Upload 16 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+bge-en-icl-paddle/lora_model_state.pdparams filter=lfs diff=lfs merge=lfs -text

bge-en-icl-paddle/added_tokens.json

@@ -0,0 +1 @@
+{"<unk>": 0, "<s>": 1, "</s>": 2, "<instruct>": 32000, "<query>": 32001, "<response>": 32002}

bge-en-icl-paddle/all_results.json

@@ -0,0 +1,7 @@
+{
+    "progress_or_epoch": 0.1482,
+    "train_loss": 0.7501962792873382,
+    "train_runtime": 35785.9312,
+    "train_samples_per_second": 0.3577,
+    "train_steps_per_second": 0.0028
+}

bge-en-icl-paddle/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "hf_models/bge-en-icl",
+  "architectures": [
+    "MistralModel"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 32768,
+  "model_type": "mistral",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "paddlenlp_version": "3.0.0b4.post20250825",
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 10000.0,
+  "sliding_window": 4096,
+  "tie_word_embeddings": false,
+  "vocab_size": 32003
+}

bge-en-icl-paddle/lora_config.json

@@ -0,0 +1,28 @@
+{
+  "base_model_name_or_path": null,
+  "do_qat": false,
+  "dtype": "float16",
+  "enable_lora_list": null,
+  "head_dim": null,
+  "lora_alpha": 64,
+  "lora_dropout": 0.1,
+  "lora_plus_scale": 1.0,
+  "lora_use_mixer": false,
+  "loraga": false,
+  "lorapro": false,
+  "merge_weights": false,
+  "pissa": false,
+  "r": 32,
+  "rslora": false,
+  "scaling": 2.0,
+  "target_modules": [
+    ".*q_proj$",
+    ".*k_proj$",
+    ".*v_proj$"
+  ],
+  "tensor_parallel_degree": 1,
+  "trainable_bias": null,
+  "trainable_modules": null,
+  "use_mora": false,
+  "use_quick_lora": false
+}

bge-en-icl-paddle/lora_model_state.pdparams

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8cdf4f779d379473820c5c0be209a886c4ce44c97d7280eb861ceb9530a6b122
+size 113314622

bge-en-icl-paddle/sentencepiece.bpe.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

bge-en-icl-paddle/special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "pad_token": "<unk>", "additional_special_tokens": [{"content": "<instruct>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, {"content": "<query>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, {"content": "<response>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}]}

bge-en-icl-paddle/tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "add_bos_token": true, "add_eos_token": true, "sp_model_kwargs": {}, "decode_with_prefix_space": false, "add_prefix_space": true, "added_tokens_decoder": {"0": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "1": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "2": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "32000": {"content": "<instruct>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "32001": {"content": "<query>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}, "32002": {"content": "<response>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true}}, "additional_special_tokens": ["<instruct>", "<query>", "<response>"], "clean_up_tokenization_spaces": false, "legacy": true, "model_max_length": 1000000000000000019884624838656, "pad_token": "<unk>", "spaces_between_special_tokens": false, "use_default_system_prompt": false, "name_or_path": "/mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle", "from_slow": true, "_from_auto": true, "special_tokens_map_file": "/mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle/special_tokens_map.json", "tokenizer_class": "LlamaTokenizer"}

bge-en-icl-paddle/train_bge-en-icl-paddle_dureader_dual.train.jsonl.log

@@ -0,0 +1,745 @@
+/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/utils/cpp_extension/extension_utils.py:718: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md
+  warnings.warn(warning_message)
+LAUNCH INFO 2025-10-28 12:24:09,237 -----------  Configuration  ----------------------
+LAUNCH INFO 2025-10-28 12:24:09,237 auto_cluster_config: 0
+LAUNCH INFO 2025-10-28 12:24:09,237 auto_parallel_config: None
+LAUNCH INFO 2025-10-28 12:24:09,237 auto_tuner_json: None
+LAUNCH INFO 2025-10-28 12:24:09,237 devices: 0,1,2,3
+LAUNCH INFO 2025-10-28 12:24:09,237 elastic_level: -1
+LAUNCH INFO 2025-10-28 12:24:09,237 elastic_timeout: 30
+LAUNCH INFO 2025-10-28 12:24:09,237 enable_gpu_log: True
+LAUNCH INFO 2025-10-28 12:24:09,237 gloo_port: 6767
+LAUNCH INFO 2025-10-28 12:24:09,237 host: None
+LAUNCH INFO 2025-10-28 12:24:09,237 ips: None
+LAUNCH INFO 2025-10-28 12:24:09,237 job_id: default
+LAUNCH INFO 2025-10-28 12:24:09,237 legacy: False
+LAUNCH INFO 2025-10-28 12:24:09,237 log_dir: log
+LAUNCH INFO 2025-10-28 12:24:09,239 log_level: INFO
+LAUNCH INFO 2025-10-28 12:24:09,239 log_overwrite: False
+LAUNCH INFO 2025-10-28 12:24:09,239 master: None
+LAUNCH INFO 2025-10-28 12:24:09,239 max_restart: 3
+LAUNCH INFO 2025-10-28 12:24:09,239 nnodes: 1
+LAUNCH INFO 2025-10-28 12:24:09,239 nproc_per_node: None
+LAUNCH INFO 2025-10-28 12:24:09,239 rank: -1
