
"loss": 1.4331,
|
| 2091 |
+
"step": 694
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 2.26,
|
| 2095 |
+
"learning_rate": 8.515166823349095e-07,
|
| 2096 |
+
"loss": 1.4642,
|
| 2097 |
+
"step": 696
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 2.27,
|
| 2101 |
+
"learning_rate": 8.463016983523627e-07,
|
| 2102 |
+
"loss": 1.4659,
|
| 2103 |
+
"step": 698
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.27,
|
| 2107 |
+
"learning_rate": 8.410909920102388e-07,
|
| 2108 |
+
"loss": 1.5181,
|
| 2109 |
+
"step": 700
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.28,
|
| 2113 |
+
"learning_rate": 8.358847083298417e-07,
|
| 2114 |
+
"loss": 1.4767,
|
| 2115 |
+
"step": 702
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 2.29,
|
| 2119 |
+
"learning_rate": 8.306829922093857e-07,
|
| 2120 |
+
"loss": 1.5135,
|
| 2121 |
+
"step": 704
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 2.29,
|
| 2125 |
+
"learning_rate": 8.254859884199641e-07,
|
| 2126 |
+
"loss": 1.4766,
|
| 2127 |
+
"step": 706
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 2.3,
|
| 2131 |
+
"learning_rate": 8.202938416015197e-07,
|
| 2132 |
+
"loss": 1.4883,
|
| 2133 |
+
"step": 708
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 2.31,
|
| 2137 |
+
"learning_rate": 8.15106696258818e-07,
|
| 2138 |
+
"loss": 1.4938,
|
| 2139 |
+
"step": 710
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 2.31,
|
| 2143 |
+
"learning_rate": 8.099246967574266e-07,
|
| 2144 |
+
"loss": 1.4874,
|
| 2145 |
+
"step": 712
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.32,
|
| 2149 |
+
"learning_rate": 8.047479873196979e-07,
|
| 2150 |
+
"loss": 1.4768,
|
| 2151 |
+
"step": 714
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.33,
|
| 2155 |
+
"learning_rate": 7.995767120207536e-07,
|
| 2156 |
+
"loss": 1.4486,
|
| 2157 |
+
"step": 716
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 2.33,
|
| 2161 |
+
"learning_rate": 7.944110147844758e-07,
|
| 2162 |
+
"loss": 1.4077,
|
| 2163 |
+
"step": 718
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 2.34,
|
| 2167 |
+
"learning_rate": 7.892510393795014e-07,
|
| 2168 |
+
"loss": 1.5044,
|
| 2169 |
+
"step": 720
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 2.35,
|
| 2173 |
+
"learning_rate": 7.840969294152211e-07,
|
| 2174 |
+
"loss": 1.443,
|
| 2175 |
+
"step": 722
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 2.35,
|
| 2179 |
+
"learning_rate": 7.789488283377807e-07,
|
| 2180 |
+
"loss": 1.4819,
|
| 2181 |
+
"step": 724
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 2.36,
|
| 2185 |
+
"learning_rate": 7.738068794260915e-07,
|
| 2186 |
+
"loss": 1.4862,
|
| 2187 |
+
"step": 726
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.37,
|
| 2191 |
+
"learning_rate": 7.68671225787841e-07,
|
| 2192 |
+
"loss": 1.3899,
|
| 2193 |
+
"step": 728
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.37,
|
| 2197 |
+
"learning_rate": 7.635420103555104e-07,
|
| 2198 |
+
"loss": 1.4501,
|
| 2199 |
+
"step": 730
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 2.38,
|
| 2203 |
+
"learning_rate": 7.584193758823965e-07,
|
| 2204 |
+
"loss": 1.499,
|
| 2205 |
+
"step": 732
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 2.39,
|
| 2209 |
+
"learning_rate": 7.533034649386384e-07,
|
| 2210 |
+
"loss": 1.4646,
|
| 2211 |
+
"step": 734
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 2.39,
|
| 2215 |
+
"learning_rate": 7.481944199072509e-07,
|
| 2216 |
+
"loss": 1.4742,
|
| 2217 |
+
"step": 736
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 2.4,
|
| 2221 |
+
"learning_rate": 7.430923829801595e-07,
|
| 2222 |
+
"loss": 1.4071,
|
| 2223 |
+
"step": 738
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 2.4,
|
| 2227 |
+
"learning_rate": 7.379974961542447e-07,
|
| 2228 |
+
"loss": 1.4704,
|
| 2229 |
+
"step": 740
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.41,
|
| 2233 |
+
"learning_rate": 7.3290990122739e-07,
|
| 2234 |
+
"loss": 1.396,
|
| 2235 |
+
"step": 742
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.42,
|
| 2239 |
+
"learning_rate": 7.278297397945355e-07,
|
| 2240 |
+
"loss": 1.5513,
|
| 2241 |
+
"step": 744
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 2.42,
|
| 2245 |
+
"learning_rate": 7.227571532437349e-07,
|
| 2246 |
+
"loss": 1.5466,
|
| 2247 |
+
"step": 746
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 2.43,
|
| 2251 |
+
"learning_rate": 7.17692282752224e-07,
|
| 2252 |
+
"loss": 1.4583,
|
| 2253 |
+
"step": 748
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 2.44,
|
| 2257 |
+
"learning_rate": 7.126352692824894e-07,
|
| 2258 |
+
"loss": 1.4533,
|
| 2259 |
+
"step": 750
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 2.44,
|
| 2263 |
+
"learning_rate": 7.075862535783453e-07,
|
| 2264 |
+
"loss": 1.4587,
|
| 2265 |
+
"step": 752
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 2.45,
|
| 2269 |
+
"learning_rate": 7.025453761610165e-07,
|
| 2270 |
+
"loss": 1.418,
|
| 2271 |
+
"step": 754
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.46,
|
| 2275 |
+
"learning_rate": 6.975127773252287e-07,
|
| 2276 |
+
"loss": 1.519,
|
| 2277 |
+
"step": 756
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.46,
|
| 2281 |
+
"learning_rate": 6.924885971353026e-07,
|
| 2282 |
+
"loss": 1.3961,
|
| 2283 |
+
"step": 758
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 2.47,
|
| 2287 |
+
"learning_rate": 6.87472975421256e-07,
|
| 2288 |
+
"loss": 1.347,
|
| 2289 |
+
"step": 760
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 2.48,
|
| 2293 |
+
"learning_rate": 6.824660517749123e-07,
|
| 2294 |
+
"loss": 1.4599,
|
| 2295 |
+
"step": 762
