Model save
Browse files- README.md +63 -0
- adapter_model.safetensors +1 -1
- all_results.json +13 -0
- config_argument.yaml +50 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1802 -0
README.md
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: hllj/zephyr-7b-beta-vi-math
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
model-index:
|
| 6 |
+
- name: sft-zephyr-7b-beta-v2
|
| 7 |
+
results: []
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 11 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 12 |
+
|
| 13 |
+
# sft-zephyr-7b-beta-v2
|
| 14 |
+
|
| 15 |
+
This model is a fine-tuned version of [hllj/zephyr-7b-beta-vi-math](https://huggingface.co/hllj/zephyr-7b-beta-vi-math) on the None dataset.
|
| 16 |
+
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 0.4432
|
| 18 |
+
|
| 19 |
+
## Model description
|
| 20 |
+
|
| 21 |
+
More information needed
|
| 22 |
+
|
| 23 |
+
## Intended uses & limitations
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Training and evaluation data
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training procedure
|
| 32 |
+
|
| 33 |
+
### Training hyperparameters
|
| 34 |
+
|
| 35 |
+
The following hyperparameters were used during training:
|
| 36 |
+
- learning_rate: 5e-05
|
| 37 |
+
- train_batch_size: 4
|
| 38 |
+
- eval_batch_size: 4
|
| 39 |
+
- seed: 42
|
| 40 |
+
- distributed_type: multi-GPU
|
| 41 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
+
- lr_scheduler_type: cosine
|
| 43 |
+
- lr_scheduler_warmup_ratio: 0.05
|
| 44 |
+
- num_epochs: 2
|
| 45 |
+
- mixed_precision_training: Native AMP
|
| 46 |
+
|
| 47 |
+
### Training results
|
| 48 |
+
|
| 49 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 50 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
| 51 |
+
| 0.3943 | 0.35 | 500 | 0.4903 |
|
| 52 |
+
| 0.3796 | 0.69 | 1000 | 0.4698 |
|
| 53 |
+
| 0.353 | 1.04 | 1500 | 0.4521 |
|
| 54 |
+
| 0.3173 | 1.39 | 2000 | 0.4463 |
|
| 55 |
+
| 0.3276 | 1.73 | 2500 | 0.4431 |
|
| 56 |
+
|
| 57 |
+
|
| 58 |
+
### Framework versions
|
| 59 |
+
|
| 60 |
+
- Transformers 4.35.2
|
| 61 |
+
- Pytorch 2.1.0
|
| 62 |
+
- Datasets 2.15.0
|
| 63 |
+
- Tokenizers 0.15.0
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 872450448
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad8bb3bca951a9d1513286b9f2d9e0ab164c4e7349a9215df143b36d9549888e
|
| 3 |
size 872450448
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_loss": 0.44321340322494507,
|
| 4 |
+
"eval_runtime": 79.4506,
|
| 5 |
+
"eval_samples": 640,
|
| 6 |
+
"eval_samples_per_second": 8.055,
|
| 7 |
+
"eval_steps_per_second": 2.014,
|
| 8 |
+
"train_loss": 0.3615408574310464,
|
| 9 |
+
"train_runtime": 5165.0048,
|
| 10 |
+
"train_samples": 5762,
|
| 11 |
+
"train_samples_per_second": 2.231,
|
| 12 |
+
"train_steps_per_second": 0.558
|
| 13 |
+
}
|
config_argument.yaml
ADDED
|
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
cache_dir: ./cache
|
| 2 |
+
ddp_find_unused_parameters: false
|
| 3 |
+
ddp_timeout: 30000
|
| 4 |
+
device_map: auto
|
| 5 |
+
do_eval: true
|
| 6 |
+
do_train: true
|
| 7 |
+
eval_steps: 500
|
| 8 |
+
evaluation_strategy: steps
|
| 9 |
+
fp16: true
|
| 10 |
+
gradient_accumulation_steps: 1
|
| 11 |
+
gradient_checkpointing: true
|
| 12 |
+
gradient_checkpointing_kwargs:
|
| 13 |
+
use_reentrant: false
|
| 14 |
+
hub_model_id: hllj/sft-zephyr-7b-beta-v2
|
| 15 |
+
hub_strategy: every_save
|
| 16 |
+
learning_rate: 5.0e-05
|
| 17 |
+
log_level: info
|
| 18 |
+
logging_first_step: true
|
| 19 |
+
logging_steps: 10
|
| 20 |
+
logging_strategy: steps
|
| 21 |
+
lora_alpha: 128
|
| 22 |
+
lora_dropout: 0.05
|
| 23 |
+
lora_r: 256
|
| 24 |
+
lora_target_modules:
|
| 25 |
+
- q_proj
|
| 26 |
+
- k_proj
|
| 27 |
+
- v_proj
|
| 28 |
+
- o_proj
|
| 29 |
+
lr_scheduler_type: cosine
|
| 30 |
+
max_seq_length: 512
|
| 31 |
+
model_name_or_path: hllj/zephyr-7b-beta-vi-math
|
| 32 |
+
model_type: auto
|
| 33 |
+
num_train_epochs: 2
|
| 34 |
+
output_dir: outputs-sft-zephyr-beta-v2
|
| 35 |
+
overwrite_output_dir: true
|
| 36 |
+
per_device_eval_batch_size: 4
|
| 37 |
+
per_device_train_batch_size: 4
|
| 38 |
+
preprocessing_num_workers: 4
|
| 39 |
+
push_to_hub: true
|
| 40 |
+
report_to: wandb
|
| 41 |
+
run_name: sft-zephyr-7b-beta-v2
|
| 42 |
+
save_steps: 500
|
| 43 |
+
save_strategy: steps
|
| 44 |
+
save_total_limit: 13
|
| 45 |
+
seed: 42
|
| 46 |
+
torch_dtype: float16
|
| 47 |
+
train_file_dir: datasets/finetune
|
| 48 |
+
use_peft: true
|
| 49 |
+
warmup_ratio: 0.05
|
| 50 |
+
weight_decay: 0.05
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"eval_loss": 0.44321340322494507,
|
| 4 |
+
"eval_runtime": 79.4506,
|
| 5 |
+
"eval_samples": 640,
|
| 6 |
+
"eval_samples_per_second": 8.055,
|
| 7 |
+
"eval_steps_per_second": 2.014
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"train_loss": 0.3615408574310464,
|
| 4 |
+
"train_runtime": 5165.0048,
|
| 5 |
+
"train_samples": 5762,
|
| 6 |
+
"train_samples_per_second": 2.231,
|
| 7 |
+
"train_steps_per_second": 0.558
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1802 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2882,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 3.4482758620689656e-07,
|
| 14 |
+
"loss": 0.8107,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 2.7586206896551725e-06,
|
| 20 |
+
"loss": 0.9975,
|
| 21 |
+
"step": 10
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 5.517241379310345e-06,
|
| 26 |
+
"loss": 0.8562,
|
| 27 |
+
"step": 20
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.02,
|
| 31 |
+
"learning_rate": 8.96551724137931e-06,
|
| 32 |
+
"loss": 0.8419,
|
| 33 |
+
"step": 30
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.03,
|
| 37 |
+
"learning_rate": 1.2413793103448277e-05,
|
| 38 |
+
"loss": 0.6876,
|
| 39 |
+
"step": 40
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.03,
|
| 43 |
+
"learning_rate": 1.586206896551724e-05,
|
| 44 |
+
"loss": 0.532,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04,
|
| 49 |
+
"learning_rate": 1.9310344827586207e-05,
|
| 50 |
+
"loss": 0.4951,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05,
|
| 55 |
+
"learning_rate": 2.2413793103448276e-05,
|
| 56 |
+
"loss": 0.5264,
|
| 57 |
+
"step": 70
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.06,
|
| 61 |
+
"learning_rate": 2.5862068965517244e-05,
|
| 62 |
+
"loss": 0.5313,
|
| 63 |
+
"step": 80
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.06,
|
| 67 |
+
"learning_rate": 2.9310344827586206e-05,
|
| 68 |
+
"loss": 0.4944,
|
| 69 |
+
"step": 90
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.07,
|
| 73 |
+
"learning_rate": 3.275862068965517e-05,
|
| 74 |
+
"loss": 0.4927,
|
| 75 |
+
"step": 100
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.08,
|
| 79 |
+
