Model save
Browse files- README.md +4 -7
- all_results.json +9 -9
- config_argument.yaml +188 -0
- eval_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +40 -1786
README.md
CHANGED
|
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss:
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
@@ -42,17 +42,14 @@ The following hyperparameters were used during training:
|
|
| 42 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 43 |
- lr_scheduler_type: cosine
|
| 44 |
- lr_scheduler_warmup_ratio: 0.05
|
| 45 |
-
-
|
| 46 |
|
| 47 |
### Training results
|
| 48 |
|
| 49 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 50 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 51 |
-
|
|
| 52 |
-
|
|
| 53 |
-
| 0.5553 | 1.03 | 1500 | 0.5688 |
|
| 54 |
-
| 0.5175 | 1.37 | 2000 | 0.5533 |
|
| 55 |
-
| 0.5614 | 1.71 | 2500 | 0.5487 |
|
| 56 |
|
| 57 |
|
| 58 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model is a fine-tuned version of [HuggingFaceH4/zephyr-7b-beta](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) on the None dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 1.1997
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 42 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 43 |
- lr_scheduler_type: cosine
|
| 44 |
- lr_scheduler_warmup_ratio: 0.05
|
| 45 |
+
- training_steps: 50
|
| 46 |
|
| 47 |
### Training results
|
| 48 |
|
| 49 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 50 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 51 |
+
| 1.5219 | 0.02 | 25 | 1.2539 |
|
| 52 |
+
| 1.3156 | 0.03 | 50 | 1.1997 |
|
|
|
|
|
|
|
|
|
|
| 53 |
|
| 54 |
|
| 55 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss":
|
| 4 |
-
"eval_runtime": 112.
|
| 5 |
"eval_samples": 650,
|
| 6 |
-
"eval_samples_per_second": 5.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
-
"train_loss":
|
| 9 |
-
"train_runtime":
|
| 10 |
"train_samples": 5845,
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 0.03,
|
| 3 |
+
"eval_loss": 1.1997405290603638,
|
| 4 |
+
"eval_runtime": 112.9062,
|
| 5 |
"eval_samples": 650,
|
| 6 |
+
"eval_samples_per_second": 5.757,
|
| 7 |
+
"eval_steps_per_second": 1.444,
|
| 8 |
+
"train_loss": 1.4715181255340577,
|
| 9 |
+
"train_runtime": 335.1031,
|
| 10 |
"train_samples": 5845,
|
| 11 |
+
"train_samples_per_second": 0.597,
|
| 12 |
+
"train_steps_per_second": 0.149
|
| 13 |
}
|
config_argument.yaml
ADDED
|
@@ -0,0 +1,188 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
!!python/tuple
|
| 2 |
+
- !!python/object:__main__.ModelArguments
|
| 3 |
+
bnb_4bit_quant_type: nf4
|
| 4 |
+
cache_dir: ./cache
|
| 5 |
+
device_map: auto
|
| 6 |
+
load_in_4bit: true
|
| 7 |
+
load_in_8bit: false
|
| 8 |
+
model_name_or_path: HuggingFaceH4/zephyr-7b-beta
|
| 9 |
+
model_revision: main
|
| 10 |
+
model_type: auto
|
| 11 |
+
neft_alpha: 0
|
| 12 |
+
rope_scaling: null
|
| 13 |
+
shift_attn: false
|
| 14 |
+
tokenizer_name_or_path: null
|
| 15 |
+
torch_dtype: float16
|
| 16 |
+
trust_remote_code: true
|
| 17 |
+
use_bnb_nested_quant: false
|
| 18 |
+
use_fast_tokenizer: false
|
| 19 |
+
use_flash_attention_2: false
|
| 20 |
+
- !!python/object:__main__.DataArguments
|
| 21 |
+
dataset_config_name: null
|
| 22 |
+
dataset_name: null
|
| 23 |
+
ignore_pad_token_for_loss: true
|
| 24 |
+
max_eval_samples: null
|
| 25 |
+
max_train_samples: null
|
| 26 |
+
overwrite_cache: false
|
| 27 |
+
preprocessing_num_workers: 4
|
| 28 |
+
template_name: vicuna
|
| 29 |
+
train_file_dir: datasets/finetune
|
| 30 |
+
validation_file_dir: null
|
| 31 |
+
validation_split_percentage: 10
|
| 32 |
+
- !!python/object:__main__.SFTConfig
|
| 33 |
+
__cached__setup_devices: !!python/object/apply:torch.device
|
| 34 |
+
- cuda
|
| 35 |
+
- 0
|
| 36 |
+
_n_gpu: 1
|
| 37 |
+
adafactor: false
|
| 38 |
+
adam_beta1: 0.9
|
| 39 |
+
adam_beta2: 0.999
|
| 40 |
+
adam_epsilon: 1.0e-08
|
| 41 |
+
auto_find_batch_size: false
|
| 42 |
+
bf16: false
|
| 43 |
+
bf16_full_eval: false
|
| 44 |
+
data_seed: null
|
| 45 |
+
dataloader_drop_last: false
|
| 46 |
+
dataloader_num_workers: 0
|
| 47 |
+
dataloader_pin_memory: true
|
| 48 |
+
ddp_backend: null
|
| 49 |
+
ddp_broadcast_buffers: null
|
| 50 |
+
ddp_bucket_cap_mb: null
|
| 51 |
+
ddp_find_unused_parameters: false
|
| 52 |
+
ddp_timeout: 30000
|
| 53 |
+
debug: []
|
| 54 |
+
deepspeed: null
|
| 55 |
+
deepspeed_plugin: null
|
| 56 |
+
disable_tqdm: false
|
| 57 |
+
dispatch_batches: null
|
| 58 |
+
distributed_state: !!python/object:accelerate.state.PartialState
|
| 59 |
+
_cpu: false
|
| 60 |
+
backend: null
|
| 61 |
+
debug: false
|
| 62 |
+
device: !!python/object/apply:torch.device
|
| 63 |
+
- cuda
|
| 64 |
+
- 0
|
| 65 |
+
distributed_type: !!python/object/apply:accelerate.utils.dataclasses.DistributedType
|
| 66 |
+
- MULTI_GPU
|
| 67 |
+
fork_launched: false
|
| 68 |
+
local_process_index: 0
|
| 69 |
+
num_processes: 1
|
| 70 |
+
process_index: 0
|
| 71 |
+
do_eval: true
|
| 72 |
+
do_predict: false
|
| 73 |
+
do_train: true
|
| 74 |
+
eval_accumulation_steps: null
|
| 75 |
+
eval_delay: 0
|
| 76 |
+
eval_steps: 25
|
| 77 |
+
evaluation_strategy: !!python/object/apply:transformers.trainer_utils.IntervalStrategy
|
| 78 |
+
- steps
|
| 79 |
+
fp16: false
|
| 80 |
+
fp16_backend: auto
|
| 81 |
+
fp16_full_eval: false
|
| 82 |
+
fp16_opt_level: O1
|
| 83 |
+
fsdp: []
|
| 84 |
+
fsdp_config:
|
| 85 |
+
min_num_params: 0
|
| 86 |
+
xla: false
|
| 87 |
+
xla_fsdp_grad_ckpt: false
|
| 88 |
+
fsdp_min_num_params: 0
|
| 89 |
+
fsdp_transformer_layer_cls_to_wrap: null
|
| 90 |
+
full_determinism: false
|
| 91 |
+
gradient_accumulation_steps: 1
|
| 92 |
+
gradient_checkpointing: true
|
| 93 |
+
gradient_checkpointing_kwargs:
|
| 94 |
+
use_reentrant: false
|
| 95 |
+
greater_is_better: null
|
| 96 |
+
group_by_length: false
|
| 97 |
+
half_precision_backend: auto
|
| 98 |
+
hub_always_push: false
|
| 99 |
+
hub_model_id: hllj/non-qa-sft-zephyr-7b-beta-v1
|
| 100 |
+
hub_private_repo: false
|
| 101 |
+
hub_strategy: !!python/object/apply:transformers.trainer_utils.HubStrategy
|
| 102 |
+
- every_save
|
| 103 |
+
hub_token: null
|
| 104 |
+
ignore_data_skip: false
|
| 105 |
+
include_inputs_for_metrics: false
|
| 106 |
+
include_tokens_per_second: false
|
| 107 |
+
jit_mode_eval: false
|
| 108 |
+
label_names: null
|
| 109 |
+
label_smoothing_factor: 0.0
|
| 110 |
+
learning_rate: 3.0e-05
|
| 111 |
+
length_column_name: length
|
| 112 |
+
load_best_model_at_end: false
|
| 113 |
+
local_rank: 0
|
| 114 |
+
log_level: info
|
| 115 |
+
log_level_replica: warning
|
| 116 |
+
log_on_each_node: true
|
| 117 |
+
logging_dir: outputs-sft-zephyr-beta-v1/runs/Nov22_05-52-29_a72e59c0abac
|
| 118 |
+
logging_first_step: true
|
| 119 |
+
logging_nan_inf_filter: true
|
| 120 |
+
logging_steps: 10
|
| 121 |
+
logging_strategy: !!python/object/apply:transformers.trainer_utils.IntervalStrategy
|
| 122 |
+
- steps
|
| 123 |
+
lr_scheduler_type: !!python/object/apply:transformers.trainer_utils.SchedulerType
|
| 124 |
+
- cosine
|
| 125 |
+
max_grad_norm: 1.0
|
| 126 |
+
max_seq_length: 512
|
| 127 |
+
max_steps: 50
|
| 128 |
+
metric_for_best_model: null
|
| 129 |
+
mp_parameters: ''
|
| 130 |
+
neftune_noise_alpha: null
|
| 131 |
+
no_cuda: false
|
| 132 |
+
num_train_epochs: 3.0
|
| 133 |
+
optim: !!python/object/apply:transformers.training_args.OptimizerNames
|
| 134 |
+
- adamw_torch
|
| 135 |
+
optim_args: null
|
| 136 |
+
output_dir: outputs-sft-zephyr-beta-v1
|
| 137 |
+
overwrite_output_dir: true
|
| 138 |
+
past_index: -1
|
| 139 |
+
per_device_eval_batch_size: 4
|
| 140 |
+
per_device_train_batch_size: 4
|
| 141 |
+
per_gpu_eval_batch_size: null
|
| 142 |
+
per_gpu_train_batch_size: null
|
| 143 |
+
prediction_loss_only: false
|
| 144 |
+
push_to_hub: true
|
| 145 |
+
push_to_hub_model_id: null
|
| 146 |
+
push_to_hub_organization: null
|
| 147 |
+
push_to_hub_token: null
|
| 148 |
+
ray_scope: last
|
| 149 |
+
remove_unused_columns: true
|
| 150 |
+
report_to:
|
| 151 |
+
- wandb
|
| 152 |
+
resume_from_checkpoint: null
|
| 153 |
+
run_name: sft-zephyr-7b-beta-v1
|
| 154 |
+
save_on_each_node: false
|
| 155 |
+
save_safetensors: true
|
| 156 |
+
save_steps: 25
|
| 157 |
+
save_strategy: !!python/object/apply:transformers.trainer_utils.IntervalStrategy
|
| 158 |
+
- steps
|
| 159 |
+
save_total_limit: 13
|
| 160 |
+
seed: 42
|
| 161 |
+
skip_memory_metrics: true
|
| 162 |
+
split_batches: false
|
| 163 |
+
tf32: null
|
| 164 |
+
torch_compile: false
|
| 165 |
+
torch_compile_backend: null
|
| 166 |
+
torch_compile_mode: null
|
| 167 |
+
torchdynamo: null
|
| 168 |
+
tpu_metrics_debug: false
|
| 169 |
+
tpu_num_cores: null
|
| 170 |
+
use_cpu: false
|
| 171 |
+
use_ipex: false
|
| 172 |
+
use_legacy_prediction_loop: false
|
| 173 |
+
use_mps_device: false
|
| 174 |
+
warmup_ratio: 0.05
|
| 175 |
+
warmup_steps: 0
|
| 176 |
+
weight_decay: 0.05
|
| 177 |
+
- !!python/object:__main__.ScriptArguments
|
| 178 |
+
lora_alpha: 16
|
| 179 |
+
lora_dropout: 0.1
|
| 180 |
+
lora_modules_to_save: null
|
| 181 |
+
lora_r: 64
|
| 182 |
+
lora_target_modules:
|
| 183 |
+
- q_proj
|
| 184 |
+
- k_proj
|
| 185 |
+
- v_proj
|
| 186 |
+
- o_proj
|
| 187 |
+
peft_path: null
|
| 188 |
+
use_peft: true
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss":
|
| 4 |
-
"eval_runtime": 112.