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 2.48,
|
| 2299 |
+
"learning_rate": 6.774679655460158e-07,
|
| 2300 |
+
"loss": 1.4537,
|
| 2301 |
+
"step": 764
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 2.49,
|
| 2305 |
+
"learning_rate": 6.724788558383527e-07,
|
| 2306 |
+
"loss": 1.4792,
|
| 2307 |
+
"step": 766
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 2.5,
|
| 2311 |
+
"learning_rate": 6.674988615058792e-07,
|
| 2312 |
+
"loss": 1.5149,
|
| 2313 |
+
"step": 768
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.5,
|
| 2317 |
+
"learning_rate": 6.625281211488591e-07,
|
| 2318 |
+
"loss": 1.457,
|
| 2319 |
+
"step": 770
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.51,
|
| 2323 |
+
"learning_rate": 6.575667731100045e-07,
|
| 2324 |
+
"loss": 1.489,
|
| 2325 |
+
"step": 772
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 2.52,
|
| 2329 |
+
"learning_rate": 6.526149554706257e-07,
|
| 2330 |
+
"loss": 1.4102,
|
| 2331 |
+
"step": 774
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 2.52,
|
| 2335 |
+
"learning_rate": 6.476728060467888e-07,
|
| 2336 |
+
"loss": 1.4224,
|
| 2337 |
+
"step": 776
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 2.53,
|
| 2341 |
+
"learning_rate": 6.427404623854802e-07,
|
| 2342 |
+
"loss": 1.4085,
|
| 2343 |
+
"step": 778
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 2.53,
|
| 2347 |
+
"learning_rate": 6.378180617607778e-07,
|
| 2348 |
+
"loss": 1.5626,
|
| 2349 |
+
"step": 780
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 2.54,
|
| 2353 |
+
"learning_rate": 6.329057411700298e-07,
|
| 2354 |
+
"loss": 1.4363,
|
| 2355 |
+
"step": 782
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.55,
|
| 2359 |
+
"learning_rate": 6.280036373300445e-07,
|
| 2360 |
+
"loss": 1.4856,
|
| 2361 |
+
"step": 784
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.55,
|
| 2365 |
+
"learning_rate": 6.231118866732828e-07,
|
| 2366 |
+
"loss": 1.4322,
|
| 2367 |
+
"step": 786
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 2.56,
|
| 2371 |
+
"learning_rate": 6.182306253440619e-07,
|
| 2372 |
+
"loss": 1.4242,
|
| 2373 |
+
"step": 788
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 2.57,
|
| 2377 |
+
"learning_rate": 6.133599891947664e-07,
|
| 2378 |
+
"loss": 1.5215,
|
| 2379 |
+
"step": 790
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 2.57,
|
| 2383 |
+
"learning_rate": 6.085001137820671e-07,
|
| 2384 |
+
"loss": 1.4654,
|
| 2385 |
+
"step": 792
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 2.58,
|
| 2389 |
+
"learning_rate": 6.036511343631488e-07,
|
| 2390 |
+
"loss": 1.5148,
|
| 2391 |
+
"step": 794
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 2.59,
|
| 2395 |
+
"learning_rate": 5.988131858919438e-07,
|
| 2396 |
+
"loss": 1.4302,
|
| 2397 |
+
"step": 796
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.59,
|
| 2401 |
+
"learning_rate": 5.939864030153793e-07,
|
| 2402 |
+
"loss": 1.5034,
|
| 2403 |
+
"step": 798
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.6,
|
| 2407 |
+
"learning_rate": 5.89170920069628e-07,
|
| 2408 |
+
"loss": 1.3958,
|
| 2409 |
+
"step": 800
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 2.61,
|
| 2413 |
+
"learning_rate": 5.84366871076369e-07,
|
| 2414 |
+
"loss": 1.4176,
|
| 2415 |
+
"step": 802
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 2.61,
|
| 2419 |
+
"learning_rate": 5.795743897390584e-07,
|
| 2420 |
+
"loss": 1.426,
|
| 2421 |
+
"step": 804
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 2.62,
|
| 2425 |
+
"learning_rate": 5.747936094392089e-07,
|
| 2426 |
+
"loss": 1.4418,
|
| 2427 |
+
"step": 806
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 2.63,
|
| 2431 |
+
"learning_rate": 5.700246632326763e-07,
|
| 2432 |
+
"loss": 1.4593,
|
| 2433 |
+
"step": 808
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 2.63,
|
| 2437 |
+
"learning_rate": 5.652676838459566e-07,
|
| 2438 |
+
"loss": 1.4742,
|
| 2439 |
+
"step": 810
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.64,
|
| 2443 |
+
"learning_rate": 5.605228036724927e-07,
|
| 2444 |
+
"loss": 1.4536,
|
| 2445 |
+
"step": 812
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.65,
|
| 2449 |
+
"learning_rate": 5.557901547689889e-07,
|
| 2450 |
+
"loss": 1.5283,
|
| 2451 |
+
"step": 814
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 2.65,
|
| 2455 |
+
"learning_rate": 5.510698688517373e-07,
|
| 2456 |
+
"loss": 1.4925,
|
| 2457 |
+
"step": 816
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 2.66,
|
| 2461 |
+
"learning_rate": 5.463620772929494e-07,
|
| 2462 |
+
"loss": 1.446,
|
| 2463 |
+
"step": 818
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 2.66,
|
| 2467 |
+
"learning_rate": 5.416669111171008e-07,
|
| 2468 |
+
"loss": 1.5124,
|
| 2469 |
+
"step": 820
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 2.67,
|
| 2473 |
+
"learning_rate": 5.369845009972869e-07,
|
| 2474 |
+
"loss": 1.4495,
|
| 2475 |
+
"step": 822
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 2.68,
|
| 2479 |
+
"learning_rate": 5.323149772515812e-07,
|
| 2480 |
+
"loss": 1.4775,
|
| 2481 |
+
"step": 824
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.68,
|
| 2485 |
+
"learning_rate": 5.276584698394137e-07,
|
| 2486 |
+
"loss": 1.3818,
|
| 2487 |
+
"step": 826
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.69,
|
| 2491 |
+
"learning_rate": 5.230151083579505e-07,
|
| 2492 |
+
"loss": 1.4227,
|
| 2493 |
+
"step": 828
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 2.7,
|
| 2497 |
+
"learning_rate": 5.183850220384873e-07,
|
| 2498 |
+
"loss": 1.527,
|
| 2499 |
+
"step": 830
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 2.7,
