Upload 5 files
Browse files- scheduler.pt +3 -0
- sft_args.json +209 -0
- trainer_state.json +3251 -0
- training_args.bin +3 -0
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8a8fcfee27e2311e99fceeb21fe6187b6a8ab14c19134e1a91af67755049cd7
|
| 3 |
+
size 1064
|
sft_args.json
ADDED
|
@@ -0,0 +1,209 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "phi3-vision-128k-instruct",
|
| 3 |
+
"model_id_or_path": "LLM-Research/Phi-3-vision-128k-instruct",
|
| 4 |
+
"model_revision": "master",
|
| 5 |
+
"sft_type": "lora",
|
| 6 |
+
"freeze_parameters": 0.0,
|
| 7 |
+
"additional_trainable_parameters": [],
|
| 8 |
+
"tuner_backend": "peft",
|
| 9 |
+
"template_type": "phi3-vl",
|
| 10 |
+
"output_dir": "D:\\_____NEW_NN\\LLM\\MiniCPM-V\\finetune\\output\\phi3-vision-128k-instruct\\v0-20240531-071942",
|
| 11 |
+
"add_output_dir_suffix": true,
|
| 12 |
+
"ddp_backend": null,
|
| 13 |
+
"ddp_find_unused_parameters": null,
|
| 14 |
+
"ddp_broadcast_buffers": null,
|
| 15 |
+
"seed": 42,
|
| 16 |
+
"resume_from_checkpoint": null,
|
| 17 |
+
"ignore_data_skip": false,
|
| 18 |
+
"dtype": "bf16",
|
| 19 |
+
"packing": false,
|
| 20 |
+
"dataset": [
|
| 21 |
+
"test2"
|
| 22 |
+
],
|
| 23 |
+
"val_dataset": [],
|
| 24 |
+
"dataset_seed": 42,
|
| 25 |
+
"dataset_test_ratio": 0.05,
|
| 26 |
+
"use_loss_scale": false,
|
| 27 |
+
"system": null,
|
| 28 |
+
"max_length": 2048,
|
| 29 |
+
"truncation_strategy": "delete",
|
| 30 |
+
"check_dataset_strategy": "none",
|
| 31 |
+
"model_name": [
|
| 32 |
+
null,
|
| 33 |
+
null
|
| 34 |
+
],
|
| 35 |
+
"model_author": [
|
| 36 |
+
null,
|
| 37 |
+
null
|
| 38 |
+
],
|
| 39 |
+
"quant_method": null,
|
| 40 |
+
"quantization_bit": 0,
|
| 41 |
+
"hqq_axis": 0,
|
| 42 |
+
"hqq_dynamic_config_path": null,
|
| 43 |
+
"bnb_4bit_comp_dtype": "bf16",
|
| 44 |
+
"bnb_4bit_quant_type": "nf4",
|
| 45 |
+
"bnb_4bit_use_double_quant": true,
|
| 46 |
+
"bnb_4bit_quant_storage": null,
|
| 47 |
+
"lora_target_modules": [
|
| 48 |
+
"v_proj",
|
| 49 |
+
"q_proj",
|
| 50 |
+
"out_proj",
|
| 51 |
+
"gate_up_proj",
|
| 52 |
+
"img_projection.0",
|
| 53 |
+
"qkv_proj",
|
| 54 |
+
"img_projection.2",
|
| 55 |
+
"fc1",
|
| 56 |
+
"k_proj",
|
| 57 |
+
"fc2",
|
| 58 |
+
"o_proj",
|
| 59 |
+
"down_proj"
|
| 60 |
+
],
|
| 61 |
+
"lora_rank": 64,
|
| 62 |
+
"lora_alpha": 64,
|
| 63 |
+
"lora_dropout_p": 0.05,
|
| 64 |
+
"lora_bias_trainable": "none",
|
| 65 |
+
"lora_modules_to_save": [],
|
| 66 |
+
"lora_dtype": null,
|
| 67 |
+
"lora_lr_ratio": null,
|
| 68 |
+
"use_rslora": false,
|
| 69 |
+
"use_dora": false,
|
| 70 |
+
"init_lora_weights": true,
|
| 71 |
+
"boft_block_size": 4,
|
| 72 |
+
"boft_block_num": 0,
|
| 73 |
+
"boft_n_butterfly_factor": 1,
|
| 74 |
+
"boft_target_modules": [
|
| 75 |
+
"DEFAULT"
|
| 76 |
+
],
|
| 77 |
+
"boft_dropout": 0.0,
|
| 78 |
+
"boft_modules_to_save": [],
|
| 79 |
+
"vera_rank": 256,
|
| 80 |
+
"vera_target_modules": [
|
| 81 |
+
"DEFAULT"
|
| 82 |
+
],
|
| 83 |
+
"vera_projection_prng_key": 0,
|
| 84 |
+
"vera_dropout": 0.0,
|
| 85 |
+
"vera_d_initial": 0.1,
|
| 86 |
+
"vera_modules_to_save": [],
|
| 87 |
+
"adapter_act": "gelu",
|
| 88 |
+
"adapter_length": 128,
|
| 89 |
+
"use_galore": false,
|
| 90 |
+
"galore_rank": 128,
|
| 91 |
+
"galore_target_modules": null,
|
| 92 |
+
"galore_update_proj_gap": 50,
|
| 93 |
+
"galore_scale": 1.0,
|
| 94 |
+
"galore_proj_type": "std",
|
| 95 |
+
"galore_optim_per_parameter": false,
|
| 96 |
+
"galore_with_embedding": false,
|
| 97 |
+
"adalora_target_r": 8,
|
| 98 |
+
"adalora_init_r": 12,
|
| 99 |
+
"adalora_tinit": 0,
|
| 100 |
+
"adalora_tfinal": 0,
|
| 101 |
+
"adalora_deltaT": 1,
|
| 102 |
+
"adalora_beta1": 0.85,
|
| 103 |
+
"adalora_beta2": 0.85,
|
| 104 |
+
"adalora_orth_reg_weight": 0.5,
|
| 105 |
+
"ia3_target_modules": [
|
| 106 |
+
"DEFAULT"
|
| 107 |
+
],
|
| 108 |
+
"ia3_feedforward_modules": [],
|
| 109 |
+
"ia3_modules_to_save": [],
|
| 110 |
+
"llamapro_num_new_blocks": 4,
|
| 111 |
+
"llamapro_num_groups": null,
|
| 112 |
+
"neftune_noise_alpha": null,
|
| 113 |
+
"neftune_backend": "transformers",
|
| 114 |
+
"lisa_activated_layers": 0,
|
| 115 |
+
"lisa_step_interval": 20,
|
| 116 |
+
"gradient_checkpointing": true,
|
| 117 |
+
"deepspeed": null,
|
| 118 |
+
"batch_size": 1,
|
| 119 |
+
"eval_batch_size": 1,
|
| 120 |
+
"num_train_epochs": 4,
|
| 121 |
+
"max_steps": -1,
|
| 122 |
+
"optim": "adamw_torch",
|
| 123 |
+
"adam_beta1": 0.9,
|
| 124 |
+
"adam_beta2": 0.999,
|
| 125 |
+
"learning_rate": 0.00015,
|
| 126 |
+
"weight_decay": 0.1,
|
| 127 |
+
"gradient_accumulation_steps": 2,
|
| 128 |
+
"max_grad_norm": 0.5,
|
| 129 |
+
"predict_with_generate": false,
|
| 130 |
+
"lr_scheduler_type": "linear",
|
| 131 |
+
"warmup_ratio": 0.05,
|
| 132 |
+
"eval_steps": 50,
|
| 133 |
+
"save_steps": 369,
|
| 134 |
+
"save_only_model": false,
|
| 135 |
+
"save_total_limit": 2,
|
| 136 |
+
"logging_steps": 5,
|
| 137 |
+
"dataloader_num_workers": 1,
|
| 138 |
+
"dataloader_pin_memory": true,
|
| 139 |
+
"dataloader_drop_last": false,
|
| 140 |
+
"push_to_hub": false,
|
| 141 |
+
"hub_model_id": null,
|
| 142 |
+
"hub_token": null,
|
| 143 |
+
"hub_private_repo": false,
|
| 144 |
+
"push_hub_strategy": "push_best",
|
| 145 |
+
"test_oom_error": false,
|
| 146 |
+
"disable_tqdm": false,
|
| 147 |
+
"lazy_tokenize": true,
|
| 148 |
+
"preprocess_num_proc": 1,
|
| 149 |
+
"use_flash_attn": null,
|
| 150 |
+
"ignore_args_error": false,
|
| 151 |
+
"check_model_is_latest": true,
|
| 152 |
+
"logging_dir": "D:\\_____NEW_NN\\LLM\\MiniCPM-V\\finetune\\output\\phi3-vision-128k-instruct\\v0-20240531-071942/runs",
|
| 153 |
+
"report_to": [
|
| 154 |
+
"tensorboard"
|
| 155 |
+
],
|
| 156 |
+
"acc_strategy": "token",
|
| 157 |
+
"save_on_each_node": true,
|
| 158 |
+
"evaluation_strategy": "steps",
|
| 159 |
+
"save_strategy": "steps",
|
| 160 |
+
"save_safetensors": true,
|
| 161 |
+
"gpu_memory_fraction": null,
|
| 162 |
+
"include_num_input_tokens_seen": false,
|
| 163 |
+
"local_repo_path": null,
|
| 164 |
+
"custom_register_path": null,
|
| 165 |
+
"custom_dataset_info": null,
|
| 166 |
+
"device_map_config_path": null,
|
| 167 |
+
"max_new_tokens": 2048,
|
| 168 |
+
"do_sample": true,
|
| 169 |
+
"temperature": 0.3,
|
| 170 |
+
"top_k": 20,
|
| 171 |
+
"top_p": 0.7,
|
| 172 |
+
"repetition_penalty": 1.0,
|
| 173 |
+
"num_beams": 1,
|
| 174 |
+
"fsdp": "",
|
| 175 |
+
"fsdp_config": null,
|
| 176 |
+
"sequence_parallel_size": 1,
|
| 177 |
+
"model_layer_cls_name": null,
|
| 178 |
+
"metric_warmup_step": 0,
|
| 179 |
+
"fsdp_num": 1,
|
| 180 |
+
"per_device_train_batch_size": null,
|
| 181 |
+
"per_device_eval_batch_size": null,
|
| 182 |
+
"self_cognition_sample": 0,
|
| 183 |
+
"train_dataset_mix_ratio": 0.0,
|
| 184 |
+
"train_dataset_mix_ds": [
|
| 185 |
+
"ms-bench"
|
| 186 |
+
],
|
| 187 |
+
"train_dataset_sample": -1,
|
| 188 |
+
"val_dataset_sample": null,
|
| 189 |
+
"safe_serialization": null,
|
| 190 |
+
"only_save_model": null,
|
| 191 |
+
"neftune_alpha": null,
|
| 192 |
+
"deepspeed_config_path": null,
|
| 193 |
+
"model_cache_dir": null,
|
| 194 |
+
"custom_train_dataset_path": [],
|
| 195 |
+
"custom_val_dataset_path": [],
|
| 196 |
+
"use_self_cognition": false,
|
| 197 |
+
"lora_use_embedding": false,
|
| 198 |
+
"lora_use_all": true,
|
| 199 |
+
"lora_m2s_use_embedding": false,
|
| 200 |
+
"lora_m2s_use_ln": false,
|
| 201 |
+
"torch_dtype": "torch.bfloat16",
|
| 202 |
+
"fp16": false,
|
| 203 |
+
"bf16": true,
|
| 204 |
+
"bnb_4bit_compute_dtype": "torch.bfloat16",
|
| 205 |
+
"load_in_4bit": false,
|
| 206 |
+
"load_in_8bit": false,
|
| 207 |
+
"train_sampler_random": true,
|
| 208 |
+
"training_args": "Seq2SeqTrainingArguments(output_dir='D:\\\\_____NEW_NN\\\\LLM\\\\MiniCPM-V\\\\finetune\\\\output\\\\phi3-vision-128k-instruct\\\\v0-20240531-071942', overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, evaluation_strategy=<IntervalStrategy.STEPS: 'steps'>, prediction_loss_only=False, per_device_train_batch_size=1, per_device_eval_batch_size=1, per_gpu_train_batch_size=None, per_gpu_eval_batch_size=None, gradient_accumulation_steps=2, eval_accumulation_steps=None, eval_delay=0, learning_rate=0.00015, weight_decay=0.1, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-08, max_grad_norm=0.5, num_train_epochs=4, max_steps=-1, lr_scheduler_type=<SchedulerType.LINEAR: 'linear'>, lr_scheduler_kwargs={}, warmup_ratio=0.05, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='D:\\\\_____NEW_NN\\\\LLM\\\\MiniCPM-V\\\\finetune\\\\output\\\\phi3-vision-128k-instruct\\\\v0-20240531-071942/runs', logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=5, logging_nan_inf_filter=True, save_strategy=<IntervalStrategy.STEPS: 'steps'>, save_steps=369, save_total_limit=2, save_safetensors=True, save_on_each_node=True, save_only_model=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=None, jit_mode_eval=False, use_ipex=False, bf16=True, fp16=False, fp16_opt_level='O1', half_precision_backend='auto', bf16_full_eval=False, fp16_full_eval=False, tf32=None, local_rank=0, ddp_backend=None, tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=50, dataloader_num_workers=1, dataloader_prefetch_factor=None, past_index=-1, run_name='D:\\\\_____NEW_NN\\\\LLM\\\\MiniCPM-V\\\\finetune\\\\output\\\\phi3-vision-128k-instruct\\\\v0-20240531-071942', disable_tqdm=False, remove_unused_columns=False, label_names=None, load_best_model_at_end=False, metric_for_best_model='loss', greater_is_better=False, ignore_data_skip=False, fsdp=[], fsdp_min_num_params=0, fsdp_config={'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=None, even_batches=True, use_seedable_sampler=True, gradient_accumulation_kwargs=None), deepspeed=None, label_smoothing_factor=0.0, optim=<OptimizerNames.ADAMW_TORCH: 'adamw_torch'>, optim_args=None, adafactor=False, group_by_length=False, length_column_name='length', report_to=['tensorboard'], ddp_find_unused_parameters=None, ddp_bucket_cap_mb=None, ddp_broadcast_buffers=None, dataloader_pin_memory=True, dataloader_persistent_workers=False, skip_memory_metrics=True, use_legacy_prediction_loop=False, push_to_hub=False, resume_from_checkpoint=None, hub_model_id=None, hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=False, hub_always_push=False, gradient_checkpointing=True, gradient_checkpointing_kwargs=None, include_inputs_for_metrics=False, eval_do_concat_batches=True, fp16_backend='auto', push_to_hub_model_id=None, push_to_hub_organization=None, push_to_hub_token=None, mp_parameters='', auto_find_batch_size=False, full_determinism=False, torchdynamo=None, ray_scope='last', ddp_timeout=1800, torch_compile=False, torch_compile_backend=None, torch_compile_mode=None, dispatch_batches=None, split_batches=None, include_tokens_per_second=False, include_num_input_tokens_seen=False, neftune_noise_alpha=None, optim_target_modules=None, sortish_sampler=True, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=GenerationConfig {\n \"do_sample\": true,\n \"eos_token_id\": 32000,\n \"max_new_tokens\": 2048,\n \"pad_token_id\": 32000,\n \"temperature\": 0.3,\n \"top_k\": 20,\n \"top_p\": 0.7\n}\n, train_sampler_random=True, push_hub_strategy='push_best', acc_strategy='token', additional_saved_files=[], metric_warmup_step=0, train_dataset_sample=738)"
|
| 209 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3251 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.89783895,
|
| 3 |
+
"best_model_checkpoint": "D:\\_____NEW_NN\\LLM\\MiniCPM-V\\finetune\\output\\phi3-vision-128k-instruct\\v0-20240531-071942\\checkpoint-1476",
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 50,
|
| 6 |
+
"global_step": 1476,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"acc": 0.47843874,
|
| 13 |
+
"epoch": 0.0027100271002710027,
|
| 14 |
+
"grad_norm": 0.921875,
|
| 15 |
+
"learning_rate": 2.027027027027027e-06,
|
| 16 |
+
"loss": 2.52092218,
|
| 17 |
+
"memory(GiB)": 12.33,
|
| 18 |
+
"step": 1,
|
| 19 |
+
"train_speed(iter/s)": 0.066334
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"acc": 0.51126236,
|
| 23 |
+
"epoch": 0.013550135501355014,
|
| 24 |
+
"grad_norm": 0.796875,
|
| 25 |
+
"learning_rate": 1.0135135135135135e-05,
|
| 26 |
+
"loss": 2.42214346,
|
| 27 |
+
"memory(GiB)": 13.38,
|
| 28 |
+
"step": 5,
|
| 29 |
+
"train_speed(iter/s)": 0.127981
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"acc": 0.48636103,
|
| 33 |
+
"epoch": 0.02710027100271003,
|
| 34 |
+
"grad_norm": 1.09375,
|
| 35 |
+
"learning_rate": 2.027027027027027e-05,
|
| 36 |
+
"loss": 2.53938828,
|
| 37 |
+
"memory(GiB)": 13.38,
|
| 38 |
+
"step": 10,
|
| 39 |
+
"train_speed(iter/s)": 0.144622
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"acc": 0.47292571,
|
| 43 |
+
"epoch": 0.04065040650406504,
|
| 44 |
+
"grad_norm": 1.1328125,
|
| 45 |
+
"learning_rate": 3.0405405405405404e-05,
|
| 46 |
+
"loss": 2.50403137,
|
| 47 |
+
"memory(GiB)": 13.38,
|
| 48 |
+
"step": 15,
|
| 49 |
+
"train_speed(iter/s)": 0.150329
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"acc": 0.52069569,
|
| 53 |
+
"epoch": 0.05420054200542006,
|
| 54 |
+
"grad_norm": 0.83984375,
|
| 55 |
+
"learning_rate": 4.054054054054054e-05,
|
| 56 |
+
"loss": 2.28312988,
|
| 57 |
+
"memory(GiB)": 13.38,
|
| 58 |
+
"step": 20,
|
| 59 |
+
"train_speed(iter/s)": 0.152832
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"acc": 0.53072515,
|
| 63 |
+
"epoch": 0.06775067750677506,
|
| 64 |
+
"grad_norm": 0.59765625,
|
| 65 |
+
"learning_rate": 5.067567567567567e-05,
|
| 66 |
+
"loss": 2.20206394,
|
| 67 |
+
"memory(GiB)": 13.38,
|
| 68 |
+
"step": 25,
|
| 69 |
+
"train_speed(iter/s)": 0.154186
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"acc": 0.52423077,
|
| 73 |
+
"epoch": 0.08130081300813008,
|
| 74 |
+
"grad_norm": 0.89453125,
|
| 75 |
+
"learning_rate": 6.081081081081081e-05,
|
| 76 |
+
"loss": 2.31712227,
|
| 77 |
+
"memory(GiB)": 14.44,
|
| 78 |
+
"step": 30,
|
| 79 |
+
"train_speed(iter/s)": 0.155403
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"acc": 0.54341574,
|
| 83 |
+
"epoch": 0.0948509485094851,
|
| 84 |
+
"grad_norm": 0.7109375,
|
| 85 |
+
"learning_rate": 7.094594594594594e-05,
|
| 86 |
+
"loss": 2.03825722,
|
| 87 |
+
"memory(GiB)": 14.44,
|
| 88 |
+
"step": 35,
|
| 89 |
+
"train_speed(iter/s)": 0.156383
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"acc": 0.56128917,
|
| 93 |
+
"epoch": 0.10840108401084012,
|
| 94 |
+
"grad_norm": 0.734375,
|
| 95 |
+
"learning_rate": 8.108108108108108e-05,
|
| 96 |
+
"loss": 1.98176575,
|
| 97 |
+
"memory(GiB)": 14.44,
|
| 98 |
+
"step": 40,
|
| 99 |
+
"train_speed(iter/s)": 0.156973
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"acc": 0.53511982,
|
| 103 |
+
"epoch": 0.12195121951219512,
|
| 104 |
+
"grad_norm": 1.1484375,
|
| 105 |
+
"learning_rate": 9.121621621621621e-05,
|
| 106 |
+
"loss": 2.07999401,
|
| 107 |
+
"memory(GiB)": 14.44,
|
| 108 |
+
"step": 45,
|
| 109 |
+
"train_speed(iter/s)": 0.157395
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"acc": 0.56112757,
|
| 113 |
+
"epoch": 0.13550135501355012,
|
| 114 |
+
"grad_norm": 0.83984375,
|
| 115 |
+
"learning_rate": 0.00010135135135135135,
|
| 116 |
+
"loss": 1.92188244,
|
| 117 |
+
"memory(GiB)": 14.44,
|
| 118 |
+
"step": 50,
|
| 119 |
+