+LAUNCH INFO 2025-10-28 12:24:09,239 run_mode: collective
+LAUNCH INFO 2025-10-28 12:24:09,239 server_num: None
+LAUNCH INFO 2025-10-28 12:24:09,239 servers: 
+LAUNCH INFO 2025-10-28 12:24:09,239 sort_ip: False
+LAUNCH INFO 2025-10-28 12:24:09,239 start_port: 6070
+LAUNCH INFO 2025-10-28 12:24:09,240 trainer_num: None
+LAUNCH INFO 2025-10-28 12:24:09,240 trainers: 
+LAUNCH INFO 2025-10-28 12:24:09,240 training_script: train.py
+LAUNCH INFO 2025-10-28 12:24:09,241 training_script_args: ['--do_train', '--query_instruction_for_retrieval', 'query: ', '--passage_instruction_for_retrieval', '', '--model_name_or_path', '/mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle', '--output_dir', 'tmp_train/bge-en-icl-paddle', '--save_steps', '999999999', '--train_data', './data/dureader_dual.train.jsonl', '--fp16_opt_level', 'O2', '--fp16', '--per_device_train_batch_size', '1', '--gradient_accumulation_steps', '32', '--recompute', '--train_group_size', '4', '--learning_rate', '1e-4', '--query_max_len', '128', '--passage_max_len', '4096', '--num_train_epochs', '1', '--logging_steps', '1', '--overwrite_output_dir', '--negatives_cross_device', '--warmup_steps', '10', '--max_steps', '100', '--do_train', '--fine_tune_type', 'lora', '--sentence_pooling_method', 'last', '--sharding', 'stage3 offload', '--use_flash_attention', '--temperature', '0.01']
+LAUNCH INFO 2025-10-28 12:24:09,241 with_gloo: 1
+LAUNCH INFO 2025-10-28 12:24:09,241 --------------------------------------------------
+LAUNCH INFO 2025-10-28 12:24:09,243 Job: default, mode collective, replicas 1[1:1], elastic False
+LAUNCH INFO 2025-10-28 12:24:09,287 Run Pod: zbwwum, replicas 4, status ready
+LAUNCH INFO 2025-10-28 12:24:09,601 Watching Pod: zbwwum, replicas 4, status running
+/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/utils/cpp_extension/extension_utils.py:718: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md
+  warnings.warn(warning_message)
+/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/_distutils_hack/__init__.py:30: UserWarning: Setuptools is replacing distutils. Support for replacing an already imported distutils is deprecated. In the future, this condition will fail. Register concerns at https://github.com/pypa/setuptools/issues/new?template=distutils-deprecation.yml
+  warnings.warn(
+/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddlenlp/trainer/training_args.py:1205: UserWarning: `offload` is not supported NOW!
+  warnings.warn("`offload` is not supported NOW!")
+[2025-10-28 12:24:14,635] [    INFO] distributed_strategy.py:335 - distributed strategy initialized
+======================= Modified FLAGS detected =======================
+FLAGS(name='FLAGS_nvidia_package_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia', default_value='')
+FLAGS(name='FLAGS_cublas_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/cublas/lib', default_value='')
+FLAGS(name='FLAGS_cusolver_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/cusolver/lib', default_value='')
+FLAGS(name='FLAGS_cusparse_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/cusparse/lib', default_value='')
+FLAGS(name='FLAGS_cupti_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/cuda_cupti/lib', default_value='')
+FLAGS(name='FLAGS_curand_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/curand/lib', default_value='')
+FLAGS(name='FLAGS_cudnn_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/cudnn/lib', default_value='')
+FLAGS(name='FLAGS_nccl_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/nccl/lib', default_value='')
+FLAGS(name='FLAGS_selected_gpus', current_value='0', default_value='')
+FLAGS(name='FLAGS_cuda_cccl_dir', current_value='/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/../nvidia/cuda_cccl/include/', default_value='')
+FLAGS(name='FLAGS_enable_pir_in_executor', current_value=True, default_value=False)
+=======================================================================
+I1028 12:24:14.636552   129 tcp_store.cc:336] create libuv server at port: 58491
+I1028 12:24:14.637319   129 tcp_utils.cc:132] Successfully connected to 172.17.0.5:58491
+I1028 12:24:15.241484   129 process_group_nccl.cc:154] ProcessGroupNCCL pg_timeout_ 1800000
+I1028 12:24:15.241554   129 process_group_nccl.cc:155] ProcessGroupNCCL nccl_comm_init_option_ 0
+[2025-10-28 12:24:15,242] [    INFO] topology.py:526 - Total 4 pipe comm group(s) create successfully!
+W1028 12:24:15.243945   129 gpu_resources.cc:114] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 12.8, Runtime API Version: 12.8
+/mnt/141nfs/zhongtianyun2023/miniconda3/envs/paddle_0905/lib/python3.10/site-packages/paddle/distributed/communication/group.py:145: UserWarning: Current global rank 0 is not in group _default_pg10
+  warnings.warn(
+[2025-10-28 12:24:15,246] [    INFO] topology.py:526 - Total 4 data comm group(s) create successfully!
+[2025-10-28 12:24:15,247] [    INFO] topology.py:526 - Total 4 model comm group(s) create successfully!
