Upload 6 files
Browse files- README.md +202 -0
- adapter_config.json +34 -0
- adapter_model.safetensors +3 -0
- config.json +70 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +3030 -0
README.md
ADDED
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: liuhaotian/llava-v1.6-mistral-7b
|
3 |
+
library_name: peft
|
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 |
+
### Framework versions
|
201 |
+
|
202 |
+
- PEFT 0.12.0
|
adapter_config.json
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": null,
|
4 |
+
"base_model_name_or_path": "liuhaotian/llava-v1.6-mistral-7b",
|
5 |
+
"bias": "none",
|
6 |
+
"fan_in_fan_out": false,
|
7 |
+
"inference_mode": true,
|
8 |
+
"init_lora_weights": true,
|
9 |
+
"layer_replication": null,
|
10 |
+
"layers_pattern": null,
|
11 |
+
"layers_to_transform": null,
|
12 |
+
"loftq_config": {},
|
13 |
+
"lora_alpha": 32,
|
14 |
+
"lora_dropout": 0.05,
|
15 |
+
"megatron_config": null,
|
16 |
+
"megatron_core": "megatron.core",
|
17 |
+
"modules_to_save": null,
|
18 |
+
"peft_type": "LORA",
|
19 |
+
"r": 16,
|
20 |
+
"rank_pattern": {},
|
21 |
+
"revision": null,
|
22 |
+
"target_modules": [
|
23 |
+
"q_proj",
|
24 |
+
"gate_proj",
|
25 |
+
"k_proj",
|
26 |
+
"down_proj",
|
27 |
+
"v_proj",
|
28 |
+
"o_proj",
|
29 |
+
"up_proj"
|
30 |
+
],
|
31 |
+
"task_type": "CAUSAL_LM",
|
32 |
+
"use_dora": false,
|
33 |
+
"use_rslora": false
|
34 |
+
}
|
adapter_model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9c8e0bdd47b52a03db1653b8df35ad8ad5644499b73400eab0eda717ccfff4d8
|
3 |
+
size 88690480
|
config.json
ADDED
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "liuhaotian/llava-v1.6-mistral-7b",
|
3 |
+
"architectures": [
|
4 |
+
"LlavaMistralForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_dropout": 0.0,
|
7 |
+
"bos_token_id": 1,
|
8 |
+
"eos_token_id": 2,
|
9 |
+
"freeze_mm_mlp_adapter": false,
|
10 |
+
"freeze_mm_vision_resampler": false,
|
11 |
+
"hidden_act": "silu",
|
12 |
+
"hidden_size": 4096,
|
13 |
+
"image_aspect_ratio": "anyres",
|
14 |
+
"image_crop_resolution": 224,
|
15 |
+
"image_grid_pinpoints": [
|
16 |
+
[
|
17 |
+
336,
|
18 |
+
672
|
19 |
+
],
|
20 |
+
[
|
21 |
+
672,
|
22 |
+
336
|
23 |
+
],
|
24 |
+
[
|
25 |
+
672,
|
26 |
+
672
|
27 |
+
],
|
28 |
+
[
|
29 |
+
1008,
|
30 |
+
336
|
31 |
+
],
|
32 |
+
[
|
33 |
+
336,
|
34 |
+
1008
|
35 |
+
]
|
36 |
+
],
|
37 |
+
"image_split_resolution": 224,
|
38 |
+
"initializer_range": 0.02,
|
39 |
+
"intermediate_size": 14336,
|
40 |
+
"max_position_embeddings": 32768,
|
41 |
+
"mm_hidden_size": 1024,
|
42 |
+
"mm_patch_merge_type": "spatial_unpad",
|
43 |
+
"mm_projector_lr": 2e-05,
|
44 |
+
"mm_projector_type": "mlp2x_gelu",
|
45 |
+
"mm_resampler_type": null,
|
46 |
+
"mm_use_im_patch_token": false,
|
47 |
+
"mm_use_im_start_end": false,
|
48 |
+
"mm_vision_select_feature": "patch",
|
49 |
+
"mm_vision_select_layer": -2,
|
50 |
+
"mm_vision_tower": "openai/clip-vit-large-patch14-336",
|
51 |
+
"mm_vision_tower_lr": 2e-06,
|
52 |
+
"model_type": "llava_mistral",
|
53 |
+
"num_attention_heads": 32,
|
54 |
+
"num_hidden_layers": 32,
|
55 |
+
"num_key_value_heads": 8,
|
56 |
+
"rms_norm_eps": 1e-05,
|
57 |
+
"rope_theta": 1000000.0,
|
58 |
+
"sliding_window": null,
|
59 |
+
"tie_word_embeddings": false,
|
60 |
+
"tokenizer_model_max_length": 4096,
|
61 |
+
"tokenizer_padding_side": "right",
|
62 |
+
"torch_dtype": "bfloat16",
|
63 |
+
"transformers_version": "4.37.2",
|
64 |
+
"tune_mm_mlp_adapter": false,
|
65 |
+
"tune_mm_vision_resampler": false,
|
66 |
+
"unfreeze_mm_vision_tower": true,
|
67 |
+
"use_cache": true,
|
68 |
+
"use_mm_proj": true,
|
69 |
+
"vocab_size": 32000
|
70 |
+
}
|
non_lora_trainables.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a4378854ebd21c1d8a0f4b978b9c126ae12b9e0395cf9c9b85dbb343946d36e3
|
3 |
+
size 41961648
|
trainer_state.json
ADDED
@@ -0,0 +1,3030 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 1.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 500,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.0,
|
13 |
+
"learning_rate": 8.000000000000001e-07,
|
14 |
+
"loss": 0.8959,
|
15 |
+
"step": 1
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.0,
|
19 |
+
"learning_rate": 1.6000000000000001e-06,
|
20 |
+
"loss": 0.8883,
|
21 |
+
"step": 2
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.01,
|
25 |
+
"learning_rate": 2.4000000000000003e-06,
|
26 |
+
"loss": 0.8056,
|
27 |
+
"step": 3
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.01,
|
31 |
+
"learning_rate": 3.2000000000000003e-06,
|
32 |
+
"loss": 0.5905,
|
33 |
+
"step": 4
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.01,
|
37 |
+
"learning_rate": 4.000000000000001e-06,
|
38 |
+
"loss": 0.7238,
|
39 |
+
"step": 5
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.01,
|
43 |
+
"learning_rate": 4.800000000000001e-06,
|
44 |
+
"loss": 0.8706,
|
45 |
+
"step": 6
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.01,
|
49 |
+
"learning_rate": 5.600000000000001e-06,
|
50 |
+
"loss": 0.7291,
|
51 |
+
"step": 7
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.02,
|
55 |
+
"learning_rate": 6.4000000000000006e-06,
|
56 |
+
"loss": 0.8379,
|
57 |
+
"step": 8
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.02,
|
61 |
+
"learning_rate": 7.2000000000000005e-06,
|
62 |
+
"loss": 0.7678,
|
63 |
+
"step": 9
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.02,
|
67 |
+
"learning_rate": 8.000000000000001e-06,
|
68 |
+
"loss": 0.5016,
|
69 |
+
"step": 10
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.02,
|
73 |
+
"learning_rate": 8.8e-06,
|
74 |
+
"loss": 0.5994,
|
75 |
+
"step": 11
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 0.02,
|
79 |
+
"learning_rate": 9.600000000000001e-06,
|
80 |
+
"loss": 0.6028,
|
81 |
+
"step": 12
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 0.03,
|
85 |
+
"learning_rate": 1.04e-05,
|
86 |
+
"loss": 0.6321,
|
87 |
+
"step": 13
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.03,
|
91 |
+
"learning_rate": 1.1200000000000001e-05,
|
92 |
+
"loss": 0.3967,
|
93 |
+
"step": 14
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.03,
|
97 |
+
"learning_rate": 1.2e-05,
|
98 |
+
"loss": 0.4705,
|
99 |
+
"step": 15
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.03,
|
103 |
+
"learning_rate": 1.2800000000000001e-05,
|
104 |
+
"loss": 0.5506,
|
105 |
+
"step": 16
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 0.03,
|
109 |
+
"learning_rate": 1.3600000000000002e-05,
|
110 |
+
"loss": 0.3876,
|
111 |
+
"step": 17
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 0.04,
|
115 |
+
"learning_rate": 1.4400000000000001e-05,
|
116 |
+
"loss": 0.4482,
|
117 |
+
"step": 18
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 0.04,
|
121 |
+
"learning_rate": 1.5200000000000002e-05,
|
122 |
+
"loss": 0.5656,
|
123 |
+
"step": 19
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 0.04,
|
127 |
+
"learning_rate": 1.6000000000000003e-05,
|
128 |
+
"loss": 0.7643,
|
129 |
+
"step": 20
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.04,
|
133 |
+
"learning_rate": 1.6800000000000002e-05,
|
134 |
+
"loss": 0.3924,
|
135 |
+
"step": 21
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.04,
|
139 |
+
"learning_rate": 1.76e-05,
|
140 |
+
"loss": 0.4228,
|
141 |
+
"step": 22
|
142 |
+
},
|
143 |
+
{
|
144 |
+
"epoch": 0.05,
|
145 |
+
"learning_rate": 1.8400000000000003e-05,
|
146 |
+
"loss": 0.3103,
|
147 |
+
"step": 23
|
148 |
+
},
|
149 |
+
{
|
150 |
+
"epoch": 0.05,
|
151 |
+
"learning_rate": 1.9200000000000003e-05,
|
152 |
+
"loss": 0.5515,
|
153 |
+
"step": 24
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 0.05,
|
157 |
+
"learning_rate": 2e-05,
|
158 |
+
"loss": 0.374,
|
159 |
+
"step": 25
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 0.05,
|
163 |
+
"learning_rate": 1.9999781283802247e-05,
|
164 |
+
"loss": 0.174,
|
165 |
+
"step": 26
|
166 |
+
},
|
167 |
+
{
|
168 |
+
"epoch": 0.05,
|
169 |
+
"learning_rate": 1.999912514477634e-05,
|
170 |
+
"loss": 0.2516,
|
171 |
+
"step": 27
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.06,
|
175 |
+
"learning_rate": 1.999803161162393e-05,
|
176 |
+
"loss": 0.4822,
|
177 |
+
"step": 28
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.06,
|
181 |
+
"learning_rate": 1.9996500732179695e-05,
|
182 |
+
"loss": 0.2868,
|
183 |
+
"step": 29
|
184 |
+
},
|
185 |
+
{
|
186 |
+
"epoch": 0.06,
|
187 |
+
"learning_rate": 1.999453257340926e-05,
|
188 |
+
"loss": 0.3294,
|
189 |
+
"step": 30
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 0.06,
|
193 |
+
"learning_rate": 1.9992127221406276e-05,
|
194 |
+
"loss": 0.4017,
|
195 |
+
"step": 31
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 0.06,
|
199 |
+
"learning_rate": 1.9989284781388617e-05,
|
200 |
+
"loss": 0.3511,
|
201 |
+
"step": 32
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 0.07,
|
205 |
+
"learning_rate": 1.9986005377693826e-05,
|
206 |
+
"loss": 0.3472,
|
207 |
+
"step": 33
|
208 |
+
},
|
209 |
+
{
|
210 |
+
"epoch": 0.07,
|
211 |
+
"learning_rate": 1.9982289153773648e-05,
|
212 |
+
"loss": 0.145,
|
213 |
+
"step": 34
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.07,
|
217 |
+
"learning_rate": 1.9978136272187745e-05,
|
218 |
+
"loss": 0.2914,
|
219 |
+
"step": 35
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.07,
|
223 |
+
"learning_rate": 1.9973546914596622e-05,
|
224 |
+
"loss": 0.403,
|
225 |
+
"step": 36
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 0.07,
|
229 |
+
"learning_rate": 1.9968521281753642e-05,
|
230 |
+
"loss": 0.2962,
|
231 |
+
"step": 37
|
232 |
+
},
|
233 |
+
{
|
234 |
+
"epoch": 0.08,
|
235 |
+
"learning_rate": 1.996305959349627e-05,
|
236 |
+
"loss": 0.3813,
|
237 |
+
"step": 38
|
238 |
+
},
|
239 |
+
{
|
240 |
+
"epoch": 0.08,
|
241 |
+
"learning_rate": 1.995716208873644e-05,
|
242 |
+
"loss": 0.2893,
|
243 |
+
"step": 39
|
244 |
+
},
|
245 |
+
{
|
246 |
+
"epoch": 0.08,
|
247 |
+
"learning_rate": 1.9950829025450116e-05,
|
248 |
+
"loss": 0.317,
|
249 |
+
"step": 40
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 0.08,
|
253 |
+
"learning_rate": 1.9944060680666e-05,
|
254 |
+
"loss": 0.2615,
|
255 |
+
"step": 41
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.08,
|
259 |
+