"train_speed(iter/s)": 0.157924
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.13550135501355012,
|
| 123 |
+
"eval_acc": 0.5618983279851376,
|
| 124 |
+
"eval_loss": 1.901513695716858,
|
| 125 |
+
"eval_runtime": 44.6741,
|
| 126 |
+
"eval_samples_per_second": 0.851,
|
| 127 |
+
"eval_steps_per_second": 0.851,
|
| 128 |
+
"step": 50
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"acc": 0.57467785,
|
| 132 |
+
"epoch": 0.14905149051490515,
|
| 133 |
+
"grad_norm": 0.9921875,
|
| 134 |
+
"learning_rate": 0.00011148648648648647,
|
| 135 |
+
"loss": 1.92710114,
|
| 136 |
+
"memory(GiB)": 15.21,
|
| 137 |
+
"step": 55,
|
| 138 |
+
"train_speed(iter/s)": 0.140365
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"acc": 0.55041471,
|
| 142 |
+
"epoch": 0.16260162601626016,
|
| 143 |
+
"grad_norm": 1.1484375,
|
| 144 |
+
"learning_rate": 0.00012162162162162162,
|
| 145 |
+
"loss": 2.09731407,
|
| 146 |
+
"memory(GiB)": 15.21,
|
| 147 |
+
"step": 60,
|
| 148 |
+
"train_speed(iter/s)": 0.141894
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"acc": 0.53853202,
|
| 152 |
+
"epoch": 0.17615176151761516,
|
| 153 |
+
"grad_norm": 1.03125,
|
| 154 |
+
"learning_rate": 0.00013175675675675675,
|
| 155 |
+
"loss": 2.01167774,
|
| 156 |
+
"memory(GiB)": 15.21,
|
| 157 |
+
"step": 65,
|
| 158 |
+
"train_speed(iter/s)": 0.143162
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"acc": 0.57625084,
|
| 162 |
+
"epoch": 0.1897018970189702,
|
| 163 |
+
"grad_norm": 0.81640625,
|
| 164 |
+
"learning_rate": 0.00014189189189189188,
|
| 165 |
+
"loss": 1.77697067,
|
| 166 |
+
"memory(GiB)": 15.21,
|
| 167 |
+
"step": 70,
|
| 168 |
+
"train_speed(iter/s)": 0.144318
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"acc": 0.57468171,
|
| 172 |
+
"epoch": 0.2032520325203252,
|
| 173 |
+
"grad_norm": 0.96484375,
|
| 174 |
+
"learning_rate": 0.00014989300998573466,
|
| 175 |
+
"loss": 1.92596684,
|
| 176 |
+
"memory(GiB)": 15.21,
|
| 177 |
+
"step": 75,
|
| 178 |
+
"train_speed(iter/s)": 0.145344
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"acc": 0.58551345,
|
| 182 |
+
"epoch": 0.21680216802168023,
|
| 183 |
+
"grad_norm": 1.0546875,
|
| 184 |
+
"learning_rate": 0.00014935805991440798,
|
| 185 |
+
"loss": 1.85387783,
|
| 186 |
+
"memory(GiB)": 15.21,
|
| 187 |
+
"step": 80,
|
| 188 |
+
"train_speed(iter/s)": 0.146049
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"acc": 0.54782548,
|
| 192 |
+
"epoch": 0.23035230352303523,
|
| 193 |
+
"grad_norm": 1.0078125,
|
| 194 |
+
"learning_rate": 0.0001488231098430813,
|
| 195 |
+
"loss": 2.04980812,
|
| 196 |
+
"memory(GiB)": 15.21,
|
| 197 |
+
"step": 85,
|
| 198 |
+
"train_speed(iter/s)": 0.146726
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"acc": 0.56594706,
|
| 202 |
+
"epoch": 0.24390243902439024,
|
| 203 |
+
"grad_norm": 1.0,
|
| 204 |
+
"learning_rate": 0.0001482881597717546,
|
| 205 |
+
"loss": 1.93099308,
|
| 206 |
+
"memory(GiB)": 15.21,
|
| 207 |
+
"step": 90,
|
| 208 |
+
"train_speed(iter/s)": 0.147437
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"acc": 0.57289362,
|
| 212 |
+
"epoch": 0.25745257452574527,
|
| 213 |
+
"grad_norm": 0.8046875,
|
| 214 |
+
"learning_rate": 0.00014775320970042795,
|
| 215 |
+
"loss": 1.84038963,
|
| 216 |
+
"memory(GiB)": 15.58,
|
| 217 |
+
"step": 95,
|
| 218 |
+
"train_speed(iter/s)": 0.147976
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"acc": 0.5938961,
|
| 222 |
+
"epoch": 0.27100271002710025,
|
| 223 |
+
"grad_norm": 1.015625,
|
| 224 |
+
"learning_rate": 0.00014721825962910127,
|
| 225 |
+
"loss": 1.71151619,
|
| 226 |
+
"memory(GiB)": 15.58,
|
| 227 |
+
"step": 100,
|
| 228 |
+
"train_speed(iter/s)": 0.148573
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.27100271002710025,
|
| 232 |
+
"eval_acc": 0.587907448066205,
|
| 233 |
+
"eval_loss": 1.7437644004821777,
|
| 234 |
+
"eval_runtime": 44.775,
|
| 235 |
+
"eval_samples_per_second": 0.849,
|
| 236 |
+
"eval_steps_per_second": 0.849,
|
| 237 |
+
"step": 100
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"acc": 0.5828393,
|
| 241 |
+
"epoch": 0.2845528455284553,
|
| 242 |
+
"grad_norm": 1.078125,
|
| 243 |
+
"learning_rate": 0.00014668330955777461,
|
| 244 |
+
"loss": 1.90272522,
|
| 245 |
+
"memory(GiB)": 15.58,
|
| 246 |
+
"step": 105,
|
| 247 |
+
"train_speed(iter/s)": 0.140241
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"acc": 0.57981925,
|
| 251 |
+
"epoch": 0.2981029810298103,
|
| 252 |
+
"grad_norm": 0.859375,
|
| 253 |
+
"learning_rate": 0.00014614835948644793,
|
| 254 |
+
"loss": 1.88997154,
|
| 255 |
+
"memory(GiB)": 15.58,
|
| 256 |
+
"step": 110,
|
| 257 |
+
"train_speed(iter/s)": 0.141057
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"acc": 0.56191244,
|
| 261 |
+
"epoch": 0.3116531165311653,
|
| 262 |
+
"grad_norm": 0.94140625,
|
| 263 |
+
"learning_rate": 0.00014561340941512125,
|
| 264 |
+
"loss": 1.92102909,
|
| 265 |
+
"memory(GiB)": 15.58,
|
| 266 |
+
"step": 115,
|
| 267 |
+
"train_speed(iter/s)": 0.141824
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"acc": 0.57420607,
|
| 271 |
+
"epoch": 0.3252032520325203,
|
| 272 |
+
"grad_norm": 1.171875,
|
| 273 |
+
"learning_rate": 0.00014507845934379456,
|
| 274 |
+
"loss": 1.80924416,
|
| 275 |
+
"memory(GiB)": 15.58,
|
| 276 |
+
"step": 120,
|
| 277 |
+
"train_speed(iter/s)": 0.14256
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"acc": 0.57703466,
|
| 281 |
+
"epoch": 0.33875338753387535,
|
| 282 |
+
"grad_norm": 0.84765625,
|
| 283 |
+
"learning_rate": 0.00014454350927246788,
|
| 284 |
+
"loss": 1.85804043,
|
| 285 |
+
"memory(GiB)": 15.58,
|
| 286 |
+
"step": 125,
|
| 287 |
+
"train_speed(iter/s)": 0.143173
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"acc": 0.58581357,
|
| 291 |
+
"epoch": 0.3523035230352303,
|
| 292 |
+
"grad_norm": 0.96484375,
|
| 293 |
+
"learning_rate": 0.0001440085592011412,
|
| 294 |
+
"loss": 1.75103779,
|
| 295 |
+
"memory(GiB)": 15.58,
|
| 296 |
+
"step": 130,
|
| 297 |
+
"train_speed(iter/s)": 0.143759
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"acc": 0.59179254,
|
| 301 |
+
"epoch": 0.36585365853658536,
|
| 302 |
+
"grad_norm": 0.93359375,
|
| 303 |
+
"learning_rate": 0.00014347360912981454,
|
| 304 |
+
"loss": 1.67790794,
|
| 305 |
+
"memory(GiB)": 15.58,
|
| 306 |
+
"step": 135,
|
| 307 |
+
"train_speed(iter/s)": 0.14428
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"acc": 0.59449296,
|
| 311 |
+
"epoch": 0.3794037940379404,
|
| 312 |
+
"grad_norm": 0.74609375,
|
| 313 |
+
"learning_rate": 0.00014293865905848786,
|
| 314 |
+
"loss": 1.72561359,
|
| 315 |
+
"memory(GiB)": 15.58,
|
| 316 |
+
"step": 140,
|
| 317 |
+
"train_speed(iter/s)": 0.144799
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"acc": 0.57711525,
|
| 321 |
+
"epoch": 0.39295392953929537,
|
| 322 |
+
"grad_norm": 1.109375,
|
| 323 |
+
"learning_rate": 0.0001424037089871612,
|
| 324 |
+
"loss": 1.79506092,
|
| 325 |
+
"memory(GiB)": 15.58,
|
| 326 |
+
"step": 145,
|
| 327 |
+
"train_speed(iter/s)": 0.14532
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"acc": 0.57541366,
|
| 331 |
+
"epoch": 0.4065040650406504,
|
| 332 |
+
"grad_norm": 5.46875,
|
| 333 |
+
"learning_rate": 0.00014186875891583452,
|
| 334 |
+
"loss": 1.80377541,
|
| 335 |
+
"memory(GiB)": 15.58,
|
| 336 |
+
"step": 150,
|
| 337 |
+
"train_speed(iter/s)": 0.145757
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.4065040650406504,
|
| 341 |
+
"eval_acc": 0.5983786522546867,
|
| 342 |
+
"eval_loss": 1.676965594291687,
|
| 343 |
+
"eval_runtime": 44.3227,
|
| 344 |
+
"eval_samples_per_second": 0.857,
|
| 345 |
+
"eval_steps_per_second": 0.857,
|
| 346 |
+
"step": 150
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"acc": 0.58702893,
|
| 350 |
+
"epoch": 0.42005420054200543,
|
| 351 |
+
"grad_norm": 1.0625,
|
| 352 |
+
"learning_rate": 0.00014133380884450783,
|
| 353 |
+
"loss": 1.86865826,
|
| 354 |
+
"memory(GiB)": 15.58,
|
| 355 |
+
"step": 155,
|
| 356 |
+
"train_speed(iter/s)": 0.140358
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"acc": 0.58541198,
|
| 360 |
+
"epoch": 0.43360433604336046,
|
| 361 |
+
"grad_norm": 1.015625,
|
| 362 |
+
"learning_rate": 0.00014079885877318115,
|
| 363 |
+
"loss": 1.82644196,
|
| 364 |
+
"memory(GiB)": 15.58,
|
| 365 |
+
"step": 160,
|
| 366 |
+
"train_speed(iter/s)": 0.140936
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"acc": 0.58856125,
|
| 370 |
+
"epoch": 0.44715447154471544,
|
| 371 |
+
"grad_norm": 0.9921875,
|
| 372 |
+
"learning_rate": 0.00014026390870185447,
|
| 373 |
+
"loss": 1.68714104,
|
| 374 |
+
"memory(GiB)": 15.58,
|
| 375 |
+
"step": 165,
|
| 376 |
+
"train_speed(iter/s)": 0.141495
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"acc": 0.59458299,
|
| 380 |
+
"epoch": 0.46070460704607047,
|
| 381 |
+
"grad_norm": 0.91015625,
|
| 382 |
+
"learning_rate": 0.0001397289586305278,
|
| 383 |
+
"loss": 1.72063637,
|
| 384 |
+
"memory(GiB)": 15.58,
|
| 385 |
+
"step": 170,
|
| 386 |
+
"train_speed(iter/s)": 0.142005
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"acc": 0.57354069,
|
| 390 |
+
"epoch": 0.4742547425474255,
|
| 391 |
+
"grad_norm": 0.7734375,
|
| 392 |
+
"learning_rate": 0.00013919400855920113,
|
| 393 |
+
"loss": 1.83291931,
|
| 394 |
+
"memory(GiB)": 15.58,
|
| 395 |
+
"step": 175,
|
| 396 |
+
"train_speed(iter/s)": 0.142509
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"acc": 0.58851786,
|
| 400 |
+
"epoch": 0.4878048780487805,
|
| 401 |
+
"grad_norm": 1.234375,
|
| 402 |
+
"learning_rate": 0.00013865905848787447,
|
| 403 |
+
"loss": 1.70540199,
|
| 404 |
+
"memory(GiB)": 15.58,
|
| 405 |
+
"step": 180,
|
| 406 |
+
"train_speed(iter/s)": 0.143002
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"acc": 0.59666262,
|
| 410 |
+
"epoch": 0.5013550135501355,
|
| 411 |
+
"grad_norm": 0.87890625,
|
| 412 |
+
"learning_rate": 0.0001381241084165478,
|
| 413 |
+
"loss": 1.70772285,
|
| 414 |
+
"memory(GiB)": 15.58,
|
| 415 |
+
"step": 185,
|
| 416 |
+
"train_speed(iter/s)": 0.143463
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"acc": 0.58001013,
|
| 420 |
+
"epoch": 0.5149051490514905,
|
| 421 |
+
"grad_norm": 0.890625,
|
| 422 |
+
"learning_rate": 0.0001375891583452211,
|
| 423 |
+
"loss": 1.73730106,
|
| 424 |
+
"memory(GiB)": 15.58,
|
| 425 |
+
"step": 190,
|
| 426 |
+
"train_speed(iter/s)": 0.143883
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"acc": 0.5934463,
|
| 430 |
+
"epoch": 0.5284552845528455,
|
| 431 |
+
"grad_norm": 1.421875,
|
| 432 |
+
"learning_rate": 0.00013705420827389442,
|
| 433 |
+
"loss": 1.68941402,
|
| 434 |
+
"memory(GiB)": 15.58,
|
| 435 |
+
"step": 195,
|
| 436 |
+
"train_speed(iter/s)": 0.144297
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"acc": 0.58476119,
|
| 440 |
+
"epoch": 0.5420054200542005,
|
| 441 |
+
"grad_norm": 0.91015625,
|
| 442 |
+
"learning_rate": 0.00013651925820256774,
|
| 443 |
+
"loss": 1.74896088,
|
| 444 |
+
"memory(GiB)": 15.58,
|
| 445 |
+
"step": 200,
|
| 446 |
+
"train_speed(iter/s)": 0.144655
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.5420054200542005,
|
| 450 |
+
"eval_acc": 0.6105387603445364,
|
| 451 |
+
"eval_loss": 1.6297248601913452,
|
| 452 |
+
"eval_runtime": 44.3331,
|
| 453 |
+
"eval_samples_per_second": 0.857,
|
| 454 |
+
"eval_steps_per_second": 0.857,
|
| 455 |
+
"step": 200
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"acc": 0.58323898,
|
| 459 |
+
"epoch": 0.5555555555555556,
|
| 460 |
+
"grad_norm": 0.79296875,
|
| 461 |
+
"learning_rate": 0.00013598430813124105,
|
| 462 |
+
"loss": 1.74196358,
|
| 463 |
+
"memory(GiB)": 15.58,
|
| 464 |
+
"step": 205,
|
| 465 |
+
"train_speed(iter/s)": 0.140612
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"acc": 0.59854908,
|
| 469 |
+
"epoch": 0.5691056910569106,
|
| 470 |
+
"grad_norm": 1.109375,
|
| 471 |
+
"learning_rate": 0.0001354493580599144,
|
| 472 |
+
"loss": 1.6279623,
|
| 473 |
+
"memory(GiB)": 15.58,
|
| 474 |
+
"step": 210,
|
| 475 |
+
"train_speed(iter/s)": 0.141053
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"acc": 0.58306313,
|
| 479 |
+
"epoch": 0.5826558265582655,
|
| 480 |
+
"grad_norm": 0.98828125,
|
| 481 |
+
"learning_rate": 0.00013491440798858771,
|
| 482 |
+
"loss": 1.85492191,
|
| 483 |
+
"memory(GiB)": 15.58,
|
| 484 |
+
"step": 215,
|
| 485 |
+
"train_speed(iter/s)": 0.141461
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"acc": 0.58454275,
|
| 489 |
+
"epoch": 0.5962059620596206,
|
| 490 |
+
"grad_norm": 0.890625,
|
| 491 |
+
"learning_rate": 0.00013437945791726106,
|
| 492 |
+
"loss": 1.75104046,
|
| 493 |
+
"memory(GiB)": 15.58,
|
| 494 |
+
"step": 220,
|
| 495 |
+
"train_speed(iter/s)": 0.141869
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"acc": 0.60898943,
|
| 499 |
+
"epoch": 0.6097560975609756,
|
| 500 |
+
"grad_norm": 0.828125,
|
| 501 |
+
"learning_rate": 0.00013384450784593437,
|
| 502 |
+
"loss": 1.57786808,
|
| 503 |
+
"memory(GiB)": 15.58,
|
| 504 |
+
"step": 225,
|
| 505 |
+
"train_speed(iter/s)": 0.142237
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"acc": 0.58954048,
|
| 509 |
+
"epoch": 0.6233062330623306,
|
| 510 |
+
"grad_norm": 1.1640625,
|
| 511 |
+
"learning_rate": 0.0001333095577746077,
|
| 512 |
+
"loss": 1.72581081,
|
| 513 |
+
"memory(GiB)": 15.58,
|
| 514 |
+
"step": 230,
|
| 515 |
+
"train_speed(iter/s)": 0.142622
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"acc": 0.59608021,
|
| 519 |
+
"epoch": 0.6368563685636857,
|
| 520 |
+
"grad_norm": 1.2265625,
|
| 521 |
+
"learning_rate": 0.000132774607703281,
|
| 522 |
+
"loss": 1.70160866,
|
| 523 |
+
"memory(GiB)": 15.58,
|
| 524 |
+
"step": 235,
|
| 525 |
+
"train_speed(iter/s)": 0.142983
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"acc": 0.57084417,
|
| 529 |
+
"epoch": 0.6504065040650406,
|
| 530 |
+
"grad_norm": 1.1640625,
|
| 531 |
+
"learning_rate": 0.00013223965763195432,
|
| 532 |
+
"loss": 1.81941319,
|
| 533 |
+
"memory(GiB)": 15.58,
|
| 534 |
+
"step": 240,
|
| 535 |
+
"train_speed(iter/s)": 0.143329
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"acc": 0.61476159,
|
| 539 |
+
"epoch": 0.6639566395663956,
|
| 540 |
+
"grad_norm": 0.9453125,
|
| 541 |
+
"learning_rate": 0.00013170470756062767,
|
| 542 |
+
"loss": 1.68505306,
|
| 543 |
+
"memory(GiB)": 15.58,
|
| 544 |
+
"step": 245,
|
| 545 |
+
"train_speed(iter/s)": 0.143652
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"acc": 0.5995626,
|
| 549 |
+
"epoch": 0.6775067750677507,
|
| 550 |
+
"grad_norm": 0.9140625,
|
| 551 |
+
"learning_rate": 0.00013116975748930098,
|
| 552 |
+
"loss": 1.77789631,
|
| 553 |
+
"memory(GiB)": 15.58,
|
| 554 |
+
"step": 250,
|
| 555 |
+
"train_speed(iter/s)": 0.144002
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6775067750677507,
|
| 559 |
+
"eval_acc": 0.61087654112481,
|
| 560 |
+
"eval_loss": 1.6108067035675049,
|
| 561 |
+
"eval_runtime": 44.1094,
|
| 562 |
+
"eval_samples_per_second": 0.861,
|
| 563 |
+
"eval_steps_per_second": 0.861,
|
| 564 |
+
"step": 250
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"acc": 0.62165804,
|
| 568 |
+
"epoch": 0.6910569105691057,
|
| 569 |
+
"grad_norm": 1.6171875,
|
| 570 |
+
"learning_rate": 0.0001306348074179743,