"learning_rate": 3.620689655172414e-05,
|
| 80 |
+
"loss": 0.483,
|
| 81 |
+
"step": 110
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.08,
|
| 85 |
+
"learning_rate": 3.965517241379311e-05,
|
| 86 |
+
"loss": 0.4766,
|
| 87 |
+
"step": 120
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09,
|
| 91 |
+
"learning_rate": 4.275862068965518e-05,
|
| 92 |
+
"loss": 0.5116,
|
| 93 |
+
"step": 130
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1,
|
| 97 |
+
"learning_rate": 4.6206896551724135e-05,
|
| 98 |
+
"loss": 0.4383,
|
| 99 |
+
"step": 140
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.1,
|
| 103 |
+
"learning_rate": 4.9655172413793107e-05,
|
| 104 |
+
"loss": 0.4489,
|
| 105 |
+
"step": 150
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.11,
|
| 109 |
+
"learning_rate": 4.999866604456691e-05,
|
| 110 |
+
"loss": 0.4921,
|
| 111 |
+
"step": 160
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.12,
|
| 115 |
+
"learning_rate": 4.999405502336266e-05,
|
| 116 |
+
"loss": 0.4698,
|
| 117 |
+
"step": 170
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.12,
|
| 121 |
+
"learning_rate": 4.998615107515172e-05,
|
| 122 |
+
"loss": 0.4356,
|
| 123 |
+
"step": 180
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.13,
|
| 127 |
+
"learning_rate": 4.9974955241266695e-05,
|
| 128 |
+
"loss": 0.5204,
|
| 129 |
+
"step": 190
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14,
|
| 133 |
+
"learning_rate": 4.996046899674091e-05,
|
| 134 |
+
"loss": 0.4585,
|
| 135 |
+
"step": 200
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.15,
|
| 139 |
+
"learning_rate": 4.994269425011402e-05,
|
| 140 |
+
"loss": 0.4721,
|
| 141 |
+
"step": 210
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.15,
|
| 145 |
+
"learning_rate": 4.9921633343180654e-05,
|
| 146 |
+
"loss": 0.4852,
|
| 147 |
+
"step": 220
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.16,
|
| 151 |
+
"learning_rate": 4.98972890506818e-05,
|
| 152 |
+
"loss": 0.4478,
|
| 153 |
+
"step": 230
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.17,
|
| 157 |
+
"learning_rate": 4.986966457993929e-05,
|
| 158 |
+
"loss": 0.4189,
|
| 159 |
+
"step": 240
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.17,
|
| 163 |
+
"learning_rate": 4.983876357043322e-05,
|
| 164 |
+
"loss": 0.439,
|
| 165 |
+
"step": 250
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.18,
|
| 169 |
+
"learning_rate": 4.980459009332246e-05,
|
| 170 |
+
"loss": 0.4436,
|
| 171 |
+
"step": 260
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.19,
|
| 175 |
+
"learning_rate": 4.976714865090827e-05,
|
| 176 |
+
"loss": 0.4002,
|
| 177 |
+
"step": 270
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.19,
|
| 181 |
+
"learning_rate": 4.972644417604117e-05,
|
| 182 |
+
"loss": 0.4599,
|
| 183 |
+
"step": 280
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.2,
|
| 187 |
+
"learning_rate": 4.9682482031470975e-05,
|
| 188 |
+
"loss": 0.4055,
|
| 189 |
+
"step": 290
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.21,
|
| 193 |
+
"learning_rate": 4.963526800914035e-05,
|
| 194 |
+
"loss": 0.4235,
|
| 195 |
+
"step": 300
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.22,
|
| 199 |
+
"learning_rate": 4.958480832942165e-05,
|
| 200 |
+
"loss": 0.397,
|
| 201 |
+
"step": 310
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.22,
|
| 205 |
+
"learning_rate": 4.9531109640297455e-05,
|
| 206 |
+
"loss": 0.4319,
|
| 207 |
+
"step": 320
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.23,
|
| 211 |
+
"learning_rate": 4.947417901648467e-05,
|
| 212 |
+
"loss": 0.4318,
|
| 213 |
+
"step": 330
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.24,
|
| 217 |
+
"learning_rate": 4.9414023958502444e-05,
|
| 218 |
+
"loss": 0.3984,
|
| 219 |
+
"step": 340
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.24,
|
| 223 |
+
"learning_rate": 4.9350652391684024e-05,
|
| 224 |
+
"loss": 0.4399,
|
| 225 |
+
"step": 350
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.25,
|
| 229 |
+
"learning_rate": 4.9284072665132576e-05,
|
| 230 |
+
"loss": 0.4207,
|
| 231 |
+
"step": 360
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.26,
|
| 235 |
+
"learning_rate": 4.92142935506212e-05,
|
| 236 |
+
"loss": 0.3951,
|
| 237 |
+
"step": 370
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.26,
|
| 241 |
+
"learning_rate": 4.9141324241437284e-05,
|
| 242 |
+
"loss": 0.3983,
|
| 243 |
+
"step": 380
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.27,
|
| 247 |
+
"learning_rate": 4.906517435117128e-05,
|
| 248 |
+
"loss": 0.4133,
|
| 249 |
+
"step": 390
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.28,
|
| 253 |
+
"learning_rate": 4.8985853912450155e-05,
|
| 254 |
+
"loss": 0.4269,
|
| 255 |
+
"step": 400
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.28,
|
| 259 |
+
"learning_rate": 4.8903373375615564e-05,
|
| 260 |
+
"loss": 0.4028,
|
| 261 |
+
"step": 410
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.29,
|
| 265 |
+
"learning_rate": 4.88177436073471e-05,
|
| 266 |
+
"loss": 0.3755,
|
| 267 |
+
"step": 420
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.3,
|
| 271 |
+
"learning_rate": 4.8728975889230576e-05,
|
| 272 |
+
"loss": 0.3991,
|
| 273 |
+
"step": 430
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.31,
|
| 277 |
+
"learning_rate": 4.86370819162717e-05,
|
| 278 |
+
"loss": 0.3806,
|
| 279 |
+
"step": 440
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.31,
|
| 283 |
+
"learning_rate": 4.8542073795355294e-05,
|
| 284 |
+
"loss": 0.4233,
|
| 285 |
+
"step": 450
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.32,
|
| 289 |
+
"learning_rate": 4.8443964043650204e-05,
|
| 290 |
+
"loss": 0.4172,
|
| 291 |
+
"step": 460
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.33,
|
| 295 |
+
"learning_rate": 4.8342765586960234e-05,
|
| 296 |
+
"loss": 0.4218,
|
| 297 |
+
"step": 470
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.33,
|
| 301 |
+
"learning_rate": 4.823849175802113e-05,
|
| 302 |
+
"loss": 0.3584,
|
| 303 |
+
"step": 480
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.34,
|
| 307 |
+
"learning_rate": 4.813115629474405e-05,
|
| 308 |
+
"loss": 0.3933,
|
| 309 |
+
"step": 490
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.35,
|
| 313 |
+
"learning_rate": 4.802077333840561e-05,
|
| 314 |
+
"loss": 0.3943,
|
| 315 |
+
"step": 500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.35,
|
| 319 |
+
"eval_loss": 0.4903396666049957,
|
| 320 |
+
"eval_runtime": 79.5297,
|
| 321 |
+
"eval_samples_per_second": 8.047,
|
| 322 |
+
"eval_steps_per_second": 2.012,
|
| 323 |
+
"step": 500
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 0.35,
|
| 327 |
+
"learning_rate": 4.7907357431784786e-05,
|
| 328 |
+
"loss": 0.3576,
|
| 329 |
+
"step": 510