+I1028 12:24:15.247094   129 process_group_nccl.cc:154] ProcessGroupNCCL pg_timeout_ 1800000
+I1028 12:24:15.247102   129 process_group_nccl.cc:155] ProcessGroupNCCL nccl_comm_init_option_ 0
+[2025-10-28 12:24:15,247] [    INFO] topology.py:526 - Total 1 sharding comm group(s) create successfully!
+I1028 12:24:15.247182   129 process_group_nccl.cc:154] ProcessGroupNCCL pg_timeout_ 1800000
+I1028 12:24:15.247188   129 process_group_nccl.cc:155] ProcessGroupNCCL nccl_comm_init_option_ 0
+[2025-10-28 12:24:15,247] [    INFO] topology.py:440 - HybridParallelInfo: rank_id: 0, mp_degree: 1, sharding_degree: 4, pp_degree: 1, dp_degree: 1, sep_degree: 1, mp_group: [0],  sharding_group: [0, 1, 2, 3], pp_group: [0], dp_group: [0], sep:group: None, check/clip group: [0, 1, 2, 3]
+[2025-10-28 12:24:15,247] [    INFO] -     +==============================================================================+
+    |                                                                              |
+    |                         DistributedStrategy Overview                         |
+    |                                                                              |
+    +==============================================================================+
+    |                        a_sync=True <-> a_sync_configs                        |
+    +------------------------------------------------------------------------------+
+    |                               k_steps                    -1                  |
+    |                     max_merge_var_num                    1                   |
+    |                       send_queue_size                    16                  |
+    |               independent_recv_thread                  False                 |
+    |         min_send_grad_num_before_recv                    1                   |
+    |                      thread_pool_size                    1                   |
+    |                       send_wait_times                    1                   |
+    |               runtime_split_send_recv                  False                 |
+    |                        launch_barrier                   True                 |
+    |             heter_worker_device_guard                   cpu                  |
+    |                        lr_decay_steps                    10                  |
+    |                            use_ps_gpu                    0                   |
+    |                         use_gpu_graph                    0                   |
+    +==============================================================================+
+    |                    Environment Flags, Communication Flags                    |
+    +------------------------------------------------------------------------------+
+    |                                  mode                    1                   |
+    |                               elastic                  False                 |
+    |                                  auto                  False                 |
+    |                   sync_nccl_allreduce                   True                 |
+    |                         nccl_comm_num                    1                   |
+    |            use_hierarchical_allreduce                  False                 |
+    |   hierarchical_allreduce_inter_nranks                    1                   |
+    |                       sync_batch_norm                  False                 |
+    |                   fuse_all_reduce_ops                   True                 |
+    |                  fuse_grad_size_in_MB                    32                  |
+    |              fuse_grad_size_in_TFLOPS                   50.0                 |
+    |               cudnn_exhaustive_search                  False                 |
+    |             conv_workspace_size_limit                   512                  |
+    |    cudnn_batchnorm_spatial_persistent                  False                 |
+    |                        fp16_allreduce                  False                 |
+    |               last_comm_group_size_MB                   1.0                  |
+    |                find_unused_parameters                  False                 |
+    |            without_graph_optimization                   True                 |
+    |                 fuse_grad_size_in_num                    8                   |
+    |                 calc_comm_same_stream                  False                 |
+    |                                   asp                  False                 |
+    |                       fuse_grad_merge                  False                 |
+    |                             semi_auto                  False                 |
+    |                            adam_d2sum                  False                 |
+    |                           auto_search                  False                 |
+    |                        heter_ccl_mode                  False                 |
+    |                         is_fl_ps_mode                  False                 |
+    |                      with_coordinator                  False                 |
+    |                            split_data                   True                 |
+    |                  downpour_table_param                    []                  |
+    |                       fs_client_param                                        |
+    +==============================================================================+
+    |                                Build Strategy                                |
+    +------------------------------------------------------------------------------+
+    |              fuse_elewise_add_act_ops                  False                 |
+    |                       fuse_bn_act_ops                  False                 |
+    |              fuse_relu_depthwise_conv                  False                 |
+    |                    fuse_broadcast_ops                  False                 |
+    |                fuse_all_optimizer_ops                  False                 |
+    |                        enable_inplace                  False                 |
+    |     enable_backward_optimizer_op_deps                   True                 |
+    |                 cache_runtime_context                  False                 |
+    |                   fuse_bn_add_act_ops                   True                 |
+    |                    enable_auto_fusion                  False                 |
+    |                          enable_addto                  False                 |
+    |              allow_cuda_graph_capture                  False                 |
+    |                       reduce_strategy                    0                   |
+    |                    fuse_gemm_epilogue                  False                 |
+    |                   debug_graphviz_path                                        |
+    |                       fused_attention                  False                 |
+    |                     fused_feedforward                  False                 |
+    |            fuse_dot_product_attention                  False                 |
+    |                          fuse_resunit                  False                 |
+    +==============================================================================+
+
+[2025-10-28 12:24:15,247] [    INFO] - The default value for the training argument `--report_to` will change in v5 (from all installed integrations to none). In v5, you will need to use `--report_to all` to get the same behavior as now. You should start updating your code and make this info disappear :-).