|
| 571 |
+
"loss": 1.57510653,
|
| 572 |
+
"memory(GiB)": 15.58,
|
| 573 |
+
"step": 255,
|
| 574 |
+
"train_speed(iter/s)": 0.140805
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"acc": 0.59346585,
|
| 578 |
+
"epoch": 0.7046070460704607,
|
| 579 |
+
"grad_norm": 1.046875,
|
| 580 |
+
"learning_rate": 0.00013009985734664764,
|
| 581 |
+
"loss": 1.76646061,
|
| 582 |
+
"memory(GiB)": 15.58,
|
| 583 |
+
"step": 260,
|
| 584 |
+
"train_speed(iter/s)": 0.141165
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"acc": 0.60632119,
|
| 588 |
+
"epoch": 0.7181571815718157,
|
| 589 |
+
"grad_norm": 1.078125,
|
| 590 |
+
"learning_rate": 0.00012956490727532096,
|
| 591 |
+
"loss": 1.59940271,
|
| 592 |
+
"memory(GiB)": 15.95,
|
| 593 |
+
"step": 265,
|
| 594 |
+
"train_speed(iter/s)": 0.141502
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"acc": 0.59642992,
|
| 598 |
+
"epoch": 0.7317073170731707,
|
| 599 |
+
"grad_norm": 1.203125,
|
| 600 |
+
"learning_rate": 0.00012902995720399428,
|
| 601 |
+
"loss": 1.6549921,
|
| 602 |
+
"memory(GiB)": 15.95,
|
| 603 |
+
"step": 270,
|
| 604 |
+
"train_speed(iter/s)": 0.141855
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"acc": 0.60280037,
|
| 608 |
+
"epoch": 0.7452574525745257,
|
| 609 |
+
"grad_norm": 1.0625,
|
| 610 |
+
"learning_rate": 0.0001284950071326676,
|
| 611 |
+
"loss": 1.63179569,
|
| 612 |
+
"memory(GiB)": 15.95,
|
| 613 |
+
"step": 275,
|
| 614 |
+
"train_speed(iter/s)": 0.142164
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"acc": 0.594034,
|
| 618 |
+
"epoch": 0.7588075880758808,
|
| 619 |
+
"grad_norm": 1.1796875,
|
| 620 |
+
"learning_rate": 0.00012796005706134094,
|
| 621 |
+
"loss": 1.61152973,
|
| 622 |
+
"memory(GiB)": 15.95,
|
| 623 |
+
"step": 280,
|
| 624 |
+
"train_speed(iter/s)": 0.142474
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"acc": 0.59763761,
|
| 628 |
+
"epoch": 0.7723577235772358,
|
| 629 |
+
"grad_norm": 1.0859375,
|
| 630 |
+
"learning_rate": 0.00012742510699001425,
|
| 631 |
+
"loss": 1.66443863,
|
| 632 |
+
"memory(GiB)": 15.95,
|
| 633 |
+
"step": 285,
|
| 634 |
+
"train_speed(iter/s)": 0.142791
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"acc": 0.58316121,
|
| 638 |
+
"epoch": 0.7859078590785907,
|
| 639 |
+
"grad_norm": 0.73828125,
|
| 640 |
+
"learning_rate": 0.00012689015691868757,
|
| 641 |
+
"loss": 1.77028389,
|
| 642 |
+
"memory(GiB)": 15.95,
|
| 643 |
+
"step": 290,
|
| 644 |
+
"train_speed(iter/s)": 0.143067
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"acc": 0.60665183,
|
| 648 |
+
"epoch": 0.7994579945799458,
|
| 649 |
+
"grad_norm": 1.25,
|
| 650 |
+
"learning_rate": 0.00012635520684736091,
|
| 651 |
+
"loss": 1.64849663,
|
| 652 |
+
"memory(GiB)": 15.95,
|
| 653 |
+
"step": 295,
|
| 654 |
+
"train_speed(iter/s)": 0.143366
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"acc": 0.55526123,
|
| 658 |
+
"epoch": 0.8130081300813008,
|
| 659 |
+
"grad_norm": 1.203125,
|
| 660 |
+
"learning_rate": 0.00012582025677603423,
|
| 661 |
+
"loss": 1.86367321,
|
| 662 |
+
"memory(GiB)": 15.95,
|
| 663 |
+
"step": 300,
|
| 664 |
+
"train_speed(iter/s)": 0.143649
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 0.8130081300813008,
|
| 668 |
+
"eval_acc": 0.6129032258064516,
|
| 669 |
+
"eval_loss": 1.5865528583526611,
|
| 670 |
+
"eval_runtime": 44.0975,
|
| 671 |
+
"eval_samples_per_second": 0.862,
|
| 672 |
+
"eval_steps_per_second": 0.862,
|
| 673 |
+
"step": 300
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"acc": 0.6019598,
|
| 677 |
+
"epoch": 0.8265582655826558,
|
| 678 |
+
"grad_norm": 1.1328125,
|
| 679 |
+
"learning_rate": 0.00012528530670470755,
|
| 680 |
+
"loss": 1.71853142,
|
| 681 |
+
"memory(GiB)": 15.95,
|
| 682 |
+
"step": 305,
|
| 683 |
+
"train_speed(iter/s)": 0.14098
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"acc": 0.6009192,
|
| 687 |
+
"epoch": 0.8401084010840109,
|
| 688 |
+
"grad_norm": 1.359375,
|
| 689 |
+
"learning_rate": 0.00012475035663338086,
|
| 690 |
+
"loss": 1.72751312,
|
| 691 |
+
"memory(GiB)": 16.34,
|
| 692 |
+
"step": 310,
|
| 693 |
+
"train_speed(iter/s)": 0.141284
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"acc": 0.58907671,
|
| 697 |
+
"epoch": 0.8536585365853658,
|
| 698 |
+
"grad_norm": 0.98046875,
|
| 699 |
+
"learning_rate": 0.0001242154065620542,
|
| 700 |
+
"loss": 1.71860523,
|
| 701 |
+
"memory(GiB)": 16.34,
|
| 702 |
+
"step": 315,
|
| 703 |
+
"train_speed(iter/s)": 0.141562
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"acc": 0.62246222,
|
| 707 |
+
"epoch": 0.8672086720867209,
|
| 708 |
+
"grad_norm": 1.0234375,
|
| 709 |
+
"learning_rate": 0.00012368045649072752,
|
| 710 |
+
"loss": 1.58847914,
|
| 711 |
+
"memory(GiB)": 16.34,
|
| 712 |
+
"step": 320,
|
| 713 |
+
"train_speed(iter/s)": 0.141848
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"acc": 0.61927052,
|
| 717 |
+
"epoch": 0.8807588075880759,
|
| 718 |
+
"grad_norm": 1.59375,
|
| 719 |
+
"learning_rate": 0.00012314550641940084,
|
| 720 |
+
"loss": 1.54207745,
|
| 721 |
+
"memory(GiB)": 16.34,
|
| 722 |
+
"step": 325,
|
| 723 |
+
"train_speed(iter/s)": 0.142133
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"acc": 0.60672359,
|
| 727 |
+
"epoch": 0.8943089430894309,
|
| 728 |
+
"grad_norm": 2.140625,
|
| 729 |
+
"learning_rate": 0.00012261055634807416,
|
| 730 |
+
"loss": 1.71789646,
|
| 731 |
+
"memory(GiB)": 16.34,
|
| 732 |
+
"step": 330,
|
| 733 |
+
"train_speed(iter/s)": 0.142392
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"acc": 0.63249903,
|
| 737 |
+
"epoch": 0.907859078590786,
|
| 738 |
+
"grad_norm": 1.1328125,
|
| 739 |
+
"learning_rate": 0.0001220756062767475,
|
| 740 |
+
"loss": 1.46934195,
|
| 741 |
+
"memory(GiB)": 16.34,
|
| 742 |
+
"step": 335,
|
| 743 |
+
"train_speed(iter/s)": 0.142658
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"acc": 0.60973873,
|
| 747 |
+
"epoch": 0.9214092140921409,
|
| 748 |
+
"grad_norm": 0.84375,
|
| 749 |
+
"learning_rate": 0.00012154065620542082,
|
| 750 |
+
"loss": 1.67951736,
|
| 751 |
+
"memory(GiB)": 16.34,
|
| 752 |
+
"step": 340,
|
| 753 |
+
"train_speed(iter/s)": 0.142914
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"acc": 0.58657136,
|
| 757 |
+
"epoch": 0.9349593495934959,
|
| 758 |
+
"grad_norm": 1.2734375,
|
| 759 |
+
"learning_rate": 0.00012100570613409413,
|
| 760 |
+
"loss": 1.68692303,
|
| 761 |
+
"memory(GiB)": 16.34,
|
| 762 |
+
"step": 345,
|
| 763 |
+
"train_speed(iter/s)": 0.143167
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"acc": 0.59965162,
|
| 767 |
+
"epoch": 0.948509485094851,
|
| 768 |
+
"grad_norm": 1.421875,
|
| 769 |
+
"learning_rate": 0.00012047075606276746,
|
| 770 |
+
"loss": 1.71949959,
|
| 771 |
+
"memory(GiB)": 16.34,
|
| 772 |
+
"step": 350,
|
| 773 |
+
"train_speed(iter/s)": 0.143387
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.948509485094851,
|
| 777 |
+
"eval_acc": 0.6171254855598717,
|
| 778 |
+
"eval_loss": 1.562721610069275,
|
| 779 |
+
"eval_runtime": 44.0566,
|
| 780 |
+
"eval_samples_per_second": 0.863,
|
| 781 |
+
"eval_steps_per_second": 0.863,
|
| 782 |
+
"step": 350
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"acc": 0.59322553,
|
| 786 |
+
"epoch": 0.962059620596206,
|
| 787 |
+
"grad_norm": 1.3046875,
|
| 788 |
+
"learning_rate": 0.0001199358059914408,
|
| 789 |
+
"loss": 1.66712799,
|
| 790 |
+
"memory(GiB)": 16.34,
|
| 791 |
+
"step": 355,
|
| 792 |
+
"train_speed(iter/s)": 0.14111
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"acc": 0.59898105,
|
| 796 |
+
"epoch": 0.975609756097561,
|
| 797 |
+
"grad_norm": 1.265625,
|
| 798 |
+
"learning_rate": 0.00011940085592011411,
|
| 799 |
+
"loss": 1.66754684,
|
| 800 |
+
"memory(GiB)": 16.34,
|
| 801 |
+
"step": 360,
|
| 802 |
+
"train_speed(iter/s)": 0.141373
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"acc": 0.60211391,
|
| 806 |
+
"epoch": 0.989159891598916,
|
| 807 |
+
"grad_norm": 1.359375,
|
| 808 |
+
"learning_rate": 0.00011886590584878744,
|
| 809 |
+
"loss": 1.67542057,
|
| 810 |
+
"memory(GiB)": 16.34,
|
| 811 |
+
"step": 365,
|
| 812 |
+
"train_speed(iter/s)": 0.141638
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"acc": 0.63076615,
|
| 816 |
+
"epoch": 1.002710027100271,
|
| 817 |
+
"grad_norm": 0.8515625,
|
| 818 |
+
"learning_rate": 0.00011833095577746076,
|
| 819 |
+
"loss": 1.51196938,
|
| 820 |
+
"memory(GiB)": 16.34,
|
| 821 |
+
"step": 370,
|
| 822 |
+
"train_speed(iter/s)": 0.141355
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"acc": 0.63542643,
|
| 826 |
+
"epoch": 1.016260162601626,
|
| 827 |
+
"grad_norm": 0.828125,
|
| 828 |
+
"learning_rate": 0.00011779600570613407,
|
| 829 |
+
"loss": 1.38132334,
|
| 830 |
+
"memory(GiB)": 16.34,
|
| 831 |
+
"step": 375,
|
| 832 |
+
"train_speed(iter/s)": 0.141603
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"acc": 0.66358423,
|
| 836 |
+
"epoch": 1.029810298102981,
|
| 837 |
+
"grad_norm": 0.859375,
|
| 838 |
+
"learning_rate": 0.00011726105563480742,
|
| 839 |
+
"loss": 1.41512909,
|
| 840 |
+
"memory(GiB)": 16.34,
|
| 841 |
+
"step": 380,
|
| 842 |
+
"train_speed(iter/s)": 0.141845
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"acc": 0.63692493,
|
| 846 |
+
"epoch": 1.043360433604336,
|
| 847 |
+
"grad_norm": 1.0859375,
|
| 848 |
+
"learning_rate": 0.00011672610556348073,
|
| 849 |
+
"loss": 1.38078823,
|
| 850 |
+
"memory(GiB)": 16.34,
|
| 851 |
+
"step": 385,
|
| 852 |
+
"train_speed(iter/s)": 0.142082
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"acc": 0.6641499,
|
| 856 |
+
"epoch": 1.056910569105691,
|
| 857 |
+
"grad_norm": 1.125,
|
| 858 |
+
"learning_rate": 0.00011619115549215406,
|
| 859 |
+
"loss": 1.3458046,
|
| 860 |
+
"memory(GiB)": 16.34,
|
| 861 |
+
"step": 390,
|
| 862 |
+
"train_speed(iter/s)": 0.142305
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"acc": 0.67276783,
|
| 866 |
+
"epoch": 1.070460704607046,
|
| 867 |
+
"grad_norm": 0.98828125,
|
| 868 |
+
"learning_rate": 0.00011565620542082738,
|
| 869 |
+
"loss": 1.2076004,
|
| 870 |
+
"memory(GiB)": 16.34,
|
| 871 |
+
"step": 395,
|
| 872 |
+
"train_speed(iter/s)": 0.142533
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"acc": 0.65821433,
|
| 876 |
+
"epoch": 1.084010840108401,
|
| 877 |
+
"grad_norm": 1.15625,
|
| 878 |
+
"learning_rate": 0.0001151212553495007,
|
| 879 |
+
"loss": 1.38641891,
|
| 880 |
+
"memory(GiB)": 16.34,
|
| 881 |
+
"step": 400,
|
| 882 |
+
"train_speed(iter/s)": 0.142742
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 1.084010840108401,
|
| 886 |
+
"eval_acc": 0.6162810336091876,
|
| 887 |
+
"eval_loss": 1.5887514352798462,
|
| 888 |
+
"eval_runtime": 44.1624,
|
| 889 |
+
"eval_samples_per_second": 0.86,
|
| 890 |
+
"eval_steps_per_second": 0.86,
|
| 891 |
+
"step": 400
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"acc": 0.63352804,
|
| 895 |
+
"epoch": 1.0975609756097562,
|
| 896 |
+
"grad_norm": 1.2421875,
|
| 897 |
+
"learning_rate": 0.00011458630527817403,
|
| 898 |
+
"loss": 1.44696302,
|
| 899 |
+
"memory(GiB)": 16.34,
|
| 900 |
+
"step": 405,
|
| 901 |
+
"train_speed(iter/s)": 0.140759
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"acc": 0.61955671,
|
| 905 |
+
"epoch": 1.1111111111111112,
|
| 906 |
+
"grad_norm": 1.5546875,
|
| 907 |
+
"learning_rate": 0.00011405135520684734,
|
| 908 |
+
"loss": 1.48915672,
|
| 909 |
+
"memory(GiB)": 16.34,
|
| 910 |
+
"step": 410,
|
| 911 |
+
"train_speed(iter/s)": 0.140987
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"acc": 0.69259667,
|
| 915 |
+
"epoch": 1.1246612466124661,
|
| 916 |
+
"grad_norm": 0.796875,
|
| 917 |
+
"learning_rate": 0.00011351640513552069,
|
| 918 |
+
"loss": 1.22069292,
|
| 919 |
+
"memory(GiB)": 16.34,
|
| 920 |
+
"step": 415,
|
| 921 |
+
"train_speed(iter/s)": 0.141191
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"acc": 0.67346158,
|
| 925 |
+
"epoch": 1.1382113821138211,
|
| 926 |
+
"grad_norm": 1.171875,
|
| 927 |
+
"learning_rate": 0.000112981455064194,
|
| 928 |
+
"loss": 1.36795778,
|
| 929 |
+
"memory(GiB)": 16.34,
|
| 930 |
+
"step": 420,
|
| 931 |
+
"train_speed(iter/s)": 0.141417
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"acc": 0.67346702,
|
| 935 |
+
"epoch": 1.151761517615176,
|
| 936 |
+
"grad_norm": 1.0390625,
|
| 937 |
+
"learning_rate": 0.00011244650499286732,
|
| 938 |
+
"loss": 1.27111759,
|
| 939 |
+
"memory(GiB)": 16.34,
|
| 940 |
+
"step": 425,
|
| 941 |
+
"train_speed(iter/s)": 0.141634
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"acc": 0.61625972,
|
| 945 |
+
"epoch": 1.165311653116531,
|
| 946 |
+
"grad_norm": 1.2578125,
|
| 947 |
+
"learning_rate": 0.00011191155492154065,
|
| 948 |
+
"loss": 1.42474003,
|
| 949 |
+
"memory(GiB)": 16.34,
|
| 950 |
+
"step": 430,
|
| 951 |
+
"train_speed(iter/s)": 0.141853
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"acc": 0.65880041,
|
| 955 |
+
"epoch": 1.1788617886178863,
|
| 956 |
+
"grad_norm": 1.2578125,
|
| 957 |
+
"learning_rate": 0.00011137660485021397,
|
| 958 |
+
"loss": 1.33981524,
|
| 959 |
+
"memory(GiB)": 16.34,
|
| 960 |
+
"step": 435,
|
| 961 |
+
"train_speed(iter/s)": 0.142048
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"acc": 0.63503499,
|
| 965 |
+
"epoch": 1.1924119241192412,
|
| 966 |
+
"grad_norm": 1.2265625,
|
| 967 |
+
"learning_rate": 0.00011084165477888728,
|
| 968 |
+
"loss": 1.44122829,
|
| 969 |
+
"memory(GiB)": 16.34,
|
| 970 |
+
"step": 440,
|
| 971 |
+
"train_speed(iter/s)": 0.142254
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"acc": 0.61518383,
|
| 975 |
+
"epoch": 1.2059620596205962,
|
| 976 |
+
"grad_norm": 1.4765625,
|
| 977 |
+
"learning_rate": 0.00011030670470756061,
|
| 978 |
+
"loss": 1.54818697,
|
| 979 |
+
"memory(GiB)": 16.34,
|
| 980 |
+
"step": 445,
|
| 981 |
+
"train_speed(iter/s)": 0.142446
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"acc": 0.68010144,
|
| 985 |
+
"epoch": 1.2195121951219512,
|
| 986 |
+
"grad_norm": 1.3828125,
|
| 987 |
+
"learning_rate": 0.00010977175463623394,
|
| 988 |
+
"loss": 1.28560524,
|
| 989 |
+
"memory(GiB)": 16.34,
|
| 990 |
+
"step": 450,
|
| 991 |
+
"train_speed(iter/s)": 0.142639
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.2195121951219512,
|
| 995 |
+
"eval_acc": 0.6230366492146597,
|
| 996 |
+
"eval_loss": 1.5483555793762207,
|
| 997 |
+
"eval_runtime": 44.0588,
|
| 998 |
+
"eval_samples_per_second": 0.862,
|
| 999 |
+
"eval_steps_per_second": 0.862,
|
| 1000 |
+
"step": 450
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"acc": 0.64718337,
|
| 1004 |
+
"epoch": 1.2330623306233062,
|
| 1005 |
+
"grad_norm": 1.4375,
|
| 1006 |
+
"learning_rate": 0.00010923680456490727,
|
| 1007 |
+
"loss": 1.3778326,
|
| 1008 |
+
"memory(GiB)": 16.34,
|
| 1009 |
+
"step": 455,
|
| 1010 |
+
"train_speed(iter/s)": 0.140879
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"acc": 0.68637676,
|
| 1014 |
+
"epoch": 1.2466124661246614,
|
| 1015 |
+
"grad_norm": 1.453125,
|
| 1016 |
+
"learning_rate": 0.00010870185449358059,
|
| 1017 |
+
"loss": 1.21860657,
|
| 1018 |
+
"memory(GiB)": 16.34,
|
| 1019 |
+
"step": 460,
|
| 1020 |
+
"train_speed(iter/s)": 0.141097
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"acc": 0.63144946,
|
| 1024 |