|
| 5 |
"eval_samples": 650,
|
| 6 |
-
"eval_samples_per_second": 5.
|
| 7 |
-
"eval_steps_per_second": 1.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 0.03,
|
| 3 |
+
"eval_loss": 1.1997405290603638,
|
| 4 |
+
"eval_runtime": 112.9062,
|
| 5 |
"eval_samples": 650,
|
| 6 |
+
"eval_samples_per_second": 5.757,
|
| 7 |
+
"eval_steps_per_second": 1.444
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 5845,
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 0.03,
|
| 3 |
+
"train_loss": 1.4715181255340577,
|
| 4 |
+
"train_runtime": 335.1031,
|
| 5 |
"train_samples": 5845,
|
| 6 |
+
"train_samples_per_second": 0.597,
|
| 7 |
+
"train_steps_per_second": 0.149
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,1826 +1,80 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"eval_steps":
|
| 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": 0.0,
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.01,
|
| 19 |
-
"learning_rate": 2.
|
| 20 |
-
"loss": 1.
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.01,
|
| 25 |
-
"learning_rate":
|
| 26 |
-
"loss": 1.
|
| 27 |
"step": 20
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.02,
|
| 31 |
-
"
|
| 32 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
"step": 30
|
| 34 |
},
|
| 35 |
{
|
| 36 |
"epoch": 0.03,
|
| 37 |
-
"learning_rate":
|
| 38 |
-
"loss": 1.
|
| 39 |
"step": 40
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.03,
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"
|
| 50 |
-
"
|
| 51 |
-
"
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
"epoch": 0.05,
|
| 55 |
-
"learning_rate": 1.4285714285714285e-05,
|
| 56 |
-
"loss": 1.3685,
|
| 57 |
-
"step": 70
|
| 58 |
-
},
|
| 59 |
-
{
|
| 60 |
-
"epoch": 0.05,
|
| 61 |
-
"learning_rate": 1.63265306122449e-05,
|
| 62 |
-
"loss": 1.3332,
|
| 63 |
-
"step": 80
|
| 64 |
-
},
|
| 65 |
-
{
|
| 66 |
-
"epoch": 0.06,
|
| 67 |
-
"learning_rate": 1.836734693877551e-05,
|
| 68 |
-
"loss": 1.1891,
|
| 69 |
-
"step": 90
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.07,
|
| 73 |
-
"learning_rate": 2.0408163265306123e-05,
|
| 74 |
-
"loss": 0.9755,
|
| 75 |
-
"step": 100
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.08,
|
| 79 |
-
"learning_rate": 2.2448979591836737e-05,
|
| 80 |
-
"loss": 0.9277,
|
| 81 |
-
"step": 110
|
| 82 |
-
},
|
| 83 |
-
{
|
| 84 |
-
"epoch": 0.08,
|
| 85 |
-
"learning_rate": 2.448979591836735e-05,
|
| 86 |
-
"loss": 0.8792,
|
| 87 |
-
"step": 120
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.09,
|
| 91 |
-
"learning_rate": 2.653061224489796e-05,
|
| 92 |
-
"loss": 0.8798,
|
| 93 |
-
"step": 130
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.1,
|
| 97 |
-
"learning_rate": 2.857142857142857e-05,
|
| 98 |
-
"loss": 0.8967,
|
| 99 |
-
"step": 140
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.1,
|
| 103 |
-
"learning_rate": 2.9999913612413297e-05,
|
| 104 |
-
"loss": 0.8726,
|
| 105 |
-
"step": 150
|
| 106 |
-
},
|
| 107 |
-
{
|
| 108 |
-
"epoch": 0.11,
|
| 109 |
-
"learning_rate": 2.9998377860774987e-05,
|
| 110 |
-
"loss": 0.8242,
|
| 111 |
-
"step": 160
|
| 112 |
-
},
|
| 113 |
-
{
|
| 114 |
-
"epoch": 0.12,
|
| 115 |
-
"learning_rate": 2.9994922611221096e-05,
|
| 116 |
-
"loss": 0.762,
|
| 117 |
-
"step": 170
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"epoch": 0.12,
|
| 121 |
-
"learning_rate": 2.998954830595574e-05,
|
| 122 |
-
"loss": 0.854,
|
| 123 |
-
"step": 180
|
| 124 |
-
},
|
| 125 |
-
{
|
| 126 |
-
"epoch": 0.13,
|
| 127 |
-
"learning_rate": 2.9982255632784477e-05,
|
| 128 |
-
"loss": 0.7793,
|
| 129 |
-
"step": 190
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.14,
|
| 133 |
-
"learning_rate": 2.9973045525026307e-05,
|
| 134 |
-
"loss": 0.7864,
|
| 135 |
-
"step": 200
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.14,
|
| 139 |
-
"learning_rate": 2.9961919161394204e-05,
|
| 140 |
-
"loss": 0.7602,
|
| 141 |
-
"step": 210
|
| 142 |
-
},
|
| 143 |
-
{
|
| 144 |
-
"epoch": 0.15,
|
| 145 |
-
"learning_rate": 2.9948877965844274e-05,
|
| 146 |
-
"loss": 0.7455,
|
| 147 |
-
"step": 220
|
| 148 |
-
},
|
| 149 |
-
{
|
| 150 |
-
"epoch": 0.16,
|
| 151 |
-
"learning_rate": 2.9933923607393534e-05,
|
| 152 |
-
"loss": 0.7094,
|
| 153 |
-
"step": 230
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.16,
|
| 157 |
-
"learning_rate": 2.991705799990627e-05,
|
| 158 |
-
"loss": 0.847,
|
| 159 |
-
"step": 240
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.17,
|
| 163 |
-
"learning_rate": 2.9898283301849137e-05,
|
| 164 |
-
"loss": 0.7693,
|
| 165 |
-
"step": 250
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 0.18,
|
| 169 |
-
"learning_rate": 2.9877601916014893e-05,
|
| 170 |
-
"loss": 0.7415,
|
| 171 |
-
"step": 260
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.18,
|
| 175 |
-
"learning_rate": 2.9855016489214902e-05,
|
| 176 |
-
"loss": 0.7504,
|
| 177 |
-
"step": 270
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.19,
|
| 181 |
-
"learning_rate": 2.983052991194041e-05,
|
| 182 |
-
"loss": 0.7186,
|
| 183 |
-
"step": 280
|
| 184 |
-
},
|
| 185 |
-
{
|
| 186 |
-
"epoch": 0.2,
|
| 187 |
-
"learning_rate": 2.980414531799256e-05,
|
| 188 |
-
"loss": 0.7272,
|
| 189 |
-
"step": 290
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.21,
|
| 193 |
-
"learning_rate": 2.977586608408141e-05,
|
| 194 |
-
"loss": 0.7452,
|
| 195 |
-
"step": 300
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.21,
|
| 199 |
-
"learning_rate": 2.974569582939371e-05,
|
| 200 |
-
"loss": 0.7392,
|
| 201 |
-
"step": 310
|
| 202 |
-
},
|
| 203 |
-
{
|
| 204 |
-
"epoch": 0.22,
|
| 205 |
-
"learning_rate": 2.9713638415129753e-05,
|
| 206 |
-
"loss": 0.7574,
|
| 207 |
-
"step": 320
|
| 208 |
-
},
|
| 209 |
-
{
|
| 210 |
-
"epoch": 0.23,
|
| 211 |
-
"learning_rate": 2.967969794400921e-05,
|
| 212 |
-
"loss": 0.7113,
|
| 213 |
-
"step": 330
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.23,
|
| 217 |
-
"learning_rate": 2.9643878759746055e-05,
|
| 218 |
-
"loss": 0.7494,
|
| 219 |
-
"step": 340
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.24,
|
| 223 |
-
"learning_rate": 2.9606185446492676e-05,
|
| 224 |
-
"loss": 0.7889,
|
| 225 |
-
"step": 350
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.25,
|
| 229 |
-
"learning_rate": 2.9566622828253155e-05,
|
| 230 |
-
"loss": 0.6575,
|
| 231 |
-
"step": 360
|
| 232 |
-
},
|
| 233 |
-
{
|
| 234 |
-
"epoch": 0.25,
|
| 235 |
-
"learning_rate": 2.9525195968265937e-05,
|
| 236 |
-
"loss": 0.7284,
|
| 237 |
-
"step": 370
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 0.26,
|
| 241 |
-
"learning_rate": 2.9481910168355798e-05,
|
| 242 |
-
"loss": 0.6623,
|
| 243 |
-
"step": 380
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"epoch": 0.27,
|
| 247 |
-
"learning_rate": 2.943677096825533e-05,
|
| 248 |
-
"loss": 0.7226,
|
| 249 |
-
"step": 390
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.27,
|
| 253 |
-
"learning_rate": 2.938978414489597e-05,
|
| 254 |
-
"loss": 0.678,
|
| 255 |
-
"step": 400
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.28,
|
| 259 |
-
"learning_rate": 2.9340955711668652e-05,
|
| 260 |
-
"loss": 0.6915,
|
| 261 |
-
"step": 410
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.29,
|
| 265 |
-
"learning_rate": 2.9290291917654223e-05,
|
| 266 |
-
"loss": 0.6835,
|
| 267 |
-
"step": 420