+[2025-10-28 12:24:15,248] [ WARNING] - Process rank: 0, device: gpu, distributed training: True, 16-bits training: True
+[2025-10-28 12:24:15,248] [    INFO] - Training/evaluation parameters RetrieverTrainingArguments(
+adam_beta1=0.9,
+adam_beta2=0.999,
+adam_epsilon=1e-08,
+amp_custom_black_list=None,
+amp_custom_white_list=None,
+amp_master_grad=False,
+auto_parallel_resume_form_hybrid_parallel=False,
+bf16=False,
+bf16_full_eval=False,
+ckpt_quant_stage=O0,
+context_parallel_degree=1,
+count_trained_tokens=False,
+data_parallel_config=,
+dataloader_drop_last=False,
+dataloader_num_workers=0,
+dataloader_shuffle=True,
+ddp_find_unused_parameters=None,
+device=gpu,
+disable_tqdm=False,
+distributed_dataloader=False,
+do_eval=False,
+do_export=False,
+do_predict=False,
+do_train=True,
+enable_auto_parallel=False,
+enable_zero_cost_checkpoint=False,
+eval_accumulation_steps=None,
+eval_steps=None,
+evaluation_strategy=IntervalStrategy.NO,
+expert_max_capacity=4294967296,
+expert_min_capacity=1,
+expert_parallel_degree=1,
+expert_tensor_parallel_degree=1,
+fine_tune_type=lora,
+fix_position_embedding=False,
+flash_device_save_steps=0,
+flatten_param_grads=False,
+force_reshard_pp=False,
+fp16=True,
+fp16_full_eval=False,
+fp16_opt_level=O2,
+fuse_sequence_parallel_allreduce=False,
+gradient_accumulation_steps=32,
+greater_is_better=None,
+hybrid_parallel_topo_order=pp_first,
+ignore_data_skip=False,
+ignore_load_lr_and_optim=False,
+ignore_save_lr_and_optim=False,
+label_names=None,
+lazy_data_processing=True,
+learning_rate=0.0001,
+load_best_model_at_end=False,
+load_sharded_model=False,
+local_rank=0,
+log_on_each_node=True,
+logging_dir=tmp_train/bge-en-icl-paddle/runs/Oct28_12-24-14_b63e119a1648,
+logging_first_step=False,
+logging_steps=1,
+logging_strategy=IntervalStrategy.STEPS,
+lr_end=1e-07,
+lr_scheduler_type=SchedulerType.LINEAR,
+margin=0.2,
+matryoshka_dims=[64, 128, 256, 512, 768],
+matryoshka_loss_weights=[1, 1, 1, 1, 1],
+max_evaluate_steps=-1,
+max_grad_norm=1.0,
+max_steps=100,
+metric_for_best_model=None,
+metrics_output_path=None,
+min_lr=0.0,
+minimum_eval_times=None,
+nccl_comm_group_config=None,
+negatives_cross_device=True,
+no_cuda=False,
+num_cycles=0.5,
+num_train_epochs=1.0,
+offload_optim=False,
+optim=OptimizerNames.ADAMW,
+ordered_save_group_size=0,
+output_dir=tmp_train/bge-en-icl-paddle,
+output_signal_dir=tmp_train/bge-en-icl-paddle,
+overwrite_output_dir=True,
+pad_token_id=0,
+past_index=-1,
+pdc_download_ckpt=False,
+pdc_download_timeout=300,
+per_device_eval_batch_size=8,
+per_device_train_batch_size=1,
+pipeline_parallel_config=,
+pipeline_parallel_degree=1,
+power=1.0,
+prediction_loss_only=False,
+recompute=True,
+refined_recompute={},
+release_grads=False,
+remove_unused_columns=True,
+report_to=['visualdl'],
+resume_from_checkpoint=None,
+run_name=tmp_train/bge-en-icl-paddle,
+save_on_each_node=False,
+save_rng_states=True,
+save_sharded_model=False,
+save_sharding_stage1_model_include_freeze_params=False,
+save_steps=999999999,
+save_strategy=IntervalStrategy.STEPS,
+save_tokenizer=True,
+save_total_limit=None,
+scale_loss=32768,
+seed=42,
+sentence_pooling_method=last,
+sep_parallel_degree=1,
+sequence_parallel=False,
+sequence_parallel_config=,
+sharding=[<ShardingOption.FULL_SHARD: 'stage3'>, <ShardingOption.OFFLOAD: 'offload'>],
+sharding_comm_buffer_size_MB=-1,
+sharding_degree=-1,
+sharding_parallel_config=,
+sharding_parallel_degree=4,
+sharding_parallel_mesh_dimension=dp,
+skip_data_intervals=None,
+skip_memory_metrics=True,
+skip_profile_timer=True,
+split_inputs_sequence_dim=True,
+split_norm_comm=False,
+temperature=0.01,
+tensor_parallel_config=,
+tensor_parallel_degree=1,
+tensorwise_offload_optimizer=False,
+to_static=False,
+unified_checkpoint=False,
+unified_checkpoint_config=,
+use_async_save=False,
+use_expert_parallel=False,
+use_inbatch_neg=False,
+use_lowprecision_moment=False,
+use_matryoshka=False,
+wandb_api_key=None,
+wandb_http_proxy=None,
+warmup_ratio=0.0,
+warmup_steps=10,
+weight_decay=0.0,
+zcc_ema_interval=1,
+zcc_pipeline_hooks_capacity_usage=0.6,
+zcc_save_ema_coef=None,
+zcc_workers_num=3,
+)
+[2025-10-28 12:24:15,248] [    INFO] - Model parameters ModelArguments(model_name_or_path='/mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle', tokenizer_name=None, normalized=True, use_flash_attention=True)
+[2025-10-28 12:24:15,248] [    INFO] - Data parameters DataArguments(train_data='./data/dureader_dual.train.jsonl', train_group_size=4, query_max_len=128, passage_max_len=4096, max_example_num_per_dataset=100000000, query_instruction_for_retrieval='query: ', passage_instruction_for_retrieval='')
+[2025-10-28 12:24:15,250] [    INFO] - The global seed is set to 42, local seed is set to 46 and random seed is set to 42.
