Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- all_results.json +7 -7
- checkpoint-28/README.md +202 -0
- checkpoint-28/adapter_config.json +28 -0
- checkpoint-28/adapter_model.safetensors +3 -0
- checkpoint-28/optimizer.pt +3 -0
- checkpoint-28/rng_state.pth +3 -0
- checkpoint-28/scheduler.pt +3 -0
- checkpoint-28/trainer_state.json +32 -0
- checkpoint-28/training_args.bin +3 -0
- runs/Jul19_09-37-10_e192c82914bf/events.out.tfevents.1721381916.e192c82914bf.162.0 +3 -0
- train_results.json +7 -7
- trainer_state.json +14 -2289
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 544841648
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be94a4f0f91667bca6db4128038f7aef8c95cf882def6649f583e59bc60e1197
|
| 3 |
size 544841648
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 1.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"total_flos": 176151436984320.0,
|
| 4 |
+
"train_loss": 1.4319358553205217,
|
| 5 |
+
"train_runtime": 19.3836,
|
| 6 |
+
"train_samples": 13,
|
| 7 |
+
"train_samples_per_second": 2.683,
|
| 8 |
+
"train_steps_per_second": 1.445
|
| 9 |
}
|
checkpoint-28/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: google/flan-t5-xl
|
| 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 0.9.0
|
checkpoint-28/adapter_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "google/flan-t5-xl",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"loftq_config": {},
|
| 12 |
+
"lora_alpha": 32,
|
| 13 |
+
"lora_dropout": 0.1,
|
| 14 |
+
"megatron_config": null,
|
| 15 |
+
"megatron_core": "megatron.core",
|
| 16 |
+
"modules_to_save": null,
|
| 17 |
+
"peft_type": "LORA",
|
| 18 |
+
"r": 8,
|
| 19 |
+
"rank_pattern": {},
|
| 20 |
+
"revision": null,
|
| 21 |
+
"target_modules": [
|
| 22 |
+
"v",
|
| 23 |
+
"q"
|
| 24 |
+
],
|
| 25 |
+
"task_type": "SEQ_2_SEQ_LM",
|
| 26 |
+
"use_dora": false,
|
| 27 |
+
"use_rslora": false
|
| 28 |
+
}
|
checkpoint-28/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:be94a4f0f91667bca6db4128038f7aef8c95cf882def6649f583e59bc60e1197
|
| 3 |
+
size 544841648
|
checkpoint-28/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:237484bf2ac3f65aafcb20b5743bf6bd3c1c1319966528a7bb64e50db8fa0b9e
|
| 3 |
+
size 37990394
|
checkpoint-28/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec7e2b5a6196fca337c8bbaa2e63ce336abdbdbf731991872374ab9a46a8bc68
|
| 3 |
+
size 14244
|
checkpoint-28/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b4ca939c39349627ebbc4f461efada993c34387eeab65975e53a7d0b481e126f
|
| 3 |
+
size 1064
|
checkpoint-28/trainer_state.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 28,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [],
|
| 11 |
+
"logging_steps": 1000,
|
| 12 |
+
"max_steps": 28,
|
| 13 |
+
"num_input_tokens_seen": 0,
|
| 14 |
+
"num_train_epochs": 4,
|
| 15 |
+
"save_steps": 1000,
|
| 16 |
+
"stateful_callbacks": {
|
| 17 |
+
"TrainerControl": {
|
| 18 |
+
"args": {
|
| 19 |
+
"should_epoch_stop": false,
|
| 20 |
+
"should_evaluate": false,
|
| 21 |
+
"should_log": false,
|
| 22 |
+
"should_save": true,
|
| 23 |
+
"should_training_stop": true
|
| 24 |
+
},
|
| 25 |
+
"attributes": {}
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"total_flos": 176151436984320.0,
|
| 29 |
+
"train_batch_size": 2,
|
| 30 |
+
"trial_name": null,
|
| 31 |
+
"trial_params": null
|
| 32 |
+
}
|
checkpoint-28/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ce91e008637b68a0fd7e49c34c924bf637601929f32fb9ab731fcc39e6c6fec
|
| 3 |
+
size 5304
|
runs/Jul19_09-37-10_e192c82914bf/events.out.tfevents.1721381916.e192c82914bf.162.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7743f94e841177b376471868029f1502887b27008bab08ea7b54f54db8e256b4
|
| 3 |
+
size 6253
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 1.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"total_flos": 176151436984320.0,
|
| 4 |
+
"train_loss": 1.4319358553205217,
|
| 5 |
+
"train_runtime": 19.3836,
|
| 6 |
+
"train_samples": 13,
|
| 7 |
+
"train_samples_per_second": 2.683,
|
| 8 |
+
"train_steps_per_second": 1.445
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,2303 +1,28 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 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
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
"epoch": 0.006140620202640466,
|
| 20 |
-
"grad_norm": 1.6946674585342407,
|
| 21 |
-
"learning_rate": 2.0408163265306121e-07,
|
| 22 |
-
"loss": 1.4115,
|
| 23 |
-
"step": 20
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.0092109303039607,
|
| 27 |
-
"grad_norm": 2.268678903579712,
|
| 28 |
-
"learning_rate": 3.0612244897959183e-07,
|
| 29 |
-
"loss": 1.5281,
|
| 30 |
-
"step": 30
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.012281240405280933,
|
| 34 |
-
"grad_norm": 1.6933097839355469,
|
| 35 |
-
"learning_rate": 4.0816326530612243e-07,
|
| 36 |
-
"loss": 1.4233,
|
| 37 |
-
"step": 40
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.015351550506601168,
|
| 41 |
-
"grad_norm": 1.591718077659607,
|
| 42 |
-
"learning_rate": 5.10204081632653e-07,
|
| 43 |
-
"loss": 1.4581,
|
| 44 |
-
"step": 50
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.0184218606079214,
|
| 48 |
-
"grad_norm": 1.4630407094955444,
|
| 49 |
-
"learning_rate": 6.122448979591837e-07,
|
| 50 |
-
"loss": 1.5041,
|
| 51 |
-
"step": 60
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.021492170709241632,
|
| 55 |
-
"grad_norm": 1.8335317373275757,
|
| 56 |
-
"learning_rate": 7.142857142857143e-07,
|
| 57 |
-
"loss": 1.4093,
|
| 58 |
-
"step": 70
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.024562480810561865,
|
| 62 |
-
"grad_norm": 1.1747108697891235,
|
| 63 |
-
"learning_rate": 8.163265306122449e-07,
|
| 64 |
-
"loss": 1.4103,
|
| 65 |
-
"step": 80
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.0276327909118821,
|
| 69 |
-
"grad_norm": 1.8263486623764038,
|
| 70 |
-
"learning_rate": 9.183673469387755e-07,
|
| 71 |
-
"loss": 1.4307,
|
| 72 |
-
"step": 90
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.030703101013202335,
|
| 76 |
-
"grad_norm": 1.6976567506790161,
|
| 77 |
-
"learning_rate": 9.993668882557771e-07,
|
| 78 |
-
"loss": 1.3716,
|
| 79 |
-
"step": 100
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.03377341111452257,
|
| 83 |
-
"grad_norm": 1.7711292505264282,
|
| 84 |
-
"learning_rate": 9.962013295346628e-07,
|
| 85 |
-
"loss": 1.5421,
|
| 86 |
-
"step": 110
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.0368437212158428,
|
| 90 |
-
"grad_norm": 1.3698827028274536,
|
| 91 |
-
"learning_rate": 9.930357708135485e-07,
|
| 92 |
-
"loss": 1.416,
|
| 93 |
-
"step": 120
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.039914031317163035,
|
| 97 |
-
"grad_norm": 2.1234116554260254,
|
| 98 |
-
"learning_rate": 9.898702120924344e-07,
|
| 99 |
-
"loss": 1.474,
|
| 100 |
-
"step": 130
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.042984341418483264,
|
| 104 |
-
"grad_norm": 1.4592777490615845,
|
| 105 |
-
"learning_rate": 9.8670465337132e-07,
|
| 106 |
-
"loss": 1.4286,
|
| 107 |
-
"step": 140
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.0460546515198035,
|
| 111 |
-
"grad_norm": 2.3222646713256836,
|
| 112 |
-
"learning_rate": 9.835390946502057e-07,
|
| 113 |
-
"loss": 1.4442,
|
| 114 |
-
"step": 150
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.04912496162112373,
|
| 118 |
-
"grad_norm": 1.4718642234802246,
|
| 119 |
-
"learning_rate": 9.803735359290914e-07,
|
| 120 |
-
"loss": 1.3608,
|
| 121 |
-
"step": 160
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.05219527172244397,
|
| 125 |
-
"grad_norm": 1.648229718208313,
|
| 126 |
-
"learning_rate": 9.772079772079773e-07,
|
| 127 |
-
"loss": 1.5419,
|
| 128 |
-
"step": 170
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.0552655818237642,
|
| 132 |
-
"grad_norm": 2.1675705909729004,
|
| 133 |
-
"learning_rate": 9.74042418486863e-07,
|
| 134 |
-
"loss": 1.3704,
|
| 135 |
-
"step": 180
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.058335891925084433,
|
| 139 |
-
"grad_norm": 1.9648547172546387,
|
| 140 |
-
"learning_rate": 9.708768597657486e-07,
|
| 141 |
-
"loss": 1.5037,
|
| 142 |
-
"step": 190
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.06140620202640467,
|
| 146 |
-
"grad_norm": 2.2735514640808105,
|
| 147 |
-
"learning_rate": 9.677113010446345e-07,
|
| 148 |
-
"loss": 1.4135,
|
| 149 |
-
"step": 200
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.0644765121277249,
|
| 153 |
-
"grad_norm": 1.4363802671432495,
|
| 154 |
-
"learning_rate": 9.6454574232352e-07,
|
| 155 |
-
"loss": 1.4943,
|
| 156 |
-
"step": 210
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.06754682222904514,
|
| 160 |
-
"grad_norm": 1.8715194463729858,
|
| 161 |
-
"learning_rate": 9.613801836024058e-07,
|
| 162 |
-
"loss": 1.3656,
|
| 163 |
-
"step": 220
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.07061713233036537,
|
| 167 |
-
"grad_norm": 1.812849521636963,
|
| 168 |
-
"learning_rate": 9.582146248812915e-07,
|
| 169 |
-
"loss": 1.3368,
|
| 170 |
-
"step": 230
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.0736874424316856,
|
| 174 |
-
"grad_norm": 1.711365818977356,
|
| 175 |
-
"learning_rate": 9.550490661601772e-07,
|
| 176 |
-
"loss": 1.344,
|
| 177 |
-
"step": 240
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.07675775253300583,
|
| 181 |
-
"grad_norm": 1.574011206626892,
|
| 182 |
-
"learning_rate": 9.51883507439063e-07,
|
| 183 |
-
"loss": 1.2724,
|
| 184 |
-
"step": 250
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.07982806263432607,
|
| 188 |
-
"grad_norm": 1.8186664581298828,
|
| 189 |
-
"learning_rate": 9.487179487179486e-07,
|
| 190 |
-
"loss": 1.3877,
|
| 191 |
-
"step": 260
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.0828983727356463,
|
| 195 |
-
"grad_norm": 2.6951262950897217,
|
| 196 |
-
"learning_rate": 9.455523899968344e-07,