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 0.29,
|
| 271 |
-
"learning_rate": 2.9237799246823677e-05,
|
| 272 |
-
"loss": 0.6486,
|
| 273 |
-
"step": 430
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 0.3,
|
| 277 |
-
"learning_rate": 2.918348441720833e-05,
|
| 278 |
-
"loss": 0.6618,
|
| 279 |
-
"step": 440
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 0.31,
|
| 283 |
-
"learning_rate": 2.9127354380040067e-05,
|
| 284 |
-
"loss": 0.7055,
|
| 285 |
-
"step": 450
|
| 286 |
-
},
|
| 287 |
-
{
|
| 288 |
-
"epoch": 0.31,
|
| 289 |
-
"learning_rate": 2.906941631886168e-05,
|
| 290 |
-
"loss": 0.6836,
|
| 291 |
-
"step": 460
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.32,
|
| 295 |
-
"learning_rate": 2.9009677648607556e-05,
|
| 296 |
-
"loss": 0.6514,
|
| 297 |
-
"step": 470
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.33,
|
| 301 |
-
"learning_rate": 2.894814601465469e-05,
|
| 302 |
-
"loss": 0.716,
|
| 303 |
-
"step": 480
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.34,
|
| 307 |
-
"learning_rate": 2.8884829291844227e-05,
|
| 308 |
-
"loss": 0.6487,
|
| 309 |
-
"step": 490
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 0.34,
|
| 313 |
-
"learning_rate": 2.8819735583473635e-05,
|
| 314 |
-
"loss": 0.6891,
|
| 315 |
-
"step": 500
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 0.34,
|
| 319 |
-
"eval_loss": 0.665991485118866,
|
| 320 |
-
"eval_runtime": 112.8957,
|
| 321 |
-
"eval_samples_per_second": 5.758,
|
| 322 |
-
"eval_steps_per_second": 1.444,
|
| 323 |
-
"step": 500
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"epoch": 0.35,
|
| 327 |
-
"learning_rate": 2.8752873220259666e-05,
|
| 328 |
-
"loss": 0.7041,
|
| 329 |
-
"step": 510
|
| 330 |
-
},
|
| 331 |
-
{
|
| 332 |
-
"epoch": 0.36,
|
| 333 |
-
"learning_rate": 2.8684250759272157e-05,
|
| 334 |
-
"loss": 0.7158,
|
| 335 |
-
"step": 520
|
| 336 |
-
},
|
| 337 |
-
{
|
| 338 |
-
"epoch": 0.36,
|
| 339 |
-
"learning_rate": 2.8613876982838903e-05,
|
| 340 |
-
"loss": 0.6794,
|
| 341 |
-
"step": 530
|
| 342 |
-
},
|
| 343 |
-
{
|
| 344 |
-
"epoch": 0.37,
|
| 345 |
-
"learning_rate": 2.85417608974217e-05,
|
| 346 |
-
"loss": 0.6874,
|
| 347 |
-
"step": 540
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"epoch": 0.38,
|
| 351 |
-
"learning_rate": 2.846791173246368e-05,
|
| 352 |
-
"loss": 0.6694,
|
| 353 |
-
"step": 550
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.38,
|
| 357 |
-
"learning_rate": 2.839233893920815e-05,
|
| 358 |
-
"loss": 0.6643,
|
| 359 |
-
"step": 560
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.39,
|
| 363 |
-
"learning_rate": 2.8315052189488983e-05,
|
| 364 |
-
"loss": 0.6691,
|
| 365 |
-
"step": 570
|
| 366 |
-
},
|
| 367 |
-
{
|
| 368 |
-
"epoch": 0.4,
|
| 369 |
-
"learning_rate": 2.8236061374492838e-05,
|
| 370 |
-
"loss": 0.6586,
|
| 371 |
-
"step": 580
|
| 372 |
-
},
|
| 373 |
-
{
|
| 374 |
-
"epoch": 0.4,
|
| 375 |
-
"learning_rate": 2.8155376603493274e-05,
|
| 376 |
-
"loss": 0.646,
|
| 377 |
-
"step": 590
|
| 378 |
-
},
|
| 379 |
-
{
|
| 380 |
-
"epoch": 0.41,
|
| 381 |
-
"learning_rate": 2.8073008202556964e-05,
|
| 382 |
-
"loss": 0.6588,
|
| 383 |
-
"step": 600
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 0.42,
|
| 387 |
-
"learning_rate": 2.7988966713222155e-05,
|
| 388 |
-
"loss": 0.6293,
|
| 389 |
-
"step": 610
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 0.42,
|
| 393 |
-
"learning_rate": 2.7903262891149568e-05,
|
| 394 |
-
"loss": 0.6398,
|
| 395 |
-
"step": 620
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.43,
|
| 399 |
-
"learning_rate": 2.7815907704745874e-05,
|
| 400 |
-
"loss": 0.6031,
|
| 401 |
-
"step": 630
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.44,
|
| 405 |
-
"learning_rate": 2.7726912333759963e-05,
|
| 406 |
-
"loss": 0.6452,
|
| 407 |
-
"step": 640
|
| 408 |
-
},
|
| 409 |
-
{
|
| 410 |
-
"epoch": 0.44,
|
| 411 |
-
"learning_rate": 2.763628816785215e-05,
|
| 412 |
-
"loss": 0.6369,
|
| 413 |
-
"step": 650
|
| 414 |
-
},
|
| 415 |
-
{
|
| 416 |
-
"epoch": 0.45,
|
| 417 |
-
"learning_rate": 2.7544046805136524e-05,
|
| 418 |
-
"loss": 0.7018,
|
| 419 |
-
"step": 660
|
| 420 |
-
},
|
| 421 |
-
{
|
| 422 |
-
"epoch": 0.46,
|
| 423 |
-
"learning_rate": 2.7450200050696616e-05,
|
| 424 |
-
"loss": 0.6782,
|
| 425 |
-
"step": 670
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 0.47,
|
| 429 |
-
"learning_rate": 2.7354759915074574e-05,
|
| 430 |
-
"loss": 0.6427,
|
| 431 |
-
"step": 680
|
| 432 |
-
},
|
| 433 |
-
{
|
| 434 |
-
"epoch": 0.47,
|
| 435 |
-
"learning_rate": 2.7257738612734065e-05,
|
| 436 |
-
"loss": 0.6457,
|
| 437 |
-
"step": 690
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.48,
|
| 441 |
-
"learning_rate": 2.715914856049705e-05,
|
| 442 |
-
"loss": 0.6016,
|
| 443 |
-
"step": 700
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.49,
|
| 447 |
-
"learning_rate": 2.7059002375954684e-05,
|
| 448 |
-
"loss": 0.6606,
|
| 449 |
-
"step": 710
|
| 450 |
-
},
|
| 451 |
-
{
|
| 452 |
-
"epoch": 0.49,
|
| 453 |
-
"learning_rate": 2.695731287585249e-05,
|
| 454 |
-
"loss": 0.6334,
|
| 455 |
-
"step": 720
|
| 456 |
-
},
|
| 457 |
-
{
|
| 458 |
-
"epoch": 0.5,
|
| 459 |
-
"learning_rate": 2.685409307445011e-05,
|
| 460 |
-
"loss": 0.6806,
|
| 461 |
-
"step": 730
|
| 462 |
-
},
|
| 463 |
-
{
|
| 464 |
-
"epoch": 0.51,
|
| 465 |
-
"learning_rate": 2.6749356181855685e-05,
|
| 466 |
-
"loss": 0.5794,
|
| 467 |
-
"step": 740
|
| 468 |
-
},
|
| 469 |
-
{
|
| 470 |
-
"epoch": 0.51,
|
| 471 |
-
"learning_rate": 2.6643115602335263e-05,
|
| 472 |
-
"loss": 0.5806,
|
| 473 |
-
"step": 750
|
| 474 |
-
},
|
| 475 |
-
{
|
| 476 |
-
"epoch": 0.52,
|
| 477 |
-
"learning_rate": 2.6535384932597288e-05,
|
| 478 |
-
"loss": 0.6126,
|
| 479 |
-
"step": 760
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.53,
|
| 483 |
-
"learning_rate": 2.6426177960052494e-05,
|
| 484 |
-
"loss": 0.6405,
|
| 485 |
-
"step": 770
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.53,
|
| 489 |
-
"learning_rate": 2.631550866104941e-05,
|
| 490 |
-
"loss": 0.644,
|
| 491 |
-
"step": 780
|
| 492 |
-
},
|
| 493 |
-
{
|
| 494 |
-
"epoch": 0.54,
|
| 495 |
-
"learning_rate": 2.6203391199085617e-05,
|
| 496 |
-
"loss": 0.6072,
|
| 497 |
-
"step": 790
|
| 498 |
-
},
|
| 499 |
-
{
|
| 500 |
-
"epoch": 0.55,
|
| 501 |
-
"learning_rate": 2.6089839922995145e-05,
|
| 502 |
-
"loss": 0.6152,
|
| 503 |
-
"step": 800
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 0.55,
|
| 507 |
-
"learning_rate": 2.5974869365112084e-05,
|
| 508 |
-
"loss": 0.6294,
|
| 509 |
-
"step": 810
|
| 510 |
-
},
|
| 511 |
-
{
|
| 512 |
-
"epoch": 0.56,
|
| 513 |
-
"learning_rate": 2.5858494239410705e-05,
|
| 514 |
-
"loss": 0.6486,
|
| 515 |
-
"step": 820
|
| 516 |
-
},
|
| 517 |
-
{
|
| 518 |
-
"epoch": 0.57,
|
| 519 |
-
"learning_rate": 2.5740729439622414e-05,
|
| 520 |
-
"loss": 0.6507,
|
| 521 |
-
"step": 830
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.57,
|
| 525 |
-
"learning_rate": 2.5621590037329604e-05,
|
| 526 |
-
"loss": 0.6327,
|
| 527 |
-
"step": 840
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.58,
|
| 531 |
-
"learning_rate": 2.550109128003681e-05,
|
| 532 |
-
"loss": 0.5947,
|
| 533 |
-
"step": 850
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 0.59,
|
| 537 |
-
"learning_rate": 2.5379248589219307e-05,
|
| 538 |
-