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.36,
|
| 333 |
+
"learning_rate": 4.779092351724694e-05,
|
| 334 |
+
"loss": 0.3631,
|
| 335 |
+
"step": 520
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.37,
|
| 339 |
+
"learning_rate": 4.767148693477517e-05,
|
| 340 |
+
"loss": 0.432,
|
| 341 |
+
"step": 530
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 0.37,
|
| 345 |
+
"learning_rate": 4.75490634199493e-05,
|
| 346 |
+
"loss": 0.4042,
|
| 347 |
+
"step": 540
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 0.38,
|
| 351 |
+
"learning_rate": 4.742366910187275e-05,
|
| 352 |
+
"loss": 0.3582,
|
| 353 |
+
"step": 550
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.39,
|
| 357 |
+
"learning_rate": 4.7295320501047536e-05,
|
| 358 |
+
"loss": 0.3894,
|
| 359 |
+
"step": 560
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4,
|
| 363 |
+
"learning_rate": 4.716403452719773e-05,
|
| 364 |
+
"loss": 0.3687,
|
| 365 |
+
"step": 570
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.4,
|
| 369 |
+
"learning_rate": 4.702982847704163e-05,
|
| 370 |
+
"loss": 0.3736,
|
| 371 |
+
"step": 580
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.41,
|
| 375 |
+
"learning_rate": 4.6892720032012953e-05,
|
| 376 |
+
"loss": 0.4158,
|
| 377 |
+
"step": 590
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.42,
|
| 381 |
+
"learning_rate": 4.6752727255931314e-05,
|
| 382 |
+
"loss": 0.3849,
|
| 383 |
+
"step": 600
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.42,
|
| 387 |
+
"learning_rate": 4.660986859262236e-05,
|
| 388 |
+
"loss": 0.4131,
|
| 389 |
+
"step": 610
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.43,
|
| 393 |
+
"learning_rate": 4.646416286348783e-05,
|
| 394 |
+
"loss": 0.3922,
|
| 395 |
+
"step": 620
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.44,
|
| 399 |
+
"learning_rate": 4.631562926502584e-05,
|
| 400 |
+
"loss": 0.4013,
|
| 401 |
+
"step": 630
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.44,
|
| 405 |
+
"learning_rate": 4.6164287366301816e-05,
|
| 406 |
+
"loss": 0.4209,
|
| 407 |
+
"step": 640
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.45,
|
| 411 |
+
"learning_rate": 4.6010157106370266e-05,
|
| 412 |
+
"loss": 0.432,
|
| 413 |
+
"step": 650
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.46,
|
| 417 |
+
"learning_rate": 4.5853258791647864e-05,
|
| 418 |
+
"loss": 0.4061,
|
| 419 |
+
"step": 660
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.46,
|
| 423 |
+
"learning_rate": 4.569361309323809e-05,
|
| 424 |
+
"loss": 0.3992,
|
| 425 |
+
"step": 670
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.47,
|
| 429 |
+
"learning_rate": 4.553124104420784e-05,
|
| 430 |
+
"loss": 0.4148,
|
| 431 |
+
"step": 680
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.48,
|
| 435 |
+
"learning_rate": 4.536616403681642e-05,
|
| 436 |
+
"loss": 0.4124,
|
| 437 |
+
"step": 690
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.49,
|
| 441 |
+
"learning_rate": 4.5198403819697045e-05,
|
| 442 |
+
"loss": 0.3989,
|
| 443 |
+
"step": 700
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.49,
|
| 447 |
+
"learning_rate": 4.5027982494991574e-05,
|
| 448 |
+
"loss": 0.3927,
|
| 449 |
+
"step": 710
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.5,
|
| 453 |
+
"learning_rate": 4.485492251543856e-05,
|
| 454 |
+
"loss": 0.415,
|
| 455 |
+
"step": 720
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.51,
|
| 459 |
+
"learning_rate": 4.467924668141512e-05,
|
| 460 |
+
"loss": 0.3861,
|
| 461 |
+
"step": 730
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.51,
|
| 465 |
+
"learning_rate": 4.450097813793305e-05,
|
| 466 |
+
"loss": 0.3943,
|
| 467 |
+
"step": 740
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.52,
|
| 471 |
+
"learning_rate": 4.432014037158949e-05,
|
| 472 |
+
"loss": 0.3947,
|
| 473 |
+
"step": 750
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.53,
|
| 477 |
+
"learning_rate": 4.4136757207472626e-05,
|
| 478 |
+
"loss": 0.4037,
|
| 479 |
+
"step": 760
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.53,
|
| 483 |
+
"learning_rate": 4.3950852806022704e-05,
|
| 484 |
+
"loss": 0.3525,
|
| 485 |
+
"step": 770
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.54,
|
| 489 |
+
"learning_rate": 4.376245165984903e-05,
|
| 490 |
+
"loss": 0.3653,
|
| 491 |
+
"step": 780
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.55,
|
| 495 |
+
"learning_rate": 4.357157859050305e-05,
|
| 496 |
+
"loss": 0.3715,
|
| 497 |
+
"step": 790
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.56,
|
| 501 |
+
"learning_rate": 4.3378258745208124e-05,
|
| 502 |
+
"loss": 0.3226,
|
| 503 |
+
"step": 800
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.56,
|
| 507 |
+
"learning_rate": 4.318251759354648e-05,
|
| 508 |
+
"loss": 0.4124,
|
| 509 |
+
"step": 810
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.57,
|
| 513 |
+
"learning_rate": 4.2984380924103614e-05,
|
| 514 |
+
"loss": 0.4113,
|
| 515 |
+
"step": 820
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.58,
|
| 519 |
+
"learning_rate": 4.278387484107067e-05,
|
| 520 |
+
"loss": 0.4211,
|
| 521 |
+
"step": 830
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.58,
|
| 525 |
+
"learning_rate": 4.258102576080526e-05,
|
| 526 |
+
"loss": 0.3897,
|
| 527 |
+
"step": 840
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.59,
|
| 531 |
+
"learning_rate": 4.237586040835117e-05,
|
| 532 |
+
"loss": 0.3771,
|
| 533 |
+
"step": 850
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.6,
|
| 537 |
+
"learning_rate": 4.2168405813917335e-05,
|
| 538 |
+
"loss": 0.3633,
|
| 539 |
+
"step": 860
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.6,
|
| 543 |
+
"learning_rate": 4.1958689309316725e-05,
|
| 544 |
+
"loss": 0.3735,
|
| 545 |
+
"step": 870
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.61,
|
| 549 |
+
"learning_rate": 4.174673852436535e-05,
|
| 550 |
+
"loss": 0.3905,
|
| 551 |
+
"step": 880
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.62,
|
| 555 |
+
"learning_rate": 4.1532581383242094e-05,
|
| 556 |
+
"loss": 0.4141,
|
| 557 |
+
"step": 890
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.62,
|
| 561 |
+
"learning_rate": 4.1316246100809784e-05,
|
| 562 |
+
"loss": 0.3535,
|
| 563 |
+
"step": 900
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.63,
|
| 567 |
+
"learning_rate": 4.109776117889789e-05,
|
| 568 |
+
"loss": 0.3709,
|
| 569 |
+
"step": 910
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.64,
|
| 573 |
+
"learning_rate": 4.087715540254747e-05,
|
| 574 |
+
"loss": 0.3693,
|
| 575 |
+
"step": 920
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.65,
|
| 579 |
+
"learning_rate": 4.065445783621876e-05,
|
| 580 |
+
"loss": 0.3672,
|
| 581 |
+