+
"epoch": 1.2601626016260163,
|
| 1025 |
+
"grad_norm": 1.0703125,
|
| 1026 |
+
"learning_rate": 0.00010816690442225392,
|
| 1027 |
+
"loss": 1.38577023,
|
| 1028 |
+
"memory(GiB)": 16.34,
|
| 1029 |
+
"step": 465,
|
| 1030 |
+
"train_speed(iter/s)": 0.141294
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"acc": 0.66895781,
|
| 1034 |
+
"epoch": 1.2737127371273713,
|
| 1035 |
+
"grad_norm": 1.296875,
|
| 1036 |
+
"learning_rate": 0.00010763195435092724,
|
| 1037 |
+
"loss": 1.29167385,
|
| 1038 |
+
"memory(GiB)": 16.34,
|
| 1039 |
+
"step": 470,
|
| 1040 |
+
"train_speed(iter/s)": 0.141488
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"acc": 0.6553885,
|
| 1044 |
+
"epoch": 1.2872628726287263,
|
| 1045 |
+
"grad_norm": 1.1484375,
|
| 1046 |
+
"learning_rate": 0.00010709700427960055,
|
| 1047 |
+
"loss": 1.35011473,
|
| 1048 |
+
"memory(GiB)": 16.34,
|
| 1049 |
+
"step": 475,
|
| 1050 |
+
"train_speed(iter/s)": 0.141687
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"acc": 0.65576148,
|
| 1054 |
+
"epoch": 1.3008130081300813,
|
| 1055 |
+
"grad_norm": 1.1796875,
|
| 1056 |
+
"learning_rate": 0.00010656205420827388,
|
| 1057 |
+
"loss": 1.39116755,
|
| 1058 |
+
"memory(GiB)": 16.34,
|
| 1059 |
+
"step": 480,
|
| 1060 |
+
"train_speed(iter/s)": 0.141863
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"acc": 0.62025633,
|
| 1064 |
+
"epoch": 1.3143631436314362,
|
| 1065 |
+
"grad_norm": 1.375,
|
| 1066 |
+
"learning_rate": 0.00010602710413694721,
|
| 1067 |
+
"loss": 1.55190315,
|
| 1068 |
+
"memory(GiB)": 16.34,
|
| 1069 |
+
"step": 485,
|
| 1070 |
+
"train_speed(iter/s)": 0.142058
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"acc": 0.65667233,
|
| 1074 |
+
"epoch": 1.3279132791327912,
|
| 1075 |
+
"grad_norm": 1.1484375,
|
| 1076 |
+
"learning_rate": 0.00010549215406562054,
|
| 1077 |
+
"loss": 1.30497828,
|
| 1078 |
+
"memory(GiB)": 16.34,
|
| 1079 |
+
"step": 490,
|
| 1080 |
+
"train_speed(iter/s)": 0.142237
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"acc": 0.6458045,
|
| 1084 |
+
"epoch": 1.3414634146341464,
|
| 1085 |
+
"grad_norm": 2.421875,
|
| 1086 |
+
"learning_rate": 0.00010495720399429386,
|
| 1087 |
+
"loss": 1.40175285,
|
| 1088 |
+
"memory(GiB)": 16.34,
|
| 1089 |
+
"step": 495,
|
| 1090 |
+
"train_speed(iter/s)": 0.14241
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"acc": 0.65004997,
|
| 1094 |
+
"epoch": 1.3550135501355014,
|
| 1095 |
+
"grad_norm": 1.265625,
|
| 1096 |
+
"learning_rate": 0.00010442225392296718,
|
| 1097 |
+
"loss": 1.31625471,
|
| 1098 |
+
"memory(GiB)": 16.34,
|
| 1099 |
+
"step": 500,
|
| 1100 |
+
"train_speed(iter/s)": 0.142583
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 1.3550135501355014,
|
| 1104 |
+
"eval_acc": 0.61932106063165,
|
| 1105 |
+
"eval_loss": 1.535814642906189,
|
| 1106 |
+
"eval_runtime": 44.07,
|
| 1107 |
+
"eval_samples_per_second": 0.862,
|
| 1108 |
+
"eval_steps_per_second": 0.862,
|
| 1109 |
+
"step": 500
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"acc": 0.65512385,
|
| 1113 |
+
"epoch": 1.3685636856368564,
|
| 1114 |
+
"grad_norm": 0.9375,
|
| 1115 |
+
"learning_rate": 0.00010388730385164051,
|
| 1116 |
+
"loss": 1.2989893,
|
| 1117 |
+
"memory(GiB)": 16.34,
|
| 1118 |
+
"step": 505,
|
| 1119 |
+
"train_speed(iter/s)": 0.140995
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"acc": 0.65444102,
|
| 1123 |
+
"epoch": 1.3821138211382114,
|
| 1124 |
+
"grad_norm": 1.5625,
|
| 1125 |
+
"learning_rate": 0.00010335235378031382,
|
| 1126 |
+
"loss": 1.32140932,
|
| 1127 |
+
"memory(GiB)": 16.34,
|
| 1128 |
+
"step": 510,
|
| 1129 |
+
"train_speed(iter/s)": 0.141172
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"acc": 0.66242504,
|
| 1133 |
+
"epoch": 1.3956639566395663,
|
| 1134 |
+
"grad_norm": 0.9140625,
|
| 1135 |
+
"learning_rate": 0.00010281740370898714,
|
| 1136 |
+
"loss": 1.34259834,
|
| 1137 |
+
"memory(GiB)": 16.34,
|
| 1138 |
+
"step": 515,
|
| 1139 |
+
"train_speed(iter/s)": 0.141336
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"acc": 0.62716827,
|
| 1143 |
+
"epoch": 1.4092140921409215,
|
| 1144 |
+
"grad_norm": 1.4296875,
|
| 1145 |
+
"learning_rate": 0.00010228245363766048,
|
| 1146 |
+
"loss": 1.46508284,
|
| 1147 |
+
"memory(GiB)": 16.34,
|
| 1148 |
+
"step": 520,
|
| 1149 |
+
"train_speed(iter/s)": 0.141508
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"acc": 0.61702657,
|
| 1153 |
+
"epoch": 1.4227642276422765,
|
| 1154 |
+
"grad_norm": 1.5546875,
|
| 1155 |
+
"learning_rate": 0.0001017475035663338,
|
| 1156 |
+
"loss": 1.43188276,
|
| 1157 |
+
"memory(GiB)": 16.34,
|
| 1158 |
+
"step": 525,
|
| 1159 |
+
"train_speed(iter/s)": 0.141681
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"acc": 0.63238263,
|
| 1163 |
+
"epoch": 1.4363143631436315,
|
| 1164 |
+
"grad_norm": 1.3671875,
|
| 1165 |
+
"learning_rate": 0.00010121255349500713,
|
| 1166 |
+
"loss": 1.52214499,
|
| 1167 |
+
"memory(GiB)": 16.34,
|
| 1168 |
+
"step": 530,
|
| 1169 |
+
"train_speed(iter/s)": 0.141854
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"acc": 0.6323854,
|
| 1173 |
+
"epoch": 1.4498644986449865,
|
| 1174 |
+
"grad_norm": 1.375,
|
| 1175 |
+
"learning_rate": 0.00010067760342368045,
|
| 1176 |
+
"loss": 1.38910236,
|
| 1177 |
+
"memory(GiB)": 16.34,
|
| 1178 |
+
"step": 535,
|
| 1179 |
+
"train_speed(iter/s)": 0.142019
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"acc": 0.60772176,
|
| 1183 |
+
"epoch": 1.4634146341463414,
|
| 1184 |
+
"grad_norm": 1.21875,
|
| 1185 |
+
"learning_rate": 0.00010014265335235376,
|
| 1186 |
+
"loss": 1.46768923,
|
| 1187 |
+
"memory(GiB)": 16.34,
|
| 1188 |
+
"step": 540,
|
| 1189 |
+
"train_speed(iter/s)": 0.14218
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"acc": 0.67315965,
|
| 1193 |
+
"epoch": 1.4769647696476964,
|
| 1194 |
+
"grad_norm": 1.34375,
|
| 1195 |
+
"learning_rate": 9.96077032810271e-05,
|
| 1196 |
+
"loss": 1.25136938,
|
| 1197 |
+
"memory(GiB)": 16.34,
|
| 1198 |
+
"step": 545,
|
| 1199 |
+
"train_speed(iter/s)": 0.142338
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"acc": 0.66408758,
|
| 1203 |
+
"epoch": 1.4905149051490514,
|
| 1204 |
+
"grad_norm": 2.046875,
|
| 1205 |
+
"learning_rate": 9.907275320970041e-05,
|
| 1206 |
+
"loss": 1.25655928,
|
| 1207 |
+
"memory(GiB)": 16.34,
|
| 1208 |
+
"step": 550,
|
| 1209 |
+
"train_speed(iter/s)": 0.142495
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 1.4905149051490514,
|
| 1213 |
+
"eval_acc": 0.6254011146765749,
|
| 1214 |
+
"eval_loss": 1.5346177816390991,
|
| 1215 |
+
"eval_runtime": 44.1805,
|
| 1216 |
+
"eval_samples_per_second": 0.86,
|
| 1217 |
+
"eval_steps_per_second": 0.86,
|
| 1218 |
+
"step": 550
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"acc": 0.68655195,
|
| 1222 |
+
"epoch": 1.5040650406504064,
|
| 1223 |
+
"grad_norm": 1.125,
|
| 1224 |
+
"learning_rate": 9.853780313837375e-05,
|
| 1225 |
+
"loss": 1.08906736,
|
| 1226 |
+
"memory(GiB)": 16.34,
|
| 1227 |
+
"step": 555,
|
| 1228 |
+
"train_speed(iter/s)": 0.141037
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"acc": 0.65898876,
|
| 1232 |
+
"epoch": 1.5176151761517616,
|
| 1233 |
+
"grad_norm": 1.4296875,
|
| 1234 |
+
"learning_rate": 9.800285306704707e-05,
|
| 1235 |
+
"loss": 1.26696377,
|
| 1236 |
+
"memory(GiB)": 16.34,
|
| 1237 |
+
"step": 560,
|
| 1238 |
+
"train_speed(iter/s)": 0.14119
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"acc": 0.63674688,
|
| 1242 |
+
"epoch": 1.5311653116531165,
|
| 1243 |
+
"grad_norm": 1.5390625,
|
| 1244 |
+
"learning_rate": 9.746790299572039e-05,
|
| 1245 |
+
"loss": 1.38277016,
|
| 1246 |
+
"memory(GiB)": 16.34,
|
| 1247 |
+
"step": 565,
|
| 1248 |
+
"train_speed(iter/s)": 0.141349
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"acc": 0.67204466,
|
| 1252 |
+
"epoch": 1.5447154471544715,
|
| 1253 |
+
"grad_norm": 1.234375,
|
| 1254 |
+
"learning_rate": 9.693295292439372e-05,
|
| 1255 |
+
"loss": 1.25243311,
|
| 1256 |
+
"memory(GiB)": 16.34,
|
| 1257 |
+
"step": 570,
|
| 1258 |
+
"train_speed(iter/s)": 0.141502
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"acc": 0.67878027,
|
| 1262 |
+
"epoch": 1.5582655826558267,
|
| 1263 |
+
"grad_norm": 1.3828125,
|
| 1264 |
+
"learning_rate": 9.639800285306703e-05,
|
| 1265 |
+
"loss": 1.18253031,
|
| 1266 |
+
"memory(GiB)": 16.34,
|
| 1267 |
+
"step": 575,
|
| 1268 |
+
"train_speed(iter/s)": 0.141663
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"acc": 0.64645357,
|
| 1272 |
+
"epoch": 1.5718157181571817,
|
| 1273 |
+
"grad_norm": 1.671875,
|
| 1274 |
+
"learning_rate": 9.586305278174036e-05,
|
| 1275 |
+
"loss": 1.39789114,
|
| 1276 |
+
"memory(GiB)": 16.34,
|
| 1277 |
+
"step": 580,
|
| 1278 |
+
"train_speed(iter/s)": 0.141812
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"acc": 0.63896599,
|
| 1282 |
+
"epoch": 1.5853658536585367,
|
| 1283 |
+
"grad_norm": 1.5546875,
|
| 1284 |
+
"learning_rate": 9.532810271041368e-05,
|
| 1285 |
+
"loss": 1.37434454,
|
| 1286 |
+
"memory(GiB)": 16.34,
|
| 1287 |
+
"step": 585,
|
| 1288 |
+
"train_speed(iter/s)": 0.141961
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"acc": 0.67930498,
|
| 1292 |
+
"epoch": 1.5989159891598916,
|
| 1293 |
+
"grad_norm": 1.3671875,
|
| 1294 |
+
"learning_rate": 9.479315263908701e-05,
|
| 1295 |
+
"loss": 1.24675446,
|
| 1296 |
+
"memory(GiB)": 16.34,
|
| 1297 |
+
"step": 590,
|
| 1298 |
+
"train_speed(iter/s)": 0.142113
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"acc": 0.66582651,
|
| 1302 |
+
"epoch": 1.6124661246612466,
|
| 1303 |
+
"grad_norm": 1.1171875,
|
| 1304 |
+
"learning_rate": 9.425820256776034e-05,
|
| 1305 |
+
"loss": 1.33937092,
|
| 1306 |
+
"memory(GiB)": 16.34,
|
| 1307 |
+
"step": 595,
|
| 1308 |
+
"train_speed(iter/s)": 0.142256
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"acc": 0.65553112,
|
| 1312 |
+
"epoch": 1.6260162601626016,
|
| 1313 |
+
"grad_norm": 1.5859375,
|
| 1314 |
+
"learning_rate": 9.372325249643366e-05,
|
| 1315 |
+
"loss": 1.25127707,
|
| 1316 |
+
"memory(GiB)": 16.34,
|
| 1317 |
+
"step": 600,
|
| 1318 |
+
"train_speed(iter/s)": 0.142394
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.6260162601626016,
|
| 1322 |
+
"eval_acc": 0.6233744299949333,
|
| 1323 |
+
"eval_loss": 1.5426762104034424,
|
| 1324 |
+
"eval_runtime": 44.3564,
|
| 1325 |
+
"eval_samples_per_second": 0.857,
|
| 1326 |
+
"eval_steps_per_second": 0.857,
|
| 1327 |
+
"step": 600
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"acc": 0.64805899,
|
| 1331 |
+
"epoch": 1.6395663956639566,
|
| 1332 |
+
"grad_norm": 1.078125,
|
| 1333 |
+
"learning_rate": 9.318830242510699e-05,
|
| 1334 |
+
"loss": 1.30654621,
|
| 1335 |
+
"memory(GiB)": 16.34,
|
| 1336 |
+
"step": 605,
|
| 1337 |
+
"train_speed(iter/s)": 0.141073
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"acc": 0.67441335,
|
| 1341 |
+
"epoch": 1.6531165311653115,
|
| 1342 |
+
"grad_norm": 1.34375,
|
| 1343 |
+
"learning_rate": 9.26533523537803e-05,
|
| 1344 |
+
"loss": 1.20785751,
|
| 1345 |
+
"memory(GiB)": 16.34,
|
| 1346 |
+
"step": 610,
|
| 1347 |
+
"train_speed(iter/s)": 0.141227
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"acc": 0.64989614,
|
| 1351 |
+
"epoch": 1.6666666666666665,
|
| 1352 |
+
"grad_norm": 2.25,
|
| 1353 |
+
"learning_rate": 9.211840228245362e-05,
|
| 1354 |
+
"loss": 1.41231976,
|
| 1355 |
+
"memory(GiB)": 16.34,
|
| 1356 |
+
"step": 615,
|
| 1357 |
+
"train_speed(iter/s)": 0.141374
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"acc": 0.65118213,
|
| 1361 |
+
"epoch": 1.6802168021680217,
|
| 1362 |
+
"grad_norm": 1.34375,
|
| 1363 |
+
"learning_rate": 9.158345221112695e-05,
|
| 1364 |
+
"loss": 1.33156195,
|
| 1365 |
+
"memory(GiB)": 16.34,
|
| 1366 |
+
"step": 620,
|
| 1367 |
+
"train_speed(iter/s)": 0.141527
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"acc": 0.68491917,
|
| 1371 |
+
"epoch": 1.6937669376693767,
|
| 1372 |
+
"grad_norm": 2.078125,
|
| 1373 |
+
"learning_rate": 9.104850213980028e-05,
|
| 1374 |
+
"loss": 1.15608921,
|
| 1375 |
+
"memory(GiB)": 16.34,
|
| 1376 |
+
"step": 625,
|
| 1377 |
+
"train_speed(iter/s)": 0.141656
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"acc": 0.6636075,
|
| 1381 |
+
"epoch": 1.7073170731707317,
|
| 1382 |
+
"grad_norm": 1.2890625,
|
| 1383 |
+
"learning_rate": 9.051355206847361e-05,
|
| 1384 |
+
"loss": 1.3377409,
|
| 1385 |
+
"memory(GiB)": 16.34,
|
| 1386 |
+
"step": 630,
|
| 1387 |
+
"train_speed(iter/s)": 0.141795
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"acc": 0.66115265,
|
| 1391 |
+
"epoch": 1.7208672086720869,
|
| 1392 |
+
"grad_norm": 1.796875,
|
| 1393 |
+
"learning_rate": 8.997860199714693e-05,
|
| 1394 |
+
"loss": 1.41666918,
|
| 1395 |
+
"memory(GiB)": 16.34,
|
| 1396 |
+
"step": 635,
|
| 1397 |
+
"train_speed(iter/s)": 0.141928
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"acc": 0.65316691,
|
| 1401 |
+
"epoch": 1.7344173441734418,
|
| 1402 |
+
"grad_norm": 1.875,
|
| 1403 |
+
"learning_rate": 8.944365192582024e-05,
|
| 1404 |
+
"loss": 1.37885714,
|
| 1405 |
+
"memory(GiB)": 16.34,
|
| 1406 |
+
"step": 640,
|
| 1407 |
+
"train_speed(iter/s)": 0.14206
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"acc": 0.63365035,
|
| 1411 |
+
"epoch": 1.7479674796747968,
|
| 1412 |
+
"grad_norm": 2.03125,
|
| 1413 |
+
"learning_rate": 8.890870185449357e-05,
|
| 1414 |
+
"loss": 1.43597651,
|
| 1415 |
+
"memory(GiB)": 16.34,
|
| 1416 |
+
"step": 645,
|
| 1417 |
+
"train_speed(iter/s)": 0.142195
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"acc": 0.6461009,
|
| 1421 |
+
"epoch": 1.7615176151761518,
|
| 1422 |
+
"grad_norm": 1.390625,
|
| 1423 |
+
"learning_rate": 8.837375178316689e-05,
|
| 1424 |
+
"loss": 1.40720987,
|
| 1425 |
+
"memory(GiB)": 16.34,
|
| 1426 |
+
"step": 650,
|
| 1427 |
+
"train_speed(iter/s)": 0.142326
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.7615176151761518,
|
| 1431 |
+
"eval_acc": 0.6223610876541125,
|
| 1432 |
+
"eval_loss": 1.5279418230056763,
|
| 1433 |
+
"eval_runtime": 44.3195,
|
| 1434 |
+
"eval_samples_per_second": 0.857,
|
| 1435 |
+
"eval_steps_per_second": 0.857,
|
| 1436 |
+
"step": 650
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"acc": 0.68952079,
|
| 1440 |
+
"epoch": 1.7750677506775068,
|
| 1441 |
+
"grad_norm": 1.5859375,
|
| 1442 |
+
"learning_rate": 8.783880171184023e-05,
|
| 1443 |
+
"loss": 1.1804883,
|
| 1444 |
+
"memory(GiB)": 16.34,
|
| 1445 |
+
"step": 655,
|
| 1446 |
+
"train_speed(iter/s)": 0.141093
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"acc": 0.63278737,
|
| 1450 |
+
"epoch": 1.7886178861788617,
|
| 1451 |
+
"grad_norm": 1.3125,
|
| 1452 |
+
"learning_rate": 8.730385164051355e-05,
|
| 1453 |
+
"loss": 1.45864544,
|
| 1454 |
+
"memory(GiB)": 16.34,
|
| 1455 |
+
"step": 660,
|
| 1456 |
+
"train_speed(iter/s)": 0.141227
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"acc": 0.6499536,
|
| 1460 |
+
"epoch": 1.8021680216802167,
|
| 1461 |
+
"grad_norm": 1.0859375,
|
| 1462 |
+
"learning_rate": 8.676890156918687e-05,
|
| 1463 |
+
"loss": 1.36629667,
|
| 1464 |
+
"memory(GiB)": 16.34,
|
| 1465 |
+
"step": 665,
|
| 1466 |
+
"train_speed(iter/s)": 0.14135
