checkpoint-800-uploaded-manually
Browse files- README.md +202 -1
- adapter_config.json +38 -0
- adapter_model.safetensors +3 -0
- bn_adapter_ultrachat_rummanvai/adapter_config.json +38 -0
- bn_adapter_ultrachat_rummanvai/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.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +41 -0
- trainer_state.json +2421 -0
- training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,204 @@
|
|
| 1 |
---
|
| 2 |
-
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
base_model: intelsense/mis2nd4th5thekushfinalyo1
|
| 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 0.8.2.dev0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": {
|
| 4 |
+
"base_model_class": "MistralForCausalLM",
|
| 5 |
+
"parent_library": "transformers.models.mistral.modeling_mistral"
|
| 6 |
+
},
|
| 7 |
+
"base_model_name_or_path": "intelsense/mis2nd4th5thekushfinalyo1",
|
| 8 |
+
"bias": "none",
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 32,
|
| 16 |
+
"lora_dropout": 0.05,
|
| 17 |
+
"megatron_config": null,
|
| 18 |
+
"megatron_core": "megatron.core",
|
| 19 |
+
"modules_to_save": [
|
| 20 |
+
"embed_tokens",
|
| 21 |
+
"lm_head"
|
| 22 |
+
],
|
| 23 |
+
"peft_type": "LORA",
|
| 24 |
+
"r": 16,
|
| 25 |
+
"rank_pattern": {},
|
| 26 |
+
"revision": null,
|
| 27 |
+
"target_modules": [
|
| 28 |
+
"v_proj",
|
| 29 |
+
"q_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"gate_proj",
|
| 32 |
+
"down_proj",
|
| 33 |
+
"o_proj",
|
| 34 |
+
"up_proj"
|
| 35 |
+
],
|
| 36 |
+
"task_type": null,
|
| 37 |
+
"use_rslora": false
|
| 38 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b9d3a3f4c10d601f28a854b60c25b2d340f2cc543f485dbee0ab20e71e69fb4
|
| 3 |
+
size 1195438880
|
bn_adapter_ultrachat_rummanvai/adapter_config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": {
|
| 4 |
+
"base_model_class": "MistralForCausalLM",
|
| 5 |
+
"parent_library": "transformers.models.mistral.modeling_mistral"
|
| 6 |
+
},
|
| 7 |
+
"base_model_name_or_path": "intelsense/mis2nd4th5thekushfinalyo1",
|
| 8 |
+
"bias": "none",
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 32,
|
| 16 |
+
"lora_dropout": 0.05,
|
| 17 |
+
"megatron_config": null,
|
| 18 |
+
"megatron_core": "megatron.core",
|
| 19 |
+
"modules_to_save": [
|
| 20 |
+
"embed_tokens",
|
| 21 |
+
"lm_head"
|
| 22 |
+
],
|
| 23 |
+
"peft_type": "LORA",
|
| 24 |
+
"r": 16,
|
| 25 |
+
"rank_pattern": {},
|
| 26 |
+
"revision": null,
|
| 27 |
+
"target_modules": [
|
| 28 |
+
"v_proj",
|
| 29 |
+
"q_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"gate_proj",
|
| 32 |
+
"down_proj",
|
| 33 |
+
"o_proj",
|
| 34 |
+
"up_proj"
|
| 35 |
+
],
|
| 36 |
+
"task_type": null,
|
| 37 |
+
"use_rslora": false
|
| 38 |
+
}
|
bn_adapter_ultrachat_rummanvai/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5be608fda6a03ff049708575bcbfc3cb86c299ab4eb6b10023092ce714ae7764
|
| 3 |
+
size 1195438880
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3129a05d62b29c8b7944e8934a70a61c40947a546db24291a6331fb79c3cad3f
|
| 3 |
+
size 1198237404
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7420ca43a47d47541ec6d0576311bd99ccc18c161c6a57cbb9d802785e1f073d
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99987b620a2745710faa7434b61852ba9418c9491edaf476c7112645db3b157c
|
| 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.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:075e1fa867e40e996ad6a23f60554d934079af5d66614905626f018756a8af5e
|
| 3 |
+
size 1557786
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": true,
|
| 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 |
+
"bos_token": "<s>",
|
| 31 |
+
"clean_up_tokenization_spaces": false,
|
| 32 |
+
"eos_token": "</s>",
|
| 33 |
+
"legacy": true,
|
| 34 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 35 |
+
"pad_token": "</s>",
|
| 36 |
+
"sp_model_kwargs": {},
|
| 37 |
+
"spaces_between_special_tokens": false,
|
| 38 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 39 |
+
"unk_token": "<unk>",
|
| 40 |
+
"use_default_system_prompt": false
|
| 41 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2421 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 29.62962962962963,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.07,
|
| 13 |
+
"learning_rate": 1.999970808109463e-05,
|
| 14 |
+
"loss": 5.1275,
|
| 15 |
+
"step": 2
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.15,
|
| 19 |
+
"learning_rate": 1.9998814492475354e-05,
|
| 20 |
+
"loss": 4.575,
|
| 21 |
+
"step": 4
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.22,
|
| 25 |
+
"learning_rate": 1.9997319220216543e-05,
|
| 26 |
+
"loss": 3.7996,
|
| 27 |
+
"step": 6
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.3,
|
| 31 |
+
"learning_rate": 1.999522235429695e-05,
|
| 32 |
+
"loss": 3.5835,
|
| 33 |
+
"step": 8
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.37,
|
| 37 |
+
"learning_rate": 1.9992524020896517e-05,
|
| 38 |
+
"loss": 3.185,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.44,
|
| 43 |
+
"learning_rate": 1.9989224382388813e-05,
|
| 44 |
+
"loss": 2.7431,
|
| 45 |
+
"step": 12
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.52,
|
| 49 |
+
"learning_rate": 1.9985323637331214e-05,
|
| 50 |
+
"loss": 2.6319,
|
| 51 |
+
"step": 14
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.59,
|
| 55 |
+
"learning_rate": 1.9980822020453007e-05,
|
| 56 |
+
"loss": 2.4273,
|
| 57 |
+
"step": 16
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.67,
|
| 61 |
+
"learning_rate": 1.9975719802641226e-05,
|
| 62 |
+
"loss": 2.3277,
|
| 63 |
+
"step": 18
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.74,
|
| 67 |
+
"learning_rate": 1.9970017290924365e-05,
|
| 68 |
+
"loss": 2.2614,
|
| 69 |
+
"step": 20
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.81,
|
| 73 |
+
"learning_rate": 1.99637148284539e-05,
|
| 74 |
+
"loss": 2.2018,
|
| 75 |
+
"step": 22
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.89,
|
| 79 |
+
"learning_rate": 1.9956812794483658e-05,
|
| 80 |
+
"loss": 2.0017,
|
| 81 |
+
"step": 24
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.96,
|
| 85 |
+
"learning_rate": 1.9949311604346965e-05,
|
| 86 |
+
"loss": 2.0208,
|
| 87 |
+
"step": 26
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.04,
|
| 91 |
+
"learning_rate": 1.994121170943167e-05,
|
| 92 |
+
"loss": 1.8773,
|
| 93 |
+
"step": 28
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.11,
|
| 97 |
+
"learning_rate": 1.993251359715298e-05,
|
| 98 |
+
"loss": 1.7932,
|
| 99 |
+
"step": 30
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.19,
|
| 103 |
+
"learning_rate": 1.992321779092414e-05,
|
| 104 |
+
"loss": 1.806,
|
| 105 |
+
"step": 32
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.26,
|
| 109 |
+
"learning_rate": 1.991332485012491e-05,
|
| 110 |
+
"loss": 1.731,
|
| 111 |
+
"step": 34
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.33,
|
| 115 |
+
"learning_rate": 1.9902835370067927e-05,
|
| 116 |
+
"loss": 1.7323,
|
| 117 |
+
"step": 36
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.41,
|
| 121 |
+
"learning_rate": 1.989174998196288e-05,
|
| 122 |
+
"loss": 1.7465,
|
| 123 |
+
"step": 38
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.48,
|
| 127 |
+
"learning_rate": 1.9880069352878508e-05,
|
| 128 |
+
"loss": 1.7926,
|
| 129 |
+
"step": 40
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.56,
|
| 133 |
+
"learning_rate": 1.986779418570249e-05,
|
| 134 |
+
"loss": 1.7291,
|
| 135 |
+
"step": 42
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.63,
|
| 139 |
+
"learning_rate": 1.9854925219099116e-05,
|
| 140 |
+
"loss": 1.716,
|
| 141 |
+
"step": 44
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 1.7,
|
| 145 |
+
"learning_rate": 1.9841463227464862e-05,
|
| 146 |
+
"loss": 1.5832,
|
| 147 |
+
"step": 46
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.78,
|
| 151 |
+
"learning_rate": 1.9827409020881775e-05,
|
| 152 |
+
"loss": 1.6131,
|
| 153 |
+
"step": 48
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.85,
|
| 157 |
+
"learning_rate": 1.9812763445068725e-05,
|
| 158 |
+
"loss": 1.6786,
|
| 159 |
+
"step": 50
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.93,
|
| 163 |
+
"learning_rate": 1.979752738133054e-05,
|
| 164 |
+
"loss": 1.5621,
|
| 165 |
+
"step": 52
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.0,
|
| 169 |
+
"learning_rate": 1.9781701746504926e-05,
|
| 170 |
+
"loss": 1.6118,
|
| 171 |
+
"step": 54
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.07,
|
| 175 |
+
"learning_rate": 1.9765287492907347e-05,
|
| 176 |
+
"loss": 1.4956,
|
| 177 |
+
"step": 56
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.15,
|
| 181 |
+
"learning_rate": 1.974828560827368e-05,
|
| 182 |
+
"loss": 1.5483,
|
| 183 |
+
"step": 58
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 2.22,
|
| 187 |
+
