Training in progress, step 24

adapter_config.json

@@ -24,12 +24,12 @@
   "revision": null,
   "target_modules": [
     "q_proj",
-    "down_proj",
+    "k_proj",
     "o_proj",
+    "gate_proj",
     "v_proj",
     "up_proj",
-    "k_proj",
-    "gate_proj"
+    "down_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:aed37a42eb32c018b99f8a2ea54f2d721d1910bc576a33053a7beb1601e0530a
+oid sha256:2fb54ace7b34bb3f78cb3eec551559350e742d0e43028ab639989fa4c52acc8a
 size 739322344

args.json

@@ -6,6 +6,7 @@
   "torch_dtype": "bfloat16",
   "attn_impl": null,
   "num_labels": null,
+  "problem_type": null,
   "rope_scaling": null,
   "device_map": null,
   "max_memory": {},
@@ -24,7 +25,7 @@
   "use_chat_template": true,
   "template_backend": "swift",
   "dataset": [
-    "/root/dataDisk/ms-swift/all_processed_logs4.jsonl"
+    "/root/dataDisk/ms-swift/mixed_logs_new.jsonl"
   ],
   "val_dataset": [],
   "split_dataset_ratio": 0.03,
@@ -78,7 +79,7 @@
   "custom_register_path": [],
   "ignore_args_error": false,
   "use_swift_lora": false,
-  "output_dir": "/root/dataDisk/ms-swift/output/v64-20250408-042513",
+  "output_dir": "/root/dataDisk/ms-swift/output/v86-20250409-074407",
   "overwrite_output_dir": false,
   "do_train": false,
   "do_eval": false,
@@ -96,7 +97,7 @@
   "learning_rate": 1e-05,
   "weight_decay": 0.01,
   "adam_beta1": 0.9,
-  "adam_beta2": 0.999,
+  "adam_beta2": 0.95,
   "adam_epsilon": 1e-08,
   "max_grad_norm": 1.0,
   "num_train_epochs": 1.0,
@@ -108,7 +109,7 @@
   "log_level": "passive",
   "log_level_replica": "warning",
   "log_on_each_node": true,
-  "logging_dir": "/root/dataDisk/ms-swift/output/v64-20250408-042513/runs",
+  "logging_dir": "/root/dataDisk/ms-swift/output/v86-20250409-074407/runs",
   "logging_strategy": "steps",
   "logging_first_step": true,
   "logging_steps": 1,
@@ -179,7 +180,7 @@
         "device": "cpu",
         "pin_memory": true
       },
-      "overlap_comm": true,
+      "overlap_comm": false,
       "contiguous_gradients": true,
       "sub_group_size": 1000000000.0,
       "reduce_bucket_size": "auto",
@@ -219,9 +220,7 @@
   "hub_private_repo": true,
   "hub_always_push": false,
   "gradient_checkpointing": true,
-  "gradient_checkpointing_kwargs": {
-    "use_reentrant": true
-  },
+  "gradient_checkpointing_kwargs": "{\"use_reentrant\": true}",
   "include_inputs_for_metrics": false,
   "include_for_metrics": [],
   "eval_do_concat_batches": true,
@@ -255,6 +254,17 @@
   "generation_max_length": null,
   "generation_num_beams": null,
   "generation_config": null,
+  "check_model": true,
+  "acc_strategy": "token",
+  "train_sampler_random": true,
+  "metric_warmup_step": 0,
+  "fsdp_num": 1,
+  "acc_steps": 1,
+  "eval_use_evalscope": false,
+  "eval_datasets": [],
+  "eval_limit": null,
+  "eval_datasets_args": null,
+  "eval_generation_config": null,
   "freeze_parameters": [],
   "freeze_parameters_ratio": 0.0,
   "trainable_parameters": [],
@@ -326,10 +336,6 @@
   "reft_intervention_type": "LoreftIntervention",
   "reft_args": null,
   "use_liger": false,
-  "model_layer_cls_name": null,
-  "metric_warmup_step": 0,
-  "fsdp_num": 1,
-  "acc_steps": 1,
   "swanlab_token": null,
   "swanlab_project": null,
   "swanlab_workspace": null,
@@ -337,22 +343,20 @@
