Training in progress, step 10

adapter_config.json

@@ -12,24 +12,24 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 256,
+  "lora_alpha": 16,
   "lora_bias": false,
   "lora_dropout": 0.1,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": [],
   "peft_type": "LORA",
-  "r": 128,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "k_proj",
     "up_proj",
     "down_proj",
+    "o_proj",
+    "v_proj",
     "gate_proj",
-    "q_proj",
-    "o_proj"
+    "k_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ba03733657146eaf762aa7b0c7782390f6c3f3d56e26f628ba0dd64ac113363
-size 1478569432
+oid sha256:be94cd510bdedcc178334ba545ff7f140b03d4078136c60121d6282b1d1cff45
+size 369699136

args.json

@@ -78,7 +78,7 @@
   "custom_register_path": [],
   "ignore_args_error": false,
   "use_swift_lora": false,
-  "output_dir": "/root/dataDisk/ms-swift/output/v39-20250329-173701",
+  "output_dir": "/root/dataDisk/ms-swift/output/v40-20250329-182358",
   "overwrite_output_dir": false,
   "do_train": false,
   "do_eval": false,
@@ -99,7 +99,7 @@
   "adam_beta2": 0.999,
   "adam_epsilon": 1e-08,
   "max_grad_norm": 1.0,
-  "num_train_epochs": 1.0,
+  "num_train_epochs": 3.0,
   "max_steps": -1,
   "lr_scheduler_type": "linear",
   "lr_scheduler_kwargs": null,
@@ -108,7 +108,7 @@
   "log_level": "passive",
   "log_level_replica": "warning",
   "log_on_each_node": true,
-  "logging_dir": "/root/dataDisk/ms-swift/output/v39-20250329-173701/runs",
+  "logging_dir": "/root/dataDisk/ms-swift/output/v40-20250329-182358/runs",
   "logging_strategy": "steps",
   "logging_first_step": true,
   "logging_steps": 1,
@@ -266,8 +266,8 @@
   ],
   "target_regex": null,
   "modules_to_save": [],
-  "lora_rank": 128,
-  "lora_alpha": 256,
+  "lora_rank": 32,
+  "lora_alpha": 16,
   "lora_dropout": 0.1,
   "lora_bias": "none",
   "lora_dtype": null,
@@ -351,8 +351,8 @@
   "local_world_size": 8,
   "model_suffix": "mistral-agentic-turbo",
   "model_info": "ModelInfo(model_type='mistral_2501', model_dir='/root/highspeedstorage/ft-volume/mistral-agentic-turbo', torch_dtype=torch.bfloat16, max_model_len=131072, quant_method=None, quant_bits=None, rope_scaling={'factor': 4.0, 'original_max_position_embeddings': 32768, 'type': 'dynamic'}, config=None, task_type='causal_lm', num_labels=None)",
-  "model_meta": "ModelMeta(model_type='mistral_2501', model_groups=[ModelGroup(models=[Model(ms_model_id='mistralai/Mistral-Small-24B-Base-2501', hf_model_id='mistralai/Mistral-Small-24B-Base-2501', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='mistralai/Mistral-Small-24B-Instruct-2501', hf_model_id='mistralai/Mistral-Small-24B-Instruct-2501', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='mistral_2501', get_function=<function get_model_tokenizer_with_flash_attn at 0x7fdeb5812c20>, model_arch='llama', architectures=['MistralForCausalLM'], additional_saved_files=[], torch_dtype=None, is_multimodal=False, is_reward=False, task_type=None, ignore_patterns=[], requires=[], tags=[])",
+  "model_meta": "ModelMeta(model_type='mistral_2501', model_groups=[ModelGroup(models=[Model(ms_model_id='mistralai/Mistral-Small-24B-Base-2501', hf_model_id='mistralai/Mistral-Small-24B-Base-2501', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='mistralai/Mistral-Small-24B-Instruct-2501', hf_model_id='mistralai/Mistral-Small-24B-Instruct-2501', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='mistral_2501', get_function=<function get_model_tokenizer_with_flash_attn at 0x7fc523c62c20>, model_arch='llama', architectures=['MistralForCausalLM'], additional_saved_files=[], torch_dtype=None, is_multimodal=False, is_reward=False, task_type=None, ignore_patterns=[], requires=[], tags=[])",