|
| 2503 |
+
"learning_rate": 5.13768339742855e-07,
|
| 2504 |
+
"loss": 1.5138,
|
| 2505 |
+
"step": 832
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 2.71,
|
| 2509 |
+
"learning_rate": 5.0916518995983e-07,
|
| 2510 |
+
"loss": 1.4676,
|
| 2511 |
+
"step": 834
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 2.72,
|
| 2515 |
+
"learning_rate": 5.045757008015606e-07,
|
| 2516 |
+
"loss": 1.4236,
|
| 2517 |
+
"step": 836
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 2.72,
|
| 2521 |
+
"learning_rate": 5.000000000000002e-07,
|
| 2522 |
+
"loss": 1.4541,
|
| 2523 |
+
"step": 838
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.73,
|
| 2527 |
+
"learning_rate": 4.954382149033532e-07,
|
| 2528 |
+
"loss": 1.4262,
|
| 2529 |
+
"step": 840
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.74,
|
| 2533 |
+
"learning_rate": 4.908904724725299e-07,
|
| 2534 |
+
"loss": 1.4582,
|
| 2535 |
+
"step": 842
|
| 2536 |
+
},
|
| 2537 |
+
{
|
| 2538 |
+
"epoch": 2.74,
|
| 2539 |
+
"learning_rate": 4.863568992776144e-07,
|
| 2540 |
+
"loss": 1.4549,
|
| 2541 |
+
"step": 844
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 2.75,
|
| 2545 |
+
"learning_rate": 4.818376214943397e-07,
|
| 2546 |
+
"loss": 1.5634,
|
| 2547 |
+
"step": 846
|
| 2548 |
+
},
|
| 2549 |
+
{
|
| 2550 |
+
"epoch": 2.76,
|
| 2551 |
+
"learning_rate": 4.773327649005777e-07,
|
| 2552 |
+
"loss": 1.4474,
|
| 2553 |
+
"step": 848
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 2.76,
|
| 2557 |
+
"learning_rate": 4.728424548728396e-07,
|
| 2558 |
+
"loss": 1.4686,
|
| 2559 |
+
"step": 850
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 2.77,
|
| 2563 |
+
"learning_rate": 4.683668163827824e-07,
|
| 2564 |
+
"loss": 1.41,
|
| 2565 |
+
"step": 852
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 2.77,
|
| 2569 |
+
"learning_rate": 4.6390597399373644e-07,
|
| 2570 |
+
"loss": 1.5725,
|
| 2571 |
+
"step": 854
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.78,
|
| 2575 |
+
"learning_rate": 4.594600518572339e-07,
|
| 2576 |
+
"loss": 1.4113,
|
| 2577 |
+
"step": 856
|
| 2578 |
+
},
|
| 2579 |
+
{
|
| 2580 |
+
"epoch": 2.79,
|
| 2581 |
+
"learning_rate": 4.550291737095555e-07,
|
| 2582 |
+
"loss": 1.5014,
|
| 2583 |
+
"step": 858
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 2.79,
|
| 2587 |
+
"learning_rate": 4.506134628682877e-07,
|
| 2588 |
+
"loss": 1.4909,
|
| 2589 |
+
"step": 860
|
| 2590 |
+
},
|
| 2591 |
+
{
|
| 2592 |
+
"epoch": 2.8,
|
| 2593 |
+
"learning_rate": 4.462130422288881e-07,
|
| 2594 |
+
"loss": 1.4286,
|
| 2595 |
+
"step": 862
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 2.81,
|
| 2599 |
+
"learning_rate": 4.4182803426126667e-07,
|
| 2600 |
+
"loss": 1.3898,
|
| 2601 |
+
"step": 864
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 2.81,
|
| 2605 |
+
"learning_rate": 4.3745856100637834e-07,
|
| 2606 |
+
"loss": 1.4199,
|
| 2607 |
+
"step": 866
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 2.82,
|
| 2611 |
+
"learning_rate": 4.331047440728227e-07,
|
| 2612 |
+
"loss": 1.4073,
|
| 2613 |
+
"step": 868
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.83,
|
| 2617 |
+
"learning_rate": 4.2876670463346333e-07,
|
| 2618 |
+
"loss": 1.4012,
|
| 2619 |
+
"step": 870
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 2.83,
|
| 2623 |
+
"learning_rate": 4.244445634220545e-07,
|
| 2624 |
+
"loss": 1.489,
|
| 2625 |
+
"step": 872
|
| 2626 |
+
},
|
| 2627 |
+
{
|
| 2628 |
+
"epoch": 2.84,
|
| 2629 |
+
"learning_rate": 4.2013844072987903e-07,
|
| 2630 |
+
"loss": 1.4733,
|
| 2631 |
+
"step": 874
|
| 2632 |
+
},
|
| 2633 |
+
{
|
| 2634 |
+
"epoch": 2.85,
|
| 2635 |
+
"learning_rate": 4.158484564024021e-07,
|
| 2636 |
+
"loss": 1.4922,
|
| 2637 |
+
"step": 876
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 2.85,
|
| 2641 |
+
"learning_rate": 4.115747298359363e-07,
|
| 2642 |
+
"loss": 1.3864,
|
| 2643 |
+
"step": 878
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 2.86,
|
| 2647 |
+
"learning_rate": 4.073173799743169e-07,
|
| 2648 |
+
"loss": 1.4948,
|
| 2649 |
+
"step": 880
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 2.87,
|
| 2653 |
+
"learning_rate": 4.030765253055928e-07,
|
| 2654 |
+
"loss": 1.5001,
|
| 2655 |
+
"step": 882
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 2.87,
|
| 2659 |
+
"learning_rate": 3.9885228385872806e-07,
|
| 2660 |
+
"loss": 1.4739,
|
| 2661 |
+
"step": 884
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"epoch": 2.88,
|
| 2665 |
+
"learning_rate": 3.946447732003174e-07,
|
| 2666 |
+
"loss": 1.49,
|
| 2667 |
+
"step": 886
|
| 2668 |
+
},
|
| 2669 |
+
{
|
| 2670 |
+
"epoch": 2.89,
|
| 2671 |
+
"learning_rate": 3.9045411043131493e-07,
|
| 2672 |
+
"loss": 1.4435,
|
| 2673 |
+
"step": 888
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 2.89,
|
| 2677 |
+
"learning_rate": 3.862804121837733e-07,
|
| 2678 |
+
"loss": 1.4894,
|
| 2679 |
+
"step": 890
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 2.9,
|
| 2683 |
+
"learning_rate": 3.821237946175986e-07,
|
| 2684 |
+
"loss": 1.3914,
|
| 2685 |
+
"step": 892
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 2.9,
|
| 2689 |
+
"learning_rate": 3.779843734173188e-07,
|
| 2690 |
+
"loss": 1.3535,
|
| 2691 |
+
"step": 894
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 2.91,
|
| 2695 |
+
"learning_rate": 3.738622637888608e-07,
|
| 2696 |
+
"loss": 1.4704,
|
| 2697 |
+
"step": 896
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 2.92,
|
| 2701 |
+
"learning_rate": 3.6975758045634695e-07,
|
| 2702 |
+
"loss": 1.5333,
|
| 2703 |
+
"step": 898
|
| 2704 |
+
},
|
| 2705 |
+
{
|
| 2706 |
+
"epoch": 2.92,