"step": 930
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.65,
|
| 585 |
+
"learning_rate": 4.042969781996203e-05,
|
| 586 |
+
"loss": 0.3624,
|
| 587 |
+
"step": 940
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 0.66,
|
| 591 |
+
"learning_rate": 4.020290496555202e-05,
|
| 592 |
+
"loss": 0.3854,
|
| 593 |
+
"step": 950
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.67,
|
| 597 |
+
"learning_rate": 3.997410915258669e-05,
|
| 598 |
+
"loss": 0.3877,
|
| 599 |
+
"step": 960
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.67,
|
| 603 |
+
"learning_rate": 3.9743340524550596e-05,
|
| 604 |
+
"loss": 0.3653,
|
| 605 |
+
"step": 970
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.68,
|
| 609 |
+
"learning_rate": 3.9510629484843536e-05,
|
| 610 |
+
"loss": 0.3426,
|
| 611 |
+
"step": 980
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.69,
|
| 615 |
+
"learning_rate": 3.9276006692774964e-05,
|
| 616 |
+
"loss": 0.4158,
|
| 617 |
+
"step": 990
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.69,
|
| 621 |
+
"learning_rate": 3.903950305952466e-05,
|
| 622 |
+
"loss": 0.3796,
|
| 623 |
+
"step": 1000
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 0.69,
|
| 627 |
+
"eval_loss": 0.4698101580142975,
|
| 628 |
+
"eval_runtime": 79.5766,
|
| 629 |
+
"eval_samples_per_second": 8.043,
|
| 630 |
+
"eval_steps_per_second": 2.011,
|
| 631 |
+
"step": 1000
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 0.7,
|
| 635 |
+
"learning_rate": 3.880114974407024e-05,
|
| 636 |
+
"loss": 0.3567,
|
| 637 |
+
"step": 1010
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 0.71,
|
| 641 |
+
"learning_rate": 3.8560978149082026e-05,
|
| 642 |
+
"loss": 0.3904,
|
| 643 |
+
"step": 1020
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.71,
|
| 647 |
+
"learning_rate": 3.831901991678576e-05,
|
| 648 |
+
"loss": 0.35,
|
| 649 |
+
"step": 1030
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.72,
|
| 653 |
+
"learning_rate": 3.807530692479382e-05,
|
| 654 |
+
"loss": 0.3931,
|
| 655 |
+
"step": 1040
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.73,
|
| 659 |
+
"learning_rate": 3.7829871281905354e-05,
|
| 660 |
+
"loss": 0.3523,
|
| 661 |
+
"step": 1050
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.74,
|
| 665 |
+
"learning_rate": 3.7582745323876076e-05,
|
| 666 |
+
"loss": 0.3798,
|
| 667 |
+
"step": 1060
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.74,
|
| 671 |
+
"learning_rate": 3.7333961609157986e-05,
|
| 672 |
+
"loss": 0.3955,
|
| 673 |
+
"step": 1070
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.75,
|
| 677 |
+
"learning_rate": 3.708355291460992e-05,
|
| 678 |
+
"loss": 0.3749,
|
| 679 |
+
"step": 1080
|
| 680 |
+
},
|
| 681 |
+
{
|
| 682 |
+
"epoch": 0.76,
|
| 683 |
+
"learning_rate": 3.683155223117922e-05,
|
| 684 |
+
"loss": 0.395,
|
| 685 |
+
"step": 1090
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.76,
|
| 689 |
+
"learning_rate": 3.657799275955522e-05,
|
| 690 |
+
"loss": 0.3852,
|
| 691 |
+
"step": 1100
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.77,
|
| 695 |
+
"learning_rate": 3.632290790579512e-05,
|
| 696 |
+
"loss": 0.3744,
|
| 697 |
+
"step": 1110
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.78,
|
| 701 |
+
"learning_rate": 3.60663312769228e-05,
|
| 702 |
+
"loss": 0.3901,
|
| 703 |
+
"step": 1120
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.78,
|
| 707 |
+
"learning_rate": 3.580829667650114e-05,
|
| 708 |
+
"loss": 0.3669,
|
| 709 |
+
"step": 1130
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.79,
|
| 713 |
+
"learning_rate": 3.554883810017844e-05,
|
| 714 |
+
"loss": 0.3951,
|
| 715 |
+
"step": 1140
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.8,
|
| 719 |
+
"learning_rate": 3.528798973120959e-05,
|
| 720 |
+
"loss": 0.3885,
|
| 721 |
+
"step": 1150
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.8,
|
| 725 |
+
"learning_rate": 3.5025785935952404e-05,
|
| 726 |
+
"loss": 0.36,
|
| 727 |
+
"step": 1160
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.81,
|
| 731 |
+
"learning_rate": 3.476226125934e-05,
|
| 732 |
+
"loss": 0.3997,
|
| 733 |
+
"step": 1170
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.82,
|
| 737 |
+
"learning_rate": 3.4497450420329485e-05,
|
| 738 |
+
"loss": 0.3966,
|
| 739 |
+
"step": 1180
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.83,
|
| 743 |
+
"learning_rate": 3.423138830732783e-05,
|
| 744 |
+
"loss": 0.367,
|
| 745 |
+
"step": 1190
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.83,
|
| 749 |
+
"learning_rate": 3.396410997359537e-05,
|
| 750 |
+
"loss": 0.3615,
|
| 751 |
+
"step": 1200
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.84,
|
| 755 |
+
"learning_rate": 3.3695650632627576e-05,
|
| 756 |
+
"loss": 0.3618,
|
| 757 |
+
"step": 1210
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.85,
|
| 761 |
+
"learning_rate": 3.342604565351573e-05,
|
| 762 |
+
"loss": 0.3984,
|
| 763 |
+
"step": 1220
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 0.85,
|
| 767 |
+
"learning_rate": 3.315533055628716e-05,
|
| 768 |
+
"loss": 0.3856,
|
| 769 |
+
"step": 1230
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.86,
|
| 773 |
+
"learning_rate": 3.288354100722545e-05,
|
| 774 |
+
"loss": 0.4113,
|
| 775 |
+
"step": 1240
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.87,
|
| 779 |
+
"learning_rate": 3.2610712814171534e-05,
|
| 780 |
+
"loss": 0.4048,
|
| 781 |
+
"step": 1250
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.87,
|
| 785 |
+
"learning_rate": 3.233688192180602e-05,
|
| 786 |
+
"loss": 0.3547,
|
| 787 |
+
"step": 1260
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.88,
|
| 791 |
+
"learning_rate": 3.206208440691356e-05,
|
| 792 |
+
"loss": 0.3758,
|
| 793 |
+
"step": 1270
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.89,
|
| 797 |
+
"learning_rate": 3.17863564736298e-05,
|
| 798 |
+
"loss": 0.3788,
|
| 799 |
+
"step": 1280
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 0.9,
|
| 803 |
+
"learning_rate": 3.150973444867151e-05,
|
| 804 |
+
"loss": 0.3839,
|
| 805 |
+
"step": 1290
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.9,
|
| 809 |
+
"learning_rate": 3.123225477655063e-05,
|
| 810 |
+
"loss": 0.355,
|
| 811 |
+
"step": 1300
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 0.91,
|
| 815 |
+
"learning_rate": 3.095395401477274e-05,
|
| 816 |
+
"loss": 0.4029,
|
| 817 |
+
"step": 1310
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.92,
|
| 821 |
+
"learning_rate": 3.067486882902069e-05,
|
| 822 |
+
"loss": 0.3823,
|
| 823 |
+
"step": 1320
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.92,
|
| 827 |
+
"learning_rate": 3.039503598832394e-05,
|
| 828 |
+
"loss": 0.3735,
|
| 829 |
+
"step": 1330