|
| 197 |
-
"loss": 1.3745,
|
| 198 |
-
"step": 270
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.08596868283696653,
|
| 202 |
-
"grad_norm": 1.8013410568237305,
|
| 203 |
-
"learning_rate": 9.423868312757202e-07,
|
| 204 |
-
"loss": 1.5363,
|
| 205 |
-
"step": 280
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.08903899293828677,
|
| 209 |
-
"grad_norm": 1.856616735458374,
|
| 210 |
-
"learning_rate": 9.392212725546059e-07,
|
| 211 |
-
"loss": 1.3532,
|
| 212 |
-
"step": 290
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.092109303039607,
|
| 216 |
-
"grad_norm": 4.05745267868042,
|
| 217 |
-
"learning_rate": 9.360557138334915e-07,
|
| 218 |
-
"loss": 1.3142,
|
| 219 |
-
"step": 300
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.09517961314092724,
|
| 223 |
-
"grad_norm": 1.9517884254455566,
|
| 224 |
-
"learning_rate": 9.328901551123773e-07,
|
| 225 |
-
"loss": 1.2781,
|
| 226 |
-
"step": 310
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.09824992324224746,
|
| 230 |
-
"grad_norm": 1.9139647483825684,
|
| 231 |
-
"learning_rate": 9.29724596391263e-07,
|
| 232 |
-
"loss": 1.3773,
|
| 233 |
-
"step": 320
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.1013202333435677,
|
| 237 |
-
"grad_norm": 2.246847152709961,
|
| 238 |
-
"learning_rate": 9.265590376701488e-07,
|
| 239 |
-
"loss": 1.3832,
|
| 240 |
-
"step": 330
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.10439054344488793,
|
| 244 |
-
"grad_norm": 1.9175080060958862,
|
| 245 |
-
"learning_rate": 9.233934789490344e-07,
|
| 246 |
-
"loss": 1.5022,
|
| 247 |
-
"step": 340
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.10746085354620817,
|
| 251 |
-
"grad_norm": 2.2456302642822266,
|
| 252 |
-
"learning_rate": 9.202279202279202e-07,
|
| 253 |
-
"loss": 1.3191,
|
| 254 |
-
"step": 350
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.1105311636475284,
|
| 258 |
-
"grad_norm": 2.491514205932617,
|
| 259 |
-
"learning_rate": 9.17062361506806e-07,
|
| 260 |
-
"loss": 1.2719,
|
| 261 |
-
"step": 360
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.11360147374884863,
|
| 265 |
-
"grad_norm": 2.05846905708313,
|
| 266 |
-
"learning_rate": 9.138968027856916e-07,
|
| 267 |
-
"loss": 1.3671,
|
| 268 |
-
"step": 370
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.11667178385016887,
|
| 272 |
-
"grad_norm": 2.2819442749023438,
|
| 273 |
-
"learning_rate": 9.107312440645773e-07,
|
| 274 |
-
"loss": 1.2372,
|
| 275 |
-
"step": 380
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.1197420939514891,
|
| 279 |
-
"grad_norm": 1.681643009185791,
|
| 280 |
-
"learning_rate": 9.075656853434631e-07,
|
| 281 |
-
"loss": 1.1861,
|
| 282 |
-
"step": 390
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 0.12281240405280934,
|
| 286 |
-
"grad_norm": 1.7323535680770874,
|
| 287 |
-
"learning_rate": 9.044001266223488e-07,
|
| 288 |
-
"loss": 1.4128,
|
| 289 |
-
"step": 400
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.12588271415412958,
|
| 293 |
-
"grad_norm": 1.7624768018722534,
|
| 294 |
-
"learning_rate": 9.012345679012346e-07,
|
| 295 |
-
"loss": 1.2962,
|
| 296 |
-
"step": 410
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.1289530242554498,
|
| 300 |
-
"grad_norm": 2.0329411029815674,
|
| 301 |
-
"learning_rate": 8.980690091801203e-07,
|
| 302 |
-
"loss": 1.3528,
|
| 303 |
-
"step": 420
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.13202333435677002,
|
| 307 |
-
"grad_norm": 2.3826355934143066,
|
| 308 |
-
"learning_rate": 8.949034504590059e-07,
|
| 309 |
-
"loss": 1.3301,
|
| 310 |
-
"step": 430
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 0.13509364445809027,
|
| 314 |
-
"grad_norm": 2.101026773452759,
|
| 315 |
-
"learning_rate": 8.917378917378917e-07,
|
| 316 |
-
"loss": 1.2829,
|
| 317 |
-
"step": 440
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 0.1381639545594105,
|
| 321 |
-
"grad_norm": 2.447965383529663,
|
| 322 |
-
"learning_rate": 8.885723330167775e-07,
|
| 323 |
-
"loss": 1.2124,
|
| 324 |
-
"step": 450
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.14123426466073075,
|
| 328 |
-
"grad_norm": 2.949786901473999,
|
| 329 |
-
"learning_rate": 8.854067742956631e-07,
|
| 330 |
-
"loss": 1.2711,
|
| 331 |
-
"step": 460
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.14430457476205097,
|
| 335 |
-
"grad_norm": 2.295283079147339,
|
| 336 |
-
"learning_rate": 8.822412155745489e-07,
|
| 337 |
-
"loss": 1.1455,
|
| 338 |
-
"step": 470
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 0.1473748848633712,
|
| 342 |
-
"grad_norm": 2.41044282913208,
|
| 343 |
-
"learning_rate": 8.790756568534346e-07,
|
| 344 |
-
"loss": 1.2558,
|
| 345 |
-
"step": 480
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.15044519496469144,
|
| 349 |
-
"grad_norm": 1.8993618488311768,
|
| 350 |
-
"learning_rate": 8.759100981323203e-07,
|
| 351 |
-
"loss": 1.2903,
|
| 352 |
-
"step": 490
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 0.15351550506601166,
|
| 356 |
-
"grad_norm": 2.800321102142334,
|
| 357 |
-
"learning_rate": 8.72744539411206e-07,
|
| 358 |
-
"loss": 1.348,
|
| 359 |
-
"step": 500
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.1565858151673319,
|
| 363 |
-
"grad_norm": 1.8116044998168945,
|
| 364 |
-
"learning_rate": 8.695789806900917e-07,
|
| 365 |
-
"loss": 1.3102,
|
| 366 |
-
"step": 510
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 0.15965612526865214,
|
| 370 |
-
"grad_norm": 2.3227620124816895,
|
| 371 |
-
"learning_rate": 8.664134219689775e-07,
|
| 372 |
-
"loss": 1.2579,
|
| 373 |
-
"step": 520
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.16272643536997236,
|
| 377 |
-
"grad_norm": 2.884206771850586,
|
| 378 |
-
"learning_rate": 8.632478632478633e-07,
|
| 379 |
-
"loss": 1.2379,
|
| 380 |
-
"step": 530
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.1657967454712926,
|
| 384 |
-
"grad_norm": 1.9436391592025757,
|
| 385 |
-
"learning_rate": 8.600823045267489e-07,
|
| 386 |
-
"loss": 1.2234,
|
| 387 |
-
"step": 540
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.16886705557261283,
|
| 391 |
-
"grad_norm": 2.247084617614746,
|
| 392 |
-
"learning_rate": 8.569167458056347e-07,
|
| 393 |
-
"loss": 1.3054,
|
| 394 |
-
"step": 550
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 0.17193736567393306,
|
| 398 |
-
"grad_norm": 1.5196994543075562,
|
| 399 |
-
"learning_rate": 8.537511870845204e-07,
|
| 400 |
-
"loss": 1.2672,
|
| 401 |
-
"step": 560
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.1750076757752533,
|
| 405 |
-
"grad_norm": 1.7843005657196045,
|
| 406 |
-
"learning_rate": 8.505856283634061e-07,
|
| 407 |
-
"loss": 1.2728,
|
| 408 |
-
"step": 570
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.17807798587657353,
|
| 412 |
-
"grad_norm": 2.1363468170166016,
|
| 413 |
-
"learning_rate": 8.474200696422918e-07,
|
| 414 |
-
"loss": 1.1388,
|
| 415 |
-
"step": 580
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.18114829597789378,
|
| 419 |
-
"grad_norm": 2.081279993057251,
|
| 420 |
-
"learning_rate": 8.442545109211775e-07,
|
| 421 |
-
"loss": 1.2837,
|
| 422 |
-
"step": 590
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 0.184218606079214,
|
| 426 |
-
"grad_norm": 2.1368634700775146,
|
| 427 |
-
"learning_rate": 8.410889522000633e-07,
|
| 428 |
-
"loss": 1.2227,
|
| 429 |
-
"step": 600
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.18728891618053423,
|
| 433 |
-
"grad_norm": 1.2492815256118774,
|
| 434 |
-
"learning_rate": 8.379233934789491e-07,
|
| 435 |
-
"loss": 1.2216,
|
| 436 |
-
"step": 610
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.19035922628185448,
|
| 440 |
-
"grad_norm": 1.7561103105545044,
|
| 441 |
-
"learning_rate": 8.347578347578346e-07,
|
| 442 |
-
"loss": 1.2899,
|
| 443 |
-
"step": 620
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.1934295363831747,
|
| 447 |
-
"grad_norm": 2.0109572410583496,
|
| 448 |
-
"learning_rate": 8.315922760367204e-07,
|
| 449 |
-
"loss": 1.2586,
|
| 450 |
-
"step": 630
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.19649984648449492,
|
| 454 |
-
"grad_norm": 1.9294732809066772,
|
| 455 |
-
"learning_rate": 8.284267173156062e-07,
|
| 456 |
-
"loss": 1.2466,
|
| 457 |
-
"step": 640
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 0.19957015658581517,
|
| 461 |
-
"grad_norm": 1.9305318593978882,
|
| 462 |
-
"learning_rate": 8.252611585944919e-07,
|
| 463 |
-
"loss": 1.2221,
|
| 464 |
-
"step": 650
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.2026404666871354,
|
| 468 |
-
"grad_norm": 1.4620651006698608,
|
| 469 |
-
"learning_rate": 8.220955998733776e-07,
|
| 470 |
-
"loss": 1.182,
|
| 471 |
-
"step": 660
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.20571077678845565,
|
| 475 |
-
"grad_norm": 2.4326343536376953,
|
| 476 |
-
"learning_rate": 8.189300411522634e-07,
|
| 477 |
-
"loss": 1.1801,
|
| 478 |
-
"step": 670
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.20878108688977587,
|
| 482 |
-
"grad_norm": 1.9332904815673828,
|
| 483 |
-
"learning_rate": 8.157644824311491e-07,
|
| 484 |
-
"loss": 1.0918,
|
| 485 |
-
"step": 680
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.2118513969910961,
|
| 489 |
-
"grad_norm": 2.345059394836426,
|
| 490 |
-
"learning_rate": 8.125989237100348e-07,
|
| 491 |
-
"loss": 1.1408,
|
| 492 |
-
"step": 690
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.21492170709241634,
|
| 496 |
-
"grad_norm": 2.2262935638427734,
|
| 497 |
-
"learning_rate": 8.094333649889204e-07,
|
| 498 |
-
"loss": 1.1688,
|
| 499 |
-
"step": 700
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.21799201719373656,
|
| 503 |
-
"grad_norm": 1.9713890552520752,
|
| 504 |
-
"learning_rate": 8.062678062678062e-07,
|
| 505 |
-
"loss": 1.1289,
|
| 506 |
-
"step": 710
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 0.2210623272950568,