|
| 2707 |
+
"learning_rate": 3.65670437658902e-07,
|
| 2708 |
+
"loss": 1.4566,
|
| 2709 |
+
"step": 900
|
| 2710 |
+
},
|
| 2711 |
+
{
|
| 2712 |
+
"epoch": 2.93,
|
| 2713 |
+
"learning_rate": 3.61600949147472e-07,
|
| 2714 |
+
"loss": 1.4203,
|
| 2715 |
+
"step": 902
|
| 2716 |
+
},
|
| 2717 |
+
{
|
| 2718 |
+
"epoch": 2.94,
|
| 2719 |
+
"learning_rate": 3.575492281816589e-07,
|
| 2720 |
+
"loss": 1.4287,
|
| 2721 |
+
"step": 904
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 2.94,
|
| 2725 |
+
"learning_rate": 3.5351538752657005e-07,
|
| 2726 |
+
"loss": 1.4287,
|
| 2727 |
+
"step": 906
|
| 2728 |
+
},
|
| 2729 |
+
{
|
| 2730 |
+
"epoch": 2.95,
|
| 2731 |
+
"learning_rate": 3.494995394496778e-07,
|
| 2732 |
+
"loss": 1.4893,
|
| 2733 |
+
"step": 908
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 2.96,
|
| 2737 |
+
"learning_rate": 3.455017957176957e-07,
|
| 2738 |
+
"loss": 1.4444,
|
| 2739 |
+
"step": 910
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 2.96,
|
| 2743 |
+
"learning_rate": 3.415222675934678e-07,
|
| 2744 |
+
"loss": 1.4595,
|
| 2745 |
+
"step": 912
|
| 2746 |
+
},
|
| 2747 |
+
{
|
| 2748 |
+
"epoch": 2.97,
|
| 2749 |
+
"learning_rate": 3.3756106583287205e-07,
|
| 2750 |
+
"loss": 1.4748,
|
| 2751 |
+
"step": 914
|
| 2752 |
+
},
|
| 2753 |
+
{
|
| 2754 |
+
"epoch": 2.98,
|
| 2755 |
+
"learning_rate": 3.3361830068173867e-07,
|
| 2756 |
+
"loss": 1.5064,
|
| 2757 |
+
"step": 916
|
| 2758 |
+
},
|
| 2759 |
+
{
|
| 2760 |
+
"epoch": 2.98,
|
| 2761 |
+
"learning_rate": 3.296940818727801e-07,
|
| 2762 |
+
"loss": 1.4905,
|
| 2763 |
+
"step": 918
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 2.99,
|
| 2767 |
+
"learning_rate": 3.2578851862253796e-07,
|
| 2768 |
+
"loss": 1.428,
|
| 2769 |
+
"step": 920
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 3.0,
|
| 2773 |
+
"learning_rate": 3.219017196283447e-07,
|
| 2774 |
+
"loss": 1.5019,
|
| 2775 |
+
"step": 922
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 3.0,
|
| 2779 |
+
"learning_rate": 3.1803379306529485e-07,
|
| 2780 |
+
"loss": 1.4214,
|
| 2781 |
+
"step": 924
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 3.01,
|
| 2785 |
+
"learning_rate": 3.1418484658323806e-07,
|
| 2786 |
+
"loss": 1.4506,
|
| 2787 |
+
"step": 926
|
| 2788 |
+
},
|
| 2789 |
+
{
|
| 2790 |
+
"epoch": 3.02,
|
| 2791 |
+
"learning_rate": 3.1035498730378176e-07,
|
| 2792 |
+
"loss": 1.4256,
|
| 2793 |
+
"step": 928
|
| 2794 |
+
},
|
| 2795 |
+
{
|
| 2796 |
+
"epoch": 3.02,
|
| 2797 |
+
"learning_rate": 3.065443218173088e-07,
|
| 2798 |
+
"loss": 1.4937,
|
| 2799 |
+
"step": 930
|
| 2800 |
+
},
|
| 2801 |
+
{
|
| 2802 |
+
"epoch": 3.03,
|
| 2803 |
+
"learning_rate": 3.027529561800117e-07,
|
| 2804 |
+
"loss": 1.3694,
|
| 2805 |
+
"step": 932
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 3.03,
|
| 2809 |
+
"learning_rate": 2.9898099591094217e-07,
|
| 2810 |
+
"loss": 1.4178,
|
| 2811 |
+
"step": 934
|
| 2812 |
+
},
|
| 2813 |
+
{
|
| 2814 |
+
"epoch": 3.04,
|
| 2815 |
+
"learning_rate": 2.952285459890718e-07,
|
| 2816 |
+
"loss": 1.4458,
|
| 2817 |
+
"step": 936
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 3.05,
|
| 2821 |
+
"learning_rate": 2.9149571085037215e-07,
|
| 2822 |
+
"loss": 1.4191,
|
| 2823 |
+
"step": 938
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 3.05,
|
| 2827 |
+
"learning_rate": 2.877825943849079e-07,
|
| 2828 |
+
"loss": 1.399,
|
| 2829 |
+
"step": 940
|
| 2830 |
+
},
|
| 2831 |
+
{
|
| 2832 |
+
"epoch": 3.06,
|
| 2833 |
+
"learning_rate": 2.8408929993394493e-07,
|
| 2834 |
+
"loss": 1.3529,
|
| 2835 |
+
"step": 942
|
| 2836 |
+
},
|
| 2837 |
+
{
|
| 2838 |
+
"epoch": 3.07,
|
| 2839 |
+
"learning_rate": 2.804159302870751e-07,
|
| 2840 |
+
"loss": 1.4003,
|
| 2841 |
+
"step": 944
|
| 2842 |
+
},
|
| 2843 |
+
{
|
| 2844 |
+
"epoch": 3.07,
|
| 2845 |
+
"learning_rate": 2.767625876793543e-07,
|
| 2846 |
+
"loss": 1.3533,
|
| 2847 |
+
"step": 946
|
| 2848 |
+
},
|
| 2849 |
+
{
|
| 2850 |
+
"epoch": 3.08,
|
| 2851 |
+
"learning_rate": 2.7312937378845767e-07,
|
| 2852 |
+
"loss": 1.4441,
|
| 2853 |
+
"step": 948
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 3.09,
|
| 2857 |
+
"learning_rate": 2.6951638973185073e-07,
|
| 2858 |
+
"loss": 1.3955,
|
| 2859 |
+
"step": 950
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 3.09,
|
| 2863 |
+
"learning_rate": 2.659237360639727e-07,
|
| 2864 |
+
"loss": 1.4242,
|
| 2865 |
+
"step": 952
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 3.1,
|
| 2869 |
+
"learning_rate": 2.6235151277344e-07,
|
| 2870 |
+
"loss": 1.3973,
|
| 2871 |
+
"step": 954
|
| 2872 |
+
},
|
| 2873 |
+
{
|
| 2874 |
+
"epoch": 3.11,
|
| 2875 |
+
"learning_rate": 2.587998192802638e-07,
|
| 2876 |
+
"loss": 1.431,
|
| 2877 |
+
"step": 956
|
| 2878 |
+
},
|
| 2879 |
+
{
|
| 2880 |
+
"epoch": 3.11,
|
| 2881 |
+
"learning_rate": 2.55268754433081e-07,
|
| 2882 |
+
"loss": 1.4513,
|
| 2883 |
+
"step": 958
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"epoch": 3.12,
|
| 2887 |
+
"learning_rate": 2.5175841650640415e-07,
|
| 2888 |
+
"loss": 1.3152,
|
| 2889 |
+
"step": 960
|
| 2890 |
+
},
|
| 2891 |
+
{
|
| 2892 |
+
"epoch": 3.13,
|
| 2893 |
+
"learning_rate": 2.482689031978872e-07,
|
| 2894 |
+
"loss": 1.4249,
|
| 2895 |
+
"step": 962
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 3.13,
|
| 2899 |
+
"learning_rate": 2.4480031162560486e-07,
|
| 2900 |
+
"loss": 1.3757,
|
| 2901 |
+
"step": 964
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 3.14,
|
| 2905 |
+
"learning_rate": 2.4135273832535087e-07,
|
| 2906 |
+
"loss": 1.4599,
|
| 2907 |
+
"step": 966
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 3.15,
|
| 2911 |
+