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"acc": 0.66262636,
|
| 1470 |
+
"epoch": 1.8157181571815717,
|
| 1471 |
+
"grad_norm": 2.8125,
|
| 1472 |
+
"learning_rate": 8.62339514978602e-05,
|
| 1473 |
+
"loss": 1.28030796,
|
| 1474 |
+
"memory(GiB)": 16.34,
|
| 1475 |
+
"step": 670,
|
| 1476 |
+
"train_speed(iter/s)": 0.141486
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"acc": 0.6478013,
|
| 1480 |
+
"epoch": 1.8292682926829267,
|
| 1481 |
+
"grad_norm": 1.3203125,
|
| 1482 |
+
"learning_rate": 8.569900142653351e-05,
|
| 1483 |
+
"loss": 1.37910089,
|
| 1484 |
+
"memory(GiB)": 16.34,
|
| 1485 |
+
"step": 675,
|
| 1486 |
+
"train_speed(iter/s)": 0.141617
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"acc": 0.65061078,
|
| 1490 |
+
"epoch": 1.8428184281842819,
|
| 1491 |
+
"grad_norm": 1.859375,
|
| 1492 |
+
"learning_rate": 8.516405135520683e-05,
|
| 1493 |
+
"loss": 1.24110394,
|
| 1494 |
+
"memory(GiB)": 16.34,
|
| 1495 |
+
"step": 680,
|
| 1496 |
+
"train_speed(iter/s)": 0.141749
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"acc": 0.66720443,
|
| 1500 |
+
"epoch": 1.8563685636856369,
|
| 1501 |
+
"grad_norm": 1.2734375,
|
| 1502 |
+
"learning_rate": 8.462910128388016e-05,
|
| 1503 |
+
"loss": 1.36949673,
|
| 1504 |
+
"memory(GiB)": 16.34,
|
| 1505 |
+
"step": 685,
|
| 1506 |
+
"train_speed(iter/s)": 0.141882
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"acc": 0.65051932,
|
| 1510 |
+
"epoch": 1.8699186991869918,
|
| 1511 |
+
"grad_norm": 1.796875,
|
| 1512 |
+
"learning_rate": 8.409415121255349e-05,
|
| 1513 |
+
"loss": 1.3470686,
|
| 1514 |
+
"memory(GiB)": 16.34,
|
| 1515 |
+
"step": 690,
|
| 1516 |
+
"train_speed(iter/s)": 0.142007
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"acc": 0.64647999,
|
| 1520 |
+
"epoch": 1.883468834688347,
|
| 1521 |
+
"grad_norm": 1.0546875,
|
| 1522 |
+
"learning_rate": 8.355920114122682e-05,
|
| 1523 |
+
"loss": 1.27561255,
|
| 1524 |
+
"memory(GiB)": 16.34,
|
| 1525 |
+
"step": 695,
|
| 1526 |
+
"train_speed(iter/s)": 0.142124
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"acc": 0.64771528,
|
| 1530 |
+
"epoch": 1.897018970189702,
|
| 1531 |
+
"grad_norm": 2.5,
|
| 1532 |
+
"learning_rate": 8.302425106990014e-05,
|
| 1533 |
+
"loss": 1.3874403,
|
| 1534 |
+
"memory(GiB)": 16.34,
|
| 1535 |
+
"step": 700,
|
| 1536 |
+
"train_speed(iter/s)": 0.142242
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.897018970189702,
|
| 1540 |
+
"eval_acc": 0.6274277993582165,
|
| 1541 |
+
"eval_loss": 1.5236802101135254,
|
| 1542 |
+
"eval_runtime": 44.4112,
|
| 1543 |
+
"eval_samples_per_second": 0.856,
|
| 1544 |
+
"eval_steps_per_second": 0.856,
|
| 1545 |
+
"step": 700
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"acc": 0.65817127,
|
| 1549 |
+
"epoch": 1.910569105691057,
|
| 1550 |
+
"grad_norm": 1.3203125,
|
| 1551 |
+
"learning_rate": 8.248930099857345e-05,
|
| 1552 |
+
"loss": 1.29632025,
|
| 1553 |
+
"memory(GiB)": 16.34,
|
| 1554 |
+
"step": 705,
|
| 1555 |
+
"train_speed(iter/s)": 0.141099
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"acc": 0.64647436,
|
| 1559 |
+
"epoch": 1.924119241192412,
|
| 1560 |
+
"grad_norm": 1.234375,
|
| 1561 |
+
"learning_rate": 8.195435092724678e-05,
|
| 1562 |
+
"loss": 1.39382238,
|
| 1563 |
+
"memory(GiB)": 16.34,
|
| 1564 |
+
"step": 710,
|
| 1565 |
+
"train_speed(iter/s)": 0.141217
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"acc": 0.65741391,
|
| 1569 |
+
"epoch": 1.937669376693767,
|
| 1570 |
+
"grad_norm": 0.98828125,
|
| 1571 |
+
"learning_rate": 8.14194008559201e-05,
|
| 1572 |
+
"loss": 1.32606802,
|
| 1573 |
+
"memory(GiB)": 16.34,
|
| 1574 |
+
"step": 715,
|
| 1575 |
+
"train_speed(iter/s)": 0.141337
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"acc": 0.65078535,
|
| 1579 |
+
"epoch": 1.951219512195122,
|
| 1580 |
+
"grad_norm": 1.6171875,
|
| 1581 |
+
"learning_rate": 8.088445078459343e-05,
|
| 1582 |
+
"loss": 1.28092451,
|
| 1583 |
+
"memory(GiB)": 16.34,
|
| 1584 |
+
"step": 720,
|
| 1585 |
+
"train_speed(iter/s)": 0.141457
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"acc": 0.64983764,
|
| 1589 |
+
"epoch": 1.9647696476964769,
|
| 1590 |
+
"grad_norm": 1.7890625,
|
| 1591 |
+
"learning_rate": 8.034950071326676e-05,
|
| 1592 |
+
"loss": 1.35295801,
|
| 1593 |
+
"memory(GiB)": 16.34,
|
| 1594 |
+
"step": 725,
|
| 1595 |
+
"train_speed(iter/s)": 0.141585
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"acc": 0.64880919,
|
| 1599 |
+
"epoch": 1.9783197831978319,
|
| 1600 |
+
"grad_norm": 1.6640625,
|
| 1601 |
+
"learning_rate": 7.981455064194009e-05,
|
| 1602 |
+
"loss": 1.38945732,
|
| 1603 |
+
"memory(GiB)": 16.34,
|
| 1604 |
+
"step": 730,
|
| 1605 |
+
"train_speed(iter/s)": 0.141703
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"acc": 0.64617214,
|
| 1609 |
+
"epoch": 1.9918699186991868,
|
| 1610 |
+
"grad_norm": 1.3671875,
|
| 1611 |
+
"learning_rate": 7.927960057061341e-05,
|
| 1612 |
+
"loss": 1.31189098,
|
| 1613 |
+
"memory(GiB)": 16.34,
|
| 1614 |
+
"step": 735,
|
| 1615 |
+
"train_speed(iter/s)": 0.141817
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"acc": 0.67255039,
|
| 1619 |
+
"epoch": 2.005420054200542,
|
| 1620 |
+
"grad_norm": 1.3359375,
|
| 1621 |
+
"learning_rate": 7.874465049928672e-05,
|
| 1622 |
+
"loss": 1.17746153,
|
| 1623 |
+
"memory(GiB)": 16.34,
|
| 1624 |
+
"step": 740,
|
| 1625 |
+
"train_speed(iter/s)": 0.141669
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"acc": 0.69897633,
|
| 1629 |
+
"epoch": 2.0189701897018972,
|
| 1630 |
+
"grad_norm": 1.1875,
|
| 1631 |
+
"learning_rate": 7.820970042796005e-05,
|
| 1632 |
+
"loss": 1.03704672,
|
| 1633 |
+
"memory(GiB)": 16.34,
|
| 1634 |
+
"step": 745,
|
| 1635 |
+
"train_speed(iter/s)": 0.141785
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"acc": 0.71377811,
|
| 1639 |
+
"epoch": 2.032520325203252,
|
| 1640 |
+
"grad_norm": 1.640625,
|
| 1641 |
+
"learning_rate": 7.767475035663337e-05,
|
| 1642 |
+
"loss": 1.04516726,
|
| 1643 |
+
"memory(GiB)": 16.34,
|
| 1644 |
+
"step": 750,
|
| 1645 |
+
"train_speed(iter/s)": 0.141905
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 2.032520325203252,
|
| 1649 |
+
"eval_acc": 0.6208410741428813,
|
| 1650 |
+
"eval_loss": 1.603255271911621,
|
| 1651 |
+
"eval_runtime": 44.4359,
|
| 1652 |
+
"eval_samples_per_second": 0.855,
|
| 1653 |
+
"eval_steps_per_second": 0.855,
|
| 1654 |
+
"step": 750
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"acc": 0.73705945,
|
| 1658 |
+
"epoch": 2.046070460704607,
|
| 1659 |
+
"grad_norm": 1.8203125,
|
| 1660 |
+
"learning_rate": 7.713980028530669e-05,
|
| 1661 |
+
"loss": 0.9383255,
|
| 1662 |
+
"memory(GiB)": 16.34,
|
| 1663 |
+
"step": 755,
|
| 1664 |
+
"train_speed(iter/s)": 0.14084
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"acc": 0.73054934,
|
| 1668 |
+
"epoch": 2.059620596205962,
|
| 1669 |
+
"grad_norm": 1.8203125,
|
| 1670 |
+
"learning_rate": 7.660485021398003e-05,
|
| 1671 |
+
"loss": 0.91379232,
|
| 1672 |
+
"memory(GiB)": 16.34,
|
| 1673 |
+
"step": 760,
|
| 1674 |
+
"train_speed(iter/s)": 0.140958
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"acc": 0.73154573,
|
| 1678 |
+
"epoch": 2.073170731707317,
|
| 1679 |
+
"grad_norm": 1.6796875,
|
| 1680 |
+
"learning_rate": 7.606990014265335e-05,
|
| 1681 |
+
"loss": 0.97700481,
|
| 1682 |
+
"memory(GiB)": 16.34,
|
| 1683 |
+
"step": 765,
|
| 1684 |
+
"train_speed(iter/s)": 0.141076
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"acc": 0.72998781,
|
| 1688 |
+
"epoch": 2.086720867208672,
|
| 1689 |
+
"grad_norm": 1.4140625,
|
| 1690 |
+
"learning_rate": 7.553495007132668e-05,
|
| 1691 |
+
"loss": 0.93853807,
|
| 1692 |
+
"memory(GiB)": 16.34,
|
| 1693 |
+
"step": 770,
|
| 1694 |
+
"train_speed(iter/s)": 0.141186
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"acc": 0.73213534,
|
| 1698 |
+
"epoch": 2.100271002710027,
|
| 1699 |
+
"grad_norm": 1.78125,
|
| 1700 |
+
"learning_rate": 7.5e-05,
|
| 1701 |
+
"loss": 0.94589176,
|
| 1702 |
+
"memory(GiB)": 16.34,
|
| 1703 |
+
"step": 775,
|
| 1704 |
+
"train_speed(iter/s)": 0.141303
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"acc": 0.73035493,
|
| 1708 |
+
"epoch": 2.113821138211382,
|
| 1709 |
+
"grad_norm": 1.6953125,
|
| 1710 |
+
"learning_rate": 7.446504992867331e-05,
|
| 1711 |
+
"loss": 0.95597754,
|
| 1712 |
+
"memory(GiB)": 16.34,
|
| 1713 |
+
"step": 780,
|
| 1714 |
+
"train_speed(iter/s)": 0.141422
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"acc": 0.73428354,
|
| 1718 |
+
"epoch": 2.127371273712737,
|
| 1719 |
+
"grad_norm": 1.484375,
|
| 1720 |
+
"learning_rate": 7.393009985734664e-05,
|
| 1721 |
+
"loss": 0.98254423,
|
| 1722 |
+
"memory(GiB)": 16.34,
|
| 1723 |
+
"step": 785,
|
| 1724 |
+
"train_speed(iter/s)": 0.141543
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"acc": 0.69750729,
|
| 1728 |
+
"epoch": 2.140921409214092,
|
| 1729 |
+
"grad_norm": 1.828125,
|
| 1730 |
+
"learning_rate": 7.339514978601997e-05,
|
| 1731 |
+
"loss": 1.10206041,
|
| 1732 |
+
"memory(GiB)": 16.34,
|
| 1733 |
+
"step": 790,
|
| 1734 |
+
"train_speed(iter/s)": 0.141651
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"acc": 0.73764381,
|
| 1738 |
+
"epoch": 2.154471544715447,
|
| 1739 |
+
"grad_norm": 1.296875,
|
| 1740 |
+
"learning_rate": 7.286019971469329e-05,
|
| 1741 |
+
"loss": 0.84075432,
|
| 1742 |
+
"memory(GiB)": 16.34,
|
| 1743 |
+
"step": 795,
|
| 1744 |
+
"train_speed(iter/s)": 0.141765
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"acc": 0.72124152,
|
| 1748 |
+
"epoch": 2.168021680216802,
|
| 1749 |
+
"grad_norm": 1.4375,
|
| 1750 |
+
"learning_rate": 7.23252496433666e-05,
|
| 1751 |
+
"loss": 1.05863771,
|
| 1752 |
+
"memory(GiB)": 16.34,
|
| 1753 |
+
"step": 800,
|
| 1754 |
+
"train_speed(iter/s)": 0.141874
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 2.168021680216802,
|
| 1758 |
+
"eval_acc": 0.6240499915554805,
|
| 1759 |
+
"eval_loss": 1.61227285861969,
|
| 1760 |
+
"eval_runtime": 44.3145,
|
| 1761 |
+
"eval_samples_per_second": 0.858,
|
| 1762 |
+
"eval_steps_per_second": 0.858,
|
| 1763 |
+
"step": 800
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"acc": 0.73518519,
|
| 1767 |
+
"epoch": 2.181571815718157,
|
| 1768 |
+
"grad_norm": 1.546875,
|
| 1769 |
+
"learning_rate": 7.179029957203993e-05,
|
| 1770 |
+
"loss": 0.96204119,
|
| 1771 |
+
"memory(GiB)": 16.34,
|
| 1772 |
+
"step": 805,
|
| 1773 |
+
"train_speed(iter/s)": 0.140881
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"acc": 0.74290161,
|
| 1777 |
+
"epoch": 2.1951219512195124,
|
| 1778 |
+
"grad_norm": 2.015625,
|
| 1779 |
+
"learning_rate": 7.125534950071326e-05,
|
| 1780 |
+
"loss": 0.92387733,
|
| 1781 |
+
"memory(GiB)": 16.34,
|
| 1782 |
+
"step": 810,
|
| 1783 |
+
"train_speed(iter/s)": 0.14021
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"acc": 0.73101602,
|
| 1787 |
+
"epoch": 2.2086720867208673,
|
| 1788 |
+
"grad_norm": 1.6015625,
|
| 1789 |
+
"learning_rate": 7.072039942938658e-05,
|
| 1790 |
+
"loss": 0.96248655,
|
| 1791 |
+
"memory(GiB)": 16.34,
|
| 1792 |
+
"step": 815,
|
| 1793 |
+
"train_speed(iter/s)": 0.140323
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"acc": 0.73887796,
|
| 1797 |
+
"epoch": 2.2222222222222223,
|
| 1798 |
+
"grad_norm": 1.796875,
|
| 1799 |
+
"learning_rate": 7.018544935805991e-05,
|
| 1800 |
+
"loss": 0.94751673,
|
| 1801 |
+
"memory(GiB)": 16.34,
|
| 1802 |
+
"step": 820,
|
| 1803 |
+
"train_speed(iter/s)": 0.140441
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"acc": 0.75793715,
|
| 1807 |
+
"epoch": 2.2357723577235773,
|
| 1808 |
+
"grad_norm": 2.015625,
|
| 1809 |
+
"learning_rate": 6.965049928673323e-05,
|
| 1810 |
+
"loss": 0.84263477,
|
| 1811 |
+
"memory(GiB)": 16.34,
|
| 1812 |
+
"step": 825,
|
| 1813 |
+
"train_speed(iter/s)": 0.140555
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"acc": 0.75873909,
|
| 1817 |
+
"epoch": 2.2493224932249323,
|
| 1818 |
+
"grad_norm": 1.953125,
|
| 1819 |
+
"learning_rate": 6.911554921540656e-05,
|
| 1820 |
+
"loss": 0.89048252,
|
| 1821 |
+
"memory(GiB)": 16.34,
|
| 1822 |
+
"step": 830,
|
| 1823 |
+
"train_speed(iter/s)": 0.14067
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"acc": 0.74456077,
|
| 1827 |
+
"epoch": 2.2628726287262872,
|
| 1828 |
+
"grad_norm": 1.7890625,
|
| 1829 |
+
"learning_rate": 6.858059914407987e-05,
|
| 1830 |
+
"loss": 0.90777779,
|
| 1831 |
+
"memory(GiB)": 16.34,
|
| 1832 |
+
"step": 835,
|
| 1833 |
+
"train_speed(iter/s)": 0.140775
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"acc": 0.75809846,
|
| 1837 |
+
"epoch": 2.2764227642276422,
|
| 1838 |
+
"grad_norm": 2.0625,
|
| 1839 |
+
"learning_rate": 6.80456490727532e-05,
|
| 1840 |
+
"loss": 0.87556753,
|
| 1841 |
+
"memory(GiB)": 16.34,
|
| 1842 |
+
"step": 840,
|
| 1843 |
+
"train_speed(iter/s)": 0.14088
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"acc": 0.73194971,
|
| 1847 |
+
"epoch": 2.289972899728997,
|
| 1848 |
+
"grad_norm": 2.765625,
|
| 1849 |
+
"learning_rate": 6.751069900142653e-05,
|
| 1850 |
+
"loss": 0.9232769,
|
| 1851 |
+
"memory(GiB)": 16.34,
|
| 1852 |
+
"step": 845,
|
| 1853 |
+
"train_speed(iter/s)": 0.140986
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"acc": 0.74470835,
|
| 1857 |
+
"epoch": 2.303523035230352,
|
| 1858 |
+
"grad_norm": 1.8515625,
|
| 1859 |
+
"learning_rate": 6.697574893009985e-05,
|
| 1860 |
+
"loss": 0.88600664,
|
| 1861 |
+
"memory(GiB)": 16.34,
|
| 1862 |
+
"step": 850,
|
| 1863 |
+
"train_speed(iter/s)": 0.141086
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 2.303523035230352,
|
| 1867 |
+
"eval_acc": 0.6174632663401453,
|
| 1868 |
+
"eval_loss": 1.660492181777954,
|
| 1869 |
+
"eval_runtime": 44.1393,
|
| 1870 |
+
"eval_samples_per_second": 0.861,
|
| 1871 |
+
"eval_steps_per_second": 0.861,
|
| 1872 |
+
"step": 850
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"acc": 0.76063514,
|
| 1876 |
+
"epoch": 2.317073170731707,
|
| 1877 |
+
"grad_norm": 2.40625,
|
| 1878 |
+
"learning_rate": 6.644079885877318e-05,
|
| 1879 |
+
"loss": 0.80236473,
|
| 1880 |
+
"memory(GiB)": 16.34,
|
| 1881 |
+
"step": 855,
|
| 1882 |
+
"train_speed(iter/s)": 0.140168
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"acc": 0.71505499,
|
| 1886 |
+
"epoch": 2.330623306233062,
|
| 1887 |
+
"grad_norm": 2.671875,
|
| 1888 |
+
"learning_rate": 6.59058487874465e-05,
|
| 1889 |
+
"loss": 1.01317081,
|
| 1890 |
+
"memory(GiB)": 16.34,
|
| 1891 |
+
"step": 860,
|
| 1892 |
+
"train_speed(iter/s)": 0.140281
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"acc": 0.73396034,
|
| 1896 |
+
"epoch": 2.3441734417344176,
|
| 1897 |
+
"grad_norm": 1.703125,
|
| 1898 |
+
"learning_rate": 6.537089871611983e-05,
|
| 1899 |
+
"loss": 0.96496754,
|
| 1900 |
+
"memory(GiB)": 16.34,
|
| 1901 |
+
"step": 865,
|
| 1902 |
+
"train_speed(iter/s)": 0.140387
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"acc": 0.74145699,
|
| 1906 |
+
"epoch": 2.3577235772357725,
|
| 1907 |
+
"grad_norm": 1.9453125,
|
| 1908 |
+
"learning_rate": 6.483594864479316e-05,
|
| 1909 |
+
"loss": 1.00490999,
|
| 1910 |
+
"memory(GiB)": 16.34,
|
| 1911 |
+
"step": 870,
|
| 1912 |
+
"train_speed(iter/s)": 0.140492
