Commit
•
7e2c7a4
1
Parent(s):
3454939
Upload 11 files
Browse files- README.md +202 -1
- adapter_config.json +31 -0
- adapter_model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +24 -0
- tokenizer.model +3 -0
- tokenizer_config.json +44 -0
- trainer_state.json +2541 -0
- training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,204 @@
|
|
| 1 |
---
|
| 2 |
-
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
base_model: mistralai/Mistral-7B-v0.1
|
| 4 |
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
|
| 201 |
+
|
| 202 |
+
### Framework versions
|
| 203 |
+
|
| 204 |
+
- PEFT 0.7.1
|
adapter_config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "mistralai/Mistral-7B-v0.1",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"loftq_config": {},
|
| 12 |
+
"lora_alpha": 256,
|
| 13 |
+
"lora_dropout": 0.05,
|
| 14 |
+
"megatron_config": null,
|
| 15 |
+
"megatron_core": "megatron.core",
|
| 16 |
+
"modules_to_save": null,
|
| 17 |
+
"peft_type": "LORA",
|
| 18 |
+
"r": 128,
|
| 19 |
+
"rank_pattern": {},
|
| 20 |
+
"revision": null,
|
| 21 |
+
"target_modules": [
|
| 22 |
+
"gate_proj",
|
| 23 |
+
"down_proj",
|
| 24 |
+
"o_proj",
|
| 25 |
+
"k_proj",
|
| 26 |
+
"up_proj",
|
| 27 |
+
"q_proj",
|
| 28 |
+
"v_proj"
|
| 29 |
+
],
|
| 30 |
+
"task_type": "CAUSAL_LM"
|
| 31 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a379098e4734f3edaebb67d1f9afc04b3463045e093d5477e623e5b55906a70
|
| 3 |
+
size 1342238560
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9eb1327f06e61475f59b83af7a0869d8bce77801bc0974f50042c65acbcf2b9f
|
| 3 |
+
size 2684632250
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e71790ef20c8decc5a8f4576215a2fb23035dc52804a90e2ecd96e8b1b65065
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9097acc6fb628680c39b6e6266f0991b3847bbf4f3c97580caf4056ead6a3e70
|
| 3 |
+
size 1064
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
|
| 3 |
+
size 493443
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
"additional_special_tokens": [],
|
| 31 |
+
"bos_token": "<s>",
|
| 32 |
+
"clean_up_tokenization_spaces": false,
|
| 33 |
+
"eos_token": "</s>",
|
| 34 |
+
"legacy": true,
|
| 35 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 36 |
+
"pad_token": "</s>",
|
| 37 |
+
"padding_side": "right",
|
| 38 |
+
"sp_model_kwargs": {},
|
| 39 |
+
"spaces_between_special_tokens": false,
|
| 40 |
+
"split_special_tokens": false,
|
| 41 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 42 |
+
"unk_token": "<unk>",
|
| 43 |
+
"use_default_system_prompt": false
|
| 44 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2541 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.7613293843210234,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2100,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 2.2727272727272728e-06,
|
| 14 |
+
"loss": 0.8517,
|
| 15 |
+
"step": 5
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 4.5454545454545455e-06,
|
| 20 |
+
"loss": 0.7344,
|
| 21 |
+
"step": 10
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 6.818181818181818e-06,
|
| 26 |
+
"loss": 0.6299,
|
| 27 |
+
"step": 15
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.02,
|
| 31 |
+
"learning_rate": 9.090909090909091e-06,
|
| 32 |
+
"loss": 0.5712,
|
| 33 |
+
"step": 20
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.02,
|
| 37 |
+
"learning_rate": 1.1363636363636365e-05,
|
| 38 |
+
"loss": 0.5381,
|
| 39 |
+
"step": 25
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.03,
|
| 43 |
+
"learning_rate": 1.3636363636363637e-05,
|
| 44 |
+
"loss": 0.522,
|
| 45 |
+
"step": 30
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03,
|
| 49 |
+
"learning_rate": 1.590909090909091e-05,
|
| 50 |
+
"loss": 0.5062,
|
| 51 |
+
"step": 35
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
+
"learning_rate": 1.8181818181818182e-05,
|
| 56 |
+
"loss": 0.4908,
|
| 57 |
+
"step": 40
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.04,
|
| 61 |
+
"learning_rate": 2.0454545454545457e-05,
|
| 62 |
+
"loss": 0.4735,
|
| 63 |
+
"step": 45
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.04,
|
| 67 |
+
"learning_rate": 2.272727272727273e-05,
|
| 68 |
+
"loss": 0.4762,
|
| 69 |
+
"step": 50
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.05,
|
| 73 |
+
"learning_rate": 2.5e-05,
|
| 74 |
+
"loss": 0.4594,
|
| 75 |
+
"step": 55
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.05,
|
| 79 |
+
"learning_rate": 2.7272727272727273e-05,
|
| 80 |
+
"loss": 0.4551,
|
| 81 |
+
"step": 60
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.05,
|
| 85 |
+
"learning_rate": 2.954545454545455e-05,
|
| 86 |
+
"loss": 0.4548,
|
| 87 |
+
"step": 65
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06,
|
| 91 |
+
"learning_rate": 3.181818181818182e-05,
|
| 92 |
+
"loss": 0.4586,
|
| 93 |
+
"step": 70
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.06,
|
| 97 |
+
"learning_rate": 3.409090909090909e-05,
|
| 98 |
+
"loss": 0.4452,
|
| 99 |
+
"step": 75
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.07,
|
| 103 |
+
"learning_rate": 3.6363636363636364e-05,
|
| 104 |
+
"loss": 0.4494,
|
| 105 |
+
"step": 80
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.07,
|
| 109 |
+
"learning_rate": 3.8636363636363636e-05,
|
| 110 |
+
"loss": 0.4385,
|
| 111 |
+
"step": 85
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.08,
|
| 115 |
+
"learning_rate": 4.0909090909090915e-05,
|
| 116 |
+
"loss": 0.4245,
|
| 117 |
+
"step": 90
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.08,
|
| 121 |
+
"learning_rate": 4.318181818181819e-05,
|
| 122 |
+
"loss": 0.441,
|
| 123 |
+
"step": 95
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.08,
|
| 127 |
+
"learning_rate": 4.545454545454546e-05,
|
| 128 |
+
"loss": 0.4316,
|
| 129 |
+
"step": 100
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.09,
|
| 133 |
+
"learning_rate": 4.772727272727273e-05,
|
| 134 |
+
"loss": 0.423,
|
| 135 |
+
"step": 105
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.09,
|
| 139 |
+
"learning_rate": 5e-05,
|
| 140 |
+
"loss": 0.4165,
|
| 141 |
+
"step": 110
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.1,
|
| 145 |
+
"learning_rate": 4.999974326096354e-05,
|
| 146 |
+
"loss": 0.4202,
|
| 147 |
+
"step": 115
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.1,
|
| 151 |
+
"learning_rate": 4.9998973049127335e-05,
|
| 152 |
+
"loss": 0.4205,
|
| 153 |
+
"step": 120
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.1,
|
| 157 |
+
"learning_rate": 4.9997689380310875e-05,
|
| 158 |
+
"loss": 0.4304,
|
| 159 |
+
"step": 125
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.11,
|
| 163 |
+
"learning_rate": 4.999589228087958e-05,
|
| 164 |
+
"loss": 0.4138,
|
| 165 |
+
"step": 130
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.11,
|
| 169 |
+
"learning_rate": 4.9993581787744303e-05,
|
| 170 |
+
"loss": 0.4202,
|
| 171 |
+
"step": 135
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.12,
|
| 175 |
+
"learning_rate": 4.999075794836054e-05,
|
| 176 |
+
"loss": 0.3964,
|
| 177 |
+
"step": 140
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.12,
|
| 181 |
+
"learning_rate": 4.9987420820727495e-05,
|
| 182 |
+
"loss": 0.4072,
|
| 183 |
+
"step": 145
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.13,
|
| 187 |
+
"learning_rate": 4.998357047338682e-05,
|
| 188 |
+
"loss": 0.4093,
|
| 189 |
+
"step": 150
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.13,
|
| 193 |
+
"learning_rate": 4.9979206985421286e-05,
|
| 194 |
+
"loss": 0.4168,
|
| 195 |
+
"step": 155
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.13,
|
| 199 |
+
"learning_rate": 4.99743304464531e-05,
|
| 200 |
+
"loss": 0.3903,
|
| 201 |
+
"step": 160
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.14,
|
| 205 |
+
"learning_rate": 4.99689409566421e-05,
|
| 206 |
+
"loss": 0.4132,
|
| 207 |
+
"step": 165
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.14,
|
| 211 |
+
"learning_rate": 4.996303862668367e-05,
|
| 212 |
+
"loss": 0.4048,
|
| 213 |
+
"step": 170
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.15,
|
| 217 |
+
"learning_rate": 4.9956623577806505e-05,
|
| 218 |
+
"loss": 0.4046,
|
| 219 |
+
"step": 175
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.15,
|
| 223 |
+
"learning_rate": 4.994969594177007e-05,
|
| 224 |
+
"loss": 0.3927,
|
| 225 |
+
"step": 180
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.16,
|
| 229 |
+
"learning_rate": 4.994225586086195e-05,
|
| 230 |
+
"loss": 0.3953,
|
| 231 |
+
"step": 185
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.16,
|
| 235 |
+
"learning_rate": 4.993430348789486e-05,
|
| 236 |
+
"loss": 0.3811,
|
| 237 |
+
"step": 190
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.16,
|
| 241 |
+
"learning_rate": 4.9925838986203584e-05,
|
| 242 |
+
"loss": 0.3765,
|
| 243 |
+
"step": 195
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.17,
|
| 247 |
+
"learning_rate": 4.991686252964155e-05,
|
| 248 |
+
"loss": 0.4027,
|
| 249 |
+
"step": 200