   "model_dir": "/root/highspeedstorage/ft-volume/mistral-agentic-turbo",
   "hub": "<class 'swift.hub.hub.HFHub'>",
-  "training_args": "Seq2SeqTrainingArguments(output_dir='/root/dataDisk/ms-swift/output/v39-20250329-173701', overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_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=8, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.01, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=1.0, max_steps=-1, lr_scheduler_type=<SchedulerType.LINEAR: 'linear'>, lr_scheduler_kwargs=None, warmup_ratio=0.1, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/root/dataDisk/ms-swift/output/v39-20250329-173701/runs', logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=1, logging_nan_inf_filter=True, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=10, save_total_limit=2, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, 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=10, dataloader_num_workers=0, dataloader_prefetch_factor=None, past_index=-1, run_name='/root/dataDisk/ms-swift/output/v39-20250329-173701', 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}, tp_size=0, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 3, 'offload_optimizer': {'device': 'cpu', 'pin_memory': True}, 'offload_param': {'device': 'cpu', 'pin_memory': True}, 'overlap_comm': True, 'contiguous_gradients': True, 'sub_group_size': 1000000000.0, 'reduce_bucket_size': 'auto', 'stage3_prefetch_bucket_size': 'auto', 'stage3_param_persistence_threshold': 'auto', 'stage3_max_live_parameters': 1000000000.0, 'stage3_max_reuse_distance': 1000000000.0, 'stage3_gather_16bit_weights_on_model_save': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, 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=True, resume_from_checkpoint=None, hub_model_id='TheAgenticAI/LLAMA-3.3-70B-Reasoning', hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=True, hub_always_push=False, gradient_checkpointing=True, gradient_checkpointing_kwargs={'use_reentrant': True}, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', evaluation_strategy='steps', 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=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, acc_strategy='token', sequence_parallel_size=1, check_model=True, train_sampler_random=True, is_encoder_decoder=False, metric_warmup_step=0, train_dataset_sample=-1, fsdp_num=1, acc_steps=1, train_type='lora', optimizer=None, local_repo_path=None, galore_config=None)"
+  "training_args": "Seq2SeqTrainingArguments(output_dir='/root/dataDisk/ms-swift/output/v40-20250329-182358', overwrite_output_dir=False, do_train=False, do_eval=True, do_predict=False, eval_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=8, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=1e-05, weight_decay=0.01, adam_beta1=0.9, adam_beta2=0.999, adam_epsilon=1e-08, max_grad_norm=1.0, num_train_epochs=3.0, max_steps=-1, lr_scheduler_type=<SchedulerType.LINEAR: 'linear'>, lr_scheduler_kwargs=None, warmup_ratio=0.1, warmup_steps=0, log_level='passive', log_level_replica='warning', log_on_each_node=True, logging_dir='/root/dataDisk/ms-swift/output/v40-20250329-182358/runs', logging_strategy=<IntervalStrategy.STEPS: 'steps'>, logging_first_step=True, logging_steps=1, logging_nan_inf_filter=True, save_strategy=<SaveStrategy.STEPS: 