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"acc": 0.76035104,
|
| 1916 |
+
"epoch": 2.3712737127371275,
|
| 1917 |
+
"grad_norm": 1.7265625,
|
| 1918 |
+
"learning_rate": 6.430099857346647e-05,
|
| 1919 |
+
"loss": 0.93969469,
|
| 1920 |
+
"memory(GiB)": 16.34,
|
| 1921 |
+
"step": 875,
|
| 1922 |
+
"train_speed(iter/s)": 0.140599
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"acc": 0.75214877,
|
| 1926 |
+
"epoch": 2.3848238482384825,
|
| 1927 |
+
"grad_norm": 1.9296875,
|
| 1928 |
+
"learning_rate": 6.376604850213979e-05,
|
| 1929 |
+
"loss": 0.9385232,
|
| 1930 |
+
"memory(GiB)": 16.34,
|
| 1931 |
+
"step": 880,
|
| 1932 |
+
"train_speed(iter/s)": 0.140705
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"acc": 0.73846526,
|
| 1936 |
+
"epoch": 2.3983739837398375,
|
| 1937 |
+
"grad_norm": 2.421875,
|
| 1938 |
+
"learning_rate": 6.323109843081312e-05,
|
| 1939 |
+
"loss": 0.95887814,
|
| 1940 |
+
"memory(GiB)": 16.34,
|
| 1941 |
+
"step": 885,
|
| 1942 |
+
"train_speed(iter/s)": 0.140812
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"acc": 0.72576594,
|
| 1946 |
+
"epoch": 2.4119241192411924,
|
| 1947 |
+
"grad_norm": 1.46875,
|
| 1948 |
+
"learning_rate": 6.269614835948645e-05,
|
| 1949 |
+
"loss": 0.95252094,
|
| 1950 |
+
"memory(GiB)": 16.34,
|
| 1951 |
+
"step": 890,
|
| 1952 |
+
"train_speed(iter/s)": 0.140912
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"acc": 0.74866586,
|
| 1956 |
+
"epoch": 2.4254742547425474,
|
| 1957 |
+
"grad_norm": 1.5546875,
|
| 1958 |
+
"learning_rate": 6.216119828815977e-05,
|
| 1959 |
+
"loss": 0.91533632,
|
| 1960 |
+
"memory(GiB)": 16.34,
|
| 1961 |
+
"step": 895,
|
| 1962 |
+
"train_speed(iter/s)": 0.141011
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"acc": 0.72289066,
|
| 1966 |
+
"epoch": 2.4390243902439024,
|
| 1967 |
+
"grad_norm": 1.375,
|
| 1968 |
+
"learning_rate": 6.162624821683308e-05,
|
| 1969 |
+
"loss": 0.996418,
|
| 1970 |
+
"memory(GiB)": 16.34,
|
| 1971 |
+
"step": 900,
|
| 1972 |
+
"train_speed(iter/s)": 0.13966
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 2.4390243902439024,
|
| 1976 |
+
"eval_acc": 0.6152676912683668,
|
| 1977 |
+
"eval_loss": 1.6503233909606934,
|
| 1978 |
+
"eval_runtime": 44.1939,
|
| 1979 |
+
"eval_samples_per_second": 0.86,
|
| 1980 |
+
"eval_steps_per_second": 0.86,
|
| 1981 |
+
"step": 900
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"acc": 0.73535914,
|
| 1985 |
+
"epoch": 2.4525745257452574,
|
| 1986 |
+
"grad_norm": 1.3125,
|
| 1987 |
+
"learning_rate": 6.109129814550641e-05,
|
| 1988 |
+
"loss": 1.00612879,
|
| 1989 |
+
"memory(GiB)": 16.34,
|
| 1990 |
+
"step": 905,
|
| 1991 |
+
"train_speed(iter/s)": 0.138805
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"acc": 0.7277998,
|
| 1995 |
+
"epoch": 2.4661246612466123,
|
| 1996 |
+
"grad_norm": 2.328125,
|
| 1997 |
+
"learning_rate": 6.0556348074179737e-05,
|
| 1998 |
+
"loss": 1.00636635,
|
| 1999 |
+
"memory(GiB)": 16.34,
|
| 2000 |
+
"step": 910,
|
| 2001 |
+
"train_speed(iter/s)": 0.138904
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"acc": 0.70795035,
|
| 2005 |
+
"epoch": 2.4796747967479673,
|
| 2006 |
+
"grad_norm": 1.6484375,
|
| 2007 |
+
"learning_rate": 6.002139800285306e-05,
|
| 2008 |
+
"loss": 0.98440151,
|
| 2009 |
+
"memory(GiB)": 16.34,
|
| 2010 |
+
"step": 915,
|
| 2011 |
+
"train_speed(iter/s)": 0.139013
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"acc": 0.7650095,
|
| 2015 |
+
"epoch": 2.4932249322493227,
|
| 2016 |
+
"grad_norm": 2.25,
|
| 2017 |
+
"learning_rate": 5.948644793152638e-05,
|
| 2018 |
+
"loss": 0.77239523,
|
| 2019 |
+
"memory(GiB)": 16.34,
|
| 2020 |
+
"step": 920,
|
| 2021 |
+
"train_speed(iter/s)": 0.139126
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"acc": 0.71504927,
|
| 2025 |
+
"epoch": 2.5067750677506773,
|
| 2026 |
+
"grad_norm": 1.6328125,
|
| 2027 |
+
"learning_rate": 5.895149786019971e-05,
|
| 2028 |
+
"loss": 1.01518288,
|
| 2029 |
+
"memory(GiB)": 16.34,
|
| 2030 |
+
"step": 925,
|
| 2031 |
+
"train_speed(iter/s)": 0.139233
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"acc": 0.71582847,
|
| 2035 |
+
"epoch": 2.5203252032520327,
|
| 2036 |
+
"grad_norm": 2.28125,
|
| 2037 |
+
"learning_rate": 5.841654778887303e-05,
|
| 2038 |
+
"loss": 0.97862291,
|
| 2039 |
+
"memory(GiB)": 16.34,
|
| 2040 |
+
"step": 930,
|
| 2041 |
+
"train_speed(iter/s)": 0.13934
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"acc": 0.74984369,
|
| 2045 |
+
"epoch": 2.5338753387533877,
|
| 2046 |
+
"grad_norm": 1.8515625,
|
| 2047 |
+
"learning_rate": 5.788159771754635e-05,
|
| 2048 |
+
"loss": 0.85021992,
|
| 2049 |
+
"memory(GiB)": 16.34,
|
| 2050 |
+
"step": 935,
|
| 2051 |
+
"train_speed(iter/s)": 0.139448
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"acc": 0.74142261,
|
| 2055 |
+
"epoch": 2.5474254742547426,
|
| 2056 |
+
"grad_norm": 1.796875,
|
| 2057 |
+
"learning_rate": 5.734664764621968e-05,
|
| 2058 |
+
"loss": 0.91843338,
|
| 2059 |
+
"memory(GiB)": 16.34,
|
| 2060 |
+
"step": 940,
|
| 2061 |
+
"train_speed(iter/s)": 0.139564
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"acc": 0.71266222,
|
| 2065 |
+
"epoch": 2.5609756097560976,
|
| 2066 |
+
"grad_norm": 1.875,
|
| 2067 |
+
"learning_rate": 5.6811697574893007e-05,
|
| 2068 |
+
"loss": 1.14017658,
|
| 2069 |
+
"memory(GiB)": 16.34,
|
| 2070 |
+
"step": 945,
|
| 2071 |
+
"train_speed(iter/s)": 0.139667
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"acc": 0.73374839,
|
| 2075 |
+
"epoch": 2.5745257452574526,
|
| 2076 |
+
"grad_norm": 1.671875,
|
| 2077 |
+
"learning_rate": 5.627674750356633e-05,
|
| 2078 |
+
"loss": 0.99979248,
|
| 2079 |
+
"memory(GiB)": 16.34,
|
| 2080 |
+
"step": 950,
|
| 2081 |
+
"train_speed(iter/s)": 0.139769
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.5745257452574526,
|
| 2085 |
+
"eval_acc": 0.6161121432190508,
|
| 2086 |
+
"eval_loss": 1.6511973142623901,
|
| 2087 |
+
"eval_runtime": 44.1881,
|
| 2088 |
+
"eval_samples_per_second": 0.86,
|
| 2089 |
+
"eval_steps_per_second": 0.86,
|
| 2090 |
+
"step": 950
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"acc": 0.74381332,
|
| 2094 |
+
"epoch": 2.5880758807588076,
|
| 2095 |
+
"grad_norm": 1.7734375,
|
| 2096 |
+
"learning_rate": 5.5741797432239646e-05,
|
| 2097 |
+
"loss": 0.93189411,
|
| 2098 |
+
"memory(GiB)": 16.34,
|
| 2099 |
+
"step": 955,
|
| 2100 |
+
"train_speed(iter/s)": 0.138972
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"acc": 0.73305674,
|
| 2104 |
+
"epoch": 2.6016260162601625,
|
| 2105 |
+
"grad_norm": 2.4375,
|
| 2106 |
+
"learning_rate": 5.5206847360912977e-05,
|
| 2107 |
+
"loss": 0.93946905,
|
| 2108 |
+
"memory(GiB)": 16.34,
|
| 2109 |
+
"step": 960,
|
| 2110 |
+
"train_speed(iter/s)": 0.139079
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"acc": 0.72961435,
|
| 2114 |
+
"epoch": 2.6151761517615175,
|
| 2115 |
+
"grad_norm": 1.84375,
|
| 2116 |
+
"learning_rate": 5.46718972895863e-05,
|
| 2117 |
+
"loss": 0.99850683,
|
| 2118 |
+
"memory(GiB)": 16.34,
|
| 2119 |
+
"step": 965,
|
| 2120 |
+
"train_speed(iter/s)": 0.139175
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"acc": 0.7133184,
|
| 2124 |
+
"epoch": 2.6287262872628725,
|
| 2125 |
+
"grad_norm": 1.96875,
|
| 2126 |
+
"learning_rate": 5.413694721825962e-05,
|
| 2127 |
+
"loss": 1.03378878,
|
| 2128 |
+
"memory(GiB)": 16.34,
|
| 2129 |
+
"step": 970,
|
| 2130 |
+
"train_speed(iter/s)": 0.139282
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"acc": 0.71164918,
|
| 2134 |
+
"epoch": 2.642276422764228,
|
| 2135 |
+
"grad_norm": 1.3828125,
|
| 2136 |
+
"learning_rate": 5.360199714693295e-05,
|
| 2137 |
+
"loss": 1.02268944,
|
| 2138 |
+
"memory(GiB)": 16.34,
|
| 2139 |
+
"step": 975,
|
| 2140 |
+
"train_speed(iter/s)": 0.139382
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"acc": 0.71824646,
|
| 2144 |
+
"epoch": 2.6558265582655824,
|
| 2145 |
+
"grad_norm": 1.84375,
|
| 2146 |
+
"learning_rate": 5.306704707560627e-05,
|
| 2147 |
+
"loss": 1.06014738,
|
| 2148 |
+
"memory(GiB)": 16.34,
|
| 2149 |
+
"step": 980,
|
| 2150 |
+
"train_speed(iter/s)": 0.139479
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"acc": 0.75523286,
|
| 2154 |
+
"epoch": 2.669376693766938,
|
| 2155 |
+
"grad_norm": 1.7734375,
|
| 2156 |
+
"learning_rate": 5.253209700427959e-05,
|
| 2157 |
+
"loss": 0.80533791,
|
| 2158 |
+
"memory(GiB)": 16.34,
|
| 2159 |
+
"step": 985,
|
| 2160 |
+
"train_speed(iter/s)": 0.139581
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"acc": 0.72265592,
|
| 2164 |
+
"epoch": 2.682926829268293,
|
| 2165 |
+
"grad_norm": 1.5546875,
|
| 2166 |
+
"learning_rate": 5.199714693295292e-05,
|
| 2167 |
+
"loss": 1.0223958,
|
| 2168 |
+
"memory(GiB)": 16.34,
|
| 2169 |
+
"step": 990,
|
| 2170 |
+
"train_speed(iter/s)": 0.139675
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"acc": 0.74640193,
|
| 2174 |
+
"epoch": 2.696476964769648,
|
| 2175 |
+
"grad_norm": 1.5546875,
|
| 2176 |
+
"learning_rate": 5.1462196861626247e-05,
|
| 2177 |
+
"loss": 0.8896265,
|
| 2178 |
+
"memory(GiB)": 16.34,
|
| 2179 |
+
"step": 995,
|
| 2180 |
+
"train_speed(iter/s)": 0.139767
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"acc": 0.75235152,
|
| 2184 |
+
"epoch": 2.710027100271003,
|
| 2185 |
+
"grad_norm": 1.6484375,
|
| 2186 |
+
"learning_rate": 5.092724679029957e-05,
|
| 2187 |
+
"loss": 0.87937946,
|
| 2188 |
+
"memory(GiB)": 16.34,
|
| 2189 |
+
"step": 1000,
|
| 2190 |
+
"train_speed(iter/s)": 0.139867
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 2.710027100271003,
|
| 2194 |
+
"eval_acc": 0.6123965546360413,
|
| 2195 |
+
"eval_loss": 1.6541699171066284,
|
| 2196 |
+
"eval_runtime": 44.2136,
|
| 2197 |
+
"eval_samples_per_second": 0.859,
|
| 2198 |
+
"eval_steps_per_second": 0.859,
|
| 2199 |
+
"step": 1000
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"acc": 0.73264089,
|
| 2203 |
+
"epoch": 2.7235772357723578,
|
| 2204 |
+
"grad_norm": 1.96875,
|
| 2205 |
+
"learning_rate": 5.0392296718972886e-05,
|
| 2206 |
+
"loss": 0.95834293,
|
| 2207 |
+
"memory(GiB)": 16.34,
|
| 2208 |
+
"step": 1005,
|
| 2209 |
+
"train_speed(iter/s)": 0.139102
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"acc": 0.76931543,
|
| 2213 |
+
"epoch": 2.7371273712737128,
|
| 2214 |
+
"grad_norm": 1.7890625,
|
| 2215 |
+
"learning_rate": 4.9857346647646217e-05,
|
| 2216 |
+
"loss": 0.82889891,
|
| 2217 |
+
"memory(GiB)": 16.34,
|
| 2218 |
+
"step": 1010,
|
| 2219 |
+
"train_speed(iter/s)": 0.139199
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"acc": 0.74891815,
|
| 2223 |
+
"epoch": 2.7506775067750677,
|
| 2224 |
+
"grad_norm": 2.65625,
|
| 2225 |
+
"learning_rate": 4.932239657631954e-05,
|
| 2226 |
+
"loss": 0.93503094,
|
| 2227 |
+
"memory(GiB)": 16.34,
|
| 2228 |
+
"step": 1015,
|
| 2229 |
+
"train_speed(iter/s)": 0.139298
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"acc": 0.72551074,
|
| 2233 |
+
"epoch": 2.7642276422764227,
|
| 2234 |
+
"grad_norm": 1.5703125,
|
| 2235 |
+
"learning_rate": 4.878744650499286e-05,
|
| 2236 |
+
"loss": 1.06086788,
|
| 2237 |
+
"memory(GiB)": 16.34,
|
| 2238 |
+
"step": 1020,
|
| 2239 |
+
"train_speed(iter/s)": 0.139383
|
| 2240 |
+
},
|
| 2241 |
+
{
|
| 2242 |
+
"acc": 0.75321589,
|
| 2243 |
+
"epoch": 2.7777777777777777,
|
| 2244 |
+
"grad_norm": 1.7734375,
|
| 2245 |
+
"learning_rate": 4.825249643366619e-05,
|
| 2246 |
+
"loss": 0.83553734,
|
| 2247 |
+
"memory(GiB)": 16.34,
|
| 2248 |
+
"step": 1025,
|
| 2249 |
+
"train_speed(iter/s)": 0.139478
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"acc": 0.73832054,
|
| 2253 |
+
"epoch": 2.7913279132791327,
|
| 2254 |
+
"grad_norm": 1.53125,
|
| 2255 |
+
"learning_rate": 4.771754636233951e-05,
|
| 2256 |
+
"loss": 0.94093418,
|
| 2257 |
+
"memory(GiB)": 16.34,
|
| 2258 |
+
"step": 1030,
|
| 2259 |
+
"train_speed(iter/s)": 0.139572
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"acc": 0.76351671,
|
| 2263 |
+
"epoch": 2.8048780487804876,
|
| 2264 |
+
"grad_norm": 1.2890625,
|
| 2265 |
+
"learning_rate": 4.718259629101283e-05,
|
| 2266 |
+
"loss": 0.84423561,
|
| 2267 |
+
"memory(GiB)": 16.34,
|
| 2268 |
+
"step": 1035,
|
| 2269 |
+
"train_speed(iter/s)": 0.139664
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"acc": 0.75100279,
|
| 2273 |
+
"epoch": 2.818428184281843,
|
| 2274 |
+
"grad_norm": 1.890625,
|
| 2275 |
+
"learning_rate": 4.6647646219686156e-05,
|
| 2276 |
+
"loss": 0.91757078,
|
| 2277 |
+
"memory(GiB)": 16.34,
|
| 2278 |
+
"step": 1040,
|
| 2279 |
+
"train_speed(iter/s)": 0.139754
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"acc": 0.73302913,
|
| 2283 |
+
"epoch": 2.8319783197831976,
|
| 2284 |
+
"grad_norm": 1.3671875,
|
| 2285 |
+
"learning_rate": 4.6112696148359487e-05,
|
| 2286 |
+
"loss": 0.94331446,
|
| 2287 |
+
"memory(GiB)": 16.34,
|
| 2288 |
+
"step": 1045,
|
| 2289 |
+
"train_speed(iter/s)": 0.139841
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"acc": 0.73944206,
|
| 2293 |
+
"epoch": 2.845528455284553,
|
| 2294 |
+
"grad_norm": 1.59375,
|
| 2295 |
+
"learning_rate": 4.55777460770328e-05,
|
| 2296 |
+
"loss": 0.98726377,
|
| 2297 |
+
"memory(GiB)": 16.34,
|
| 2298 |
+
"step": 1050,
|
| 2299 |
+
"train_speed(iter/s)": 0.139931
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.845528455284553,
|
| 2303 |
+
"eval_acc": 0.6145921297078196,
|
| 2304 |
+
"eval_loss": 1.6571751832962036,
|
| 2305 |
+
"eval_runtime": 44.3705,
|
| 2306 |
+
"eval_samples_per_second": 0.856,
|
| 2307 |
+
"eval_steps_per_second": 0.856,
|
| 2308 |
+
"step": 1050
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"acc": 0.73031135,
|
| 2312 |
+
"epoch": 2.859078590785908,
|
| 2313 |
+
"grad_norm": 2.25,
|
| 2314 |
+
"learning_rate": 4.5042796005706126e-05,
|
| 2315 |
+
"loss": 0.95681438,
|
| 2316 |
+
"memory(GiB)": 16.34,
|
| 2317 |
+
"step": 1055,
|
| 2318 |
+
"train_speed(iter/s)": 0.139201
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"acc": 0.71694627,
|
| 2322 |
+
"epoch": 2.872628726287263,
|
| 2323 |
+
"grad_norm": 1.6328125,
|
| 2324 |
+
"learning_rate": 4.4507845934379456e-05,
|
| 2325 |
+
"loss": 1.02266083,
|
| 2326 |
+
"memory(GiB)": 16.34,
|
| 2327 |
+
"step": 1060,
|
| 2328 |
+
"train_speed(iter/s)": 0.139294
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"acc": 0.72791638,
|
| 2332 |
+
"epoch": 2.886178861788618,
|
| 2333 |
+
"grad_norm": 1.7578125,
|
| 2334 |
+
"learning_rate": 4.397289586305278e-05,
|
| 2335 |
+
"loss": 0.99341927,
|
| 2336 |
+
"memory(GiB)": 16.34,
|
| 2337 |
+
"step": 1065,
|
| 2338 |
+
"train_speed(iter/s)": 0.139385
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"acc": 0.74939594,
|
| 2342 |
+
"epoch": 2.899728997289973,
|
| 2343 |
+
"grad_norm": 1.921875,
|
| 2344 |
+
"learning_rate": 4.34379457917261e-05,
|
| 2345 |
+
"loss": 0.91077061,
|
| 2346 |
+
"memory(GiB)": 16.34,
|
| 2347 |
+
"step": 1070,
|
| 2348 |
+
"train_speed(iter/s)": 0.139478
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"acc": 0.72694654,
|
| 2352 |
+
"epoch": 2.913279132791328,
|
| 2353 |
+
"grad_norm": 1.7265625,
|
| 2354 |
+
"learning_rate": 4.290299572039942e-05,
|
| 2355 |
+
"loss": 0.98774853,
|
| 2356 |
+
"memory(GiB)": 16.34,
|
| 2357 |
+
"step": 1075,
|
| 2358 |
+