|
| 510 |
-
"grad_norm": 2.036687135696411,
|
| 511 |
-
"learning_rate": 8.03102247546692e-07,
|
| 512 |
-
"loss": 1.0987,
|
| 513 |
-
"step": 720
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.22413263739637704,
|
| 517 |
-
"grad_norm": 2.875053644180298,
|
| 518 |
-
"learning_rate": 7.999366888255777e-07,
|
| 519 |
-
"loss": 1.1466,
|
| 520 |
-
"step": 730
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.22720294749769726,
|
| 524 |
-
"grad_norm": 1.509848952293396,
|
| 525 |
-
"learning_rate": 7.967711301044634e-07,
|
| 526 |
-
"loss": 1.1387,
|
| 527 |
-
"step": 740
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.2302732575990175,
|
| 531 |
-
"grad_norm": 1.844084620475769,
|
| 532 |
-
"learning_rate": 7.936055713833491e-07,
|
| 533 |
-
"loss": 1.2311,
|
| 534 |
-
"step": 750
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.23334356770033773,
|
| 538 |
-
"grad_norm": 1.9919129610061646,
|
| 539 |
-
"learning_rate": 7.904400126622348e-07,
|
| 540 |
-
"loss": 1.2468,
|
| 541 |
-
"step": 760
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.23641387780165796,
|
| 545 |
-
"grad_norm": 1.9697023630142212,
|
| 546 |
-
"learning_rate": 7.872744539411206e-07,
|
| 547 |
-
"loss": 1.1679,
|
| 548 |
-
"step": 770
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.2394841879029782,
|
| 552 |
-
"grad_norm": 1.8078380823135376,
|
| 553 |
-
"learning_rate": 7.841088952200063e-07,
|
| 554 |
-
"loss": 1.1574,
|
| 555 |
-
"step": 780
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.24255449800429843,
|
| 559 |
-
"grad_norm": 1.8240193128585815,
|
| 560 |
-
"learning_rate": 7.80943336498892e-07,
|
| 561 |
-
"loss": 1.2098,
|
| 562 |
-
"step": 790
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 0.24562480810561868,
|
| 566 |
-
"grad_norm": 2.1345813274383545,
|
| 567 |
-
"learning_rate": 7.777777777777778e-07,
|
| 568 |
-
"loss": 1.1673,
|
| 569 |
-
"step": 800
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.2486951182069389,
|
| 573 |
-
"grad_norm": 2.6286094188690186,
|
| 574 |
-
"learning_rate": 7.746122190566636e-07,
|
| 575 |
-
"loss": 1.1209,
|
| 576 |
-
"step": 810
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.25176542830825915,
|
| 580 |
-
"grad_norm": 2.819676637649536,
|
| 581 |
-
"learning_rate": 7.714466603355491e-07,
|
| 582 |
-
"loss": 1.1418,
|
| 583 |
-
"step": 820
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.25483573840957935,
|
| 587 |
-
"grad_norm": 3.308809995651245,
|
| 588 |
-
"learning_rate": 7.682811016144349e-07,
|
| 589 |
-
"loss": 1.0108,
|
| 590 |
-
"step": 830
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.2579060485108996,
|
| 594 |
-
"grad_norm": 2.4087624549865723,
|
| 595 |
-
"learning_rate": 7.651155428933206e-07,
|
| 596 |
-
"loss": 1.1166,
|
| 597 |
-
"step": 840
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.26097635861221985,
|
| 601 |
-
"grad_norm": 2.6052870750427246,
|
| 602 |
-
"learning_rate": 7.619499841722064e-07,
|
| 603 |
-
"loss": 1.1701,
|
| 604 |
-
"step": 850
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.26404666871354004,
|
| 608 |
-
"grad_norm": 1.932104468345642,
|
| 609 |
-
"learning_rate": 7.587844254510921e-07,
|
| 610 |
-
"loss": 1.1451,
|
| 611 |
-
"step": 860
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.2671169788148603,
|
| 615 |
-
"grad_norm": 1.9061886072158813,
|
| 616 |
-
"learning_rate": 7.556188667299778e-07,
|
| 617 |
-
"loss": 1.2128,
|
| 618 |
-
"step": 870
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 0.27018728891618055,
|
| 622 |
-
"grad_norm": 2.2594542503356934,
|
| 623 |
-
"learning_rate": 7.524533080088635e-07,
|
| 624 |
-
"loss": 1.2324,
|
| 625 |
-
"step": 880
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.27325759901750074,
|
| 629 |
-
"grad_norm": 2.510110855102539,
|
| 630 |
-
"learning_rate": 7.492877492877493e-07,
|
| 631 |
-
"loss": 1.1386,
|
| 632 |
-
"step": 890
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 0.276327909118821,
|
| 636 |
-
"grad_norm": 2.064319610595703,
|
| 637 |
-
"learning_rate": 7.461221905666349e-07,
|
| 638 |
-
"loss": 1.1055,
|
| 639 |
-
"step": 900
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.27939821922014124,
|
| 643 |
-
"grad_norm": 2.702468156814575,
|
| 644 |
-
"learning_rate": 7.429566318455207e-07,
|
| 645 |
-
"loss": 1.174,
|
| 646 |
-
"step": 910
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 0.2824685293214615,
|
| 650 |
-
"grad_norm": 2.1893582344055176,
|
| 651 |
-
"learning_rate": 7.397910731244065e-07,
|
| 652 |
-
"loss": 1.0903,
|
| 653 |
-
"step": 920
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.2855388394227817,
|
| 657 |
-
"grad_norm": 3.4419305324554443,
|
| 658 |
-
"learning_rate": 7.366255144032922e-07,
|
| 659 |
-
"loss": 1.0286,
|
| 660 |
-
"step": 930
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 0.28860914952410194,
|
| 664 |
-
"grad_norm": 1.8508418798446655,
|
| 665 |
-
"learning_rate": 7.334599556821778e-07,
|
| 666 |
-
"loss": 1.2314,
|
| 667 |
-
"step": 940
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 0.2916794596254222,
|
| 671 |
-
"grad_norm": 2.510098457336426,
|
| 672 |
-
"learning_rate": 7.302943969610635e-07,
|
| 673 |
-
"loss": 1.1322,
|
| 674 |
-
"step": 950
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 0.2947497697267424,
|
| 678 |
-
"grad_norm": 2.328416347503662,
|
| 679 |
-
"learning_rate": 7.271288382399493e-07,
|
| 680 |
-
"loss": 1.025,
|
| 681 |
-
"step": 960
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 0.29782007982806263,
|
| 685 |
-
"grad_norm": 2.1068317890167236,
|
| 686 |
-
"learning_rate": 7.239632795188351e-07,
|
| 687 |
-
"loss": 1.0147,
|
| 688 |
-
"step": 970
|
| 689 |
-
},
|
| 690 |
-
{
|
| 691 |
-
"epoch": 0.3008903899293829,
|
| 692 |
-
"grad_norm": 1.842786192893982,
|
| 693 |
-
"learning_rate": 7.207977207977207e-07,
|
| 694 |
-
"loss": 1.0978,
|
| 695 |
-
"step": 980
|
| 696 |
-
},
|
| 697 |
-
{
|
| 698 |
-
"epoch": 0.3039607000307031,
|
| 699 |
-
"grad_norm": 1.274963140487671,
|
| 700 |
-
"learning_rate": 7.176321620766065e-07,
|
| 701 |
-
"loss": 0.9776,
|
| 702 |
-
"step": 990
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 0.30703101013202333,
|
| 706 |
-
"grad_norm": 2.67863392829895,
|
| 707 |
-
"learning_rate": 7.144666033554923e-07,
|
| 708 |
-
"loss": 1.1785,
|
| 709 |
-
"step": 1000
|
| 710 |
-
},
|
| 711 |
-
{
|
| 712 |
-
"epoch": 0.3101013202333436,
|
| 713 |
-
"grad_norm": 1.860278844833374,
|
| 714 |
-
"learning_rate": 7.113010446343779e-07,
|
| 715 |
-
"loss": 1.2182,
|
| 716 |
-
"step": 1010
|
| 717 |
-
},
|
| 718 |
-
{
|
| 719 |
-
"epoch": 0.3131716303346638,
|
| 720 |
-
"grad_norm": 1.9876034259796143,
|
| 721 |
-
"learning_rate": 7.081354859132636e-07,
|
| 722 |
-
"loss": 1.166,
|
| 723 |
-
"step": 1020
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 0.316241940435984,
|
| 727 |
-
"grad_norm": 2.1570093631744385,
|
| 728 |
-
"learning_rate": 7.049699271921494e-07,
|
| 729 |
-
"loss": 1.0908,
|
| 730 |
-
"step": 1030
|
| 731 |
-
},
|
| 732 |
-
{
|
| 733 |
-
"epoch": 0.3193122505373043,
|
| 734 |
-
"grad_norm": 3.1807639598846436,
|
| 735 |
-
"learning_rate": 7.018043684710351e-07,
|
| 736 |
-
"loss": 1.0003,
|
| 737 |
-
"step": 1040
|
| 738 |
-
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 0.3223825606386245,
|
| 741 |
-
"grad_norm": 3.396947145462036,
|
| 742 |
-
"learning_rate": 6.986388097499209e-07,
|
| 743 |
-
"loss": 1.1049,
|
| 744 |
-
"step": 1050
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 0.3254528707399447,
|
| 748 |
-
"grad_norm": 2.2146832942962646,
|
| 749 |
-
"learning_rate": 6.954732510288066e-07,
|
| 750 |
-
"loss": 1.0274,
|
| 751 |
-
"step": 1060
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.328523180841265,
|
| 755 |
-
"grad_norm": 1.499233365058899,
|
| 756 |
-
"learning_rate": 6.923076923076922e-07,
|
| 757 |
-
"loss": 0.9735,
|
| 758 |
-
"step": 1070
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.3315934909425852,
|
| 762 |
-
"grad_norm": 1.6564570665359497,
|
| 763 |
-
"learning_rate": 6.89142133586578e-07,
|
| 764 |
-
"loss": 1.1102,
|
| 765 |
-
"step": 1080
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 0.3346638010439054,
|
| 769 |
-
"grad_norm": 1.7835310697555542,
|
| 770 |
-
"learning_rate": 6.859765748654637e-07,
|
| 771 |
-
"loss": 1.0519,
|
| 772 |
-
"step": 1090
|
| 773 |
-
},
|
| 774 |
-
{
|
| 775 |
-
"epoch": 0.33773411114522567,
|
| 776 |
-
"grad_norm": 2.079343795776367,
|
| 777 |
-
"learning_rate": 6.828110161443494e-07,
|
| 778 |
-
"loss": 1.0187,
|
| 779 |
-
"step": 1100
|
| 780 |
-
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 0.3408044212465459,
|
| 783 |
-
"grad_norm": 2.1256511211395264,
|
| 784 |
-
"learning_rate": 6.796454574232352e-07,
|
| 785 |
-
"loss": 1.0023,
|
| 786 |
-
"step": 1110
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 0.3438747313478661,
|
| 790 |
-
"grad_norm": 2.375253915786743,
|
| 791 |
-
"learning_rate": 6.764798987021209e-07,
|
| 792 |
-
"loss": 1.2046,
|
| 793 |
-
"step": 1120
|
| 794 |
-
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 0.34694504144918636,
|
| 797 |
-
"grad_norm": 2.2755587100982666,
|
| 798 |
-
"learning_rate": 6.733143399810067e-07,
|
| 799 |
-
"loss": 1.1124,
|
| 800 |
-
"step": 1130
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 0.3500153515505066,
|
| 804 |
-
"grad_norm": 1.930176019668579,
|
| 805 |
-
"learning_rate": 6.701487812598923e-07,
|
| 806 |
-
"loss": 0.9806,
|
| 807 |
-
"step": 1140
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 0.3530856616518268,
|
| 811 |
-
"grad_norm": 1.805470585823059,
|
| 812 |
-
"learning_rate": 6.66983222538778e-07,
|
| 813 |
-
"loss": 1.0932,
|
| 814 |
-
"step": 1150
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 0.35615597175314706,
|
| 818 |
-
"grad_norm": 1.877125859260559,
|
| 819 |
-
"learning_rate": 6.638176638176638e-07,