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.17,
|
| 253 |
+
"learning_rate": 4.990737430257731e-05,
|
| 254 |
+
"loss": 0.3872,
|
| 255 |
+
"step": 205
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.18,
|
| 259 |
+
"learning_rate": 4.989737449989073e-05,
|
| 260 |
+
"loss": 0.3941,
|
| 261 |
+
"step": 210
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.18,
|
| 265 |
+
"learning_rate": 4.988686332696897e-05,
|
| 266 |
+
"loss": 0.3769,
|
| 267 |
+
"step": 215
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.18,
|
| 271 |
+
"learning_rate": 4.987584099970231e-05,
|
| 272 |
+
"loss": 0.3948,
|
| 273 |
+
"step": 220
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.19,
|
| 277 |
+
"learning_rate": 4.9864307744479696e-05,
|
| 278 |
+
"loss": 0.3926,
|
| 279 |
+
"step": 225
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.19,
|
| 283 |
+
"learning_rate": 4.9852263798184065e-05,
|
| 284 |
+
"loss": 0.3905,
|
| 285 |
+
"step": 230
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.2,
|
| 289 |
+
"learning_rate": 4.983970940818751e-05,
|
| 290 |
+
"loss": 0.3645,
|
| 291 |
+
"step": 235
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.2,
|
| 295 |
+
"learning_rate": 4.9826644832346184e-05,
|
| 296 |
+
"loss": 0.3845,
|
| 297 |
+
"step": 240
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.21,
|
| 301 |
+
"learning_rate": 4.981307033899503e-05,
|
| 302 |
+
"loss": 0.3629,
|
| 303 |
+
"step": 245
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.21,
|
| 307 |
+
"learning_rate": 4.9798986206942235e-05,
|
| 308 |
+
"loss": 0.3636,
|
| 309 |
+
"step": 250
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.21,
|
| 313 |
+
"learning_rate": 4.97843927254635e-05,
|
| 314 |
+
"loss": 0.373,
|
| 315 |
+
"step": 255
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.22,
|
| 319 |
+
"learning_rate": 4.976929019429616e-05,
|
| 320 |
+
"loss": 0.3639,
|
| 321 |
+
"step": 260
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.22,
|
| 325 |
+
"learning_rate": 4.975367892363294e-05,
|
| 326 |
+
"loss": 0.377,
|
| 327 |
+
"step": 265
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.23,
|
| 331 |
+
"learning_rate": 4.973755923411566e-05,
|
| 332 |
+
"loss": 0.3733,
|
| 333 |
+
"step": 270
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.23,
|
| 337 |
+
"learning_rate": 4.972093145682859e-05,
|
| 338 |
+
"loss": 0.3685,
|
| 339 |
+
"step": 275
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.23,
|
| 343 |
+
"learning_rate": 4.970379593329171e-05,
|
| 344 |
+
"loss": 0.3733,
|
| 345 |
+
"step": 280
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.24,
|
| 349 |
+
"learning_rate": 4.968615301545362e-05,
|
| 350 |
+
"loss": 0.3571,
|
| 351 |
+
"step": 285
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.24,
|
| 355 |
+
"learning_rate": 4.96680030656844e-05,
|
| 356 |
+
"loss": 0.3446,
|
| 357 |
+
"step": 290
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.25,
|
| 361 |
+
"learning_rate": 4.9649346456768085e-05,
|
| 362 |
+
"loss": 0.3608,
|
| 363 |
+
"step": 295
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.25,
|
| 367 |
+
"learning_rate": 4.9630183571895075e-05,
|
| 368 |
+
"loss": 0.3463,
|
| 369 |
+
"step": 300
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.26,
|
| 373 |
+
"learning_rate": 4.9610514804654195e-05,
|
| 374 |
+
"loss": 0.3669,
|
| 375 |
+
"step": 305
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.26,
|
| 379 |
+
"learning_rate": 4.95903405590247e-05,
|
| 380 |
+
"loss": 0.3549,
|
| 381 |
+
"step": 310
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.26,
|
| 385 |
+
"learning_rate": 4.956966124936787e-05,
|
| 386 |
+
"loss": 0.345,
|
| 387 |
+
"step": 315
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.27,
|
| 391 |
+
"learning_rate": 4.954847730041862e-05,
|
| 392 |
+
"loss": 0.346,
|
| 393 |
+
"step": 320
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.27,
|
| 397 |
+
"learning_rate": 4.952678914727665e-05,
|
| 398 |
+
"loss": 0.3579,
|
| 399 |
+
"step": 325
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.28,
|
| 403 |
+
"learning_rate": 4.950459723539763e-05,
|
| 404 |
+
"loss": 0.352,
|
| 405 |
+
"step": 330
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.28,
|
| 409 |
+
"learning_rate": 4.948190202058395e-05,
|
| 410 |
+
"loss": 0.3605,
|
| 411 |
+
"step": 335
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.29,
|
| 415 |
+
"learning_rate": 4.9458703968975425e-05,
|
| 416 |
+
"loss": 0.3575,
|
| 417 |
+
"step": 340
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.29,
|
| 421 |
+
"learning_rate": 4.943500355703968e-05,
|
| 422 |
+
"loss": 0.3551,
|
| 423 |
+
"step": 345
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.29,
|
| 427 |
+
"learning_rate": 4.94108012715624e-05,
|
| 428 |
+
"loss": 0.3419,
|
| 429 |
+
"step": 350
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.3,
|
| 433 |
+
"learning_rate": 4.938609760963729e-05,
|
| 434 |
+
"loss": 0.3469,
|
| 435 |
+
"step": 355
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.3,
|
| 439 |
+
"learning_rate": 4.9360893078655904e-05,
|
| 440 |
+
"loss": 0.3629,
|
| 441 |
+
"step": 360
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.31,
|
| 445 |
+
"learning_rate": 4.93351881962972e-05,
|
| 446 |
+
"loss": 0.3504,
|
| 447 |
+
"step": 365
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.31,
|
| 451 |
+
"learning_rate": 4.930898349051693e-05,
|
| 452 |
+
"loss": 0.3298,
|
| 453 |
+
"step": 370
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.31,
|
| 457 |
+
"learning_rate": 4.9282279499536735e-05,
|
| 458 |
+
"loss": 0.3287,
|
| 459 |
+
"step": 375
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.32,
|
| 463 |
+
"learning_rate": 4.92550767718332e-05,
|
| 464 |
+
"loss": 0.3621,
|
| 465 |
+
"step": 380
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.32,
|
| 469 |
+
"learning_rate": 4.922737586612648e-05,
|
| 470 |
+
"loss": 0.3363,
|
| 471 |
+
"step": 385
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.33,
|
| 475 |
+
"learning_rate": 4.919917735136889e-05,
|
| 476 |
+
"loss": 0.3387,
|
| 477 |
+
"step": 390
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.33,
|
| 481 |
+
"learning_rate": 4.9170481806733174e-05,
|
| 482 |
+
"loss": 0.3391,
|
| 483 |
+
"step": 395
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.34,
|
| 487 |
+
"learning_rate": 4.914128982160067e-05,
|
| 488 |
+
"loss": 0.3481,
|
| 489 |
+
"step": 400
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.34,
|
| 493 |
+
"learning_rate": 4.911160199554913e-05,
|
| 494 |
+
"loss": 0.323,
|
| 495 |
+
"step": 405
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.34,
|
| 499 |
+
"learning_rate": 4.908141893834047e-05,
|
| 500 |
+
"loss": 0.3333,
|
| 501 |
+
"step": 410
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.35,
|
| 505 |
+
"learning_rate": 4.905074126990822e-05,
|
| 506 |
+
"loss": 0.3343,
|
| 507 |
+
"step": 415
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.35,
|
| 511 |
+
"learning_rate": 4.9019569620344775e-05,
|
| 512 |
+
"loss": 0.321,
|
| 513 |
+
"step": 420
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.36,
|
| 517 |
+
"learning_rate": 4.898790462988848e-05,
|
| 518 |
+
"loss": 0.342,
|
| 519 |
+
"step": 425
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.36,
|
| 523 |
+
"learning_rate": 4.895574694891046e-05,
|
| 524 |
+
"loss": 0.3157,
|
| 525 |
+
"step": 430
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.36,
|
| 529 |
+
"learning_rate": 4.892309723790129e-05,
|
| 530 |
+
"loss": 0.3399,
|
| 531 |
+
"step": 435
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.37,
|
| 535 |
+
"learning_rate": 4.8889956167457384e-05,
|
| 536 |
+
"loss": 0.3386,
|
| 537 |
+
"step": 440
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.37,
|
| 541 |
+
"learning_rate": 4.885632441826727e-05,
|
| 542 |
+
"loss": 0.3327,
|
| 543 |
+
"step": 445
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.38,
|
| 547 |
+
"learning_rate": 4.8822202681097574e-05,
|
| 548 |
+
"loss": 0.3405,
|
| 549 |
+
"step": 450
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.38,
|
| 553 |
+
"learning_rate": 4.878759165677885e-05,
|
| 554 |
+
"loss": 0.3366,
|
| 555 |
+
"step": 455
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.39,
|
| 559 |
+
"learning_rate": 4.875249205619118e-05,
|
| 560 |
+
"loss": 0.3316,
|
| 561 |
+
"step": 460
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.39,
|
| 565 |
+
"learning_rate": 4.871690460024958e-05,
|
| 566 |
+
"loss": 0.3396,
|
| 567 |
+
"step": 465
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.39,
|
| 571 |
+
"learning_rate": 4.8680830019889186e-05,
|
| 572 |
+
"loss": 0.3425,
|
| 573 |
+
"step": 470
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.4,
|
| 577 |
+
"learning_rate": 4.864426905605022e-05,
|
| 578 |
+
"loss": 0.3353,
|
| 579 |
+
"step": 475
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.4,
|
| 583 |
+
"learning_rate": 4.860722245966283e-05,
|
| 584 |
+
"loss": 0.3086,
|
| 585 |
+
"step": 480