   "swanlab_mode": "cloud",
   "add_version": true,
   "resume_only_model": false,
-  "check_model": true,
   "create_checkpoint_symlink": false,
   "packing": false,
   "lazy_tokenize": false,
   "loss_type": null,
   "optimizer": null,
   "metric": null,
-  "acc_strategy": "token",
   "zero_hpz_partition_size": null,
   "rank": 0,
   "global_world_size": 8,
   "local_world_size": 8,
   "model_suffix": "mistral-4-2-2025",
   "model_info": "ModelInfo(model_type='mistral_2501', model_dir='/root/highspeedstorage/ft-volume/mistral-4-2-2025', 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 0x7ff4b04223b0>, 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 0x7ff14c4d28c0>, 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-4-2-2025",
   "hub": "<class 'swift.hub.hub.HFHub'>",
-  "training_args": "Seq2SeqTrainingArguments(output_dir='/root/dataDisk/ms-swift/output/v64-20250408-042513', 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/v64-20250408-042513/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/v64-20250408-042513', 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/v86-20250409-074407', 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=0.95, 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/v86-20250409-074407/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/v86-20250409-074407', 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': False, '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, check_model=True, acc_strategy='token', train_sampler_random=True, metric_warmup_step=0, fsdp_num=1, acc_steps=1, eval_use_evalscope=False, eval_datasets=[], eval_limit=None, eval_datasets_args=None, eval_generation_config=None, train_type='lora', optimizer=None, local_repo_path=None, galore_config=None)"
 }

logging.jsonl

@@ -1,33 +1,28 @@
-{"loss": 0.4804574, "token_acc": 0.8626087, "grad_norm": 11.75211048, "learning_rate": 3.33e-06, "memory(GiB)": 65.07, "train_speed(iter/s)": 0.016335, "epoch": 0.03404255, "global_step/max_steps": "1/29", "percentage": "3.45%", "elapsed_time": "54s", "remaining_time": "25m 21s"}
-{"loss": 0.45759296, "token_acc": 0.85855502, "grad_norm": 10.19118881, "learning_rate": 6.67e-06, "memory(GiB)": 66.32, "train_speed(iter/s)": 0.018782, "epoch": 0.06808511, "global_step/max_steps": "2/29", "percentage": "6.90%", "elapsed_time": "1m 39s", "remaining_time": "22m 24s"}
-{"loss": 0.38502353, "token_acc": 0.89037485, "grad_norm": 6.74299908, "learning_rate": 1e-05, "memory(GiB)": 89.3, "train_speed(iter/s)": 0.02022, "epoch": 0.10212766, "global_step/max_steps": "3/29", "percentage": "10.34%", "elapsed_time": "2m 21s", "remaining_time": "20m 26s"}
-{"loss": 0.34768325, "token_acc": 0.89727589, "grad_norm": 6.76169825, "learning_rate": 9.62e-06, "memory(GiB)": 89.3, "train_speed(iter/s)": 0.020858, "epoch": 0.13617021, "global_step/max_steps": "4/29", "percentage": "13.79%", "elapsed_time": "3m 4s", "remaining_time": "19m 15s"}
-{"loss": 0.34051374, "token_acc": 0.91092169, "grad_norm": 4.81446934, "learning_rate": 9.23e-06, "memory(GiB)": 89.3, "train_speed(iter/s)": 0.021584, "epoch": 0.17021277, "global_step/max_steps": "5/29", "percentage": "17.24%", "elapsed_time": "3m 44s", "remaining_time": "17m 58s"}