"learning_rate": 2.3792627924795038e-07,
|
| 2912 |
+
"loss": 1.3955,
|
| 2913 |
+
"step": 968
|
| 2914 |
+
},
|
| 2915 |
+
{
|
| 2916 |
+
"epoch": 3.15,
|
| 2917 |
+
"learning_rate": 2.3452102975659028e-07,
|
| 2918 |
+
"loss": 1.3862,
|
| 2919 |
+
"step": 970
|
| 2920 |
+
},
|
| 2921 |
+
{
|
| 2922 |
+
"epoch": 3.16,
|
| 2923 |
+
"learning_rate": 2.3113708462416503e-07,
|
| 2924 |
+
"loss": 1.411,
|
| 2925 |
+
"step": 972
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 3.16,
|
| 2929 |
+
"learning_rate": 2.277745380306383e-07,
|
| 2930 |
+
"loss": 1.4113,
|
| 2931 |
+
"step": 974
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 3.17,
|
| 2935 |
+
"learning_rate": 2.2443348356042203e-07,
|
| 2936 |
+
"loss": 1.4988,
|
| 2937 |
+
"step": 976
|
| 2938 |
+
},
|
| 2939 |
+
{
|
| 2940 |
+
"epoch": 3.18,
|
| 2941 |
+
"learning_rate": 2.2111401419977304e-07,
|
| 2942 |
+
"loss": 1.4613,
|
| 2943 |
+
"step": 978
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 3.18,
|
| 2947 |
+
"learning_rate": 2.178162223342035e-07,
|
| 2948 |
+
"loss": 1.4141,
|
| 2949 |
+
"step": 980
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 3.19,
|
| 2953 |
+
"learning_rate": 2.1454019974590954e-07,
|
| 2954 |
+
"loss": 1.4747,
|
| 2955 |
+
"step": 982
|
| 2956 |
+
},
|
| 2957 |
+
{
|
| 2958 |
+
"epoch": 3.2,
|
| 2959 |
+
"learning_rate": 2.1128603761121922e-07,
|
| 2960 |
+
"loss": 1.408,
|
| 2961 |
+
"step": 984
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 3.2,
|
| 2965 |
+
"learning_rate": 2.0805382649805225e-07,
|
| 2966 |
+
"loss": 1.4159,
|
| 2967 |
+
"step": 986
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 3.21,
|
| 2971 |
+
"learning_rate": 2.048436563634005e-07,
|
| 2972 |
+
"loss": 1.4257,
|
| 2973 |
+
"step": 988
|
| 2974 |
+
},
|
| 2975 |
+
{
|
| 2976 |
+
"epoch": 3.22,
|
| 2977 |
+
"learning_rate": 2.0165561655082518e-07,
|
| 2978 |
+
"loss": 1.4474,
|
| 2979 |
+
"step": 990
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 3.22,
|
| 2983 |
+
"learning_rate": 1.9848979578796865e-07,
|
| 2984 |
+
"loss": 1.4697,
|
| 2985 |
+
"step": 992
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 3.23,
|
| 2989 |
+
"learning_rate": 1.9534628218408588e-07,
|
| 2990 |
+
"loss": 1.4834,
|
| 2991 |
+
"step": 994
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 3.24,
|
| 2995 |
+
"learning_rate": 1.922251632275923e-07,
|
| 2996 |
+
"loss": 1.4003,
|
| 2997 |
+
"step": 996
|
| 2998 |
+
},
|
| 2999 |
+
{
|
| 3000 |
+
"epoch": 3.24,
|
| 3001 |
+
"learning_rate": 1.8912652578362853e-07,
|
| 3002 |
+
"loss": 1.3778,
|
| 3003 |
+
"step": 998
|
| 3004 |
+
},
|
| 3005 |
+
{
|
| 3006 |
+
"epoch": 3.25,
|
| 3007 |
+
"learning_rate": 1.860504560916436e-07,
|
| 3008 |
+
"loss": 1.4669,
|
| 3009 |
+
"step": 1000
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 3.26,
|
| 3013 |
+
"learning_rate": 1.8299703976299353e-07,
|
| 3014 |
+
"loss": 1.4135,
|
| 3015 |
+
"step": 1002
|
| 3016 |
+
},
|
| 3017 |
+
{
|
| 3018 |
+
"epoch": 3.26,
|
| 3019 |
+
"learning_rate": 1.7996636177855928e-07,
|
| 3020 |
+
"loss": 1.4599,
|
| 3021 |
+
"step": 1004
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 3.27,
|
| 3025 |
+
"learning_rate": 1.7695850648638244e-07,
|
| 3026 |
+
"loss": 1.4178,
|
| 3027 |
+
"step": 1006
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 3.28,
|
| 3031 |
+
"learning_rate": 1.739735575993162e-07,
|
| 3032 |
+
"loss": 1.4049,
|
| 3033 |
+
"step": 1008
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 3.28,
|
| 3037 |
+
"learning_rate": 1.7101159819269583e-07,
|
| 3038 |
+
"loss": 1.4162,
|
| 3039 |
+
"step": 1010
|
| 3040 |
+
},
|
| 3041 |
+
{
|
| 3042 |
+
"epoch": 3.29,
|
| 3043 |
+
"learning_rate": 1.68072710702028e-07,
|
| 3044 |
+
"loss": 1.4924,
|
| 3045 |
+
"step": 1012
|
| 3046 |
+
},
|
| 3047 |
+
{
|
| 3048 |
+
"epoch": 3.29,
|
| 3049 |
+
"learning_rate": 1.6515697692069487e-07,
|
| 3050 |
+
"loss": 1.4934,
|
| 3051 |
+
"step": 1014
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 3.3,
|
| 3055 |
+
"learning_rate": 1.6226447799767772e-07,
|
| 3056 |
+
"loss": 1.4499,
|
| 3057 |
+
"step": 1016
|
| 3058 |
+
},
|
| 3059 |
+
{
|
| 3060 |
+
"epoch": 3.31,
|
| 3061 |
+
"learning_rate": 1.593952944353003e-07,
|
| 3062 |
+
"loss": 1.3859,
|
| 3063 |
+
"step": 1018
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 3.31,
|
| 3067 |
+
"learning_rate": 1.565495060869857e-07,
|
| 3068 |
+
"loss": 1.3451,
|
| 3069 |
+
"step": 1020
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 3.32,
|
| 3073 |
+
"learning_rate": 1.5372719215503582e-07,
|
| 3074 |
+
"loss": 1.4642,
|
| 3075 |
+
"step": 1022
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 3.33,
|
| 3079 |
+
"learning_rate": 1.5092843118842646e-07,
|
| 3080 |
+
"loss": 1.4595,
|
| 3081 |
+
"step": 1024
|
| 3082 |
+
},
|
| 3083 |
+
{
|
| 3084 |
+
"epoch": 3.33,
|
| 3085 |
+
"learning_rate": 1.4815330108062075e-07,
|
| 3086 |
+
"loss": 1.4066,
|
| 3087 |
+
"step": 1026
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 3.34,
|
| 3091 |
+
"learning_rate": 1.4540187906740241e-07,
|
| 3092 |
+
"loss": 1.4162,
|
| 3093 |
+
"step": 1028
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 3.35,
|
| 3097 |
+
"learning_rate": 1.426742417247251e-07,
|
| 3098 |
+
"loss": 1.428,
|
| 3099 |
+
"step": 1030
|
| 3100 |
+
},
|
| 3101 |
+
{
|
| 3102 |
+
"epoch": 3.35,
|
| 3103 |
+
"learning_rate": 1.3997046496658116e-07,
|
| 3104 |
+
"loss": 1.3849,
|
| 3105 |
+
"step": 1032
|
| 3106 |
+
},
|
| 3107 |
+
{
|
| 3108 |
+
"epoch": 3.36,
|
| 3109 |
+
"learning_rate": 1.3729062404289017e-07,
|
| 3110 |
+
"loss": 1.4919,
|
| 3111 |
+
"step": 1034
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 3.37,