"learning_rate": 1.993685735045343e-05,
|
260 |
+
"loss": 0.3313,
|
261 |
+
"step": 42
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.09,
|
265 |
+
"learning_rate": 1.9929219349909393e-05,
|
266 |
+
"loss": 0.258,
|
267 |
+
"step": 43
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 0.09,
|
271 |
+
"learning_rate": 1.9921147013144782e-05,
|
272 |
+
"loss": 0.213,
|
273 |
+
"step": 44
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 0.09,
|
277 |
+
"learning_rate": 1.9912640693269754e-05,
|
278 |
+
"loss": 0.3592,
|
279 |
+
"step": 45
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 0.09,
|
283 |
+
"learning_rate": 1.9903700762378303e-05,
|
284 |
+
"loss": 0.2213,
|
285 |
+
"step": 46
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 0.09,
|
289 |
+
"learning_rate": 1.989432761153196e-05,
|
290 |
+
"loss": 0.3759,
|
291 |
+
"step": 47
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 0.1,
|
295 |
+
"learning_rate": 1.9884521650742718e-05,
|
296 |
+
"loss": 0.1658,
|
297 |
+
"step": 48
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.1,
|
301 |
+
"learning_rate": 1.9874283308955058e-05,
|
302 |
+
"loss": 0.1732,
|
303 |
+
"step": 49
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 0.1,
|
307 |
+
"learning_rate": 1.9863613034027224e-05,
|
308 |
+
"loss": 0.2279,
|
309 |
+
"step": 50
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 0.1,
|
313 |
+
"learning_rate": 1.985251129271161e-05,
|
314 |
+
"loss": 0.1859,
|
315 |
+
"step": 51
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 0.1,
|
319 |
+
"learning_rate": 1.9840978570634338e-05,
|
320 |
+
"loss": 0.5561,
|
321 |
+
"step": 52
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 0.11,
|
325 |
+
"learning_rate": 1.982901537227404e-05,
|
326 |
+
"loss": 0.1094,
|
327 |
+
"step": 53
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 0.11,
|
331 |
+
"learning_rate": 1.9816622220939762e-05,
|
332 |
+
"loss": 0.16,
|
333 |
+
"step": 54
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 0.11,
|
337 |
+
"learning_rate": 1.9803799658748096e-05,
|
338 |
+
"loss": 0.2063,
|
339 |
+
"step": 55
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.11,
|
343 |
+
"learning_rate": 1.9790548246599447e-05,
|
344 |
+
"loss": 0.2317,
|
345 |
+
"step": 56
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 0.11,
|
349 |
+
"learning_rate": 1.9776868564153517e-05,
|
350 |
+
"loss": 0.2567,
|
351 |
+
"step": 57
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 0.12,
|
355 |
+
"learning_rate": 1.976276120980393e-05,
|
356 |
+
"loss": 0.1225,
|
357 |
+
"step": 58
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"epoch": 0.12,
|
361 |
+
"learning_rate": 1.9748226800652062e-05,
|
362 |
+
"loss": 0.1664,
|
363 |
+
"step": 59
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 0.12,
|
367 |
+
"learning_rate": 1.973326597248006e-05,
|
368 |
+
"loss": 0.2462,
|
369 |
+
"step": 60
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 0.12,
|
373 |
+
"learning_rate": 1.9717879379723012e-05,
|
374 |
+
"loss": 0.1357,
|
375 |
+
"step": 61
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 0.12,
|
379 |
+
"learning_rate": 1.9702067695440333e-05,
|
380 |
+
"loss": 0.2895,
|
381 |
+
"step": 62
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 0.13,
|
385 |
+
"learning_rate": 1.9685831611286312e-05,
|
386 |
+
"loss": 0.1432,
|
387 |
+
"step": 63
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 0.13,
|
391 |
+
"learning_rate": 1.966917183747987e-05,
|
392 |
+
"loss": 0.9037,
|
393 |
+
"step": 64
|
394 |
+
},
|
395 |
+
{
|
396 |
+
"epoch": 0.13,
|
397 |
+
"learning_rate": 1.9652089102773487e-05,
|
398 |
+
"loss": 0.1122,
|
399 |
+
"step": 65
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 0.13,
|
403 |
+
"learning_rate": 1.9634584154421316e-05,
|
404 |
+
"loss": 0.3454,
|
405 |
+
"step": 66
|
406 |
+
},
|
407 |
+
{
|
408 |
+
"epoch": 0.13,
|
409 |
+
"learning_rate": 1.9616657758146503e-05,
|
410 |
+
"loss": 0.2993,
|
411 |
+
"step": 67
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 0.14,
|
415 |
+
"learning_rate": 1.95983106981077e-05,
|
416 |
+
"loss": 0.294,
|
417 |
+
"step": 68
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 0.14,
|
421 |
+
"learning_rate": 1.957954377686475e-05,
|
422 |
+
"loss": 0.2309,
|
423 |
+
"step": 69
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 0.14,
|
427 |
+
"learning_rate": 1.9560357815343577e-05,
|
428 |
+
"loss": 0.4046,
|
429 |
+
"step": 70
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.14,
|
433 |
+
"learning_rate": 1.95407536528003e-05,
|
434 |
+
"loss": 0.5102,
|
435 |
+
"step": 71
|
436 |
+
},
|
437 |
+
{
|
438 |
+
"epoch": 0.14,
|
439 |
+
"learning_rate": 1.95407536528003e-05,
|
440 |
+
"loss": 0.302,
|
441 |
+
"step": 72
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 0.15,
|
445 |
+
"learning_rate": 1.9520732146784493e-05,
|
446 |
+
"loss": 0.1975,
|
447 |
+
"step": 73
|
448 |
+
},
|
449 |
+
{
|
450 |
+
"epoch": 0.15,
|
451 |
+
"learning_rate": 1.9500294173101687e-05,
|
452 |
+
"loss": 0.5097,
|
453 |
+
"step": 74
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 0.15,
|
457 |
+
"learning_rate": 1.947944062577507e-05,
|
458 |
+
"loss": 0.1714,
|
459 |
+
"step": 75
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 0.15,
|
463 |
+
"learning_rate": 1.9458172417006347e-05,
|
464 |
+
"loss": 0.3065,
|
465 |
+
"step": 76
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 0.15,
|
469 |
+
"learning_rate": 1.9436490477135877e-05,
|
470 |
+
"loss": 0.3434,
|
471 |
+
"step": 77
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 0.16,
|
475 |
+
"learning_rate": 1.941439575460195e-05,
|
476 |
+
"loss": 0.2067,
|
477 |
+
"step": 78
|
478 |
+
},
|
479 |
+
{
|
480 |
+
"epoch": 0.16,
|
481 |
+
"learning_rate": 1.93918892158993e-05,
|
482 |
+
"loss": 0.1366,
|
483 |
+
"step": 79
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 0.16,
|
487 |
+
"learning_rate": 1.9368971845536844e-05,
|
488 |
+
"loss": 0.2233,
|
489 |
+
"step": 80
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 0.16,
|
493 |
+
"learning_rate": 1.934564464599461e-05,
|
494 |
+
"loss": 0.1644,
|
495 |
+
"step": 81
|
496 |
+
},
|
497 |
+
{
|
498 |
+
"epoch": 0.16,
|
499 |
+
"learning_rate": 1.9321908637679868e-05,
|
500 |
+
"loss": 0.3666,
|
501 |
+
"step": 82
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 0.17,
|
505 |
+
"learning_rate": 1.9297764858882516e-05,
|
506 |
+
"loss": 0.184,
|
507 |
+
"step": 83
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 0.17,
|
511 |
+
"learning_rate": 1.9273214365729655e-05,
|
512 |
+
"loss": 0.6584,
|
513 |
+
"step": 84
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 0.17,
|
517 |
+
"learning_rate": 1.924825823213939e-05,
|
518 |
+
"loss": 0.2562,
|
519 |
+
"step": 85
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 0.17,
|
523 |
+
"learning_rate": 1.922289754977385e-05,
|
524 |
+
"loss": 0.0571,
|
525 |
+
"step": 86
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 0.17,
|
529 |
+
"learning_rate": 1.9197133427991437e-05,
|
530 |
+
"loss": 0.1578,
|
531 |
+
"step": 87
|
532 |
+
},
|
533 |
+
{
|
534 |
+
"epoch": 0.18,
|
535 |
+
"learning_rate": 1.9170966993798302e-05,
|
536 |
+
"loss": 0.1214,
|
537 |
+
"step": 88
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 0.18,
|
541 |
+
"learning_rate": 1.9144399391799043e-05,
|
542 |
+
"loss": 0.1772,
|
543 |
+
"step": 89
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 0.18,
|
547 |
+
"learning_rate": 1.911743178414665e-05,
|
548 |
+
"loss": 0.249,
|
549 |
+
"step": 90
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.18,
|
553 |
+
"learning_rate": 1.909006535049163e-05,
|
554 |
+
"loss": 0.1092,
|
555 |
+
"step": 91
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 0.18,
|
559 |
+
"learning_rate": 1.9062301287930448e-05,
|
560 |
+
"loss": 0.1524,
|
561 |
+
"step": 92
|
562 |
+
},
|
563 |
+
{
|
564 |
+
"epoch": 0.19,
|
565 |
+
"learning_rate": 1.903414081095315e-05,
|
566 |
+
"loss": 0.1331,
|
567 |
+
"step": 93
|
568 |
+
},
|
569 |
+
{
|
570 |
+
"epoch": 0.19,
|
571 |
+
"learning_rate": 1.9005585151390224e-05,
|
572 |
+
"loss": 0.4028,
|
573 |
+
"step": 94
|
574 |
+
},
|
575 |
+
{
|
576 |
+
"epoch": 0.19,
|
577 |
+
"learning_rate": 1.897663555835872e-05,
|
578 |
+
"loss": 0.1218,
|
579 |
+
"step": 95
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 0.19,
|
583 |
+
"learning_rate": 1.8947293298207637e-05,
|
584 |
+
"loss": 0.1068,
|
585 |
+
"step": 96
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 0.19,
|
589 |
+
"learning_rate": 1.8917559654462474e-05,
|
590 |
+
"loss": 0.1773,
|
591 |
+
"step": 97
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 0.2,
|
595 |
+
"learning_rate": 1.8887435927769137e-05,
|
596 |
+
"loss": 0.1191,
|
597 |
+
"step": 98
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 0.2,
|
601 |
+
"learning_rate": 1.8856923435837024e-05,
|
602 |
+
"loss": 0.5611,
|
603 |
+
"step": 99
|
604 |
+
},
|
605 |
+
{
|
606 |
+
"epoch": 0.2,
|
607 |
+
"learning_rate": 1.8826023513381372e-05,
|
608 |
+
"loss": 0.1207,
|
609 |
+
"step": 100
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 0.2,
|
613 |
+
"learning_rate": 1.879473751206489e-05,
|
614 |
+
"loss": 0.1215,
|
615 |
+
"step": 101
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 0.2,
|
619 |
+
"learning_rate": 1.8763066800438638e-05,
|
620 |
+
"loss": 0.2272,
|
621 |
+
"step": 102
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 0.21,
|
625 |
+
"learning_rate": 1.8731012763882132e-05,
|
626 |
+
"loss": 0.2555,
|
627 |
+
"step": 103
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 0.21,
|
631 |
+
"learning_rate": 1.8698576804542775e-05,
|
632 |
+
"loss": 0.0687,
|
633 |
+
"step": 104
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 0.21,
|
637 |
+
"learning_rate": 1.8665760341274505e-05,
|
638 |
+
"loss": 0.2113,
|
639 |
+
"step": 105
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.21,
|
643 |
+
"learning_rate": 1.863256480957574e-05,
|
644 |
+
"loss": 0.2414,
|
645 |
+
"step": 106
|
646 |
+
},
|
647 |
+
{
|
648 |
+
"epoch": 0.21,
|
649 |
+
"learning_rate": 1.859899166152657e-05,
|
650 |
+
"loss": 0.2,
|
651 |
+
"step": 107
|
652 |
+
},
|
653 |
+
{
|
654 |
+
"epoch": 0.22,
|
655 |
+
"learning_rate": 1.856504236572526e-05,