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.93,
|
| 833 |
+
"learning_rate": 3.011449236021428e-05,
|
| 834 |
+
"loss": 0.3794,
|
| 835 |
+
"step": 1340
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.94,
|
| 839 |
+
"learning_rate": 2.9833274905868635e-05,
|
| 840 |
+
"loss": 0.3865,
|
| 841 |
+
"step": 1350
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.94,
|
| 845 |
+
"learning_rate": 2.955142067523945e-05,
|
| 846 |
+
"loss": 0.3723,
|
| 847 |
+
"step": 1360
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.95,
|
| 851 |
+
"learning_rate": 2.9268966802173436e-05,
|
| 852 |
+
"loss": 0.3733,
|
| 853 |
+
"step": 1370
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 0.96,
|
| 857 |
+
"learning_rate": 2.898595049951926e-05,
|
| 858 |
+
"loss": 0.3802,
|
| 859 |
+
"step": 1380
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.96,
|
| 863 |
+
"learning_rate": 2.8702409054224762e-05,
|
| 864 |
+
"loss": 0.3808,
|
| 865 |
+
"step": 1390
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.97,
|
| 869 |
+
"learning_rate": 2.8418379822424514e-05,
|
| 870 |
+
"loss": 0.3666,
|
| 871 |
+
"step": 1400
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.98,
|
| 875 |
+
"learning_rate": 2.8133900224518167e-05,
|
| 876 |
+
"loss": 0.4061,
|
| 877 |
+
"step": 1410
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.99,
|
| 881 |
+
"learning_rate": 2.784900774024041e-05,
|
| 882 |
+
"loss": 0.3445,
|
| 883 |
+
"step": 1420
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 0.99,
|
| 887 |
+
"learning_rate": 2.7563739903723024e-05,
|
| 888 |
+
"loss": 0.363,
|
| 889 |
+
"step": 1430
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 1.0,
|
| 893 |
+
"learning_rate": 2.7278134298549883e-05,
|
| 894 |
+
"loss": 0.3969,
|
| 895 |
+
"step": 1440
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 1.01,
|
| 899 |
+
"learning_rate": 2.6992228552805342e-05,
|
| 900 |
+
"loss": 0.3143,
|
| 901 |
+
"step": 1450
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 1.01,
|
| 905 |
+
"learning_rate": 2.6706060334116777e-05,
|
| 906 |
+
"loss": 0.3433,
|
| 907 |
+
"step": 1460
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.02,
|
| 911 |
+
"learning_rate": 2.6419667344691974e-05,
|
| 912 |
+
"loss": 0.3077,
|
| 913 |
+
"step": 1470
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.03,
|
| 917 |
+
"learning_rate": 2.613308731635189e-05,
|
| 918 |
+
"loss": 0.3177,
|
| 919 |
+
"step": 1480
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.03,
|
| 923 |
+
"learning_rate": 2.5846358005559584e-05,
|
| 924 |
+
"loss": 0.3146,
|
| 925 |
+
"step": 1490
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 1.04,
|
| 929 |
+
"learning_rate": 2.5559517188445825e-05,
|
| 930 |
+
"loss": 0.353,
|
| 931 |
+
"step": 1500
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 1.04,
|
| 935 |
+
"eval_loss": 0.4521051347255707,
|
| 936 |
+
"eval_runtime": 79.5515,
|
| 937 |
+
"eval_samples_per_second": 8.045,
|
| 938 |
+
"eval_steps_per_second": 2.011,
|
| 939 |
+
"step": 1500
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 1.05,
|
| 943 |
+
"learning_rate": 2.527260265583216e-05,
|
| 944 |
+
"loss": 0.3193,
|
| 945 |
+
"step": 1510
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 1.05,
|
| 949 |
+
"learning_rate": 2.4985652208252055e-05,
|
| 950 |
+
"loss": 0.3137,
|
| 951 |
+
"step": 1520
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 1.06,
|
| 955 |
+
"learning_rate": 2.4698703650970705e-05,
|
| 956 |
+
"loss": 0.2973,
|
| 957 |
+
"step": 1530
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.07,
|
| 961 |
+
"learning_rate": 2.4411794789004254e-05,
|
| 962 |
+
"loss": 0.3087,
|
| 963 |
+
"step": 1540
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.08,
|
| 967 |
+
"learning_rate": 2.412496342213905e-05,
|
| 968 |
+
"loss": 0.2978,
|
| 969 |
+
"step": 1550
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.08,
|
| 973 |
+
"learning_rate": 2.3838247339951595e-05,
|
| 974 |
+
"loss": 0.3012,
|
| 975 |
+
"step": 1560
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.09,
|
| 979 |
+
"learning_rate": 2.355168431682982e-05,
|
| 980 |
+
"loss": 0.3112,
|
| 981 |
+
"step": 1570
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 1.1,
|
| 985 |
+
"learning_rate": 2.3265312106996367e-05,
|
| 986 |
+
"loss": 0.3282,
|
| 987 |
+
"step": 1580
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 1.1,
|
| 991 |
+
"learning_rate": 2.2979168439534505e-05,
|
| 992 |
+
"loss": 0.3003,
|
| 993 |
+
"step": 1590
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.11,
|
| 997 |
+
"learning_rate": 2.2693291013417453e-05,
|
| 998 |
+
"loss": 0.3309,
|
| 999 |
+
"step": 1600
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.12,
|
| 1003 |
+
"learning_rate": 2.2407717492541526e-05,
|
| 1004 |
+
"loss": 0.3306,
|
| 1005 |
+
"step": 1610
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.12,
|
| 1009 |
+
"learning_rate": 2.2122485500763997e-05,
|
| 1010 |
+
"loss": 0.2955,
|
| 1011 |
+
"step": 1620
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.13,
|
| 1015 |
+
"learning_rate": 2.1837632616946266e-05,
|
| 1016 |
+
"loss": 0.3153,
|
| 1017 |
+
"step": 1630
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.14,
|
| 1021 |
+
"learning_rate": 2.1553196370002824e-05,
|
| 1022 |
+
"loss": 0.3004,
|
| 1023 |
+
"step": 1640
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1.15,
|
| 1027 |
+
"learning_rate": 2.1269214233956922e-05,
|
| 1028 |
+
"loss": 0.3089,
|
| 1029 |
+
"step": 1650
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 1.15,
|
| 1033 |
+
"learning_rate": 2.0985723623003433e-05,
|
| 1034 |
+
"loss": 0.3178,
|
| 1035 |
+
"step": 1660
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 1.16,
|
| 1039 |
+
"learning_rate": 2.070276188657959e-05,
|
| 1040 |
+
"loss": 0.3203,
|
| 1041 |
+
"step": 1670
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.17,
|
| 1045 |
+
"learning_rate": 2.042036630444421e-05,
|
| 1046 |
+
"loss": 0.3192,
|
| 1047 |
+
"step": 1680
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.17,
|
| 1051 |
+
"learning_rate": 2.0138574081766224e-05,
|
| 1052 |
+
"loss": 0.3271,
|
| 1053 |
+
"step": 1690
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.18,
|
| 1057 |
+
"learning_rate": 1.9857422344222895e-05,
|
| 1058 |
+
"loss": 0.304,
|
| 1059 |
+
"step": 1700
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.19,
|
| 1063 |
+
"learning_rate": 1.9576948133108585e-05,
|
| 1064 |
+
"loss": 0.3135,
|
| 1065 |
+
"step": 1710
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 1.19,
|
| 1069 |
+
"learning_rate": 1.929718840045463e-05,
|
| 1070 |
+
"loss": 0.3074,
|
| 1071 |
+
"step": 1720
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 1.2,
|
| 1075 |
+
"learning_rate": 1.9018180004160967e-05,
|
| 1076 |