-{"loss": 0.30721647, "token_acc": 0.90557205, "grad_norm": 3.24823046, "learning_rate": 8.85e-06, "memory(GiB)": 90.55, "train_speed(iter/s)": 0.022381, "epoch": 0.20425532, "global_step/max_steps": "6/29", "percentage": "20.69%", "elapsed_time": "4m 21s", "remaining_time": "16m 41s"}
-{"loss": 0.26436692, "token_acc": 0.90980206, "grad_norm": 2.52645493, "learning_rate": 8.46e-06, "memory(GiB)": 90.55, "train_speed(iter/s)": 0.022355, "epoch": 0.23829787, "global_step/max_steps": "7/29", "percentage": "24.14%", "elapsed_time": "5m 6s", "remaining_time": "16m 2s"}
-{"loss": 0.26473376, "token_acc": 0.92074431, "grad_norm": 3.90788126, "learning_rate": 8.08e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022451, "epoch": 0.27234043, "global_step/max_steps": "8/29", "percentage": "27.59%", "elapsed_time": "5m 49s", "remaining_time": "15m 17s"}
-{"loss": 0.25180489, "token_acc": 0.92015499, "grad_norm": 3.09036565, "learning_rate": 7.69e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022339, "epoch": 0.30638298, "global_step/max_steps": "9/29", "percentage": "31.03%", "elapsed_time": "6m 35s", "remaining_time": "14m 39s"}
-{"loss": 0.25393221, "token_acc": 0.9164224, "grad_norm": 2.27255797, "learning_rate": 7.31e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022295, "epoch": 0.34042553, "global_step/max_steps": "10/29", "percentage": "34.48%", "elapsed_time": "7m 21s", "remaining_time": "13m 59s"}
-{"eval_loss": 0.24552706, "eval_token_acc": 0.91472783, "eval_runtime": 13.041, "eval_samples_per_second": 4.371, "eval_steps_per_second": 0.613, "epoch": 0.34042553, "global_step/max_steps": "10/29", "percentage": "34.48%", "elapsed_time": "7m 34s", "remaining_time": "14m 23s"}
-{"loss": 0.25681421, "token_acc": 0.91740926, "grad_norm": 2.15936017, "learning_rate": 6.92e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.021734, "epoch": 0.37446809, "global_step/max_steps": "11/29", "percentage": "37.93%", "elapsed_time": "8m 19s", "remaining_time": "13m 36s"}
-{"loss": 0.24919307, "token_acc": 0.91749503, "grad_norm": 2.12157631, "learning_rate": 6.54e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022075, "epoch": 0.40851064, "global_step/max_steps": "12/29", "percentage": "41.38%", "elapsed_time": "8m 56s", "remaining_time": "12m 40s"}
-{"loss": 0.23583913, "token_acc": 0.91674937, "grad_norm": 1.95219505, "learning_rate": 6.15e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022257, "epoch": 0.44255319, "global_step/max_steps": "13/29", "percentage": "44.83%", "elapsed_time": "9m 37s", "remaining_time": "11m 50s"}
-{"loss": 0.23160429, "token_acc": 0.90663566, "grad_norm": 2.19393158, "learning_rate": 5.77e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022163, "epoch": 0.47659574, "global_step/max_steps": "14/29", "percentage": "48.28%", "elapsed_time": "10m 24s", "remaining_time": "11m 9s"}
-{"loss": 0.21596812, "token_acc": 0.9109509, "grad_norm": 2.00559306, "learning_rate": 5.38e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022383, "epoch": 0.5106383, "global_step/max_steps": "15/29", "percentage": "51.72%", "elapsed_time": "11m 3s", "remaining_time": "10m 19s"}
-{"loss": 0.21437249, "token_acc": 0.93086992, "grad_norm": 1.88266635, "learning_rate": 5e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.0223, "epoch": 0.54468085, "global_step/max_steps": "16/29", "percentage": "55.17%", "elapsed_time": "11m 50s", "remaining_time": "9m 37s"}