|
656 |
+
"loss": 0.2686,
|
657 |
+
"step": 108
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 0.22,
|
661 |
+
"learning_rate": 1.8530718407223976e-05,
|
662 |
+
"loss": 0.1561,
|
663 |
+
"step": 109
|
664 |
+
},
|
665 |
+
{
|
666 |
+
"epoch": 0.22,
|
667 |
+
"learning_rate": 1.849602128746387e-05,
|
668 |
+
"loss": 0.1295,
|
669 |
+
"step": 110
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 0.22,
|
673 |
+
"learning_rate": 1.8460952524209355e-05,
|
674 |
+
"loss": 0.1129,
|
675 |
+
"step": 111
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 0.22,
|
679 |
+
"learning_rate": 1.8425513651481748e-05,
|
680 |
+
"loss": 0.1517,
|
681 |
+
"step": 112
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 0.23,
|
685 |
+
"learning_rate": 1.8389706219492147e-05,
|
686 |
+
"loss": 0.2548,
|
687 |
+
"step": 113
|
688 |
+
},
|
689 |
+
{
|
690 |
+
"epoch": 0.23,
|
691 |
+
"learning_rate": 1.8353531794573623e-05,
|
692 |
+
"loss": 0.1433,
|
693 |
+
"step": 114
|
694 |
+
},
|
695 |
+
{
|
696 |
+
"epoch": 0.23,
|
697 |
+
"learning_rate": 1.831699195911272e-05,
|
698 |
+
"loss": 0.3476,
|
699 |
+
"step": 115
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 0.23,
|
703 |
+
"learning_rate": 1.8280088311480203e-05,
|
704 |
+
"loss": 0.3045,
|
705 |
+
"step": 116
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"epoch": 0.23,
|
709 |
+
"learning_rate": 1.8242822465961177e-05,
|
710 |
+
"loss": 0.3731,
|
711 |
+
"step": 117
|
712 |
+
},
|
713 |
+
{
|
714 |
+
"epoch": 0.24,
|
715 |
+
"learning_rate": 1.8205196052684445e-05,
|
716 |
+
"loss": 0.0941,
|
717 |
+
"step": 118
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 0.24,
|
721 |
+
"learning_rate": 1.8167210717551224e-05,
|
722 |
+
"loss": 0.2446,
|
723 |
+
"step": 119
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 0.24,
|
727 |
+
"learning_rate": 1.8128868122163125e-05,
|
728 |
+
"loss": 0.4302,
|
729 |
+
"step": 120
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 0.24,
|
733 |
+
"learning_rate": 1.8090169943749477e-05,
|
734 |
+
"loss": 0.1842,
|
735 |
+
"step": 121
|
736 |
+
},
|
737 |
+
{
|
738 |
+
"epoch": 0.24,
|
739 |
+
"learning_rate": 1.8051117875093974e-05,
|
740 |
+
"loss": 0.2067,
|
741 |
+
"step": 122
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 0.25,
|
745 |
+
"learning_rate": 1.8011713624460608e-05,
|
746 |
+
"loss": 0.3245,
|
747 |
+
"step": 123
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"epoch": 0.25,
|
751 |
+
"learning_rate": 1.797195891551896e-05,
|
752 |
+
"loss": 0.215,
|
753 |
+
"step": 124
|
754 |
+
},
|
755 |
+
{
|
756 |
+
"epoch": 0.25,
|
757 |
+
"learning_rate": 1.793185548726878e-05,
|
758 |
+
"loss": 0.1254,
|
759 |
+
"step": 125
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.25,
|
763 |
+
"learning_rate": 1.789140509396394e-05,
|
764 |
+
"loss": 0.2598,
|
765 |
+
"step": 126
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 0.25,
|
769 |
+
"learning_rate": 1.785060950503568e-05,
|
770 |
+
"loss": 0.1237,
|
771 |
+
"step": 127
|
772 |
+
},
|
773 |
+
{
|
774 |
+
"epoch": 0.26,
|
775 |
+
"learning_rate": 1.780947050501522e-05,
|
776 |
+
"loss": 0.1599,
|
777 |
+
"step": 128
|
778 |
+
},
|
779 |
+
{
|
780 |
+
"epoch": 0.26,
|
781 |
+
"learning_rate": 1.7767989893455696e-05,
|
782 |
+
"loss": 0.3742,
|
783 |
+
"step": 129
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 0.26,
|
787 |
+
"learning_rate": 1.7726169484853438e-05,
|
788 |
+
"loss": 0.2924,
|
789 |
+
"step": 130
|
790 |
+
},
|
791 |
+
{
|
792 |
+
"epoch": 0.26,
|
793 |
+
"learning_rate": 1.7684011108568593e-05,
|
794 |
+
"loss": 0.1479,
|
795 |
+
"step": 131
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 0.26,
|
799 |
+
"learning_rate": 1.7641516608745114e-05,
|
800 |
+
"loss": 0.1032,
|
801 |
+
"step": 132
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 0.27,
|
805 |
+
"learning_rate": 1.759868784423009e-05,
|
806 |
+
"loss": 0.3311,
|
807 |
+
"step": 133
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 0.27,
|
811 |
+
"learning_rate": 1.7555526688492418e-05,
|
812 |
+
"loss": 0.1297,
|
813 |
+
"step": 134
|
814 |
+
},
|
815 |
+
{
|
816 |
+
"epoch": 0.27,
|
817 |
+
"learning_rate": 1.7512035029540887e-05,
|
818 |
+
"loss": 0.2778,
|
819 |
+
"step": 135
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 0.27,
|
823 |
+
"learning_rate": 1.7468214769841542e-05,
|
824 |
+
"loss": 0.2048,
|
825 |
+
"step": 136
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 0.27,
|
829 |
+
"learning_rate": 1.74240678262345e-05,
|
830 |
+
"loss": 0.2699,
|
831 |
+
"step": 137
|
832 |
+
},
|
833 |
+
{
|
834 |
+
"epoch": 0.28,
|
835 |
+
"learning_rate": 1.7379596129850098e-05,
|
836 |
+
"loss": 0.3385,
|
837 |
+
"step": 138
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 0.28,
|
841 |
+
"learning_rate": 1.73348016260244e-05,
|
842 |
+
"loss": 0.281,
|
843 |
+
"step": 139
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 0.28,
|
847 |
+
"learning_rate": 1.7289686274214116e-05,
|
848 |
+
"loss": 0.1776,
|
849 |
+
"step": 140
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.28,
|
853 |
+
"learning_rate": 1.7244252047910893e-05,
|
854 |
+
"loss": 0.1584,
|
855 |
+
"step": 141
|
856 |
+
},
|
857 |
+
{
|
858 |
+
"epoch": 0.28,
|
859 |
+
"learning_rate": 1.7198500934554966e-05,
|
860 |
+
"loss": 0.1673,
|
861 |
+
"step": 142
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 0.29,
|
865 |
+
"learning_rate": 1.7152434935448257e-05,
|
866 |
+
"loss": 0.1588,
|
867 |
+
"step": 143
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 0.29,
|
871 |
+
"learning_rate": 1.7106056065666793e-05,
|
872 |
+
"loss": 0.3462,
|
873 |
+
"step": 144
|
874 |
+
},
|
875 |
+
{
|
876 |
+
"epoch": 0.29,
|
877 |
+
"learning_rate": 1.705936635397259e-05,
|
878 |
+
"loss": 0.2498,
|
879 |
+
"step": 145
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 0.29,
|
883 |
+
"learning_rate": 1.7012367842724887e-05,
|
884 |
+
"loss": 0.3301,
|
885 |
+
"step": 146
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 0.29,
|
889 |
+
"learning_rate": 1.6965062587790823e-05,
|
890 |
+
"loss": 0.2263,
|
891 |
+
"step": 147
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 0.3,
|
895 |
+
"learning_rate": 1.6917452658455496e-05,
|
896 |
+
"loss": 0.1131,
|
897 |
+
"step": 148
|
898 |
+
},
|
899 |
+
{
|
900 |
+
"epoch": 0.3,
|
901 |
+
"learning_rate": 1.6869540137331445e-05,
|
902 |
+
"loss": 0.2109,
|
903 |
+
"step": 149
|
904 |
+
},
|
905 |
+
{
|
906 |
+
"epoch": 0.3,
|
907 |
+
"learning_rate": 1.6821327120267567e-05,
|
908 |
+
"loss": 0.1655,
|
909 |
+
"step": 150
|
910 |
+
},
|
911 |
+
{
|
912 |
+
"epoch": 0.3,
|
913 |
+
"learning_rate": 1.6772815716257414e-05,
|
914 |
+
"loss": 0.3201,
|
915 |
+
"step": 151
|
916 |
+
},
|
917 |
+
{
|
918 |
+
"epoch": 0.3,
|
919 |
+
"learning_rate": 1.6724008047346946e-05,
|
920 |
+
"loss": 0.1607,
|
921 |
+
"step": 152
|
922 |
+
},
|
923 |
+
{
|
924 |
+
"epoch": 0.31,
|
925 |
+
"learning_rate": 1.667490624854173e-05,
|
926 |
+
"loss": 0.2095,
|
927 |
+
"step": 153
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 0.31,
|
931 |
+
"learning_rate": 1.66255124677135e-05,
|
932 |
+
"loss": 0.5483,
|
933 |
+
"step": 154
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 0.31,
|
937 |
+
"learning_rate": 1.6575828865506246e-05,
|
938 |
+
"loss": 0.18,
|
939 |
+
"step": 155
|
940 |
+
},
|
941 |
+
{
|
942 |
+
"epoch": 0.31,
|
943 |
+
"learning_rate": 1.6525857615241686e-05,
|
944 |
+
"loss": 0.0545,
|
945 |
+
"step": 156
|
946 |
+
},
|
947 |
+
{
|
948 |
+
"epoch": 0.31,
|
949 |
+
"learning_rate": 1.647560090282419e-05,
|
950 |
+
"loss": 0.3574,
|
951 |
+
"step": 157
|
952 |
+
},
|
953 |
+
{
|
954 |
+
"epoch": 0.32,
|
955 |
+
"learning_rate": 1.6425060926645168e-05,
|
956 |
+
"loss": 0.3601,
|
957 |
+
"step": 158
|
958 |
+
},
|
959 |
+
{
|
960 |
+
"epoch": 0.32,
|
961 |
+
"learning_rate": 1.63742398974869e-05,
|
962 |
+
"loss": 0.278,
|
963 |
+
"step": 159
|
964 |
+
},
|
965 |
+
{
|
966 |
+
"epoch": 0.32,
|
967 |
+
"learning_rate": 1.6323140038425842e-05,
|
968 |
+
"loss": 0.3027,
|
969 |
+
"step": 160
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 0.32,
|
973 |
+
"learning_rate": 1.6271763584735373e-05,
|
974 |
+
"loss": 0.4648,
|
975 |
+
"step": 161
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 0.32,
|
979 |
+
"learning_rate": 1.622011278378801e-05,
|
980 |
+
"loss": 0.0562,
|
981 |
+
"step": 162
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 0.33,
|
985 |
+
"learning_rate": 1.616818989495711e-05,
|
986 |
+
"loss": 0.1884,
|
987 |
+
"step": 163
|
988 |
+
},
|
989 |
+
{
|
990 |
+
"epoch": 0.33,
|
991 |
+
"learning_rate": 1.6115997189518043e-05,
|
992 |
+
"loss": 0.1693,
|
993 |
+
"step": 164
|
994 |
+
},
|
995 |
+
{
|
996 |
+
"epoch": 0.33,
|
997 |
+
"learning_rate": 1.6063536950548825e-05,
|
998 |
+
"loss": 0.1806,
|
999 |
+
"step": 165
|
1000 |
+
},
|
1001 |
+
{
|
1002 |
+
"epoch": 0.33,
|
1003 |
+
"learning_rate": 1.6010811472830253e-05,
|
1004 |
+
"loss": 0.4201,
|
1005 |
+
"step": 166
|
1006 |
+
},
|
1007 |
+
{
|
1008 |
+
"epoch": 0.33,
|
1009 |
+
"learning_rate": 1.595782306274553e-05,
|
1010 |
+
"loss": 0.2366,
|
1011 |
+
"step": 167
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 0.34,
|
1015 |
+
"learning_rate": 1.5904574038179372e-05,
|
1016 |
+
"loss": 0.0898,
|
1017 |
+
"step": 168
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 0.34,
|
1021 |
+
"learning_rate": 1.5851066728416617e-05,
|
1022 |
+
"loss": 0.0674,
|
1023 |
+
"step": 169
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 0.34,
|
1027 |
+
"learning_rate": 1.5797303474040332e-05,
|
1028 |
+
"loss": 0.1518,
|
1029 |
+
"step": 170
|
1030 |
+
},
|
1031 |
+
{
|
1032 |
+
"epoch": 0.34,
|
1033 |
+
"learning_rate": 1.5743286626829437e-05,
|
1034 |
+
"loss": 0.6562,
|
1035 |
+
"step": 171
|
1036 |
+
},
|
1037 |
+
{
|
1038 |
+
"epoch": 0.34,
|
1039 |
+
"learning_rate": 1.5689018549655815e-05,
|
1040 |
+
"loss": 0.1482,
|
1041 |
+
"step": 172
|
1042 |
+
},
|
1043 |
+
{
|
1044 |
+
"epoch": 0.35,
|
1045 |
+
"learning_rate": 1.5634501616380967e-05,
|
1046 |
+
"loss": 0.396,
|
1047 |
+
"step": 173
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 0.35,
|
1051 |
+
"learning_rate": 1.5579738211752165e-05,
|
1052 |
+
"loss": 0.0992,
|
1053 |
+
"step": 174