|
| 820 |
-
"loss": 1.0703,
|
| 821 |
-
"step": 1160
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 0.3592262818544673,
|
| 825 |
-
"grad_norm": 1.6196848154067993,
|
| 826 |
-
"learning_rate": 6.606521050965496e-07,
|
| 827 |
-
"loss": 1.14,
|
| 828 |
-
"step": 1170
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 0.36229659195578756,
|
| 832 |
-
"grad_norm": 2.2336461544036865,
|
| 833 |
-
"learning_rate": 6.574865463754352e-07,
|
| 834 |
-
"loss": 1.0706,
|
| 835 |
-
"step": 1180
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 0.36536690205710776,
|
| 839 |
-
"grad_norm": 2.2776904106140137,
|
| 840 |
-
"learning_rate": 6.54320987654321e-07,
|
| 841 |
-
"loss": 0.9214,
|
| 842 |
-
"step": 1190
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 0.368437212158428,
|
| 846 |
-
"grad_norm": 1.7691733837127686,
|
| 847 |
-
"learning_rate": 6.511554289332066e-07,
|
| 848 |
-
"loss": 1.0746,
|
| 849 |
-
"step": 1200
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 0.37150752225974826,
|
| 853 |
-
"grad_norm": 2.1708719730377197,
|
| 854 |
-
"learning_rate": 6.479898702120924e-07,
|
| 855 |
-
"loss": 1.0607,
|
| 856 |
-
"step": 1210
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 0.37457783236106845,
|
| 860 |
-
"grad_norm": 1.6801633834838867,
|
| 861 |
-
"learning_rate": 6.448243114909781e-07,
|
| 862 |
-
"loss": 1.0381,
|
| 863 |
-
"step": 1220
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 0.3776481424623887,
|
| 867 |
-
"grad_norm": 2.653822422027588,
|
| 868 |
-
"learning_rate": 6.416587527698638e-07,
|
| 869 |
-
"loss": 1.0508,
|
| 870 |
-
"step": 1230
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 0.38071845256370895,
|
| 874 |
-
"grad_norm": 2.0415942668914795,
|
| 875 |
-
"learning_rate": 6.384931940487496e-07,
|
| 876 |
-
"loss": 1.0113,
|
| 877 |
-
"step": 1240
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 0.38378876266502915,
|
| 881 |
-
"grad_norm": 1.906172752380371,
|
| 882 |
-
"learning_rate": 6.353276353276354e-07,
|
| 883 |
-
"loss": 0.9856,
|
| 884 |
-
"step": 1250
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 0.3868590727663494,
|
| 888 |
-
"grad_norm": 1.486282229423523,
|
| 889 |
-
"learning_rate": 6.321620766065209e-07,
|
| 890 |
-
"loss": 0.9518,
|
| 891 |
-
"step": 1260
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 0.38992938286766965,
|
| 895 |
-
"grad_norm": 1.6384541988372803,
|
| 896 |
-
"learning_rate": 6.289965178854067e-07,
|
| 897 |
-
"loss": 0.9473,
|
| 898 |
-
"step": 1270
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 0.39299969296898984,
|
| 902 |
-
"grad_norm": 1.7707164287567139,
|
| 903 |
-
"learning_rate": 6.258309591642925e-07,
|
| 904 |
-
"loss": 1.0343,
|
| 905 |
-
"step": 1280
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 0.3960700030703101,
|
| 909 |
-
"grad_norm": 2.7890567779541016,
|
| 910 |
-
"learning_rate": 6.226654004431782e-07,
|
| 911 |
-
"loss": 0.9901,
|
| 912 |
-
"step": 1290
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 0.39914031317163035,
|
| 916 |
-
"grad_norm": 3.0801589488983154,
|
| 917 |
-
"learning_rate": 6.194998417220639e-07,
|
| 918 |
-
"loss": 0.9611,
|
| 919 |
-
"step": 1300
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 0.40221062327295054,
|
| 923 |
-
"grad_norm": 1.4281337261199951,
|
| 924 |
-
"learning_rate": 6.163342830009497e-07,
|
| 925 |
-
"loss": 1.0773,
|
| 926 |
-
"step": 1310
|
| 927 |
-
},
|
| 928 |
-
{
|
| 929 |
-
"epoch": 0.4052809333742708,
|
| 930 |
-
"grad_norm": 2.135127544403076,
|
| 931 |
-
"learning_rate": 6.131687242798354e-07,
|
| 932 |
-
"loss": 0.9285,
|
| 933 |
-
"step": 1320
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 0.40835124347559104,
|
| 937 |
-
"grad_norm": 2.215280532836914,
|
| 938 |
-
"learning_rate": 6.100031655587211e-07,
|
| 939 |
-
"loss": 0.9311,
|
| 940 |
-
"step": 1330
|
| 941 |
-
},
|
| 942 |
-
{
|
| 943 |
-
"epoch": 0.4114215535769113,
|
| 944 |
-
"grad_norm": 1.7742196321487427,
|
| 945 |
-
"learning_rate": 6.068376068376067e-07,
|
| 946 |
-
"loss": 0.9504,
|
| 947 |
-
"step": 1340
|
| 948 |
-
},
|
| 949 |
-
{
|
| 950 |
-
"epoch": 0.4144918636782315,
|
| 951 |
-
"grad_norm": 1.6954290866851807,
|
| 952 |
-
"learning_rate": 6.036720481164925e-07,
|
| 953 |
-
"loss": 0.9653,
|
| 954 |
-
"step": 1350
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 0.41756217377955174,
|
| 958 |
-
"grad_norm": 2.5020253658294678,
|
| 959 |
-
"learning_rate": 6.005064893953783e-07,
|
| 960 |
-
"loss": 0.8404,
|
| 961 |
-
"step": 1360
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 0.420632483880872,
|
| 965 |
-
"grad_norm": 1.6127808094024658,
|
| 966 |
-
"learning_rate": 5.97340930674264e-07,
|
| 967 |
-
"loss": 1.0783,
|
| 968 |
-
"step": 1370
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 0.4237027939821922,
|
| 972 |
-
"grad_norm": 1.9128029346466064,
|
| 973 |
-
"learning_rate": 5.941753719531497e-07,
|
| 974 |
-
"loss": 0.9965,
|
| 975 |
-
"step": 1380
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 0.42677310408351243,
|
| 979 |
-
"grad_norm": 1.7032874822616577,
|
| 980 |
-
"learning_rate": 5.910098132320354e-07,
|
| 981 |
-
"loss": 0.9816,
|
| 982 |
-
"step": 1390
|
| 983 |
-
},
|
| 984 |
-
{
|
| 985 |
-
"epoch": 0.4298434141848327,
|
| 986 |
-
"grad_norm": 2.7620840072631836,
|
| 987 |
-
"learning_rate": 5.878442545109211e-07,
|
| 988 |
-
"loss": 0.9708,
|
| 989 |
-
"step": 1400
|
| 990 |
-
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 0.4329137242861529,
|
| 993 |
-
"grad_norm": 1.8601688146591187,
|
| 994 |
-
"learning_rate": 5.846786957898069e-07,
|
| 995 |
-
"loss": 1.0241,
|
| 996 |
-
"step": 1410
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 0.43598403438747313,
|
| 1000 |
-
"grad_norm": 2.5212254524230957,
|
| 1001 |
-
"learning_rate": 5.815131370686926e-07,
|
| 1002 |
-
"loss": 0.9622,
|
| 1003 |
-
"step": 1420
|
| 1004 |
-
},
|
| 1005 |
-
{
|
| 1006 |
-
"epoch": 0.4390543444887934,
|
| 1007 |
-
"grad_norm": 1.5986229181289673,
|
| 1008 |
-
"learning_rate": 5.783475783475783e-07,
|
| 1009 |
-
"loss": 1.018,
|
| 1010 |
-
"step": 1430
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 0.4421246545901136,
|
| 1014 |
-
"grad_norm": 1.858396053314209,
|
| 1015 |
-
"learning_rate": 5.751820196264641e-07,
|
| 1016 |
-
"loss": 0.9908,
|
| 1017 |
-
"step": 1440
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 0.4451949646914338,
|
| 1021 |
-
"grad_norm": 1.8508849143981934,
|
| 1022 |
-
"learning_rate": 5.720164609053498e-07,
|
| 1023 |
-
"loss": 0.95,
|
| 1024 |
-
"step": 1450
|
| 1025 |
-
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 0.4482652747927541,
|
| 1028 |
-
"grad_norm": 1.8969439268112183,
|
| 1029 |
-
"learning_rate": 5.688509021842354e-07,
|
| 1030 |
-
"loss": 0.9342,
|
| 1031 |
-
"step": 1460
|
| 1032 |
-
},
|
| 1033 |
-
{
|
| 1034 |
-
"epoch": 0.4513355848940743,
|
| 1035 |
-
"grad_norm": 1.766987919807434,
|
| 1036 |
-
"learning_rate": 5.656853434631212e-07,
|
| 1037 |
-
"loss": 1.0226,
|
| 1038 |
-
"step": 1470
|
| 1039 |
-
},
|
| 1040 |
-
{
|
| 1041 |
-
"epoch": 0.4544058949953945,
|
| 1042 |
-
"grad_norm": 1.9654399156570435,
|
| 1043 |
-
"learning_rate": 5.625197847420069e-07,
|
| 1044 |
-
"loss": 0.8694,
|
| 1045 |
-
"step": 1480
|
| 1046 |
-
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 0.45747620509671477,
|
| 1049 |
-
"grad_norm": 2.982687473297119,
|
| 1050 |
-
"learning_rate": 5.593542260208927e-07,
|
| 1051 |
-
"loss": 0.8769,
|
| 1052 |
-
"step": 1490
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 0.460546515198035,
|
| 1056 |
-
"grad_norm": 2.498600721359253,
|
| 1057 |
-
"learning_rate": 5.561886672997784e-07,
|
| 1058 |
-
"loss": 1.0117,
|
| 1059 |
-
"step": 1500
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 0.4636168252993552,
|
| 1063 |
-
"grad_norm": 2.473336696624756,
|
| 1064 |
-
"learning_rate": 5.530231085786641e-07,
|
| 1065 |
-
"loss": 0.8705,
|
| 1066 |
-
"step": 1510
|
| 1067 |
-
},
|
| 1068 |
-
{
|
| 1069 |
-
"epoch": 0.46668713540067547,
|
| 1070 |
-
"grad_norm": 2.05454421043396,
|
| 1071 |
-
"learning_rate": 5.498575498575498e-07,
|
| 1072 |
-
"loss": 0.9998,
|
| 1073 |
-
"step": 1520
|
| 1074 |
-
},
|
| 1075 |
-
{
|
| 1076 |
-
"epoch": 0.4697574455019957,
|
| 1077 |
-
"grad_norm": 2.6380317211151123,
|
| 1078 |
-
"learning_rate": 5.466919911364356e-07,
|
| 1079 |
-
"loss": 0.9217,
|
| 1080 |
-
"step": 1530
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 0.4728277556033159,
|
| 1084 |
-
"grad_norm": 3.275218963623047,
|
| 1085 |
-
"learning_rate": 5.435264324153212e-07,
|
| 1086 |
-
"loss": 0.932,
|
| 1087 |
-
"step": 1540
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 0.47589806570463616,
|
| 1091 |
-
"grad_norm": 1.7377240657806396,
|
| 1092 |
-
"learning_rate": 5.40360873694207e-07,
|
| 1093 |
-
"loss": 0.9129,
|
| 1094 |
-
"step": 1550
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 0.4789683758059564,
|
| 1098 |
-
"grad_norm": 2.8902907371520996,
|
| 1099 |
-
"learning_rate": 5.371953149730927e-07,
|
| 1100 |
-
"loss": 0.9501,
|
| 1101 |
-
"step": 1560
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 0.4820386859072766,
|
| 1105 |
-
"grad_norm": 1.725958228111267,
|
| 1106 |
-
"learning_rate": 5.340297562519785e-07,
|
| 1107 |
-
"loss": 0.9022,
|
| 1108 |
-
"step": 1570
|
| 1109 |
-
},
|
| 1110 |
-
{
|
| 1111 |
-
"epoch": 0.48510899600859686,
|
| 1112 |
-
"grad_norm": 1.855343222618103,
|
| 1113 |
-
"learning_rate": 5.308641975308642e-07,
|
| 1114 |
-
"loss": 0.9266,
|
| 1115 |
-
"step": 1580
|
| 1116 |
-
},
|
| 1117 |
-
{
|
| 1118 |
-
"epoch": 0.4881793061099171,
|
| 1119 |
-
"grad_norm": 1.868709921836853,
|
| 1120 |
-
"learning_rate": 5.276986388097498e-07,
|
| 1121 |
-
"loss": 0.8821,
|
| 1122 |
-
"step": 1590
|
| 1123 |
-
},
|
| 1124 |
-
{
|
| 1125 |
-
"epoch": 0.49124961621123736,
|
| 1126 |
-
"grad_norm": 3.291313886642456,