+
"loss": 0.3077,
|
| 1077 |
+
"step": 1730
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.21,
|
| 1081 |
+
"learning_rate": 1.8739959703140167e-05,
|
| 1082 |
+
"loss": 0.3066,
|
| 1083 |
+
"step": 1740
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.21,
|
| 1087 |
+
"learning_rate": 1.846256415247449e-05,
|
| 1088 |
+
"loss": 0.302,
|
| 1089 |
+
"step": 1750
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.22,
|
| 1093 |
+
"learning_rate": 1.8186029898586646e-05,
|
| 1094 |
+
"loss": 0.3125,
|
| 1095 |
+
"step": 1760
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.23,
|
| 1099 |
+
"learning_rate": 1.791039337442489e-05,
|
| 1100 |
+
"loss": 0.3242,
|
| 1101 |
+
"step": 1770
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.24,
|
| 1105 |
+
"learning_rate": 1.7635690894662986e-05,
|
| 1106 |
+
"loss": 0.2919,
|
| 1107 |
+
"step": 1780
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 1.24,
|
| 1111 |
+
"learning_rate": 1.7361958650915844e-05,
|
| 1112 |
+
"loss": 0.2989,
|
| 1113 |
+
"step": 1790
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 1.25,
|
| 1117 |
+
"learning_rate": 1.7089232706971313e-05,
|
| 1118 |
+
"loss": 0.2763,
|
| 1119 |
+
"step": 1800
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.26,
|
| 1123 |
+
"learning_rate": 1.6817548994038864e-05,
|
| 1124 |
+
"loss": 0.3313,
|
| 1125 |
+
"step": 1810
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 1.26,
|
| 1129 |
+
"learning_rate": 1.6546943306015667e-05,
|
| 1130 |
+
"loss": 0.3257,
|
| 1131 |
+
"step": 1820
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.27,
|
| 1135 |
+
"learning_rate": 1.6277451294770834e-05,
|
| 1136 |
+
"loss": 0.294,
|
| 1137 |
+
"step": 1830
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.28,
|
| 1141 |
+
"learning_rate": 1.6009108465448332e-05,
|
| 1142 |
+
"loss": 0.3138,
|
| 1143 |
+
"step": 1840
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.28,
|
| 1147 |
+
"learning_rate": 1.5741950171789245e-05,
|
| 1148 |
+
"loss": 0.3267,
|
| 1149 |
+
"step": 1850
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 1.29,
|
| 1153 |
+
"learning_rate": 1.5476011611473944e-05,
|
| 1154 |
+
"loss": 0.2859,
|
| 1155 |
+
"step": 1860
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 1.3,
|
| 1159 |
+
"learning_rate": 1.52113278214849e-05,
|
| 1160 |
+
"loss": 0.3092,
|
| 1161 |
+
"step": 1870
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.3,
|
| 1165 |
+
"learning_rate": 1.4947933673490567e-05,
|
| 1166 |
+
"loss": 0.3094,
|
| 1167 |
+
"step": 1880
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 1.31,
|
| 1171 |
+
"learning_rate": 1.468586386925113e-05,
|
| 1172 |
+
"loss": 0.3053,
|
| 1173 |
+
"step": 1890
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.32,
|
| 1177 |
+
"learning_rate": 1.4425152936046568e-05,
|
| 1178 |
+
"loss": 0.3089,
|
| 1179 |
+
"step": 1900
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.33,
|
| 1183 |
+
"learning_rate": 1.4165835222127804e-05,
|
| 1184 |
+
"loss": 0.2878,
|
| 1185 |
+
"step": 1910
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.33,
|
| 1189 |
+
"learning_rate": 1.3907944892191294e-05,
|
| 1190 |
+
"loss": 0.296,
|
| 1191 |
+
"step": 1920
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 1.34,
|
| 1195 |
+
"learning_rate": 1.3651515922877933e-05,
|
| 1196 |
+
"loss": 0.2979,
|
| 1197 |
+
"step": 1930
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 1.35,
|
| 1201 |
+
"learning_rate": 1.3396582098296705e-05,
|
| 1202 |
+
"loss": 0.3373,
|
| 1203 |
+
"step": 1940
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.35,
|
| 1207 |
+
"learning_rate": 1.3143177005573626e-05,
|
| 1208 |
+
"loss": 0.305,
|
| 1209 |
+
"step": 1950
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 1.36,
|
| 1213 |
+
"learning_rate": 1.2891334030426722e-05,
|
| 1214 |
+
"loss": 0.3245,
|
| 1215 |
+
"step": 1960
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.37,
|
| 1219 |
+
"learning_rate": 1.2641086352767506e-05,
|
| 1220 |
+
"loss": 0.3012,
|
| 1221 |
+
"step": 1970
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.37,
|
| 1225 |
+
"learning_rate": 1.2392466942329616e-05,
|
| 1226 |
+
"loss": 0.316,
|
| 1227 |
+
"step": 1980
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.38,
|
| 1231 |
+
"learning_rate": 1.2145508554325043e-05,
|
| 1232 |
+
"loss": 0.2949,
|
| 1233 |
+
"step": 1990
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 1.39,
|
| 1237 |
+
"learning_rate": 1.1900243725128715e-05,
|
| 1238 |
+
"loss": 0.3173,
|
| 1239 |
+
"step": 2000
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 1.39,
|
| 1243 |
+
"eval_loss": 0.44626814126968384,
|
| 1244 |
+
"eval_runtime": 79.5481,
|
| 1245 |
+
"eval_samples_per_second": 8.045,
|
| 1246 |
+
"eval_steps_per_second": 2.011,
|
| 1247 |
+
"step": 2000
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 1.39,
|
| 1251 |
+
"learning_rate": 1.1656704767991902e-05,
|
| 1252 |
+
"loss": 0.3123,
|
| 1253 |
+
"step": 2010
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 1.4,
|
| 1257 |
+
"learning_rate": 1.1414923768784953e-05,
|
| 1258 |
+
"loss": 0.3117,
|
| 1259 |
+
"step": 2020
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.41,
|
| 1263 |
+
"learning_rate": 1.117493258177006e-05,
|
| 1264 |
+
"loss": 0.2914,
|
| 1265 |
+
"step": 2030
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 1.42,
|
| 1269 |
+
"learning_rate": 1.0936762825404492e-05,
|
| 1270 |
+
"loss": 0.2955,
|
| 1271 |
+
"step": 2040
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.42,
|
| 1275 |
+
"learning_rate": 1.0700445878174969e-05,
|
| 1276 |
+
"loss": 0.2983,
|
| 1277 |
+
"step": 2050
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.43,
|
| 1281 |
+
"learning_rate": 1.0466012874463507e-05,
|
| 1282 |
+
"loss": 0.31,
|
| 1283 |
+
"step": 2060
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.44,
|
| 1287 |
+
"learning_rate": 1.0233494700445573e-05,
|
| 1288 |
+
"loss": 0.295,
|
| 1289 |
+
"step": 2070
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 1.44,
|
| 1293 |
+
"learning_rate": 1.0002921990020885e-05,
|
| 1294 |
+
"loss": 0.3042,
|
| 1295 |
+
"step": 2080
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 1.45,
|
| 1299 |
+
"learning_rate": 9.774325120777405e-06,
|
| 1300 |
+
"loss": 0.3037,
|
| 1301 |
+
"step": 2090
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.46,
|
| 1305 |
+
"learning_rate": 9.547734209989168e-06,
|
| 1306 |
+
"loss": 0.2866,
|
| 1307 |
+
"step": 2100
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 1.46,
|
| 1311 |
+
"learning_rate": 9.323179110648361e-06,
|
| 1312 |
+
"loss": 0.2923,
|
| 1313 |
+
"step": 2110
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.47,
|
| 1317 |
+
"learning_rate": 9.100689407532278e-06,
|
| 1318 |
+
"loss": 0.3533,