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 0.35,
|
1057 |
+
"learning_rate": 1.5524730731298136e-05,
|
1058 |
+
"loss": 0.1566,
|
1059 |
+
"step": 175
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.35,
|
1063 |
+
"learning_rate": 1.5469481581224274e-05,
|
1064 |
+
"loss": 0.186,
|
1065 |
+
"step": 176
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 0.35,
|
1069 |
+
"learning_rate": 1.541399317830738e-05,
|
1070 |
+
"loss": 0.372,
|
1071 |
+
"step": 177
|
1072 |
+
},
|
1073 |
+
{
|
1074 |
+
"epoch": 0.36,
|
1075 |
+
"learning_rate": 1.5358267949789968e-05,
|
1076 |
+
"loss": 0.0843,
|
1077 |
+
"step": 178
|
1078 |
+
},
|
1079 |
+
{
|
1080 |
+
"epoch": 0.36,
|
1081 |
+
"learning_rate": 1.530230833327405e-05,
|
1082 |
+
"loss": 0.2175,
|
1083 |
+
"step": 179
|
1084 |
+
},
|
1085 |
+
{
|
1086 |
+
"epoch": 0.36,
|
1087 |
+
"learning_rate": 1.524611677661454e-05,
|
1088 |
+
"loss": 0.4303,
|
1089 |
+
"step": 180
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 0.36,
|
1093 |
+
"learning_rate": 1.5189695737812153e-05,
|
1094 |
+
"loss": 0.3197,
|
1095 |
+
"step": 181
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 0.36,
|
1099 |
+
"learning_rate": 1.5133047684905916e-05,
|
1100 |
+
"loss": 0.0897,
|
1101 |
+
"step": 182
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 0.37,
|
1105 |
+
"learning_rate": 1.5076175095865171e-05,
|
1106 |
+
"loss": 0.2646,
|
1107 |
+
"step": 183
|
1108 |
+
},
|
1109 |
+
{
|
1110 |
+
"epoch": 0.37,
|
1111 |
+
"learning_rate": 1.5019080458481203e-05,
|
1112 |
+
"loss": 0.1229,
|
1113 |
+
"step": 184
|
1114 |
+
},
|
1115 |
+
{
|
1116 |
+
"epoch": 0.37,
|
1117 |
+
"learning_rate": 1.4961766270258422e-05,
|
1118 |
+
"loss": 0.1209,
|
1119 |
+
"step": 185
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 0.37,
|
1123 |
+
"learning_rate": 1.4904235038305084e-05,
|
1124 |
+
"loss": 0.1069,
|
1125 |
+
"step": 186
|
1126 |
+
},
|
1127 |
+
{
|
1128 |
+
"epoch": 0.37,
|
1129 |
+
"learning_rate": 1.4846489279223653e-05,
|
1130 |
+
"loss": 0.3356,
|
1131 |
+
"step": 187
|
1132 |
+
},
|
1133 |
+
{
|
1134 |
+
"epoch": 0.38,
|
1135 |
+
"learning_rate": 1.4788531519000696e-05,
|
1136 |
+
"loss": 0.1611,
|
1137 |
+
"step": 188
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 0.38,
|
1141 |
+
"learning_rate": 1.473036429289641e-05,
|
1142 |
+
"loss": 0.3134,
|
1143 |
+
"step": 189
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 0.38,
|
1147 |
+
"learning_rate": 1.4671990145333697e-05,
|
1148 |
+
"loss": 0.4353,
|
1149 |
+
"step": 190
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 0.38,
|
1153 |
+
"learning_rate": 1.461341162978688e-05,
|
1154 |
+
"loss": 0.2424,
|
1155 |
+
"step": 191
|
1156 |
+
},
|
1157 |
+
{
|
1158 |
+
"epoch": 0.38,
|
1159 |
+
"learning_rate": 1.4554631308669993e-05,
|
1160 |
+
"loss": 0.1279,
|
1161 |
+
"step": 192
|
1162 |
+
},
|
1163 |
+
{
|
1164 |
+
"epoch": 0.39,
|
1165 |
+
"learning_rate": 1.4495651753224706e-05,
|
1166 |
+
"loss": 0.1231,
|
1167 |
+
"step": 193
|
1168 |
+
},
|
1169 |
+
{
|
1170 |
+
"epoch": 0.39,
|
1171 |
+
"learning_rate": 1.4436475543407843e-05,
|
1172 |
+
"loss": 0.1808,
|
1173 |
+
"step": 194
|
1174 |
+
},
|
1175 |
+
{
|
1176 |
+
"epoch": 0.39,
|
1177 |
+
"learning_rate": 1.437710526777852e-05,
|
1178 |
+
"loss": 0.1154,
|
1179 |
+
"step": 195
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 0.39,
|
1183 |
+
"learning_rate": 1.4317543523384928e-05,
|
1184 |
+
"loss": 0.2613,
|
1185 |
+
"step": 196
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 0.39,
|
1189 |
+
"learning_rate": 1.4257792915650728e-05,
|
1190 |
+
"loss": 0.1583,
|
1191 |
+
"step": 197
|
1192 |
+
},
|
1193 |
+
{
|
1194 |
+
"epoch": 0.4,
|
1195 |
+
"learning_rate": 1.4197856058261059e-05,
|
1196 |
+
"loss": 0.2394,
|
1197 |
+
"step": 198
|
1198 |
+
},
|
1199 |
+
{
|
1200 |
+
"epoch": 0.4,
|
1201 |
+
"learning_rate": 1.4137735573048232e-05,
|
1202 |
+
"loss": 0.0783,
|
1203 |
+
"step": 199
|
1204 |
+
},
|
1205 |
+
{
|
1206 |
+
"epoch": 0.4,
|
1207 |
+
"learning_rate": 1.4077434089877038e-05,
|
1208 |
+
"loss": 0.1302,
|
1209 |
+
"step": 200
|
1210 |
+
},
|
1211 |
+
{
|
1212 |
+
"epoch": 0.4,
|
1213 |
+
"learning_rate": 1.4016954246529697e-05,
|
1214 |
+
"loss": 0.1186,
|
1215 |
+
"step": 201
|
1216 |
+
},
|
1217 |
+
{
|
1218 |
+
"epoch": 0.4,
|
1219 |
+
"learning_rate": 1.3956298688590484e-05,
|
1220 |
+
"loss": 0.0853,
|
1221 |
+
"step": 202
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 0.41,
|
1225 |
+
"learning_rate": 1.3895470069330003e-05,
|
1226 |
+
"loss": 0.2058,
|
1227 |
+
"step": 203
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 0.41,
|
1231 |
+
"learning_rate": 1.3834471049589117e-05,
|
1232 |
+
"loss": 0.2419,
|
1233 |
+
"step": 204
|
1234 |
+
},
|
1235 |
+
{
|
1236 |
+
"epoch": 0.41,
|
1237 |
+
"learning_rate": 1.377330429766256e-05,
|
1238 |
+
"loss": 0.2448,
|
1239 |
+
"step": 205
|
1240 |
+
},
|
1241 |
+
{
|
1242 |
+
"epoch": 0.41,
|
1243 |
+
"learning_rate": 1.3711972489182208e-05,
|
1244 |
+
"loss": 0.0812,
|
1245 |
+
"step": 206
|
1246 |
+
},
|
1247 |
+
{
|
1248 |
+
"epoch": 0.41,
|
1249 |
+
"learning_rate": 1.3650478307000059e-05,
|
1250 |
+
"loss": 0.0932,
|
1251 |
+
"step": 207
|
1252 |
+
},
|
1253 |
+
{
|
1254 |
+
"epoch": 0.42,
|
1255 |
+
"learning_rate": 1.3588824441070852e-05,
|
1256 |
+
"loss": 0.2578,
|
1257 |
+
"step": 208
|
1258 |
+
},
|
1259 |
+
{
|
1260 |
+
"epoch": 0.42,
|
1261 |
+
"learning_rate": 1.3527013588334415e-05,
|
1262 |
+
"loss": 0.1267,
|
1263 |
+
"step": 209
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 0.42,
|
1267 |
+
"learning_rate": 1.3465048452597682e-05,
|
1268 |
+
"loss": 0.1509,
|
1269 |
+
"step": 210
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 0.42,
|
1273 |
+
"learning_rate": 1.3402931744416432e-05,
|
1274 |
+
"loss": 0.2086,
|
1275 |
+
"step": 211
|
1276 |
+
},
|
1277 |
+
{
|
1278 |
+
"epoch": 0.42,
|
1279 |
+
"learning_rate": 1.3340666180976713e-05,
|
1280 |
+
"loss": 0.1317,
|
1281 |
+
"step": 212
|
1282 |
+
},
|
1283 |
+
{
|
1284 |
+
"epoch": 0.43,
|
1285 |
+
"learning_rate": 1.3340666180976713e-05,
|
1286 |
+
"loss": 0.096,
|
1287 |
+
"step": 213
|
1288 |
+
},
|
1289 |
+
{
|
1290 |
+
"epoch": 0.43,
|
1291 |
+
"learning_rate": 1.3278254485975977e-05,
|
1292 |
+
"loss": 0.3035,
|
1293 |
+
"step": 214
|
1294 |
+
},
|
1295 |
+
{
|
1296 |
+
"epoch": 0.43,
|
1297 |
+
"learning_rate": 1.3215699389503956e-05,
|
1298 |
+
"loss": 0.3848,
|
1299 |
+
"step": 215
|
1300 |
+
},
|
1301 |
+
{
|
1302 |
+
"epoch": 0.43,
|
1303 |
+
"learning_rate": 1.3153003627923217e-05,
|
1304 |
+
"loss": 0.2722,
|
1305 |
+
"step": 216
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 0.43,
|
1309 |
+
"learning_rate": 1.3090169943749475e-05,
|
1310 |
+
"loss": 0.2982,
|
1311 |
+
"step": 217
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 0.44,
|
1315 |
+
"learning_rate": 1.3027201085531633e-05,
|
1316 |
+
"loss": 0.0712,
|
1317 |
+
"step": 218
|
1318 |
+
},
|
1319 |
+
{
|
1320 |
+
"epoch": 0.44,
|
1321 |
+
"learning_rate": 1.2964099807731539e-05,
|
1322 |
+
"loss": 0.0559,
|
1323 |
+
"step": 219
|
1324 |
+
},
|
1325 |
+
{
|
1326 |
+
"epoch": 0.44,
|
1327 |
+
"learning_rate": 1.2900868870603502e-05,
|
1328 |
+
"loss": 0.1888,
|
1329 |
+
"step": 220
|
1330 |
+
},
|
1331 |
+
{
|
1332 |
+
"epoch": 0.44,
|
1333 |
+
"learning_rate": 1.283751104007355e-05,
|
1334 |
+
"loss": 0.1306,
|
1335 |
+
"step": 221
|
1336 |
+
},
|
1337 |
+
{
|
1338 |
+
"epoch": 0.44,
|
1339 |
+
"learning_rate": 1.2774029087618448e-05,
|
1340 |
+
"loss": 0.1362,
|
1341 |
+
"step": 222
|
1342 |
+
},
|
1343 |
+
{
|
1344 |
+
"epoch": 0.45,
|
1345 |
+
"learning_rate": 1.2710425790144445e-05,
|
1346 |
+
"loss": 0.0681,
|
1347 |
+
"step": 223
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 0.45,
|
1351 |
+
"learning_rate": 1.2646703929865817e-05,
|
1352 |
+
"loss": 0.1985,
|
1353 |
+
"step": 224
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 0.45,
|
1357 |
+
"learning_rate": 1.2582866294183167e-05,
|
1358 |
+
"loss": 0.3362,
|
1359 |
+
"step": 225
|
1360 |
+
},
|
1361 |
+
{
|
1362 |
+
"epoch": 0.45,
|
1363 |
+
"learning_rate": 1.2518915675561482e-05,
|
1364 |
+
"loss": 0.266,
|
1365 |
+
"step": 226
|
1366 |
+
},
|
1367 |
+
{
|
1368 |
+
"epoch": 0.45,
|
1369 |
+
"learning_rate": 1.2454854871407993e-05,
|
1370 |
+
"loss": 0.2089,
|
1371 |
+
"step": 227
|
1372 |
+
},
|
1373 |
+
{
|
1374 |
+
"epoch": 0.46,
|
1375 |
+
"learning_rate": 1.2390686683949799e-05,
|
1376 |
+
"loss": 0.2582,
|
1377 |
+
"step": 228
|
1378 |
+
},
|
1379 |
+
{
|
1380 |
+
"epoch": 0.46,
|
1381 |
+
"learning_rate": 1.2326413920111304e-05,
|
1382 |
+
"loss": 0.2865,
|
1383 |
+
"step": 229
|
1384 |
+
},
|
1385 |
+
{
|
1386 |
+
"epoch": 0.46,
|
1387 |
+
"learning_rate": 1.2262039391391405e-05,
|
1388 |
+
"loss": 0.1277,
|
1389 |
+
"step": 230
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 0.46,
|
1393 |
+
"learning_rate": 1.2197565913740531e-05,
|
1394 |
+
"loss": 0.2532,
|
1395 |
+
"step": 231
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 0.46,
|
1399 |
+
"learning_rate": 1.213299630743747e-05,
|
1400 |
+
"loss": 0.1533,
|
1401 |
+
"step": 232
|
1402 |
+
},
|
1403 |
+
{
|
1404 |
+
"epoch": 0.47,
|
1405 |
+
"learning_rate": 1.2068333396965968e-05,
|
1406 |
+
"loss": 0.0608,
|
1407 |
+
"step": 233
|
1408 |
+
},
|
1409 |
+
{
|
1410 |
+
"epoch": 0.47,
|
1411 |
+
"learning_rate": 1.2003580010891214e-05,
|
1412 |
+
"loss": 0.2867,
|
1413 |
+
"step": 234
|
1414 |
+
},
|
1415 |
+
{
|
1416 |
+
"epoch": 0.47,
|
1417 |
+
"learning_rate": 1.1938738981736084e-05,
|
1418 |
+
"loss": 0.2761,
|
1419 |
+
"step": 235
|
1420 |
+
},
|
1421 |
+
{
|
1422 |
+
"epoch": 0.47,
|
1423 |
+
"learning_rate": 1.187381314585725e-05,
|
1424 |
+
"loss": 0.2299,
|
1425 |
+
"step": 236
|
1426 |
+
},
|
1427 |
+
{
|
1428 |
+
"epoch": 0.47,
|
1429 |
+
"learning_rate": 1.1808805343321102e-05,
|
1430 |
+
"loss": 0.4514,
|
1431 |
+
"step": 237
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 0.48,
|
1435 |
+
"learning_rate": 1.1743718417779518e-05,
|
1436 |
+
"loss": 0.2327,
|
1437 |
+
"step": 238
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 0.48,
|
1441 |
+
"learning_rate": 1.1678555216345478e-05,
|
1442 |
+
"loss": 0.5068,
|
1443 |
+
"step": 239
|
1444 |
+
},
|
1445 |
+
{
|
1446 |
+
"epoch": 0.48,
|
1447 |
+