+[2025-10-28 12:24:15,263] [    INFO] - We are using <class 'paddlenlp.transformers.mistral.modeling.MistralModel'> to load '/mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle'.
+[2025-10-28 12:24:15,263] [    INFO] - Loading configuration file /mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle/config.json
+[2025-10-28 12:24:15,266] [    INFO] - Loading weights file /mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle/model_state.pdparams.index.json
+file: /mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle/model_state.pdparams.index.json is paddle weight.
+
+Loading checkpoint shards:   0%|          | 0/8 [00:00<?, ?it/s]
+Loading checkpoint shards:  12%|█▎        | 1/8 [00:11<01:17, 11.09s/it]
+Loading checkpoint shards:  25%|██▌       | 2/8 [00:22<01:06, 11.08s/it]
+Loading checkpoint shards:  38%|███▊      | 3/8 [00:33<00:55, 11.07s/it]
+Loading checkpoint shards:  50%|█████     | 4/8 [00:43<00:43, 10.90s/it]
+Loading checkpoint shards:  62%|██████▎   | 5/8 [00:54<00:32, 10.88s/it]
+Loading checkpoint shards:  75%|███████▌  | 6/8 [01:05<00:21, 10.78s/it]
+Loading checkpoint shards:  88%|████████▊ | 7/8 [01:15<00:10, 10.74s/it]
+Loading checkpoint shards: 100%|██████████| 8/8 [01:19<00:00,  8.32s/it]
+Loading checkpoint shards: 100%|██████████| 8/8 [01:19<00:00,  9.89s/it]
+[2025-10-28 12:25:34,791] [    INFO] - All model checkpoint weights were used when initializing MistralModel.
+
+[2025-10-28 12:25:34,791] [    INFO] - All the weights of MistralModel were initialized from the model checkpoint at /mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle.
+If your task is similar to the task the model of the checkpoint was trained on, you can already use MistralModel for predictions without further training.
+[2025-10-28 12:25:34,794] [    INFO] - Loading configuration file /mnt/141nfs/lizhuoqun/hf_models/bge-en-icl-paddle/config.json
+[2025-10-28 12:25:34,952] [ WARNING] - Reset tensor_parallel_degree of lora_config to 1.
+[2025-10-28 12:25:34,952] [    INFO] - Mark only lora and trainable_module as trainable.
+[2025-10-28 12:25:34,974] [   DEBUG] - Frozen parameters: 2.13e+10 || Trainable parameters:5.66e+07 || Total parameters:2.14e+10|| Trainable:0.26%
+[2025-10-28 12:25:35,009] [    INFO] - The global seed is set to 42, local seed is set to 46 and random seed is set to 42.
+[2025-10-28 12:25:35,067] [    INFO] - max_steps is given, it will override any value given in num_train_epochs
+[2025-10-28 12:25:35,067] [    INFO] - Using half precision
+[2025-10-28 12:25:35,087] [   DEBUG] - ============================================================
+[2025-10-28 12:25:35,087] [   DEBUG] -     Training Configuration Arguments    
+[2025-10-28 12:25:35,087] [   DEBUG] - paddle commit id              : dc312219f1498234f178a3b6a93e9f1f4ddebcb8
+[2025-10-28 12:25:35,087] [   DEBUG] - paddlenlp commit id           : 0bb61071d051adf6207199d6493b08079d58f8b2.dirty
+[2025-10-28 12:25:35,089] [   DEBUG] - _no_sync_in_gradient_accumulation: True
+[2025-10-28 12:25:35,090] [   DEBUG] - adam_beta1                    : 0.9
+[2025-10-28 12:25:35,108] [   DEBUG] - adam_beta2                    : 0.999
+[2025-10-28 12:25:35,108] [   DEBUG] - adam_epsilon                  : 1e-08
+[2025-10-28 12:25:35,108] [   DEBUG] - amp_custom_black_list         : None
+[2025-10-28 12:25:35,110] [   DEBUG] - amp_custom_white_list         : None
+[2025-10-28 12:25:35,110] [   DEBUG] - amp_master_grad               : False
+[2025-10-28 12:25:35,110] [   DEBUG] - auto_parallel_resume_form_hybrid_parallel: False
+[2025-10-28 12:25:35,111] [   DEBUG] - bf16                          : False
+[2025-10-28 12:25:35,111] [   DEBUG] - bf16_full_eval                : False
+[2025-10-28 12:25:35,111] [   DEBUG] - ckpt_quant_stage              : O0
+[2025-10-28 12:25:35,112] [   DEBUG] - context_parallel_degree       : 1
+[2025-10-28 12:25:35,112] [   DEBUG] - count_trained_tokens          : False
+[2025-10-28 12:25:35,112] [   DEBUG] - current_device                : gpu:0
+[2025-10-28 12:25:35,112] [   DEBUG] - data_parallel_config          : 
+[2025-10-28 12:25:35,112] [   DEBUG] - data_parallel_degree          : 1