-{"loss": 0.21812737, "token_acc": 0.9273279, "grad_norm": 3.35490966, "learning_rate": 4.62e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022206, "epoch": 0.5787234, "global_step/max_steps": "17/29", "percentage": "58.62%", "elapsed_time": "12m 38s", "remaining_time": "8m 55s"}
-{"loss": 0.22592279, "token_acc": 0.91987076, "grad_norm": 2.69795418, "learning_rate": 4.23e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022245, "epoch": 0.61276596, "global_step/max_steps": "18/29", "percentage": "62.07%", "elapsed_time": "13m 22s", "remaining_time": "8m 10s"}
-{"loss": 0.21260472, "token_acc": 0.9241944, "grad_norm": 2.05847311, "learning_rate": 3.85e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022284, "epoch": 0.64680851, "global_step/max_steps": "19/29", "percentage": "65.52%", "elapsed_time": "14m 5s", "remaining_time": "7m 25s"}
-{"loss": 0.23483819, "token_acc": 0.90957684, "grad_norm": 1.9506942, "learning_rate": 3.46e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022228, "epoch": 0.68085106, "global_step/max_steps": "20/29", "percentage": "68.97%", "elapsed_time": "14m 52s", "remaining_time": "6m 41s"}
-{"eval_loss": 0.20867875, "eval_token_acc": 0.92511696, "eval_runtime": 13.0105, "eval_samples_per_second": 4.381, "eval_steps_per_second": 0.615, "epoch": 0.68085106, "global_step/max_steps": "20/29", "percentage": "68.97%", "elapsed_time": "15m 5s", "remaining_time": "6m 47s"}
-{"loss": 0.20101674, "token_acc": 0.92574786, "grad_norm": 1.76422858, "learning_rate": 3.08e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.02181, "epoch": 0.71489362, "global_step/max_steps": "21/29", "percentage": "72.41%", "elapsed_time": "15m 55s", "remaining_time": "6m 4s"}
-{"loss": 0.20781831, "token_acc": 0.93246598, "grad_norm": 2.15241098, "learning_rate": 2.69e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.0219, "epoch": 0.74893617, "global_step/max_steps": "22/29", "percentage": "75.86%", "elapsed_time": "16m 37s", "remaining_time": "5m 17s"}
-{"loss": 0.20801759, "token_acc": 0.93892848, "grad_norm": 2.22540712, "learning_rate": 2.31e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.021942, "epoch": 0.78297872, "global_step/max_steps": "23/29", "percentage": "79.31%", "elapsed_time": "17m 21s", "remaining_time": "4m 31s"}
-{"loss": 0.18198422, "token_acc": 0.94633731, "grad_norm": 1.63598311, "learning_rate": 1.92e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.021974, "epoch": 0.81702128, "global_step/max_steps": "24/29", "percentage": "82.76%", "elapsed_time": "18m 5s", "remaining_time": "3m 46s"}
-{"loss": 0.19455627, "token_acc": 0.92534722, "grad_norm": 2.62162328, "learning_rate": 1.54e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.02211, "epoch": 0.85106383, "global_step/max_steps": "25/29", "percentage": "86.21%", "elapsed_time": "18m 43s", "remaining_time": "2m 59s"}
-{"loss": 0.19514789, "token_acc": 0.94586842, "grad_norm": 1.7049104, "learning_rate": 1.15e-06, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022122, "epoch": 0.88510638, "global_step/max_steps": "26/29", "percentage": "89.66%", "elapsed_time": "19m 28s", "remaining_time": "2m 14s"}