"learning_rate": 1.1613318589468512e-05,
|
1448 |
+
"loss": 0.183,
|
1449 |
+
"step": 240
|
1450 |
+
},
|
1451 |
+
{
|
1452 |
+
"epoch": 0.48,
|
1453 |
+
"learning_rate": 1.1548011390810016e-05,
|
1454 |
+
"loss": 0.4153,
|
1455 |
+
"step": 241
|
1456 |
+
},
|
1457 |
+
{
|
1458 |
+
"epoch": 0.48,
|
1459 |
+
"learning_rate": 1.148263647711842e-05,
|
1460 |
+
"loss": 0.1517,
|
1461 |
+
"step": 242
|
1462 |
+
},
|
1463 |
+
{
|
1464 |
+
"epoch": 0.49,
|
1465 |
+
"learning_rate": 1.1417196708104244e-05,
|
1466 |
+
"loss": 0.2708,
|
1467 |
+
"step": 243
|
1468 |
+
},
|
1469 |
+
{
|
1470 |
+
"epoch": 0.49,
|
1471 |
+
"learning_rate": 1.135169494631497e-05,
|
1472 |
+
"loss": 0.0495,
|
1473 |
+
"step": 244
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 0.49,
|
1477 |
+
"learning_rate": 1.1286134057009862e-05,
|
1478 |
+
"loss": 0.0739,
|
1479 |
+
"step": 245
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 0.49,
|
1483 |
+
"learning_rate": 1.1220516908034602e-05,
|
1484 |
+
"loss": 0.3282,
|
1485 |
+
"step": 246
|
1486 |
+
},
|
1487 |
+
{
|
1488 |
+
"epoch": 0.49,
|
1489 |
+
"learning_rate": 1.1154846369695864e-05,
|
1490 |
+
"loss": 0.0478,
|
1491 |
+
"step": 247
|
1492 |
+
},
|
1493 |
+
{
|
1494 |
+
"epoch": 0.5,
|
1495 |
+
"learning_rate": 1.1089125314635727e-05,
|
1496 |
+
"loss": 0.2256,
|
1497 |
+
"step": 248
|
1498 |
+
},
|
1499 |
+
{
|
1500 |
+
"epoch": 0.5,
|
1501 |
+
"learning_rate": 1.1023356617706051e-05,
|
1502 |
+
"loss": 0.107,
|
1503 |
+
"step": 249
|
1504 |
+
},
|
1505 |
+
{
|
1506 |
+
"epoch": 0.5,
|
1507 |
+
"learning_rate": 1.0957543155842703e-05,
|
1508 |
+
"loss": 0.0797,
|
1509 |
+
"step": 250
|
1510 |
+
},
|
1511 |
+
{
|
1512 |
+
"epoch": 0.5,
|
1513 |
+
"learning_rate": 1.0891687807939707e-05,
|
1514 |
+
"loss": 0.07,
|
1515 |
+
"step": 251
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 0.5,
|
1519 |
+
"learning_rate": 1.0825793454723325e-05,
|
1520 |
+
"loss": 0.3117,
|
1521 |
+
"step": 252
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 0.51,
|
1525 |
+
"learning_rate": 1.0759862978626032e-05,
|
1526 |
+
"loss": 0.2769,
|
1527 |
+
"step": 253
|
1528 |
+
},
|
1529 |
+
{
|
1530 |
+
"epoch": 0.51,
|
1531 |
+
"learning_rate": 1.0693899263660442e-05,
|
1532 |
+
"loss": 0.1839,
|
1533 |
+
"step": 254
|
1534 |
+
},
|
1535 |
+
{
|
1536 |
+
"epoch": 0.51,
|
1537 |
+
"learning_rate": 1.0627905195293135e-05,
|
1538 |
+
"loss": 0.0428,
|
1539 |
+
"step": 255
|
1540 |
+
},
|
1541 |
+
{
|
1542 |
+
"epoch": 0.51,
|
1543 |
+
"learning_rate": 1.0561883660318456e-05,
|
1544 |
+
"loss": 0.0903,
|
1545 |
+
"step": 256
|
1546 |
+
},
|
1547 |
+
{
|
1548 |
+
"epoch": 0.51,
|
1549 |
+
"learning_rate": 1.0495837546732224e-05,
|
1550 |
+
"loss": 0.2216,
|
1551 |
+
"step": 257
|
1552 |
+
},
|
1553 |
+
{
|
1554 |
+
"epoch": 0.52,
|
1555 |
+
"learning_rate": 1.0429769743605406e-05,
|
1556 |
+
"loss": 0.0941,
|
1557 |
+
"step": 258
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 0.52,
|
1561 |
+
"learning_rate": 1.0363683140957745e-05,
|
1562 |
+
"loss": 0.5135,
|
1563 |
+
"step": 259
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 0.52,
|
1567 |
+
"learning_rate": 1.0297580629631324e-05,
|
1568 |
+
"loss": 0.1504,
|
1569 |
+
"step": 260
|
1570 |
+
},
|
1571 |
+
{
|
1572 |
+
"epoch": 0.52,
|
1573 |
+
"learning_rate": 1.023146510116414e-05,
|
1574 |
+
"loss": 0.1991,
|
1575 |
+
"step": 261
|
1576 |
+
},
|
1577 |
+
{
|
1578 |
+
"epoch": 0.52,
|
1579 |
+
"learning_rate": 1.0165339447663586e-05,
|
1580 |
+
"loss": 0.0628,
|
1581 |
+
"step": 262
|
1582 |
+
},
|
1583 |
+
{
|
1584 |
+
"epoch": 0.53,
|
1585 |
+
"learning_rate": 1.0099206561679964e-05,
|
1586 |
+
"loss": 0.298,
|
1587 |
+
"step": 263
|
1588 |
+
},
|
1589 |
+
{
|
1590 |
+
"epoch": 0.53,
|
1591 |
+
"learning_rate": 1.0033069336079952e-05,
|
1592 |
+
"loss": 0.5587,
|
1593 |
+
"step": 264
|
1594 |
+
},
|
1595 |
+
{
|
1596 |
+
"epoch": 0.53,
|
1597 |
+
"learning_rate": 9.96693066392005e-06,
|
1598 |
+
"loss": 0.0739,
|
1599 |
+
"step": 265
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 0.53,
|
1603 |
+
"learning_rate": 9.900793438320037e-06,
|
1604 |
+
"loss": 0.3523,
|
1605 |
+
"step": 266
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 0.53,
|
1609 |
+
"learning_rate": 9.834660552336415e-06,
|
1610 |
+
"loss": 0.1706,
|
1611 |
+
"step": 267
|
1612 |
+
},
|
1613 |
+
{
|
1614 |
+
"epoch": 0.54,
|
1615 |
+
"learning_rate": 9.768534898835864e-06,
|
1616 |
+
"loss": 0.2783,
|
1617 |
+
"step": 268
|
1618 |
+
},
|
1619 |
+
{
|
1620 |
+
"epoch": 0.54,
|
1621 |
+
"learning_rate": 9.702419370368677e-06,
|
1622 |
+
"loss": 0.2093,
|
1623 |
+
"step": 269
|
1624 |
+
},
|
1625 |
+
{
|
1626 |
+
"epoch": 0.54,
|
1627 |
+
"learning_rate": 9.636316859042258e-06,
|
1628 |
+
"loss": 0.2205,
|
1629 |
+
"step": 270
|
1630 |
+
},
|
1631 |
+
{
|
1632 |
+
"epoch": 0.54,
|
1633 |
+
"learning_rate": 9.570230256394595e-06,
|
1634 |
+
"loss": 0.1987,
|
1635 |
+
"step": 271
|
1636 |
+
},
|
1637 |
+
{
|
1638 |
+
"epoch": 0.54,
|
1639 |
+
"learning_rate": 9.504162453267776e-06,
|
1640 |
+
"loss": 0.1426,
|
1641 |
+
"step": 272
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 0.55,
|
1645 |
+
"learning_rate": 9.438116339681546e-06,
|
1646 |
+
"loss": 0.0697,
|
1647 |
+
"step": 273
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 0.55,
|
1651 |
+
"learning_rate": 9.372094804706867e-06,
|
1652 |
+
"loss": 0.1873,
|
1653 |
+
"step": 274
|
1654 |
+
},
|
1655 |
+
{
|
1656 |
+
"epoch": 0.55,
|
1657 |
+
"learning_rate": 9.30610073633956e-06,
|
1658 |
+
"loss": 0.3201,
|
1659 |
+
"step": 275
|
1660 |
+
},
|
1661 |
+
{
|
1662 |
+
"epoch": 0.55,
|
1663 |
+
"learning_rate": 9.24013702137397e-06,
|
1664 |
+
"loss": 0.1485,
|
1665 |
+
"step": 276
|
1666 |
+
},
|
1667 |
+
{
|
1668 |
+
"epoch": 0.55,
|
1669 |
+
"learning_rate": 9.174206545276678e-06,
|
1670 |
+
"loss": 0.0733,
|
1671 |
+
"step": 277
|
1672 |
+
},
|
1673 |
+
{
|
1674 |
+
"epoch": 0.56,
|
1675 |
+
"learning_rate": 9.108312192060298e-06,
|
1676 |
+
"loss": 0.4001,
|
1677 |
+
"step": 278
|
1678 |
+
},
|
1679 |
+
{
|
1680 |
+
"epoch": 0.56,
|
1681 |
+
"learning_rate": 9.042456844157299e-06,
|
1682 |
+
"loss": 0.1089,
|
1683 |
+
"step": 279
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 0.56,
|
1687 |
+
"learning_rate": 8.97664338229395e-06,
|
1688 |
+
"loss": 0.0931,
|
1689 |
+
"step": 280
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 0.56,
|
1693 |
+
"learning_rate": 8.910874685364275e-06,
|
1694 |
+
"loss": 0.5483,
|
1695 |
+
"step": 281
|
1696 |
+
},
|
1697 |
+
{
|
1698 |
+
"epoch": 0.56,
|
1699 |
+
"learning_rate": 8.84515363030414e-06,
|
1700 |
+
"loss": 0.071,
|
1701 |
+
"step": 282
|
1702 |
+
},
|
1703 |
+
{
|
1704 |
+
"epoch": 0.57,
|
1705 |
+
"learning_rate": 8.7794830919654e-06,
|
1706 |
+
"loss": 0.2319,
|
1707 |
+
"step": 283
|
1708 |
+
},
|
1709 |
+
{
|
1710 |
+
"epoch": 0.57,
|
1711 |
+
"learning_rate": 8.713865942990143e-06,
|
1712 |
+
"loss": 0.264,
|
1713 |
+
"step": 284
|
1714 |
+
},
|
1715 |
+
{
|
1716 |
+
"epoch": 0.57,
|
1717 |
+
"learning_rate": 8.648305053685035e-06,
|
1718 |
+
"loss": 0.3833,
|
1719 |
+
"step": 285
|
1720 |
+
},
|
1721 |
+
{
|
1722 |
+
"epoch": 0.57,
|
1723 |
+
"learning_rate": 8.582803291895758e-06,
|
1724 |
+
"loss": 0.3883,
|
1725 |
+
"step": 286
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 0.57,
|
1729 |
+
"learning_rate": 8.51736352288158e-06,
|
1730 |
+
"loss": 0.1494,
|
1731 |
+
"step": 287
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 0.58,
|
1735 |
+
"learning_rate": 8.451988609189987e-06,
|
1736 |
+
"loss": 0.1005,
|
1737 |
+
"step": 288
|
1738 |
+
},
|
1739 |
+
{
|
1740 |
+
"epoch": 0.58,
|
1741 |
+
"learning_rate": 8.386681410531493e-06,
|
1742 |
+
"loss": 0.1057,
|
1743 |
+
"step": 289
|
1744 |
+
},
|
1745 |
+
{
|
1746 |
+
"epoch": 0.58,
|
1747 |
+
"learning_rate": 8.321444783654524e-06,
|
1748 |
+
"loss": 0.1763,
|
1749 |
+
"step": 290
|
1750 |
+
},
|
1751 |
+
{
|
1752 |
+
"epoch": 0.58,
|
1753 |
+
"learning_rate": 8.256281582220486e-06,
|
1754 |
+
"loss": 0.2966,
|
1755 |
+
"step": 291
|
1756 |
+
},
|
1757 |
+
{
|
1758 |
+
"epoch": 0.58,
|
1759 |
+
"learning_rate": 8.191194656678905e-06,
|
1760 |
+
"loss": 0.12,
|
1761 |
+
"step": 292
|
1762 |
+
},
|
1763 |
+
{
|
1764 |
+
"epoch": 0.59,
|
1765 |
+
"learning_rate": 8.126186854142752e-06,
|
1766 |
+
"loss": 0.4565,
|
1767 |
+
"step": 293
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 0.59,
|
1771 |
+
"learning_rate": 8.06126101826392e-06,
|
1772 |
+
"loss": 0.1963,
|
1773 |
+
"step": 294
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 0.59,
|
1777 |
+
"learning_rate": 7.99641998910879e-06,
|
1778 |
+
"loss": 0.0081,
|
1779 |
+
"step": 295
|
1780 |
+
},
|
1781 |
+
{
|
1782 |
+
"epoch": 0.59,
|
1783 |
+
"learning_rate": 7.931666603034034e-06,
|
1784 |
+
"loss": 0.0224,
|
1785 |
+
"step": 296
|
1786 |
+
},
|
1787 |
+
{
|
1788 |
+
"epoch": 0.59,
|
1789 |
+
"learning_rate": 7.867003692562533e-06,
|
1790 |
+
"loss": 0.2567,
|
1791 |
+
"step": 297
|
1792 |
+
},
|
1793 |
+
{
|
1794 |
+
"epoch": 0.6,
|
1795 |
+
"learning_rate": 7.80243408625947e-06,
|
1796 |
+
"loss": 0.0473,
|
1797 |
+
"step": 298
|
1798 |
+
},
|
1799 |
+
{
|
1800 |
+
"epoch": 0.6,
|
1801 |
+
"learning_rate": 7.7379606086086e-06,
|
1802 |
+
"loss": 0.5308,
|
1803 |
+
"step": 299
|
1804 |
+
},
|
1805 |
+
{
|
1806 |
+
"epoch": 0.6,
|
1807 |
+
"learning_rate": 7.673586079888699e-06,
|
1808 |
+
"loss": 0.0563,
|
1809 |
+
"step": 300
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 0.6,
|
1813 |
+
"learning_rate": 7.6093133160502e-06,
|
1814 |
+
"loss": 0.1527,
|
1815 |
+
"step": 301
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 0.6,
|
1819 |
+
"learning_rate": 7.545145128592009e-06,
|
1820 |
+
"loss": 0.6643,
|
1821 |
+
"step": 302
|
1822 |
+
},
|
1823 |
+
{
|
1824 |
+
"epoch": 0.61,
|
1825 |
+
"learning_rate": 7.481084324438521e-06,
|
1826 |
+
"loss": 0.162,
|
1827 |
+
"step": 303
|
1828 |
+
},
|
1829 |
+
{
|
1830 |
+
"epoch": 0.61,
|
1831 |
+
"learning_rate": 7.4171337058168365e-06,
|
1832 |
+
"loss": 0.0417,
|
1833 |
+
"step": 304
|
1834 |
+
},
|
1835 |
+
{
|
1836 |
+
"epoch": 0.61,
|
1837 |
+
"learning_rate": 7.353296070134186e-06,
|
1838 |
+
"loss": 0.361,
|
1839 |
+
"step": 305
|
1840 |
+
},
|
1841 |
+
{
|
1842 |
+
"epoch": 0.61,
|
1843 |
+