|
| 3115 |
+
"learning_rate": 1.346347935374028e-07,
|
| 3116 |
+
"loss": 1.445,
|
| 3117 |
+
"step": 1036
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 3.37,
|
| 3121 |
+
"learning_rate": 1.3200304736562638e-07,
|
| 3122 |
+
"loss": 1.4546,
|
| 3123 |
+
"step": 1038
|
| 3124 |
+
},
|
| 3125 |
+
{
|
| 3126 |
+
"epoch": 3.38,
|
| 3127 |
+
"learning_rate": 1.2939545877276726e-07,
|
| 3128 |
+
"loss": 1.5042,
|
| 3129 |
+
"step": 1040
|
| 3130 |
+
},
|
| 3131 |
+
{
|
| 3132 |
+
"epoch": 3.39,
|
| 3133 |
+
"learning_rate": 1.268121003316921e-07,
|
| 3134 |
+
"loss": 1.4369,
|
| 3135 |
+
"step": 1042
|
| 3136 |
+
},
|
| 3137 |
+
{
|
| 3138 |
+
"epoch": 3.39,
|
| 3139 |
+
"learning_rate": 1.242530439409082e-07,
|
| 3140 |
+
"loss": 1.3854,
|
| 3141 |
+
"step": 1044
|
| 3142 |
+
},
|
| 3143 |
+
{
|
| 3144 |
+
"epoch": 3.4,
|
| 3145 |
+
"learning_rate": 1.2171836082256316e-07,
|
| 3146 |
+
"loss": 1.3882,
|
| 3147 |
+
"step": 1046
|
| 3148 |
+
},
|
| 3149 |
+
{
|
| 3150 |
+
"epoch": 3.41,
|
| 3151 |
+
"learning_rate": 1.1920812152046145e-07,
|
| 3152 |
+
"loss": 1.4492,
|
| 3153 |
+
"step": 1048
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 3.41,
|
| 3157 |
+
"learning_rate": 1.1672239589810184e-07,
|
| 3158 |
+
"loss": 1.4256,
|
| 3159 |
+
"step": 1050
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 3.42,
|
| 3163 |
+
"learning_rate": 1.1426125313673285e-07,
|
| 3164 |
+
"loss": 1.4191,
|
| 3165 |
+
"step": 1052
|
| 3166 |
+
},
|
| 3167 |
+
{
|
| 3168 |
+
"epoch": 3.42,
|
| 3169 |
+
"learning_rate": 1.1182476173342715e-07,
|
| 3170 |
+
"loss": 1.4539,
|
| 3171 |
+
"step": 1054
|
| 3172 |
+
},
|
| 3173 |
+
{
|
| 3174 |
+
"epoch": 3.43,
|
| 3175 |
+
"learning_rate": 1.0941298949917576e-07,
|
| 3176 |
+
"loss": 1.352,
|
| 3177 |
+
"step": 1056
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 3.44,
|
| 3181 |
+
"learning_rate": 1.070260035570002e-07,
|
| 3182 |
+
"loss": 1.4514,
|
| 3183 |
+
"step": 1058
|
| 3184 |
+
},
|
| 3185 |
+
{
|
| 3186 |
+
"epoch": 3.44,
|
| 3187 |
+
"learning_rate": 1.0466387034008405e-07,
|
| 3188 |
+
"loss": 1.4568,
|
| 3189 |
+
"step": 1060
|
| 3190 |
+
},
|
| 3191 |
+
{
|
| 3192 |
+
"epoch": 3.45,
|
| 3193 |
+
"learning_rate": 1.0232665558992549e-07,
|
| 3194 |
+
"loss": 1.4527,
|
| 3195 |
+
"step": 1062
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 3.46,
|
| 3199 |
+
"learning_rate": 1.000144243545058e-07,
|
| 3200 |
+
"loss": 1.4312,
|
| 3201 |
+
"step": 1064
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 3.46,
|
| 3205 |
+
"learning_rate": 9.772724098648011e-08,
|
| 3206 |
+
"loss": 1.4556,
|
| 3207 |
+
"step": 1066
|
| 3208 |
+
},
|
| 3209 |
+
{
|
| 3210 |
+
"epoch": 3.47,
|
| 3211 |
+
"learning_rate": 9.546516914138603e-08,
|
| 3212 |
+
"loss": 1.4276,
|
| 3213 |
+
"step": 1068
|
| 3214 |
+
},
|
| 3215 |
+
{
|
| 3216 |
+
"epoch": 3.48,
|
| 3217 |
+
"learning_rate": 9.322827177587212e-08,
|
| 3218 |
+
"loss": 1.4723,
|
| 3219 |
+
"step": 1070
|
| 3220 |
+
},
|
| 3221 |
+
{
|
| 3222 |
+
"epoch": 3.48,
|
| 3223 |
+
"learning_rate": 9.101661114594606e-08,
|
| 3224 |
+
"loss": 1.3846,
|
| 3225 |
+
"step": 1072
|
| 3226 |
+
},
|
| 3227 |
+
{
|
| 3228 |
+
"epoch": 3.49,
|
| 3229 |
+
"learning_rate": 8.883024880524081e-08,
|
| 3230 |
+
"loss": 1.4027,
|
| 3231 |
+
"step": 1074
|
| 3232 |
+
},
|
| 3233 |
+
{
|
| 3234 |
+
"epoch": 3.5,
|
| 3235 |
+
"learning_rate": 8.66692456033029e-08,
|
| 3236 |
+
"loss": 1.4113,
|
| 3237 |
+
"step": 1076
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 3.5,
|
| 3241 |
+
"learning_rate": 8.453366168389797e-08,
|
| 3242 |
+
"loss": 1.4566,
|
| 3243 |
+
"step": 1078
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 3.51,
|
| 3247 |
+
"learning_rate": 8.242355648333765e-08,
|
| 3248 |
+
"loss": 1.4043,
|
| 3249 |
+
"step": 1080
|
| 3250 |
+
},
|
| 3251 |
+
{
|
| 3252 |
+
"epoch": 3.52,
|
| 3253 |
+
"learning_rate": 8.033898872882394e-08,
|
| 3254 |
+
"loss": 1.4555,
|
| 3255 |
+
"step": 1082
|
| 3256 |
+
},
|
| 3257 |
+
{
|
| 3258 |
+
"epoch": 3.52,
|
| 3259 |
+
"learning_rate": 7.828001643681681e-08,
|
| 3260 |
+
"loss": 1.3643,
|
| 3261 |
+
"step": 1084
|
| 3262 |
+
},
|
| 3263 |
+
{
|
| 3264 |
+
"epoch": 3.53,
|
| 3265 |
+
"learning_rate": 7.624669691141783e-08,
|
| 3266 |
+
"loss": 1.4216,
|
| 3267 |
+
"step": 1086
|
| 3268 |
+
},
|
| 3269 |
+
{
|
| 3270 |
+
"epoch": 3.54,
|
| 3271 |
+
"learning_rate": 7.423908674277579e-08,
|
| 3272 |
+
"loss": 1.4274,
|
| 3273 |
+
"step": 1088
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 3.54,
|
| 3277 |
+
"learning_rate": 7.225724180551251e-08,
|
| 3278 |
+
"loss": 1.3348,
|
| 3279 |
+
"step": 1090
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 3.55,
|
| 3283 |
+
"learning_rate": 7.030121725716632e-08,
|
| 3284 |
+
"loss": 1.4746,
|
| 3285 |
+
"step": 1092
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 3.55,
|
| 3289 |
+
"learning_rate": 6.837106753665823e-08,
|
| 3290 |
+
"loss": 1.4818,
|
| 3291 |
+
"step": 1094
|
| 3292 |
+
},
|
| 3293 |
+
{
|
| 3294 |
+
"epoch": 3.56,
|
| 3295 |
+
"learning_rate": 6.6466846362776e-08,
|
| 3296 |
+
"loss": 1.4001,
|
| 3297 |
+
"step": 1096
|
| 3298 |
+
},
|
| 3299 |
+
{
|
| 3300 |
+
"epoch": 3.57,
|
| 3301 |
+
"learning_rate": 6.458860673267941e-08,
|
| 3302 |
+
"loss": 1.4383,
|
| 3303 |
+
"step": 1098
|
| 3304 |
+
},
|
| 3305 |
+
{
|
| 3306 |
+
"epoch": 3.57,
|
| 3307 |
+
"learning_rate": 6.273640092042575e-08,
|
| 3308 |
+
"loss": 1.5331,
|
| 3309 |
+
"step": 1100
|
| 3310 |
+
},
|
| 3311 |
+
{
|
| 3312 |
+
"epoch": 3.58,
|
| 3313 |
+
"learning_rate": 6.091028047551373e-08,
|
| 3314 |
+
"loss": 1.3691,
|
| 3315 |
+
"step": 1102
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 3.59,
