Training in progress, step 5

adapter_config.json

@@ -10,23 +10,23 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 8,
+  "lora_alpha": 64,
   "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": [],
   "peft_type": "LORA",
-  "r": 4,
+  "r": 32,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
     "q_proj",
-    "up_proj",
+    "k_proj",
     "v_proj",
     "gate_proj",
-    "o_proj",
-    "k_proj",
-    "down_proj"
+    "down_proj",
+    "up_proj",
+    "o_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:58bbf1f598f54c33aeae5460a72ec5b60ed494959fc210d9b2409c26b035e59a
-size 42002136
+oid sha256:6f3bfba78311d985134ad71ce8702aa4cfe5b5ac72273271d93d42a105e23866
+size 335604696

args.json

@@ -71,7 +71,7 @@
   "custom_register_path": [],
   "ignore_args_error": false,
   "use_swift_lora": false,
-  "output_dir": "/workspace/output/v5-20250108-223508",
+  "output_dir": "/workspace/output/v6-20250108-223705",
   "overwrite_output_dir": false,
   "do_train": false,
   "do_eval": false,
@@ -101,7 +101,7 @@
   "log_level": "passive",
   "log_level_replica": "warning",
   "log_on_each_node": true,
-  "logging_dir": "/workspace/output/v5-20250108-223508/runs",
+  "logging_dir": "/workspace/output/v6-20250108-223705/runs",
   "logging_strategy": "steps",
   "logging_first_step": true,
   "logging_steps": 1,
@@ -221,8 +221,8 @@
   ],
   "target_regex": null,
   "modules_to_save": [],
-  "lora_rank": 4,
-  "lora_alpha": 8,
+  "lora_rank": 32,
+  "lora_alpha": 64,
   "lora_dropout": 0.05,
   "lora_bias": "none",
   "lora_dtype": null,
@@ -299,8 +299,8 @@
   "local_world_size": 1,
   "model_suffix": "Llama-3.1-8B-Instruct",
   "model_info": "ModelInfo(model_type='llama3_1', model_dir='/root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B-Instruct/snapshots/0e9e39f249a16976918f6564b8830bc894c89659', torch_dtype=torch.bfloat16, max_model_len=131072, quant_method='bnb', quant_bits=4, config={'factor': 8.0, 'low_freq_factor': 1.0, 'high_freq_factor': 4.0, 'original_max_position_embeddings': 8192, 'rope_type': 'llama3'}, task_type='causal_lm', num_labels=None)",
-  "model_meta": "ModelMeta(model_type='llama3_1', model_groups=[ModelGroup(models=[Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct', hf_model_id='meta-llama/Meta-Llama-3.1-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct', hf_model_id='meta-llama/Meta-Llama-3.1-70B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct', hf_model_id='meta-llama/Meta-Llama-3.1-405B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B', hf_model_id='meta-llama/Meta-Llama-3.1-8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B', hf_model_id='meta-llama/Meta-Llama-3.1-70B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B', hf_model_id='meta-llama/Meta-Llama-3.1-405B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-FP8', hf_model_id='meta-llama/Meta-Llama-3.1-70B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-FP8', hf_model_id='meta-llama/Meta-Llama-3.1-405B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct-BNB-NF4', hf_model_id='hugging-quants/Meta-Llama-3.1-8B-Instruct-BNB-NF4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-bnb-4bit', hf_model_id='unsloth/Meta-Llama-3.1-70B-Instruct-bnb-4bit', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-BNB-NF4', hf_model_id='hugging-quants/Meta-Llama-3.1-405B-Instruct-BNB-NF4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-GPTQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-70B-Instruct-GPTQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-GPTQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-405B-Instruct-GPTQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct-AWQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-AWQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-70B-Instruct-AWQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-AWQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-405B-Instruct-AWQ-INT4', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='AI-ModelScope/Llama-3.1-Nemotron-70B-Instruct-HF', hf_model_id='nvidia/Llama-3.1-Nemotron-70B-Instruct-HF', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='llama3_2', get_function=<function get_model_tokenizer_with_flash_attn at 0x75e08c533ac0>, model_arch='llama', architectures=['LlamaForCausalLM'], additional_saved_files=[], torch_dtype=None, is_multimodal=False, is_reward=False, ignore_patterns=[], requires=['transformers>=4.43'], tags=[])",