"loss": 0.6366,
|
| 539 |
-
"step": 860
|
| 540 |
-
},
|
| 541 |
-
{
|
| 542 |
-
"epoch": 0.6,
|
| 543 |
-
"learning_rate": 2.525607755834948e-05,
|
| 544 |
-
"loss": 0.6608,
|
| 545 |
-
"step": 870
|
| 546 |
-
},
|
| 547 |
-
{
|
| 548 |
-
"epoch": 0.6,
|
| 549 |
-
"learning_rate": 2.513159395090117e-05,
|
| 550 |
-
"loss": 0.6861,
|
| 551 |
-
"step": 880
|
| 552 |
-
},
|
| 553 |
-
{
|
| 554 |
-
"epoch": 0.61,
|
| 555 |
-
"learning_rate": 2.5005813698332257e-05,
|
| 556 |
-
"loss": 0.6377,
|
| 557 |
-
"step": 890
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 0.62,
|
| 561 |
-
"learning_rate": 2.4878752898045725e-05,
|
| 562 |
-
"loss": 0.601,
|
| 563 |
-
"step": 900
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.62,
|
| 567 |
-
"learning_rate": 2.475042781132953e-05,
|
| 568 |
-
"loss": 0.6185,
|
| 569 |
-
"step": 910
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.63,
|
| 573 |
-
"learning_rate": 2.462085486127547e-05,
|
| 574 |
-
"loss": 0.6199,
|
| 575 |
-
"step": 920
|
| 576 |
-
},
|
| 577 |
-
{
|
| 578 |
-
"epoch": 0.64,
|
| 579 |
-
"learning_rate": 2.449005063067736e-05,
|
| 580 |
-
"loss": 0.6137,
|
| 581 |
-
"step": 930
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 0.64,
|
| 585 |
-
"learning_rate": 2.4358031859908747e-05,
|
| 586 |
-
"loss": 0.6004,
|
| 587 |
-
"step": 940
|
| 588 |
-
},
|
| 589 |
-
{
|
| 590 |
-
"epoch": 0.65,
|
| 591 |
-
"learning_rate": 2.4224815444780473e-05,
|
| 592 |
-
"loss": 0.6245,
|
| 593 |
-
"step": 950
|
| 594 |
-
},
|
| 595 |
-
{
|
| 596 |
-
"epoch": 0.66,
|
| 597 |
-
"learning_rate": 2.4090418434378348e-05,
|
| 598 |
-
"loss": 0.6535,
|
| 599 |
-
"step": 960
|
| 600 |
-
},
|
| 601 |
-
{
|
| 602 |
-
"epoch": 0.66,
|
| 603 |
-
"learning_rate": 2.3954858028881206e-05,
|
| 604 |
-
"loss": 0.594,
|
| 605 |
-
"step": 970
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.67,
|
| 609 |
-
"learning_rate": 2.3818151577359593e-05,
|
| 610 |
-
"loss": 0.605,
|
| 611 |
-
"step": 980
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.68,
|
| 615 |
-
"learning_rate": 2.3680316575555474e-05,
|
| 616 |
-
"loss": 0.6185,
|
| 617 |
-
"step": 990
|
| 618 |
-
},
|
| 619 |
-
{
|
| 620 |
-
"epoch": 0.68,
|
| 621 |
-
"learning_rate": 2.3541370663643073e-05,
|
| 622 |
-
"loss": 0.6423,
|
| 623 |
-
"step": 1000
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 0.68,
|
| 627 |
-
"eval_loss": 0.6074596047401428,
|
| 628 |
-
"eval_runtime": 112.8513,
|
| 629 |
-
"eval_samples_per_second": 5.76,
|
| 630 |
-
"eval_steps_per_second": 1.444,
|
| 631 |
-
"step": 1000
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 0.69,
|
| 635 |
-
"learning_rate": 2.3401331623971307e-05,
|
| 636 |
-
"loss": 0.6404,
|
| 637 |
-
"step": 1010
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 0.7,
|
| 641 |
-
"learning_rate": 2.3260217378787985e-05,
|
| 642 |
-
"loss": 0.5879,
|
| 643 |
-
"step": 1020
|
| 644 |
-
},
|
| 645 |
-
{
|
| 646 |
-
"epoch": 0.7,
|
| 647 |
-
"learning_rate": 2.3118045987946115e-05,
|
| 648 |
-
"loss": 0.6582,
|
| 649 |
-
"step": 1030
|
| 650 |
-
},
|
| 651 |
-
{
|
| 652 |
-
"epoch": 0.71,
|
| 653 |
-
"learning_rate": 2.297483564659259e-05,
|
| 654 |
-
"loss": 0.6132,
|
| 655 |
-
"step": 1040
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 0.72,
|
| 659 |
-
"learning_rate": 2.2830604682839577e-05,
|
| 660 |
-
"loss": 0.5921,
|
| 661 |
-
"step": 1050
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.73,
|
| 665 |
-
"learning_rate": 2.2685371555418865e-05,
|
| 666 |
-
"loss": 0.5874,
|
| 667 |
-
"step": 1060
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 0.73,
|
| 671 |
-
"learning_rate": 2.2539154851319534e-05,
|
| 672 |
-
"loss": 0.6708,
|
| 673 |
-
"step": 1070
|
| 674 |
-
},
|
| 675 |
-
{
|
| 676 |
-
"epoch": 0.74,
|
| 677 |
-
"learning_rate": 2.2391973283409148e-05,
|
| 678 |
-
"loss": 0.6474,
|
| 679 |
-
"step": 1080
|
| 680 |
-
},
|
| 681 |
-
{
|
| 682 |
-
"epoch": 0.75,
|
| 683 |
-
"learning_rate": 2.2243845688038912e-05,
|
| 684 |
-
"loss": 0.5853,
|
| 685 |
-
"step": 1090
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 0.75,
|
| 689 |
-
"learning_rate": 2.2094791022632964e-05,
|
| 690 |
-
"loss": 0.602,
|
| 691 |
-
"step": 1100
|
| 692 |
-
},
|
| 693 |
-
{
|
| 694 |
-
"epoch": 0.76,
|
| 695 |
-
"learning_rate": 2.194482836326222e-05,
|
| 696 |
-
"loss": 0.6012,
|
| 697 |
-
"step": 1110
|
| 698 |
-
},
|
| 699 |
-
{
|
| 700 |
-
"epoch": 0.77,
|
| 701 |
-
"learning_rate": 2.179397690220298e-05,
|
| 702 |
-
"loss": 0.5837,
|
| 703 |
-
"step": 1120
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 0.77,
|
| 707 |
-
"learning_rate": 2.164225594548072e-05,
|
| 708 |
-
"loss": 0.6277,
|
| 709 |
-
"step": 1130
|
| 710 |
-
},
|
| 711 |
-
{
|
| 712 |
-
"epoch": 0.78,
|
| 713 |
-
"learning_rate": 2.1489684910399287e-05,
|
| 714 |
-
"loss": 0.548,
|
| 715 |
-
"step": 1140
|
| 716 |
-
},
|
| 717 |
-
{
|
| 718 |
-
"epoch": 0.79,
|
| 719 |
-
"learning_rate": 2.1336283323055877e-05,
|
| 720 |
-
"loss": 0.6275,
|
| 721 |
-
"step": 1150
|
| 722 |
-
},
|
| 723 |
-
{
|
| 724 |
-
"epoch": 0.79,
|
| 725 |
-
"learning_rate": 2.118207081584207e-05,
|
| 726 |
-
"loss": 0.6045,
|
| 727 |
-
"step": 1160
|
| 728 |
-
},
|
| 729 |
-
{
|
| 730 |
-
"epoch": 0.8,
|
| 731 |
-
"learning_rate": 2.102706712493125e-05,
|
| 732 |
-
"loss": 0.579,
|
| 733 |
-
"step": 1170
|
| 734 |
-
},
|
| 735 |
-
{
|
| 736 |
-
"epoch": 0.81,
|
| 737 |
-
"learning_rate": 2.0871292087752813e-05,
|
| 738 |
-
"loss": 0.6307,
|
| 739 |
-
"step": 1180
|
| 740 |
-
},
|
| 741 |
-
{
|
| 742 |
-
"epoch": 0.81,
|
| 743 |
-
"learning_rate": 2.0714765640453295e-05,
|
| 744 |
-
"loss": 0.6708,
|
| 745 |
-
"step": 1190
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 0.82,
|
| 749 |
-
"learning_rate": 2.055750781534502e-05,
|
| 750 |
-
"loss": 0.5567,
|
| 751 |
-
"step": 1200
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.83,
|
| 755 |
-
"learning_rate": 2.0399538738342276e-05,
|
| 756 |
-
"loss": 0.6077,
|
| 757 |
-
"step": 1210
|
| 758 |
-
},
|
| 759 |
-
{
|
| 760 |
-
"epoch": 0.83,
|
| 761 |
-
"learning_rate": 2.0240878626385656e-05,
|
| 762 |
-
"loss": 0.5914,
|
| 763 |
-
"step": 1220
|
| 764 |
-
},
|
| 765 |
-
{
|
| 766 |
-
"epoch": 0.84,
|
| 767 |
-
"learning_rate": 2.0081547784854636e-05,
|
| 768 |
-
"loss": 0.5752,
|
| 769 |
-
"step": 1230
|
| 770 |
-
},
|
| 771 |
-
{
|
| 772 |
-
"epoch": 0.85,
|
| 773 |
-
"learning_rate": 1.9921566604968935e-05,
|
| 774 |
-
"loss": 0.5845,
|
| 775 |
-
"step": 1240
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 0.85,
|
| 779 |
-
"learning_rate": 1.9760955561178792e-05,
|
| 780 |
-
"loss": 0.5957,
|
| 781 |
-
"step": 1250
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 0.86,
|
| 785 |
-
"learning_rate": 1.9599735208544676e-05,
|
| 786 |
-
"loss": 0.6019,
|
| 787 |
-
"step": 1260
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 0.87,
|
| 791 |
-
"learning_rate": 1.943792618010662e-05,
|
| 792 |
-
"loss": 0.6482,
|
| 793 |
-
"step": 1270
|
| 794 |
-
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 0.88,
|
| 797 |
-
"learning_rate": 1.9275549184243597e-05,
|
| 798 |
-
"loss": 0.5898,
|
| 799 |
-
"step": 1280
|
| 800 |
-
},
|
| 801 |
-
{
|
| 802 |
-
"epoch": 0.88,
|
| 803 |
-
"learning_rate": 1.9112625002023276e-05,
|
| 804 |
-
"loss": 0.5559,
|
| 805 |
-
"step": 1290
|
| 806 |
},
|
| 807 |
{
|
| 808 |
-
"epoch": 0.