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.41,
|
| 589 |
+
"learning_rate": 4.8569690991631605e-05,
|
| 590 |
+
"loss": 0.3217,
|
| 591 |
+
"step": 485
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.41,
|
| 595 |
+
"learning_rate": 4.8531675422819975e-05,
|
| 596 |
+
"loss": 0.348,
|
| 597 |
+
"step": 490
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.42,
|
| 601 |
+
"learning_rate": 4.84931765340344e-05,
|
| 602 |
+
"loss": 0.3145,
|
| 603 |
+
"step": 495
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.42,
|
| 607 |
+
"learning_rate": 4.8454195116008264e-05,
|
| 608 |
+
"loss": 0.3201,
|
| 609 |
+
"step": 500
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.42,
|
| 613 |
+
"learning_rate": 4.8414731969385725e-05,
|
| 614 |
+
"loss": 0.3334,
|
| 615 |
+
"step": 505
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 0.43,
|
| 619 |
+
"learning_rate": 4.8374787904705185e-05,
|
| 620 |
+
"loss": 0.3282,
|
| 621 |
+
"step": 510
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.43,
|
| 625 |
+
"learning_rate": 4.833436374238271e-05,
|
| 626 |
+
"loss": 0.3309,
|
| 627 |
+
"step": 515
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.44,
|
| 631 |
+
"learning_rate": 4.829346031269513e-05,
|
| 632 |
+
"loss": 0.3007,
|
| 633 |
+
"step": 520
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.44,
|
| 637 |
+
"learning_rate": 4.825207845576303e-05,
|
| 638 |
+
"loss": 0.3224,
|
| 639 |
+
"step": 525
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.44,
|
| 643 |
+
"learning_rate": 4.8210219021533434e-05,
|
| 644 |
+
"loss": 0.3265,
|
| 645 |
+
"step": 530
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 0.45,
|
| 649 |
+
"learning_rate": 4.816788286976243e-05,
|
| 650 |
+
"loss": 0.324,
|
| 651 |
+
"step": 535
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.45,
|
| 655 |
+
"learning_rate": 4.8125070869997416e-05,
|
| 656 |
+
"loss": 0.3143,
|
| 657 |
+
"step": 540
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.46,
|
| 661 |
+
"learning_rate": 4.808178390155934e-05,
|
| 662 |
+
"loss": 0.3159,
|
| 663 |
+
"step": 545
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.46,
|
| 667 |
+
"learning_rate": 4.8038022853524564e-05,
|
| 668 |
+
"loss": 0.3139,
|
| 669 |
+
"step": 550
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.47,
|
| 673 |
+
"learning_rate": 4.799378862470663e-05,
|
| 674 |
+
"loss": 0.3255,
|
| 675 |
+
"step": 555
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.47,
|
| 679 |
+
"learning_rate": 4.7949082123637797e-05,
|
| 680 |
+
"loss": 0.3301,
|
| 681 |
+
"step": 560
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.47,
|
| 685 |
+
"learning_rate": 4.790390426855039e-05,
|
| 686 |
+
"loss": 0.3108,
|
| 687 |
+
"step": 565
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.48,
|
| 691 |
+
"learning_rate": 4.785825598735793e-05,
|
| 692 |
+
"loss": 0.3272,
|
| 693 |
+
"step": 570
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.48,
|
| 697 |
+
"learning_rate": 4.781213821763606e-05,
|
| 698 |
+
"loss": 0.3175,
|
| 699 |
+
"step": 575
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.49,
|
| 703 |
+
"learning_rate": 4.7765551906603346e-05,
|
| 704 |
+
"loss": 0.3145,
|
| 705 |
+
"step": 580
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.49,
|
| 709 |
+
"learning_rate": 4.771849801110173e-05,
|
| 710 |
+
"loss": 0.3123,
|
| 711 |
+
"step": 585
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.49,
|
| 715 |
+
"learning_rate": 4.7670977497576984e-05,
|
| 716 |
+
"loss": 0.3119,
|
| 717 |
+
"step": 590
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5,
|
| 721 |
+
"learning_rate": 4.7622991342058755e-05,
|
| 722 |
+
"loss": 0.2877,
|
| 723 |
+
"step": 595
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.5,
|
| 727 |
+
"learning_rate": 4.757454053014059e-05,
|
| 728 |
+
"loss": 0.3158,
|
| 729 |
+
"step": 600
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.51,
|
| 733 |
+
"learning_rate": 4.752562605695968e-05,
|
| 734 |
+
"loss": 0.3138,
|
| 735 |
+
"step": 605
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.51,
|
| 739 |
+
"learning_rate": 4.74762489271764e-05,
|
| 740 |
+
"loss": 0.3116,
|
| 741 |
+
"step": 610
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.52,
|
| 745 |
+
"learning_rate": 4.7426410154953684e-05,
|
| 746 |
+
"loss": 0.2884,
|
| 747 |
+
"step": 615
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.52,
|
| 751 |
+
"learning_rate": 4.737611076393621e-05,
|
| 752 |
+
"loss": 0.3022,
|
| 753 |
+
"step": 620
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.52,
|
| 757 |
+
"learning_rate": 4.7325351787229336e-05,
|
| 758 |
+
"loss": 0.305,
|
| 759 |
+
"step": 625
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.53,
|
| 763 |
+
"learning_rate": 4.727413426737793e-05,
|
| 764 |
+
"loss": 0.2897,
|
| 765 |
+
"step": 630
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.53,
|
| 769 |
+
"learning_rate": 4.7222459256344945e-05,
|
| 770 |
+
"loss": 0.3071,
|
| 771 |
+
"step": 635
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.54,
|
| 775 |
+
"learning_rate": 4.717032781548977e-05,
|
| 776 |
+
"loss": 0.2932,
|
| 777 |
+
"step": 640
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.54,
|
| 781 |
+
"learning_rate": 4.7117741015546466e-05,
|
| 782 |
+
"loss": 0.3022,
|
| 783 |
+
"step": 645
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.55,
|
| 787 |
+
"learning_rate": 4.7064699936601797e-05,
|
| 788 |
+
"loss": 0.2993,
|
| 789 |
+
"step": 650
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.55,
|
| 793 |
+
"learning_rate": 4.7011205668073e-05,
|
| 794 |
+
"loss": 0.3062,
|
| 795 |
+
"step": 655
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.55,
|
| 799 |
+
"learning_rate": 4.695725930868543e-05,
|
| 800 |
+
"loss": 0.2954,
|
| 801 |
+
"step": 660
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.56,
|
| 805 |
+
"learning_rate": 4.6902861966449984e-05,
|
| 806 |
+
"loss": 0.2911,
|
| 807 |
+
"step": 665
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.56,
|
| 811 |
+
"learning_rate": 4.684801475864038e-05,
|
| 812 |
+
"loss": 0.2939,
|
| 813 |
+
"step": 670
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.57,
|
| 817 |
+
"learning_rate": 4.6792718811770145e-05,
|
| 818 |
+
"loss": 0.2938,
|
| 819 |
+
"step": 675
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.57,
|
| 823 |
+
"learning_rate": 4.6736975261569536e-05,
|
| 824 |
+
"loss": 0.3219,
|
| 825 |
+
"step": 680
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.57,
|
| 829 |
+
"learning_rate": 4.668078525296218e-05,
|
| 830 |
+
"loss": 0.2895,
|
| 831 |
+
"step": 685
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.58,
|
| 835 |
+
"learning_rate": 4.662414994004157e-05,
|
| 836 |
+
"loss": 0.2827,
|
| 837 |
+
"step": 690
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.58,
|
| 841 |
+
"learning_rate": 4.656707048604737e-05,
|
| 842 |
+
"loss": 0.2926,
|
| 843 |
+
"step": 695
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.59,
|
| 847 |
+
"learning_rate": 4.650954806334148e-05,
|
| 848 |
+
"loss": 0.2956,
|
| 849 |
+
"step": 700
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.59,
|
| 853 |
+
"learning_rate": 4.645158385338403e-05,
|
| 854 |
+
"loss": 0.2985,
|
| 855 |
+
"step": 705
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.6,
|
| 859 |
+
"learning_rate": 4.6393179046709047e-05,
|
| 860 |
+
"loss": 0.2973,
|
| 861 |
+
"step": 710
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.6,
|
| 865 |
+
"learning_rate": 4.633433484290003e-05,
|
| 866 |
+
"loss": 0.2807,
|
| 867 |
+
"step": 715
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.6,
|
| 871 |
+
"learning_rate": 4.6275052450565314e-05,
|
| 872 |
+
"loss": 0.2908,
|
| 873 |
+
"step": 720
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.61,
|
| 877 |
+
"learning_rate": 4.621533308731324e-05,
|
| 878 |
+
"loss": 0.2898,
|
| 879 |
+
"step": 725
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.61,
|
| 883 |
+
"learning_rate": 4.615517797972717e-05,
|
| 884 |
+
"loss": 0.2905,
|
| 885 |
+
"step": 730
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.62,
|
| 889 |
+
"learning_rate": 4.609458836334023e-05,
|
| 890 |
+
"loss": 0.2965,
|
| 891 |
+
"step": 735
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.62,
|
| 895 |
+
"learning_rate": 4.603356548261001e-05,
|
| 896 |
+
"loss": 0.2864,
|
| 897 |
+
"step": 740
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.62,
|
| 901 |
+
"learning_rate": 4.5972110590892956e-05,
|
| 902 |
+
"loss": 0.2808,
|
| 903 |
+
"step": 745
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.63,
|
| 907 |
+
"learning_rate": 4.5910224950418636e-05,
|
| 908 |
+
"loss": 0.2791,
|
| 909 |
+
"step": 750
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.63,
|
| 913 |
+
"learning_rate": 4.584790983226383e-05,
|
| 914 |
+
"loss": 0.2999,
|
| 915 |
+
"step": 755
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.64,
|
| 919 |
+
"learning_rate": 4.578516651632642e-05,
|
| 920 |
+