"learning_rate": 7.28957420985556e-06,
|
1844 |
+
"loss": 0.2668,
|
1845 |
+
"step": 306
|
1846 |
+
},
|
1847 |
+
{
|
1848 |
+
"epoch": 0.61,
|
1849 |
+
"learning_rate": 7.225970912381557e-06,
|
1850 |
+
"loss": 0.1708,
|
1851 |
+
"step": 307
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 0.62,
|
1855 |
+
"learning_rate": 7.16248895992645e-06,
|
1856 |
+
"loss": 0.0891,
|
1857 |
+
"step": 308
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 0.62,
|
1861 |
+
"learning_rate": 7.099131129396501e-06,
|
1862 |
+
"loss": 0.2068,
|
1863 |
+
"step": 309
|
1864 |
+
},
|
1865 |
+
{
|
1866 |
+
"epoch": 0.62,
|
1867 |
+
"learning_rate": 7.035900192268464e-06,
|
1868 |
+
"loss": 0.2319,
|
1869 |
+
"step": 310
|
1870 |
+
},
|
1871 |
+
{
|
1872 |
+
"epoch": 0.62,
|
1873 |
+
"learning_rate": 6.972798914468369e-06,
|
1874 |
+
"loss": 0.2967,
|
1875 |
+
"step": 311
|
1876 |
+
},
|
1877 |
+
{
|
1878 |
+
"epoch": 0.62,
|
1879 |
+
"learning_rate": 6.909830056250527e-06,
|
1880 |
+
"loss": 0.109,
|
1881 |
+
"step": 312
|
1882 |
+
},
|
1883 |
+
{
|
1884 |
+
"epoch": 0.63,
|
1885 |
+
"learning_rate": 6.846996372076786e-06,
|
1886 |
+
"loss": 0.0556,
|
1887 |
+
"step": 313
|
1888 |
+
},
|
1889 |
+
{
|
1890 |
+
"epoch": 0.63,
|
1891 |
+
"learning_rate": 6.784300610496049e-06,
|
1892 |
+
"loss": 0.2744,
|
1893 |
+
"step": 314
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 0.63,
|
1897 |
+
"learning_rate": 6.721745514024023e-06,
|
1898 |
+
"loss": 0.3622,
|
1899 |
+
"step": 315
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 0.63,
|
1903 |
+
"learning_rate": 6.659333819023291e-06,
|
1904 |
+
"loss": 0.1994,
|
1905 |
+
"step": 316
|
1906 |
+
},
|
1907 |
+
{
|
1908 |
+
"epoch": 0.63,
|
1909 |
+
"learning_rate": 6.59706825558357e-06,
|
1910 |
+
"loss": 0.4396,
|
1911 |
+
"step": 317
|
1912 |
+
},
|
1913 |
+
{
|
1914 |
+
"epoch": 0.64,
|
1915 |
+
"learning_rate": 6.534951547402322e-06,
|
1916 |
+
"loss": 0.3304,
|
1917 |
+
"step": 318
|
1918 |
+
},
|
1919 |
+
{
|
1920 |
+
"epoch": 0.64,
|
1921 |
+
"learning_rate": 6.47298641166559e-06,
|
1922 |
+
"loss": 0.3576,
|
1923 |
+
"step": 319
|
1924 |
+
},
|
1925 |
+
{
|
1926 |
+
"epoch": 0.64,
|
1927 |
+
"learning_rate": 6.411175558929152e-06,
|
1928 |
+
"loss": 0.1696,
|
1929 |
+
"step": 320
|
1930 |
+
},
|
1931 |
+
{
|
1932 |
+
"epoch": 0.64,
|
1933 |
+
"learning_rate": 6.349521692999945e-06,
|
1934 |
+
"loss": 0.072,
|
1935 |
+
"step": 321
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 0.64,
|
1939 |
+
"learning_rate": 6.2880275108177915e-06,
|
1940 |
+
"loss": 0.568,
|
1941 |
+
"step": 322
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 0.65,
|
1945 |
+
"learning_rate": 6.2266957023374434e-06,
|
1946 |
+
"loss": 0.3337,
|
1947 |
+
"step": 323
|
1948 |
+
},
|
1949 |
+
{
|
1950 |
+
"epoch": 0.65,
|
1951 |
+
"learning_rate": 6.165528950410884e-06,
|
1952 |
+
"loss": 0.0888,
|
1953 |
+
"step": 324
|
1954 |
+
},
|
1955 |
+
{
|
1956 |
+
"epoch": 0.65,
|
1957 |
+
"learning_rate": 6.10452993067e-06,
|
1958 |
+
"loss": 0.2292,
|
1959 |
+
"step": 325
|
1960 |
+
},
|
1961 |
+
{
|
1962 |
+
"epoch": 0.65,
|
1963 |
+
"learning_rate": 6.04370131140952e-06,
|
1964 |
+
"loss": 0.1889,
|
1965 |
+
"step": 326
|
1966 |
+
},
|
1967 |
+
{
|
1968 |
+
"epoch": 0.65,
|
1969 |
+
"learning_rate": 5.983045753470308e-06,
|
1970 |
+
"loss": 0.0938,
|
1971 |
+
"step": 327
|
1972 |
+
},
|
1973 |
+
{
|
1974 |
+
"epoch": 0.66,
|
1975 |
+
"learning_rate": 5.922565910122967e-06,
|
1976 |
+
"loss": 0.3774,
|
1977 |
+
"step": 328
|
1978 |
+
},
|
1979 |
+
{
|
1980 |
+
"epoch": 0.66,
|
1981 |
+
"learning_rate": 5.862264426951768e-06,
|
1982 |
+
"loss": 0.1941,
|
1983 |
+
"step": 329
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 0.66,
|
1987 |
+
"learning_rate": 5.802143941738945e-06,
|
1988 |
+
"loss": 0.1338,
|
1989 |
+
"step": 330
|
1990 |
+
},
|
1991 |
+
{
|
1992 |
+
"epoch": 0.66,
|
1993 |
+
"learning_rate": 5.742207084349274e-06,
|
1994 |
+
"loss": 0.3389,
|
1995 |
+
"step": 331
|
1996 |
+
},
|
1997 |
+
{
|
1998 |
+
"epoch": 0.66,
|
1999 |
+
"learning_rate": 5.6824564766150724e-06,
|
2000 |
+
"loss": 0.1251,
|
2001 |
+
"step": 332
|
2002 |
+
},
|
2003 |
+
{
|
2004 |
+
"epoch": 0.67,
|
2005 |
+
"learning_rate": 5.622894732221482e-06,
|
2006 |
+
"loss": 0.2994,
|
2007 |
+
"step": 333
|
2008 |
+
},
|
2009 |
+
{
|
2010 |
+
"epoch": 0.67,
|
2011 |
+
"learning_rate": 5.563524456592163e-06,
|
2012 |
+
"loss": 0.1573,
|
2013 |
+
"step": 334
|
2014 |
+
},
|
2015 |
+
{
|
2016 |
+
"epoch": 0.67,
|
2017 |
+
"learning_rate": 5.504348246775299e-06,
|
2018 |
+
"loss": 0.1989,
|
2019 |
+
"step": 335
|
2020 |
+
},
|
2021 |
+
{
|
2022 |
+
"epoch": 0.67,
|
2023 |
+
"learning_rate": 5.445368691330008e-06,
|
2024 |
+
"loss": 0.1285,
|
2025 |
+
"step": 336
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 0.67,
|
2029 |
+
"learning_rate": 5.386588370213124e-06,
|
2030 |
+
"loss": 0.0982,
|
2031 |
+
"step": 337
|
2032 |
+
},
|
2033 |
+
{
|
2034 |
+
"epoch": 0.68,
|
2035 |
+
"learning_rate": 5.328009854666303e-06,
|
2036 |
+
"loss": 0.2538,
|
2037 |
+
"step": 338
|
2038 |
+
},
|
2039 |
+
{
|
2040 |
+
"epoch": 0.68,
|
2041 |
+
"learning_rate": 5.269635707103593e-06,
|
2042 |
+
"loss": 0.0733,
|
2043 |
+
"step": 339
|
2044 |
+
},
|
2045 |
+
{
|
2046 |
+
"epoch": 0.68,
|
2047 |
+
"learning_rate": 5.211468480999304e-06,
|
2048 |
+
"loss": 0.5152,
|
2049 |
+
"step": 340
|
2050 |
+
},
|
2051 |
+
{
|
2052 |
+
"epoch": 0.68,
|
2053 |
+
"learning_rate": 5.153510720776354e-06,
|
2054 |
+
"loss": 0.1063,
|
2055 |
+
"step": 341
|
2056 |
+
},
|
2057 |
+
{
|
2058 |
+
"epoch": 0.68,
|
2059 |
+
"learning_rate": 5.095764961694923e-06,
|
2060 |
+
"loss": 0.4609,
|
2061 |
+
"step": 342
|
2062 |
+
},
|
2063 |
+
{
|
2064 |
+
"epoch": 0.69,
|
2065 |
+
"learning_rate": 5.0382337297415775e-06,
|
2066 |
+
"loss": 0.1152,
|
2067 |
+
"step": 343
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 0.69,
|
2071 |
+
"learning_rate": 4.980919541518796e-06,
|
2072 |
+
"loss": 0.2863,
|
2073 |
+
"step": 344
|
2074 |
+
},
|
2075 |
+
{
|
2076 |
+
"epoch": 0.69,
|
2077 |
+
"learning_rate": 4.92382490413483e-06,
|
2078 |
+
"loss": 0.1122,
|
2079 |
+
"step": 345
|
2080 |
+
},
|
2081 |
+
{
|
2082 |
+
"epoch": 0.69,
|
2083 |
+
"learning_rate": 4.866952315094088e-06,
|
2084 |
+
"loss": 0.0646,
|
2085 |
+
"step": 346
|
2086 |
+
},
|
2087 |
+
{
|
2088 |
+
"epoch": 0.69,
|
2089 |
+
"learning_rate": 4.8103042621878515e-06,
|
2090 |
+
"loss": 0.1197,
|
2091 |
+
"step": 347
|
2092 |
+
},
|
2093 |
+
{
|
2094 |
+
"epoch": 0.7,
|
2095 |
+
"learning_rate": 4.753883223385467e-06,
|
2096 |
+
"loss": 0.0571,
|
2097 |
+
"step": 348
|
2098 |
+
},
|
2099 |
+
{
|
2100 |
+
"epoch": 0.7,
|
2101 |
+
"learning_rate": 4.6976916667259555e-06,
|
2102 |
+
"loss": 0.2951,
|
2103 |
+
"step": 349
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 0.7,
|
2107 |
+
"learning_rate": 4.641732050210032e-06,
|
2108 |
+
"loss": 0.1941,
|
2109 |
+
"step": 350
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 0.7,
|
2113 |
+
"learning_rate": 4.58600682169262e-06,
|
2114 |
+
"loss": 0.1809,
|
2115 |
+
"step": 351
|
2116 |
+
},
|
2117 |
+
{
|
2118 |
+
"epoch": 0.7,
|
2119 |
+
"learning_rate": 4.530518418775734e-06,
|
2120 |
+
"loss": 0.4181,
|
2121 |
+
"step": 352
|
2122 |
+
},
|
2123 |
+
{
|
2124 |
+
"epoch": 0.71,
|
2125 |
+
"learning_rate": 4.4752692687018685e-06,
|
2126 |
+
"loss": 0.0959,
|
2127 |
+
"step": 353
|
2128 |
+
},
|
2129 |
+
{
|
2130 |
+
"epoch": 0.71,
|
2131 |
+
"learning_rate": 4.420261788247841e-06,
|
2132 |
+
"loss": 0.1603,
|
2133 |
+
"step": 354
|
2134 |
+
},
|
2135 |
+
{
|
2136 |
+
"epoch": 0.71,
|
2137 |
+
"learning_rate": 4.365498383619036e-06,
|
2138 |
+
"loss": 0.361,
|
2139 |
+
"step": 355
|
2140 |
+
},
|
2141 |
+
{
|
2142 |
+
"epoch": 0.71,
|
2143 |
+
"learning_rate": 4.3109814503441894e-06,
|
2144 |
+
"loss": 0.311,
|
2145 |
+
"step": 356
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 0.71,
|
2149 |
+
"learning_rate": 4.256713373170565e-06,
|
2150 |
+
"loss": 0.1056,
|
2151 |
+
"step": 357
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 0.72,
|
2155 |
+
"learning_rate": 4.202696525959667e-06,
|
2156 |
+
"loss": 0.1141,
|
2157 |
+
"step": 358
|
2158 |
+
},
|
2159 |
+
{
|
2160 |
+
"epoch": 0.72,
|
2161 |
+
"learning_rate": 4.148933271583385e-06,
|
2162 |
+
"loss": 0.0853,
|
2163 |
+
"step": 359
|
2164 |
+
},
|
2165 |
+
{
|
2166 |
+
"epoch": 0.72,
|
2167 |
+
"learning_rate": 4.0954259618206295e-06,
|
2168 |
+
"loss": 0.6156,
|
2169 |
+
"step": 360
|
2170 |
+
},
|
2171 |
+
{
|
2172 |
+
"epoch": 0.72,
|
2173 |
+
"learning_rate": 4.042176937254474e-06,
|
2174 |
+
"loss": 0.1291,
|
2175 |
+
"step": 361
|
2176 |
+
},
|
2177 |
+
{
|
2178 |
+
"epoch": 0.72,
|
2179 |
+
"learning_rate": 3.989188527169749e-06,
|
2180 |
+
"loss": 0.0533,
|
2181 |
+
"step": 362
|
2182 |
+
},
|
2183 |
+
{
|
2184 |
+
"epoch": 0.73,
|
2185 |
+
"learning_rate": 3.936463049451179e-06,
|
2186 |
+
"loss": 0.1032,
|
2187 |
+
"step": 363
|
2188 |
+
},
|
2189 |
+
{
|
2190 |
+
"epoch": 0.73,
|
2191 |
+
"learning_rate": 3.884002810481959e-06,
|
2192 |
+
"loss": 0.2597,
|
2193 |
+
"step": 364
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 0.73,
|
2197 |
+
"learning_rate": 3.83181010504289e-06,
|
2198 |
+
"loss": 0.2671,
|
2199 |
+
"step": 365
|
2200 |
+
},
|
2201 |
+
{
|
2202 |
+
"epoch": 0.73,
|
2203 |
+
"learning_rate": 3.7798872162119948e-06,
|
2204 |
+
"loss": 0.2067,
|
2205 |
+
"step": 366
|
2206 |
+
},
|
2207 |
+
{
|
2208 |
+
"epoch": 0.73,
|
2209 |
+
"learning_rate": 3.72823641526463e-06,
|
2210 |
+
"loss": 0.0933,
|
2211 |
+
"step": 367
|
2212 |
+
},
|
2213 |
+
{
|
2214 |
+
"epoch": 0.74,
|
2215 |
+
"learning_rate": 3.676859961574162e-06,
|
2216 |
+
"loss": 0.0428,
|
2217 |
+
"step": 368
|
2218 |
+
},
|
2219 |
+
{
|
2220 |
+
"epoch": 0.74,
|
2221 |
+
"learning_rate": 3.625760102513103e-06,
|
2222 |
+
"loss": 0.7957,
|
2223 |
+
"step": 369
|
2224 |
+
},
|
2225 |
+
{
|
2226 |
+
"epoch": 0.74,
|
2227 |
+
"learning_rate": 3.5749390733548382e-06,
|
2228 |
+
"loss": 0.1478,
|
2229 |
+
"step": 370
|
2230 |
+
},
|
2231 |
+
{
|
2232 |
+
"epoch": 0.74,
|
2233 |
+
"learning_rate": 3.5243990971758124e-06,
|
2234 |
+
"loss": 0.3287,
|
2235 |
+
"step": 371
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 0.74,
|
2239 |
+
"learning_rate": 3.4741423847583134e-06,