"learning_rate": 1.97306971157008e-05,
|
| 188 |
+
"loss": 1.4634,
|
| 189 |
+
"step": 60
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 2.3,
|
| 193 |
+
"learning_rate": 1.9712523073585e-05,
|
| 194 |
+
"loss": 1.5822,
|
| 195 |
+
"step": 62
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.37,
|
| 199 |
+
"learning_rate": 1.969376457555831e-05,
|
| 200 |
+
"loss": 1.4255,
|
| 201 |
+
"step": 64
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 2.44,
|
| 205 |
+
"learning_rate": 1.967442275042269e-05,
|
| 206 |
+
"loss": 1.5571,
|
| 207 |
+
"step": 66
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 2.52,
|
| 211 |
+
"learning_rate": 1.9654498762082096e-05,
|
| 212 |
+
"loss": 1.4493,
|
| 213 |
+
"step": 68
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 2.59,
|
| 217 |
+
"learning_rate": 1.963399380947244e-05,
|
| 218 |
+
"loss": 1.4962,
|
| 219 |
+
"step": 70
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 2.67,
|
| 223 |
+
"learning_rate": 1.9612909126489457e-05,
|
| 224 |
+
"loss": 1.445,
|
| 225 |
+
"step": 72
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 2.74,
|
| 229 |
+
"learning_rate": 1.9591245981914446e-05,
|
| 230 |
+
"loss": 1.5671,
|
| 231 |
+
"step": 74
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 2.81,
|
| 235 |
+
"learning_rate": 1.9569005679337905e-05,
|
| 236 |
+
"loss": 1.4461,
|
| 237 |
+
"step": 76
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.89,
|
| 241 |
+
"learning_rate": 1.9546189557081124e-05,
|
| 242 |
+
"loss": 1.4698,
|
| 243 |
+
"step": 78
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 2.96,
|
| 247 |
+
"learning_rate": 1.952279898811561e-05,
|
| 248 |
+
"loss": 1.5083,
|
| 249 |
+
"step": 80
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 3.04,
|
| 253 |
+
"learning_rate": 1.949883537998049e-05,
|
| 254 |
+
"loss": 1.382,
|
| 255 |
+
"step": 82
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 3.11,
|
| 259 |
+
"learning_rate": 1.947430017469783e-05,
|
| 260 |
+
"loss": 1.3836,
|
| 261 |
+
"step": 84
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 3.19,
|
| 265 |
+
"learning_rate": 1.9449194848685807e-05,
|
| 266 |
+
"loss": 1.386,
|
| 267 |
+
"step": 86
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.26,
|
| 271 |
+
"learning_rate": 1.9423520912669913e-05,
|
| 272 |
+
"loss": 1.4059,
|
| 273 |
+
"step": 88
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 3.33,
|
| 277 |
+
"learning_rate": 1.9397279911592034e-05,
|
| 278 |
+
"loss": 1.3551,
|
| 279 |
+
"step": 90
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 3.41,
|
| 283 |
+
"learning_rate": 1.9370473424517463e-05,
|
| 284 |
+
"loss": 1.2524,
|
| 285 |
+
"step": 92
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 3.48,
|
| 289 |
+
"learning_rate": 1.9343103064539915e-05,
|
| 290 |
+
"loss": 1.3327,
|
| 291 |
+
"step": 94
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 3.56,
|
| 295 |
+
"learning_rate": 1.9315170478684423e-05,
|
| 296 |
+
"loss": 1.3513,
|
| 297 |
+
"step": 96
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 3.63,
|
| 301 |
+
"learning_rate": 1.928667734780825e-05,
|
| 302 |
+
"loss": 1.3488,
|
| 303 |
+
"step": 98
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 3.7,
|
| 307 |
+
"learning_rate": 1.9257625386499716e-05,
|
| 308 |
+
"loss": 1.372,
|
| 309 |
+
"step": 100
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 3.78,
|
| 313 |
+
"learning_rate": 1.922801634297507e-05,
|
| 314 |
+
"loss": 1.298,
|
| 315 |
+
"step": 102
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 3.85,
|
| 319 |
+
"learning_rate": 1.9197851998973237e-05,
|
| 320 |
+
"loss": 1.3144,
|
| 321 |
+
"step": 104
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 3.93,
|
| 325 |
+
"learning_rate": 1.916713416964864e-05,
|
| 326 |
+
"loss": 1.3531,
|
| 327 |
+
"step": 106
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 4.0,
|
| 331 |
+
"learning_rate": 1.9135864703461943e-05,
|
| 332 |
+
"loss": 1.3806,
|
| 333 |
+
"step": 108
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 4.07,
|
| 337 |
+
"learning_rate": 1.910404548206884e-05,
|
| 338 |
+
"loss": 1.2281,
|
| 339 |
+
"step": 110
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.15,
|
| 343 |
+
"learning_rate": 1.907167842020683e-05,
|
| 344 |
+
"loss": 1.2588,
|
| 345 |
+
"step": 112
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 4.22,
|
| 349 |
+
"learning_rate": 1.903876546557996e-05,
|
| 350 |
+
"loss": 1.2666,
|
| 351 |
+
"step": 114
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 4.3,
|
| 355 |
+
"learning_rate": 1.9005308598741662e-05,
|
| 356 |
+
"loss": 1.213,
|
| 357 |
+
"step": 116
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 4.37,
|
| 361 |
+
"learning_rate": 1.897130983297555e-05,
|
| 362 |
+
"loss": 1.187,
|
| 363 |
+
"step": 118
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 4.44,
|
| 367 |
+
"learning_rate": 1.8936771214174292e-05,
|
| 368 |
+
"loss": 1.1696,
|
| 369 |
+
"step": 120
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 4.52,
|
| 373 |
+
"learning_rate": 1.8901694820716456e-05,
|
| 374 |
+
"loss": 1.1778,
|
| 375 |
+
"step": 122
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 4.59,
|
| 379 |
+
"learning_rate": 1.886608276334148e-05,
|
| 380 |
+
"loss": 1.2148,
|
| 381 |
+
"step": 124
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 4.67,
|
| 385 |
+
"learning_rate": 1.882993718502265e-05,
|
| 386 |
+
"loss": 1.1771,
|
| 387 |
+
"step": 126
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 4.74,
|
| 391 |
+
"learning_rate": 1.8793260260838127e-05,
|
| 392 |
+
"loss": 1.2162,
|
| 393 |
+
"step": 128
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 4.81,
|
| 397 |
+
"learning_rate": 1.875605419784009e-05,
|
| 398 |
+
"loss": 1.213,
|
| 399 |
+
"step": 130
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 4.89,
|
| 403 |
+
"learning_rate": 1.8718321234921895e-05,
|
| 404 |
+
"loss": 1.2071,
|
| 405 |
+
"step": 132
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 4.96,
|
| 409 |
+
"learning_rate": 1.868006364268337e-05,
|
| 410 |
+
"loss": 1.1868,
|
| 411 |
+
"step": 134
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 5.04,
|
| 415 |
+
"learning_rate": 1.8641283723294177e-05,
|
| 416 |
+
"loss": 1.1492,
|
| 417 |
+
"step": 136
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 5.11,
|
| 421 |
+
"learning_rate": 1.8601983810355265e-05,
|
| 422 |
+
"loss": 1.1094,
|
| 423 |
+
"step": 138
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 5.19,
|
| 427 |
+
"learning_rate": 1.8562166268758462e-05,
|
| 428 |
+
"loss": 1.1069,
|
| 429 |
+
"step": 140
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 5.26,
|
| 433 |
+
"learning_rate": 1.8521833494544153e-05,
|
| 434 |
+
"loss": 1.1004,
|
| 435 |
+
"step": 142
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 5.33,
|
| 439 |
+
"learning_rate": 1.848098791475711e-05,
|
| 440 |
+
"loss": 1.0538,
|
| 441 |
+
"step": 144
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 5.41,
|
| 445 |
+
"learning_rate": 1.8439631987300418e-05,
|
| 446 |
+
"loss": 1.0556,
|
| 447 |
+
"step": 146
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 5.48,
|
| 451 |
+
"learning_rate": 1.8397768200787613e-05,
|
| 452 |
+
"loss": 1.0134,
|
| 453 |
+
"step": 148
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 5.56,
|
| 457 |
+
"learning_rate": 1.8355399074392875e-05,
|
| 458 |
+
"loss": 1.0615,
|
| 459 |
+
"step": 150
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 5.63,
|
| 463 |
+
"learning_rate": 1.8312527157699488e-05,
|
| 464 |
+
"loss": 1.0209,
|
| 465 |
+
"step": 152
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 5.7,
|
| 469 |
+
"learning_rate": 1.826915503054637e-05,
|
| 470 |
+
"loss": 1.0104,
|
| 471 |
+
"step": 154
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 5.78,
|
| 475 |
+
"learning_rate": 1.8225285302872872e-05,
|
| 476 |
+
"loss": 1.0017,
|
| 477 |
+
"step": 156
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 5.85,
|
| 481 |
+
"learning_rate": 1.818092061456169e-05,
|
| 482 |
+
"loss": 1.0131,
|
| 483 |
+
"step": 158
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 5.93,
|
| 487 |
+
"learning_rate": 1.8136063635280036e-05,
|
| 488 |
+
"loss": 1.0151,
|
| 489 |
+
"step": 160
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 6.0,
|
| 493 |
+
"learning_rate": 1.8090717064318958e-05,
|
| 494 |
+
"loss": 0.9734,
|
| 495 |
+
"step": 162
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 6.07,
|
| 499 |
+
"learning_rate": 1.804488363043094e-05,
|
| 500 |
+
"loss": 0.8721,
|
| 501 |
+
"step": 164
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 6.15,
|
| 505 |
+
"learning_rate": 1.799856609166568e-05,
|
| 506 |
+
"loss": 0.885,
|
| 507 |
+
"step": 166
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 6.22,
|
| 511 |
+
"learning_rate": 1.7951767235204142e-05,
|
| 512 |
+
"loss": 0.8312,
|
| 513 |