"loss": 0.2816,
|
| 921 |
+
"step": 760
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.64,
|
| 925 |
+
"learning_rate": 4.572199629129907e-05,
|
| 926 |
+
"loss": 0.2672,
|
| 927 |
+
"step": 765
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.65,
|
| 931 |
+
"learning_rate": 4.56584004546428e-05,
|
| 932 |
+
"loss": 0.2899,
|
| 933 |
+
"step": 770
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.65,
|
| 937 |
+
"learning_rate": 4.5594380312560325e-05,
|
| 938 |
+
"loss": 0.2807,
|
| 939 |
+
"step": 775
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.65,
|
| 943 |
+
"learning_rate": 4.5529937179969206e-05,
|
| 944 |
+
"loss": 0.271,
|
| 945 |
+
"step": 780
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.66,
|
| 949 |
+
"learning_rate": 4.546507238047486e-05,
|
| 950 |
+
"loss": 0.2807,
|
| 951 |
+
"step": 785
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.66,
|
| 955 |
+
"learning_rate": 4.539978724634338e-05,
|
| 956 |
+
"loss": 0.2837,
|
| 957 |
+
"step": 790
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.67,
|
| 961 |
+
"learning_rate": 4.533408311847417e-05,
|
| 962 |
+
"loss": 0.2899,
|
| 963 |
+
"step": 795
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.67,
|
| 967 |
+
"learning_rate": 4.526796134637237e-05,
|
| 968 |
+
"loss": 0.2836,
|
| 969 |
+
"step": 800
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.68,
|
| 973 |
+
"learning_rate": 4.5201423288121205e-05,
|
| 974 |
+
"loss": 0.2717,
|
| 975 |
+
"step": 805
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.68,
|
| 979 |
+
"learning_rate": 4.5134470310354015e-05,
|
| 980 |
+
"loss": 0.2741,
|
| 981 |
+
"step": 810
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.68,
|
| 985 |
+
"learning_rate": 4.506710378822625e-05,
|
| 986 |
+
"loss": 0.2935,
|
| 987 |
+
"step": 815
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.69,
|
| 991 |
+
"learning_rate": 4.499932510538718e-05,
|
| 992 |
+
"loss": 0.2754,
|
| 993 |
+
"step": 820
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.69,
|
| 997 |
+
"learning_rate": 4.493113565395152e-05,
|
| 998 |
+
"loss": 0.2817,
|
| 999 |
+
"step": 825
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.7,
|
| 1003 |
+
"learning_rate": 4.4862536834470776e-05,
|
| 1004 |
+
"loss": 0.2834,
|
| 1005 |
+
"step": 830
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.7,
|
| 1009 |
+
"learning_rate": 4.479353005590454e-05,
|
| 1010 |
+
"loss": 0.2695,
|
| 1011 |
+
"step": 835
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.7,
|
| 1015 |
+
"learning_rate": 4.472411673559152e-05,
|
| 1016 |
+
"loss": 0.2659,
|
| 1017 |
+
"step": 840
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.71,
|
| 1021 |
+
"learning_rate": 4.465429829922043e-05,
|
| 1022 |
+
"loss": 0.2658,
|
| 1023 |
+
"step": 845
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.71,
|
| 1027 |
+
"learning_rate": 4.458407618080073e-05,
|
| 1028 |
+
"loss": 0.2736,
|
| 1029 |
+
"step": 850
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.72,
|
| 1033 |
+
"learning_rate": 4.4513451822633126e-05,
|
| 1034 |
+
"loss": 0.2705,
|
| 1035 |
+
"step": 855
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.72,
|
| 1039 |
+
"learning_rate": 4.444242667527999e-05,
|
| 1040 |
+
"loss": 0.2766,
|
| 1041 |
+
"step": 860
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.73,
|
| 1045 |
+
"learning_rate": 4.437100219753557e-05,
|
| 1046 |
+
"loss": 0.2816,
|
| 1047 |
+
"step": 865
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.73,
|
| 1051 |
+
"learning_rate": 4.429917985639598e-05,
|
| 1052 |
+
"loss": 0.2815,
|
| 1053 |
+
"step": 870
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.73,
|
| 1057 |
+
"learning_rate": 4.4226961127029126e-05,
|
| 1058 |
+
"loss": 0.265,
|
| 1059 |
+
"step": 875
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.74,
|
| 1063 |
+
"learning_rate": 4.4154347492744355e-05,
|
| 1064 |
+
"loss": 0.2723,
|
| 1065 |
+
"step": 880
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.74,
|
| 1069 |
+
"learning_rate": 4.408134044496203e-05,
|
| 1070 |
+
"loss": 0.2589,
|
| 1071 |
+
"step": 885
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.75,
|
| 1075 |
+
"learning_rate": 4.400794148318288e-05,
|
| 1076 |
+
"loss": 0.2714,
|
| 1077 |
+
"step": 890
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.75,
|
| 1081 |
+
"learning_rate": 4.39341521149572e-05,
|
| 1082 |
+
"loss": 0.2532,
|
| 1083 |
+
"step": 895
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.75,
|
| 1087 |
+
"learning_rate": 4.38599738558539e-05,
|
| 1088 |
+
"loss": 0.2723,
|
| 1089 |
+
"step": 900
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.76,
|
| 1093 |
+
"learning_rate": 4.378540822942936e-05,
|
| 1094 |
+
"loss": 0.2679,
|
| 1095 |
+
"step": 905
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.76,
|
| 1099 |
+
"learning_rate": 4.371045676719614e-05,
|
| 1100 |
+
"loss": 0.269,
|
| 1101 |
+
"step": 910
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.77,
|
| 1105 |
+
"learning_rate": 4.3635121008591546e-05,
|
| 1106 |
+
"loss": 0.2576,
|
| 1107 |
+
"step": 915
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.77,
|
| 1111 |
+
"learning_rate": 4.355940250094598e-05,
|
| 1112 |
+
"loss": 0.2618,
|
| 1113 |
+
"step": 920
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.78,
|
| 1117 |
+
"learning_rate": 4.348330279945117e-05,
|
| 1118 |
+
"loss": 0.2667,
|
| 1119 |
+
"step": 925
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.78,
|
| 1123 |
+
"learning_rate": 4.340682346712825e-05,
|
| 1124 |
+
"loss": 0.2715,
|
| 1125 |
+
"step": 930
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.78,
|
| 1129 |
+
"learning_rate": 4.3329966074795625e-05,
|
| 1130 |
+
"loss": 0.2541,
|
| 1131 |
+
"step": 935
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.79,
|
| 1135 |
+
"learning_rate": 4.325273220103672e-05,
|
| 1136 |
+
"loss": 0.2718,
|
| 1137 |
+
"step": 940
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.79,
|
| 1141 |
+
"learning_rate": 4.3175123432167555e-05,
|
| 1142 |
+
"loss": 0.2625,
|
| 1143 |
+
"step": 945
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.8,
|
| 1147 |
+
"learning_rate": 4.309714136220419e-05,
|
| 1148 |
+
"loss": 0.2614,
|
| 1149 |
+
"step": 950
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.8,
|
| 1153 |
+
"learning_rate": 4.301878759282993e-05,
|
| 1154 |
+
"loss": 0.2703,
|
| 1155 |
+
"step": 955
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 0.81,
|
| 1159 |
+
"learning_rate": 4.294006373336248e-05,
|
| 1160 |
+
"loss": 0.2588,
|
| 1161 |
+
"step": 960
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.81,
|
| 1165 |
+
"learning_rate": 4.2860971400720875e-05,
|
| 1166 |
+
"loss": 0.2642,
|
| 1167 |
+
"step": 965
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.81,
|
| 1171 |
+
"learning_rate": 4.2781512219392236e-05,
|
| 1172 |
+
"loss": 0.253,
|
| 1173 |
+
"step": 970
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.82,
|
| 1177 |
+
"learning_rate": 4.270168782139849e-05,
|
| 1178 |
+
"loss": 0.2568,
|
| 1179 |
+
"step": 975
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.82,
|
| 1183 |
+
"learning_rate": 4.262149984626271e-05,
|
| 1184 |
+
"loss": 0.2511,
|
| 1185 |
+
"step": 980
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.83,
|
| 1189 |
+
"learning_rate": 4.254094994097563e-05,
|
| 1190 |
+
"loss": 0.263,
|
| 1191 |
+
"step": 985
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 0.83,
|
| 1195 |
+
"learning_rate": 4.24600397599616e-05,
|
| 1196 |
+
"loss": 0.2707,
|
| 1197 |
+
"step": 990
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 0.83,
|
| 1201 |
+
"learning_rate": 4.2378770965044816e-05,
|
| 1202 |
+
"loss": 0.2606,
|
| 1203 |
+
"step": 995
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.84,
|
| 1207 |
+
"learning_rate": 4.229714522541504e-05,
|
| 1208 |
+
"loss": 0.2535,
|
| 1209 |
+
"step": 1000
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.84,
|
| 1213 |
+
"learning_rate": 4.221516421759334e-05,
|
| 1214 |
+
"loss": 0.2584,
|
| 1215 |
+
"step": 1005
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.85,
|
| 1219 |
+
"learning_rate": 4.213282962539774e-05,
|
| 1220 |
+
"loss": 0.2651,
|
| 1221 |
+
"step": 1010
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.85,
|
| 1225 |
+
"learning_rate": 4.2050143139908555e-05,
|
| 1226 |
+
"loss": 0.259,
|
| 1227 |
+
"step": 1015
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.86,
|
| 1231 |
+
"learning_rate": 4.1967106459433655e-05,
|
| 1232 |
+
"loss": 0.2498,
|
| 1233 |
+
"step": 1020
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 0.86,
|
| 1237 |
+
"learning_rate": 4.188372128947363e-05,
|
| 1238 |
+
"loss": 0.2667,
|
| 1239 |
+
"step": 1025
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.86,
|
| 1243 |
+
"learning_rate": 4.1799989342686745e-05,
|
| 1244 |
+
"loss": 0.2547,
|
| 1245 |
+
"step": 1030
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.87,
|
| 1249 |
+
"learning_rate": 4.1715912338853746e-05,
|
| 1250 |