+[2025-10-28 12:25:35,112] [   DEBUG] - data_parallel_rank            : 0
+[2025-10-28 12:25:35,112] [   DEBUG] - dataloader_drop_last          : False
+[2025-10-28 12:25:35,112] [   DEBUG] - dataloader_num_workers        : 0
+[2025-10-28 12:25:35,112] [   DEBUG] - dataloader_shuffle            : True
+[2025-10-28 12:25:35,112] [   DEBUG] - dataset_rank                  : 0
+[2025-10-28 12:25:35,112] [   DEBUG] - dataset_world_size            : 4
+[2025-10-28 12:25:35,112] [   DEBUG] - ddp_find_unused_parameters    : None
+[2025-10-28 12:25:35,112] [   DEBUG] - device                        : gpu
+[2025-10-28 12:25:35,112] [   DEBUG] - disable_tqdm                  : False
+[2025-10-28 12:25:35,112] [   DEBUG] - distributed_dataloader        : False
+[2025-10-28 12:25:35,112] [   DEBUG] - do_eval                       : False
+[2025-10-28 12:25:35,112] [   DEBUG] - do_export                     : False
+[2025-10-28 12:25:35,112] [   DEBUG] - do_predict                    : False
+[2025-10-28 12:25:35,113] [   DEBUG] - do_train                      : True
+[2025-10-28 12:25:35,113] [   DEBUG] - enable_auto_parallel          : False
+[2025-10-28 12:25:35,113] [   DEBUG] - enable_zero_cost_checkpoint   : False
+[2025-10-28 12:25:35,113] [   DEBUG] - eval_accumulation_steps       : None
+[2025-10-28 12:25:35,113] [   DEBUG] - eval_batch_size               : 8
+[2025-10-28 12:25:35,113] [   DEBUG] - eval_steps                    : None
+[2025-10-28 12:25:35,113] [   DEBUG] - evaluation_strategy           : IntervalStrategy.NO
+[2025-10-28 12:25:35,113] [   DEBUG] - expert_max_capacity           : 4294967296
+[2025-10-28 12:25:35,113] [   DEBUG] - expert_min_capacity           : 1
+[2025-10-28 12:25:35,113] [   DEBUG] - expert_parallel_degree        : 1
+[2025-10-28 12:25:35,113] [   DEBUG] - expert_tensor_parallel_degree : 1
+[2025-10-28 12:25:35,113] [   DEBUG] - fine_tune_type                : lora
+[2025-10-28 12:25:35,113] [   DEBUG] - fix_position_embedding        : False
+[2025-10-28 12:25:35,113] [   DEBUG] - flash_device_save_steps       : 0
+[2025-10-28 12:25:35,113] [   DEBUG] - flatten_param_grads           : False
+[2025-10-28 12:25:35,113] [   DEBUG] - force_reshard_pp              : False
+[2025-10-28 12:25:35,113] [   DEBUG] - fp16                          : True
+[2025-10-28 12:25:35,113] [   DEBUG] - fp16_full_eval                : False
+[2025-10-28 12:25:35,113] [   DEBUG] - fp16_opt_level                : O2
+[2025-10-28 12:25:35,113] [   DEBUG] - fuse_sequence_parallel_allreduce: False
+[2025-10-28 12:25:35,114] [   DEBUG] - gradient_accumulation_steps   : 32
+[2025-10-28 12:25:35,114] [   DEBUG] - greater_is_better             : None
+[2025-10-28 12:25:35,114] [   DEBUG] - hybrid_parallel_topo_order    : pp_first
+[2025-10-28 12:25:35,114] [   DEBUG] - ignore_data_skip              : False
+[2025-10-28 12:25:35,114] [   DEBUG] - ignore_load_lr_and_optim      : False
+[2025-10-28 12:25:35,114] [   DEBUG] - ignore_save_lr_and_optim      : False
+[2025-10-28 12:25:35,114] [   DEBUG] - label_names                   : None
+[2025-10-28 12:25:35,114] [   DEBUG] - lazy_data_processing          : True
+[2025-10-28 12:25:35,114] [   DEBUG] - learning_rate                 : 0.0001
+[2025-10-28 12:25:35,114] [   DEBUG] - load_best_model_at_end        : False
+[2025-10-28 12:25:35,114] [   DEBUG] - load_sharded_model            : False
+[2025-10-28 12:25:35,114] [   DEBUG] - local_process_index           : 0
+[2025-10-28 12:25:35,114] [   DEBUG] - local_rank                    : 0
+[2025-10-28 12:25:35,114] [   DEBUG] - log_level                     : -1
+[2025-10-28 12:25:35,114] [   DEBUG] - log_level_replica             : -1
+[2025-10-28 12:25:35,114] [   DEBUG] - log_on_each_node              : True
+[2025-10-28 12:25:35,114] [   DEBUG] - logging_dir                   : tmp_train/bge-en-icl-paddle/runs/Oct28_12-24-14_b63e119a1648
+[2025-10-28 12:25:35,114] [   DEBUG] - logging_first_step            : False
+[2025-10-28 12:25:35,114] [   DEBUG] - logging_steps                 : 1
+[2025-10-28 12:25:35,114] [   DEBUG] - logging_strategy              : IntervalStrategy.STEPS
+[2025-10-28 12:25:35,115] [   DEBUG] - logical_process_index         : 0