+  "model_meta": "ModelMeta(model_type='llama3_1', model_groups=[ModelGroup(models=[Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct', hf_model_id='meta-llama/Meta-Llama-3.1-8B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct', hf_model_id='meta-llama/Meta-Llama-3.1-70B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct', hf_model_id='meta-llama/Meta-Llama-3.1-405B-Instruct', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B', hf_model_id='meta-llama/Meta-Llama-3.1-8B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B', hf_model_id='meta-llama/Meta-Llama-3.1-70B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B', hf_model_id='meta-llama/Meta-Llama-3.1-405B', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-FP8', hf_model_id='meta-llama/Meta-Llama-3.1-70B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-FP8', hf_model_id='meta-llama/Meta-Llama-3.1-405B-Instruct-FP8', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct-BNB-NF4', hf_model_id='hugging-quants/Meta-Llama-3.1-8B-Instruct-BNB-NF4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-bnb-4bit', hf_model_id='unsloth/Meta-Llama-3.1-70B-Instruct-bnb-4bit', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-BNB-NF4', hf_model_id='hugging-quants/Meta-Llama-3.1-405B-Instruct-BNB-NF4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-GPTQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-70B-Instruct-GPTQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-GPTQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-405B-Instruct-GPTQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-8B-Instruct-AWQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-70B-Instruct-AWQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-70B-Instruct-AWQ-INT4', model_path=None, ms_revision=None, hf_revision=None), Model(ms_model_id='LLM-Research/Meta-Llama-3.1-405B-Instruct-AWQ-INT4', hf_model_id='hugging-quants/Meta-Llama-3.1-405B-Instruct-AWQ-INT4', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[]), ModelGroup(models=[Model(ms_model_id='AI-ModelScope/Llama-3.1-Nemotron-70B-Instruct-HF', hf_model_id='nvidia/Llama-3.1-Nemotron-70B-Instruct-HF', model_path=None, ms_revision=None, hf_revision=None)], ignore_patterns=None, requires=None, tags=[])], template='llama3_2', get_function=<function get_model_tokenizer_with_flash_attn at 0x7e836a8f7ac0>, model_arch='llama', architectures=['LlamaForCausalLM'], additional_saved_files=[], torch_dtype=None, is_multimodal=False, is_reward=False, ignore_patterns=[], requires=['transformers>=4.43'], tags=[])",
   "model_dir": "/root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B-Instruct/snapshots/0e9e39f249a16976918f6564b8830bc894c89659",
   "hub": "<class 'swift.hub.hub.HFHub'>",
-  "training_args": "Seq2SeqTrainingArguments(output_dir='/workspace/output/v5-20250108-223508', 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=1, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=0.0003, 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='/workspace/output/v5-20250108-223508/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=5, 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='nccl', tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=100, dataloader_num_workers=0, dataloader_prefetch_factor=None, past_index=-1, run_name='/workspace/output/v5-20250108-223508', 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=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), 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=['comet_ml'], 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='jacpetro/llama-agentic-sft', hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=None, hub_always_push=False, gradient_checkpointing=True, gradient_checkpointing_kwargs={'use_reentrant': False}, 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, galore_config=None)"
+  "training_args": "Seq2SeqTrainingArguments(output_dir='/workspace/output/v6-20250108-223705', 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=1, eval_accumulation_steps=None, eval_delay=0, torch_empty_cache_steps=None, learning_rate=0.0003, 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='/workspace/output/v6-20250108-223705/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=5, 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='nccl', tpu_num_cores=None, tpu_metrics_debug=False, debug=[], dataloader_drop_last=False, eval_steps=100, dataloader_num_workers=0, dataloader_prefetch_factor=None, past_index=-1, run_name='/workspace/output/v6-20250108-223705', 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=False, even_batches=True, use_seedable_sampler=True, non_blocking=False, gradient_accumulation_kwargs=None, use_configured_state=False), 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=['comet_ml'], 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='jacpetro/llama-agentic-sft', hub_strategy=<HubStrategy.EVERY_SAVE: 'every_save'>, hub_token=None, hub_private_repo=None, hub_always_push=False, gradient_checkpointing=True, gradient_checkpointing_kwargs={'use_reentrant': False}, 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, galore_config=None)"
 }