+
"loss": 0.2589,
|
| 1251 |
+
"step": 1035
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.87,
|
| 1255 |
+
"learning_rate": 4.163149200484254e-05,
|
| 1256 |
+
"loss": 0.2551,
|
| 1257 |
+
"step": 1040
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.88,
|
| 1261 |
+
"learning_rate": 4.154673007457275e-05,
|
| 1262 |
+
"loss": 0.2551,
|
| 1263 |
+
"step": 1045
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.88,
|
| 1267 |
+
"learning_rate": 4.146162828898008e-05,
|
| 1268 |
+
"loss": 0.2585,
|
| 1269 |
+
"step": 1050
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.88,
|
| 1273 |
+
"learning_rate": 4.137618839598057e-05,
|
| 1274 |
+
"loss": 0.2555,
|
| 1275 |
+
"step": 1055
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 0.89,
|
| 1279 |
+
"learning_rate": 4.129041215043467e-05,
|
| 1280 |
+
"loss": 0.2409,
|
| 1281 |
+
"step": 1060
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 0.89,
|
| 1285 |
+
"learning_rate": 4.120430131411124e-05,
|
| 1286 |
+
"loss": 0.2469,
|
| 1287 |
+
"step": 1065
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.9,
|
| 1291 |
+
"learning_rate": 4.1117857655651346e-05,
|
| 1292 |
+
"loss": 0.2563,
|
| 1293 |
+
"step": 1070
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.9,
|
| 1297 |
+
"learning_rate": 4.103108295053189e-05,
|
| 1298 |
+
"loss": 0.2419,
|
| 1299 |
+
"step": 1075
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.91,
|
| 1303 |
+
"learning_rate": 4.094397898102921e-05,
|
| 1304 |
+
"loss": 0.2514,
|
| 1305 |
+
"step": 1080
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.91,
|
| 1309 |
+
"learning_rate": 4.085654753618245e-05,
|
| 1310 |
+
"loss": 0.2504,
|
| 1311 |
+
"step": 1085
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.91,
|
| 1315 |
+
"learning_rate": 4.076879041175682e-05,
|
| 1316 |
+
"loss": 0.2555,
|
| 1317 |
+
"step": 1090
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.92,
|
| 1321 |
+
"learning_rate": 4.068070941020665e-05,
|
| 1322 |
+
"loss": 0.2542,
|
| 1323 |
+
"step": 1095
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 0.92,
|
| 1327 |
+
"learning_rate": 4.059230634063848e-05,
|
| 1328 |
+
"loss": 0.2432,
|
| 1329 |
+
"step": 1100
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.93,
|
| 1333 |
+
"learning_rate": 4.0503583018773804e-05,
|
| 1334 |
+
"loss": 0.2513,
|
| 1335 |
+
"step": 1105
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.93,
|
| 1339 |
+
"learning_rate": 4.041454126691186e-05,
|
| 1340 |
+
"loss": 0.2379,
|
| 1341 |
+
"step": 1110
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.94,
|
| 1345 |
+
"learning_rate": 4.0325182913892126e-05,
|
| 1346 |
+
"loss": 0.2603,
|
| 1347 |
+
"step": 1115
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.94,
|
| 1351 |
+
"learning_rate": 4.023550979505678e-05,
|
| 1352 |
+
"loss": 0.2459,
|
| 1353 |
+
"step": 1120
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.94,
|
| 1357 |
+
"learning_rate": 4.014552375221306e-05,
|
| 1358 |
+
"loss": 0.2349,
|
| 1359 |
+
"step": 1125
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.95,
|
| 1363 |
+
"learning_rate": 4.0055226633595336e-05,
|
| 1364 |
+
"loss": 0.2575,
|
| 1365 |
+
"step": 1130
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 0.95,
|
| 1369 |
+
"learning_rate": 3.996462029382725e-05,
|
| 1370 |
+
"loss": 0.2424,
|
| 1371 |
+
"step": 1135
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.96,
|
| 1375 |
+
"learning_rate": 3.987370659388353e-05,
|
| 1376 |
+
"loss": 0.2495,
|
| 1377 |
+
"step": 1140
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.96,
|
| 1381 |
+
"learning_rate": 3.9782487401051844e-05,
|
| 1382 |
+
"loss": 0.2293,
|
| 1383 |
+
"step": 1145
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.96,
|
| 1387 |
+
"learning_rate": 3.96909645888944e-05,
|
| 1388 |
+
"loss": 0.2377,
|
| 1389 |
+
"step": 1150
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.97,
|
| 1393 |
+
"learning_rate": 3.95991400372095e-05,
|
| 1394 |
+
"loss": 0.2331,
|
| 1395 |
+
"step": 1155
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.97,
|
| 1399 |
+
"learning_rate": 3.950701563199289e-05,
|
| 1400 |
+
"loss": 0.2422,
|
| 1401 |
+
"step": 1160
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.98,
|
| 1405 |
+
"learning_rate": 3.9414593265399046e-05,
|
| 1406 |
+
"loss": 0.2422,
|
| 1407 |
+
"step": 1165
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 0.98,
|
| 1411 |
+
"learning_rate": 3.932187483570233e-05,
|
| 1412 |
+
"loss": 0.2425,
|
| 1413 |
+
"step": 1170
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.99,
|
| 1417 |
+
"learning_rate": 3.922886224725796e-05,
|
| 1418 |
+
"loss": 0.244,
|
| 1419 |
+
"step": 1175
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.99,
|
| 1423 |
+
"learning_rate": 3.913555741046291e-05,
|
| 1424 |
+
"loss": 0.2462,
|
| 1425 |
+
"step": 1180
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.99,
|
| 1429 |
+
"learning_rate": 3.904196224171671e-05,
|
| 1430 |
+
"loss": 0.2382,
|
| 1431 |
+
"step": 1185
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.0,
|
| 1435 |
+
"learning_rate": 3.894807866338203e-05,
|
| 1436 |
+
"loss": 0.2341,
|
| 1437 |
+
"step": 1190
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.0,
|
| 1441 |
+
"learning_rate": 3.8853908603745216e-05,
|
| 1442 |
+
"loss": 0.2041,
|
| 1443 |
+
"step": 1195
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 1.01,
|
| 1447 |
+
"learning_rate": 3.875945399697671e-05,
|
| 1448 |
+
"loss": 0.1927,
|
| 1449 |
+
"step": 1200
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 1.01,
|
| 1453 |
+
"learning_rate": 3.8664716783091284e-05,
|
| 1454 |
+
"loss": 0.1869,
|
| 1455 |
+
"step": 1205
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.01,
|
| 1459 |
+
"learning_rate": 3.856969890790823e-05,
|
| 1460 |
+
"loss": 0.1833,
|
| 1461 |
+
"step": 1210
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 1.02,
|
| 1465 |
+
"learning_rate": 3.8474402323011336e-05,
|
| 1466 |
+
"loss": 0.1889,
|
| 1467 |
+
"step": 1215
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.02,
|
| 1471 |
+
"learning_rate": 3.8378828985708906e-05,
|
| 1472 |
+
"loss": 0.1748,
|
| 1473 |
+
"step": 1220
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.03,
|
| 1477 |
+
"learning_rate": 3.828298085899346e-05,
|
| 1478 |
+
"loss": 0.1777,
|
| 1479 |
+
"step": 1225
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.03,
|
| 1483 |
+
"learning_rate": 3.8186859911501424e-05,
|
| 1484 |
+
"loss": 0.1838,
|
| 1485 |
+
"step": 1230
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 1.04,
|
| 1489 |
+
"learning_rate": 3.809046811747278e-05,
|
| 1490 |
+
"loss": 0.1909,
|
| 1491 |
+
"step": 1235
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 1.04,
|
| 1495 |
+
"learning_rate": 3.7993807456710436e-05,
|
| 1496 |
+
"loss": 0.1898,
|
| 1497 |
+
"step": 1240
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.04,
|
| 1501 |
+
"learning_rate": 3.789687991453957e-05,
|
| 1502 |
+
"loss": 0.1817,
|
| 1503 |
+
"step": 1245
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.05,
|
| 1507 |
+
"learning_rate": 3.779968748176689e-05,
|
| 1508 |
+
"loss": 0.1918,
|
| 1509 |
+
"step": 1250
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.05,
|
| 1513 |
+
"learning_rate": 3.7702232154639716e-05,
|
| 1514 |
+
"loss": 0.1769,
|
| 1515 |
+
"step": 1255
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.06,
|
| 1519 |
+
"learning_rate": 3.7604515934805e-05,
|
| 1520 |
+
"loss": 0.1843,
|
| 1521 |
+
"step": 1260
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.06,
|
| 1525 |
+
"learning_rate": 3.750654082926818e-05,
|
| 1526 |
+
"loss": 0.1838,
|
| 1527 |
+
"step": 1265
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 1.07,
|
| 1531 |
+
"learning_rate": 3.7408308850351996e-05,
|
| 1532 |
+
"loss": 0.1838,
|
| 1533 |
+
"step": 1270
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1.07,
|
| 1537 |
+
"learning_rate": 3.7309822015655154e-05,
|
| 1538 |
+
"loss": 0.1814,
|
| 1539 |
+
"step": 1275
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 1.07,
|
| 1543 |
+
"learning_rate": 3.7211082348010836e-05,
|
| 1544 |
+
"loss": 0.1816,
|
| 1545 |
+
"step": 1280
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 1.08,
|
| 1549 |
+
"learning_rate": 3.711209187544521e-05,
|
| 1550 |
+
"loss": 0.1841,
|
| 1551 |
+
"step": 1285
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.08,
|
| 1555 |
+
"learning_rate": 3.7012852631135784e-05,
|
| 1556 |
+
"loss": 0.1821,
|
| 1557 |
+
"step": 1290
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.09,
|
| 1561 |
+
"learning_rate": 3.6913366653369575e-05,
|
| 1562 |
+
"loss": 0.1886,
|
| 1563 |
+
"step": 1295
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.09,
|
| 1567 |
+
"learning_rate": 3.681363598550131e-05,
|
| 1568 |
+
"loss": 0.1857,
|
| 1569 |
+
"step": 1300
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 1.09,
|
| 1573 |
+
"learning_rate": 3.6713662675911445e-05,
|
| 1574 |
+
"loss": 0.1808,
|
| 1575 |
+
"step": 1305