-{"loss": 0.19851573, "token_acc": 0.91834848, "grad_norm": 1.9396106, "learning_rate": 7.7e-07, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022067, "epoch": 0.91914894, "global_step/max_steps": "27/29", "percentage": "93.10%", "elapsed_time": "20m 16s", "remaining_time": "1m 30s"}
-{"loss": 0.18254727, "token_acc": 0.93301435, "grad_norm": 1.87893045, "learning_rate": 3.8e-07, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022131, "epoch": 0.95319149, "global_step/max_steps": "28/29", "percentage": "96.55%", "elapsed_time": "20m 58s", "remaining_time": "44s"}
-{"loss": 0.192018, "token_acc": 0.92320897, "grad_norm": 1.8398602, "learning_rate": 0.0, "memory(GiB)": 115.29, "train_speed(iter/s)": 0.022196, "epoch": 0.98723404, "global_step/max_steps": "29/29", "percentage": "100.00%", "elapsed_time": "21m 39s", "remaining_time": "0s"}
-{"eval_loss": 0.19650522, "eval_token_acc": 0.92934145, "eval_runtime": 13.0396, "eval_samples_per_second": 4.371, "eval_steps_per_second": 0.614, "epoch": 0.98723404, "global_step/max_steps": "29/29", "percentage": "100.00%", "elapsed_time": "21m 52s", "remaining_time": "0s"}
-{"train_runtime": 1317.4623, "train_samples_per_second": 1.423, "train_steps_per_second": 0.022, "total_flos": 39523818274816.0, "train_loss": 0.25552522, "epoch": 0.98723404, "global_step/max_steps": "29/29", "percentage": "100.00%", "elapsed_time": "21m 57s", "remaining_time": "0s"}
+{"loss": 0.46143222, "token_acc": 0.86343383, "grad_norm": 9.3185215, "learning_rate": 3.33e-06, "memory(GiB)": 54.87, "train_speed(iter/s)": 0.015237, "epoch": 0.04102564, "global_step/max_steps": "1/24", "percentage": "4.17%", "elapsed_time": "55s", "remaining_time": "21m 7s"}
+{"loss": 0.47108185, "token_acc": 0.85968093, "grad_norm": 8.64310455, "learning_rate": 6.67e-06, "memory(GiB)": 54.87, "train_speed(iter/s)": 0.018995, "epoch": 0.08205128, "global_step/max_steps": "2/24", "percentage": "8.33%", "elapsed_time": "1m 34s", "remaining_time": "17m 22s"}
+{"loss": 0.42076147, "token_acc": 0.85514685, "grad_norm": 5.66566086, "learning_rate": 1e-05, "memory(GiB)": 54.87, "train_speed(iter/s)": 0.020169, "epoch": 0.12307692, "global_step/max_steps": "3/24", "percentage": "12.50%", "elapsed_time": "2m 18s", "remaining_time": "16m 7s"}
+{"loss": 0.37250823, "token_acc": 0.88401355, "grad_norm": 6.63496256, "learning_rate": 9.52e-06, "memory(GiB)": 76.52, "train_speed(iter/s)": 0.019331, "epoch": 0.16410256, "global_step/max_steps": "4/24", "percentage": "16.67%", "elapsed_time": "3m 16s", "remaining_time": "16m 21s"}
+{"loss": 0.34375307, "token_acc": 0.87457379, "grad_norm": 4.0149045, "learning_rate": 9.05e-06, "memory(GiB)": 76.52, "train_speed(iter/s)": 0.019574, "epoch": 0.20512821, "global_step/max_steps": "5/24", "percentage": "20.83%", "elapsed_time": "4m 4s", "remaining_time": "15m 30s"}
+{"loss": 0.35020918, "token_acc": 0.8917227, "grad_norm": 3.02704263, "learning_rate": 8.57e-06, "memory(GiB)": 76.52, "train_speed(iter/s)": 0.020114, "epoch": 0.24615385, "global_step/max_steps": "6/24", "percentage": "25.00%", "elapsed_time": "4m 47s", "remaining_time": "14m 23s"}
+{"loss": 0.31043854, "token_acc": 0.90935031, "grad_norm": 2.69551635, "learning_rate": 8.1e-06, "memory(GiB)": 76.52, "train_speed(iter/s)": 0.02063, "epoch": 0.28717949, "global_step/max_steps": "7/24", "percentage": "29.17%", "elapsed_time": "5m 28s", "remaining_time": "13m 18s"}