'steps'>, save_steps=10, save_total_limit=2, save_safetensors=True, save_on_each_node=False, save_only_model=False, restore_callback_states_from_checkpoint=False, no_cuda=False, use_cpu=False, use_mps_device=False, seed=42, data_seed=42, 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=10, dataloader_num_workers=0, dataloader_prefetch_factor=None, past_index=-1, run_name='/root/dataDisk/ms-swift/output/v40-20250329-182358', 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}, tp_size=0, fsdp_transformer_layer_cls_to_wrap=None, accelerator_config=AcceleratorConfig(split_batches=False, dispatch_batches=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), deepspeed={'fp16': {'enabled': 'auto', 'loss_scale': 0, 'loss_scale_window': 1000, 'initial_scale_power': 16, 'hysteresis': 2, 'min_loss_scale': 1}, 'bf16': {'enabled': 'auto'}, 'zero_optimization': {'stage': 3, 'offload_optimizer': {'device': 'cpu', 'pin_memory': True}, 'offload_param': {'device': 'cpu', 'pin_memory': True}, 'overlap_comm': True, 'contiguous_gradients': True, 'sub_group_size': 1000000000.0, 'reduce_bucket_size': 'auto', 'stage3_prefetch_bucket_size': 'auto', 'stage3_param_persistence_threshold': 'auto', 'stage3_max_live_parameters': 1000000000.0, 'stage3_max_reuse_distance': 1000000000.0, 'stage3_gather_16bit_weights_on_model_save': True}, 'gradient_accumulation_steps': 'auto', 'gradient_clipping': 'auto', 'steps_per_print': 2000, 'train_batch_size': 'auto', 'train_micro_batch_size_per_gpu': 'auto', 'wall_clock_breakdown': False}, 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=True, resume_from_checkpoint=None, hub_model_id='TheAgenticAI/LLAMA-3.3-70B-Reasoning', hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=True, hub_always_push=False, gradient_checkpointing=True, gradient_checkpointing_kwargs={'use_reentrant': True}, include_inputs_for_metrics=False, include_for_metrics=[], eval_do_concat_batches=True, fp16_backend='auto', evaluation_strategy='steps', 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=None, include_num_input_tokens_seen=None, neftune_noise_alpha=None, optim_target_modules=None, batch_eval_metrics=False, eval_on_start=False, use_liger_kernel=False, eval_use_gather_object=False, average_tokens_across_devices=None, sortish_sampler=False, predict_with_generate=False, generation_max_length=None, generation_num_beams=None, generation_config=None, acc_strategy='token', sequence_parallel_size=1, check_model=True, train_sampler_random=True, is_encoder_decoder=False, metric_warmup_step=0, train_dataset_sample=-1, fsdp_num=1, acc_steps=1, train_type='lora', optimizer=None, local_repo_path=None, galore_config=None)"
 }

logging.jsonl

@@ -1,16 +1,11 @@
-{"loss": 0.4534499, "token_acc": 0.85207855, "grad_norm": 13.34440231, "learning_rate": 5e-06, "memory(GiB)": 59.77, "train_speed(iter/s)": 0.017298, "epoch": 0.0733945, "global_step/max_steps": "1/13", "percentage": "7.69%", "elapsed_time": "50s", "remaining_time": "10m 4s"}
-{"loss": 0.48438707, "token_acc": 0.84307111, "grad_norm": 14.0229454, "learning_rate": 1e-05, "memory(GiB)": 97.94, "train_speed(iter/s)": 0.022243, "epoch": 0.14678899, "global_step/max_steps": "2/13", "percentage": "15.38%", "elapsed_time": "1m 22s", "remaining_time": "7m 33s"}
-{"loss": 0.39558721, "token_acc": 0.87623521, "grad_norm": 17.53300667, "learning_rate": 9.09e-06, "memory(GiB)": 100.44, "train_speed(iter/s)": 0.023304, "epoch": 0.22018349, "global_step/max_steps": "3/13", "percentage": "23.08%", "elapsed_time": "2m 1s", "remaining_time": "6m 44s"}