+
"step": 168
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 6.3,
|
| 517 |
+
"learning_rate": 1.79044898771908e-05,
|
| 518 |
+
"loss": 0.8543,
|
| 519 |
+
"step": 170
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 6.37,
|
| 523 |
+
"learning_rate": 1.7856736862564208e-05,
|
| 524 |
+
"loss": 0.8782,
|
| 525 |
+
"step": 172
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 6.44,
|
| 529 |
+
"learning_rate": 1.7808511064885798e-05,
|
| 530 |
+
"loss": 0.8549,
|
| 531 |
+
"step": 174
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 6.52,
|
| 535 |
+
"learning_rate": 1.7759815386166945e-05,
|
| 536 |
+
"loss": 0.795,
|
| 537 |
+
"step": 176
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 6.59,
|
| 541 |
+
"learning_rate": 1.771065275669436e-05,
|
| 542 |
+
"loss": 0.883,
|
| 543 |
+
"step": 178
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 6.67,
|
| 547 |
+
"learning_rate": 1.766102613485375e-05,
|
| 548 |
+
"loss": 0.8116,
|
| 549 |
+
"step": 180
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 6.74,
|
| 553 |
+
"learning_rate": 1.761093850695178e-05,
|
| 554 |
+
"loss": 0.8215,
|
| 555 |
+
"step": 182
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 6.81,
|
| 559 |
+
"learning_rate": 1.75603928870364e-05,
|
| 560 |
+
"loss": 0.8416,
|
| 561 |
+
"step": 184
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 6.89,
|
| 565 |
+
"learning_rate": 1.7509392316715445e-05,
|
| 566 |
+
"loss": 0.7994,
|
| 567 |
+
"step": 186
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 6.96,
|
| 571 |
+
"learning_rate": 1.7457939864973625e-05,
|
| 572 |
+
"loss": 0.8312,
|
| 573 |
+
"step": 188
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 7.04,
|
| 577 |
+
"learning_rate": 1.7406038627987835e-05,
|
| 578 |
+
"loss": 0.7368,
|
| 579 |
+
"step": 190
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 7.11,
|
| 583 |
+
"learning_rate": 1.7353691728940852e-05,
|
| 584 |
+
"loss": 0.6201,
|
| 585 |
+
"step": 192
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 7.19,
|
| 589 |
+
"learning_rate": 1.7300902317833377e-05,
|
| 590 |
+
"loss": 0.6334,
|
| 591 |
+
"step": 194
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 7.26,
|
| 595 |
+
"learning_rate": 1.724767357129451e-05,
|
| 596 |
+
"loss": 0.6199,
|
| 597 |
+
"step": 196
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 7.33,
|
| 601 |
+
"learning_rate": 1.7194008692390566e-05,
|
| 602 |
+
"loss": 0.6208,
|
| 603 |
+
"step": 198
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 7.41,
|
| 607 |
+
"learning_rate": 1.7139910910432353e-05,
|
| 608 |
+
"loss": 0.6039,
|
| 609 |
+
"step": 200
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 7.48,
|
| 613 |
+
"learning_rate": 1.708538348078082e-05,
|
| 614 |
+
"loss": 0.5737,
|
| 615 |
+
"step": 202
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 7.56,
|
| 619 |
+
"learning_rate": 1.7030429684651203e-05,
|
| 620 |
+
"loss": 0.6267,
|
| 621 |
+
"step": 204
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 7.63,
|
| 625 |
+
"learning_rate": 1.6975052828915534e-05,
|
| 626 |
+
"loss": 0.6654,
|
| 627 |
+
"step": 206
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 7.7,
|
| 631 |
+
"learning_rate": 1.6919256245903675e-05,
|
| 632 |
+
"loss": 0.6041,
|
| 633 |
+
"step": 208
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 7.78,
|
| 637 |
+
"learning_rate": 1.686304329320278e-05,
|
| 638 |
+
"loss": 0.6349,
|
| 639 |
+
"step": 210
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 7.85,
|
| 643 |
+
"learning_rate": 1.6806417353455267e-05,
|
| 644 |
+
"loss": 0.6032,
|
| 645 |
+
"step": 212
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 7.93,
|
| 649 |
+
"learning_rate": 1.674938183415525e-05,
|
| 650 |
+
"loss": 0.6501,
|
| 651 |
+
"step": 214
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 8.0,
|
| 655 |
+
"learning_rate": 1.6691940167443487e-05,
|
| 656 |
+
"loss": 0.5961,
|
| 657 |
+
"step": 216
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 8.07,
|
| 661 |
+
"learning_rate": 1.6634095809900873e-05,
|
| 662 |
+
"loss": 0.4455,
|
| 663 |
+
"step": 218
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 8.15,
|
| 667 |
+
"learning_rate": 1.657585224234042e-05,
|
| 668 |
+
"loss": 0.4305,
|
| 669 |
+
"step": 220
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 8.22,
|
| 673 |
+
"learning_rate": 1.6517212969597793e-05,
|
| 674 |
+
"loss": 0.4517,
|
| 675 |
+
"step": 222
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 8.3,
|
| 679 |
+
"learning_rate": 1.6458181520320426e-05,
|
| 680 |
+
"loss": 0.4096,
|
| 681 |
+
"step": 224
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 8.37,
|
| 685 |
+
"learning_rate": 1.639876144675516e-05,
|
| 686 |
+
"loss": 0.4079,
|
| 687 |
+
"step": 226
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 8.44,
|
| 691 |
+
"learning_rate": 1.6338956324534514e-05,
|
| 692 |
+
"loss": 0.4153,
|
| 693 |
+
"step": 228
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 8.52,
|
| 697 |
+
"learning_rate": 1.6278769752461477e-05,
|
| 698 |
+
"loss": 0.3869,
|
| 699 |
+
"step": 230
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 8.59,
|
| 703 |
+
"learning_rate": 1.6218205352292984e-05,
|
| 704 |
+
"loss": 0.3988,
|
| 705 |
+
"step": 232
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 8.67,
|
| 709 |
+
"learning_rate": 1.6157266768521967e-05,
|
| 710 |
+
"loss": 0.4212,
|
| 711 |
+
"step": 234
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 8.74,
|
| 715 |
+
"learning_rate": 1.6095957668158043e-05,
|
| 716 |
+
"loss": 0.4137,
|
| 717 |
+
"step": 236
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 8.81,
|
| 721 |
+
"learning_rate": 1.6034281740506846e-05,
|
| 722 |
+
"loss": 0.4265,
|
| 723 |
+
"step": 238
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 8.89,
|
| 727 |
+
"learning_rate": 1.5972242696948024e-05,
|
| 728 |
+
"loss": 0.3991,
|
| 729 |
+
"step": 240
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 8.96,
|
| 733 |
+
"learning_rate": 1.59098442707119e-05,
|
| 734 |
+
"loss": 0.384,
|
| 735 |
+
"step": 242
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 9.04,
|
| 739 |
+
"learning_rate": 1.584709021665485e-05,
|
| 740 |
+
"loss": 0.3623,
|
| 741 |
+
"step": 244
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 9.11,
|
| 745 |
+
"learning_rate": 1.5783984311033296e-05,
|
| 746 |
+
"loss": 0.2732,
|
| 747 |
+
"step": 246
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 9.19,
|
| 751 |
+
"learning_rate": 1.5720530351276542e-05,
|
| 752 |
+
"loss": 0.2461,
|
| 753 |
+
"step": 248
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 9.26,
|
| 757 |
+
"learning_rate": 1.565673215575819e-05,
|
| 758 |
+
"loss": 0.266,
|
| 759 |
+
"step": 250
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 9.33,
|
| 763 |
+
"learning_rate": 1.5592593563566424e-05,
|
| 764 |
+
"loss": 0.2524,
|
| 765 |
+
"step": 252
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 9.41,
|
| 769 |
+
"learning_rate": 1.552811843427295e-05,
|
| 770 |
+
"loss": 0.2587,
|
| 771 |
+
"step": 254
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 9.48,
|
| 775 |
+
"learning_rate": 1.5463310647700777e-05,
|
| 776 |
+
"loss": 0.2538,
|
| 777 |
+
"step": 256
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 9.56,
|
| 781 |
+
"learning_rate": 1.5398174103690717e-05,
|
| 782 |
+
"loss": 0.2541,
|
| 783 |
+
"step": 258
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 9.63,
|
| 787 |
+
"learning_rate": 1.5332712721866735e-05,
|
| 788 |
+
"loss": 0.2695,
|
| 789 |
+
"step": 260
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 9.7,
|
| 793 |
+
"learning_rate": 1.5266930441400066e-05,
|
| 794 |
+
"loss": 0.2887,
|
| 795 |
+
"step": 262
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 9.78,
|
| 799 |
+
"learning_rate": 1.5200831220772187e-05,
|
| 800 |
+
"loss": 0.2753,
|
| 801 |
+
"step": 264
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 9.85,
|
| 805 |
+
"learning_rate": 1.5134419037536603e-05,
|
| 806 |
+
"loss": 0.2613,
|
| 807 |
+
"step": 266
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 9.93,
|
| 811 |
+
"learning_rate": 1.5067697888079496e-05,
|
| 812 |
+
"loss": 0.2655,
|
| 813 |
+
"step": 268
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 10.0,
|
| 817 |
+
"learning_rate": 1.5000671787379257e-05,
|
| 818 |
+
"loss": 0.2828,
|
| 819 |
+
"step": 270
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 10.07,
|
| 823 |
+
"learning_rate": 1.4933344768764857e-05,
|
| 824 |
+
"loss": 0.2013,
|
| 825 |
+
"step": 272
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 10.15,
|
| 829 |
+
"learning_rate": 1.4865720883673153e-05,
|
| 830 |
+
"loss": 0.194,
|
| 831 |
+
"step": 274
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 10.22,
|
| 835 |
+
"learning_rate": 1.4797804201405099e-05,
|
| 836 |
+
"loss": 0.1917,
|
| 837 |
+
"step": 276