|
| 1319 |
+
"step": 2120
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.48,
|
| 1323 |
+
"learning_rate": 8.880294413305523e-06,
|
| 1324 |
+
"loss": 0.3055,
|
| 1325 |
+
"step": 2130
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.49,
|
| 1329 |
+
"learning_rate": 8.662023164658125e-06,
|
| 1330 |
+
"loss": 0.2817,
|
| 1331 |
+
"step": 2140
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.49,
|
| 1335 |
+
"learning_rate": 8.445904418480044e-06,
|
| 1336 |
+
"loss": 0.2986,
|
| 1337 |
+
"step": 2150
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 1.5,
|
| 1341 |
+
"learning_rate": 8.231966648072432e-06,
|
| 1342 |
+
"loss": 0.2918,
|
| 1343 |
+
"step": 2160
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.51,
|
| 1347 |
+
"learning_rate": 8.020238039396352e-06,
|
| 1348 |
+
"loss": 0.3322,
|
| 1349 |
+
"step": 2170
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.51,
|
| 1353 |
+
"learning_rate": 7.810746487359351e-06,
|
| 1354 |
+
"loss": 0.3345,
|
| 1355 |
+
"step": 2180
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.52,
|
| 1359 |
+
"learning_rate": 7.603519592140293e-06,
|
| 1360 |
+
"loss": 0.29,
|
| 1361 |
+
"step": 2190
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.53,
|
| 1365 |
+
"learning_rate": 7.398584655553109e-06,
|
| 1366 |
+
"loss": 0.2871,
|
| 1367 |
+
"step": 2200
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.53,
|
| 1371 |
+
"learning_rate": 7.195968677449827e-06,
|
| 1372 |
+
"loss": 0.2861,
|
| 1373 |
+
"step": 2210
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 1.54,
|
| 1377 |
+
"learning_rate": 6.995698352163402e-06,
|
| 1378 |
+
"loss": 0.3197,
|
| 1379 |
+
"step": 2220
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 1.55,
|
| 1383 |
+
"learning_rate": 6.7978000649907335e-06,
|
| 1384 |
+
"loss": 0.3123,
|
| 1385 |
+
"step": 2230
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.55,
|
| 1389 |
+
"learning_rate": 6.602299888716479e-06,
|
| 1390 |
+
"loss": 0.3117,
|
| 1391 |
+
"step": 2240
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.56,
|
| 1395 |
+
"learning_rate": 6.409223580178009e-06,
|
| 1396 |
+
"loss": 0.3292,
|
| 1397 |
+
"step": 2250
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.57,
|
| 1401 |
+
"learning_rate": 6.218596576871963e-06,
|
| 1402 |
+
"loss": 0.2987,
|
| 1403 |
+
"step": 2260
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.58,
|
| 1407 |
+
"learning_rate": 6.0304439936029135e-06,
|
| 1408 |
+
"loss": 0.2915,
|
| 1409 |
+
"step": 2270
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.58,
|
| 1413 |
+
"learning_rate": 5.84479061917452e-06,
|
| 1414 |
+
"loss": 0.2838,
|
| 1415 |
+
"step": 2280
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 1.59,
|
| 1419 |
+
"learning_rate": 5.661660913123673e-06,
|
| 1420 |
+
"loss": 0.3064,
|
| 1421 |
+
"step": 2290
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 1.6,
|
| 1425 |
+
"learning_rate": 5.481079002497947e-06,
|
| 1426 |
+
"loss": 0.3264,
|
| 1427 |
+
"step": 2300
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.6,
|
| 1431 |
+
"learning_rate": 5.30306867867692e-06,
|
| 1432 |
+
"loss": 0.2947,
|
| 1433 |
+
"step": 2310
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 1.61,
|
| 1437 |
+
"learning_rate": 5.127653394237717e-06,
|
| 1438 |
+
"loss": 0.2978,
|
| 1439 |
+
"step": 2320
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.62,
|
| 1443 |
+
"learning_rate": 4.9548562598651286e-06,
|
| 1444 |
+
"loss": 0.2788,
|
| 1445 |
+
"step": 2330
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.62,
|
| 1449 |
+
"learning_rate": 4.78470004130685e-06,
|
| 1450 |
+
"loss": 0.3119,
|
| 1451 |
+
"step": 2340
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.63,
|
| 1455 |
+
"learning_rate": 4.61720715637411e-06,
|
| 1456 |
+
"loss": 0.2923,
|
| 1457 |
+
"step": 2350
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 1.64,
|
| 1461 |
+
"learning_rate": 4.452399671988175e-06,
|
| 1462 |
+
"loss": 0.2904,
|
| 1463 |
+
"step": 2360
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 1.64,
|
| 1467 |
+
"learning_rate": 4.290299301273054e-06,
|
| 1468 |
+
"loss": 0.2994,
|
| 1469 |
+
"step": 2370
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.65,
|
| 1473 |
+
"learning_rate": 4.130927400694831e-06,
|
| 1474 |
+
"loss": 0.3022,
|
| 1475 |
+
"step": 2380
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 1.66,
|
| 1479 |
+
"learning_rate": 3.974304967248002e-06,
|
| 1480 |
+
"loss": 0.3162,
|
| 1481 |
+
"step": 2390
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.67,
|
| 1485 |
+
"learning_rate": 3.820452635689126e-06,
|
| 1486 |
+
"loss": 0.3139,
|
| 1487 |
+
"step": 2400
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.67,
|
| 1491 |
+
"learning_rate": 3.6693906758182462e-06,
|
| 1492 |
+
"loss": 0.3179,
|
| 1493 |
+
"step": 2410
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.68,
|
| 1497 |
+
"learning_rate": 3.521138989808373e-06,
|
| 1498 |
+
"loss": 0.3531,
|
| 1499 |
+
"step": 2420
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 1.69,
|
| 1503 |
+
"learning_rate": 3.3757171095834124e-06,
|
| 1504 |
+
"loss": 0.2865,
|
| 1505 |
+
"step": 2430
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.69,
|
| 1509 |
+
"learning_rate": 3.2331441942448442e-06,
|
| 1510 |
+
"loss": 0.3204,
|
| 1511 |
+
"step": 2440
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.7,
|
| 1515 |
+
"learning_rate": 3.093439027547562e-06,
|
| 1516 |
+
"loss": 0.2953,
|
| 1517 |
+
"step": 2450
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.71,
|
| 1521 |
+
"learning_rate": 2.9566200154251407e-06,
|
| 1522 |
+
"loss": 0.2895,
|
| 1523 |
+
"step": 2460
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.71,
|
| 1527 |
+
"learning_rate": 2.8227051835648722e-06,
|
| 1528 |
+
"loss": 0.3028,
|
| 1529 |
+
"step": 2470
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.72,
|
| 1533 |
+
"learning_rate": 2.691712175032912e-06,
|
| 1534 |
+
"loss": 0.2979,
|
| 1535 |
+
"step": 2480
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.73,
|
| 1539 |
+
"learning_rate": 2.5636582479498517e-06,
|
| 1540 |
+
"loss": 0.2922,
|
| 1541 |
+
"step": 2490
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 1.73,
|
| 1545 |
+
"learning_rate": 2.4385602732169692e-06,
|
| 1546 |
+
"loss": 0.3276,
|
| 1547 |
+
"step": 2500
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1.73,
|
| 1551 |
+
"eval_loss": 0.44305261969566345,
|
| 1552 |
+
"eval_runtime": 79.5937,
|
| 1553 |
+
"eval_samples_per_second": 8.041,
|
| 1554 |
+
"eval_steps_per_second": 2.01,
|
| 1555 |
+
"step": 2500
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 1.74,
|
| 1559 |
+
"learning_rate": 2.3164347322935128e-06,
|
| 1560 |
+
"loss": 0.324,
|
| 1561 |
+
"step": 2510