-{"loss": 0.36164382, "token_acc": 0.88360022, "grad_norm": 8.0753746, "learning_rate": 8.18e-06, "memory(GiB)": 100.44, "train_speed(iter/s)": 0.024548, "epoch": 0.29357798, "global_step/max_steps": "4/13", "percentage": "30.77%", "elapsed_time": "2m 35s", "remaining_time": "5m 49s"}
-{"loss": 0.28607693, "token_acc": 0.89519958, "grad_norm": 4.6801033, "learning_rate": 7.27e-06, "memory(GiB)": 100.44, "train_speed(iter/s)": 0.02483, "epoch": 0.36697248, "global_step/max_steps": "5/13", "percentage": "38.46%", "elapsed_time": "3m 13s", "remaining_time": "5m 10s"}
-{"loss": 0.24699311, "token_acc": 0.92354802, "grad_norm": 4.40268326, "learning_rate": 6.36e-06, "memory(GiB)": 118.38, "train_speed(iter/s)": 0.024078, "epoch": 0.44036697, "global_step/max_steps": "6/13", "percentage": "46.15%", "elapsed_time": "4m 1s", "remaining_time": "4m 42s"}
-{"loss": 0.20653641, "token_acc": 0.92372492, "grad_norm": 4.65564156, "learning_rate": 5.45e-06, "memory(GiB)": 118.38, "train_speed(iter/s)": 0.024314, "epoch": 0.51376147, "global_step/max_steps": "7/13", "percentage": "53.85%", "elapsed_time": "4m 40s", "remaining_time": "4m 0s"}
-{"loss": 0.17608178, "token_acc": 0.93256823, "grad_norm": 4.18540716, "learning_rate": 4.55e-06, "memory(GiB)": 118.38, "train_speed(iter/s)": 0.024257, "epoch": 0.58715596, "global_step/max_steps": "8/13", "percentage": "61.54%", "elapsed_time": "5m 22s", "remaining_time": "3m 21s"}
-{"loss": 0.17281821, "token_acc": 0.94667173, "grad_norm": 4.15249395, "learning_rate": 3.64e-06, "memory(GiB)": 118.38, "train_speed(iter/s)": 0.024331, "epoch": 0.66055046, "global_step/max_steps": "9/13", "percentage": "69.23%", "elapsed_time": "6m 2s", "remaining_time": "2m 41s"}
-{"loss": 0.1793212, "token_acc": 0.96042697, "grad_norm": 4.5847702, "learning_rate": 2.73e-06, "memory(GiB)": 131.34, "train_speed(iter/s)": 0.023399, "epoch": 0.73394495, "global_step/max_steps": "10/13", "percentage": "76.92%", "elapsed_time": "6m 59s", "remaining_time": "2m 5s"}
-{"eval_loss": 0.17151441, "eval_token_acc": 0.94895582, "eval_runtime": 5.227, "eval_samples_per_second": 4.974, "eval_steps_per_second": 0.765, "epoch": 0.73394495, "global_step/max_steps": "10/13", "percentage": "76.92%", "elapsed_time": "7m 5s", "remaining_time": "2m 7s"}
-{"loss": 0.15501738, "token_acc": 0.9503762, "grad_norm": 4.98172522, "learning_rate": 1.82e-06, "memory(GiB)": 131.34, "train_speed(iter/s)": 0.022982, "epoch": 0.80733945, "global_step/max_steps": "11/13", "percentage": "84.62%", "elapsed_time": "7m 51s", "remaining_time": "1m 25s"}
-{"loss": 0.14941445, "token_acc": 0.94386522, "grad_norm": 3.96140695, "learning_rate": 9.1e-07, "memory(GiB)": 131.34, "train_speed(iter/s)": 0.023428, "epoch": 0.88073394, "global_step/max_steps": "12/13", "percentage": "92.31%", "elapsed_time": "8m 24s", "remaining_time": "42s"}
-{"loss": 0.16274551, "token_acc": 0.94945698, "grad_norm": 4.06107044, "learning_rate": 0.0, "memory(GiB)": 131.34, "train_speed(iter/s)": 0.023585, "epoch": 0.95412844, "global_step/max_steps": "13/13", "percentage": "100.00%", "elapsed_time": "9m 3s", "remaining_time": "0s"}
-{"eval_loss": 0.15031654, "eval_token_acc": 0.95677917, "eval_runtime": 5.2198, "eval_samples_per_second": 4.981, "eval_steps_per_second": 0.766, "epoch": 0.95412844, "global_step/max_steps": "13/13", "percentage": "100.00%", "elapsed_time": "9m 8s", "remaining_time": "0s"}
-{"train_runtime": 555.0275, "train_samples_per_second": 1.567, "train_steps_per_second": 0.023, "total_flos": 14280794439680.0, "train_loss": 0.26385177, "epoch": 0.95412844, "global_step/max_steps": "13/13", "percentage": "100.00%", "elapsed_time": "9m 15s", "remaining_time": "0s"}