"train_speed(iter/s)": 0.139564
|
| 2359 |
+
},
|
| 2360 |
+
{
|
| 2361 |
+
"acc": 0.70588508,
|
| 2362 |
+
"epoch": 2.926829268292683,
|
| 2363 |
+
"grad_norm": 2.15625,
|
| 2364 |
+
"learning_rate": 4.236804564907275e-05,
|
| 2365 |
+
"loss": 1.07887812,
|
| 2366 |
+
"memory(GiB)": 16.34,
|
| 2367 |
+
"step": 1080,
|
| 2368 |
+
"train_speed(iter/s)": 0.139657
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"acc": 0.74654303,
|
| 2372 |
+
"epoch": 2.940379403794038,
|
| 2373 |
+
"grad_norm": 1.7109375,
|
| 2374 |
+
"learning_rate": 4.183309557774607e-05,
|
| 2375 |
+
"loss": 0.91062069,
|
| 2376 |
+
"memory(GiB)": 16.34,
|
| 2377 |
+
"step": 1085,
|
| 2378 |
+
"train_speed(iter/s)": 0.139743
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"acc": 0.73493595,
|
| 2382 |
+
"epoch": 2.953929539295393,
|
| 2383 |
+
"grad_norm": 2.03125,
|
| 2384 |
+
"learning_rate": 4.1298145506419396e-05,
|
| 2385 |
+
"loss": 0.92819033,
|
| 2386 |
+
"memory(GiB)": 16.34,
|
| 2387 |
+
"step": 1090,
|
| 2388 |
+
"train_speed(iter/s)": 0.139829
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"acc": 0.71466756,
|
| 2392 |
+
"epoch": 2.9674796747967482,
|
| 2393 |
+
"grad_norm": 2.109375,
|
| 2394 |
+
"learning_rate": 4.0763195435092727e-05,
|
| 2395 |
+
"loss": 1.01220913,
|
| 2396 |
+
"memory(GiB)": 16.34,
|
| 2397 |
+
"step": 1095,
|
| 2398 |
+
"train_speed(iter/s)": 0.139916
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"acc": 0.7607831,
|
| 2402 |
+
"epoch": 2.9810298102981028,
|
| 2403 |
+
"grad_norm": 2.375,
|
| 2404 |
+
"learning_rate": 4.022824536376604e-05,
|
| 2405 |
+
"loss": 0.8505785,
|
| 2406 |
+
"memory(GiB)": 16.34,
|
| 2407 |
+
"step": 1100,
|
| 2408 |
+
"train_speed(iter/s)": 0.140004
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 2.9810298102981028,
|
| 2412 |
+
"eval_acc": 0.6206721837527445,
|
| 2413 |
+
"eval_loss": 1.6525288820266724,
|
| 2414 |
+
"eval_runtime": 44.2124,
|
| 2415 |
+
"eval_samples_per_second": 0.859,
|
| 2416 |
+
"eval_steps_per_second": 0.859,
|
| 2417 |
+
"step": 1100
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"acc": 0.74116454,
|
| 2421 |
+
"epoch": 2.994579945799458,
|
| 2422 |
+
"grad_norm": 2.65625,
|
| 2423 |
+
"learning_rate": 3.9693295292439366e-05,
|
| 2424 |
+
"loss": 0.92264805,
|
| 2425 |
+
"memory(GiB)": 16.34,
|
| 2426 |
+
"step": 1105,
|
| 2427 |
+
"train_speed(iter/s)": 0.139311
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"acc": 0.8182869,
|
| 2431 |
+
"epoch": 3.008130081300813,
|
| 2432 |
+
"grad_norm": 1.3671875,
|
| 2433 |
+
"learning_rate": 3.915834522111269e-05,
|
| 2434 |
+
"loss": 0.6927434,
|
| 2435 |
+
"memory(GiB)": 16.34,
|
| 2436 |
+
"step": 1110,
|
| 2437 |
+
"train_speed(iter/s)": 0.13923
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"acc": 0.83515587,
|
| 2441 |
+
"epoch": 3.021680216802168,
|
| 2442 |
+
"grad_norm": 1.9140625,
|
| 2443 |
+
"learning_rate": 3.862339514978602e-05,
|
| 2444 |
+
"loss": 0.58284421,
|
| 2445 |
+
"memory(GiB)": 16.34,
|
| 2446 |
+
"step": 1115,
|
| 2447 |
+
"train_speed(iter/s)": 0.139319
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"acc": 0.84008894,
|
| 2451 |
+
"epoch": 3.035230352303523,
|
| 2452 |
+
"grad_norm": 1.8828125,
|
| 2453 |
+
"learning_rate": 3.808844507845934e-05,
|
| 2454 |
+
"loss": 0.59128432,
|
| 2455 |
+
"memory(GiB)": 16.34,
|
| 2456 |
+
"step": 1120,
|
| 2457 |
+
"train_speed(iter/s)": 0.139405
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"acc": 0.84298267,
|
| 2461 |
+
"epoch": 3.048780487804878,
|
| 2462 |
+
"grad_norm": 2.4375,
|
| 2463 |
+
"learning_rate": 3.755349500713266e-05,
|
| 2464 |
+
"loss": 0.56685705,
|
| 2465 |
+
"memory(GiB)": 16.34,
|
| 2466 |
+
"step": 1125,
|
| 2467 |
+
"train_speed(iter/s)": 0.139493
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"acc": 0.83680372,
|
| 2471 |
+
"epoch": 3.062330623306233,
|
| 2472 |
+
"grad_norm": 3.171875,
|
| 2473 |
+
"learning_rate": 3.701854493580599e-05,
|
| 2474 |
+
"loss": 0.55599914,
|
| 2475 |
+
"memory(GiB)": 16.34,
|
| 2476 |
+
"step": 1130,
|
| 2477 |
+
"train_speed(iter/s)": 0.139571
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"acc": 0.81269236,
|
| 2481 |
+
"epoch": 3.075880758807588,
|
| 2482 |
+
"grad_norm": 2.40625,
|
| 2483 |
+
"learning_rate": 3.648359486447931e-05,
|
| 2484 |
+
"loss": 0.63413367,
|
| 2485 |
+
"memory(GiB)": 16.34,
|
| 2486 |
+
"step": 1135,
|
| 2487 |
+
"train_speed(iter/s)": 0.139659
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"acc": 0.8300642,
|
| 2491 |
+
"epoch": 3.089430894308943,
|
| 2492 |
+
"grad_norm": 2.71875,
|
| 2493 |
+
"learning_rate": 3.5948644793152636e-05,
|
| 2494 |
+
"loss": 0.59288554,
|
| 2495 |
+
"memory(GiB)": 16.34,
|
| 2496 |
+
"step": 1140,
|
| 2497 |
+
"train_speed(iter/s)": 0.139738
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"acc": 0.84225779,
|
| 2501 |
+
"epoch": 3.102981029810298,
|
| 2502 |
+
"grad_norm": 2.703125,
|
| 2503 |
+
"learning_rate": 3.541369472182596e-05,
|
| 2504 |
+
"loss": 0.57104263,
|
| 2505 |
+
"memory(GiB)": 16.34,
|
| 2506 |
+
"step": 1145,
|
| 2507 |
+
"train_speed(iter/s)": 0.13982
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"acc": 0.86128368,
|
| 2511 |
+
"epoch": 3.116531165311653,
|
| 2512 |
+
"grad_norm": 3.0,
|
| 2513 |
+
"learning_rate": 3.487874465049928e-05,
|
| 2514 |
+
"loss": 0.47776127,
|
| 2515 |
+
"memory(GiB)": 16.34,
|
| 2516 |
+
"step": 1150,
|
| 2517 |
+
"train_speed(iter/s)": 0.139903
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 3.116531165311653,
|
| 2521 |
+
"eval_acc": 0.6091876372234419,
|
| 2522 |
+
"eval_loss": 1.9067459106445312,
|
| 2523 |
+
"eval_runtime": 44.213,
|
| 2524 |
+
"eval_samples_per_second": 0.859,
|
| 2525 |
+
"eval_steps_per_second": 0.859,
|
| 2526 |
+
"step": 1150
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"acc": 0.81826935,
|
| 2530 |
+
"epoch": 3.130081300813008,
|
| 2531 |
+
"grad_norm": 2.234375,
|
| 2532 |
+
"learning_rate": 3.4343794579172606e-05,
|
| 2533 |
+
"loss": 0.64635777,
|
| 2534 |
+
"memory(GiB)": 16.34,
|
| 2535 |
+
"step": 1155,
|
| 2536 |
+
"train_speed(iter/s)": 0.13924
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"acc": 0.79423141,
|
| 2540 |
+
"epoch": 3.1436314363143634,
|
| 2541 |
+
"grad_norm": 2.859375,
|
| 2542 |
+
"learning_rate": 3.380884450784593e-05,
|
| 2543 |
+
"loss": 0.71202464,
|
| 2544 |
+
"memory(GiB)": 16.34,
|
| 2545 |
+
"step": 1160,
|
| 2546 |
+
"train_speed(iter/s)": 0.139319
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"acc": 0.81948729,
|
| 2550 |
+
"epoch": 3.1571815718157183,
|
| 2551 |
+
"grad_norm": 2.046875,
|
| 2552 |
+
"learning_rate": 3.327389443651925e-05,
|
| 2553 |
+
"loss": 0.62771091,
|
| 2554 |
+
"memory(GiB)": 16.34,
|
| 2555 |
+
"step": 1165,
|
| 2556 |
+
"train_speed(iter/s)": 0.1394
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"acc": 0.85081501,
|
| 2560 |
+
"epoch": 3.1707317073170733,
|
| 2561 |
+
"grad_norm": 2.125,
|
| 2562 |
+
"learning_rate": 3.2738944365192576e-05,
|
| 2563 |
+
"loss": 0.55290155,
|
| 2564 |
+
"memory(GiB)": 16.34,
|
| 2565 |
+
"step": 1170,
|
| 2566 |
+
"train_speed(iter/s)": 0.139484
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"acc": 0.84410248,
|
| 2570 |
+
"epoch": 3.1842818428184283,
|
| 2571 |
+
"grad_norm": 2.140625,
|
| 2572 |
+
"learning_rate": 3.2203994293865906e-05,
|
| 2573 |
+
"loss": 0.57252893,
|
| 2574 |
+
"memory(GiB)": 16.34,
|
| 2575 |
+
"step": 1175,
|
| 2576 |
+
"train_speed(iter/s)": 0.139567
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"acc": 0.78565025,
|
| 2580 |
+
"epoch": 3.1978319783197833,
|
| 2581 |
+
"grad_norm": 3.203125,
|
| 2582 |
+
"learning_rate": 3.166904422253922e-05,
|
| 2583 |
+
"loss": 0.73501797,
|
| 2584 |
+
"memory(GiB)": 16.34,
|
| 2585 |
+
"step": 1180,
|
| 2586 |
+
"train_speed(iter/s)": 0.13965
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"acc": 0.78738356,
|
| 2590 |
+
"epoch": 3.2113821138211383,
|
| 2591 |
+
"grad_norm": 2.953125,
|
| 2592 |
+
"learning_rate": 3.113409415121255e-05,
|
| 2593 |
+
"loss": 0.72265806,
|
| 2594 |
+
"memory(GiB)": 16.34,
|
| 2595 |
+
"step": 1185,
|
| 2596 |
+
"train_speed(iter/s)": 0.139732
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"acc": 0.818221,
|
| 2600 |
+
"epoch": 3.2249322493224932,
|
| 2601 |
+
"grad_norm": 2.21875,
|
| 2602 |
+
"learning_rate": 3.0599144079885876e-05,
|
| 2603 |
+
"loss": 0.61062384,
|
| 2604 |
+
"memory(GiB)": 16.34,
|
| 2605 |
+
"step": 1190,
|
| 2606 |
+
"train_speed(iter/s)": 0.139813
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"acc": 0.82556448,
|
| 2610 |
+
"epoch": 3.238482384823848,
|
| 2611 |
+
"grad_norm": 2.3125,
|
| 2612 |
+
"learning_rate": 3.00641940085592e-05,
|
| 2613 |
+
"loss": 0.61080799,
|
| 2614 |
+
"memory(GiB)": 16.34,
|
| 2615 |
+
"step": 1195,
|
| 2616 |
+
"train_speed(iter/s)": 0.139893
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"acc": 0.82586126,
|
| 2620 |
+
"epoch": 3.252032520325203,
|
| 2621 |
+
"grad_norm": 2.34375,
|
| 2622 |
+
"learning_rate": 2.952924393723252e-05,
|
| 2623 |
+
"loss": 0.6164794,
|
| 2624 |
+
"memory(GiB)": 16.34,
|
| 2625 |
+
"step": 1200,
|
| 2626 |
+
"train_speed(iter/s)": 0.139974
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 3.252032520325203,
|
| 2630 |
+
"eval_acc": 0.6044587062996115,
|
| 2631 |
+
"eval_loss": 1.8586076498031616,
|
| 2632 |
+
"eval_runtime": 44.1645,
|
| 2633 |
+
"eval_samples_per_second": 0.86,
|
| 2634 |
+
"eval_steps_per_second": 0.86,
|
| 2635 |
+
"step": 1200
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"acc": 0.82835131,
|
| 2639 |
+
"epoch": 3.265582655826558,
|
| 2640 |
+
"grad_norm": 6.46875,
|
| 2641 |
+
"learning_rate": 2.8994293865905846e-05,
|
| 2642 |
+
"loss": 0.61753302,
|
| 2643 |
+
"memory(GiB)": 16.34,
|
| 2644 |
+
"step": 1205,
|
| 2645 |
+
"train_speed(iter/s)": 0.139339
|
| 2646 |
+
},
|
| 2647 |
+
{
|
| 2648 |
+
"acc": 0.82603951,
|
| 2649 |
+
"epoch": 3.279132791327913,
|
| 2650 |
+
"grad_norm": 2.90625,
|
| 2651 |
+
"learning_rate": 2.8459343794579173e-05,
|
| 2652 |
+
"loss": 0.58360481,
|
| 2653 |
+
"memory(GiB)": 16.34,
|
| 2654 |
+
"step": 1210,
|
| 2655 |
+
"train_speed(iter/s)": 0.139423
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"acc": 0.79192553,
|
| 2659 |
+
"epoch": 3.292682926829268,
|
| 2660 |
+
"grad_norm": 2.5,
|
| 2661 |
+
"learning_rate": 2.7924393723252493e-05,
|
| 2662 |
+
"loss": 0.75090132,
|
| 2663 |
+
"memory(GiB)": 16.34,
|
| 2664 |
+
"step": 1215,
|
| 2665 |
+
"train_speed(iter/s)": 0.1395
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"acc": 0.83583679,
|
| 2669 |
+
"epoch": 3.306233062330623,
|
| 2670 |
+
"grad_norm": 2.15625,
|
| 2671 |
+
"learning_rate": 2.738944365192582e-05,
|
| 2672 |
+
"loss": 0.56217456,
|
| 2673 |
+
"memory(GiB)": 16.34,
|
| 2674 |
+
"step": 1220,
|
| 2675 |
+
"train_speed(iter/s)": 0.139581
|
| 2676 |
+
},
|
| 2677 |
+
{
|
| 2678 |
+
"acc": 0.80481911,
|
| 2679 |
+
"epoch": 3.3197831978319785,
|
| 2680 |
+
"grad_norm": 4.125,
|
| 2681 |
+
"learning_rate": 2.685449358059914e-05,
|
| 2682 |
+
"loss": 0.73073688,
|
| 2683 |
+
"memory(GiB)": 16.34,
|
| 2684 |
+
"step": 1225,
|
| 2685 |
+
"train_speed(iter/s)": 0.139659
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"acc": 0.8173975,
|
| 2689 |
+
"epoch": 3.3333333333333335,
|
| 2690 |
+
"grad_norm": 3.1875,
|
| 2691 |
+
"learning_rate": 2.6319543509272466e-05,
|
| 2692 |
+
"loss": 0.59854188,
|
| 2693 |
+
"memory(GiB)": 16.34,
|
| 2694 |
+
"step": 1230,
|
| 2695 |
+
"train_speed(iter/s)": 0.139735
|
| 2696 |
+
},
|
| 2697 |
+
{
|
| 2698 |
+
"acc": 0.88763266,
|
| 2699 |
+
"epoch": 3.3468834688346885,
|
| 2700 |
+
"grad_norm": 2.078125,
|
| 2701 |
+
"learning_rate": 2.5784593437945793e-05,
|
| 2702 |
+
"loss": 0.3878314,
|
| 2703 |
+
"memory(GiB)": 16.34,
|
| 2704 |
+
"step": 1235,
|
| 2705 |
+
"train_speed(iter/s)": 0.139826
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"acc": 0.82199507,
|
| 2709 |
+
"epoch": 3.3604336043360434,
|
| 2710 |
+
"grad_norm": 2.28125,
|
| 2711 |
+
"learning_rate": 2.5249643366619113e-05,
|
| 2712 |
+
"loss": 0.65350094,
|
| 2713 |
+
"memory(GiB)": 16.34,
|
| 2714 |
+
"step": 1240,
|
| 2715 |
+
"train_speed(iter/s)": 0.139901
|
| 2716 |
+
},
|
| 2717 |
+
{
|
| 2718 |
+
"acc": 0.83115234,
|
| 2719 |
+
"epoch": 3.3739837398373984,
|
| 2720 |
+
"grad_norm": 2.484375,
|
| 2721 |
+
"learning_rate": 2.471469329529244e-05,
|
| 2722 |
+
"loss": 0.61815515,
|
| 2723 |
+
"memory(GiB)": 16.34,
|
| 2724 |
+
"step": 1245,
|
| 2725 |
+
"train_speed(iter/s)": 0.13998
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"acc": 0.81266117,
|
| 2729 |
+
"epoch": 3.3875338753387534,
|
| 2730 |
+
"grad_norm": 1.78125,
|
| 2731 |
+
"learning_rate": 2.417974322396576e-05,
|
| 2732 |
+
"loss": 0.65289016,
|
| 2733 |
+
"memory(GiB)": 16.34,
|
| 2734 |
+
"step": 1250,
|
| 2735 |
+
"train_speed(iter/s)": 0.140052
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 3.3875338753387534,
|
| 2739 |
+
"eval_acc": 0.6037831447390644,
|
| 2740 |
+
"eval_loss": 1.893083095550537,
|
| 2741 |
+
"eval_runtime": 44.1639,
|
| 2742 |
+
"eval_samples_per_second": 0.86,
|
| 2743 |
+
"eval_steps_per_second": 0.86,
|
| 2744 |
+
"step": 1250
|
| 2745 |
+
},
|
| 2746 |
+
{
|
| 2747 |
+
"acc": 0.84776039,
|
| 2748 |
+
"epoch": 3.4010840108401084,
|
| 2749 |
+
"grad_norm": 2.046875,
|
| 2750 |
+
"learning_rate": 2.3644793152639086e-05,
|
| 2751 |
+
"loss": 0.53162961,
|
| 2752 |
+
"memory(GiB)": 16.34,
|
| 2753 |
+
"step": 1255,
|
| 2754 |
+
"train_speed(iter/s)": 0.139443
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"acc": 0.84318447,
|
| 2758 |
+
"epoch": 3.4146341463414633,
|
| 2759 |
+
"grad_norm": 2.234375,
|
| 2760 |
+
"learning_rate": 2.3109843081312406e-05,
|
| 2761 |
+
"loss": 0.52226324,
|
| 2762 |
+
"memory(GiB)": 16.34,
|
| 2763 |
+
"step": 1260,
|
| 2764 |
+
"train_speed(iter/s)": 0.139515
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"acc": 0.80521078,
|
| 2768 |
+
"epoch": 3.4281842818428183,
|
| 2769 |
+
"grad_norm": 2.25,
|
| 2770 |
+
"learning_rate": 2.2574893009985733e-05,
|
| 2771 |
+
"loss": 0.62338996,
|
| 2772 |
+
"memory(GiB)": 16.34,
|
| 2773 |
+
"step": 1265,
|
| 2774 |
+
"train_speed(iter/s)": 0.13959
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"acc": 0.83542995,
|
| 2778 |
+
"epoch": 3.4417344173441733,
|
| 2779 |
+
"grad_norm": 3.0625,
|
| 2780 |
+
"learning_rate": 2.203994293865906e-05,
|
| 2781 |
+
"loss": 0.60965805,
|
| 2782 |
+
"memory(GiB)": 16.34,
|
| 2783 |
+
"step": 1270,
|
| 2784 |
+
"train_speed(iter/s)": 0.139638
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"acc": 0.8333952,
|
| 2788 |
+
"epoch": 3.4552845528455283,
|
| 2789 |
+
"grad_norm": 2.625,
|
| 2790 |
+
"learning_rate": 2.150499286733238e-05,
|
| 2791 |
+
"loss": 0.60031776,
|
| 2792 |
+
"memory(GiB)": 16.34,
|
| 2793 |
+
"step": 1275,
|
| 2794 |
+
"train_speed(iter/s)": 0.139705
|
| 2795 |
+
},
|
| 2796 |
+
{
|
| 2797 |
+
"acc": 0.84367867,
|
| 2798 |
+
"epoch": 3.4688346883468837,
|
| 2799 |
+
"grad_norm": 1.859375,
|
| 2800 |
+
"learning_rate": 2.0970042796005706e-05,
|
| 2801 |
+
"loss": 0.54981174,
|
| 2802 |
+
"memory(GiB)": 16.34,
|
| 2803 |
+
"step": 1280,
|
| 2804 |
+
"train_speed(iter/s)": 0.13978