|
| 3319 |
+
"learning_rate": 5.9110296221449915e-08,
|
| 3320 |
+
"loss": 1.4393,
|
| 3321 |
+
"step": 1104
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 3.59,
|
| 3325 |
+
"learning_rate": 5.733649825433384e-08,
|
| 3326 |
+
"loss": 1.3983,
|
| 3327 |
+
"step": 1106
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 3.6,
|
| 3331 |
+
"learning_rate": 5.558893594146386e-08,
|
| 3332 |
+
"loss": 1.3864,
|
| 3333 |
+
"step": 1108
|
| 3334 |
+
},
|
| 3335 |
+
{
|
| 3336 |
+
"epoch": 3.61,
|
| 3337 |
+
"learning_rate": 5.386765791996228e-08,
|
| 3338 |
+
"loss": 1.3904,
|
| 3339 |
+
"step": 1110
|
| 3340 |
+
},
|
| 3341 |
+
{
|
| 3342 |
+
"epoch": 3.61,
|
| 3343 |
+
"learning_rate": 5.217271209542384e-08,
|
| 3344 |
+
"loss": 1.4557,
|
| 3345 |
+
"step": 1112
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 3.62,
|
| 3349 |
+
"learning_rate": 5.050414564058003e-08,
|
| 3350 |
+
"loss": 1.4285,
|
| 3351 |
+
"step": 1114
|
| 3352 |
+
},
|
| 3353 |
+
{
|
| 3354 |
+
"epoch": 3.63,
|
| 3355 |
+
"learning_rate": 4.8862004993987624e-08,
|
| 3356 |
+
"loss": 1.4477,
|
| 3357 |
+
"step": 1116
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 3.63,
|
| 3361 |
+
"learning_rate": 4.724633585873627e-08,
|
| 3362 |
+
"loss": 1.4061,
|
| 3363 |
+
"step": 1118
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 3.64,
|
| 3367 |
+
"learning_rate": 4.565718320117562e-08,
|
| 3368 |
+
"loss": 1.4155,
|
| 3369 |
+
"step": 1120
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 3.65,
|
| 3373 |
+
"learning_rate": 4.409459124966453e-08,
|
| 3374 |
+
"loss": 1.4423,
|
| 3375 |
+
"step": 1122
|
| 3376 |
+
},
|
| 3377 |
+
{
|
| 3378 |
+
"epoch": 3.65,
|
| 3379 |
+
"learning_rate": 4.255860349334006e-08,
|
| 3380 |
+
"loss": 1.4827,
|
| 3381 |
+
"step": 1124
|
| 3382 |
+
},
|
| 3383 |
+
{
|
| 3384 |
+
"epoch": 3.66,
|
| 3385 |
+
"learning_rate": 4.104926268090691e-08,
|
| 3386 |
+
"loss": 1.4305,
|
| 3387 |
+
"step": 1126
|
| 3388 |
+
},
|
| 3389 |
+
{
|
| 3390 |
+
"epoch": 3.67,
|
| 3391 |
+
"learning_rate": 3.95666108194479e-08,
|
| 3392 |
+
"loss": 1.3561,
|
| 3393 |
+
"step": 1128
|
| 3394 |
+
},
|
| 3395 |
+
{
|
| 3396 |
+
"epoch": 3.67,
|
| 3397 |
+
"learning_rate": 3.811068917325444e-08,
|
| 3398 |
+
"loss": 1.475,
|
| 3399 |
+
"step": 1130
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 3.68,
|
| 3403 |
+
"learning_rate": 3.668153826267861e-08,
|
| 3404 |
+
"loss": 1.4641,
|
| 3405 |
+
"step": 1132
|
| 3406 |
+
},
|
| 3407 |
+
{
|
| 3408 |
+
"epoch": 3.68,
|
| 3409 |
+
"learning_rate": 3.527919786300526e-08,
|
| 3410 |
+
"loss": 1.3561,
|
| 3411 |
+
"step": 1134
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 3.69,
|
| 3415 |
+
"learning_rate": 3.3903707003344774e-08,
|
| 3416 |
+
"loss": 1.3383,
|
| 3417 |
+
"step": 1136
|
| 3418 |
+
},
|
| 3419 |
+
{
|
| 3420 |
+
"epoch": 3.7,
|
| 3421 |
+
"learning_rate": 3.255510396554651e-08,
|
| 3422 |
+
"loss": 1.4744,
|
| 3423 |
+
"step": 1138
|
| 3424 |
+
},
|
| 3425 |
+
{
|
| 3426 |
+
"epoch": 3.7,
|
| 3427 |
+
"learning_rate": 3.1233426283134854e-08,
|
| 3428 |
+
"loss": 1.3745,
|
| 3429 |
+
"step": 1140
|
| 3430 |
+
},
|
| 3431 |
+
{
|
| 3432 |
+
"epoch": 3.71,
|
| 3433 |
+
"learning_rate": 2.9938710740262884e-08,
|
| 3434 |
+
"loss": 1.4124,
|
| 3435 |
+
"step": 1142
|
| 3436 |
+
},
|
| 3437 |
+
{
|
| 3438 |
+
"epoch": 3.72,
|
| 3439 |
+
"learning_rate": 2.8670993370689233e-08,
|
| 3440 |
+
"loss": 1.3834,
|
| 3441 |
+
"step": 1144
|
| 3442 |
+
},
|
| 3443 |
+
{
|
| 3444 |
+
"epoch": 3.72,
|
| 3445 |
+
"learning_rate": 2.7430309456775934e-08,
|
| 3446 |
+
"loss": 1.4374,
|
| 3447 |
+
"step": 1146
|
| 3448 |
+
},
|
| 3449 |
+
{
|
| 3450 |
+
"epoch": 3.73,
|
| 3451 |
+
"learning_rate": 2.6216693528505195e-08,
|
| 3452 |
+
"loss": 1.5022,
|
| 3453 |
+
"step": 1148
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 3.74,
|
| 3457 |
+
"learning_rate": 2.5030179362519498e-08,
|
| 3458 |
+
"loss": 1.4534,
|
| 3459 |
+
"step": 1150
|
| 3460 |
+
},
|
| 3461 |
+
{
|
| 3462 |
+
"epoch": 3.74,
|
| 3463 |
+
"learning_rate": 2.387079998118069e-08,
|
| 3464 |
+
"loss": 1.42,
|
| 3465 |
+
"step": 1152
|
| 3466 |
+
},
|
| 3467 |
+
{
|
| 3468 |
+
"epoch": 3.75,
|
| 3469 |
+
"learning_rate": 2.2738587651651487e-08,
|
| 3470 |
+
"loss": 1.4001,
|
| 3471 |
+
"step": 1154
|
| 3472 |
+
},
|
| 3473 |
+
{
|
| 3474 |
+
"epoch": 3.76,
|
| 3475 |
+
"learning_rate": 2.1633573884997315e-08,
|
| 3476 |
+
"loss": 1.3889,
|
| 3477 |
+
"step": 1156
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 3.76,
|
| 3481 |
+
"learning_rate": 2.0555789435309225e-08,
|
| 3482 |
+
"loss": 1.4377,
|
| 3483 |
+
"step": 1158
|
| 3484 |
+
},
|
| 3485 |
+
{
|
| 3486 |
+
"epoch": 3.77,
|
| 3487 |
+
"learning_rate": 1.950526429884769e-08,
|
| 3488 |
+
"loss": 1.434,
|
| 3489 |
+
"step": 1160
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 3.78,
|
| 3493 |
+
"learning_rate": 1.8482027713208615e-08,
|
| 3494 |
+
"loss": 1.5097,
|
| 3495 |
+
"step": 1162
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 3.78,
|
| 3499 |
+
"learning_rate": 1.748610815650875e-08,
|
| 3500 |
+
"loss": 1.4127,
|
| 3501 |
+
"step": 1164
|
| 3502 |
+
},
|
| 3503 |
+
{
|
| 3504 |
+
"epoch": 3.79,
|
| 3505 |
+
"learning_rate": 1.6517533346593226e-08,
|
| 3506 |
+
"loss": 1.4517,
|
| 3507 |
+
"step": 1166
|
| 3508 |
+
},
|
| 3509 |
+
{
|
| 3510 |
+
"epoch": 3.8,
|
| 3511 |
+
"learning_rate": 1.5576330240264614e-08,
|
| 3512 |
+
"loss": 1.3508,
|
| 3513 |
+
"step": 1168
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 3.8,
|
| 3517 |
+
"learning_rate": 1.4662525032532292e-08,
|
| 3518 |
+
"loss": 1.4554,
|
| 3519 |
+
"step": 1170
|
| 3520 |
+
},
|
| 3521 |
+
{
|
| 3522 |