|
| 1127 |
-
"learning_rate": 5.245330800886356e-07,
|
| 1128 |
-
"loss": 0.9109,
|
| 1129 |
-
"step": 1600
|
| 1130 |
-
},
|
| 1131 |
-
{
|
| 1132 |
-
"epoch": 0.49431992631255756,
|
| 1133 |
-
"grad_norm": 2.1387109756469727,
|
| 1134 |
-
"learning_rate": 5.213675213675214e-07,
|
| 1135 |
-
"loss": 0.9324,
|
| 1136 |
-
"step": 1610
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 0.4973902364138778,
|
| 1140 |
-
"grad_norm": 1.3971613645553589,
|
| 1141 |
-
"learning_rate": 5.18201962646407e-07,
|
| 1142 |
-
"loss": 0.9861,
|
| 1143 |
-
"step": 1620
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 0.500460546515198,
|
| 1147 |
-
"grad_norm": 1.6886439323425293,
|
| 1148 |
-
"learning_rate": 5.150364039252928e-07,
|
| 1149 |
-
"loss": 0.8934,
|
| 1150 |
-
"step": 1630
|
| 1151 |
-
},
|
| 1152 |
-
{
|
| 1153 |
-
"epoch": 0.5035308566165183,
|
| 1154 |
-
"grad_norm": 2.2841763496398926,
|
| 1155 |
-
"learning_rate": 5.118708452041786e-07,
|
| 1156 |
-
"loss": 0.9155,
|
| 1157 |
-
"step": 1640
|
| 1158 |
-
},
|
| 1159 |
-
{
|
| 1160 |
-
"epoch": 0.5066011667178385,
|
| 1161 |
-
"grad_norm": 3.6136739253997803,
|
| 1162 |
-
"learning_rate": 5.087052864830642e-07,
|
| 1163 |
-
"loss": 0.8794,
|
| 1164 |
-
"step": 1650
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 0.5096714768191587,
|
| 1168 |
-
"grad_norm": 2.126060962677002,
|
| 1169 |
-
"learning_rate": 5.055397277619499e-07,
|
| 1170 |
-
"loss": 0.897,
|
| 1171 |
-
"step": 1660
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 0.512741786920479,
|
| 1175 |
-
"grad_norm": 1.4978396892547607,
|
| 1176 |
-
"learning_rate": 5.023741690408357e-07,
|
| 1177 |
-
"loss": 0.8214,
|
| 1178 |
-
"step": 1670
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 0.5158120970217992,
|
| 1182 |
-
"grad_norm": 1.483849287033081,
|
| 1183 |
-
"learning_rate": 4.992086103197214e-07,
|
| 1184 |
-
"loss": 0.9382,
|
| 1185 |
-
"step": 1680
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 0.5188824071231194,
|
| 1189 |
-
"grad_norm": 1.8234091997146606,
|
| 1190 |
-
"learning_rate": 4.960430515986072e-07,
|
| 1191 |
-
"loss": 1.016,
|
| 1192 |
-
"step": 1690
|
| 1193 |
-
},
|
| 1194 |
-
{
|
| 1195 |
-
"epoch": 0.5219527172244397,
|
| 1196 |
-
"grad_norm": 1.4991623163223267,
|
| 1197 |
-
"learning_rate": 4.928774928774928e-07,
|
| 1198 |
-
"loss": 0.9616,
|
| 1199 |
-
"step": 1700
|
| 1200 |
-
},
|
| 1201 |
-
{
|
| 1202 |
-
"epoch": 0.5250230273257599,
|
| 1203 |
-
"grad_norm": 2.0311944484710693,
|
| 1204 |
-
"learning_rate": 4.897119341563785e-07,
|
| 1205 |
-
"loss": 0.8459,
|
| 1206 |
-
"step": 1710
|
| 1207 |
-
},
|
| 1208 |
-
{
|
| 1209 |
-
"epoch": 0.5280933374270801,
|
| 1210 |
-
"grad_norm": 1.9169743061065674,
|
| 1211 |
-
"learning_rate": 4.865463754352643e-07,
|
| 1212 |
-
"loss": 0.8369,
|
| 1213 |
-
"step": 1720
|
| 1214 |
-
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 0.5311636475284004,
|
| 1217 |
-
"grad_norm": 2.482090473175049,
|
| 1218 |
-
"learning_rate": 4.833808167141501e-07,
|
| 1219 |
-
"loss": 0.8837,
|
| 1220 |
-
"step": 1730
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 0.5342339576297206,
|
| 1224 |
-
"grad_norm": 2.389390230178833,
|
| 1225 |
-
"learning_rate": 4.802152579930357e-07,
|
| 1226 |
-
"loss": 0.8079,
|
| 1227 |
-
"step": 1740
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 0.5373042677310408,
|
| 1231 |
-
"grad_norm": 1.729292392730713,
|
| 1232 |
-
"learning_rate": 4.770496992719214e-07,
|
| 1233 |
-
"loss": 0.9238,
|
| 1234 |
-
"step": 1750
|
| 1235 |
-
},
|
| 1236 |
-
{
|
| 1237 |
-
"epoch": 0.5403745778323611,
|
| 1238 |
-
"grad_norm": 2.0271520614624023,
|
| 1239 |
-
"learning_rate": 4.738841405508072e-07,
|
| 1240 |
-
"loss": 0.8741,
|
| 1241 |
-
"step": 1760
|
| 1242 |
-
},
|
| 1243 |
-
{
|
| 1244 |
-
"epoch": 0.5434448879336813,
|
| 1245 |
-
"grad_norm": 1.5236482620239258,
|
| 1246 |
-
"learning_rate": 4.707185818296929e-07,
|
| 1247 |
-
"loss": 0.9811,
|
| 1248 |
-
"step": 1770
|
| 1249 |
-
},
|
| 1250 |
-
{
|
| 1251 |
-
"epoch": 0.5465151980350015,
|
| 1252 |
-
"grad_norm": 1.9456485509872437,
|
| 1253 |
-
"learning_rate": 4.6755302310857865e-07,
|
| 1254 |
-
"loss": 0.8036,
|
| 1255 |
-
"step": 1780
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 0.5495855081363218,
|
| 1259 |
-
"grad_norm": 1.8617053031921387,
|
| 1260 |
-
"learning_rate": 4.643874643874643e-07,
|
| 1261 |
-
"loss": 0.9359,
|
| 1262 |
-
"step": 1790
|
| 1263 |
-
},
|
| 1264 |
-
{
|
| 1265 |
-
"epoch": 0.552655818237642,
|
| 1266 |
-
"grad_norm": 1.4658828973770142,
|
| 1267 |
-
"learning_rate": 4.612219056663501e-07,
|
| 1268 |
-
"loss": 1.0115,
|
| 1269 |
-
"step": 1800
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 0.5557261283389623,
|
| 1273 |
-
"grad_norm": 3.7697947025299072,
|
| 1274 |
-
"learning_rate": 4.580563469452358e-07,
|
| 1275 |
-
"loss": 0.9872,
|
| 1276 |
-
"step": 1810
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 0.5587964384402825,
|
| 1280 |
-
"grad_norm": 1.5588624477386475,
|
| 1281 |
-
"learning_rate": 4.5489078822412155e-07,
|
| 1282 |
-
"loss": 0.885,
|
| 1283 |
-
"step": 1820
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 0.5618667485416027,
|
| 1287 |
-
"grad_norm": 2.504701852798462,
|
| 1288 |
-
"learning_rate": 4.5172522950300727e-07,
|
| 1289 |
-
"loss": 0.8898,
|
| 1290 |
-
"step": 1830
|
| 1291 |
-
},
|
| 1292 |
-
{
|
| 1293 |
-
"epoch": 0.564937058642923,
|
| 1294 |
-
"grad_norm": 1.5156068801879883,
|
| 1295 |
-
"learning_rate": 4.48559670781893e-07,
|
| 1296 |
-
"loss": 0.8852,
|
| 1297 |
-
"step": 1840
|
| 1298 |
-
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 0.5680073687442432,
|
| 1301 |
-
"grad_norm": 1.6788108348846436,
|
| 1302 |
-
"learning_rate": 4.453941120607787e-07,
|
| 1303 |
-
"loss": 0.9429,
|
| 1304 |
-
"step": 1850
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 0.5710776788455634,
|
| 1308 |
-
"grad_norm": 1.6702663898468018,
|
| 1309 |
-
"learning_rate": 4.422285533396644e-07,
|
| 1310 |
-
"loss": 0.8841,
|
| 1311 |
-
"step": 1860
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 0.5741479889468837,
|
| 1315 |
-
"grad_norm": 1.7993017435073853,
|
| 1316 |
-
"learning_rate": 4.390629946185502e-07,
|
| 1317 |
-
"loss": 0.8859,
|
| 1318 |
-
"step": 1870
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 0.5772182990482039,
|
| 1322 |
-
"grad_norm": 1.976618766784668,
|
| 1323 |
-
"learning_rate": 4.358974358974359e-07,
|
| 1324 |
-
"loss": 0.9707,
|
| 1325 |
-
"step": 1880
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 0.5802886091495241,
|
| 1329 |
-
"grad_norm": 1.4222575426101685,
|
| 1330 |
-
"learning_rate": 4.3273187717632157e-07,
|
| 1331 |
-
"loss": 0.8285,
|
| 1332 |
-
"step": 1890
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 0.5833589192508444,
|
| 1336 |
-
"grad_norm": 1.9608110189437866,
|
| 1337 |
-
"learning_rate": 4.295663184552073e-07,
|
| 1338 |
-
"loss": 0.8951,
|
| 1339 |
-
"step": 1900
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 0.5864292293521646,
|
| 1343 |
-
"grad_norm": 1.8060814142227173,
|
| 1344 |
-
"learning_rate": 4.264007597340931e-07,
|
| 1345 |
-
"loss": 0.9487,
|
| 1346 |
-
"step": 1910
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 0.5894995394534848,
|
| 1350 |
-
"grad_norm": 2.0148568153381348,
|
| 1351 |
-
"learning_rate": 4.2323520101297875e-07,
|
| 1352 |
-
"loss": 0.872,
|
| 1353 |
-
"step": 1920
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 0.5925698495548051,
|
| 1357 |
-
"grad_norm": 2.4442741870880127,
|
| 1358 |
-
"learning_rate": 4.2006964229186447e-07,
|
| 1359 |
-
"loss": 0.8681,
|
| 1360 |
-
"step": 1930
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 0.5956401596561253,
|
| 1364 |
-
"grad_norm": 2.112091541290283,
|
| 1365 |
-
"learning_rate": 4.1690408357075025e-07,
|
| 1366 |
-
"loss": 0.8961,
|
| 1367 |
-
"step": 1940
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 0.5987104697574455,
|
| 1371 |
-
"grad_norm": 2.6487627029418945,
|
| 1372 |
-
"learning_rate": 4.13738524849636e-07,
|
| 1373 |
-
"loss": 0.8829,
|
| 1374 |
-
"step": 1950
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 0.6017807798587658,
|
| 1378 |
-
"grad_norm": 1.5637843608856201,
|
| 1379 |
-
"learning_rate": 4.1057296612852165e-07,
|
| 1380 |
-
"loss": 0.9355,
|
| 1381 |
-
"step": 1960
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 0.604851089960086,
|
| 1385 |
-
"grad_norm": 1.4878854751586914,
|
| 1386 |
-
"learning_rate": 4.0740740740740737e-07,
|
| 1387 |
-
"loss": 0.9063,
|
| 1388 |
-
"step": 1970
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 0.6079214000614062,
|
| 1392 |
-
"grad_norm": 2.0287206172943115,
|
| 1393 |
-
"learning_rate": 4.0424184868629315e-07,
|
| 1394 |
-
"loss": 0.9765,
|
| 1395 |
-
"step": 1980
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 0.6109917101627265,
|
| 1399 |
-
"grad_norm": 1.8464608192443848,
|
| 1400 |
-
"learning_rate": 4.010762899651788e-07,
|
| 1401 |
-
"loss": 1.0074,
|
| 1402 |
-
"step": 1990
|
| 1403 |
-
},
|
| 1404 |
-
{
|
| 1405 |
-
"epoch": 0.6140620202640467,
|
| 1406 |
-
"grad_norm": 2.208984851837158,
|
| 1407 |
-
"learning_rate": 3.9791073124406455e-07,
|
| 1408 |
-
"loss": 0.9381,
|
| 1409 |
-
"step": 2000
|
| 1410 |
-
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 0.6171323303653669,
|
| 1413 |
-
"grad_norm": 3.5903429985046387,
|
| 1414 |
-
"learning_rate": 3.9474517252295033e-07,
|
| 1415 |
-
"loss": 0.9285,
|
| 1416 |
-
"step": 2010
|
| 1417 |
-
},
|
| 1418 |
-
{
|
| 1419 |
-
"epoch": 0.6202026404666872,
|
| 1420 |
-
"grad_norm": 1.8227884769439697,
|
| 1421 |
-
"learning_rate": 3.91579613801836e-07,
|
| 1422 |
-
"loss": 0.8484,
|
| 1423 |
-
"step": 2020
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 0.6232729505680074,
|
| 1427 |
-
"grad_norm": 1.8685295581817627,
|
| 1428 |
-
"learning_rate": 3.884140550807217e-07,
|