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"acc": 0.8121892,
|
| 2808 |
+
"epoch": 3.4823848238482387,
|
| 2809 |
+
"grad_norm": 2.046875,
|
| 2810 |
+
"learning_rate": 2.0435092724679026e-05,
|
| 2811 |
+
"loss": 0.67483497,
|
| 2812 |
+
"memory(GiB)": 16.34,
|
| 2813 |
+
"step": 1285,
|
| 2814 |
+
"train_speed(iter/s)": 0.139851
|
| 2815 |
+
},
|
| 2816 |
+
{
|
| 2817 |
+
"acc": 0.87883768,
|
| 2818 |
+
"epoch": 3.4959349593495936,
|
| 2819 |
+
"grad_norm": 1.6875,
|
| 2820 |
+
"learning_rate": 1.9900142653352353e-05,
|
| 2821 |
+
"loss": 0.45111437,
|
| 2822 |
+
"memory(GiB)": 16.34,
|
| 2823 |
+
"step": 1290,
|
| 2824 |
+
"train_speed(iter/s)": 0.139924
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"acc": 0.81952734,
|
| 2828 |
+
"epoch": 3.5094850948509486,
|
| 2829 |
+
"grad_norm": 2.640625,
|
| 2830 |
+
"learning_rate": 1.9365192582025676e-05,
|
| 2831 |
+
"loss": 0.66096244,
|
| 2832 |
+
"memory(GiB)": 16.34,
|
| 2833 |
+
"step": 1295,
|
| 2834 |
+
"train_speed(iter/s)": 0.139995
|
| 2835 |
+
},
|
| 2836 |
+
{
|
| 2837 |
+
"acc": 0.82709446,
|
| 2838 |
+
"epoch": 3.5230352303523036,
|
| 2839 |
+
"grad_norm": 2.140625,
|
| 2840 |
+
"learning_rate": 1.8830242510699e-05,
|
| 2841 |
+
"loss": 0.58278303,
|
| 2842 |
+
"memory(GiB)": 16.34,
|
| 2843 |
+
"step": 1300,
|
| 2844 |
+
"train_speed(iter/s)": 0.140068
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 3.5230352303523036,
|
| 2848 |
+
"eval_acc": 0.603276473568654,
|
| 2849 |
+
"eval_loss": 1.8986655473709106,
|
| 2850 |
+
"eval_runtime": 44.3091,
|
| 2851 |
+
"eval_samples_per_second": 0.858,
|
| 2852 |
+
"eval_steps_per_second": 0.858,
|
| 2853 |
+
"step": 1300
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"acc": 0.83689537,
|
| 2857 |
+
"epoch": 3.5365853658536586,
|
| 2858 |
+
"grad_norm": 1.9453125,
|
| 2859 |
+
"learning_rate": 1.8295292439372323e-05,
|
| 2860 |
+
"loss": 0.55435977,
|
| 2861 |
+
"memory(GiB)": 16.34,
|
| 2862 |
+
"step": 1305,
|
| 2863 |
+
"train_speed(iter/s)": 0.139476
|
| 2864 |
+
},
|
| 2865 |
+
{
|
| 2866 |
+
"acc": 0.8525279,
|
| 2867 |
+
"epoch": 3.5501355013550135,
|
| 2868 |
+
"grad_norm": 2.21875,
|
| 2869 |
+
"learning_rate": 1.776034236804565e-05,
|
| 2870 |
+
"loss": 0.51831579,
|
| 2871 |
+
"memory(GiB)": 16.34,
|
| 2872 |
+
"step": 1310,
|
| 2873 |
+
"train_speed(iter/s)": 0.139549
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"acc": 0.83099527,
|
| 2877 |
+
"epoch": 3.5636856368563685,
|
| 2878 |
+
"grad_norm": 1.8828125,
|
| 2879 |
+
"learning_rate": 1.7225392296718973e-05,
|
| 2880 |
+
"loss": 0.61766572,
|
| 2881 |
+
"memory(GiB)": 16.34,
|
| 2882 |
+
"step": 1315,
|
| 2883 |
+
"train_speed(iter/s)": 0.139617
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"acc": 0.82115297,
|
| 2887 |
+
"epoch": 3.5772357723577235,
|
| 2888 |
+
"grad_norm": 2.765625,
|
| 2889 |
+
"learning_rate": 1.6690442225392296e-05,
|
| 2890 |
+
"loss": 0.62109137,
|
| 2891 |
+
"memory(GiB)": 16.34,
|
| 2892 |
+
"step": 1320,
|
| 2893 |
+
"train_speed(iter/s)": 0.13969
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"acc": 0.85132771,
|
| 2897 |
+
"epoch": 3.5907859078590785,
|
| 2898 |
+
"grad_norm": 2.421875,
|
| 2899 |
+
"learning_rate": 1.615549215406562e-05,
|
| 2900 |
+
"loss": 0.55242,
|
| 2901 |
+
"memory(GiB)": 16.34,
|
| 2902 |
+
"step": 1325,
|
| 2903 |
+
"train_speed(iter/s)": 0.139763
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"acc": 0.82085381,
|
| 2907 |
+
"epoch": 3.6043360433604335,
|
| 2908 |
+
"grad_norm": 2.8125,
|
| 2909 |
+
"learning_rate": 1.5620542082738943e-05,
|
| 2910 |
+
"loss": 0.62965093,
|
| 2911 |
+
"memory(GiB)": 16.34,
|
| 2912 |
+
"step": 1330,
|
| 2913 |
+
"train_speed(iter/s)": 0.139837
|
| 2914 |
+
},
|
| 2915 |
+
{
|
| 2916 |
+
"acc": 0.85592861,
|
| 2917 |
+
"epoch": 3.617886178861789,
|
| 2918 |
+
"grad_norm": 1.875,
|
| 2919 |
+
"learning_rate": 1.5085592011412266e-05,
|
| 2920 |
+
"loss": 0.54980545,
|
| 2921 |
+
"memory(GiB)": 16.34,
|
| 2922 |
+
"step": 1335,
|
| 2923 |
+
"train_speed(iter/s)": 0.139907
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"acc": 0.80786858,
|
| 2927 |
+
"epoch": 3.6314363143631434,
|
| 2928 |
+
"grad_norm": 2.328125,
|
| 2929 |
+
"learning_rate": 1.4550641940085591e-05,
|
| 2930 |
+
"loss": 0.67223382,
|
| 2931 |
+
"memory(GiB)": 16.34,
|
| 2932 |
+
"step": 1340,
|
| 2933 |
+
"train_speed(iter/s)": 0.139974
|
| 2934 |
+
},
|
| 2935 |
+
{
|
| 2936 |
+
"acc": 0.85045481,
|
| 2937 |
+
"epoch": 3.644986449864499,
|
| 2938 |
+
"grad_norm": 2.078125,
|
| 2939 |
+
"learning_rate": 1.4015691868758914e-05,
|
| 2940 |
+
"loss": 0.50758219,
|
| 2941 |
+
"memory(GiB)": 16.34,
|
| 2942 |
+
"step": 1345,
|
| 2943 |
+
"train_speed(iter/s)": 0.140054
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"acc": 0.85536547,
|
| 2947 |
+
"epoch": 3.658536585365854,
|
| 2948 |
+
"grad_norm": 1.7734375,
|
| 2949 |
+
"learning_rate": 1.348074179743224e-05,
|
| 2950 |
+
"loss": 0.53356686,
|
| 2951 |
+
"memory(GiB)": 16.34,
|
| 2952 |
+
"step": 1350,
|
| 2953 |
+
"train_speed(iter/s)": 0.140124
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 3.658536585365854,
|
| 2957 |
+
"eval_acc": 0.6010808984968755,
|
| 2958 |
+
"eval_loss": 1.8972593545913696,
|
| 2959 |
+
"eval_runtime": 44.2915,
|
| 2960 |
+
"eval_samples_per_second": 0.858,
|
| 2961 |
+
"eval_steps_per_second": 0.858,
|
| 2962 |
+
"step": 1350
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"acc": 0.8181016,
|
| 2966 |
+
"epoch": 3.6720867208672088,
|
| 2967 |
+
"grad_norm": 2.3125,
|
| 2968 |
+
"learning_rate": 1.2945791726105563e-05,
|
| 2969 |
+
"loss": 0.59544349,
|
| 2970 |
+
"memory(GiB)": 16.34,
|
| 2971 |
+
"step": 1355,
|
| 2972 |
+
"train_speed(iter/s)": 0.139555
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"acc": 0.87799835,
|
| 2976 |
+
"epoch": 3.6856368563685638,
|
| 2977 |
+
"grad_norm": 2.03125,
|
| 2978 |
+
"learning_rate": 1.2410841654778886e-05,
|
| 2979 |
+
"loss": 0.42857742,
|
| 2980 |
+
"memory(GiB)": 16.34,
|
| 2981 |
+
"step": 1360,
|
| 2982 |
+
"train_speed(iter/s)": 0.139631
|
| 2983 |
+
},
|
| 2984 |
+
{
|
| 2985 |
+
"acc": 0.81485729,
|
| 2986 |
+
"epoch": 3.6991869918699187,
|
| 2987 |
+
"grad_norm": 1.65625,
|
| 2988 |
+
"learning_rate": 1.187589158345221e-05,
|
| 2989 |
+
"loss": 0.64030995,
|
| 2990 |
+
"memory(GiB)": 16.34,
|
| 2991 |
+
"step": 1365,
|
| 2992 |
+
"train_speed(iter/s)": 0.139703
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"acc": 0.82011805,
|
| 2996 |
+
"epoch": 3.7127371273712737,
|
| 2997 |
+
"grad_norm": 2.125,
|
| 2998 |
+
"learning_rate": 1.1340941512125534e-05,
|
| 2999 |
+
"loss": 0.63428659,
|
| 3000 |
+
"memory(GiB)": 16.34,
|
| 3001 |
+
"step": 1370,
|
| 3002 |
+
"train_speed(iter/s)": 0.139771
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"acc": 0.78548884,
|
| 3006 |
+
"epoch": 3.7262872628726287,
|
| 3007 |
+
"grad_norm": 3.984375,
|
| 3008 |
+
"learning_rate": 1.0805991440798858e-05,
|
| 3009 |
+
"loss": 0.74558139,
|
| 3010 |
+
"memory(GiB)": 16.34,
|
| 3011 |
+
"step": 1375,
|
| 3012 |
+
"train_speed(iter/s)": 0.139836
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"acc": 0.85822716,
|
| 3016 |
+
"epoch": 3.7398373983739837,
|
| 3017 |
+
"grad_norm": 2.328125,
|
| 3018 |
+
"learning_rate": 1.0271041369472183e-05,
|
| 3019 |
+
"loss": 0.47486768,
|
| 3020 |
+
"memory(GiB)": 16.34,
|
| 3021 |
+
"step": 1380,
|
| 3022 |
+
"train_speed(iter/s)": 0.139902
|
| 3023 |
+
},
|
| 3024 |
+
{
|
| 3025 |
+
"acc": 0.83469105,
|
| 3026 |
+
"epoch": 3.7533875338753386,
|
| 3027 |
+
"grad_norm": 2.53125,
|
| 3028 |
+
"learning_rate": 9.736091298145506e-06,
|
| 3029 |
+
"loss": 0.56590233,
|
| 3030 |
+
"memory(GiB)": 16.34,
|
| 3031 |
+
"step": 1385,
|
| 3032 |
+
"train_speed(iter/s)": 0.139971
|
| 3033 |
+
},
|
| 3034 |
+
{
|
| 3035 |
+
"acc": 0.80486965,
|
| 3036 |
+
"epoch": 3.7669376693766936,
|
| 3037 |
+
"grad_norm": 3.28125,
|
| 3038 |
+
"learning_rate": 9.20114122681883e-06,
|
| 3039 |
+
"loss": 0.68021841,
|
| 3040 |
+
"memory(GiB)": 16.34,
|
| 3041 |
+
"step": 1390,
|
| 3042 |
+
"train_speed(iter/s)": 0.140036
|
| 3043 |
+
},
|
| 3044 |
+
{
|
| 3045 |
+
"acc": 0.84560328,
|
| 3046 |
+
"epoch": 3.7804878048780486,
|
| 3047 |
+
"grad_norm": 2.421875,
|
| 3048 |
+
"learning_rate": 8.666191155492154e-06,
|
| 3049 |
+
"loss": 0.52926989,
|
| 3050 |
+
"memory(GiB)": 16.34,
|
| 3051 |
+
"step": 1395,
|
| 3052 |
+
"train_speed(iter/s)": 0.140102
|
| 3053 |
+
},
|
| 3054 |
+
{
|
| 3055 |
+
"acc": 0.83154497,
|
| 3056 |
+
"epoch": 3.794037940379404,
|
| 3057 |
+
"grad_norm": 2.671875,
|
| 3058 |
+
"learning_rate": 8.131241084165478e-06,
|
| 3059 |
+
"loss": 0.5866107,
|
| 3060 |
+
"memory(GiB)": 16.34,
|
| 3061 |
+
"step": 1400,
|
| 3062 |
+
"train_speed(iter/s)": 0.14017
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 3.794037940379404,
|
| 3066 |
+
"eval_acc": 0.6036142543489276,
|
| 3067 |
+
"eval_loss": 1.8983793258666992,
|
| 3068 |
+
"eval_runtime": 44.4268,
|
| 3069 |
+
"eval_samples_per_second": 0.855,
|
| 3070 |
+
"eval_steps_per_second": 0.855,
|
| 3071 |
+
"step": 1400
|
| 3072 |
+
},
|
| 3073 |
+
{
|
| 3074 |
+
"acc": 0.84980259,
|
| 3075 |
+
"epoch": 3.8075880758807585,
|
| 3076 |
+
"grad_norm": 2.859375,
|
| 3077 |
+
"learning_rate": 7.596291012838801e-06,
|
| 3078 |
+
"loss": 0.56505013,
|
| 3079 |
+
"memory(GiB)": 16.34,
|
| 3080 |
+
"step": 1405,
|
| 3081 |
+
"train_speed(iter/s)": 0.139621
|
| 3082 |
+
},
|
| 3083 |
+
{
|
| 3084 |
+
"acc": 0.80126667,
|
| 3085 |
+
"epoch": 3.821138211382114,
|
| 3086 |
+
"grad_norm": 2.21875,
|
| 3087 |
+
"learning_rate": 7.061340941512125e-06,
|
| 3088 |
+
"loss": 0.67650762,
|
| 3089 |
+
"memory(GiB)": 16.34,
|
| 3090 |
+
"step": 1410,
|
| 3091 |
+
"train_speed(iter/s)": 0.139688
|
| 3092 |
+
},
|
| 3093 |
+
{
|
| 3094 |
+
"acc": 0.79974804,
|
| 3095 |
+
"epoch": 3.834688346883469,
|
| 3096 |
+
"grad_norm": 2.3125,
|
| 3097 |
+
"learning_rate": 6.5263908701854486e-06,
|
| 3098 |
+
"loss": 0.743855,
|
| 3099 |
+
"memory(GiB)": 16.34,
|
| 3100 |
+
"step": 1415,
|
| 3101 |
+
"train_speed(iter/s)": 0.139749
|
| 3102 |
+
},
|
| 3103 |
+
{
|
| 3104 |
+
"acc": 0.84023552,
|
| 3105 |
+
"epoch": 3.848238482384824,
|
| 3106 |
+
"grad_norm": 2.234375,
|
| 3107 |
+
"learning_rate": 5.991440798858773e-06,
|
| 3108 |
+
"loss": 0.57722979,
|
| 3109 |
+
"memory(GiB)": 16.34,
|
| 3110 |
+
"step": 1420,
|
| 3111 |
+
"train_speed(iter/s)": 0.139813
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"acc": 0.82472792,
|
| 3115 |
+
"epoch": 3.861788617886179,
|
| 3116 |
+
"grad_norm": 2.484375,
|
| 3117 |
+
"learning_rate": 5.456490727532097e-06,
|
| 3118 |
+
"loss": 0.56691737,
|
| 3119 |
+
"memory(GiB)": 16.34,
|
| 3120 |
+
"step": 1425,
|
| 3121 |
+
"train_speed(iter/s)": 0.139877
|
| 3122 |
+
},
|
| 3123 |
+
{
|
| 3124 |
+
"acc": 0.83573132,
|
| 3125 |
+
"epoch": 3.875338753387534,
|
| 3126 |
+
"grad_norm": 1.9765625,
|
| 3127 |
+
"learning_rate": 4.92154065620542e-06,
|
| 3128 |
+
"loss": 0.60298834,
|
| 3129 |
+
"memory(GiB)": 16.34,
|
| 3130 |
+
"step": 1430,
|
| 3131 |
+
"train_speed(iter/s)": 0.13994
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"acc": 0.83143454,
|
| 3135 |
+
"epoch": 3.888888888888889,
|
| 3136 |
+
"grad_norm": 1.9453125,
|
| 3137 |
+
"learning_rate": 4.386590584878744e-06,
|
| 3138 |
+
"loss": 0.61832671,
|
| 3139 |
+
"memory(GiB)": 16.34,
|
| 3140 |
+
"step": 1435,
|
| 3141 |
+
"train_speed(iter/s)": 0.140003
|
| 3142 |
+
},
|
| 3143 |
+
{
|
| 3144 |
+
"acc": 0.83209105,
|
| 3145 |
+
"epoch": 3.902439024390244,
|
| 3146 |
+
"grad_norm": 3.265625,
|
| 3147 |
+
"learning_rate": 3.851640513552068e-06,
|
| 3148 |
+
"loss": 0.6038147,
|
| 3149 |
+
"memory(GiB)": 16.34,
|
| 3150 |
+
"step": 1440,
|
| 3151 |
+
"train_speed(iter/s)": 0.140065
|
| 3152 |
+
},
|
| 3153 |
+
{
|
| 3154 |
+
"acc": 0.79715924,
|
| 3155 |
+
"epoch": 3.915989159891599,
|
| 3156 |
+
"grad_norm": 2.015625,
|
| 3157 |
+
"learning_rate": 3.316690442225392e-06,
|
| 3158 |
+
"loss": 0.70867634,
|
| 3159 |
+
"memory(GiB)": 16.34,
|
| 3160 |
+
"step": 1445,
|
| 3161 |
+
"train_speed(iter/s)": 0.140125
|
| 3162 |
+
},
|
| 3163 |
+
{
|
| 3164 |
+
"acc": 0.85556803,
|
| 3165 |
+
"epoch": 3.9295392953929538,
|
| 3166 |
+
"grad_norm": 1.7734375,
|
| 3167 |
+
"learning_rate": 2.781740370898716e-06,
|
| 3168 |
+
"loss": 0.52723417,
|
| 3169 |
+
"memory(GiB)": 16.34,
|
| 3170 |
+
"step": 1450,
|
| 3171 |
+
"train_speed(iter/s)": 0.140182
|
| 3172 |
+
},
|
| 3173 |
+
{
|
| 3174 |
+
"epoch": 3.9295392953929538,
|
| 3175 |
+
"eval_acc": 0.6029386927883803,
|
| 3176 |
+
"eval_loss": 1.8965932130813599,
|
| 3177 |
+
"eval_runtime": 44.4125,
|
| 3178 |
+
"eval_samples_per_second": 0.856,
|
| 3179 |
+
"eval_steps_per_second": 0.856,
|
| 3180 |
+
"step": 1450
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"acc": 0.83306837,
|
| 3184 |
+
"epoch": 3.943089430894309,
|
| 3185 |
+
"grad_norm": 2.25,
|
| 3186 |
+
"learning_rate": 2.2467902995720398e-06,
|
| 3187 |
+
"loss": 0.55906692,
|
| 3188 |
+
"memory(GiB)": 16.34,
|
| 3189 |
+
"step": 1455,
|
| 3190 |
+
"train_speed(iter/s)": 0.139645
|
| 3191 |
+
},
|
| 3192 |
+
{
|
| 3193 |
+
"acc": 0.81574478,
|
| 3194 |
+
"epoch": 3.9566395663956637,
|
| 3195 |
+
"grad_norm": 2.6875,
|
| 3196 |
+
"learning_rate": 1.7118402282453637e-06,
|
| 3197 |
+
"loss": 0.62746248,
|
| 3198 |
+
"memory(GiB)": 16.34,
|
| 3199 |
+
"step": 1460,
|
| 3200 |
+
"train_speed(iter/s)": 0.139708
|
| 3201 |
+
},
|
| 3202 |
+
{
|
| 3203 |
+
"acc": 0.83905201,
|
| 3204 |
+
"epoch": 3.970189701897019,
|
| 3205 |
+
"grad_norm": 2.46875,
|
| 3206 |
+
"learning_rate": 1.1768901569186875e-06,
|
| 3207 |
+
"loss": 0.56978045,
|
| 3208 |
+
"memory(GiB)": 16.34,
|
| 3209 |
+
"step": 1465,
|
| 3210 |
+
"train_speed(iter/s)": 0.139771
|
| 3211 |
+
},
|
| 3212 |
+
{
|
| 3213 |
+
"acc": 0.85569115,
|
| 3214 |
+
"epoch": 3.983739837398374,
|
| 3215 |
+
"grad_norm": 1.90625,
|
| 3216 |
+
"learning_rate": 6.419400855920114e-07,
|
| 3217 |
+
"loss": 0.48882861,
|
| 3218 |
+
"memory(GiB)": 16.34,
|
| 3219 |
+
"step": 1470,
|
| 3220 |
+
"train_speed(iter/s)": 0.139834
|
| 3221 |
+
},
|
| 3222 |
+
{
|
| 3223 |
+
"acc": 0.82850809,
|
| 3224 |
+
"epoch": 3.997289972899729,
|
| 3225 |
+
"grad_norm": 1.828125,
|
| 3226 |
+
"learning_rate": 1.0699001426533523e-07,
|
| 3227 |
+
"loss": 0.56869435,
|
| 3228 |
+
"memory(GiB)": 16.34,
|
| 3229 |
+
"step": 1475,
|
| 3230 |
+
"train_speed(iter/s)": 0.139901
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 4.0,
|
| 3234 |
+
"eval_acc": 0.6037831447390644,
|
| 3235 |
+
"eval_loss": 1.8978389501571655,
|
| 3236 |
+
"eval_runtime": 44.2826,
|
| 3237 |
+
"eval_samples_per_second": 0.858,
|
| 3238 |
+
"eval_steps_per_second": 0.858,
|
| 3239 |
+
"step": 1476
|
| 3240 |
+
}
|
| 3241 |
+
],
|
| 3242 |
+
"logging_steps": 5,
|
| 3243 |
+
"max_steps": 1476,
|
| 3244 |
+
"num_input_tokens_seen": 0,
|
| 3245 |
+
"num_train_epochs": 4,
|
| 3246 |
+
"save_steps": 369,
|
| 3247 |
+
"total_flos": 1.991427194376192e+17,
|
| 3248 |
+
"train_batch_size": 1,
|
| 3249 |
+
"trial_name": null,
|
| 3250 |
+
"trial_params": null
|
| 3251 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e164dfb4f067ebefb257c06a7010f5749952eef991c3155070cfd3ae753eb94f
|
| 3 |
+
size 7032
|