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 10.3,
|
| 841 |
+
"learning_rate": 1.472959880888086e-05,
|
| 842 |
+
"loss": 0.1783,
|
| 843 |
+
"step": 278
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 10.37,
|
| 847 |
+
"learning_rate": 1.4661108810393883e-05,
|
| 848 |
+
"loss": 0.1992,
|
| 849 |
+
"step": 280
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 10.44,
|
| 853 |
+
"learning_rate": 1.4592338327363917e-05,
|
| 854 |
+
"loss": 0.1804,
|
| 855 |
+
"step": 282
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 10.52,
|
| 859 |
+
"learning_rate": 1.4523291498089021e-05,
|
| 860 |
+
"loss": 0.1834,
|
| 861 |
+
"step": 284
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 10.59,
|
| 865 |
+
"learning_rate": 1.4453972477496509e-05,
|
| 866 |
+
"loss": 0.1938,
|
| 867 |
+
"step": 286
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 10.67,
|
| 871 |
+
"learning_rate": 1.4384385436892963e-05,
|
| 872 |
+
"loss": 0.1756,
|
| 873 |
+
"step": 288
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 10.74,
|
| 877 |
+
"learning_rate": 1.4314534563713186e-05,
|
| 878 |
+
"loss": 0.1928,
|
| 879 |
+
"step": 290
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 10.81,
|
| 883 |
+
"learning_rate": 1.4244424061268239e-05,
|
| 884 |
+
"loss": 0.2066,
|
| 885 |
+
"step": 292
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 10.89,
|
| 889 |
+
"learning_rate": 1.4174058148492515e-05,
|
| 890 |
+
"loss": 0.2115,
|
| 891 |
+
"step": 294
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 10.96,
|
| 895 |
+
"learning_rate": 1.4103441059689843e-05,
|
| 896 |
+
"loss": 0.2111,
|
| 897 |
+
"step": 296
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 11.04,
|
| 901 |
+
"learning_rate": 1.4032577044278695e-05,
|
| 902 |
+
"loss": 0.1766,
|
| 903 |
+
"step": 298
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 11.11,
|
| 907 |
+
"learning_rate": 1.3961470366536472e-05,
|
| 908 |
+
"loss": 0.1426,
|
| 909 |
+
"step": 300
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 11.19,
|
| 913 |
+
"learning_rate": 1.3890125305342905e-05,
|
| 914 |
+
"loss": 0.1705,
|
| 915 |
+
"step": 302
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 11.26,
|
| 919 |
+
"learning_rate": 1.3818546153922567e-05,
|
| 920 |
+
"loss": 0.1591,
|
| 921 |
+
"step": 304
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 11.33,
|
| 925 |
+
"learning_rate": 1.374673721958653e-05,
|
| 926 |
+
"loss": 0.1579,
|
| 927 |
+
"step": 306
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 11.41,
|
| 931 |
+
"learning_rate": 1.3674702823473165e-05,
|
| 932 |
+
"loss": 0.1579,
|
| 933 |
+
"step": 308
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 11.48,
|
| 937 |
+
"learning_rate": 1.3602447300288114e-05,
|
| 938 |
+
"loss": 0.1607,
|
| 939 |
+
"step": 310
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 11.56,
|
| 943 |
+
"learning_rate": 1.3529974998043453e-05,
|
| 944 |
+
"loss": 0.1569,
|
| 945 |
+
"step": 312
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 11.63,
|
| 949 |
+
"learning_rate": 1.345729027779605e-05,
|
| 950 |
+
"loss": 0.1809,
|
| 951 |
+
"step": 314
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 11.7,
|
| 955 |
+
"learning_rate": 1.3384397513385126e-05,
|
| 956 |
+
"loss": 0.1573,
|
| 957 |
+
"step": 316
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 11.78,
|
| 961 |
+
"learning_rate": 1.3311301091169067e-05,
|
| 962 |
+
"loss": 0.1686,
|
| 963 |
+
"step": 318
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 11.85,
|
| 967 |
+
"learning_rate": 1.3238005409761468e-05,
|
| 968 |
+
"loss": 0.1827,
|
| 969 |
+
"step": 320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 11.93,
|
| 973 |
+
"learning_rate": 1.3164514879766443e-05,
|
| 974 |
+
"loss": 0.1791,
|
| 975 |
+
"step": 322
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 12.0,
|
| 979 |
+
"learning_rate": 1.309083392351322e-05,
|
| 980 |
+
"loss": 0.1692,
|
| 981 |
+
"step": 324
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 12.07,
|
| 985 |
+
"learning_rate": 1.3016966974790018e-05,
|
| 986 |
+
"loss": 0.1352,
|
| 987 |
+
"step": 326
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 12.15,
|
| 991 |
+
"learning_rate": 1.294291847857725e-05,
|
| 992 |
+
"loss": 0.1432,
|
| 993 |
+
"step": 328
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 12.22,
|
| 997 |
+
"learning_rate": 1.2868692890780036e-05,
|
| 998 |
+
"loss": 0.139,
|
| 999 |
+
"step": 330
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 12.3,
|
| 1003 |
+
"learning_rate": 1.2794294677960073e-05,
|
| 1004 |
+
"loss": 0.1614,
|
| 1005 |
+
"step": 332
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 12.37,
|
| 1009 |
+
"learning_rate": 1.2719728317066853e-05,
|
| 1010 |
+
"loss": 0.1562,
|
| 1011 |
+
"step": 334
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 12.44,
|
| 1015 |
+
"learning_rate": 1.2644998295168257e-05,
|
| 1016 |
+
"loss": 0.1419,
|
| 1017 |
+
"step": 336
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 12.52,
|
| 1021 |
+
"learning_rate": 1.2570109109180561e-05,
|
| 1022 |
+
"loss": 0.1649,
|
| 1023 |
+
"step": 338
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 12.59,
|
| 1027 |
+
"learning_rate": 1.249506526559781e-05,
|
| 1028 |
+
"loss": 0.1539,
|
| 1029 |
+
"step": 340
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 12.67,
|
| 1033 |
+
"learning_rate": 1.2419871280220646e-05,
|
| 1034 |
+
"loss": 0.1572,
|
| 1035 |
+
"step": 342
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 12.74,
|
| 1039 |
+
"learning_rate": 1.2344531677884572e-05,
|
| 1040 |
+
"loss": 0.1558,
|
| 1041 |
+
"step": 344
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 12.81,
|
| 1045 |
+
"learning_rate": 1.2269050992187666e-05,
|
| 1046 |
+
"loss": 0.1521,
|
| 1047 |
+
"step": 346
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 12.89,
|
| 1051 |
+
"learning_rate": 1.219343376521776e-05,
|
| 1052 |
+
"loss": 0.153,
|
| 1053 |
+
"step": 348
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 12.96,
|
| 1057 |
+
"learning_rate": 1.211768454727913e-05,
|
| 1058 |
+
"loss": 0.1582,
|
| 1059 |
+
"step": 350
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 13.04,
|
| 1063 |
+
"learning_rate": 1.2041807896618676e-05,
|
| 1064 |
+
"loss": 0.1504,
|
| 1065 |
+
"step": 352
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 13.11,
|
| 1069 |
+
"learning_rate": 1.1965808379151622e-05,
|
| 1070 |
+
"loss": 0.1386,
|
| 1071 |
+
"step": 354
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 13.19,
|
| 1075 |
+
"learning_rate": 1.1889690568186767e-05,
|
| 1076 |
+
"loss": 0.1438,
|
| 1077 |
+
"step": 356
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 13.26,
|
| 1081 |
+
"learning_rate": 1.1813459044151276e-05,
|
| 1082 |
+
"loss": 0.1299,
|
| 1083 |
+
"step": 358
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 13.33,
|
| 1087 |
+
"learning_rate": 1.1737118394315052e-05,
|
| 1088 |
+
"loss": 0.1398,
|
| 1089 |
+
"step": 360
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 13.41,
|
| 1093 |
+
"learning_rate": 1.1660673212514708e-05,
|
| 1094 |
+
"loss": 0.1417,
|
| 1095 |
+
"step": 362
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 13.48,
|
| 1099 |
+
"learning_rate": 1.1584128098877112e-05,
|
| 1100 |
+
"loss": 0.1463,
|
| 1101 |
+
"step": 364
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 13.56,
|
| 1105 |
+
"learning_rate": 1.1507487659542578e-05,
|
| 1106 |
+
"loss": 0.1483,
|
| 1107 |
+
"step": 366
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 13.63,
|
| 1111 |
+
"learning_rate": 1.1430756506387699e-05,
|
| 1112 |
+
"loss": 0.1514,
|
| 1113 |
+
"step": 368
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 13.7,
|
| 1117 |
+
"learning_rate": 1.135393925674781e-05,
|
| 1118 |
+
"loss": 0.1428,
|
| 1119 |
+
"step": 370
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 13.78,
|
| 1123 |
+
"learning_rate": 1.1277040533139149e-05,
|
| 1124 |
+
"loss": 0.1588,
|
| 1125 |
+
"step": 372
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 13.85,
|
| 1129 |
+
"learning_rate": 1.1200064962980692e-05,
|
| 1130 |
+
"loss": 0.156,
|
| 1131 |
+
"step": 374
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 13.93,
|
| 1135 |
+
"learning_rate": 1.1123017178315692e-05,
|
| 1136 |
+
"loss": 0.1455,
|
| 1137 |
+
"step": 376
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 14.0,
|
| 1141 |
+
"learning_rate": 1.104590181553294e-05,
|
| 1142 |
+
"loss": 0.1627,
|
| 1143 |
+
"step": 378
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 14.07,
|
| 1147 |
+
"learning_rate": 1.0968723515087786e-05,
|
| 1148 |
+
"loss": 0.1357,
|
| 1149 |
+
"step": 380
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 14.15,
|
| 1153 |
+
"learning_rate": 1.0891486921222877e-05,
|
| 1154 |
+
"loss": 0.1405,
|
| 1155 |
+
"step": 382
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 14.22,