| 1429 |
-
"loss": 0.8183,
|
| 1430 |
-
"step": 2030
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 0.6263432606693276,
|
| 1434 |
-
"grad_norm": 1.505017876625061,
|
| 1435 |
-
"learning_rate": 3.8524849635960745e-07,
|
| 1436 |
-
"loss": 0.9065,
|
| 1437 |
-
"step": 2040
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 0.6294135707706479,
|
| 1441 |
-
"grad_norm": 1.7611987590789795,
|
| 1442 |
-
"learning_rate": 3.820829376384932e-07,
|
| 1443 |
-
"loss": 0.8511,
|
| 1444 |
-
"step": 2050
|
| 1445 |
-
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 0.632483880871968,
|
| 1448 |
-
"grad_norm": 1.9972463846206665,
|
| 1449 |
-
"learning_rate": 3.789173789173789e-07,
|
| 1450 |
-
"loss": 0.8686,
|
| 1451 |
-
"step": 2060
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 0.6355541909732882,
|
| 1455 |
-
"grad_norm": 2.034118890762329,
|
| 1456 |
-
"learning_rate": 3.757518201962646e-07,
|
| 1457 |
-
"loss": 0.7841,
|
| 1458 |
-
"step": 2070
|
| 1459 |
-
},
|
| 1460 |
-
{
|
| 1461 |
-
"epoch": 0.6386245010746086,
|
| 1462 |
-
"grad_norm": 1.953757882118225,
|
| 1463 |
-
"learning_rate": 3.725862614751503e-07,
|
| 1464 |
-
"loss": 0.893,
|
| 1465 |
-
"step": 2080
|
| 1466 |
-
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 0.6416948111759287,
|
| 1469 |
-
"grad_norm": 2.0740292072296143,
|
| 1470 |
-
"learning_rate": 3.694207027540361e-07,
|
| 1471 |
-
"loss": 0.8787,
|
| 1472 |
-
"step": 2090
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.644765121277249,
|
| 1476 |
-
"grad_norm": 1.808714509010315,
|
| 1477 |
-
"learning_rate": 3.662551440329218e-07,
|
| 1478 |
-
"loss": 0.8674,
|
| 1479 |
-
"step": 2100
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.6478354313785692,
|
| 1483 |
-
"grad_norm": 2.792292594909668,
|
| 1484 |
-
"learning_rate": 3.630895853118075e-07,
|
| 1485 |
-
"loss": 0.7246,
|
| 1486 |
-
"step": 2110
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 0.6509057414798894,
|
| 1490 |
-
"grad_norm": 1.9077739715576172,
|
| 1491 |
-
"learning_rate": 3.5992402659069325e-07,
|
| 1492 |
-
"loss": 0.8878,
|
| 1493 |
-
"step": 2120
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 0.6539760515812097,
|
| 1497 |
-
"grad_norm": 2.024315357208252,
|
| 1498 |
-
"learning_rate": 3.56758467869579e-07,
|
| 1499 |
-
"loss": 0.8197,
|
| 1500 |
-
"step": 2130
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 0.65704636168253,
|
| 1504 |
-
"grad_norm": 1.799088954925537,
|
| 1505 |
-
"learning_rate": 3.535929091484647e-07,
|
| 1506 |
-
"loss": 0.7852,
|
| 1507 |
-
"step": 2140
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 0.6601166717838501,
|
| 1511 |
-
"grad_norm": 2.362877368927002,
|
| 1512 |
-
"learning_rate": 3.504273504273504e-07,
|
| 1513 |
-
"loss": 0.8204,
|
| 1514 |
-
"step": 2150
|
| 1515 |
-
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 0.6631869818851704,
|
| 1518 |
-
"grad_norm": 1.3777751922607422,
|
| 1519 |
-
"learning_rate": 3.4726179170623615e-07,
|
| 1520 |
-
"loss": 0.8348,
|
| 1521 |
-
"step": 2160
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 0.6662572919864906,
|
| 1525 |
-
"grad_norm": 1.9549330472946167,
|
| 1526 |
-
"learning_rate": 3.440962329851219e-07,
|
| 1527 |
-
"loss": 0.8312,
|
| 1528 |
-
"step": 2170
|
| 1529 |
-
},
|
| 1530 |
-
{
|
| 1531 |
-
"epoch": 0.6693276020878108,
|
| 1532 |
-
"grad_norm": 1.8688315153121948,
|
| 1533 |
-
"learning_rate": 3.4093067426400755e-07,
|
| 1534 |
-
"loss": 0.8096,
|
| 1535 |
-
"step": 2180
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 0.6723979121891311,
|
| 1539 |
-
"grad_norm": 1.5635504722595215,
|
| 1540 |
-
"learning_rate": 3.3776511554289333e-07,
|
| 1541 |
-
"loss": 0.9582,
|
| 1542 |
-
"step": 2190
|
| 1543 |
-
},
|
| 1544 |
-
{
|
| 1545 |
-
"epoch": 0.6754682222904513,
|
| 1546 |
-
"grad_norm": 1.500148892402649,
|
| 1547 |
-
"learning_rate": 3.3459955682177905e-07,
|
| 1548 |
-
"loss": 0.8907,
|
| 1549 |
-
"step": 2200
|
| 1550 |
-
},
|
| 1551 |
-
{
|
| 1552 |
-
"epoch": 0.6785385323917715,
|
| 1553 |
-
"grad_norm": 1.767208456993103,
|
| 1554 |
-
"learning_rate": 3.314339981006647e-07,
|
| 1555 |
-
"loss": 0.912,
|
| 1556 |
-
"step": 2210
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 0.6816088424930918,
|
| 1560 |
-
"grad_norm": 2.737919807434082,
|
| 1561 |
-
"learning_rate": 3.2826843937955045e-07,
|
| 1562 |
-
"loss": 0.7511,
|
| 1563 |
-
"step": 2220
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 0.684679152594412,
|
| 1567 |
-
"grad_norm": 1.8129867315292358,
|
| 1568 |
-
"learning_rate": 3.2510288065843623e-07,
|
| 1569 |
-
"loss": 0.8095,
|
| 1570 |
-
"step": 2230
|
| 1571 |
-
},
|
| 1572 |
-
{
|
| 1573 |
-
"epoch": 0.6877494626957322,
|
| 1574 |
-
"grad_norm": 1.8896024227142334,
|
| 1575 |
-
"learning_rate": 3.219373219373219e-07,
|
| 1576 |
-
"loss": 0.7891,
|
| 1577 |
-
"step": 2240
|
| 1578 |
-
},
|
| 1579 |
-
{
|
| 1580 |
-
"epoch": 0.6908197727970525,
|
| 1581 |
-
"grad_norm": 1.7065414190292358,
|
| 1582 |
-
"learning_rate": 3.187717632162076e-07,
|
| 1583 |
-
"loss": 0.8119,
|
| 1584 |
-
"step": 2250
|
| 1585 |
-
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 0.6938900828983727,
|
| 1588 |
-
"grad_norm": 1.7079863548278809,
|
| 1589 |
-
"learning_rate": 3.1560620449509335e-07,
|
| 1590 |
-
"loss": 0.7894,
|
| 1591 |
-
"step": 2260
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 0.6969603929996929,
|
| 1595 |
-
"grad_norm": 1.4625552892684937,
|
| 1596 |
-
"learning_rate": 3.1244064577397913e-07,
|
| 1597 |
-
"loss": 0.8951,
|
| 1598 |
-
"step": 2270
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 0.7000307031010132,
|
| 1602 |
-
"grad_norm": 1.7978609800338745,
|
| 1603 |
-
"learning_rate": 3.092750870528648e-07,
|
| 1604 |
-
"loss": 0.898,
|
| 1605 |
-
"step": 2280
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 0.7031010132023334,
|
| 1609 |
-
"grad_norm": 2.16448974609375,
|
| 1610 |
-
"learning_rate": 3.0610952833175053e-07,
|
| 1611 |
-
"loss": 0.8961,
|
| 1612 |
-
"step": 2290
|
| 1613 |
-
},
|
| 1614 |
-
{
|
| 1615 |
-
"epoch": 0.7061713233036536,
|
| 1616 |
-
"grad_norm": 1.6174335479736328,
|
| 1617 |
-
"learning_rate": 3.029439696106363e-07,
|
| 1618 |
-
"loss": 0.7894,
|
| 1619 |
-
"step": 2300
|
| 1620 |
-
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 0.7092416334049739,
|
| 1623 |
-
"grad_norm": 2.177285671234131,
|
| 1624 |
-
"learning_rate": 2.99778410889522e-07,
|
| 1625 |
-
"loss": 0.873,
|
| 1626 |
-
"step": 2310
|
| 1627 |
-
},
|
| 1628 |
-
{
|
| 1629 |
-
"epoch": 0.7123119435062941,
|
| 1630 |
-
"grad_norm": 2.4392199516296387,
|
| 1631 |
-
"learning_rate": 2.966128521684077e-07,
|
| 1632 |
-
"loss": 0.869,
|
| 1633 |
-
"step": 2320
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 0.7153822536076143,
|
| 1637 |
-
"grad_norm": 2.829726219177246,
|
| 1638 |
-
"learning_rate": 2.9344729344729343e-07,
|
| 1639 |
-
"loss": 0.8769,
|
| 1640 |
-
"step": 2330
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 0.7184525637089346,
|
| 1644 |
-
"grad_norm": 2.4519991874694824,
|
| 1645 |
-
"learning_rate": 2.9028173472617915e-07,
|
| 1646 |
-
"loss": 0.7622,
|
| 1647 |
-
"step": 2340
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 0.7215228738102548,
|
| 1651 |
-
"grad_norm": 2.479762077331543,
|
| 1652 |
-
"learning_rate": 2.871161760050649e-07,
|
| 1653 |
-
"loss": 0.8554,
|
| 1654 |
-
"step": 2350
|
| 1655 |
-
},
|
| 1656 |
-
{
|
| 1657 |
-
"epoch": 0.7245931839115751,
|
| 1658 |
-
"grad_norm": 1.7480168342590332,
|
| 1659 |
-
"learning_rate": 2.839506172839506e-07,
|
| 1660 |
-
"loss": 0.8004,
|
| 1661 |
-
"step": 2360
|
| 1662 |
-
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 0.7276634940128953,
|
| 1665 |
-
"grad_norm": 1.6796847581863403,
|
| 1666 |
-
"learning_rate": 2.8078505856283633e-07,
|
| 1667 |
-
"loss": 0.7557,
|
| 1668 |
-
"step": 2370
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 0.7307338041142155,
|
| 1672 |
-
"grad_norm": 1.8005452156066895,
|
| 1673 |
-
"learning_rate": 2.7761949984172205e-07,
|
| 1674 |
-
"loss": 0.7349,
|
| 1675 |
-
"step": 2380
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 0.7338041142155358,
|
| 1679 |
-
"grad_norm": 1.5013504028320312,
|
| 1680 |
-
"learning_rate": 2.744539411206078e-07,
|
| 1681 |
-
"loss": 0.7608,
|
| 1682 |
-
"step": 2390
|
| 1683 |
-
},
|
| 1684 |
-
{
|
| 1685 |
-
"epoch": 0.736874424316856,
|
| 1686 |
-
"grad_norm": 1.9756343364715576,
|
| 1687 |
-
"learning_rate": 2.7128838239949345e-07,
|
| 1688 |
-
"loss": 0.7542,
|
| 1689 |
-
"step": 2400
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 0.7399447344181762,
|
| 1693 |
-
"grad_norm": 1.6326251029968262,
|
| 1694 |
-
"learning_rate": 2.6812282367837923e-07,
|
| 1695 |
-
"loss": 0.8952,
|
| 1696 |
-
"step": 2410
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 0.7430150445194965,
|
| 1700 |
-
"grad_norm": 2.549936056137085,
|
| 1701 |
-
"learning_rate": 2.6495726495726495e-07,
|
| 1702 |
-
"loss": 0.9441,
|
| 1703 |
-
"step": 2420
|
| 1704 |
-
},
|
| 1705 |
-
{
|
| 1706 |
-
"epoch": 0.7460853546208167,
|
| 1707 |
-
"grad_norm": 2.071596622467041,
|
| 1708 |
-
"learning_rate": 2.617917062361507e-07,
|
| 1709 |
-
"loss": 0.8192,
|
| 1710 |
-
"step": 2430
|
| 1711 |
-
},
|
| 1712 |
-
{
|
| 1713 |
-
"epoch": 0.7491556647221369,
|
| 1714 |
-
"grad_norm": 2.2240512371063232,
|
| 1715 |
-
"learning_rate": 2.586261475150364e-07,
|
| 1716 |
-
"loss": 0.8697,
|
| 1717 |
-
"step": 2440
|
| 1718 |
-
},
|
| 1719 |
-
{
|
| 1720 |
-
"epoch": 0.7522259748234572,
|
| 1721 |
-
"grad_norm": 2.4233920574188232,
|
| 1722 |
-
"learning_rate": 2.5546058879392213e-07,
|
| 1723 |
-
"loss": 0.8084,
|
| 1724 |
-
"step": 2450
|
| 1725 |
-
},
|
| 1726 |
-
{
|
| 1727 |
-
"epoch": 0.7552962849247774,
|
| 1728 |
-
"grad_norm": 1.7786365747451782,
|
| 1729 |
-
"learning_rate": 2.5229503007280786e-07,
|
| 1730 |
-
"loss": 0.8905,
|