|
2240 |
+
"loss": 0.1473,
|
2241 |
+
"step": 372
|
2242 |
+
},
|
2243 |
+
{
|
2244 |
+
"epoch": 0.75,
|
2245 |
+
"learning_rate": 3.4241711344937557e-06,
|
2246 |
+
"loss": 0.0655,
|
2247 |
+
"step": 373
|
2248 |
+
},
|
2249 |
+
{
|
2250 |
+
"epoch": 0.75,
|
2251 |
+
"learning_rate": 3.3744875322865035e-06,
|
2252 |
+
"loss": 0.3421,
|
2253 |
+
"step": 374
|
2254 |
+
},
|
2255 |
+
{
|
2256 |
+
"epoch": 0.75,
|
2257 |
+
"learning_rate": 3.3250937514582758e-06,
|
2258 |
+
"loss": 0.0579,
|
2259 |
+
"step": 375
|
2260 |
+
},
|
2261 |
+
{
|
2262 |
+
"epoch": 0.75,
|
2263 |
+
"learning_rate": 3.2759919526530536e-06,
|
2264 |
+
"loss": 0.4689,
|
2265 |
+
"step": 376
|
2266 |
+
},
|
2267 |
+
{
|
2268 |
+
"epoch": 0.75,
|
2269 |
+
"learning_rate": 3.2271842837425917e-06,
|
2270 |
+
"loss": 0.3456,
|
2271 |
+
"step": 377
|
2272 |
+
},
|
2273 |
+
{
|
2274 |
+
"epoch": 0.76,
|
2275 |
+
"learning_rate": 3.178672879732435e-06,
|
2276 |
+
"loss": 0.2375,
|
2277 |
+
"step": 378
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 0.76,
|
2281 |
+
"learning_rate": 3.1304598626685544e-06,
|
2282 |
+
"loss": 0.0537,
|
2283 |
+
"step": 379
|
2284 |
+
},
|
2285 |
+
{
|
2286 |
+
"epoch": 0.76,
|
2287 |
+
"learning_rate": 3.0825473415445073e-06,
|
2288 |
+
"loss": 0.368,
|
2289 |
+
"step": 380
|
2290 |
+
},
|
2291 |
+
{
|
2292 |
+
"epoch": 0.76,
|
2293 |
+
"learning_rate": 3.034937412209178e-06,
|
2294 |
+
"loss": 0.1693,
|
2295 |
+
"step": 381
|
2296 |
+
},
|
2297 |
+
{
|
2298 |
+
"epoch": 0.76,
|
2299 |
+
"learning_rate": 2.9876321572751143e-06,
|
2300 |
+
"loss": 0.0516,
|
2301 |
+
"step": 382
|
2302 |
+
},
|
2303 |
+
{
|
2304 |
+
"epoch": 0.77,
|
2305 |
+
"learning_rate": 2.9406336460274144e-06,
|
2306 |
+
"loss": 0.2091,
|
2307 |
+
"step": 383
|
2308 |
+
},
|
2309 |
+
{
|
2310 |
+
"epoch": 0.77,
|
2311 |
+
"learning_rate": 2.8939439343332086e-06,
|
2312 |
+
"loss": 0.0812,
|
2313 |
+
"step": 384
|
2314 |
+
},
|
2315 |
+
{
|
2316 |
+
"epoch": 0.77,
|
2317 |
+
"learning_rate": 2.847565064551747e-06,
|
2318 |
+
"loss": 0.5318,
|
2319 |
+
"step": 385
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 0.77,
|
2323 |
+
"learning_rate": 2.8014990654450325e-06,
|
2324 |
+
"loss": 0.1424,
|
2325 |
+
"step": 386
|
2326 |
+
},
|
2327 |
+
{
|
2328 |
+
"epoch": 0.77,
|
2329 |
+
"learning_rate": 2.7557479520891104e-06,
|
2330 |
+
"loss": 0.2126,
|
2331 |
+
"step": 387
|
2332 |
+
},
|
2333 |
+
{
|
2334 |
+
"epoch": 0.78,
|
2335 |
+
"learning_rate": 2.7103137257858867e-06,
|
2336 |
+
"loss": 0.0954,
|
2337 |
+
"step": 388
|
2338 |
+
},
|
2339 |
+
{
|
2340 |
+
"epoch": 0.78,
|
2341 |
+
"learning_rate": 2.6651983739756026e-06,
|
2342 |
+
"loss": 0.2334,
|
2343 |
+
"step": 389
|
2344 |
+
},
|
2345 |
+
{
|
2346 |
+
"epoch": 0.78,
|
2347 |
+
"learning_rate": 2.6204038701499056e-06,
|
2348 |
+
"loss": 0.2642,
|
2349 |
+
"step": 390
|
2350 |
+
},
|
2351 |
+
{
|
2352 |
+
"epoch": 0.78,
|
2353 |
+
"learning_rate": 2.5759321737655017e-06,
|
2354 |
+
"loss": 0.2249,
|
2355 |
+
"step": 391
|
2356 |
+
},
|
2357 |
+
{
|
2358 |
+
"epoch": 0.78,
|
2359 |
+
"learning_rate": 2.5317852301584642e-06,
|
2360 |
+
"loss": 0.1975,
|
2361 |
+
"step": 392
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 0.79,
|
2365 |
+
"learning_rate": 2.487964970459118e-06,
|
2366 |
+
"loss": 0.0917,
|
2367 |
+
"step": 393
|
2368 |
+
},
|
2369 |
+
{
|
2370 |
+
"epoch": 0.79,
|
2371 |
+
"learning_rate": 2.4444733115075823e-06,
|
2372 |
+
"loss": 0.0956,
|
2373 |
+
"step": 394
|
2374 |
+
},
|
2375 |
+
{
|
2376 |
+
"epoch": 0.79,
|
2377 |
+
"learning_rate": 2.401312155769916e-06,
|
2378 |
+
"loss": 0.0797,
|
2379 |
+
"step": 395
|
2380 |
+
},
|
2381 |
+
{
|
2382 |
+
"epoch": 0.79,
|
2383 |
+
"learning_rate": 2.3584833912548887e-06,
|
2384 |
+
"loss": 0.3195,
|
2385 |
+
"step": 396
|
2386 |
+
},
|
2387 |
+
{
|
2388 |
+
"epoch": 0.79,
|
2389 |
+
"learning_rate": 2.315988891431412e-06,
|
2390 |
+
"loss": 0.3054,
|
2391 |
+
"step": 397
|
2392 |
+
},
|
2393 |
+
{
|
2394 |
+
"epoch": 0.8,
|
2395 |
+
"learning_rate": 2.2738305151465646e-06,
|
2396 |
+
"loss": 0.1219,
|
2397 |
+
"step": 398
|
2398 |
+
},
|
2399 |
+
{
|
2400 |
+
"epoch": 0.8,
|
2401 |
+
"learning_rate": 2.2320101065443055e-06,
|
2402 |
+
"loss": 0.1354,
|
2403 |
+
"step": 399
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 0.8,
|
2407 |
+
"learning_rate": 2.190529494984782e-06,
|
2408 |
+
"loss": 0.1462,
|
2409 |
+
"step": 400
|
2410 |
+
},
|
2411 |
+
{
|
2412 |
+
"epoch": 0.8,
|
2413 |
+
"learning_rate": 2.149390494964323e-06,
|
2414 |
+
"loss": 0.072,
|
2415 |
+
"step": 401
|
2416 |
+
},
|
2417 |
+
{
|
2418 |
+
"epoch": 0.8,
|
2419 |
+
"learning_rate": 2.1085949060360654e-06,
|
2420 |
+
"loss": 0.4922,
|
2421 |
+
"step": 402
|
2422 |
+
},
|
2423 |
+
{
|
2424 |
+
"epoch": 0.81,
|
2425 |
+
"learning_rate": 2.0681445127312218e-06,
|
2426 |
+
"loss": 0.0753,
|
2427 |
+
"step": 403
|
2428 |
+
},
|
2429 |
+
{
|
2430 |
+
"epoch": 0.81,
|
2431 |
+
"learning_rate": 2.0280410844810426e-06,
|
2432 |
+
"loss": 0.109,
|
2433 |
+
"step": 404
|
2434 |
+
},
|
2435 |
+
{
|
2436 |
+
"epoch": 0.81,
|
2437 |
+
"learning_rate": 1.988286375539391e-06,
|
2438 |
+
"loss": 0.1175,
|
2439 |
+
"step": 405
|
2440 |
+
},
|
2441 |
+
{
|
2442 |
+
"epoch": 0.81,
|
2443 |
+
"learning_rate": 1.9488821249060297e-06,
|
2444 |
+
"loss": 0.0664,
|
2445 |
+
"step": 406
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 0.81,
|
2449 |
+
"learning_rate": 1.9098300562505266e-06,
|
2450 |
+
"loss": 0.0725,
|
2451 |
+
"step": 407
|
2452 |
+
},
|
2453 |
+
{
|
2454 |
+
"epoch": 0.82,
|
2455 |
+
"learning_rate": 1.8711318778368792e-06,
|
2456 |
+
"loss": 0.3384,
|
2457 |
+
"step": 408
|
2458 |
+
},
|
2459 |
+
{
|
2460 |
+
"epoch": 0.82,
|
2461 |
+
"learning_rate": 1.8327892824487792e-06,
|
2462 |
+
"loss": 0.1832,
|
2463 |
+
"step": 409
|
2464 |
+
},
|
2465 |
+
{
|
2466 |
+
"epoch": 0.82,
|
2467 |
+
"learning_rate": 1.7948039473155553e-06,
|
2468 |
+
"loss": 0.4133,
|
2469 |
+
"step": 410
|
2470 |
+
},
|
2471 |
+
{
|
2472 |
+
"epoch": 0.82,
|
2473 |
+
"learning_rate": 1.7571775340388275e-06,
|
2474 |
+
"loss": 0.2031,
|
2475 |
+
"step": 411
|
2476 |
+
},
|
2477 |
+
{
|
2478 |
+
"epoch": 0.82,
|
2479 |
+
"learning_rate": 1.7199116885197996e-06,
|
2480 |
+
"loss": 0.0999,
|
2481 |
+
"step": 412
|
2482 |
+
},
|
2483 |
+
{
|
2484 |
+
"epoch": 0.83,
|
2485 |
+
"learning_rate": 1.6830080408872852e-06,
|
2486 |
+
"loss": 0.1729,
|
2487 |
+
"step": 413
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 0.83,
|
2491 |
+
"learning_rate": 1.646468205426377e-06,
|
2492 |
+
"loss": 0.0638,
|
2493 |
+
"step": 414
|
2494 |
+
},
|
2495 |
+
{
|
2496 |
+
"epoch": 0.83,
|
2497 |
+
"learning_rate": 1.6102937805078544e-06,
|
2498 |
+
"loss": 0.2655,
|
2499 |
+
"step": 415
|
2500 |
+
},
|
2501 |
+
{
|
2502 |
+
"epoch": 0.83,
|
2503 |
+
"learning_rate": 1.5744863485182537e-06,
|
2504 |
+
"loss": 0.1536,
|
2505 |
+
"step": 416
|
2506 |
+
},
|
2507 |
+
{
|
2508 |
+
"epoch": 0.83,
|
2509 |
+
"learning_rate": 1.5390474757906449e-06,
|
2510 |
+
"loss": 0.3617,
|
2511 |
+
"step": 417
|
2512 |
+
},
|
2513 |
+
{
|
2514 |
+
"epoch": 0.84,
|
2515 |
+
"learning_rate": 1.5039787125361327e-06,
|
2516 |
+
"loss": 0.3315,
|
2517 |
+
"step": 418
|
2518 |
+
},
|
2519 |
+
{
|
2520 |
+
"epoch": 0.84,
|
2521 |
+
"learning_rate": 1.4692815927760274e-06,
|
2522 |
+
"loss": 0.052,
|
2523 |
+
"step": 419
|
2524 |
+
},
|
2525 |
+
{
|
2526 |
+
"epoch": 0.84,
|
2527 |
+
"learning_rate": 1.4349576342747462e-06,
|
2528 |
+
"loss": 0.2473,
|
2529 |
+
"step": 420
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 0.84,
|
2533 |
+
"learning_rate": 1.4010083384734308e-06,
|
2534 |
+
"loss": 0.0813,
|
2535 |
+
"step": 421
|
2536 |
+
},
|
2537 |
+
{
|
2538 |
+
"epoch": 0.84,
|
2539 |
+
"learning_rate": 1.367435190424261e-06,
|
2540 |
+
"loss": 0.1755,
|
2541 |
+
"step": 422
|
2542 |
+
},
|
2543 |
+
{
|
2544 |
+
"epoch": 0.85,
|
2545 |
+
"learning_rate": 1.3342396587254957e-06,
|
2546 |
+
"loss": 0.1983,
|
2547 |
+
"step": 423
|
2548 |
+
},
|
2549 |
+
{
|
2550 |
+
"epoch": 0.85,
|
2551 |
+
"learning_rate": 1.3014231954572287e-06,
|
2552 |
+
"loss": 0.1303,
|
2553 |
+
"step": 424
|
2554 |
+
},
|
2555 |
+
{
|
2556 |
+
"epoch": 0.85,
|
2557 |
+
"learning_rate": 1.26898723611787e-06,
|
2558 |
+
"loss": 0.3482,
|
2559 |
+
"step": 425
|
2560 |
+
},
|
2561 |
+
{
|
2562 |
+
"epoch": 0.85,
|
2563 |
+
"learning_rate": 1.2369331995613664e-06,
|
2564 |
+
"loss": 0.0799,
|
2565 |
+
"step": 426
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 0.85,
|
2569 |
+
"learning_rate": 1.2052624879351105e-06,
|
2570 |
+
"loss": 0.1366,
|
2571 |
+
"step": 427
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 0.86,
|
2575 |
+
"learning_rate": 1.1739764866186309e-06,
|
2576 |
+
"loss": 0.0649,
|
2577 |
+
"step": 428
|
2578 |
+
},
|
2579 |
+
{
|
2580 |
+
"epoch": 0.86,
|
2581 |
+
"learning_rate": 1.143076564162977e-06,
|
2582 |
+
"loss": 0.073,
|
2583 |
+
"step": 429
|
2584 |
+
},
|
2585 |
+
{
|
2586 |
+
"epoch": 0.86,
|
2587 |
+
"learning_rate": 1.112564072230863e-06,
|
2588 |
+
"loss": 0.1632,
|
2589 |
+
"step": 430
|
2590 |
+
},
|
2591 |
+
{
|
2592 |
+
"epoch": 0.86,
|
2593 |
+
"learning_rate": 1.0824403455375287e-06,
|
2594 |
+
"loss": 0.056,
|
2595 |
+
"step": 431
|
2596 |
+
},
|
2597 |
+
{
|
2598 |
+
"epoch": 0.86,
|
2599 |
+
"learning_rate": 1.0527067017923654e-06,
|
2600 |
+
"loss": 0.161,
|
2601 |
+
"step": 432
|
2602 |
+
},
|
2603 |
+
{
|
2604 |
+
"epoch": 0.87,
|
2605 |
+
"learning_rate": 1.0233644416412792e-06,
|
2606 |
+
"loss": 0.0551,
|
2607 |
+
"step": 433
|
2608 |
+
},
|
2609 |
+
{
|
2610 |
+
"epoch": 0.87,
|
2611 |
+
"learning_rate": 9.944148486097793e-07,
|
2612 |
+
"loss": 0.0524,
|
2613 |
+
"step": 434
|
2614 |
+
},
|
2615 |
+
{
|
2616 |
+
"epoch": 0.87,
|
2617 |
+
"learning_rate": 9.658591890468515e-07,
|
2618 |
+
"loss": 0.0555,
|
2619 |
+
"step": 435
|
2620 |
+
},
|
2621 |
+
{
|
2622 |
+
"epoch": 0.87,
|
2623 |
+
"learning_rate": 9.376987120695547e-07,
|
2624 |
+
"loss": 0.4678,
|
2625 |
+
"step": 436
|
2626 |
+
},
|
2627 |
+
{
|
2628 |
+
"epoch": 0.87,
|
2629 |
+
"learning_rate": 9.09934649508375e-07,
|
2630 |
+
"loss": 0.3581,
|
2631 |
+
"step": 437
|
2632 |
+
},
|
2633 |
+
{
|
2634 |
+
"epoch": 0.88,
|
2635 |
+
"learning_rate": 8.825682158533555e-07,