|
| 809 |
-
"
|
| 810 |
-
"
|
| 811 |
-
"
|
| 812 |
-
|
| 813 |
-
|
| 814 |
-
"
|
| 815 |
-
"learning_rate": 1.8785218550258373e-05,
|
| 816 |
-
"loss": 0.6049,
|
| 817 |
-
"step": 1310
|
| 818 |
-
},
|
| 819 |
-
{
|
| 820 |
-
"epoch": 0.9,
|
| 821 |
-
"learning_rate": 1.862077818231193e-05,
|
| 822 |
-
"loss": 0.5744,
|
| 823 |
-
"step": 1320
|
| 824 |
-
},
|
| 825 |
-
{
|
| 826 |
-
"epoch": 0.91,
|
| 827 |
-
"learning_rate": 1.845587442584185e-05,
|
| 828 |
-
"loss": 0.6184,
|
| 829 |
-
"step": 1330
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 0.92,
|
| 833 |
-
"learning_rate": 1.8290528385291558e-05,
|
| 834 |
-
"loss": 0.5656,
|
| 835 |
-
"step": 1340
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 0.92,
|
| 839 |
-
"learning_rate": 1.8124761221708136e-05,
|
| 840 |
-
"loss": 0.5297,
|
| 841 |
-
"step": 1350
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 0.93,
|
| 845 |
-
"learning_rate": 1.7958594150034165e-05,
|
| 846 |
-
"loss": 0.5681,
|
| 847 |
-
"step": 1360
|
| 848 |
-
},
|
| 849 |
-
{
|
| 850 |
-
"epoch": 0.94,
|
| 851 |
-
"learning_rate": 1.7792048436392584e-05,
|
| 852 |
-
"loss": 0.6133,
|
| 853 |
-
"step": 1370
|
| 854 |
-
},
|
| 855 |
-
{
|
| 856 |
-
"epoch": 0.94,
|
| 857 |
-
"learning_rate": 1.7625145395365098e-05,
|
| 858 |
-
"loss": 0.5877,
|
| 859 |
-
"step": 1380
|
| 860 |
-
},
|
| 861 |
-
{
|
| 862 |
-
"epoch": 0.95,
|
| 863 |
-
"learning_rate": 1.7457906387264288e-05,
|
| 864 |
-
"loss": 0.5887,
|
| 865 |
-
"step": 1390
|
| 866 |
-
},
|
| 867 |
-
{
|
| 868 |
-
"epoch": 0.96,
|
| 869 |
-
"learning_rate": 1.729035281539993e-05,
|
| 870 |
-
"loss": 0.5648,
|
| 871 |
-
"step": 1400
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.96,
|
| 875 |
-
"learning_rate": 1.7122506123339775e-05,
|
| 876 |
-
"loss": 0.551,
|
| 877 |
-
"step": 1410
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 0.97,
|
| 881 |
-
"learning_rate": 1.6954387792165222e-05,
|
| 882 |
-
"loss": 0.5537,
|
| 883 |
-
"step": 1420
|
| 884 |
-
},
|
| 885 |
-
{
|
| 886 |
-
"epoch": 0.98,
|
| 887 |
-
"learning_rate": 1.6786019337722136e-05,
|
| 888 |
-
"loss": 0.542,
|
| 889 |
-
"step": 1430
|
| 890 |
-
},
|
| 891 |
-
{
|
| 892 |
-
"epoch": 0.98,
|
| 893 |
-
"learning_rate": 1.6617422307867238e-05,
|
| 894 |
-
"loss": 0.5466,
|
| 895 |
-
"step": 1440
|
| 896 |
-
},
|
| 897 |
-
{
|
| 898 |
-
"epoch": 0.99,
|
| 899 |
-
"learning_rate": 1.6448618279710438e-05,
|
| 900 |
-
"loss": 0.5977,
|
| 901 |
-
"step": 1450
|
| 902 |
-
},
|
| 903 |
-
{
|
| 904 |
-
"epoch": 1.0,
|
| 905 |
-
"learning_rate": 1.6279628856853338e-05,
|
| 906 |
-
"loss": 0.5897,
|
| 907 |
-
"step": 1460
|
| 908 |
-
},
|
| 909 |
-
{
|
| 910 |
-
"epoch": 1.01,
|
| 911 |
-
"learning_rate": 1.611047566662445e-05,
|
| 912 |
-
"loss": 0.5539,
|
| 913 |
-
"step": 1470
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 1.01,
|
| 917 |
-
"learning_rate": 1.5941180357311268e-05,
|
| 918 |
-
"loss": 0.5411,
|
| 919 |
-
"step": 1480
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 1.02,
|
| 923 |
-
"learning_rate": 1.5771764595389764e-05,
|
| 924 |
-
"loss": 0.5479,
|
| 925 |
-
"step": 1490
|
| 926 |
-
},
|
| 927 |
-
{
|
| 928 |
-
"epoch": 1.03,
|
| 929 |
-
"learning_rate": 1.560225006275145e-05,
|
| 930 |
-
"loss": 0.5553,
|
| 931 |
-
"step": 1500
|
| 932 |
-
},
|
| 933 |
-
{
|
| 934 |
-
"epoch": 1.03,
|
| 935 |
-
"eval_loss": 0.5688494443893433,
|
| 936 |
-
"eval_runtime": 112.9699,
|
| 937 |
-
"eval_samples_per_second": 5.754,
|
| 938 |
-
"eval_steps_per_second": 1.443,
|
| 939 |
-
"step": 1500
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 1.03,
|
| 943 |
-
"learning_rate": 1.543265845392858e-05,
|
| 944 |
-
"loss": 0.535,
|
| 945 |
-
"step": 1510
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 1.04,
|
| 949 |
-
"learning_rate": 1.5263011473317623e-05,
|
| 950 |
-
"loss": 0.531,
|
| 951 |
-
"step": 1520
|
| 952 |
-
},
|
| 953 |
-
{
|
| 954 |
-
"epoch": 1.05,
|
| 955 |
-
"learning_rate": 1.509333083240157e-05,
|
| 956 |
-
"loss": 0.568,
|
| 957 |
-
"step": 1530
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 1.05,
|
| 961 |
-
"learning_rate": 1.4923638246971261e-05,
|
| 962 |
-
"loss": 0.5375,
|
| 963 |
-
"step": 1540
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 1.06,
|
| 967 |
-
"learning_rate": 1.4753955434346188e-05,
|
| 968 |
-
"loss": 0.5868,
|
| 969 |
-
"step": 1550
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 1.07,
|
| 973 |
-
"learning_rate": 1.458430411059513e-05,
|
| 974 |
-
"loss": 0.5549,
|
| 975 |
-
"step": 1560
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 1.07,
|
| 979 |
-
"learning_rate": 1.4414705987756907e-05,
|
| 980 |
-
"loss": 0.5595,
|
| 981 |
-
"step": 1570
|
| 982 |
-
},
|
| 983 |
-
{
|
| 984 |
-
"epoch": 1.08,
|
| 985 |
-
"learning_rate": 1.4245182771061667e-05,
|
| 986 |
-
"loss": 0.5578,
|
| 987 |
-
"step": 1580
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 1.09,
|
| 991 |
-
"learning_rate": 1.407575615615303e-05,
|
| 992 |
-
"loss": 0.538,
|
| 993 |
-
"step": 1590
|
| 994 |
-
},
|
| 995 |
-
{
|
| 996 |
-
"epoch": 1.09,
|
| 997 |
-
"learning_rate": 1.3906447826311502e-05,
|
| 998 |
-
"loss": 0.5574,
|
| 999 |
-
"step": 1600
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 1.1,
|
| 1003 |
-
"learning_rate": 1.3737279449679414e-05,
|
| 1004 |
-
"loss": 0.5667,
|
| 1005 |
-
"step": 1610
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 1.11,
|
| 1009 |
-
"learning_rate": 1.3568272676487834e-05,
|
| 1010 |
-
"loss": 0.5564,
|
| 1011 |
-
"step": 1620
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 1.11,
|
| 1015 |
-
"learning_rate": 1.3399449136285751e-05,
|
| 1016 |
-
"loss": 0.5589,
|
| 1017 |
-
"step": 1630
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 1.12,
|
| 1021 |
-
"learning_rate": 1.3230830435171951e-05,
|
| 1022 |
-
"loss": 0.5636,
|
| 1023 |
-
"step": 1640
|
| 1024 |
-
},
|
| 1025 |
-
{
|
| 1026 |
-
"epoch": 1.13,
|
| 1027 |
-
"learning_rate": 1.3062438153029829e-05,
|
| 1028 |
-
"loss": 0.5677,
|
| 1029 |
-
"step": 1650
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 1.14,
|
| 1033 |
-
"learning_rate": 1.2894293840765583e-05,
|
| 1034 |
-
"loss": 0.5321,
|
| 1035 |
-
"step": 1660
|
| 1036 |
-
},
|
| 1037 |
-
{
|
| 1038 |
-
"epoch": 1.14,
|
| 1039 |
-
"learning_rate": 1.272641901755015e-05,
|
| 1040 |
-
"loss": 0.5664,
|
| 1041 |
-
"step": 1670
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 1.15,
|
| 1045 |
-
"learning_rate": 1.2558835168065162e-05,
|
| 1046 |
-
"loss": 0.6372,
|
| 1047 |
-
"step": 1680
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 1.16,
|
| 1051 |
-
"learning_rate": 1.2391563739753316e-05,
|
| 1052 |
-
"loss": 0.5377,
|
| 1053 |
-
"step": 1690
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 1.16,
|
| 1057 |
-
"learning_rate": 1.2224626140073533e-05,
|
| 1058 |
-
"loss": 0.5974,
|
| 1059 |
-
"step": 1700
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 1.17,
|
| 1063 |
-
"learning_rate": 1.2058043733761228e-05,
|
| 1064 |
-
"loss": 0.5924,
|
| 1065 |
-
"step": 1710
|
| 1066 |
-
},
|
| 1067 |
-
{
|
| 1068 |
-
"epoch": 1.18,
|
| 1069 |
-
"learning_rate": 1.1891837840094038e-05,
|
| 1070 |
-
"loss": 0.5149,
|
| 1071 |
-
"step": 1720
|
| 1072 |
-
},
|
| 1073 |
-
{
|
| 1074 |
-
"epoch": 1.18,
|
| 1075 |
-
"learning_rate": 1.1726029730163372e-05,
|
| 1076 |
-
"loss": 0.5486,
|
| 1077 |
-
"step": 1730
|
| 1078 |
-
},
|
| 1079 |
-
{
|
| 1080 |
-
"epoch": 1.19,
|
| 1081 |
-
"learning_rate": 1.156064062415212e-05,
|
| 1082 |
-
"loss": 0.5161,
|
| 1083 |
-
"step": 1740
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 1.2,
|
| 1087 |
-