+{"loss": 0.3027271, "token_acc": 0.89195308, "grad_norm": 2.27069497, "learning_rate": 7.62e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.020872, "epoch": 0.32820513, "global_step/max_steps": "8/24", "percentage": "33.33%", "elapsed_time": "6m 12s", "remaining_time": "12m 25s"}
+{"loss": 0.28800282, "token_acc": 0.91445856, "grad_norm": 2.58754706, "learning_rate": 7.14e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.020954, "epoch": 0.36923077, "global_step/max_steps": "9/24", "percentage": "37.50%", "elapsed_time": "6m 58s", "remaining_time": "11m 38s"}
+{"loss": 0.28545466, "token_acc": 0.90576357, "grad_norm": 2.16132617, "learning_rate": 6.67e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.021314, "epoch": 0.41025641, "global_step/max_steps": "10/24", "percentage": "41.67%", "elapsed_time": "7m 38s", "remaining_time": "10m 42s"}
+{"eval_loss": 0.29409501, "eval_token_acc": 0.90527162, "eval_runtime": 10.0421, "eval_samples_per_second": 4.78, "eval_steps_per_second": 0.597, "epoch": 0.41025641, "global_step/max_steps": "10/24", "percentage": "41.67%", "elapsed_time": "7m 48s", "remaining_time": "10m 56s"}
+{"loss": 0.29755145, "token_acc": 0.89969161, "grad_norm": 2.20071864, "learning_rate": 6.19e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.02072, "epoch": 0.45128205, "global_step/max_steps": "11/24", "percentage": "45.83%", "elapsed_time": "8m 40s", "remaining_time": "10m 14s"}
+{"loss": 0.28594711, "token_acc": 0.90274928, "grad_norm": 2.26474857, "learning_rate": 5.71e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.020675, "epoch": 0.49230769, "global_step/max_steps": "12/24", "percentage": "50.00%", "elapsed_time": "9m 29s", "remaining_time": "9m 29s"}
+{"loss": 0.27647626, "token_acc": 0.89691489, "grad_norm": 2.19667244, "learning_rate": 5.24e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.020771, "epoch": 0.53333333, "global_step/max_steps": "13/24", "percentage": "54.17%", "elapsed_time": "10m 15s", "remaining_time": "8m 40s"}
+{"loss": 0.28978908, "token_acc": 0.92168009, "grad_norm": 2.21953773, "learning_rate": 4.76e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.020869, "epoch": 0.57435897, "global_step/max_steps": "14/24", "percentage": "58.33%", "elapsed_time": "11m 0s", "remaining_time": "7m 51s"}
+{"loss": 0.25644785, "token_acc": 0.91911697, "grad_norm": 1.99253571, "learning_rate": 4.29e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.021005, "epoch": 0.61538462, "global_step/max_steps": "15/24", "percentage": "62.50%", "elapsed_time": "11m 43s", "remaining_time": "7m 2s"}
+{"loss": 0.25205338, "token_acc": 0.92406639, "grad_norm": 1.91407692, "learning_rate": 3.81e-06, "memory(GiB)": 102.1, "train_speed(iter/s)": 0.020998, "epoch": 0.65641026, "global_step/max_steps": "16/24", "percentage": "66.67%", "elapsed_time": "12m 31s", "remaining_time": "6m 15s"}
+{"loss": 0.25136927, "token_acc": 0.92661265, "grad_norm": 1.94001377, "learning_rate": 3.33e-06, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.02106, "epoch": 0.6974359, "global_step/max_steps": "17/24", "percentage": "70.83%", "elapsed_time": "13m 16s", "remaining_time": "5m 28s"}