| 1731 |
-
"step": 2460
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 0.7583665950260976,
|
| 1735 |
-
"grad_norm": 1.9785822629928589,
|
| 1736 |
-
"learning_rate": 2.491294713516936e-07,
|
| 1737 |
-
"loss": 0.7836,
|
| 1738 |
-
"step": 2470
|
| 1739 |
-
},
|
| 1740 |
-
{
|
| 1741 |
-
"epoch": 0.7614369051274179,
|
| 1742 |
-
"grad_norm": 1.8973301649093628,
|
| 1743 |
-
"learning_rate": 2.4596391263057925e-07,
|
| 1744 |
-
"loss": 0.8684,
|
| 1745 |
-
"step": 2480
|
| 1746 |
-
},
|
| 1747 |
-
{
|
| 1748 |
-
"epoch": 0.7645072152287381,
|
| 1749 |
-
"grad_norm": 1.6130449771881104,
|
| 1750 |
-
"learning_rate": 2.4279835390946503e-07,
|
| 1751 |
-
"loss": 0.7865,
|
| 1752 |
-
"step": 2490
|
| 1753 |
-
},
|
| 1754 |
-
{
|
| 1755 |
-
"epoch": 0.7675775253300583,
|
| 1756 |
-
"grad_norm": 1.6121731996536255,
|
| 1757 |
-
"learning_rate": 2.396327951883507e-07,
|
| 1758 |
-
"loss": 0.8248,
|
| 1759 |
-
"step": 2500
|
| 1760 |
-
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 0.7706478354313786,
|
| 1763 |
-
"grad_norm": 1.6531673669815063,
|
| 1764 |
-
"learning_rate": 2.3646723646723645e-07,
|
| 1765 |
-
"loss": 0.6813,
|
| 1766 |
-
"step": 2510
|
| 1767 |
-
},
|
| 1768 |
-
{
|
| 1769 |
-
"epoch": 0.7737181455326988,
|
| 1770 |
-
"grad_norm": 2.1497576236724854,
|
| 1771 |
-
"learning_rate": 2.3330167774612215e-07,
|
| 1772 |
-
"loss": 0.8148,
|
| 1773 |
-
"step": 2520
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 0.776788455634019,
|
| 1777 |
-
"grad_norm": 1.9006414413452148,
|
| 1778 |
-
"learning_rate": 2.301361190250079e-07,
|
| 1779 |
-
"loss": 0.7169,
|
| 1780 |
-
"step": 2530
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 0.7798587657353393,
|
| 1784 |
-
"grad_norm": 1.6517983675003052,
|
| 1785 |
-
"learning_rate": 2.2697056030389363e-07,
|
| 1786 |
-
"loss": 0.8157,
|
| 1787 |
-
"step": 2540
|
| 1788 |
-
},
|
| 1789 |
-
{
|
| 1790 |
-
"epoch": 0.7829290758366595,
|
| 1791 |
-
"grad_norm": 1.6298216581344604,
|
| 1792 |
-
"learning_rate": 2.2380500158277936e-07,
|
| 1793 |
-
"loss": 0.762,
|
| 1794 |
-
"step": 2550
|
| 1795 |
-
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 0.7859993859379797,
|
| 1798 |
-
"grad_norm": 1.198767066001892,
|
| 1799 |
-
"learning_rate": 2.2063944286166508e-07,
|
| 1800 |
-
"loss": 0.8029,
|
| 1801 |
-
"step": 2560
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 0.7890696960393,
|
| 1805 |
-
"grad_norm": 1.9983083009719849,
|
| 1806 |
-
"learning_rate": 2.1747388414055078e-07,
|
| 1807 |
-
"loss": 0.7524,
|
| 1808 |
-
"step": 2570
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 0.7921400061406202,
|
| 1812 |
-
"grad_norm": 1.9805610179901123,
|
| 1813 |
-
"learning_rate": 2.1430832541943653e-07,
|
| 1814 |
-
"loss": 0.7348,
|
| 1815 |
-
"step": 2580
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 0.7952103162419404,
|
| 1819 |
-
"grad_norm": 2.2282845973968506,
|
| 1820 |
-
"learning_rate": 2.1114276669832223e-07,
|
| 1821 |
-
"loss": 0.7699,
|
| 1822 |
-
"step": 2590
|
| 1823 |
-
},
|
| 1824 |
-
{
|
| 1825 |
-
"epoch": 0.7982806263432607,
|
| 1826 |
-
"grad_norm": 1.6694320440292358,
|
| 1827 |
-
"learning_rate": 2.0797720797720798e-07,
|
| 1828 |
-
"loss": 0.716,
|
| 1829 |
-
"step": 2600
|
| 1830 |
-
},
|
| 1831 |
-
{
|
| 1832 |
-
"epoch": 0.8013509364445809,
|
| 1833 |
-
"grad_norm": 2.17622971534729,
|
| 1834 |
-
"learning_rate": 2.048116492560937e-07,
|
| 1835 |
-
"loss": 0.7734,
|
| 1836 |
-
"step": 2610
|
| 1837 |
-
},
|
| 1838 |
-
{
|
| 1839 |
-
"epoch": 0.8044212465459011,
|
| 1840 |
-
"grad_norm": 1.7180266380310059,
|
| 1841 |
-
"learning_rate": 2.016460905349794e-07,
|
| 1842 |
-
"loss": 0.8293,
|
| 1843 |
-
"step": 2620
|
| 1844 |
-
},
|
| 1845 |
-
{
|
| 1846 |
-
"epoch": 0.8074915566472214,
|
| 1847 |
-
"grad_norm": 2.1564178466796875,
|
| 1848 |
-
"learning_rate": 1.9848053181386516e-07,
|
| 1849 |
-
"loss": 0.746,
|
| 1850 |
-
"step": 2630
|
| 1851 |
-
},
|
| 1852 |
-
{
|
| 1853 |
-
"epoch": 0.8105618667485416,
|
| 1854 |
-
"grad_norm": 1.8624786138534546,
|
| 1855 |
-
"learning_rate": 1.9531497309275086e-07,
|
| 1856 |
-
"loss": 0.759,
|
| 1857 |
-
"step": 2640
|
| 1858 |
-
},
|
| 1859 |
-
{
|
| 1860 |
-
"epoch": 0.8136321768498619,
|
| 1861 |
-
"grad_norm": 3.1279971599578857,
|
| 1862 |
-
"learning_rate": 1.9214941437163658e-07,
|
| 1863 |
-
"loss": 0.7752,
|
| 1864 |
-
"step": 2650
|
| 1865 |
-
},
|
| 1866 |
-
{
|
| 1867 |
-
"epoch": 0.8167024869511821,
|
| 1868 |
-
"grad_norm": 2.345259189605713,
|
| 1869 |
-
"learning_rate": 1.889838556505223e-07,
|
| 1870 |
-
"loss": 0.7453,
|
| 1871 |
-
"step": 2660
|
| 1872 |
-
},
|
| 1873 |
-
{
|
| 1874 |
-
"epoch": 0.8197727970525023,
|
| 1875 |
-
"grad_norm": 2.008528709411621,
|
| 1876 |
-
"learning_rate": 1.8581829692940803e-07,
|
| 1877 |
-
"loss": 0.7598,
|
| 1878 |
-
"step": 2670
|
| 1879 |
-
},
|
| 1880 |
-
{
|
| 1881 |
-
"epoch": 0.8228431071538226,
|
| 1882 |
-
"grad_norm": 2.444312334060669,
|
| 1883 |
-
"learning_rate": 1.8265273820829376e-07,
|
| 1884 |
-
"loss": 0.8075,
|
| 1885 |
-
"step": 2680
|
| 1886 |
-
},
|
| 1887 |
-
{
|
| 1888 |
-
"epoch": 0.8259134172551428,
|
| 1889 |
-
"grad_norm": 2.321638345718384,
|
| 1890 |
-
"learning_rate": 1.7948717948717948e-07,
|
| 1891 |
-
"loss": 0.7458,
|
| 1892 |
-
"step": 2690
|
| 1893 |
-
},
|
| 1894 |
-
{
|
| 1895 |
-
"epoch": 0.828983727356463,
|
| 1896 |
-
"grad_norm": 1.5578280687332153,
|
| 1897 |
-
"learning_rate": 1.763216207660652e-07,
|
| 1898 |
-
"loss": 0.8062,
|
| 1899 |
-
"step": 2700
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 0.8320540374577833,
|
| 1903 |
-
"grad_norm": 2.1762735843658447,
|
| 1904 |
-
"learning_rate": 1.7315606204495093e-07,
|
| 1905 |
-
"loss": 0.7908,
|
| 1906 |
-
"step": 2710
|
| 1907 |
-
},
|
| 1908 |
-
{
|
| 1909 |
-
"epoch": 0.8351243475591035,
|
| 1910 |
-
"grad_norm": 1.9740185737609863,
|
| 1911 |
-
"learning_rate": 1.6999050332383666e-07,
|
| 1912 |
-
"loss": 0.7735,
|
| 1913 |
-
"step": 2720
|
| 1914 |
-
},
|
| 1915 |
-
{
|
| 1916 |
-
"epoch": 0.8381946576604237,
|
| 1917 |
-
"grad_norm": 1.6252373456954956,
|
| 1918 |
-
"learning_rate": 1.6682494460272236e-07,
|
| 1919 |
-
"loss": 0.8089,
|
| 1920 |
-
"step": 2730
|
| 1921 |
-
},
|
| 1922 |
-
{
|
| 1923 |
-
"epoch": 0.841264967761744,
|
| 1924 |
-
"grad_norm": 2.2141306400299072,
|
| 1925 |
-
"learning_rate": 1.636593858816081e-07,
|
| 1926 |
-
"loss": 0.8623,
|
| 1927 |
-
"step": 2740
|
| 1928 |
-
},
|
| 1929 |
-
{
|
| 1930 |
-
"epoch": 0.8443352778630642,
|
| 1931 |
-
"grad_norm": 2.028036117553711,
|
| 1932 |
-
"learning_rate": 1.604938271604938e-07,
|
| 1933 |
-
"loss": 0.8883,
|
| 1934 |
-
"step": 2750
|
| 1935 |
-
},
|
| 1936 |
-
{
|
| 1937 |
-
"epoch": 0.8474055879643844,
|
| 1938 |
-
"grad_norm": 1.9973171949386597,
|
| 1939 |
-
"learning_rate": 1.5732826843937956e-07,
|
| 1940 |
-
"loss": 0.8589,
|
| 1941 |
-
"step": 2760
|
| 1942 |
-
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 0.8504758980657047,
|
| 1945 |
-
"grad_norm": 2.2462198734283447,
|
| 1946 |
-
"learning_rate": 1.5416270971826526e-07,
|
| 1947 |
-
"loss": 0.8815,
|
| 1948 |
-
"step": 2770
|
| 1949 |
-
},
|
| 1950 |
-
{
|
| 1951 |
-
"epoch": 0.8535462081670249,
|
| 1952 |
-
"grad_norm": 1.4654901027679443,
|
| 1953 |
-
"learning_rate": 1.5099715099715098e-07,
|
| 1954 |
-
"loss": 0.7683,
|
| 1955 |
-
"step": 2780
|
| 1956 |
-
},
|
| 1957 |
-
{
|
| 1958 |
-
"epoch": 0.8566165182683451,
|
| 1959 |
-
"grad_norm": 1.7927677631378174,
|
| 1960 |
-
"learning_rate": 1.4783159227603673e-07,
|
| 1961 |
-
"loss": 0.802,
|
| 1962 |
-
"step": 2790
|
| 1963 |
-
},
|
| 1964 |
-
{
|
| 1965 |
-
"epoch": 0.8596868283696654,
|
| 1966 |
-
"grad_norm": 2.216456890106201,
|
| 1967 |
-
"learning_rate": 1.4466603355492243e-07,
|
| 1968 |
-
"loss": 0.6807,
|
| 1969 |
-
"step": 2800
|
| 1970 |
-
},
|
| 1971 |
-
{
|
| 1972 |
-
"epoch": 0.8627571384709856,
|
| 1973 |
-
"grad_norm": 2.353961229324341,
|
| 1974 |
-
"learning_rate": 1.4150047483380816e-07,
|
| 1975 |
-
"loss": 0.8638,
|
| 1976 |
-
"step": 2810
|
| 1977 |
-
},
|
| 1978 |
-
{
|
| 1979 |
-
"epoch": 0.8658274485723058,
|
| 1980 |
-
"grad_norm": 1.4108779430389404,
|
| 1981 |
-
"learning_rate": 1.3833491611269388e-07,
|
| 1982 |
-
"loss": 0.7982,
|
| 1983 |
-
"step": 2820
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 0.8688977586736261,
|
| 1987 |
-
"grad_norm": 2.021253824234009,
|
| 1988 |
-
"learning_rate": 1.351693573915796e-07,
|
| 1989 |
-
"loss": 0.771,
|
| 1990 |
-
"step": 2830
|
| 1991 |
-
},
|
| 1992 |
-
{
|
| 1993 |
-
"epoch": 0.8719680687749463,
|
| 1994 |
-
"grad_norm": 1.6902941465377808,
|
| 1995 |
-
"learning_rate": 1.3200379867046533e-07,
|
| 1996 |
-
"loss": 0.7988,
|
| 1997 |
-
"step": 2840
|
| 1998 |
-
},
|
| 1999 |
-
{
|
| 2000 |
-
"epoch": 0.8750383788762665,
|
| 2001 |
-
"grad_norm": 2.057300329208374,
|
| 2002 |
-
"learning_rate": 1.2883823994935106e-07,
|
| 2003 |
-
"loss": 0.7589,
|
| 2004 |
-
"step": 2850
|
| 2005 |
-
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 0.8781086889775868,
|
| 2008 |
-
"grad_norm": 2.0052621364593506,
|
| 2009 |
-
"learning_rate": 1.2567268122823676e-07,
|
| 2010 |
-
"loss": 0.7285,
|
| 2011 |
-
"step": 2860
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 0.881178999078907,
|
| 2015 |
-
"grad_norm": 1.7238054275512695,
|
| 2016 |
-
"learning_rate": 1.225071225071225e-07,
|
| 2017 |
-
"loss": 0.7748,
|
| 2018 |
-
"step": 2870
|
| 2019 |
-
},
|
| 2020 |
-
{
|
| 2021 |
-
"epoch": 0.8842493091802271,
|
| 2022 |
-
"grad_norm": 1.6604859828948975,
|
| 2023 |
-
"learning_rate": 1.1934156378600823e-07,
|
| 2024 |
-
"loss": 0.7975,
|
| 2025 |
-
"step": 2880
|
| 2026 |
-
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 0.8873196192815475,
|
| 2029 |
-
"grad_norm": 2.2870965003967285,
|
| 2030 |
-
"learning_rate": 1.1617600506489395e-07,
|
| 2031 |
-
"loss": 0.822,
|
| 2032 |
-
"step": 2890
|
| 2033 |