|
| 1159 |
+
"learning_rate": 1.0814196681688695e-05,
|
| 1160 |
+
"loss": 0.1307,
|
| 1161 |
+
"step": 384
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 14.3,
|
| 1165 |
+
"learning_rate": 1.0736857447463887e-05,
|
| 1166 |
+
"loss": 0.121,
|
| 1167 |
+
"step": 386
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 14.37,
|
| 1171 |
+
"learning_rate": 1.065947387247536e-05,
|
| 1172 |
+
"loss": 0.1381,
|
| 1173 |
+
"step": 388
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 14.44,
|
| 1177 |
+
"learning_rate": 1.0582050613318276e-05,
|
| 1178 |
+
"loss": 0.1364,
|
| 1179 |
+
"step": 390
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 14.52,
|
| 1183 |
+
"learning_rate": 1.0504592328975783e-05,
|
| 1184 |
+
"loss": 0.139,
|
| 1185 |
+
"step": 392
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 14.59,
|
| 1189 |
+
"learning_rate": 1.0427103680538706e-05,
|
| 1190 |
+
"loss": 0.137,
|
| 1191 |
+
"step": 394
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 14.67,
|
| 1195 |
+
"learning_rate": 1.0349589330925039e-05,
|
| 1196 |
+
"loss": 0.1381,
|
| 1197 |
+
"step": 396
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 14.74,
|
| 1201 |
+
"learning_rate": 1.0272053944599358e-05,
|
| 1202 |
+
"loss": 0.1438,
|
| 1203 |
+
"step": 398
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 14.81,
|
| 1207 |
+
"learning_rate": 1.0194502187292128e-05,
|
| 1208 |
+
"loss": 0.135,
|
| 1209 |
+
"step": 400
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 14.89,
|
| 1213 |
+
"learning_rate": 1.0116938725718956e-05,
|
| 1214 |
+
"loss": 0.1435,
|
| 1215 |
+
"step": 402
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 14.96,
|
| 1219 |
+
"learning_rate": 1.0039368227299753e-05,
|
| 1220 |
+
"loss": 0.1518,
|
| 1221 |
+
"step": 404
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 15.04,
|
| 1225 |
+
"learning_rate": 9.961795359877873e-06,
|
| 1226 |
+
"loss": 0.1295,
|
| 1227 |
+
"step": 406
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 15.11,
|
| 1231 |
+
"learning_rate": 9.884224791439229e-06,
|
| 1232 |
+
"loss": 0.1267,
|
| 1233 |
+
"step": 408
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 15.19,
|
| 1237 |
+
"learning_rate": 9.806661189831397e-06,
|
| 1238 |
+
"loss": 0.1278,
|
| 1239 |
+
"step": 410
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 15.26,
|
| 1243 |
+
"learning_rate": 9.729109222482708e-06,
|
| 1244 |
+
"loss": 0.1344,
|
| 1245 |
+
"step": 412
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 15.33,
|
| 1249 |
+
"learning_rate": 9.651573556121408e-06,
|
| 1250 |
+
"loss": 0.1359,
|
| 1251 |
+
"step": 414
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 15.41,
|
| 1255 |
+
"learning_rate": 9.574058856494819e-06,
|
| 1256 |
+
"loss": 0.1313,
|
| 1257 |
+
"step": 416
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 15.48,
|
| 1261 |
+
"learning_rate": 9.496569788088572e-06,
|
| 1262 |
+
"loss": 0.1369,
|
| 1263 |
+
"step": 418
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 15.56,
|
| 1267 |
+
"learning_rate": 9.419111013845939e-06,
|
| 1268 |
+
"loss": 0.1224,
|
| 1269 |
+
"step": 420
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 15.63,
|
| 1273 |
+
"learning_rate": 9.341687194887216e-06,
|
| 1274 |
+
"loss": 0.1433,
|
| 1275 |
+
"step": 422
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 15.7,
|
| 1279 |
+
"learning_rate": 9.264302990229253e-06,
|
| 1280 |
+
"loss": 0.1263,
|
| 1281 |
+
"step": 424
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 15.78,
|
| 1285 |
+
"learning_rate": 9.186963056505082e-06,
|
| 1286 |
+
"loss": 0.1449,
|
| 1287 |
+
"step": 426
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 15.85,
|
| 1291 |
+
"learning_rate": 9.109672047683723e-06,
|
| 1292 |
+
"loss": 0.1351,
|
| 1293 |
+
"step": 428
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 15.93,
|
| 1297 |
+
"learning_rate": 9.032434614790101e-06,
|
| 1298 |
+
"loss": 0.1391,
|
| 1299 |
+
"step": 430
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 16.0,
|
| 1303 |
+
"learning_rate": 8.955255405625195e-06,
|
| 1304 |
+
"loss": 0.1392,
|
| 1305 |
+
"step": 432
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 16.07,
|
| 1309 |
+
"learning_rate": 8.878139064486332e-06,
|
| 1310 |
+
"loss": 0.1294,
|
| 1311 |
+
"step": 434
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 16.15,
|
| 1315 |
+
"learning_rate": 8.801090231887733e-06,
|
| 1316 |
+
"loss": 0.1151,
|
| 1317 |
+
"step": 436
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 16.22,
|
| 1321 |
+
"learning_rate": 8.724113544281248e-06,
|
| 1322 |
+
"loss": 0.113,
|
| 1323 |
+
"step": 438
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 16.3,
|
| 1327 |
+
"learning_rate": 8.647213633777381e-06,
|
| 1328 |
+
"loss": 0.1272,
|
| 1329 |
+
"step": 440
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 16.37,
|
| 1333 |
+
"learning_rate": 8.570395127866521e-06,
|
| 1334 |
+
"loss": 0.1246,
|
| 1335 |
+
"step": 442
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 16.44,
|
| 1339 |
+
"learning_rate": 8.49366264914051e-06,
|
| 1340 |
+
"loss": 0.1322,
|
| 1341 |
+
"step": 444
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 16.52,
|
| 1345 |
+
"learning_rate": 8.417020815014442e-06,
|
| 1346 |
+
"loss": 0.1351,
|
| 1347 |
+
"step": 446
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 16.59,
|
| 1351 |
+
"learning_rate": 8.340474237448851e-06,
|
| 1352 |
+
"loss": 0.1286,
|
| 1353 |
+
"step": 448
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 16.67,
|
| 1357 |
+
"learning_rate": 8.264027522672134e-06,
|
| 1358 |
+
"loss": 0.1317,
|
| 1359 |
+
"step": 450
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 16.74,
|
| 1363 |
+
"learning_rate": 8.187685270903419e-06,
|
| 1364 |
+
"loss": 0.1277,
|
| 1365 |
+
"step": 452
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 16.81,
|
| 1369 |
+
"learning_rate": 8.1114520760757e-06,
|
| 1370 |
+
"loss": 0.1361,
|
| 1371 |
+
"step": 454
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 16.89,
|
| 1375 |
+
"learning_rate": 8.035332525559436e-06,
|
| 1376 |
+
"loss": 0.1389,
|
| 1377 |
+
"step": 456
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 16.96,
|
| 1381 |
+
"learning_rate": 7.959331199886464e-06,
|
| 1382 |
+
"loss": 0.1382,
|
| 1383 |
+
"step": 458
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 17.04,
|
| 1387 |
+
"learning_rate": 7.88345267247439e-06,
|
| 1388 |
+
"loss": 0.1342,
|
| 1389 |
+
"step": 460
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 17.11,
|
| 1393 |
+
"learning_rate": 7.807701509351382e-06,
|
| 1394 |
+
"loss": 0.1158,
|
| 1395 |
+
"step": 462
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 17.19,
|
| 1399 |
+
"learning_rate": 7.732082268881377e-06,
|
| 1400 |
+
"loss": 0.1207,
|
| 1401 |
+
"step": 464
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 17.26,
|
| 1405 |
+
"learning_rate": 7.656599501489827e-06,
|
| 1406 |
+
"loss": 0.116,
|
| 1407 |
+
"step": 466
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 17.33,
|
| 1411 |
+
"learning_rate": 7.581257749389826e-06,
|
| 1412 |
+
"loss": 0.1275,
|
| 1413 |
+
"step": 468
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 17.41,
|
| 1417 |
+
"learning_rate": 7.506061546308818e-06,
|
| 1418 |
+
"loss": 0.1182,
|
| 1419 |
+
"step": 470
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 17.48,
|
| 1423 |
+
"learning_rate": 7.431015417215751e-06,
|
| 1424 |
+
"loss": 0.1328,
|
| 1425 |
+
"step": 472
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 17.56,
|
| 1429 |
+
"learning_rate": 7.356123878048806e-06,
|
| 1430 |
+
"loss": 0.1206,
|
| 1431 |
+
"step": 474
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 17.63,
|
| 1435 |
+
"learning_rate": 7.2813914354436325e-06,
|
| 1436 |
+
"loss": 0.125,
|
| 1437 |
+
"step": 476
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 17.7,
|
| 1441 |
+
"learning_rate": 7.206822586462168e-06,
|
| 1442 |
+
"loss": 0.1229,
|
| 1443 |
+
"step": 478
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 17.78,
|
| 1447 |
+
"learning_rate": 7.132421818322015e-06,
|
| 1448 |
+
"loss": 0.1301,
|
| 1449 |
+
"step": 480
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 17.85,
|
| 1453 |
+
"learning_rate": 7.058193608126439e-06,
|
| 1454 |
+
"loss": 0.1367,
|
| 1455 |
+
"step": 482
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 17.93,
|
| 1459 |
+
"learning_rate": 6.984142422594935e-06,
|
| 1460 |
+
"loss": 0.1368,
|
| 1461 |
+
"step": 484
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 18.0,
|
| 1465 |
+
"learning_rate": 6.910272717794462e-06,
|
| 1466 |
+
"loss": 0.1279,
|
| 1467 |
+
"step": 486
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 18.07,
|
| 1471 |
+
"learning_rate": 6.836588938871271e-06,
|
| 1472 |
+
"loss": 0.1116,
|
| 1473 |
+
"step": 488
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 18.15,