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 1.1,
|
| 1579 |
+
"learning_rate": 3.661344877796407e-05,
|
| 1580 |
+
"loss": 0.1799,
|
| 1581 |
+
"step": 1310
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.1,
|
| 1585 |
+
"learning_rate": 3.6512996349964744e-05,
|
| 1586 |
+
"loss": 0.186,
|
| 1587 |
+
"step": 1315
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.11,
|
| 1591 |
+
"learning_rate": 3.6412307455118246e-05,
|
| 1592 |
+
"loss": 0.187,
|
| 1593 |
+
"step": 1320
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.11,
|
| 1597 |
+
"learning_rate": 3.631138416148615e-05,
|
| 1598 |
+
"loss": 0.1856,
|
| 1599 |
+
"step": 1325
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.12,
|
| 1603 |
+
"learning_rate": 3.62102285419444e-05,
|
| 1604 |
+
"loss": 0.1843,
|
| 1605 |
+
"step": 1330
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.12,
|
| 1609 |
+
"learning_rate": 3.6108842674140706e-05,
|
| 1610 |
+
"loss": 0.1709,
|
| 1611 |
+
"step": 1335
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.12,
|
| 1615 |
+
"learning_rate": 3.600722864045185e-05,
|
| 1616 |
+
"loss": 0.1773,
|
| 1617 |
+
"step": 1340
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 1.13,
|
| 1621 |
+
"learning_rate": 3.5905388527940976e-05,
|
| 1622 |
+
"loss": 0.1917,
|
| 1623 |
+
"step": 1345
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.13,
|
| 1627 |
+
"learning_rate": 3.580332442831466e-05,
|
| 1628 |
+
"loss": 0.1775,
|
| 1629 |
+
"step": 1350
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.14,
|
| 1633 |
+
"learning_rate": 3.570103843788e-05,
|
| 1634 |
+
"loss": 0.1812,
|
| 1635 |
+
"step": 1355
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.14,
|
| 1639 |
+
"learning_rate": 3.559853265750152e-05,
|
| 1640 |
+
"loss": 0.1812,
|
| 1641 |
+
"step": 1360
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.14,
|
| 1645 |
+
"learning_rate": 3.549580919255805e-05,
|
| 1646 |
+
"loss": 0.1827,
|
| 1647 |
+
"step": 1365
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.15,
|
| 1651 |
+
"learning_rate": 3.539287015289945e-05,
|
| 1652 |
+
"loss": 0.1889,
|
| 1653 |
+
"step": 1370
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 1.15,
|
| 1657 |
+
"learning_rate": 3.528971765280332e-05,
|
| 1658 |
+
"loss": 0.1733,
|
| 1659 |
+
"step": 1375
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.16,
|
| 1663 |
+
"learning_rate": 3.518635381093154e-05,
|
| 1664 |
+
"loss": 0.1835,
|
| 1665 |
+
"step": 1380
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.16,
|
| 1669 |
+
"learning_rate": 3.508278075028675e-05,
|
| 1670 |
+
"loss": 0.1787,
|
| 1671 |
+
"step": 1385
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.17,
|
| 1675 |
+
"learning_rate": 3.497900059816879e-05,
|
| 1676 |
+
"loss": 0.1797,
|
| 1677 |
+
"step": 1390
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 1.17,
|
| 1681 |
+
"learning_rate": 3.4875015486130947e-05,
|
| 1682 |
+
"loss": 0.1791,
|
| 1683 |
+
"step": 1395
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.17,
|
| 1687 |
+
"learning_rate": 3.477082754993624e-05,
|
| 1688 |
+
"loss": 0.1722,
|
| 1689 |
+
"step": 1400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.18,
|
| 1693 |
+
"learning_rate": 3.466643892951346e-05,
|
| 1694 |
+
"loss": 0.1807,
|
| 1695 |
+
"step": 1405
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 1.18,
|
| 1699 |
+
"learning_rate": 3.4561851768913346e-05,
|
| 1700 |
+
"loss": 0.1817,
|
| 1701 |
+
"step": 1410
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 1.19,
|
| 1705 |
+
"learning_rate": 3.445706821626443e-05,
|
| 1706 |
+
"loss": 0.1789,
|
| 1707 |
+
"step": 1415
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.19,
|
| 1711 |
+
"learning_rate": 3.435209042372898e-05,
|
| 1712 |
+
"loss": 0.1935,
|
| 1713 |
+
"step": 1420
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 1.2,
|
| 1717 |
+
"learning_rate": 3.4246920547458794e-05,
|
| 1718 |
+
"loss": 0.1802,
|
| 1719 |
+
"step": 1425
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 1.2,
|
| 1723 |
+
"learning_rate": 3.4141560747550875e-05,
|
| 1724 |
+
"loss": 0.1834,
|
| 1725 |
+
"step": 1430
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.2,
|
| 1729 |
+
"learning_rate": 3.40360131880031e-05,
|
| 1730 |
+
"loss": 0.1833,
|
| 1731 |
+
"step": 1435
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.21,
|
| 1735 |
+
"learning_rate": 3.393028003666978e-05,
|
| 1736 |
+
"loss": 0.1764,
|
| 1737 |
+
"step": 1440
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 1.21,
|
| 1741 |
+
"learning_rate": 3.3824363465217105e-05,
|
| 1742 |
+
"loss": 0.1763,
|
| 1743 |
+
"step": 1445
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 1.22,
|
| 1747 |
+
"learning_rate": 3.3718265649078545e-05,
|
| 1748 |
+
"loss": 0.1679,
|
| 1749 |
+
"step": 1450
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.22,
|
| 1753 |
+
"learning_rate": 3.361198876741019e-05,
|
| 1754 |
+
"loss": 0.1739,
|
| 1755 |
+
"step": 1455
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 1.22,
|
| 1759 |
+
"learning_rate": 3.3505535003045994e-05,
|
| 1760 |
+
"loss": 0.1798,
|
| 1761 |
+
"step": 1460
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 1.23,
|
| 1765 |
+
"learning_rate": 3.339890654245288e-05,
|
| 1766 |
+
"loss": 0.185,
|
| 1767 |
+
"step": 1465
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.23,
|
| 1771 |
+
"learning_rate": 3.329210557568593e-05,
|
| 1772 |
+
"loss": 0.1712,
|
| 1773 |
+
"step": 1470
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.24,
|
| 1777 |
+
"learning_rate": 3.3185134296343314e-05,
|
| 1778 |
+
"loss": 0.1705,
|
| 1779 |
+
"step": 1475
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 1.24,
|
| 1783 |
+
"learning_rate": 3.3077994901521284e-05,
|
| 1784 |
+
"loss": 0.1781,
|
| 1785 |
+
"step": 1480
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 1.25,
|
| 1789 |
+
"learning_rate": 3.297068959176906e-05,
|
| 1790 |
+
"loss": 0.17,
|
| 1791 |
+
"step": 1485
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 1.25,
|
| 1795 |
+
"learning_rate": 3.286322057104356e-05,
|
| 1796 |
+
"loss": 0.1769,
|
| 1797 |
+
"step": 1490
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 1.25,
|
| 1801 |
+
"learning_rate": 3.275559004666424e-05,
|
| 1802 |
+
"loss": 0.1757,
|
| 1803 |
+
"step": 1495
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 1.26,
|
| 1807 |
+
"learning_rate": 3.264780022926765e-05,
|
| 1808 |
+
"loss": 0.1785,
|
| 1809 |
+
"step": 1500
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.26,
|
| 1813 |
+
"learning_rate": 3.2539853332762106e-05,
|
| 1814 |
+
"loss": 0.1708,
|
| 1815 |
+
"step": 1505
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.27,
|
| 1819 |
+
"learning_rate": 3.2431751574282175e-05,
|
| 1820 |
+
"loss": 0.1803,
|
| 1821 |
+
"step": 1510
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 1.27,
|
| 1825 |
+
"learning_rate": 3.232349717414318e-05,
|
| 1826 |
+
"loss": 0.173,
|
| 1827 |
+
"step": 1515
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.27,
|
| 1831 |
+
"learning_rate": 3.221509235579554e-05,
|
| 1832 |
+
"loss": 0.1838,
|
| 1833 |
+
"step": 1520
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.28,
|
| 1837 |
+
"learning_rate": 3.210653934577914e-05,
|
| 1838 |
+
"loss": 0.167,
|
| 1839 |
+
"step": 1525
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.28,
|
| 1843 |
+
"learning_rate": 3.19978403736776e-05,
|
| 1844 |
+
"loss": 0.1777,
|
| 1845 |
+
"step": 1530
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.29,
|
| 1849 |
+
"learning_rate": 3.1888997672072476e-05,
|
| 1850 |
+
"loss": 0.1822,
|
| 1851 |
+
"step": 1535
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.29,
|
| 1855 |
+
"learning_rate": 3.178001347649738e-05,
|
| 1856 |
+
"loss": 0.172,
|
| 1857 |
+
"step": 1540
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.3,
|
| 1861 |
+
"learning_rate": 3.167089002539212e-05,
|
| 1862 |
+
"loss": 0.1813,
|
| 1863 |
+
"step": 1545
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 1.3,
|
| 1867 |
+
"learning_rate": 3.156162956005666e-05,
|
| 1868 |
+
"loss": 0.1768,
|
| 1869 |
+
"step": 1550
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 1.3,
|
| 1873 |
+
"learning_rate": 3.1452234324605126e-05,
|
| 1874 |
+
"loss": 0.1788,
|
| 1875 |
+
"step": 1555
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 1.31,
|
| 1879 |
+
"learning_rate": 3.13427065659197e-05,
|
| 1880 |
+
"loss": 0.1749,
|
| 1881 |
+
"step": 1560
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 1.31,
|
| 1885 |
+
"learning_rate": 3.1233048533604494e-05,
|
| 1886 |
+
"loss": 0.1736,
|
| 1887 |
+
"step": 1565
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 1.32,
|
| 1891 |
+
"learning_rate": 3.112326247993931e-05,
|
| 1892 |
+
"loss": 0.1856,
|
| 1893 |
+
"step": 1570
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.32,
|
| 1897 |
+
"learning_rate": 3.1013350659833394e-05,
|
| 1898 |
+
"loss": 0.1844,
|
| 1899 |
+
"step": 1575
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.33,