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 1.75,
|
| 1565 |
+
"learning_rate": 2.1972977150253064e-06,
|
| 1566 |
+
"loss": 0.3013,
|
| 1567 |
+
"step": 2520
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 1.76,
|
| 1571 |
+
"learning_rate": 2.0811649175249497e-06,
|
| 1572 |
+
"loss": 0.2888,
|
| 1573 |
+
"step": 2530
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 1.76,
|
| 1577 |
+
"learning_rate": 1.9680516401038435e-06,
|
| 1578 |
+
"loss": 0.3062,
|
| 1579 |
+
"step": 2540
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 1.77,
|
| 1583 |
+
"learning_rate": 1.8579727852564287e-06,
|
| 1584 |
+
"loss": 0.329,
|
| 1585 |
+
"step": 2550
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.78,
|
| 1589 |
+
"learning_rate": 1.750942855696802e-06,
|
| 1590 |
+
"loss": 0.3261,
|
| 1591 |
+
"step": 2560
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.78,
|
| 1595 |
+
"learning_rate": 1.6469759524479933e-06,
|
| 1596 |
+
"loss": 0.3179,
|
| 1597 |
+
"step": 2570
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 1.79,
|
| 1601 |
+
"learning_rate": 1.5460857729841877e-06,
|
| 1602 |
+
"loss": 0.3144,
|
| 1603 |
+
"step": 2580
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 1.8,
|
| 1607 |
+
"learning_rate": 1.448285609426106e-06,
|
| 1608 |
+
"loss": 0.333,
|
| 1609 |
+
"step": 2590
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 1.8,
|
| 1613 |
+
"learning_rate": 1.3535883467897914e-06,
|
| 1614 |
+
"loss": 0.3049,
|
| 1615 |
+
"step": 2600
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.81,
|
| 1619 |
+
"learning_rate": 1.2620064612890142e-06,
|
| 1620 |
+
"loss": 0.3349,
|
| 1621 |
+
"step": 2610
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 1.82,
|
| 1625 |
+
"learning_rate": 1.1735520186915678e-06,
|
| 1626 |
+
"loss": 0.2748,
|
| 1627 |
+
"step": 2620
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.83,
|
| 1631 |
+
"learning_rate": 1.0882366727296173e-06,
|
| 1632 |
+
"loss": 0.3011,
|
| 1633 |
+
"step": 2630
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.83,
|
| 1637 |
+
"learning_rate": 1.0060716635643313e-06,
|
| 1638 |
+
"loss": 0.3385,
|
| 1639 |
+
"step": 2640
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 1.84,
|
| 1643 |
+
"learning_rate": 9.270678163050217e-07,
|
| 1644 |
+
"loss": 0.2985,
|
| 1645 |
+
"step": 2650
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.85,
|
| 1649 |
+
"learning_rate": 8.512355395829341e-07,
|
| 1650 |
+
"loss": 0.2894,
|
| 1651 |
+
"step": 2660
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 1.85,
|
| 1655 |
+
"learning_rate": 7.785848241799564e-07,
|
| 1656 |
+
"loss": 0.2569,
|
| 1657 |
+
"step": 2670
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 1.86,
|
| 1661 |
+
"learning_rate": 7.09125241712319e-07,
|
| 1662 |
+
"loss": 0.3214,
|
| 1663 |
+
"step": 2680
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 1.87,
|
| 1667 |
+
"learning_rate": 6.42865943369561e-07,
|
| 1668 |
+
"loss": 0.2708,
|
| 1669 |
+
"step": 2690
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.87,
|
| 1673 |
+
"learning_rate": 5.798156587088877e-07,
|
| 1674 |
+
"loss": 0.3013,
|
| 1675 |
+
"step": 2700
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.88,
|
| 1679 |
+
"learning_rate": 5.199826945050473e-07,
|
| 1680 |
+
"loss": 0.3242,
|
| 1681 |
+
"step": 2710
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 1.89,
|
| 1685 |
+
"learning_rate": 4.6337493365592733e-07,
|
| 1686 |
+
"loss": 0.305,
|
| 1687 |
+
"step": 2720
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 1.89,
|
| 1691 |
+
"learning_rate": 4.0999983414401754e-07,
|
| 1692 |
+
"loss": 0.3078,
|
| 1693 |
+
"step": 2730
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 1.9,
|
| 1697 |
+
"learning_rate": 3.598644280538049e-07,
|
| 1698 |
+
"loss": 0.3247,
|
| 1699 |
+
"step": 2740
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 1.91,
|
| 1703 |
+
"learning_rate": 3.129753206453201e-07,
|
| 1704 |
+
"loss": 0.2903,
|
| 1705 |
+
"step": 2750
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.92,
|
| 1709 |
+
"learning_rate": 2.6933868948390284e-07,
|
| 1710 |
+
"loss": 0.3032,
|
| 1711 |
+
"step": 2760
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.92,
|
| 1715 |
+
"learning_rate": 2.2896028362633383e-07,
|
| 1716 |
+
"loss": 0.3086,
|
| 1717 |
+
"step": 2770
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.93,
|
| 1721 |
+
"learning_rate": 1.9184542286337092e-07,
|
| 1722 |
+
"loss": 0.3085,
|
| 1723 |
+
"step": 2780
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 1.94,
|
| 1727 |
+
"learning_rate": 1.5799899701890443e-07,
|
| 1728 |
+
"loss": 0.3126,
|
| 1729 |
+
"step": 2790
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 1.94,
|
| 1733 |
+
"learning_rate": 1.2742546530572496e-07,
|
| 1734 |
+
"loss": 0.328,
|
| 1735 |
+
"step": 2800
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.95,
|
| 1739 |
+
"learning_rate": 1.0012885573802145e-07,
|
| 1740 |
+
"loss": 0.2739,
|
| 1741 |
+
"step": 2810
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 1.96,
|
| 1745 |
+
"learning_rate": 7.611276460069428e-08,
|
| 1746 |
+
"loss": 0.2971,
|
| 1747 |
+
"step": 2820
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 1.96,
|
| 1751 |
+
"learning_rate": 5.538035597556779e-08,
|
| 1752 |
+
"loss": 0.3123,
|
| 1753 |
+
"step": 2830
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.97,
|
| 1757 |
+
"learning_rate": 3.7934361324512556e-08,
|
| 1758 |
+
"loss": 0.3296,
|
| 1759 |
+
"step": 2840
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.98,
|
| 1763 |
+
"learning_rate": 2.377707912958327e-08,
|
| 1764 |
+
"loss": 0.2975,
|
| 1765 |
+
"step": 2850
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 1.98,
|
| 1769 |
+
"learning_rate": 1.2910374590194285e-08,
|
| 1770 |
+
"loss": 0.3368,
|
| 1771 |
+
"step": 2860
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 1.99,
|
| 1775 |
+
"learning_rate": 5.335679377391212e-09,
|
| 1776 |
+
"loss": 0.3273,
|
| 1777 |
+
"step": 2870
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 2.0,
|
| 1781 |
+
"learning_rate": 1.0539914452267762e-09,
|
| 1782 |
+
"loss": 0.2978,
|
| 1783 |
+
"step": 2880
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 2.0,
|
| 1787 |
+
"step": 2882,
|
| 1788 |
+
"total_flos": 2.5530693253595136e+17,
|
| 1789 |
+
"train_loss": 0.3615408574310464,
|
| 1790 |
+
"train_runtime": 5165.0048,
|
| 1791 |
+
"train_samples_per_second": 2.231,
|
| 1792 |
+
"train_steps_per_second": 0.558
|
| 1793 |
+
}
|
| 1794 |
+
],
|
| 1795 |
+
"logging_steps": 10,
|
| 1796 |
+
"max_steps": 2882,
|
| 1797 |
+
"num_train_epochs": 2,
|
| 1798 |
+
"save_steps": 500,
|
| 1799 |
+
"total_flos": 2.5530693253595136e+17,
|
| 1800 |
+
"trial_name": null,
|
| 1801 |
+
"trial_params": null
|
| 1802 |
+
}
|