+{"loss": 0.4534499, "token_acc": 0.85207855, "grad_norm": 0.82650959, "learning_rate": 2.5e-06, "memory(GiB)": 61.41, "train_speed(iter/s)": 0.017978, "epoch": 0.0733945, "global_step/max_steps": "1/39", "percentage": "2.56%", "elapsed_time": "49s", "remaining_time": "31m 30s"}
+{"loss": 0.48438707, "token_acc": 0.84307111, "grad_norm": 0.8687833, "learning_rate": 5e-06, "memory(GiB)": 97.64, "train_speed(iter/s)": 0.022952, "epoch": 0.14678899, "global_step/max_steps": "2/39", "percentage": "5.13%", "elapsed_time": "1m 21s", "remaining_time": "25m 3s"}
+{"loss": 0.46338826, "token_acc": 0.85836282, "grad_norm": 0.87502807, "learning_rate": 7.5e-06, "memory(GiB)": 97.64, "train_speed(iter/s)": 0.023986, "epoch": 0.22018349, "global_step/max_steps": "3/39", "percentage": "7.69%", "elapsed_time": "1m 59s", "remaining_time": "23m 50s"}
+{"loss": 0.46354297, "token_acc": 0.86267256, "grad_norm": 0.79912186, "learning_rate": 1e-05, "memory(GiB)": 98.89, "train_speed(iter/s)": 0.02525, "epoch": 0.29357798, "global_step/max_steps": "4/39", "percentage": "10.26%", "elapsed_time": "2m 32s", "remaining_time": "22m 14s"}
+{"loss": 0.42857072, "token_acc": 0.84789872, "grad_norm": 0.75490576, "learning_rate": 9.71e-06, "memory(GiB)": 98.89, "train_speed(iter/s)": 0.025505, "epoch": 0.36697248, "global_step/max_steps": "5/39", "percentage": "12.82%", "elapsed_time": "3m 10s", "remaining_time": "21m 33s"}
+{"loss": 0.41332674, "token_acc": 0.86822497, "grad_norm": 0.746373, "learning_rate": 9.43e-06, "memory(GiB)": 116.84, "train_speed(iter/s)": 0.024698, "epoch": 0.44036697, "global_step/max_steps": "6/39", "percentage": "15.38%", "elapsed_time": "3m 57s", "remaining_time": "21m 43s"}
+{"loss": 0.41638315, "token_acc": 0.85066626, "grad_norm": 0.70349604, "learning_rate": 9.14e-06, "memory(GiB)": 116.84, "train_speed(iter/s)": 0.024925, "epoch": 0.51376147, "global_step/max_steps": "7/39", "percentage": "17.95%", "elapsed_time": "4m 34s", "remaining_time": "20m 56s"}
+{"loss": 0.36732465, "token_acc": 0.8665636, "grad_norm": 0.59028256, "learning_rate": 8.86e-06, "memory(GiB)": 116.84, "train_speed(iter/s)": 0.024855, "epoch": 0.58715596, "global_step/max_steps": "8/39", "percentage": "20.51%", "elapsed_time": "5m 15s", "remaining_time": "20m 24s"}
+{"loss": 0.35201401, "token_acc": 0.89277345, "grad_norm": 0.55272579, "learning_rate": 8.57e-06, "memory(GiB)": 118.09, "train_speed(iter/s)": 0.024923, "epoch": 0.66055046, "global_step/max_steps": "9/39", "percentage": "23.08%", "elapsed_time": "5m 55s", "remaining_time": "19m 44s"}
+{"loss": 0.37256077, "token_acc": 0.89065348, "grad_norm": 0.80991906, "learning_rate": 8.29e-06, "memory(GiB)": 131.04, "train_speed(iter/s)": 0.023943, "epoch": 0.73394495, "global_step/max_steps": "10/39", "percentage": "25.64%", "elapsed_time": "6m 51s", "remaining_time": "19m 54s"}
+{"eval_loss": 0.41130996, "eval_token_acc": 0.87739136, "eval_runtime": 5.218, "eval_samples_per_second": 4.983, "eval_steps_per_second": 0.767, "epoch": 0.73394495, "global_step/max_steps": "10/39", "percentage": "25.64%", "elapsed_time": "6m 57s", "remaining_time": "20m 9s"}

runs/events.out.tfevents.1743272706.notebook-67532d59-da58-11ef-a92b-d22a634ff48d-0.273350.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae0a7672f995088369cf3fc1f1bc28547955abe9723402ed5b66539f6a394156
+size 10951

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:704ebc33ac965782c742b0bc0cd5f2835ccacc8ae48ea2301abf1c4c74d4b18c
+oid sha256:f23d182131a82fe05dbe7c6c7e4f6b7e2c0cba0e7ceb1a0b479f09bea96b1c2c
 size 8248