|
| 1903 |
+
"learning_rate": 3.090331533077913e-05,
|
| 1904 |
+
"loss": 0.176,
|
| 1905 |
+
"step": 1580
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.33,
|
| 1909 |
+
"learning_rate": 3.0793158752805675e-05,
|
| 1910 |
+
"loss": 0.1662,
|
| 1911 |
+
"step": 1585
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 1.33,
|
| 1915 |
+
"learning_rate": 3.0682883188432505e-05,
|
| 1916 |
+
"loss": 0.1712,
|
| 1917 |
+
"step": 1590
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 1.34,
|
| 1921 |
+
"learning_rate": 3.0572490902623006e-05,
|
| 1922 |
+
"loss": 0.1682,
|
| 1923 |
+
"step": 1595
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.34,
|
| 1927 |
+
"learning_rate": 3.0461984162737906e-05,
|
| 1928 |
+
"loss": 0.1702,
|
| 1929 |
+
"step": 1600
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 1.35,
|
| 1933 |
+
"learning_rate": 3.0351365238488723e-05,
|
| 1934 |
+
"loss": 0.1678,
|
| 1935 |
+
"step": 1605
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.35,
|
| 1939 |
+
"learning_rate": 3.0240636401891124e-05,
|
| 1940 |
+
"loss": 0.1782,
|
| 1941 |
+
"step": 1610
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.35,
|
| 1945 |
+
"learning_rate": 3.01297999272183e-05,
|
| 1946 |
+
"loss": 0.1753,
|
| 1947 |
+
"step": 1615
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.36,
|
| 1951 |
+
"learning_rate": 3.001885809095424e-05,
|
| 1952 |
+
"loss": 0.1862,
|
| 1953 |
+
"step": 1620
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 1.36,
|
| 1957 |
+
"learning_rate": 2.9907813171746935e-05,
|
| 1958 |
+
"loss": 0.1756,
|
| 1959 |
+
"step": 1625
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 1.37,
|
| 1963 |
+
"learning_rate": 2.9796667450361636e-05,
|
| 1964 |
+
"loss": 0.1762,
|
| 1965 |
+
"step": 1630
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 1.37,
|
| 1969 |
+
"learning_rate": 2.9685423209633995e-05,
|
| 1970 |
+
"loss": 0.1731,
|
| 1971 |
+
"step": 1635
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 1.38,
|
| 1975 |
+
"learning_rate": 2.957408273442313e-05,
|
| 1976 |
+
"loss": 0.1714,
|
| 1977 |
+
"step": 1640
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.38,
|
| 1981 |
+
"learning_rate": 2.9462648311564745e-05,
|
| 1982 |
+
"loss": 0.1809,
|
| 1983 |
+
"step": 1645
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.38,
|
| 1987 |
+
"learning_rate": 2.9351122229824146e-05,
|
| 1988 |
+
"loss": 0.177,
|
| 1989 |
+
"step": 1650
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 1.39,
|
| 1993 |
+
"learning_rate": 2.923950677984925e-05,
|
| 1994 |
+
"loss": 0.1771,
|
| 1995 |
+
"step": 1655
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 1.39,
|
| 1999 |
+
"learning_rate": 2.912780425412349e-05,
|
| 2000 |
+
"loss": 0.1748,
|
| 2001 |
+
"step": 1660
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 1.4,
|
| 2005 |
+
"learning_rate": 2.9016016946918777e-05,
|
| 2006 |
+
"loss": 0.1812,
|
| 2007 |
+
"step": 1665
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 1.4,
|
| 2011 |
+
"learning_rate": 2.890414715424836e-05,
|
| 2012 |
+
"loss": 0.1765,
|
| 2013 |
+
"step": 1670
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 1.4,
|
| 2017 |
+
"learning_rate": 2.8792197173819645e-05,
|
| 2018 |
+
"loss": 0.1751,
|
| 2019 |
+
"step": 1675
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.41,
|
| 2023 |
+
"learning_rate": 2.8680169304987052e-05,
|
| 2024 |
+
"loss": 0.1863,
|
| 2025 |
+
"step": 1680
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.41,
|
| 2029 |
+
"learning_rate": 2.856806584870475e-05,
|
| 2030 |
+
"loss": 0.1822,
|
| 2031 |
+
"step": 1685
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 1.42,
|
| 2035 |
+
"learning_rate": 2.8455889107479404e-05,
|
| 2036 |
+
"loss": 0.1707,
|
| 2037 |
+
"step": 1690
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 1.42,
|
| 2041 |
+
"learning_rate": 2.8343641385322894e-05,
|
| 2042 |
+
"loss": 0.1768,
|
| 2043 |
+
"step": 1695
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 1.43,
|
| 2047 |
+
"learning_rate": 2.8231324987704978e-05,
|
| 2048 |
+
"loss": 0.1767,
|
| 2049 |
+
"step": 1700
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 1.43,
|
| 2053 |
+
"learning_rate": 2.8118942221505957e-05,
|
| 2054 |
+
"loss": 0.179,
|
| 2055 |
+
"step": 1705
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 1.43,
|
| 2059 |
+
"learning_rate": 2.8006495394969268e-05,
|
| 2060 |
+
"loss": 0.1721,
|
| 2061 |
+
"step": 1710
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.44,
|
| 2065 |
+
"learning_rate": 2.7893986817654117e-05,
|
| 2066 |
+
"loss": 0.1645,
|
| 2067 |
+
"step": 1715
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.44,
|
| 2071 |
+
"learning_rate": 2.7781418800387992e-05,
|
| 2072 |
+
"loss": 0.1763,
|
| 2073 |
+
"step": 1720
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 1.45,
|
| 2077 |
+
"learning_rate": 2.766879365521925e-05,
|
| 2078 |
+
"loss": 0.1783,
|
| 2079 |
+
"step": 1725
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.45,
|
| 2083 |
+
"learning_rate": 2.7556113695369585e-05,
|
| 2084 |
+
"loss": 0.1731,
|
| 2085 |
+
"step": 1730
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 1.46,
|
| 2089 |
+
"learning_rate": 2.744338123518653e-05,
|
| 2090 |
+
"loss": 0.1638,
|
| 2091 |
+
"step": 1735
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 1.46,
|
| 2095 |
+
"learning_rate": 2.733059859009596e-05,
|
| 2096 |
+
"loss": 0.172,
|
| 2097 |
+
"step": 1740
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 1.46,
|
| 2101 |
+
"learning_rate": 2.7217768076554473e-05,
|
| 2102 |
+
"loss": 0.1775,
|
| 2103 |
+
"step": 1745
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.47,
|
| 2107 |
+
"learning_rate": 2.7104892012001854e-05,
|
| 2108 |
+
"loss": 0.1751,
|
| 2109 |
+
"step": 1750
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.47,
|
| 2113 |
+
"learning_rate": 2.6991972714813485e-05,
|
| 2114 |
+
"loss": 0.174,
|
| 2115 |
+
"step": 1755
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 1.48,
|
| 2119 |
+
"learning_rate": 2.6879012504252677e-05,
|
| 2120 |
+
"loss": 0.1674,
|
| 2121 |
+
"step": 1760
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 1.48,
|
| 2125 |
+
"learning_rate": 2.676601370042307e-05,
|
| 2126 |
+
"loss": 0.1748,
|
| 2127 |
+
"step": 1765
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 1.48,
|
| 2131 |
+
"learning_rate": 2.6652978624221002e-05,
|
| 2132 |
+
"loss": 0.1728,
|
| 2133 |
+
"step": 1770
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 1.49,
|
| 2137 |
+
"learning_rate": 2.6539909597287798e-05,
|
| 2138 |
+
"loss": 0.1736,
|
| 2139 |
+
"step": 1775
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 1.49,
|
| 2143 |
+
"learning_rate": 2.642680894196209e-05,
|
| 2144 |
+
"loss": 0.162,
|
| 2145 |
+
"step": 1780
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.5,
|
| 2149 |
+
"learning_rate": 2.6313678981232143e-05,
|
| 2150 |
+
"loss": 0.1701,
|
| 2151 |
+
"step": 1785
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.5,
|
| 2155 |
+
"learning_rate": 2.620052203868813e-05,
|
| 2156 |
+
"loss": 0.1813,
|
| 2157 |
+
"step": 1790
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 1.51,
|
| 2161 |
+
"learning_rate": 2.608734043847439e-05,
|
| 2162 |
+
"loss": 0.171,
|
| 2163 |
+
"step": 1795
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 1.51,
|
| 2167 |
+
"learning_rate": 2.5974136505241737e-05,
|
| 2168 |
+
"loss": 0.1654,
|
| 2169 |
+
"step": 1800
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 1.51,
|
| 2173 |
+
"learning_rate": 2.586091256409966e-05,
|
| 2174 |
+
"loss": 0.168,
|
| 2175 |
+
"step": 1805
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 1.52,
|
| 2179 |
+
"learning_rate": 2.574767094056862e-05,
|
| 2180 |
+
"loss": 0.1726,
|
| 2181 |
+
"step": 1810
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 1.52,
|
| 2185 |
+
"learning_rate": 2.5634413960532217e-05,
|
| 2186 |
+
"loss": 0.167,
|
| 2187 |
+
"step": 1815
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.53,
|
| 2191 |
+
"learning_rate": 2.5521143950189508e-05,
|
| 2192 |
+
"loss": 0.1712,
|
| 2193 |
+
"step": 1820
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.53,
|
| 2197 |
+
"learning_rate": 2.5407863236007147e-05,
|
| 2198 |
+
"loss": 0.1631,
|
| 2199 |
+
"step": 1825
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 1.53,
|
| 2203 |
+
"learning_rate": 2.5294574144671644e-05,
|
| 2204 |
+
"loss": 0.1762,
|
| 2205 |
+
"step": 1830
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 1.54,
|
| 2209 |
+
"learning_rate": 2.5181279003041575e-05,
|
| 2210 |
+
"loss": 0.1739,
|
| 2211 |
+
"step": 1835
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 1.54,
|
| 2215 |
+
"learning_rate": 2.5067980138099777e-05,
|
| 2216 |
+
"loss": 0.1724,
|
| 2217 |
+
"step": 1840
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 1.55,
|
| 2221 |
+
"learning_rate": 2.4954679876905576e-05,
|
| 2222 |
+
"loss": 0.1673,
|
| 2223 |
+
"step": 1845
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 1.55,