|
| 1477 |
+
"learning_rate": 6.763095519783447e-06,
|
| 1478 |
+
"loss": 0.1124,
|
| 1479 |
+
"step": 490
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 18.22,
|
| 1483 |
+
"learning_rate": 6.689796883034059e-06,
|
| 1484 |
+
"loss": 0.1192,
|
| 1485 |
+
"step": 492
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 18.3,
|
| 1489 |
+
"learning_rate": 6.616697439405073e-06,
|
| 1490 |
+
"loss": 0.1209,
|
| 1491 |
+
"step": 494
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 18.37,
|
| 1495 |
+
"learning_rate": 6.54380158769189e-06,
|
| 1496 |
+
"loss": 0.1154,
|
| 1497 |
+
"step": 496
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 18.44,
|
| 1501 |
+
"learning_rate": 6.471113714438679e-06,
|
| 1502 |
+
"loss": 0.1261,
|
| 1503 |
+
"step": 498
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 18.52,
|
| 1507 |
+
"learning_rate": 6.398638193674396e-06,
|
| 1508 |
+
"loss": 0.1325,
|
| 1509 |
+
"step": 500
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 18.59,
|
| 1513 |
+
"learning_rate": 6.326379386649579e-06,
|
| 1514 |
+
"loss": 0.1204,
|
| 1515 |
+
"step": 502
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 18.67,
|
| 1519 |
+
"learning_rate": 6.2543416415739175e-06,
|
| 1520 |
+
"loss": 0.1184,
|
| 1521 |
+
"step": 504
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 18.74,
|
| 1525 |
+
"learning_rate": 6.182529293354578e-06,
|
| 1526 |
+
"loss": 0.1292,
|
| 1527 |
+
"step": 506
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 18.81,
|
| 1531 |
+
"learning_rate": 6.110946663335366e-06,
|
| 1532 |
+
"loss": 0.129,
|
| 1533 |
+
"step": 508
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 18.89,
|
| 1537 |
+
"learning_rate": 6.0395980590366685e-06,
|
| 1538 |
+
"loss": 0.1234,
|
| 1539 |
+
"step": 510
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 18.96,
|
| 1543 |
+
"learning_rate": 5.968487773896281e-06,
|
| 1544 |
+
"loss": 0.1337,
|
| 1545 |
+
"step": 512
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 19.04,
|
| 1549 |
+
"learning_rate": 5.897620087010998e-06,
|
| 1550 |
+
"loss": 0.1263,
|
| 1551 |
+
"step": 514
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 19.11,
|
| 1555 |
+
"learning_rate": 5.826999262879164e-06,
|
| 1556 |
+
"loss": 0.1153,
|
| 1557 |
+
"step": 516
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 19.19,
|
| 1561 |
+
"learning_rate": 5.7566295511440195e-06,
|
| 1562 |
+
"loss": 0.1218,
|
| 1563 |
+
"step": 518
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 19.26,
|
| 1567 |
+
"learning_rate": 5.68651518633801e-06,
|
| 1568 |
+
"loss": 0.1215,
|
| 1569 |
+
"step": 520
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 19.33,
|
| 1573 |
+
"learning_rate": 5.616660387627925e-06,
|
| 1574 |
+
"loss": 0.1173,
|
| 1575 |
+
"step": 522
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 19.41,
|
| 1579 |
+
"learning_rate": 5.547069358561064e-06,
|
| 1580 |
+
"loss": 0.1194,
|
| 1581 |
+
"step": 524
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 19.48,
|
| 1585 |
+
"learning_rate": 5.47774628681224e-06,
|
| 1586 |
+
"loss": 0.1169,
|
| 1587 |
+
"step": 526
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 19.56,
|
| 1591 |
+
"learning_rate": 5.4086953439318024e-06,
|
| 1592 |
+
"loss": 0.1179,
|
| 1593 |
+
"step": 528
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 19.63,
|
| 1597 |
+
"learning_rate": 5.339920685094613e-06,
|
| 1598 |
+
"loss": 0.1214,
|
| 1599 |
+
"step": 530
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 19.7,
|
| 1603 |
+
"learning_rate": 5.27142644885001e-06,
|
| 1604 |
+
"loss": 0.1222,
|
| 1605 |
+
"step": 532
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 19.78,
|
| 1609 |
+
"learning_rate": 5.203216756872746e-06,
|
| 1610 |
+
"loss": 0.1205,
|
| 1611 |
+
"step": 534
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 19.85,
|
| 1615 |
+
"learning_rate": 5.135295713714998e-06,
|
| 1616 |
+
"loss": 0.1186,
|
| 1617 |
+
"step": 536
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 19.93,
|
| 1621 |
+
"learning_rate": 5.067667406559349e-06,
|
| 1622 |
+
"loss": 0.1257,
|
| 1623 |
+
"step": 538
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 20.0,
|
| 1627 |
+
"learning_rate": 5.000335904972846e-06,
|
| 1628 |
+
"loss": 0.1201,
|
| 1629 |
+
"step": 540
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 20.07,
|
| 1633 |
+
"learning_rate": 4.933305260662111e-06,
|
| 1634 |
+
"loss": 0.117,
|
| 1635 |
+
"step": 542
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 20.15,
|
| 1639 |
+
"learning_rate": 4.8665795072295455e-06,
|
| 1640 |
+
"loss": 0.1169,
|
| 1641 |
+
"step": 544
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 20.22,
|
| 1645 |
+
"learning_rate": 4.800162659930563e-06,
|
| 1646 |
+
"loss": 0.1121,
|
| 1647 |
+
"step": 546
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 20.3,
|
| 1651 |
+
"learning_rate": 4.734058715432019e-06,
|
| 1652 |
+
"loss": 0.1111,
|
| 1653 |
+
"step": 548
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 20.37,
|
| 1657 |
+
"learning_rate": 4.668271651571691e-06,
|
| 1658 |
+
"loss": 0.1177,
|
| 1659 |
+
"step": 550
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 20.44,
|
| 1663 |
+
"learning_rate": 4.602805427118878e-06,
|
| 1664 |
+
"loss": 0.1091,
|
| 1665 |
+
"step": 552
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 20.52,
|
| 1669 |
+
"learning_rate": 4.537663981536239e-06,
|
| 1670 |
+
"loss": 0.1134,
|
| 1671 |
+
"step": 554
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 20.59,
|
| 1675 |
+
"learning_rate": 4.4728512347426835e-06,
|
| 1676 |
+
"loss": 0.1226,
|
| 1677 |
+
"step": 556
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 20.67,
|
| 1681 |
+
"learning_rate": 4.408371086877516e-06,
|
| 1682 |
+
"loss": 0.1233,
|
| 1683 |
+
"step": 558
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 20.74,
|
| 1687 |
+
"learning_rate": 4.344227418065728e-06,
|
| 1688 |
+
"loss": 0.121,
|
| 1689 |
+
"step": 560
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 20.81,
|
| 1693 |
+
"learning_rate": 4.280424088184534e-06,
|
| 1694 |
+
"loss": 0.114,
|
| 1695 |
+
"step": 562
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 20.89,
|
| 1699 |
+
"learning_rate": 4.216964936631054e-06,
|
| 1700 |
+
"loss": 0.1245,
|
| 1701 |
+
"step": 564
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 20.96,
|
| 1705 |
+
"learning_rate": 4.153853782091335e-06,
|
| 1706 |
+
"loss": 0.1194,
|
| 1707 |
+
"step": 566
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 21.04,
|
| 1711 |
+
"learning_rate": 4.0910944223105155e-06,
|
| 1712 |
+
"loss": 0.1161,
|
| 1713 |
+
"step": 568
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 21.11,
|
| 1717 |
+
"learning_rate": 4.028690633864312e-06,
|
| 1718 |
+
"loss": 0.1107,
|
| 1719 |
+
"step": 570
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 21.19,
|
| 1723 |
+
"learning_rate": 3.966646171931754e-06,
|
| 1724 |
+
"loss": 0.1113,
|
| 1725 |
+
"step": 572
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 21.26,
|
| 1729 |
+
"learning_rate": 3.904964770069233e-06,
|
| 1730 |
+
"loss": 0.114,
|
| 1731 |
+
"step": 574
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 21.33,
|
| 1735 |
+
"learning_rate": 3.843650139985807e-06,
|
| 1736 |
+
"loss": 0.1113,
|
| 1737 |
+
"step": 576
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 21.41,
|
| 1741 |
+
"learning_rate": 3.7827059713198646e-06,
|
| 1742 |
+
"loss": 0.1151,
|
| 1743 |
+
"step": 578
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 21.48,
|
| 1747 |
+
"learning_rate": 3.7221359314170933e-06,
|
| 1748 |
+
"loss": 0.1147,
|
| 1749 |
+
"step": 580
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 21.56,
|
| 1753 |
+
"learning_rate": 3.661943665109796e-06,
|
| 1754 |
+
"loss": 0.1212,
|
| 1755 |
+
"step": 582
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 21.63,
|
| 1759 |
+
"learning_rate": 3.602132794497558e-06,
|
| 1760 |
+
"loss": 0.1211,
|
| 1761 |
+
"step": 584
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 21.7,
|
| 1765 |
+
"learning_rate": 3.5427069187293007e-06,
|
| 1766 |
+
"loss": 0.1139,
|
| 1767 |
+
"step": 586
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 21.78,
|
| 1771 |
+
"learning_rate": 3.4836696137866755e-06,
|
| 1772 |
+
"loss": 0.1138,
|
| 1773 |
+
"step": 588
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 21.85,
|
| 1777 |
+
"learning_rate": 3.425024432268894e-06,
|
| 1778 |
+
"loss": 0.1219,
|
| 1779 |
+
"step": 590
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 21.93,
|
| 1783 |
+
"learning_rate": 3.3667749031789554e-06,
|
| 1784 |
+
"loss": 0.1186,
|
| 1785 |
+
"step": 592
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 22.0,
|
| 1789 |
+
"learning_rate": 3.308924531711265e-06,
|
| 1790 |
+
"loss": 0.1188,
|
| 1791 |
+
"step": 594
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 22.07,