+[2025-10-28 12:25:35,115] [   DEBUG] - lr_end                        : 1e-07
+[2025-10-28 12:25:35,115] [   DEBUG] - lr_scheduler_type             : SchedulerType.LINEAR
+[2025-10-28 12:25:35,115] [   DEBUG] - margin                        : 0.2
+[2025-10-28 12:25:35,115] [   DEBUG] - matryoshka_dims               : [64, 128, 256, 512, 768]
+[2025-10-28 12:25:35,115] [   DEBUG] - matryoshka_loss_weights       : [1, 1, 1, 1, 1]
+[2025-10-28 12:25:35,115] [   DEBUG] - max_evaluate_steps            : -1
+[2025-10-28 12:25:35,115] [   DEBUG] - max_grad_norm                 : 1.0
+[2025-10-28 12:25:35,115] [   DEBUG] - max_steps                     : 100
+[2025-10-28 12:25:35,115] [   DEBUG] - metric_for_best_model         : None
+[2025-10-28 12:25:35,115] [   DEBUG] - metrics_output_path           : None
+[2025-10-28 12:25:35,115] [   DEBUG] - min_lr                        : 0.0
+[2025-10-28 12:25:35,115] [   DEBUG] - minimum_eval_times            : None
+[2025-10-28 12:25:35,115] [   DEBUG] - nccl_comm_group_config        : None
+[2025-10-28 12:25:35,115] [   DEBUG] - negatives_cross_device        : True
+[2025-10-28 12:25:35,115] [   DEBUG] - no_cuda                       : False
+[2025-10-28 12:25:35,115] [   DEBUG] - num_cycles                    : 0.5
+[2025-10-28 12:25:35,115] [   DEBUG] - num_train_epochs              : 1.0
+[2025-10-28 12:25:35,115] [   DEBUG] - offload_optim                 : False
+[2025-10-28 12:25:35,116] [   DEBUG] - optim                         : OptimizerNames.ADAMW
+[2025-10-28 12:25:35,116] [   DEBUG] - optimizer_name_suffix         : shard00
+[2025-10-28 12:25:35,116] [   DEBUG] - ordered_save_group_size       : 0
+[2025-10-28 12:25:35,116] [   DEBUG] - output_dir                    : tmp_train/bge-en-icl-paddle
+[2025-10-28 12:25:35,116] [   DEBUG] - output_signal_dir             : tmp_train/bge-en-icl-paddle
+[2025-10-28 12:25:35,116] [   DEBUG] - overwrite_output_dir          : True
+[2025-10-28 12:25:35,116] [   DEBUG] - pad_token_id                  : 0
+[2025-10-28 12:25:35,116] [   DEBUG] - past_index                    : -1
+[2025-10-28 12:25:35,116] [   DEBUG] - pdc_download_ckpt             : False
+[2025-10-28 12:25:35,116] [   DEBUG] - pdc_download_timeout          : 300
+[2025-10-28 12:25:35,116] [   DEBUG] - per_device_eval_batch_size    : 8
+[2025-10-28 12:25:35,116] [   DEBUG] - per_device_train_batch_size   : 1
+[2025-10-28 12:25:35,116] [   DEBUG] - pipeline_parallel_config      : 
+[2025-10-28 12:25:35,116] [   DEBUG] - pipeline_parallel_degree      : 1
+[2025-10-28 12:25:35,116] [   DEBUG] - pipeline_parallel_rank        : 0
+[2025-10-28 12:25:35,116] [   DEBUG] - power                         : 1.0
+[2025-10-28 12:25:35,116] [   DEBUG] - prediction_loss_only          : False
+[2025-10-28 12:25:35,116] [   DEBUG] - process_index                 : 0
+[2025-10-28 12:25:35,117] [   DEBUG] - recompute                     : True
+[2025-10-28 12:25:35,117] [   DEBUG] - refined_recompute             : {}
+[2025-10-28 12:25:35,117] [   DEBUG] - release_grads                 : False
+[2025-10-28 12:25:35,117] [   DEBUG] - remove_unused_columns         : True
+[2025-10-28 12:25:35,117] [   DEBUG] - report_to                     : ['visualdl']
+[2025-10-28 12:25:35,117] [   DEBUG] - resume_from_checkpoint        : None
+[2025-10-28 12:25:35,117] [   DEBUG] - run_name                      : tmp_train/bge-en-icl-paddle
+[2025-10-28 12:25:35,117] [   DEBUG] - save_on_each_node             : False
+[2025-10-28 12:25:35,117] [   DEBUG] - save_rng_states               : True
+[2025-10-28 12:25:35,117] [   DEBUG] - save_sharded_model            : False
+[2025-10-28 12:25:35,117] [   DEBUG] - save_sharding_stage1_model_include_freeze_params: False
+[2025-10-28 12:25:35,117] [   DEBUG] - save_steps                    : 999999999
+[2025-10-28 12:25:35,117] [   DEBUG] - save_strategy                 : IntervalStrategy.STEPS
+[2025-10-28 12:25:35,117] [   DEBUG] - save_tokenizer                : True
+[2025-10-28 12:25:35,117] [   DEBUG] - save_total_limit              : None
+[2025-10-28 12:25:35,117] [   DEBUG] - scale_loss                    : 32768
+[2025-10-28 12:25:35,117] [   DEBUG] - seed                          : 42