|
| 2227 |
+
"learning_rate": 2.4841380546546953e-05,
|
| 2228 |
+
"loss": 0.1685,
|
| 2229 |
+
"step": 1850
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 1.56,
|
| 2233 |
+
"learning_rate": 2.4728084474092775e-05,
|
| 2234 |
+
"loss": 0.1638,
|
| 2235 |
+
"step": 1855
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.56,
|
| 2239 |
+
"learning_rate": 2.461479398654501e-05,
|
| 2240 |
+
"loss": 0.1809,
|
| 2241 |
+
"step": 1860
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 1.56,
|
| 2245 |
+
"learning_rate": 2.45015114107909e-05,
|
| 2246 |
+
"loss": 0.1732,
|
| 2247 |
+
"step": 1865
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 1.57,
|
| 2251 |
+
"learning_rate": 2.43882390735552e-05,
|
| 2252 |
+
"loss": 0.1656,
|
| 2253 |
+
"step": 1870
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 1.57,
|
| 2257 |
+
"learning_rate": 2.427497930135237e-05,
|
| 2258 |
+
"loss": 0.1669,
|
| 2259 |
+
"step": 1875
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 1.58,
|
| 2263 |
+
"learning_rate": 2.4161734420438782e-05,
|
| 2264 |
+
"loss": 0.1638,
|
| 2265 |
+
"step": 1880
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 1.58,
|
| 2269 |
+
"learning_rate": 2.4048506756764973e-05,
|
| 2270 |
+
"loss": 0.165,
|
| 2271 |
+
"step": 1885
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 1.59,
|
| 2275 |
+
"learning_rate": 2.393529863592784e-05,
|
| 2276 |
+
"loss": 0.1758,
|
| 2277 |
+
"step": 1890
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.59,
|
| 2281 |
+
"learning_rate": 2.382211238312288e-05,
|
| 2282 |
+
"loss": 0.1663,
|
| 2283 |
+
"step": 1895
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 1.59,
|
| 2287 |
+
"learning_rate": 2.3708950323096482e-05,
|
| 2288 |
+
"loss": 0.1718,
|
| 2289 |
+
"step": 1900
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 1.6,
|
| 2293 |
+
"learning_rate": 2.359581478009809e-05,
|
| 2294 |
+
"loss": 0.163,
|
| 2295 |
+
"step": 1905
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 1.6,
|
| 2299 |
+
"learning_rate": 2.3482708077832522e-05,
|
| 2300 |
+
"loss": 0.1637,
|
| 2301 |
+
"step": 1910
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 1.61,
|
| 2305 |
+
"learning_rate": 2.336963253941225e-05,
|
| 2306 |
+
"loss": 0.1703,
|
| 2307 |
+
"step": 1915
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 1.61,
|
| 2311 |
+
"learning_rate": 2.325659048730965e-05,
|
| 2312 |
+
"loss": 0.167,
|
| 2313 |
+
"step": 1920
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 1.61,
|
| 2317 |
+
"learning_rate": 2.314358424330932e-05,
|
| 2318 |
+
"loss": 0.1603,
|
| 2319 |
+
"step": 1925
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.62,
|
| 2323 |
+
"learning_rate": 2.3030616128460407e-05,
|
| 2324 |
+
"loss": 0.166,
|
| 2325 |
+
"step": 1930
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 1.62,
|
| 2329 |
+
"learning_rate": 2.29176884630289e-05,
|
| 2330 |
+
"loss": 0.1633,
|
| 2331 |
+
"step": 1935
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 1.63,
|
| 2335 |
+
"learning_rate": 2.2804803566450008e-05,
|
| 2336 |
+
"loss": 0.1666,
|
| 2337 |
+
"step": 1940
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 1.63,
|
| 2341 |
+
"learning_rate": 2.2691963757280494e-05,
|
| 2342 |
+
"loss": 0.1712,
|
| 2343 |
+
"step": 1945
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 1.64,
|
| 2347 |
+
"learning_rate": 2.257917135315105e-05,
|
| 2348 |
+
"loss": 0.1633,
|
| 2349 |
+
"step": 1950
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 1.64,
|
| 2353 |
+
"learning_rate": 2.2466428670718766e-05,
|
| 2354 |
+
"loss": 0.159,
|
| 2355 |
+
"step": 1955
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 1.64,
|
| 2359 |
+
"learning_rate": 2.2353738025619426e-05,
|
| 2360 |
+
"loss": 0.1679,
|
| 2361 |
+
"step": 1960
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 1.65,
|
| 2365 |
+
"learning_rate": 2.2241101732420044e-05,
|
| 2366 |
+
"loss": 0.1726,
|
| 2367 |
+
"step": 1965
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 1.65,
|
| 2371 |
+
"learning_rate": 2.2128522104571303e-05,
|
| 2372 |
+
"loss": 0.1699,
|
| 2373 |
+
"step": 1970
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 1.66,
|
| 2377 |
+
"learning_rate": 2.201600145436001e-05,
|
| 2378 |
+
"loss": 0.1655,
|
| 2379 |
+
"step": 1975
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 1.66,
|
| 2383 |
+
"learning_rate": 2.1903542092861624e-05,
|
| 2384 |
+
"loss": 0.1685,
|
| 2385 |
+
"step": 1980
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 1.66,
|
| 2389 |
+
"learning_rate": 2.1791146329892805e-05,
|
| 2390 |
+
"loss": 0.1644,
|
| 2391 |
+
"step": 1985
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 1.67,
|
| 2395 |
+
"learning_rate": 2.1678816473963944e-05,
|
| 2396 |
+
"loss": 0.1774,
|
| 2397 |
+
"step": 1990
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 1.67,
|
| 2401 |
+
"learning_rate": 2.1566554832231746e-05,
|
| 2402 |
+
"loss": 0.1674,
|
| 2403 |
+
"step": 1995
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.68,
|
| 2407 |
+
"learning_rate": 2.1454363710451888e-05,
|
| 2408 |
+
"loss": 0.1729,
|
| 2409 |
+
"step": 2000
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 1.68,
|
| 2413 |
+
"learning_rate": 2.13422454129316e-05,
|
| 2414 |
+
"loss": 0.1687,
|
| 2415 |
+
"step": 2005
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 1.69,
|
| 2419 |
+
"learning_rate": 2.1230202242482362e-05,
|
| 2420 |
+
"loss": 0.1664,
|
| 2421 |
+
"step": 2010
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 1.69,
|
| 2425 |
+
"learning_rate": 2.1118236500372654e-05,
|
| 2426 |
+
"loss": 0.157,
|
| 2427 |
+
"step": 2015
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 1.69,
|
| 2431 |
+
"learning_rate": 2.100635048628059e-05,
|
| 2432 |
+
"loss": 0.1665,
|
| 2433 |
+
"step": 2020
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 1.7,
|
| 2437 |
+
"learning_rate": 2.0894546498246783e-05,
|
| 2438 |
+
"loss": 0.1721,
|
| 2439 |
+
"step": 2025
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 1.7,
|
| 2443 |
+
"learning_rate": 2.0782826832627083e-05,
|
| 2444 |
+
"loss": 0.1598,
|
| 2445 |
+
"step": 2030
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 1.71,
|
| 2449 |
+
"learning_rate": 2.0671193784045426e-05,
|
| 2450 |
+
"loss": 0.1729,
|
| 2451 |
+
"step": 2035
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 1.71,
|
| 2455 |
+
"learning_rate": 2.0559649645346728e-05,
|
| 2456 |
+
"loss": 0.1717,
|
| 2457 |
+
"step": 2040
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 1.72,
|
| 2461 |
+
"learning_rate": 2.044819670754976e-05,
|
| 2462 |
+
"loss": 0.1584,
|
| 2463 |
+
"step": 2045
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 1.72,
|
| 2467 |
+
"learning_rate": 2.03368372598001e-05,
|
| 2468 |
+
"loss": 0.1644,
|
| 2469 |
+
"step": 2050
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 1.72,
|
| 2473 |
+
"learning_rate": 2.0225573589323154e-05,
|
| 2474 |
+
"loss": 0.172,
|
| 2475 |
+
"step": 2055
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 1.73,
|
| 2479 |
+
"learning_rate": 2.0114407981377114e-05,
|
| 2480 |
+
"loss": 0.1697,
|
| 2481 |
+
"step": 2060
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 1.73,
|
| 2485 |
+
"learning_rate": 2.0003342719206067e-05,
|
| 2486 |
+
"loss": 0.1627,
|
| 2487 |
+
"step": 2065
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 1.74,
|
| 2491 |
+
"learning_rate": 1.989238008399309e-05,
|
| 2492 |
+
"loss": 0.1715,
|
| 2493 |
+
"step": 2070
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 1.74,
|
| 2497 |
+
"learning_rate": 1.978152235481338e-05,
|
| 2498 |
+
"loss": 0.1548,
|
| 2499 |
+
"step": 2075
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 1.74,
|
| 2503 |
+
"learning_rate": 1.9670771808587455e-05,
|
| 2504 |
+
"loss": 0.1638,
|
| 2505 |
+
"step": 2080
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 1.75,
|
| 2509 |
+
"learning_rate": 1.9560130720034422e-05,
|
| 2510 |
+
"loss": 0.1605,
|
| 2511 |
+
"step": 2085
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 1.75,
|
| 2515 |
+
"learning_rate": 1.9449601361625176e-05,
|
| 2516 |
+
"loss": 0.1678,
|
| 2517 |
+
"step": 2090
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 1.76,
|
| 2521 |
+
"learning_rate": 1.9339186003535808e-05,
|
| 2522 |
+
"loss": 0.1629,
|
| 2523 |
+
"step": 2095
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 1.76,
|
| 2527 |
+
"learning_rate": 1.9228886913600928e-05,
|
| 2528 |
+
"loss": 0.1594,
|
| 2529 |
+
"step": 2100
|
| 2530 |
+
}
|
| 2531 |
+
],
|
| 2532 |
+
"logging_steps": 5,
|
| 2533 |
+
"max_steps": 3576,
|
| 2534 |
+
"num_input_tokens_seen": 0,
|
| 2535 |
+
"num_train_epochs": 3,
|
| 2536 |
+
"save_steps": 100,
|
| 2537 |
+
"total_flos": 7.281171626033873e+18,
|
| 2538 |
+
"train_batch_size": 1,
|
| 2539 |
+
"trial_name": null,
|
| 2540 |
+
"trial_params": null
|
| 2541 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a4732aa590837b51f08705b71b31d3bc572baba911c6c2c8cf9340cbaf2268f
|
| 3 |
+
size 4920
|