|
| 1795 |
+
"learning_rate": 3.251476799040725e-06,
|
| 1796 |
+
"loss": 0.1145,
|
| 1797 |
+
"step": 596
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 22.15,
|
| 1801 |
+
"learning_rate": 3.194435162113245e-06,
|
| 1802 |
+
"loss": 0.1152,
|
| 1803 |
+
"step": 598
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 22.22,
|
| 1807 |
+
"learning_rate": 3.137803053437725e-06,
|
| 1808 |
+
"loss": 0.1144,
|
| 1809 |
+
"step": 600
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 22.3,
|
| 1813 |
+
"learning_rate": 3.0815838808794907e-06,
|
| 1814 |
+
"loss": 0.1073,
|
| 1815 |
+
"step": 602
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 22.37,
|
| 1819 |
+
"learning_rate": 3.0257810274552444e-06,
|
| 1820 |
+
"loss": 0.1045,
|
| 1821 |
+
"step": 604
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 22.44,
|
| 1825 |
+
"learning_rate": 2.970397851129467e-06,
|
| 1826 |
+
"loss": 0.1181,
|
| 1827 |
+
"step": 606
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 22.52,
|
| 1831 |
+
"learning_rate": 2.9154376846123633e-06,
|
| 1832 |
+
"loss": 0.1146,
|
| 1833 |
+
"step": 608
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 22.59,
|
| 1837 |
+
"learning_rate": 2.86090383515931e-06,
|
| 1838 |
+
"loss": 0.1165,
|
| 1839 |
+
"step": 610
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 22.67,
|
| 1843 |
+
"learning_rate": 2.8067995843718544e-06,
|
| 1844 |
+
"loss": 0.1164,
|
| 1845 |
+
"step": 612
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 22.74,
|
| 1849 |
+
"learning_rate": 2.753128188000208e-06,
|
| 1850 |
+
"loss": 0.1131,
|
| 1851 |
+
"step": 614
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 22.81,
|
| 1855 |
+
"learning_rate": 2.6998928757473764e-06,
|
| 1856 |
+
"loss": 0.1138,
|
| 1857 |
+
"step": 616
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 22.89,
|
| 1861 |
+
"learning_rate": 2.64709685107477e-06,
|
| 1862 |
+
"loss": 0.1168,
|
| 1863 |
+
"step": 618
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 22.96,
|
| 1867 |
+
"learning_rate": 2.594743291009456e-06,
|
| 1868 |
+
"loss": 0.1178,
|
| 1869 |
+
"step": 620
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 23.04,
|
| 1873 |
+
"learning_rate": 2.54283534595297e-06,
|
| 1874 |
+
"loss": 0.1173,
|
| 1875 |
+
"step": 622
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 23.11,
|
| 1879 |
+
"learning_rate": 2.491376139491756e-06,
|
| 1880 |
+
"loss": 0.1119,
|
| 1881 |
+
"step": 624
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 23.19,
|
| 1885 |
+
"learning_rate": 2.440368768209167e-06,
|
| 1886 |
+
"loss": 0.1075,
|
| 1887 |
+
"step": 626
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 23.26,
|
| 1891 |
+
"learning_rate": 2.3898163014991694e-06,
|
| 1892 |
+
"loss": 0.1128,
|
| 1893 |
+
"step": 628
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 23.33,
|
| 1897 |
+
"learning_rate": 2.3397217813816107e-06,
|
| 1898 |
+
"loss": 0.1125,
|
| 1899 |
+
"step": 630
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 23.41,
|
| 1903 |
+
"learning_rate": 2.2900882223191766e-06,
|
| 1904 |
+
"loss": 0.118,
|
| 1905 |
+
"step": 632
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 23.48,
|
| 1909 |
+
"learning_rate": 2.240918611035988e-06,
|
| 1910 |
+
"loss": 0.107,
|
| 1911 |
+
"step": 634
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 23.56,
|
| 1915 |
+
"learning_rate": 2.1922159063378876e-06,
|
| 1916 |
+
"loss": 0.1201,
|
| 1917 |
+
"step": 636
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 23.63,
|
| 1921 |
+
"learning_rate": 2.1439830389343695e-06,
|
| 1922 |
+
"loss": 0.1083,
|
| 1923 |
+
"step": 638
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 23.7,
|
| 1927 |
+
"learning_rate": 2.0962229112622403e-06,
|
| 1928 |
+
"loss": 0.1152,
|
| 1929 |
+
"step": 640
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 23.78,
|
| 1933 |
+
"learning_rate": 2.0489383973109643e-06,
|
| 1934 |
+
"loss": 0.113,
|
| 1935 |
+
"step": 642
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 23.85,
|
| 1939 |
+
"learning_rate": 2.0021323424496962e-06,
|
| 1940 |
+
"loss": 0.1149,
|
| 1941 |
+
"step": 644
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 23.93,
|
| 1945 |
+
"learning_rate": 1.955807563256097e-06,
|
| 1946 |
+
"loss": 0.111,
|
| 1947 |
+
"step": 646
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 24.0,
|
| 1951 |
+
"learning_rate": 1.9099668473468113e-06,
|
| 1952 |
+
"loss": 0.1164,
|
| 1953 |
+
"step": 648
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 24.07,
|
| 1957 |
+
"learning_rate": 1.8646129532097391e-06,
|
| 1958 |
+
"loss": 0.1129,
|
| 1959 |
+
"step": 650
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 24.15,
|
| 1963 |
+
"learning_rate": 1.8197486100380357e-06,
|
| 1964 |
+
"loss": 0.1089,
|
| 1965 |
+
"step": 652
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 24.22,
|
| 1969 |
+
"learning_rate": 1.775376517565891e-06,
|
| 1970 |
+
"loss": 0.116,
|
| 1971 |
+
"step": 654
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 24.3,
|
| 1975 |
+
"learning_rate": 1.7314993459060492e-06,
|
| 1976 |
+
"loss": 0.1121,
|
| 1977 |
+
"step": 656
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 24.37,
|
| 1981 |
+
"learning_rate": 1.6881197353891643e-06,
|
| 1982 |
+
"loss": 0.1118,
|
| 1983 |
+
"step": 658
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 24.44,
|
| 1987 |
+
"learning_rate": 1.645240296404892e-06,
|
| 1988 |
+
"loss": 0.11,
|
| 1989 |
+
"step": 660
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 24.52,
|
| 1993 |
+
"learning_rate": 1.6028636092448168e-06,
|
| 1994 |
+
"loss": 0.1134,
|
| 1995 |
+
"step": 662
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 24.59,
|
| 1999 |
+
"learning_rate": 1.5609922239471842e-06,
|
| 2000 |
+
"loss": 0.1079,
|
| 2001 |
+
"step": 664
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 24.67,
|
| 2005 |
+
"learning_rate": 1.5196286601434508e-06,
|
| 2006 |
+
"loss": 0.1128,
|
| 2007 |
+
"step": 666
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 24.74,
|
| 2011 |
+
"learning_rate": 1.4787754069066574e-06,
|
| 2012 |
+
"loss": 0.1102,
|
| 2013 |
+
"step": 668
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 24.81,
|
| 2017 |
+
"learning_rate": 1.4384349226016559e-06,
|
| 2018 |
+
"loss": 0.1159,
|
| 2019 |
+
"step": 670
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 24.89,
|
| 2023 |
+
"learning_rate": 1.3986096347371692e-06,
|
| 2024 |
+
"loss": 0.1145,
|
| 2025 |
+
"step": 672
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 24.96,
|
| 2029 |
+
"learning_rate": 1.3593019398197216e-06,
|
| 2030 |
+
"loss": 0.1113,
|
| 2031 |
+
"step": 674
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 25.04,
|
| 2035 |
+
"learning_rate": 1.3205142032094188e-06,
|
| 2036 |
+
"loss": 0.1113,
|
| 2037 |
+
"step": 676
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 25.11,
|
| 2041 |
+
"learning_rate": 1.2822487589776234e-06,
|
| 2042 |
+
"loss": 0.1088,
|
| 2043 |
+
"step": 678
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 25.19,
|
| 2047 |
+
"learning_rate": 1.2445079097664914e-06,
|
| 2048 |
+
"loss": 0.1127,
|
| 2049 |
+
"step": 680
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 25.26,
|
| 2053 |
+
"learning_rate": 1.2072939266504068e-06,
|
| 2054 |
+
"loss": 0.1066,
|
| 2055 |
+
"step": 682
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 25.33,
|
| 2059 |
+
"learning_rate": 1.1706090489993348e-06,
|
| 2060 |
+
"loss": 0.1193,
|
| 2061 |
+
"step": 684
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 25.41,
|
| 2065 |
+
"learning_rate": 1.134455484344047e-06,
|
| 2066 |
+
"loss": 0.108,
|
| 2067 |
+
"step": 686
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 25.48,
|
| 2071 |
+
"learning_rate": 1.098835408243295e-06,
|
| 2072 |
+
"loss": 0.1083,
|
| 2073 |
+
"step": 688
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 25.56,
|
| 2077 |
+
"learning_rate": 1.0637509641528876e-06,
|
| 2078 |
+
"loss": 0.112,
|
| 2079 |
+
"step": 690
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 25.63,
|
| 2083 |
+
"learning_rate": 1.0292042632967158e-06,
|
| 2084 |
+
"loss": 0.1157,
|
| 2085 |
+
"step": 692
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 25.7,
|
| 2089 |
+
"learning_rate": 9.95197384539698e-07,
|
| 2090 |
+
"loss": 0.1128,
|
| 2091 |
+
"step": 694
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 25.78,
|
| 2095 |
+
"learning_rate": 9.61732374262696e-07,
|
| 2096 |
+
"loss": 0.1123,
|
| 2097 |
+
"step": 696
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 25.85,
|
| 2101 |
+
"learning_rate": 9.288112462393605e-07,
|
| 2102 |
+
"loss": 0.1111,
|
| 2103 |
+
"step": 698
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 25.93,
|
| 2107 |
+
"learning_rate": 8.964359815149593e-07,
|
| 2108 |
+
"loss": 0.1194,
|
| 2109 |
+
"step": 700