"learning_rate": 1.1395691688618916e-05,
|
| 1088 |
-
"loss": 0.5317,
|
| 1089 |
-
"step": 1750
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 1.2,
|
| 1093 |
-
"learning_rate": 1.1231204033789202e-05,
|
| 1094 |
-
"loss": 0.5456,
|
| 1095 |
-
"step": 1760
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 1.21,
|
| 1099 |
-
"learning_rate": 1.1067198710853537e-05,
|
| 1100 |
-
"loss": 0.4997,
|
| 1101 |
-
"step": 1770
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 1.22,
|
| 1105 |
-
"learning_rate": 1.0903696709273497e-05,
|
| 1106 |
-
"loss": 0.5274,
|
| 1107 |
-
"step": 1780
|
| 1108 |
-
},
|
| 1109 |
-
{
|
| 1110 |
-
"epoch": 1.22,
|
| 1111 |
-
"learning_rate": 1.074071895409537e-05,
|
| 1112 |
-
"loss": 0.5395,
|
| 1113 |
-
"step": 1790
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 1.23,
|
| 1117 |
-
"learning_rate": 1.0578286303272224e-05,
|
| 1118 |
-
"loss": 0.5754,
|
| 1119 |
-
"step": 1800
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 1.24,
|
| 1123 |
-
"learning_rate": 1.0416419544994449e-05,
|
| 1124 |
-
"loss": 0.5593,
|
| 1125 |
-
"step": 1810
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 1.24,
|
| 1129 |
-
"learning_rate": 1.0255139395029315e-05,
|
| 1130 |
-
"loss": 0.5635,
|
| 1131 |
-
"step": 1820
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 1.25,
|
| 1135 |
-
"learning_rate": 1.009446649406974e-05,
|
| 1136 |
-
"loss": 0.5796,
|
| 1137 |
-
"step": 1830
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 1.26,
|
| 1141 |
-
"learning_rate": 9.934421405092689e-06,
|
| 1142 |
-
"loss": 0.5489,
|
| 1143 |
-
"step": 1840
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 1.27,
|
| 1147 |
-
"learning_rate": 9.775024610727527e-06,
|
| 1148 |
-
"loss": 0.5869,
|
| 1149 |
-
"step": 1850
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 1.27,
|
| 1153 |
-
"learning_rate": 9.616296510634619e-06,
|
| 1154 |
-
"loss": 0.5463,
|
| 1155 |
-
"step": 1860
|
| 1156 |
-
},
|
| 1157 |
-
{
|
| 1158 |
-
"epoch": 1.28,
|
| 1159 |
-
"learning_rate": 9.458257418894607e-06,
|
| 1160 |
-
"loss": 0.5725,
|
| 1161 |
-
"step": 1870
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 1.29,
|
| 1165 |
-
"learning_rate": 9.300927561408574e-06,
|
| 1166 |
-
"loss": 0.5694,
|
| 1167 |
-
"step": 1880
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 1.29,
|
| 1171 |
-
"learning_rate": 9.14432707330957e-06,
|
| 1172 |
-
"loss": 0.5721,
|
| 1173 |
-
"step": 1890
|
| 1174 |
-
},
|
| 1175 |
-
{
|
| 1176 |
-
"epoch": 1.3,
|
| 1177 |
-
"learning_rate": 8.988475996385653e-06,
|
| 1178 |
-
"loss": 0.5423,
|
| 1179 |
-
"step": 1900
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 1.31,
|
| 1183 |
-
"learning_rate": 8.833394276514976e-06,
|
| 1184 |
-
"loss": 0.5376,
|
| 1185 |
-
"step": 1910
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 1.31,
|
| 1189 |
-
"learning_rate": 8.679101761113099e-06,
|
| 1190 |
-
"loss": 0.5102,
|
| 1191 |
-
"step": 1920
|
| 1192 |
-
},
|
| 1193 |
-
{
|
| 1194 |
-
"epoch": 1.32,
|
| 1195 |
-
"learning_rate": 8.525618196592897e-06,
|
| 1196 |
-
"loss": 0.5698,
|
| 1197 |
-
"step": 1930
|
| 1198 |
-
},
|
| 1199 |
-
{
|
| 1200 |
-
"epoch": 1.33,
|
| 1201 |
-
"learning_rate": 8.372963225837405e-06,
|
| 1202 |
-
"loss": 0.4899,
|
| 1203 |
-
"step": 1940
|
| 1204 |
-
},
|
| 1205 |
-
{
|
| 1206 |
-
"epoch": 1.33,
|
| 1207 |
-
"learning_rate": 8.221156385685939e-06,
|
| 1208 |
-
"loss": 0.5326,
|
| 1209 |
-
"step": 1950
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 1.34,
|
| 1213 |
-
"learning_rate": 8.070217104433746e-06,
|
| 1214 |
-
"loss": 0.5562,
|
| 1215 |
-
"step": 1960
|
| 1216 |
-
},
|
| 1217 |
-
{
|
| 1218 |
-
"epoch": 1.35,
|
| 1219 |
-
"learning_rate": 7.920164699345559e-06,
|
| 1220 |
-
"loss": 0.542,
|
| 1221 |
-
"step": 1970
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 1.35,
|
| 1225 |
-
"learning_rate": 7.771018374183367e-06,
|
| 1226 |
-
"loss": 0.5584,
|
| 1227 |
-
"step": 1980
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 1.36,
|
| 1231 |
-
"learning_rate": 7.622797216748732e-06,
|
| 1232 |
-
"loss": 0.559,
|
| 1233 |
-
"step": 1990
|
| 1234 |
-
},
|
| 1235 |
-
{
|
| 1236 |
-
"epoch": 1.37,
|
| 1237 |
-
"learning_rate": 7.475520196439913e-06,
|
| 1238 |
-
"loss": 0.5175,
|
| 1239 |
-
"step": 2000
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 1.37,
|
| 1243 |
-
"eval_loss": 0.5532566905021667,
|
| 1244 |
-
"eval_runtime": 112.9998,
|
| 1245 |
-
"eval_samples_per_second": 5.752,
|
| 1246 |
-
"eval_steps_per_second": 1.442,
|
| 1247 |
-
"step": 2000
|
| 1248 |
-
},
|
| 1249 |
-
{
|
| 1250 |
-
"epoch": 1.37,
|
| 1251 |
-
"learning_rate": 7.329206161824133e-06,
|
| 1252 |
-
"loss": 0.5066,
|
| 1253 |
-
"step": 2010
|
| 1254 |
-
},
|
| 1255 |
-
{
|
| 1256 |
-
"epoch": 1.38,
|
| 1257 |
-
"learning_rate": 7.183873838225359e-06,
|
| 1258 |
-
"loss": 0.5446,
|
| 1259 |
-
"step": 2020
|
| 1260 |
-
},
|
| 1261 |
-
{
|
| 1262 |
-
"epoch": 1.39,
|
| 1263 |
-
"learning_rate": 7.039541825327827e-06,
|
| 1264 |
-
"loss": 0.5875,
|
| 1265 |
-
"step": 2030
|
| 1266 |
-
},
|
| 1267 |
-
{
|
| 1268 |
-
"epoch": 1.4,
|
| 1269 |
-
"learning_rate": 6.896228594795646e-06,
|
| 1270 |
-
"loss": 0.5211,
|
| 1271 |
-
"step": 2040
|
| 1272 |
-
},
|
| 1273 |
-
{
|
| 1274 |
-
"epoch": 1.4,
|
| 1275 |
-
"learning_rate": 6.753952487908767e-06,
|
| 1276 |
-
"loss": 0.5785,
|
| 1277 |
-
"step": 2050
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 1.41,
|
| 1281 |
-
"learning_rate": 6.612731713215683e-06,
|
| 1282 |
-
"loss": 0.5338,
|
| 1283 |
-
"step": 2060
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 1.42,
|
| 1287 |
-
"learning_rate": 6.472584344203087e-06,
|
| 1288 |
-
"loss": 0.5539,
|
| 1289 |
-
"step": 2070
|
| 1290 |
-
},
|
| 1291 |
-
{
|
| 1292 |
-
"epoch": 1.42,
|
| 1293 |
-
"learning_rate": 6.3335283169828236e-06,
|
| 1294 |
-
"loss": 0.5763,
|
| 1295 |
-
"step": 2080
|
| 1296 |
-
},
|
| 1297 |
-
{
|
| 1298 |
-
"epoch": 1.43,
|
| 1299 |
-
"learning_rate": 6.195581427996395e-06,
|
| 1300 |
-
"loss": 0.5044,
|
| 1301 |
-
"step": 2090
|
| 1302 |
-
},
|
| 1303 |
-
{
|
| 1304 |
-
"epoch": 1.44,
|
| 1305 |
-
"learning_rate": 6.058761331737406e-06,
|
| 1306 |
-
"loss": 0.5509,
|
| 1307 |
-
"step": 2100
|
| 1308 |
-
},
|
| 1309 |
-
{
|
| 1310 |
-
"epoch": 1.44,
|
| 1311 |
-
"learning_rate": 5.923085538492093e-06,
|
| 1312 |
-
"loss": 0.5212,
|
| 1313 |
-
"step": 2110
|
| 1314 |
-
},
|
| 1315 |
-
{
|
| 1316 |
-
"epoch": 1.45,
|
| 1317 |
-
"learning_rate": 5.788571412098394e-06,
|
| 1318 |
-
"loss": 0.5772,
|
| 1319 |
-
"step": 2120
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 1.46,
|
| 1323 |
-
"learning_rate": 5.655236167723671e-06,
|
| 1324 |
-
"loss": 0.5368,
|
| 1325 |
-
"step": 2130
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 1.46,
|
| 1329 |
-
"learning_rate": 5.5230968696615565e-06,
|
| 1330 |
-
"loss": 0.5327,
|
| 1331 |
-
"step": 2140
|
| 1332 |
-
},
|
| 1333 |
-
{
|
| 1334 |
-
"epoch": 1.47,
|
| 1335 |
-
"learning_rate": 5.392170429148018e-06,
|
| 1336 |
-
"loss": 0.5411,
|
| 1337 |
-
"step": 2150
|
| 1338 |
-
},
|
| 1339 |
-
{
|
| 1340 |
-
"epoch": 1.48,
|
| 1341 |
-
"learning_rate": 5.26247360219706e-06,
|
| 1342 |
-
"loss": 0.5354,
|
| 1343 |
-
"step": 2160
|
| 1344 |
-
},
|
| 1345 |
-
{
|
| 1346 |
-
"epoch": 1.48,
|
| 1347 |
-
"learning_rate": 5.134022987456326e-06,
|
| 1348 |
-
"loss": 0.5276,
|
| 1349 |
-
"step": 2170
|
| 1350 |
-
},
|
| 1351 |
-
{
|
| 1352 |
-
"epoch": 1.49,
|
| 1353 |
-
"learning_rate": 5.006835024082726e-06,
|
| 1354 |
-
"loss": 0.5055,
|
| 1355 |
-
"step": 2180
|
| 1356 |
-
},
|
| 1357 |
-
{
|
| 1358 |
-
"epoch": 1.5,
|