+{"loss": 0.27359912, "token_acc": 0.89710177, "grad_norm": 2.43866014, "learning_rate": 2.86e-06, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.021257, "epoch": 0.73846154, "global_step/max_steps": "18/24", "percentage": "75.00%", "elapsed_time": "13m 56s", "remaining_time": "4m 38s"}
+{"loss": 0.23825151, "token_acc": 0.92749731, "grad_norm": 2.5236938, "learning_rate": 2.38e-06, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.021291, "epoch": 0.77948718, "global_step/max_steps": "19/24", "percentage": "79.17%", "elapsed_time": "14m 41s", "remaining_time": "3m 52s"}
+{"loss": 0.25170776, "token_acc": 0.89211481, "grad_norm": 3.14535904, "learning_rate": 1.9e-06, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.021329, "epoch": 0.82051282, "global_step/max_steps": "20/24", "percentage": "83.33%", "elapsed_time": "15m 27s", "remaining_time": "3m 5s"}
+{"eval_loss": 0.26594558, "eval_token_acc": 0.91252329, "eval_runtime": 10.0357, "eval_samples_per_second": 4.783, "eval_steps_per_second": 0.598, "epoch": 0.82051282, "global_step/max_steps": "20/24", "percentage": "83.33%", "elapsed_time": "15m 37s", "remaining_time": "3m 7s"}
+{"loss": 0.23205811, "token_acc": 0.90668009, "grad_norm": 3.71121573, "learning_rate": 1.43e-06, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.020958, "epoch": 0.86153846, "global_step/max_steps": "21/24", "percentage": "87.50%", "elapsed_time": "16m 31s", "remaining_time": "2m 21s"}
+{"loss": 0.21498692, "token_acc": 0.93693359, "grad_norm": 1.96844864, "learning_rate": 9.5e-07, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.020997, "epoch": 0.9025641, "global_step/max_steps": "22/24", "percentage": "91.67%", "elapsed_time": "17m 17s", "remaining_time": "1m 34s"}
+{"loss": 0.26505035, "token_acc": 0.88642392, "grad_norm": 2.83930326, "learning_rate": 4.8e-07, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.021149, "epoch": 0.94358974, "global_step/max_steps": "23/24", "percentage": "95.83%", "elapsed_time": "17m 56s", "remaining_time": "46s"}
+{"loss": 0.21959037, "token_acc": 0.92977191, "grad_norm": 1.96380556, "learning_rate": 0.0, "memory(GiB)": 117.53, "train_speed(iter/s)": 0.021197, "epoch": 0.98461538, "global_step/max_steps": "24/24", "percentage": "100.00%", "elapsed_time": "18m 41s", "remaining_time": "0s"}
+{"eval_loss": 0.26333836, "eval_token_acc": 0.91394805, "eval_runtime": 10.028, "eval_samples_per_second": 4.787, "eval_steps_per_second": 0.598, "epoch": 0.98461538, "global_step/max_steps": "24/24", "percentage": "100.00%", "elapsed_time": "18m 51s", "remaining_time": "0s"}
+{"train_runtime": 1136.4427, "train_samples_per_second": 1.367, "train_steps_per_second": 0.021, "total_flos": 31015965818880.0, "train_loss": 0.30046865, "epoch": 0.98461538, "global_step/max_steps": "24/24", "percentage": "100.00%", "elapsed_time": "18m 56s", "remaining_time": "0s"}

runs/events.out.tfevents.1744184797.notebook-67532d59-da58-11ef-a92b-d22a634ff48d-0.764350.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f925306561322d4eb7eae7769c840b360b5e63315ef17062a5267cd9dced322
+size 17275

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a176fb931638787b2131d3501b67ccad00744282697a5323fe90c1ebc5a55ef7
+oid sha256:78b20a459ec0b059f157a8336084f7a4ac2ab61e83d745ab8217d29a3a2a6529
 size 8248