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 26.0,
|
| 2113 |
+
"learning_rate": 8.646085282871597e-07,
|
| 2114 |
+
"loss": 0.1103,
|
| 2115 |
+
"step": 702
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 26.07,
|
| 2119 |
+
"learning_rate": 8.33330801788812e-07,
|
| 2120 |
+
"loss": 0.113,
|
| 2121 |
+
"step": 704
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 26.15,
|
| 2125 |
+
"learning_rate": 8.026046841726687e-07,
|
| 2126 |
+
"loss": 0.1165,
|
| 2127 |
+
"step": 706
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 26.22,
|
| 2131 |
+
"learning_rate": 7.724320243981587e-07,
|
| 2132 |
+
"loss": 0.1231,
|
| 2133 |
+
"step": 708
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 26.3,
|
| 2137 |
+
"learning_rate": 7.42814638120104e-07,
|
| 2138 |
+
"loss": 0.1117,
|
| 2139 |
+
"step": 710
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 26.37,
|
| 2143 |
+
"learning_rate": 7.137543075794651e-07,
|
| 2144 |
+
"loss": 0.1102,
|
| 2145 |
+
"step": 712
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 26.44,
|
| 2149 |
+
"learning_rate": 6.852527814960986e-07,
|
| 2150 |
+
"loss": 0.1106,
|
| 2151 |
+
"step": 714
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 26.52,
|
| 2155 |
+
"learning_rate": 6.573117749635294e-07,
|
| 2156 |
+
"loss": 0.1136,
|
| 2157 |
+
"step": 716
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 26.59,
|
| 2161 |
+
"learning_rate": 6.299329693457279e-07,
|
| 2162 |
+
"loss": 0.1118,
|
| 2163 |
+
"step": 718
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 26.67,
|
| 2167 |
+
"learning_rate": 6.031180121759572e-07,
|
| 2168 |
+
"loss": 0.1072,
|
| 2169 |
+
"step": 720
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 26.74,
|
| 2173 |
+
"learning_rate": 5.768685170576105e-07,
|
| 2174 |
+
"loss": 0.1102,
|
| 2175 |
+
"step": 722
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 26.81,
|
| 2179 |
+
"learning_rate": 5.511860635671229e-07,
|
| 2180 |
+
"loss": 0.1089,
|
| 2181 |
+
"step": 724
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 26.89,
|
| 2185 |
+
"learning_rate": 5.260721971589211e-07,
|
| 2186 |
+
"loss": 0.1124,
|
| 2187 |
+
"step": 726
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 26.96,
|
| 2191 |
+
"learning_rate": 5.015284290724165e-07,
|
| 2192 |
+
"loss": 0.1043,
|
| 2193 |
+
"step": 728
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 27.04,
|
| 2197 |
+
"learning_rate": 4.775562362410724e-07,
|
| 2198 |
+
"loss": 0.1137,
|
| 2199 |
+
"step": 730
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 27.11,
|
| 2203 |
+
"learning_rate": 4.5415706120352287e-07,
|
| 2204 |
+
"loss": 0.1112,
|
| 2205 |
+
"step": 732
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 27.19,
|
| 2209 |
+
"learning_rate": 4.3133231201677894e-07,
|
| 2210 |
+
"loss": 0.1094,
|
| 2211 |
+
"step": 734
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 27.26,
|
| 2215 |
+
"learning_rate": 4.0908336217147957e-07,
|
| 2216 |
+
"loss": 0.1145,
|
| 2217 |
+
"step": 736
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 27.33,
|
| 2221 |
+
"learning_rate": 3.8741155050926237e-07,
|
| 2222 |
+
"loss": 0.1128,
|
| 2223 |
+
"step": 738
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 27.41,
|
| 2227 |
+
"learning_rate": 3.6631818114218007e-07,
|
| 2228 |
+
"loss": 0.1109,
|
| 2229 |
+
"step": 740
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 27.48,
|
| 2233 |
+
"learning_rate": 3.4580452337423465e-07,
|
| 2234 |
+
"loss": 0.1109,
|
| 2235 |
+
"step": 742
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 27.56,
|
| 2239 |
+
"learning_rate": 3.258718116249937e-07,
|
| 2240 |
+
"loss": 0.1139,
|
| 2241 |
+
"step": 744
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 27.63,
|
| 2245 |
+
"learning_rate": 3.0652124535531125e-07,
|
| 2246 |
+
"loss": 0.1066,
|
| 2247 |
+
"step": 746
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 27.7,
|
| 2251 |
+
"learning_rate": 2.8775398899514304e-07,
|
| 2252 |
+
"loss": 0.1104,
|
| 2253 |
+
"step": 748
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 27.78,
|
| 2257 |
+
"learning_rate": 2.695711718734839e-07,
|
| 2258 |
+
"loss": 0.1088,
|
| 2259 |
+
"step": 750
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 27.85,
|
| 2263 |
+
"learning_rate": 2.5197388815040634e-07,
|
| 2264 |
+
"loss": 0.1112,
|
| 2265 |
+
"step": 752
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 27.93,
|
| 2269 |
+
"learning_rate": 2.3496319675121892e-07,
|
| 2270 |
+
"loss": 0.1123,
|
| 2271 |
+
"step": 754
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 28.0,
|
| 2275 |
+
"learning_rate": 2.185401213027405e-07,
|
| 2276 |
+
"loss": 0.1143,
|
| 2277 |
+
"step": 756
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 28.07,
|
| 2281 |
+
"learning_rate": 2.0270565007171726e-07,
|
| 2282 |
+
"loss": 0.1103,
|
| 2283 |
+
"step": 758
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 28.15,
|
| 2287 |
+
"learning_rate": 1.8746073590533487e-07,
|
| 2288 |
+
"loss": 0.1096,
|
| 2289 |
+
"step": 760
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 28.22,
|
| 2293 |
+
"learning_rate": 1.7280629617389187e-07,
|
| 2294 |
+
"loss": 0.1085,
|
| 2295 |
+
"step": 762
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 28.3,
|
| 2299 |
+
"learning_rate": 1.5874321271559634e-07,
|
| 2300 |
+
"loss": 0.1052,
|
| 2301 |
+
"step": 764
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 28.37,
|
| 2305 |
+
"learning_rate": 1.4527233178349255e-07,
|
| 2306 |
+
"loss": 0.1165,
|
| 2307 |
+
"step": 766
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 28.44,
|
| 2311 |
+
"learning_rate": 1.323944639945507e-07,
|
| 2312 |
+
"loss": 0.1179,
|
| 2313 |
+
"step": 768
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 28.52,
|
| 2317 |
+
"learning_rate": 1.2011038428087263e-07,
|
| 2318 |
+
"loss": 0.1173,
|
| 2319 |
+
"step": 770
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 28.59,
|
| 2323 |
+
"learning_rate": 1.0842083184307239e-07,
|
| 2324 |
+
"loss": 0.1095,
|
| 2325 |
+
"step": 772
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 28.67,
|
| 2329 |
+
"learning_rate": 9.732651010578408e-08,
|
| 2330 |
+
"loss": 0.1135,
|
| 2331 |
+
"step": 774
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 28.74,
|
| 2335 |
+
"learning_rate": 8.682808667534348e-08,
|
| 2336 |
+
"loss": 0.1166,
|
| 2337 |
+
"step": 776
|
| 2338 |
+
},
|
| 2339 |
+
{
|
| 2340 |
+
"epoch": 28.81,
|
| 2341 |
+
"learning_rate": 7.692619329960571e-08,
|
| 2342 |
+
"loss": 0.1044,
|
| 2343 |
+
"step": 778
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 28.89,
|
| 2347 |
+
"learning_rate": 6.762142582993458e-08,
|
| 2348 |
+
"loss": 0.1179,
|
| 2349 |
+
"step": 780
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 28.96,
|
| 2353 |
+
"learning_rate": 5.8914344185344584e-08,
|
| 2354 |
+
"loss": 0.1084,
|
| 2355 |
+
"step": 782
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 29.04,
|
| 2359 |
+
"learning_rate": 5.0805472318806726e-08,
|
| 2360 |
+
"loss": 0.1074,
|
| 2361 |
+
"step": 784
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 29.11,
|
| 2365 |
+
"learning_rate": 4.329529818572265e-08,
|
| 2366 |
+
"loss": 0.1149,
|
| 2367 |
+
"step": 786
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 29.19,
|
| 2371 |
+
"learning_rate": 3.6384273714560324e-08,
|
| 2372 |
+
"loss": 0.1215,
|
| 2373 |
+
"step": 788
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 29.26,
|
| 2377 |
+
"learning_rate": 3.0072814779656956e-08,
|
| 2378 |
+
"loss": 0.1111,
|
| 2379 |
+
"step": 790
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 29.33,
|
| 2383 |
+
"learning_rate": 2.4361301176195617e-08,
|
| 2384 |
+
"loss": 0.1036,
|
| 2385 |
+
"step": 792
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 29.41,
|
| 2389 |
+
"learning_rate": 1.9250076597351343e-08,
|
| 2390 |
+
"loss": 0.1069,
|
| 2391 |
+
"step": 794
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 29.48,
|
| 2395 |
+
"learning_rate": 1.4739448613607655e-08,
|
| 2396 |
+
"loss": 0.1097,
|
| 2397 |
+
"step": 796
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 29.56,
|
| 2401 |
+
"learning_rate": 1.082968865424805e-08,
|
| 2402 |
+
"loss": 0.1123,
|
| 2403 |
+
"step": 798
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 29.63,
|
| 2407 |
+
"learning_rate": 7.521031991024608e-09,
|
| 2408 |
+
"loss": 0.1128,
|
| 2409 |
+
"step": 800
|
| 2410 |
+
}
|
| 2411 |
+
],
|
| 2412 |
+
"logging_steps": 2,
|
| 2413 |
+
"max_steps": 810,
|
| 2414 |
+
"num_input_tokens_seen": 0,
|
| 2415 |
+
"num_train_epochs": 30,
|
| 2416 |
+
"save_steps": 200,
|
| 2417 |
+
"total_flos": 1.2489159543555686e+17,
|
| 2418 |
+
"train_batch_size": 32,
|
| 2419 |
+
"trial_name": null,
|
| 2420 |
+
"trial_params": null
|
| 2421 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93dfa1077deca1c8244626c88c997e5aa915062db76196a9d7f4cd3b37fb56e0
|
| 3 |
+
size 4792
|