+
"epoch": 3.81,
|
| 3523 |
+
"learning_rate": 1.3776143155883491e-08,
|
| 3524 |
+
"loss": 1.4146,
|
| 3525 |
+
"step": 1172
|
| 3526 |
+
},
|
| 3527 |
+
{
|
| 3528 |
+
"epoch": 3.81,
|
| 3529 |
+
"learning_rate": 1.2917209279575625e-08,
|
| 3530 |
+
"loss": 1.4899,
|
| 3531 |
+
"step": 1174
|
| 3532 |
+
},
|
| 3533 |
+
{
|
| 3534 |
+
"epoch": 3.82,
|
| 3535 |
+
"learning_rate": 1.208574730894929e-08,
|
| 3536 |
+
"loss": 1.3946,
|
| 3537 |
+
"step": 1176
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 3.83,
|
| 3541 |
+
"learning_rate": 1.128178038476324e-08,
|
| 3542 |
+
"loss": 1.3951,
|
| 3543 |
+
"step": 1178
|
| 3544 |
+
},
|
| 3545 |
+
{
|
| 3546 |
+
"epoch": 3.83,
|
| 3547 |
+
"learning_rate": 1.0505330882550678e-08,
|
| 3548 |
+
"loss": 1.3615,
|
| 3549 |
+
"step": 1180
|
| 3550 |
+
},
|
| 3551 |
+
{
|
| 3552 |
+
"epoch": 3.84,
|
| 3553 |
+
"learning_rate": 9.756420411995647e-09,
|
| 3554 |
+
"loss": 1.5179,
|
| 3555 |
+
"step": 1182
|
| 3556 |
+
},
|
| 3557 |
+
{
|
| 3558 |
+
"epoch": 3.85,
|
| 3559 |
+
"learning_rate": 9.035069816332619e-09,
|
| 3560 |
+
"loss": 1.4684,
|
| 3561 |
+
"step": 1184
|
| 3562 |
+
},
|
| 3563 |
+
{
|
| 3564 |
+
"epoch": 3.85,
|
| 3565 |
+
"learning_rate": 8.341299171765514e-09,
|
| 3566 |
+
"loss": 1.422,
|
| 3567 |
+
"step": 1186
|
| 3568 |
+
},
|
| 3569 |
+
{
|
| 3570 |
+
"epoch": 3.86,
|
| 3571 |
+
"learning_rate": 7.675127786909485e-09,
|
| 3572 |
+
"loss": 1.4584,
|
| 3573 |
+
"step": 1188
|
| 3574 |
+
},
|
| 3575 |
+
{
|
| 3576 |
+
"epoch": 3.87,
|
| 3577 |
+
"learning_rate": 7.036574202253343e-09,
|
| 3578 |
+
"loss": 1.4359,
|
| 3579 |
+
"step": 1190
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 3.87,
|
| 3583 |
+
"learning_rate": 6.425656189643747e-09,
|
| 3584 |
+
"loss": 1.4749,
|
| 3585 |
+
"step": 1192
|
| 3586 |
+
},
|
| 3587 |
+
{
|
| 3588 |
+
"epoch": 3.88,
|
| 3589 |
+
"learning_rate": 5.842390751790049e-09,
|
| 3590 |
+
"loss": 1.3864,
|
| 3591 |
+
"step": 1194
|
| 3592 |
+
},
|
| 3593 |
+
{
|
| 3594 |
+
"epoch": 3.89,
|
| 3595 |
+
"learning_rate": 5.286794121791782e-09,
|
| 3596 |
+
"loss": 1.4197,
|
| 3597 |
+
"step": 1196
|
| 3598 |
+
},
|
| 3599 |
+
{
|
| 3600 |
+
"epoch": 3.89,
|
| 3601 |
+
"learning_rate": 4.758881762686462e-09,
|
| 3602 |
+
"loss": 1.4282,
|
| 3603 |
+
"step": 1198
|
| 3604 |
+
},
|
| 3605 |
+
{
|
| 3606 |
+
"epoch": 3.9,
|
| 3607 |
+
"learning_rate": 4.2586683670193805e-09,
|
| 3608 |
+
"loss": 1.4736,
|
| 3609 |
+
"step": 1200
|
| 3610 |
+
},
|
| 3611 |
+
{
|
| 3612 |
+
"epoch": 3.91,
|
| 3613 |
+
"learning_rate": 3.786167856434375e-09,
|
| 3614 |
+
"loss": 1.447,
|
| 3615 |
+
"step": 1202
|
| 3616 |
+
},
|
| 3617 |
+
{
|
| 3618 |
+
"epoch": 3.91,
|
| 3619 |
+
"learning_rate": 3.3413933812870276e-09,
|
| 3620 |
+
"loss": 1.4103,
|
| 3621 |
+
"step": 1204
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 3.92,
|
| 3625 |
+
"learning_rate": 2.9243573202778484e-09,
|
| 3626 |
+
"loss": 1.4619,
|
| 3627 |
+
"step": 1206
|
| 3628 |
+
},
|
| 3629 |
+
{
|
| 3630 |
+
"epoch": 3.93,
|
| 3631 |
+
"learning_rate": 2.5350712801084363e-09,
|
| 3632 |
+
"loss": 1.4229,
|
| 3633 |
+
"step": 1208
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 3.93,
|
| 3637 |
+
"learning_rate": 2.173546095158074e-09,
|
| 3638 |
+
"loss": 1.403,
|
| 3639 |
+
"step": 1210
|
| 3640 |
+
},
|
| 3641 |
+
{
|
| 3642 |
+
"epoch": 3.94,
|
| 3643 |
+
"learning_rate": 1.8397918271824131e-09,
|
| 3644 |
+
"loss": 1.448,
|
| 3645 |
+
"step": 1212
|
| 3646 |
+
},
|
| 3647 |
+
{
|
| 3648 |
+
"epoch": 3.94,
|
| 3649 |
+
"learning_rate": 1.5338177650332517e-09,
|
| 3650 |
+
"loss": 1.4532,
|
| 3651 |
+
"step": 1214
|
| 3652 |
+
},
|
| 3653 |
+
{
|
| 3654 |
+
"epoch": 3.95,
|
| 3655 |
+
"learning_rate": 1.255632424400077e-09,
|
| 3656 |
+
"loss": 1.4438,
|
| 3657 |
+
"step": 1216
|
| 3658 |
+
},
|
| 3659 |
+
{
|
| 3660 |
+
"epoch": 3.96,
|
| 3661 |
+
"learning_rate": 1.0052435475731425e-09,
|
| 3662 |
+
"loss": 1.4381,
|
| 3663 |
+
"step": 1218
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 3.96,
|
| 3667 |
+
"learning_rate": 7.826581032279734e-10,
|
| 3668 |
+
"loss": 1.3884,
|
| 3669 |
+
"step": 1220
|
| 3670 |
+
},
|
| 3671 |
+
{
|
| 3672 |
+
"epoch": 3.97,
|
| 3673 |
+
"learning_rate": 5.878822862310784e-10,
|
| 3674 |
+
"loss": 1.3937,
|
| 3675 |
+
"step": 1222
|
| 3676 |
+
},
|
| 3677 |
+
{
|
| 3678 |
+
"epoch": 3.98,
|
| 3679 |
+
"learning_rate": 4.209215174681979e-10,
|
| 3680 |
+
"loss": 1.4549,
|
| 3681 |
+
"step": 1224
|
| 3682 |
+
},
|
| 3683 |
+
{
|
| 3684 |
+
"epoch": 3.98,
|
| 3685 |
+
"learning_rate": 2.8178044369286945e-10,
|
| 3686 |
+
"loss": 1.3781,
|
| 3687 |
+
"step": 1226
|
| 3688 |
+
},
|
| 3689 |
+
{
|
| 3690 |
+
"epoch": 3.99,
|
| 3691 |
+
"learning_rate": 1.7046293739741979e-10,
|
| 3692 |
+
"loss": 1.45,
|
| 3693 |
+
"step": 1228
|
| 3694 |
+
},
|
| 3695 |
+
{
|
| 3696 |
+
"epoch": 3.99,
|
| 3697 |
+
"step": 1228,
|
| 3698 |
+
"total_flos": 5.36747403420631e+16,
|
| 3699 |
+
"train_loss": 1.5255624304376907,
|
| 3700 |
+
"train_runtime": 12464.0287,
|
| 3701 |
+
"train_samples_per_second": 6.32,
|
| 3702 |
+
"train_steps_per_second": 0.099
|
| 3703 |
+
}
|
| 3704 |
+
],
|
| 3705 |
+
"logging_steps": 2,
|
| 3706 |
+
"max_steps": 1228,
|
| 3707 |
+
"num_input_tokens_seen": 0,
|
| 3708 |
+
"num_train_epochs": 4,
|
| 3709 |
+
"save_steps": 9000,
|
| 3710 |
+
"total_flos": 5.36747403420631e+16,
|
| 3711 |
+
"train_batch_size": 2,
|
| 3712 |
+
"trial_name": null,
|
| 3713 |
+
"trial_params": null
|
| 3714 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03e2624f18858318179d865ff040c26dce2f960778809053c2a4321d9ea58af0
|
| 3 |
+
size 5883
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|