| 1359 |
-
"learning_rate": 4.880925989638598e-06,
|
| 1360 |
-
"loss": 0.5838,
|
| 1361 |
-
"step": 2190
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 1.5,
|
| 1365 |
-
"learning_rate": 4.75631199800848e-06,
|
| 1366 |
-
"loss": 0.5915,
|
| 1367 |
-
"step": 2200
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 1.51,
|
| 1371 |
-
"learning_rate": 4.633008997336863e-06,
|
| 1372 |
-
"loss": 0.4826,
|
| 1373 |
-
"step": 2210
|
| 1374 |
-
},
|
| 1375 |
-
{
|
| 1376 |
-
"epoch": 1.52,
|
| 1377 |
-
"learning_rate": 4.5110327679871335e-06,
|
| 1378 |
-
"loss": 0.562,
|
| 1379 |
-
"step": 2220
|
| 1380 |
-
},
|
| 1381 |
-
{
|
| 1382 |
-
"epoch": 1.53,
|
| 1383 |
-
"learning_rate": 4.3903989205219805e-06,
|
| 1384 |
-
"loss": 0.5399,
|
| 1385 |
-
"step": 2230
|
| 1386 |
-
},
|
| 1387 |
-
{
|
| 1388 |
-
"epoch": 1.53,
|
| 1389 |
-
"learning_rate": 4.2711228937055865e-06,
|
| 1390 |
-
"loss": 0.5079,
|
| 1391 |
-
"step": 2240
|
| 1392 |
-
},
|
| 1393 |
-
{
|
| 1394 |
-
"epoch": 1.54,
|
| 1395 |
-
"learning_rate": 4.1532199525277285e-06,
|
| 1396 |
-
"loss": 0.541,
|
| 1397 |
-
"step": 2250
|
| 1398 |
-
},
|
| 1399 |
-
{
|
| 1400 |
-
"epoch": 1.55,
|
| 1401 |
-
"learning_rate": 4.0367051862501955e-06,
|
| 1402 |
-
"loss": 0.5536,
|
| 1403 |
-
"step": 2260
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 1.55,
|
| 1407 |
-
"learning_rate": 3.921593506475625e-06,
|
| 1408 |
-
"loss": 0.5456,
|
| 1409 |
-
"step": 2270
|
| 1410 |
-
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 1.56,
|
| 1413 |
-
"learning_rate": 3.807899645239149e-06,
|
| 1414 |
-
"loss": 0.5251,
|
| 1415 |
-
"step": 2280
|
| 1416 |
-
},
|
| 1417 |
-
{
|
| 1418 |
-
"epoch": 1.57,
|
| 1419 |
-
"learning_rate": 3.695638153122936e-06,
|
| 1420 |
-
"loss": 0.5232,
|
| 1421 |
-
"step": 2290
|
| 1422 |
-
},
|
| 1423 |
-
{
|
| 1424 |
-
"epoch": 1.57,
|
| 1425 |
-
"learning_rate": 3.5848233973940525e-06,
|
| 1426 |
-
"loss": 0.5597,
|
| 1427 |
-
"step": 2300
|
| 1428 |
-
},
|
| 1429 |
-
{
|
| 1430 |
-
"epoch": 1.58,
|
| 1431 |
-
"learning_rate": 3.4754695601656887e-06,
|
| 1432 |
-
"loss": 0.5475,
|
| 1433 |
-
"step": 2310
|
| 1434 |
-
},
|
| 1435 |
-
{
|
| 1436 |
-
"epoch": 1.59,
|
| 1437 |
-
"learning_rate": 3.367590636582165e-06,
|
| 1438 |
-
"loss": 0.557,
|
| 1439 |
-
"step": 2320
|
| 1440 |
-
},
|
| 1441 |
-
{
|
| 1442 |
-
"epoch": 1.59,
|
| 1443 |
-
"learning_rate": 3.2612004330277988e-06,
|
| 1444 |
-
"loss": 0.4989,
|
| 1445 |
-
"step": 2330
|
| 1446 |
-
},
|
| 1447 |
-
{
|
| 1448 |
-
"epoch": 1.6,
|
| 1449 |
-
"learning_rate": 3.156312565359975e-06,
|
| 1450 |
-
"loss": 0.5602,
|
| 1451 |
-
"step": 2340
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 1.61,
|
| 1455 |
-
"learning_rate": 3.0529404571665865e-06,
|
| 1456 |
-
"loss": 0.5619,
|
| 1457 |
-
"step": 2350
|
| 1458 |
-
},
|
| 1459 |
-
{
|
| 1460 |
-
"epoch": 1.61,
|
| 1461 |
-
"learning_rate": 2.9510973380480584e-06,
|
| 1462 |
-
"loss": 0.537,
|
| 1463 |
-
"step": 2360
|
| 1464 |
-
},
|
| 1465 |
-
{
|
| 1466 |
-
"epoch": 1.62,
|
| 1467 |
-
"learning_rate": 2.850796241924237e-06,
|
| 1468 |
-
"loss": 0.5252,
|
| 1469 |
-
"step": 2370
|
| 1470 |
-
},
|
| 1471 |
-
{
|
| 1472 |
-
"epoch": 1.63,
|
| 1473 |
-
"learning_rate": 2.7520500053662927e-06,
|
| 1474 |
-
"loss": 0.5549,
|
| 1475 |
-
"step": 2380
|
| 1476 |
-
},
|
| 1477 |
-
{
|
| 1478 |
-
"epoch": 1.63,
|
| 1479 |
-
"learning_rate": 2.654871265953916e-06,
|
| 1480 |
-
"loss": 0.57,
|
| 1481 |
-
"step": 2390
|
| 1482 |
-
},
|
| 1483 |
-
{
|
| 1484 |
-
"epoch": 1.64,
|
| 1485 |
-
"learning_rate": 2.559272460657915e-06,
|
| 1486 |
-
"loss": 0.5821,
|
| 1487 |
-
"step": 2400
|
| 1488 |
-
},
|
| 1489 |
-
{
|
| 1490 |
-
"epoch": 1.65,
|
| 1491 |
-
"learning_rate": 2.4652658242485547e-06,
|
| 1492 |
-
"loss": 0.5291,
|
| 1493 |
-
"step": 2410
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 1.66,
|
| 1497 |
-
"learning_rate": 2.372863387729749e-06,
|
| 1498 |
-
"loss": 0.5418,
|
| 1499 |
-
"step": 2420
|
| 1500 |
-
},
|
| 1501 |
-
{
|
| 1502 |
-
"epoch": 1.66,
|
| 1503 |
-
"learning_rate": 2.282076976799303e-06,
|
| 1504 |
-
"loss": 0.6195,
|
| 1505 |
-
"step": 2430
|
| 1506 |
-
},
|
| 1507 |
-
{
|
| 1508 |
-
"epoch": 1.67,
|
| 1509 |
-
"learning_rate": 2.1929182103354907e-06,
|
| 1510 |
-
"loss": 0.4792,
|
| 1511 |
-
"step": 2440
|
| 1512 |
-
},
|
| 1513 |
-
{
|
| 1514 |
-
"epoch": 1.68,
|
| 1515 |
-
"learning_rate": 2.105398498910033e-06,
|
| 1516 |
-
"loss": 0.5733,
|
| 1517 |
-
"step": 2450
|
| 1518 |
-
},
|
| 1519 |
-
{
|
| 1520 |
-
"epoch": 1.68,
|
| 1521 |
-
"learning_rate": 2.0195290433277987e-06,
|
| 1522 |
-
"loss": 0.5846,
|
| 1523 |
-
"step": 2460
|
| 1524 |
-
},
|
| 1525 |
-
{
|
| 1526 |
-
"epoch": 1.69,
|
| 1527 |
-
"learning_rate": 1.935320833193291e-06,
|
| 1528 |
-
"loss": 0.5232,
|
| 1529 |
-
"step": 2470
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 1.7,
|
| 1533 |
-
"learning_rate": 1.852784645504223e-06,
|
| 1534 |
-
"loss": 0.537,
|
| 1535 |
-
"step": 2480
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 1.7,
|
| 1539 |
-
"learning_rate": 1.771931043272257e-06,
|
| 1540 |
-
"loss": 0.5164,
|
| 1541 |
-
"step": 2490
|
| 1542 |
-
},
|
| 1543 |
-
{
|
| 1544 |
-
"epoch": 1.71,
|
| 1545 |
-
"learning_rate": 1.6927703741711437e-06,
|
| 1546 |
-
"loss": 0.5614,
|
| 1547 |
-
"step": 2500
|
| 1548 |
-
},
|
| 1549 |
-
{
|
| 1550 |
-
"epoch": 1.71,
|
| 1551 |
-
"eval_loss": 0.5487334728240967,
|
| 1552 |
-
"eval_runtime": 112.8609,
|
| 1553 |
-
"eval_samples_per_second": 5.759,
|
| 1554 |
-
"eval_steps_per_second": 1.444,
|
| 1555 |
-
"step": 2500
|
| 1556 |
-
},
|
| 1557 |
-
{
|
| 1558 |
-
"epoch": 1.72,
|
| 1559 |
-
"learning_rate": 1.6153127692124298e-06,
|
| 1560 |
-
"loss": 0.6014,
|
| 1561 |
-
"step": 2510
|
| 1562 |
-
},
|
| 1563 |
-
{
|
| 1564 |
-
"epoch": 1.72,
|
| 1565 |
-
"learning_rate": 1.5395681414488938e-06,
|
| 1566 |
-
"loss": 0.5534,
|
| 1567 |
-
"step": 2520
|
| 1568 |
-
},
|
| 1569 |
-
{
|
| 1570 |
-
"epoch": 1.73,
|
| 1571 |
-
"learning_rate": 1.4655461847058633e-06,
|
| 1572 |
-
"loss": 0.5652,
|
| 1573 |
-
"step": 2530
|
| 1574 |
-
},
|
| 1575 |
-
{
|
| 1576 |
-
"epoch": 1.74,
|
| 1577 |
-
"learning_rate": 1.3932563723405833e-06,
|
| 1578 |
-
"loss": 0.6083,
|
| 1579 |
-
"step": 2540
|
| 1580 |
-
},
|
| 1581 |
-
{
|
| 1582 |
-
"epoch": 1.74,
|
| 1583 |
-
"learning_rate": 1.3227079560298305e-06,
|
| 1584 |
-
"loss": 0.5197,
|
| 1585 |
-
"step": 2550
|
| 1586 |
-
},
|
| 1587 |
-
{
|
| 1588 |
-
"epoch": 1.75,
|
| 1589 |
-
"learning_rate": 1.2539099645858776e-06,
|
| 1590 |
-
"loss": 0.5637,
|
| 1591 |
-
"step": 2560
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 1.76,
|
| 1595 |
-
"learning_rate": 1.1868712028009782e-06,
|
| 1596 |
-
"loss": 0.4674,
|
| 1597 |
-
"step": 2570
|
| 1598 |
-
},
|
| 1599 |
-
{
|
| 1600 |
-
"epoch": 1.76,
|
| 1601 |
-
"learning_rate": 1.1216002503205213e-06,
|
| 1602 |
-
"loss": 0.5429,
|
| 1603 |
-
"step": 2580
|
| 1604 |
-
},
|
| 1605 |
-
{
|
| 1606 |
-
"epoch": 1.77,
|
| 1607 |
-
"learning_rate": 1.0581054605450153e-06,
|
| 1608 |
-
"loss": 0.5204,
|
| 1609 |
-
"step": 2590
|
| 1610 |
-
},
|
| 1611 |
-
{
|
| 1612 |
-
"epoch": 1.78,
|
| 1613 |
-
"learning_rate": 9.963949595610117e-07,
|
| 1614 |
-
"loss": 0.5476,
|
| 1615 |
-
"step": 2600
|
| 1616 |
-
},
|
| 1617 |
-
{
|
| 1618 |
-
"epoch": 1.79,
|
| 1619 |
-
"learning_rate": 9.364766451011236e-07,
|
| 1620 |
-
"loss": 0.5862,
|
| 1621 |
-
"step": 2610
|
| 1622 |
-
},
|
| 1623 |
-
{
|
| 1624 |
-
"epoch": 1.79,
|
| 1625 |
-
"learning_rate": 8.783581855332817e-07,
|
| 1626 |
-
"loss": 0.5291,
|
| 1627 |
-
"step": 2620
|
| 1628 |
-
},
|
| 1629 |
-
{
|