|
2636 |
+
"loss": 0.13,
|
2637 |
+
"step": 438
|
2638 |
+
},
|
2639 |
+
{
|
2640 |
+
"epoch": 0.88,
|
2641 |
+
"learning_rate": 8.55600608200956e-07,
|
2642 |
+
"loss": 0.3564,
|
2643 |
+
"step": 439
|
2644 |
+
},
|
2645 |
+
{
|
2646 |
+
"epoch": 0.88,
|
2647 |
+
"learning_rate": 8.290330062017015e-07,
|
2648 |
+
"loss": 0.2288,
|
2649 |
+
"step": 440
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 0.88,
|
2653 |
+
"learning_rate": 8.02866572008566e-07,
|
2654 |
+
"loss": 0.5636,
|
2655 |
+
"step": 441
|
2656 |
+
},
|
2657 |
+
{
|
2658 |
+
"epoch": 0.88,
|
2659 |
+
"learning_rate": 7.771024502261526e-07,
|
2660 |
+
"loss": 0.3019,
|
2661 |
+
"step": 442
|
2662 |
+
},
|
2663 |
+
{
|
2664 |
+
"epoch": 0.89,
|
2665 |
+
"learning_rate": 7.517417678606121e-07,
|
2666 |
+
"loss": 0.4403,
|
2667 |
+
"step": 443
|
2668 |
+
},
|
2669 |
+
{
|
2670 |
+
"epoch": 0.89,
|
2671 |
+
"learning_rate": 7.267856342703461e-07,
|
2672 |
+
"loss": 0.1987,
|
2673 |
+
"step": 444
|
2674 |
+
},
|
2675 |
+
{
|
2676 |
+
"epoch": 0.89,
|
2677 |
+
"learning_rate": 7.022351411174866e-07,
|
2678 |
+
"loss": 0.1701,
|
2679 |
+
"step": 445
|
2680 |
+
},
|
2681 |
+
{
|
2682 |
+
"epoch": 0.89,
|
2683 |
+
"learning_rate": 6.780913623201346e-07,
|
2684 |
+
"loss": 0.2195,
|
2685 |
+
"step": 446
|
2686 |
+
},
|
2687 |
+
{
|
2688 |
+
"epoch": 0.89,
|
2689 |
+
"learning_rate": 6.543553540053926e-07,
|
2690 |
+
"loss": 0.351,
|
2691 |
+
"step": 447
|
2692 |
+
},
|
2693 |
+
{
|
2694 |
+
"epoch": 0.9,
|
2695 |
+
"learning_rate": 6.310281544631547e-07,
|
2696 |
+
"loss": 0.0955,
|
2697 |
+
"step": 448
|
2698 |
+
},
|
2699 |
+
{
|
2700 |
+
"epoch": 0.9,
|
2701 |
+
"learning_rate": 6.081107841007006e-07,
|
2702 |
+
"loss": 0.2191,
|
2703 |
+
"step": 449
|
2704 |
+
},
|
2705 |
+
{
|
2706 |
+
"epoch": 0.9,
|
2707 |
+
"learning_rate": 5.856042453980526e-07,
|
2708 |
+
"loss": 0.2032,
|
2709 |
+
"step": 450
|
2710 |
+
},
|
2711 |
+
{
|
2712 |
+
"epoch": 0.9,
|
2713 |
+
"learning_rate": 5.63509522864123e-07,
|
2714 |
+
"loss": 0.3534,
|
2715 |
+
"step": 451
|
2716 |
+
},
|
2717 |
+
{
|
2718 |
+
"epoch": 0.9,
|
2719 |
+
"learning_rate": 5.418275829936537e-07,
|
2720 |
+
"loss": 0.1667,
|
2721 |
+
"step": 452
|
2722 |
+
},
|
2723 |
+
{
|
2724 |
+
"epoch": 0.91,
|
2725 |
+
"learning_rate": 5.205593742249326e-07,
|
2726 |
+
"loss": 0.1871,
|
2727 |
+
"step": 453
|
2728 |
+
},
|
2729 |
+
{
|
2730 |
+
"epoch": 0.91,
|
2731 |
+
"learning_rate": 4.997058268983135e-07,
|
2732 |
+
"loss": 0.3943,
|
2733 |
+
"step": 454
|
2734 |
+
},
|
2735 |
+
{
|
2736 |
+
"epoch": 0.91,
|
2737 |
+
"learning_rate": 4.792678532155115e-07,
|
2738 |
+
"loss": 0.4139,
|
2739 |
+
"step": 455
|
2740 |
+
},
|
2741 |
+
{
|
2742 |
+
"epoch": 0.91,
|
2743 |
+
"learning_rate": 4.592463471997022e-07,
|
2744 |
+
"loss": 0.0439,
|
2745 |
+
"step": 456
|
2746 |
+
},
|
2747 |
+
{
|
2748 |
+
"epoch": 0.91,
|
2749 |
+
"learning_rate": 4.396421846564236e-07,
|
2750 |
+
"loss": 0.3162,
|
2751 |
+
"step": 457
|
2752 |
+
},
|
2753 |
+
{
|
2754 |
+
"epoch": 0.92,
|
2755 |
+
"learning_rate": 4.204562231352516e-07,
|
2756 |
+
"loss": 0.2204,
|
2757 |
+
"step": 458
|
2758 |
+
},
|
2759 |
+
{
|
2760 |
+
"epoch": 0.92,
|
2761 |
+
"learning_rate": 4.016893018922996e-07,
|
2762 |
+
"loss": 0.0393,
|
2763 |
+
"step": 459
|
2764 |
+
},
|
2765 |
+
{
|
2766 |
+
"epoch": 0.92,
|
2767 |
+
"learning_rate": 3.83342241853496e-07,
|
2768 |
+
"loss": 0.0726,
|
2769 |
+
"step": 460
|
2770 |
+
},
|
2771 |
+
{
|
2772 |
+
"epoch": 0.92,
|
2773 |
+
"learning_rate": 3.6541584557868604e-07,
|
2774 |
+
"loss": 0.2546,
|
2775 |
+
"step": 461
|
2776 |
+
},
|
2777 |
+
{
|
2778 |
+
"epoch": 0.92,
|
2779 |
+
"learning_rate": 3.4791089722651437e-07,
|
2780 |
+
"loss": 0.1004,
|
2781 |
+
"step": 462
|
2782 |
+
},
|
2783 |
+
{
|
2784 |
+
"epoch": 0.93,
|
2785 |
+
"learning_rate": 3.3082816252012927e-07,
|
2786 |
+
"loss": 0.3242,
|
2787 |
+
"step": 463
|
2788 |
+
},
|
2789 |
+
{
|
2790 |
+
"epoch": 0.93,
|
2791 |
+
"learning_rate": 3.1416838871368925e-07,
|
2792 |
+
"loss": 0.1316,
|
2793 |
+
"step": 464
|
2794 |
+
},
|
2795 |
+
{
|
2796 |
+
"epoch": 0.93,
|
2797 |
+
"learning_rate": 2.9793230455966936e-07,
|
2798 |
+
"loss": 0.0682,
|
2799 |
+
"step": 465
|
2800 |
+
},
|
2801 |
+
{
|
2802 |
+
"epoch": 0.93,
|
2803 |
+
"learning_rate": 2.8212062027698995e-07,
|
2804 |
+
"loss": 0.0692,
|
2805 |
+
"step": 466
|
2806 |
+
},
|
2807 |
+
{
|
2808 |
+
"epoch": 0.93,
|
2809 |
+
"learning_rate": 2.667340275199426e-07,
|
2810 |
+
"loss": 0.2308,
|
2811 |
+
"step": 467
|
2812 |
+
},
|
2813 |
+
{
|
2814 |
+
"epoch": 0.94,
|
2815 |
+
"learning_rate": 2.5177319934793995e-07,
|
2816 |
+
"loss": 0.1639,
|
2817 |
+
"step": 468
|
2818 |
+
},
|
2819 |
+
{
|
2820 |
+
"epoch": 0.94,
|
2821 |
+
"learning_rate": 2.3723879019607376e-07,
|
2822 |
+
"loss": 0.3444,
|
2823 |
+
"step": 469
|
2824 |
+
},
|
2825 |
+
{
|
2826 |
+
"epoch": 0.94,
|
2827 |
+
"learning_rate": 2.2313143584648423e-07,
|
2828 |
+
"loss": 0.1779,
|
2829 |
+
"step": 470
|
2830 |
+
},
|
2831 |
+
{
|
2832 |
+
"epoch": 0.94,
|
2833 |
+
"learning_rate": 2.0945175340055356e-07,
|
2834 |
+
"loss": 0.3289,
|
2835 |
+
"step": 471
|
2836 |
+
},
|
2837 |
+
{
|
2838 |
+
"epoch": 0.94,
|
2839 |
+
"learning_rate": 1.9620034125190645e-07,
|
2840 |
+
"loss": 0.3088,
|
2841 |
+
"step": 472
|
2842 |
+
},
|
2843 |
+
{
|
2844 |
+
"epoch": 0.95,
|
2845 |
+
"learning_rate": 1.833777790602398e-07,
|
2846 |
+
"loss": 0.4307,
|
2847 |
+
"step": 473
|
2848 |
+
},
|
2849 |
+
{
|
2850 |
+
"epoch": 0.95,
|
2851 |
+
"learning_rate": 1.7098462772596302e-07,
|
2852 |
+
"loss": 0.233,
|
2853 |
+
"step": 474
|
2854 |
+
},
|
2855 |
+
{
|
2856 |
+
"epoch": 0.95,
|
2857 |
+
"learning_rate": 1.5902142936566333e-07,
|
2858 |
+
"loss": 0.2091,
|
2859 |
+
"step": 475
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 0.95,
|
2863 |
+
"learning_rate": 1.474887072883935e-07,
|
2864 |
+
"loss": 0.0338,
|
2865 |
+
"step": 476
|
2866 |
+
},
|
2867 |
+
{
|
2868 |
+
"epoch": 0.95,
|
2869 |
+
"learning_rate": 1.3638696597277678e-07,
|
2870 |
+
"loss": 0.2794,
|
2871 |
+
"step": 477
|
2872 |
+
},
|
2873 |
+
{
|
2874 |
+
"epoch": 0.96,
|
2875 |
+
"learning_rate": 1.2571669104494254e-07,
|
2876 |
+
"loss": 0.1669,
|
2877 |
+
"step": 478
|
2878 |
+
},
|
2879 |
+
{
|
2880 |
+
"epoch": 0.96,
|
2881 |
+
"learning_rate": 1.154783492572853e-07,
|
2882 |
+
"loss": 0.2617,
|
2883 |
+
"step": 479
|
2884 |
+
},
|
2885 |
+
{
|
2886 |
+
"epoch": 0.96,
|
2887 |
+
"learning_rate": 1.0567238846803995e-07,
|
2888 |
+
"loss": 0.1099,
|
2889 |
+
"step": 480
|
2890 |
+
},
|
2891 |
+
{
|
2892 |
+
"epoch": 0.96,
|
2893 |
+
"learning_rate": 9.629923762170091e-08,
|
2894 |
+
"loss": 0.2098,
|
2895 |
+
"step": 481
|
2896 |
+
},
|
2897 |
+
{
|
2898 |
+
"epoch": 0.96,
|
2899 |
+
"learning_rate": 8.735930673024806e-08,
|
2900 |
+
"loss": 0.13,
|
2901 |
+
"step": 482
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 0.97,
|
2905 |
+
"learning_rate": 7.885298685522235e-08,
|
2906 |
+
"loss": 0.1898,
|
2907 |
+
"step": 483
|
2908 |
+
},
|
2909 |
+
{
|
2910 |
+
"epoch": 0.97,
|
2911 |
+
"learning_rate": 7.078065009060941e-08,
|
2912 |
+
"loss": 0.0568,
|
2913 |
+
"step": 484
|
2914 |
+
},
|
2915 |
+
{
|
2916 |
+
"epoch": 0.97,
|
2917 |
+
"learning_rate": 6.314264954657257e-08,
|
2918 |
+
"loss": 0.1637,
|
2919 |
+
"step": 485
|
2920 |
+
},
|
2921 |
+
{
|
2922 |
+
"epoch": 0.97,
|
2923 |
+
"learning_rate": 5.5939319333998546e-08,
|
2924 |
+
"loss": 0.16,
|
2925 |
+
"step": 486
|
2926 |
+
},
|
2927 |
+
{
|
2928 |
+
"epoch": 0.97,
|
2929 |
+
"learning_rate": 4.9170974549885844e-08,
|
2930 |
+
"loss": 0.1184,
|
2931 |
+
"step": 487
|
2932 |
+
},
|
2933 |
+
{
|
2934 |
+
"epoch": 0.98,
|
2935 |
+
"learning_rate": 4.2837911263562406e-08,
|
2936 |
+
"loss": 0.0857,
|
2937 |
+
"step": 488
|
2938 |
+
},
|
2939 |
+
{
|
2940 |
+
"epoch": 0.98,
|
2941 |
+
"learning_rate": 3.694040650373154e-08,
|
2942 |
+
"loss": 0.2791,
|
2943 |
+
"step": 489
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 0.98,
|
2947 |
+
"learning_rate": 3.147871824635717e-08,
|
2948 |
+
"loss": 0.1298,
|
2949 |
+
"step": 490
|
2950 |
+
},
|
2951 |
+
{
|
2952 |
+
"epoch": 0.98,
|
2953 |
+
"learning_rate": 2.645308540337843e-08,
|
2954 |
+
"loss": 0.1037,
|
2955 |
+
"step": 491
|
2956 |
+
},
|
2957 |
+
{
|
2958 |
+
"epoch": 0.98,
|
2959 |
+
"learning_rate": 2.1863727812254653e-08,
|
2960 |
+
"loss": 0.0524,
|
2961 |
+
"step": 492
|
2962 |
+
},
|
2963 |
+
{
|
2964 |
+
"epoch": 0.99,
|
2965 |
+
"learning_rate": 1.7710846226355327e-08,
|
2966 |
+
"loss": 0.0859,
|
2967 |
+
"step": 493
|
2968 |
+
},
|
2969 |
+
{
|
2970 |
+
"epoch": 0.99,
|
2971 |
+
"learning_rate": 1.3994622306173766e-08,
|
2972 |
+
"loss": 0.1147,
|
2973 |
+
"step": 494
|
2974 |
+
},
|
2975 |
+
{
|
2976 |
+
"epoch": 0.99,
|
2977 |
+
"learning_rate": 1.0715218611384582e-08,
|
2978 |
+
"loss": 0.114,
|
2979 |
+
"step": 495
|
2980 |
+
},
|
2981 |
+
{
|
2982 |
+
"epoch": 0.99,
|
2983 |
+
"learning_rate": 7.872778593728258e-09,
|
2984 |
+
"loss": 0.1548,
|
2985 |
+
"step": 496
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 0.99,
|
2989 |
+
"learning_rate": 5.467426590739511e-09,
|
2990 |
+
"loss": 0.0877,
|
2991 |
+
"step": 497
|
2992 |
+
},
|
2993 |
+
{
|
2994 |
+
"epoch": 1.0,
|
2995 |
+
"learning_rate": 3.499267820307184e-09,
|
2996 |
+
"loss": 0.1934,
|
2997 |
+
"step": 498
|
2998 |
+
},
|
2999 |
+
{
|
3000 |
+
"epoch": 1.0,
|
3001 |
+
"learning_rate": 1.9683883760723834e-09,
|
3002 |
+
"loss": 0.1438,
|
3003 |
+
"step": 499
|
3004 |
+
},
|
3005 |
+
{
|
3006 |
+
"epoch": 1.0,
|
3007 |
+
"learning_rate": 8.748552236603758e-10,
|
3008 |
+
"loss": 0.0584,
|
3009 |
+
"step": 500
|
3010 |
+
},
|
3011 |
+
{
|
3012 |
+
"epoch": 1.0,
|
3013 |
+
"step": 500,
|
3014 |
+
"total_flos": 20389429714944.0,
|
3015 |
+
"train_loss": 0.23822226324118673,
|
3016 |
+
"train_runtime": 2064.6478,
|
3017 |
+
"train_samples_per_second": 0.969,
|
3018 |
+
"train_steps_per_second": 0.242
|
3019 |
+
}
|
3020 |
+
],
|
3021 |
+
"logging_steps": 1.0,
|
3022 |
+
"max_steps": 500,
|
3023 |
+
"num_input_tokens_seen": 0,
|
3024 |
+
"num_train_epochs": 1,
|
3025 |
+
"save_steps": 500,
|
3026 |
+
"total_flos": 20389429714944.0,
|
3027 |
+
"train_batch_size": 4,
|
3028 |
+
"trial_name": null,
|
3029 |
+
"trial_params": null
|
3030 |
+
}
|