-
},
|
| 2034 |
-
{
|
| 2035 |
-
"epoch": 0.8903899293828677,
|
| 2036 |
-
"grad_norm": 1.709179162979126,
|
| 2037 |
-
"learning_rate": 1.1301044634377967e-07,
|
| 2038 |
-
"loss": 0.7735,
|
| 2039 |
-
"step": 2900
|
| 2040 |
-
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 0.893460239484188,
|
| 2043 |
-
"grad_norm": 1.6187310218811035,
|
| 2044 |
-
"learning_rate": 1.0984488762266538e-07,
|
| 2045 |
-
"loss": 0.8106,
|
| 2046 |
-
"step": 2910
|
| 2047 |
-
},
|
| 2048 |
-
{
|
| 2049 |
-
"epoch": 0.8965305495855082,
|
| 2050 |
-
"grad_norm": 1.4298169612884521,
|
| 2051 |
-
"learning_rate": 1.0667932890155112e-07,
|
| 2052 |
-
"loss": 0.6901,
|
| 2053 |
-
"step": 2920
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 0.8996008596868283,
|
| 2057 |
-
"grad_norm": 1.541094183921814,
|
| 2058 |
-
"learning_rate": 1.0351377018043685e-07,
|
| 2059 |
-
"loss": 0.775,
|
| 2060 |
-
"step": 2930
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 0.9026711697881487,
|
| 2064 |
-
"grad_norm": 1.4745653867721558,
|
| 2065 |
-
"learning_rate": 1.0034821145932257e-07,
|
| 2066 |
-
"loss": 0.9154,
|
| 2067 |
-
"step": 2940
|
| 2068 |
-
},
|
| 2069 |
-
{
|
| 2070 |
-
"epoch": 0.9057414798894688,
|
| 2071 |
-
"grad_norm": 1.3007853031158447,
|
| 2072 |
-
"learning_rate": 9.718265273820828e-08,
|
| 2073 |
-
"loss": 0.7063,
|
| 2074 |
-
"step": 2950
|
| 2075 |
-
},
|
| 2076 |
-
{
|
| 2077 |
-
"epoch": 0.908811789990789,
|
| 2078 |
-
"grad_norm": 1.819677472114563,
|
| 2079 |
-
"learning_rate": 9.401709401709401e-08,
|
| 2080 |
-
"loss": 0.7288,
|
| 2081 |
-
"step": 2960
|
| 2082 |
-
},
|
| 2083 |
-
{
|
| 2084 |
-
"epoch": 0.9118821000921093,
|
| 2085 |
-
"grad_norm": 1.7065274715423584,
|
| 2086 |
-
"learning_rate": 9.085153529597973e-08,
|
| 2087 |
-
"loss": 0.7891,
|
| 2088 |
-
"step": 2970
|
| 2089 |
-
},
|
| 2090 |
-
{
|
| 2091 |
-
"epoch": 0.9149524101934295,
|
| 2092 |
-
"grad_norm": 1.572373390197754,
|
| 2093 |
-
"learning_rate": 8.768597657486546e-08,
|
| 2094 |
-
"loss": 0.802,
|
| 2095 |
-
"step": 2980
|
| 2096 |
-
},
|
| 2097 |
-
{
|
| 2098 |
-
"epoch": 0.9180227202947497,
|
| 2099 |
-
"grad_norm": 1.870667815208435,
|
| 2100 |
-
"learning_rate": 8.452041785375117e-08,
|
| 2101 |
-
"loss": 0.767,
|
| 2102 |
-
"step": 2990
|
| 2103 |
-
},
|
| 2104 |
-
{
|
| 2105 |
-
"epoch": 0.92109303039607,
|
| 2106 |
-
"grad_norm": 1.924141764640808,
|
| 2107 |
-
"learning_rate": 8.135485913263691e-08,
|
| 2108 |
-
"loss": 0.8292,
|
| 2109 |
-
"step": 3000
|
| 2110 |
-
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 0.9241633404973902,
|
| 2113 |
-
"grad_norm": 1.7546625137329102,
|
| 2114 |
-
"learning_rate": 7.818930041152264e-08,
|
| 2115 |
-
"loss": 0.7928,
|
| 2116 |
-
"step": 3010
|
| 2117 |
-
},
|
| 2118 |
-
{
|
| 2119 |
-
"epoch": 0.9272336505987104,
|
| 2120 |
-
"grad_norm": 2.3460092544555664,
|
| 2121 |
-
"learning_rate": 7.502374169040836e-08,
|
| 2122 |
-
"loss": 0.8534,
|
| 2123 |
-
"step": 3020
|
| 2124 |
-
},
|
| 2125 |
-
{
|
| 2126 |
-
"epoch": 0.9303039607000307,
|
| 2127 |
-
"grad_norm": 2.752711296081543,
|
| 2128 |
-
"learning_rate": 7.185818296929407e-08,
|
| 2129 |
-
"loss": 0.7184,
|
| 2130 |
-
"step": 3030
|
| 2131 |
-
},
|
| 2132 |
-
{
|
| 2133 |
-
"epoch": 0.9333742708013509,
|
| 2134 |
-
"grad_norm": 1.901767373085022,
|
| 2135 |
-
"learning_rate": 6.86926242481798e-08,
|
| 2136 |
-
"loss": 0.7939,
|
| 2137 |
-
"step": 3040
|
| 2138 |
-
},
|
| 2139 |
-
{
|
| 2140 |
-
"epoch": 0.9364445809026711,
|
| 2141 |
-
"grad_norm": 2.0688488483428955,
|
| 2142 |
-
"learning_rate": 6.552706552706552e-08,
|
| 2143 |
-
"loss": 0.7265,
|
| 2144 |
-
"step": 3050
|
| 2145 |
-
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 0.9395148910039914,
|
| 2148 |
-
"grad_norm": 1.7756198644638062,
|
| 2149 |
-
"learning_rate": 6.236150680595125e-08,
|
| 2150 |
-
"loss": 0.7499,
|
| 2151 |
-
"step": 3060
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 0.9425852011053116,
|
| 2155 |
-
"grad_norm": 1.5019465684890747,
|
| 2156 |
-
"learning_rate": 5.919594808483697e-08,
|
| 2157 |
-
"loss": 0.8274,
|
| 2158 |
-
"step": 3070
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 0.9456555112066318,
|
| 2162 |
-
"grad_norm": 1.6010905504226685,
|
| 2163 |
-
"learning_rate": 5.60303893637227e-08,
|
| 2164 |
-
"loss": 0.7893,
|
| 2165 |
-
"step": 3080
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 0.9487258213079521,
|
| 2169 |
-
"grad_norm": 1.6651114225387573,
|
| 2170 |
-
"learning_rate": 5.286483064260842e-08,
|
| 2171 |
-
"loss": 0.7241,
|
| 2172 |
-
"step": 3090
|
| 2173 |
-
},
|
| 2174 |
-
{
|
| 2175 |
-
"epoch": 0.9517961314092723,
|
| 2176 |
-
"grad_norm": 2.506920576095581,
|
| 2177 |
-
"learning_rate": 4.969927192149414e-08,
|
| 2178 |
-
"loss": 0.8166,
|
| 2179 |
-
"step": 3100
|
| 2180 |
-
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 0.9548664415105925,
|
| 2183 |
-
"grad_norm": 2.0352845191955566,
|
| 2184 |
-
"learning_rate": 4.653371320037987e-08,
|
| 2185 |
-
"loss": 0.7791,
|
| 2186 |
-
"step": 3110
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 0.9579367516119128,
|
| 2190 |
-
"grad_norm": 2.4386065006256104,
|
| 2191 |
-
"learning_rate": 4.336815447926559e-08,
|
| 2192 |
-
"loss": 0.7289,
|
| 2193 |
-
"step": 3120
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 0.961007061713233,
|
| 2197 |
-
"grad_norm": 1.4857659339904785,
|
| 2198 |
-
"learning_rate": 4.020259575815131e-08,
|
| 2199 |
-
"loss": 0.8678,
|
| 2200 |
-
"step": 3130
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 0.9640773718145532,
|
| 2204 |
-
"grad_norm": 1.967745065689087,
|
| 2205 |
-
"learning_rate": 3.7037037037037036e-08,
|
| 2206 |
-
"loss": 0.7746,
|
| 2207 |
-
"step": 3140
|
| 2208 |
-
},
|
| 2209 |
-
{
|
| 2210 |
-
"epoch": 0.9671476819158735,
|
| 2211 |
-
"grad_norm": 1.7988271713256836,
|
| 2212 |
-
"learning_rate": 3.387147831592276e-08,
|
| 2213 |
-
"loss": 0.7677,
|
| 2214 |
-
"step": 3150
|
| 2215 |
-
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 0.9702179920171937,
|
| 2218 |
-
"grad_norm": 1.9551514387130737,
|
| 2219 |
-
"learning_rate": 3.070591959480849e-08,
|
| 2220 |
-
"loss": 0.8411,
|
| 2221 |
-
"step": 3160
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 0.9732883021185139,
|
| 2225 |
-
"grad_norm": 2.3268728256225586,
|
| 2226 |
-
"learning_rate": 2.7540360873694205e-08,
|
| 2227 |
-
"loss": 0.8246,
|
| 2228 |
-
"step": 3170
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 0.9763586122198342,
|
| 2232 |
-
"grad_norm": 1.7236260175704956,
|
| 2233 |
-
"learning_rate": 2.4374802152579927e-08,
|
| 2234 |
-
"loss": 0.8471,
|
| 2235 |
-
"step": 3180
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 0.9794289223211544,
|
| 2239 |
-
"grad_norm": 1.8868488073349,
|
| 2240 |
-
"learning_rate": 2.1209243431465652e-08,
|
| 2241 |
-
"loss": 0.7894,
|
| 2242 |
-
"step": 3190
|
| 2243 |
-
},
|
| 2244 |
-
{
|
| 2245 |
-
"epoch": 0.9824992324224747,
|
| 2246 |
-
"grad_norm": 2.07332706451416,
|
| 2247 |
-
"learning_rate": 1.8043684710351374e-08,
|
| 2248 |
-
"loss": 0.7217,
|
| 2249 |
-
"step": 3200
|
| 2250 |
-
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 0.9855695425237949,
|
| 2253 |
-
"grad_norm": 1.6548092365264893,
|
| 2254 |
-
"learning_rate": 1.48781259892371e-08,
|
| 2255 |
-
"loss": 0.8436,
|
| 2256 |
-
"step": 3210
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 0.9886398526251151,
|
| 2260 |
-
"grad_norm": 1.4612842798233032,
|
| 2261 |
-
"learning_rate": 1.1712567268122825e-08,
|
| 2262 |
-
"loss": 0.7211,
|
| 2263 |
-
"step": 3220
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 0.9917101627264354,
|
| 2267 |
-
"grad_norm": 2.267146587371826,
|
| 2268 |
-
"learning_rate": 8.547008547008548e-09,
|
| 2269 |
-
"loss": 0.8614,
|
| 2270 |
-
"step": 3230
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 0.9947804728277556,
|
| 2274 |
-
"grad_norm": 2.2766923904418945,
|
| 2275 |
-
"learning_rate": 5.38144982589427e-09,
|
| 2276 |
-
"loss": 0.8921,
|
| 2277 |
-
"step": 3240
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 0.9978507829290758,
|
| 2281 |
-
"grad_norm": 2.107379674911499,
|
| 2282 |
-
"learning_rate": 2.2158911047799933e-09,
|
| 2283 |
-
"loss": 0.7466,
|
| 2284 |
-
"step": 3250
|
| 2285 |
-
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 1.0,
|
| 2288 |
-
"step": 3257,
|
| 2289 |
-
"total_flos": 2.214849998094336e+16,
|
| 2290 |
-
"train_loss": 0.9982479734950758,
|
| 2291 |
-
"train_runtime": 2064.7675,
|
| 2292 |
-
"train_samples_per_second": 3.154,
|
| 2293 |
-
"train_steps_per_second": 1.577
|
| 2294 |
}
|
| 2295 |
],
|
| 2296 |
-
"logging_steps":
|
| 2297 |
-
"max_steps":
|
| 2298 |
"num_input_tokens_seen": 0,
|
| 2299 |
-
"num_train_epochs":
|
| 2300 |
-
"save_steps":
|
| 2301 |
"stateful_callbacks": {
|
| 2302 |
"TrainerControl": {
|
| 2303 |
"args": {
|
|
@@ -2310,7 +35,7 @@
|
|
| 2310 |
"attributes": {}
|
| 2311 |
}
|
| 2312 |
},
|
| 2313 |
-
"total_flos":
|
| 2314 |
"train_batch_size": 2,
|
| 2315 |
"trial_name": null,
|
| 2316 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 28,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 4.0,
|
| 13 |
+
"step": 28,
|
| 14 |
+
"total_flos": 176151436984320.0,
|
| 15 |
+
"train_loss": 1.4319358553205217,
|
| 16 |
+
"train_runtime": 19.3836,
|
| 17 |
+
"train_samples_per_second": 2.683,
|
| 18 |
+
"train_steps_per_second": 1.445
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
}
|
| 20 |
],
|
| 21 |
+
"logging_steps": 1000,
|
| 22 |
+
"max_steps": 28,
|
| 23 |
"num_input_tokens_seen": 0,
|
| 24 |
+
"num_train_epochs": 4,
|
| 25 |
+
"save_steps": 1000,
|
| 26 |
"stateful_callbacks": {
|
| 27 |
"TrainerControl": {
|
| 28 |
"args": {
|
|
|
|
| 35 |
"attributes": {}
|
| 36 |
}
|
| 37 |
},
|
| 38 |
+
"total_flos": 176151436984320.0,
|
| 39 |
"train_batch_size": 2,
|
| 40 |
"trial_name": null,
|
| 41 |
"trial_params": null
|