| 1630 |
-
"epoch": 1.8,
|
| 1631 |
-
"learning_rate": 8.220470188793128e-07,
|
| 1632 |
-
"loss": 0.5299,
|
| 1633 |
-
"step": 2630
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 1.81,
|
| 1637 |
-
"learning_rate": 7.675503518630428e-07,
|
| 1638 |
-
"loss": 0.5563,
|
| 1639 |
-
"step": 2640
|
| 1640 |
-
},
|
| 1641 |
-
{
|
| 1642 |
-
"epoch": 1.81,
|
| 1643 |
-
"learning_rate": 7.148751589879599e-07,
|
| 1644 |
-
"loss": 0.5179,
|
| 1645 |
-
"step": 2650
|
| 1646 |
-
},
|
| 1647 |
-
{
|
| 1648 |
-
"epoch": 1.82,
|
| 1649 |
-
"learning_rate": 6.640281816446248e-07,
|
| 1650 |
-
"loss": 0.547,
|
| 1651 |
-
"step": 2660
|
| 1652 |
-
},
|
| 1653 |
-
{
|
| 1654 |
-
"epoch": 1.83,
|
| 1655 |
-
"learning_rate": 6.150159272479044e-07,
|
| 1656 |
-
"loss": 0.5231,
|
| 1657 |
-
"step": 2670
|
| 1658 |
-
},
|
| 1659 |
-
{
|
| 1660 |
-
"epoch": 1.83,
|
| 1661 |
-
"learning_rate": 5.678446684041416e-07,
|
| 1662 |
-
"loss": 0.5496,
|
| 1663 |
-
"step": 2680
|
| 1664 |
-
},
|
| 1665 |
-
{
|
| 1666 |
-
"epoch": 1.84,
|
| 1667 |
-
"learning_rate": 5.225204421083934e-07,
|
| 1668 |
-
"loss": 0.5732,
|
| 1669 |
-
"step": 2690
|
| 1670 |
-
},
|
| 1671 |
-
{
|
| 1672 |
-
"epoch": 1.85,
|
| 1673 |
-
"learning_rate": 4.790490489718147e-07,
|
| 1674 |
-
"loss": 0.536,
|
| 1675 |
-
"step": 2700
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 1.85,
|
| 1679 |
-
"learning_rate": 4.3743605247929307e-07,
|
| 1680 |
-
"loss": 0.5526,
|
| 1681 |
-
"step": 2710
|
| 1682 |
-
},
|
| 1683 |
-
{
|
| 1684 |
-
"epoch": 1.86,
|
| 1685 |
-
"learning_rate": 3.9768677827742073e-07,
|
| 1686 |
-
"loss": 0.5145,
|
| 1687 |
-
"step": 2720
|
| 1688 |
-
},
|
| 1689 |
-
{
|
| 1690 |
-
"epoch": 1.87,
|
| 1691 |
-
"learning_rate": 3.598063134929341e-07,
|
| 1692 |
-
"loss": 0.5583,
|
| 1693 |
-
"step": 2730
|
| 1694 |
-
},
|
| 1695 |
-
{
|
| 1696 |
-
"epoch": 1.87,
|
| 1697 |
-
"learning_rate": 3.2379950608164845e-07,
|
| 1698 |
-
"loss": 0.5408,
|
| 1699 |
-
"step": 2740
|
| 1700 |
-
},
|
| 1701 |
-
{
|
| 1702 |
-
"epoch": 1.88,
|
| 1703 |
-
"learning_rate": 2.8967096420802364e-07,
|
| 1704 |
-
"loss": 0.5287,
|
| 1705 |
-
"step": 2750
|
| 1706 |
-
},
|
| 1707 |
-
{
|
| 1708 |
-
"epoch": 1.89,
|
| 1709 |
-
"learning_rate": 2.5742505565539974e-07,
|
| 1710 |
-
"loss": 0.5565,
|
| 1711 |
-
"step": 2760
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 1.89,
|
| 1715 |
-
"learning_rate": 2.270659072670156e-07,
|
| 1716 |
-
"loss": 0.5068,
|
| 1717 |
-
"step": 2770
|
| 1718 |
-
},
|
| 1719 |
-
{
|
| 1720 |
-
"epoch": 1.9,
|
| 1721 |
-
"learning_rate": 1.9859740441784958e-07,
|
| 1722 |
-
"loss": 0.5354,
|
| 1723 |
-
"step": 2780
|
| 1724 |
-
},
|
| 1725 |
-
{
|
| 1726 |
-
"epoch": 1.91,
|
| 1727 |
-
"learning_rate": 1.720231905173675e-07,
|
| 1728 |
-
"loss": 0.5375,
|
| 1729 |
-
"step": 2790
|
| 1730 |
-
},
|
| 1731 |
-
{
|
| 1732 |
-
"epoch": 1.92,
|
| 1733 |
-
"learning_rate": 1.4734666654324046e-07,
|
| 1734 |
-
"loss": 0.5401,
|
| 1735 |
-
"step": 2800
|
| 1736 |
-
},
|
| 1737 |
-
{
|
| 1738 |
-
"epoch": 1.92,
|
| 1739 |
-
"learning_rate": 1.2457099060608156e-07,
|
| 1740 |
-
"loss": 0.5922,
|
| 1741 |
-
"step": 2810
|
| 1742 |
-
},
|
| 1743 |
-
{
|
| 1744 |
-
"epoch": 1.93,
|
| 1745 |
-
"learning_rate": 1.036990775452773e-07,
|
| 1746 |
-
"loss": 0.5462,
|
| 1747 |
-
"step": 2820
|
| 1748 |
-
},
|
| 1749 |
-
{
|
| 1750 |
-
"epoch": 1.94,
|
| 1751 |
-
"learning_rate": 8.473359855593609e-08,
|
| 1752 |
-
"loss": 0.5514,
|
| 1753 |
-
"step": 2830
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 1.94,
|
| 1757 |
-
"learning_rate": 6.767698084703677e-08,
|
| 1758 |
-
"loss": 0.5318,
|
| 1759 |
-
"step": 2840
|
| 1760 |
-
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 1.95,
|
| 1763 |
-
"learning_rate": 5.253140733078865e-08,
|
| 1764 |
-
"loss": 0.5045,
|
| 1765 |
-
"step": 2850
|
| 1766 |
-
},
|
| 1767 |
-
{
|
| 1768 |
-
"epoch": 1.96,
|
| 1769 |
-
"learning_rate": 3.929881634326005e-08,
|
| 1770 |
-
"loss": 0.5767,
|
| 1771 |
-
"step": 2860
|
| 1772 |
-
},
|
| 1773 |
-
{
|
| 1774 |
-
"epoch": 1.96,
|
| 1775 |
-
"learning_rate": 2.7980901396313374e-08,
|
| 1776 |
-
"loss": 0.5566,
|
| 1777 |
-
"step": 2870
|
| 1778 |
-
},
|
| 1779 |
-
{
|
| 1780 |
-
"epoch": 1.97,
|
| 1781 |
-
"learning_rate": 1.8579110960865175e-08,
|
| 1782 |
-
"loss": 0.5318,
|
| 1783 |
-
"step": 2880
|
| 1784 |
-
},
|
| 1785 |
-
{
|
| 1786 |
-
"epoch": 1.98,
|
| 1787 |
-
"learning_rate": 1.1094648281516073e-08,
|
| 1788 |
-
"loss": 0.5267,
|
| 1789 |
-
"step": 2890
|
| 1790 |
-
},
|
| 1791 |
-
{
|
| 1792 |
-
"epoch": 1.98,
|
| 1793 |
-
"learning_rate": 5.528471222552289e-09,
|
| 1794 |
-
"loss": 0.5307,
|
| 1795 |
-
"step": 2900
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 1.99,
|
| 1799 |
-
"learning_rate": 1.8812921453603693e-09,
|
| 1800 |
-
"loss": 0.5309,
|
| 1801 |
-
"step": 2910
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 2.0,
|
| 1805 |
-
"learning_rate": 1.5357781726010878e-10,
|
| 1806 |
-
"loss": 0.5018,
|
| 1807 |
-
"step": 2920
|
| 1808 |
-
},
|
| 1809 |
-
{
|
| 1810 |
-
"epoch": 2.0,
|
| 1811 |
-
"step": 2924,
|
| 1812 |
-
"total_flos": 2.288755460150395e+17,
|
| 1813 |
-
"train_loss": 0.628653114618257,
|
| 1814 |
-
"train_runtime": 6305.9173,
|
| 1815 |
-
"train_samples_per_second": 1.854,
|
| 1816 |
-
"train_steps_per_second": 0.464
|
| 1817 |
}
|
| 1818 |
],
|
| 1819 |
"logging_steps": 10,
|
| 1820 |
-
"max_steps":
|
| 1821 |
-
"num_train_epochs":
|
| 1822 |
-
"save_steps":
|
| 1823 |
-
"total_flos":
|
| 1824 |
"trial_name": null,
|
| 1825 |
"trial_params": null
|
| 1826 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.03419972640218878,
|
| 5 |
+
"eval_steps": 25,
|
| 6 |
+
"global_step": 50,
|
| 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": 9.999999999999999e-06,
|
| 14 |
+
"loss": 1.4953,
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 2.838778253789822e-05,
|
| 20 |
+
"loss": 1.7137,
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 0.01,
|
| 25 |
+
"learning_rate": 2.1314021436425026e-05,
|
| 26 |
+
"loss": 1.5219,
|
| 27 |
"step": 20
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.02,
|
| 31 |
+
"eval_loss": 1.2538621425628662,
|
| 32 |
+
"eval_runtime": 112.7562,
|
| 33 |
+
"eval_samples_per_second": 5.765,
|
| 34 |
+
"eval_steps_per_second": 1.446,
|
| 35 |
+
"step": 25
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.02,
|
| 39 |
+
"learning_rate": 1.1522697745987076e-05,
|
| 40 |
+
"loss": 1.4272,
|
| 41 |
"step": 30
|
| 42 |
},
|
| 43 |
{
|
| 44 |
"epoch": 0.03,
|
| 45 |
+
"learning_rate": 3.2280092208200853e-06,
|
| 46 |
+
"loss": 1.401,
|
| 47 |
"step": 40
|
| 48 |
},
|
| 49 |
{
|
| 50 |
"epoch": 0.03,
|
| 51 |
+
"learning_rate": 0.0,
|
| 52 |
+
"loss": 1.3156,
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
+
"epoch": 0.03,
|
| 57 |
+
"eval_loss": 1.1997405290603638,
|
| 58 |
+
"eval_runtime": 113.4502,
|
| 59 |
+
"eval_samples_per_second": 5.729,
|
| 60 |
+
"eval_steps_per_second": 1.437,
|
| 61 |
+
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
},
|
| 63 |
{
|
| 64 |
+
"epoch": 0.03,
|
| 65 |
+
"step": 50,
|
| 66 |
+
"total_flos": 3919242130882560.0,
|
| 67 |
+
"train_loss": 1.4715181255340577,
|
| 68 |
+
"train_runtime": 335.1031,
|
| 69 |
+
"train_samples_per_second": 0.597,
|
| 70 |
+
"train_steps_per_second": 0.149
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
}
|
| 72 |
],
|
| 73 |
"logging_steps": 10,
|
| 74 |
+
"max_steps": 50,
|
| 75 |
+
"num_train_epochs": 1,
|
| 76 |
+
"save_steps": 25,
|
| 77 |
+
"total_flos": 3919242130882560.0,
|
| 78 |
"trial_name": null,
|
| 79 |
"trial_params": null
|
| 80 |
}
|