logging.jsonl

@@ -1,6 +1,11 @@
-{"loss": 1.51931369, "token_acc": 0.65133395, "grad_norm": 0.5639717, "learning_rate": 8.11e-06, "memory(GiB)": 11.56, "train_speed(iter/s)": 0.103188, "epoch": 0.00813008, "global_step/max_steps": "1/369", "percentage": "0.27%", "elapsed_time": "3s", "remaining_time": "24m 14s"}
-{"loss": 1.25695169, "token_acc": 0.70556827, "grad_norm": 0.51285839, "learning_rate": 1.622e-05, "memory(GiB)": 11.56, "train_speed(iter/s)": 0.150757, "epoch": 0.01626016, "global_step/max_steps": "2/369", "percentage": "0.54%", "elapsed_time": "7s", "remaining_time": "23m 1s"}
-{"loss": 1.51376784, "token_acc": 0.65096953, "grad_norm": 0.50419378, "learning_rate": 2.432e-05, "memory(GiB)": 11.56, "train_speed(iter/s)": 0.178341, "epoch": 0.02439024, "global_step/max_steps": "3/369", "percentage": "0.81%", "elapsed_time": "11s", "remaining_time": "22m 31s"}
-{"loss": 1.44488692, "token_acc": 0.65977444, "grad_norm": 0.32598454, "learning_rate": 3.243e-05, "memory(GiB)": 11.56, "train_speed(iter/s)": 0.196312, "epoch": 0.03252033, "global_step/max_steps": "4/369", "percentage": "1.08%", "elapsed_time": "14s", "remaining_time": "22m 15s"}
-{"loss": 1.1349889, "token_acc": 0.72309028, "grad_norm": 0.54342312, "learning_rate": 4.054e-05, "memory(GiB)": 11.56, "train_speed(iter/s)": 0.208955, "epoch": 0.04065041, "global_step/max_steps": "5/369", "percentage": "1.36%", "elapsed_time": "18s", "remaining_time": "22m 4s"}
-{"loss": 1.64801562, "token_acc": 0.6032849, "grad_norm": 0.45636648, "learning_rate": 4.865e-05, "memory(GiB)": 11.56, "train_speed(iter/s)": 0.216078, "epoch": 0.04878049, "global_step/max_steps": "6/369", "percentage": "1.63%", "elapsed_time": "22s", "remaining_time": "22m 12s"}
+{"loss": 1.51931369, "token_acc": 0.65133395, "grad_norm": 1.74212694, "learning_rate": 8.11e-06, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.102087, "epoch": 0.00813008, "global_step/max_steps": "1/369", "percentage": "0.27%", "elapsed_time": "4s", "remaining_time": "25m 43s"}
+{"loss": 1.25695169, "token_acc": 0.70556827, "grad_norm": 1.53716826, "learning_rate": 1.622e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.146031, "epoch": 0.01626016, "global_step/max_steps": "2/369", "percentage": "0.54%", "elapsed_time": "8s", "remaining_time": "24m 45s"}
+{"loss": 1.51197827, "token_acc": 0.64889197, "grad_norm": 1.5586139, "learning_rate": 2.432e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.168869, "epoch": 0.02439024, "global_step/max_steps": "3/369", "percentage": "0.81%", "elapsed_time": "12s", "remaining_time": "24m 44s"}
+{"loss": 1.44245231, "token_acc": 0.6585213, "grad_norm": 1.05729473, "learning_rate": 3.243e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.183824, "epoch": 0.03252033, "global_step/max_steps": "4/369", "percentage": "1.08%", "elapsed_time": "16s", "remaining_time": "24m 34s"}
+{"loss": 1.11332512, "token_acc": 0.72395833, "grad_norm": 1.61427522, "learning_rate": 4.054e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.196698, "epoch": 0.04065041, "global_step/max_steps": "5/369", "percentage": "1.36%", "elapsed_time": "19s", "remaining_time": "24m 2s"}
+{"loss": 1.62357724, "token_acc": 0.60518004, "grad_norm": 1.17167616, "learning_rate": 4.865e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.196566, "epoch": 0.04878049, "global_step/max_steps": "6/369", "percentage": "1.63%", "elapsed_time": "24s", "remaining_time": "25m 7s"}
+{"loss": 0.70930511, "token_acc": 0.83585201, "grad_norm": 0.6108681, "learning_rate": 5.676e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.202937, "epoch": 0.05691057, "global_step/max_steps": "7/369", "percentage": "1.90%", "elapsed_time": "28s", "remaining_time": "24m 54s"}
+{"loss": 0.7276203, "token_acc": 0.83107275, "grad_norm": 0.76370066, "learning_rate": 6.486e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.2095, "epoch": 0.06504065, "global_step/max_steps": "8/369", "percentage": "2.17%", "elapsed_time": "32s", "remaining_time": "24m 30s"}
+{"loss": 1.29925573, "token_acc": 0.6878089, "grad_norm": 1.19990838, "learning_rate": 7.297e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.214803, "epoch": 0.07317073, "global_step/max_steps": "9/369", "percentage": "2.44%", "elapsed_time": "36s", "remaining_time": "24m 11s"}
+{"loss": 1.11077285, "token_acc": 0.72534014, "grad_norm": 0.97156239, "learning_rate": 8.108e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.219705, "epoch": 0.08130081, "global_step/max_steps": "10/369", "percentage": "2.71%", "elapsed_time": "39s", "remaining_time": "23m 52s"}
+{"loss": 0.97970724, "token_acc": 0.73512748, "grad_norm": 0.94852263, "learning_rate": 8.919e-05, "memory(GiB)": 12.82, "train_speed(iter/s)": 0.219112, "epoch": 0.08943089, "global_step/max_steps": "11/369", "percentage": "2.98%", "elapsed_time": "44s", "remaining_time": "24m 11s"}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:29d7a1e2470267b2cbfc2e018516909dbdd972bbf3631610d543ea0848243e2f
+oid sha256:8724e39a27cd66df3ce48fb6d9f107fddb4ac0cd690dcde29bfa67001aa97d5c
 size 5880