+[2025-10-28 12:25:35,117] [   DEBUG] - sentence_pooling_method       : last
+[2025-10-28 12:25:35,117] [   DEBUG] - sep_parallel_degree           : 1
+[2025-10-28 12:25:35,117] [   DEBUG] - sequence_parallel             : False
+[2025-10-28 12:25:35,118] [   DEBUG] - sequence_parallel_config      : 
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding                      : [<ShardingOption.FULL_SHARD: 'stage3'>, <ShardingOption.OFFLOAD: 'offload'>]
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding_comm_buffer_size_MB  : -1
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding_degree               : -1
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding_parallel_config      : 
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding_parallel_degree      : 4
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding_parallel_mesh_dimension: dp
+[2025-10-28 12:25:35,118] [   DEBUG] - sharding_parallel_rank        : 0
+[2025-10-28 12:25:35,118] [   DEBUG] - should_load_dataset           : True
+[2025-10-28 12:25:35,118] [   DEBUG] - should_load_sharding_stage1_model: False
+[2025-10-28 12:25:35,118] [   DEBUG] - should_log                    : True
+[2025-10-28 12:25:35,118] [   DEBUG] - should_save                   : True
+[2025-10-28 12:25:35,118] [   DEBUG] - should_save_model_state       : True
+[2025-10-28 12:25:35,118] [   DEBUG] - should_save_model_with_tensor_fusion: False
+[2025-10-28 12:25:35,118] [   DEBUG] - should_save_sharding_stage1_model: False
+[2025-10-28 12:25:35,118] [   DEBUG] - skip_data_intervals           : None
+[2025-10-28 12:25:35,118] [   DEBUG] - skip_memory_metrics           : True
+[2025-10-28 12:25:35,118] [   DEBUG] - skip_profile_timer            : True
+[2025-10-28 12:25:35,118] [   DEBUG] - split_inputs_sequence_dim     : True
+[2025-10-28 12:25:35,119] [   DEBUG] - split_norm_comm               : False
+[2025-10-28 12:25:35,119] [   DEBUG] - temperature                   : 0.01
+[2025-10-28 12:25:35,119] [   DEBUG] - tensor_parallel_config        : 
+[2025-10-28 12:25:35,119] [   DEBUG] - tensor_parallel_degree        : 1
+[2025-10-28 12:25:35,119] [   DEBUG] - tensor_parallel_rank          : 0
+[2025-10-28 12:25:35,119] [   DEBUG] - tensorwise_offload_optimizer  : False
+[2025-10-28 12:25:35,119] [   DEBUG] - to_static                     : False
+[2025-10-28 12:25:35,119] [   DEBUG] - train_batch_size              : 1
+[2025-10-28 12:25:35,119] [   DEBUG] - unified_checkpoint            : False
+[2025-10-28 12:25:35,119] [   DEBUG] - unified_checkpoint_config     : 
+[2025-10-28 12:25:35,119] [   DEBUG] - use_async_save                : False
+[2025-10-28 12:25:35,119] [   DEBUG] - use_expert_parallel           : False
+[2025-10-28 12:25:35,119] [   DEBUG] - use_hybrid_parallel           : True
+[2025-10-28 12:25:35,119] [   DEBUG] - use_inbatch_neg               : False
+[2025-10-28 12:25:35,119] [   DEBUG] - use_lowprecision_moment       : False
+[2025-10-28 12:25:35,119] [   DEBUG] - use_matryoshka                : False
+[2025-10-28 12:25:35,119] [   DEBUG] - wandb_api_key                 : None
+[2025-10-28 12:25:35,119] [   DEBUG] - wandb_http_proxy              : None
+[2025-10-28 12:25:35,119] [   DEBUG] - warmup_ratio                  : 0.0
+[2025-10-28 12:25:35,119] [   DEBUG] - warmup_steps                  : 10
+[2025-10-28 12:25:35,120] [   DEBUG] - weight_decay                  : 0.0
+[2025-10-28 12:25:35,120] [   DEBUG] - weight_name_suffix            : 
+[2025-10-28 12:25:35,120] [   DEBUG] - world_size                    : 4
+[2025-10-28 12:25:35,120] [   DEBUG] - zcc_ema_interval              : 1
+[2025-10-28 12:25:35,120] [   DEBUG] - zcc_pipeline_hooks_capacity_usage: 0.6
+[2025-10-28 12:25:35,120] [   DEBUG] - zcc_save_ema_coef             : None
+[2025-10-28 12:25:35,120] [   DEBUG] - zcc_workers_num               : 3
+[2025-10-28 12:25:35,120] [   DEBUG] - 
+[2025-10-28 12:25:35,120] [    INFO] - Starting training from resume_from_checkpoint : None
+W1028 12:25:35.277778   129 nccl_comm_context.cc:70] ncclCommInitRankConfigMemOpt is not supported.
+[2025-10-28 12:25:36,009] [ WARNING] group_sharded.py:148 - the input of scaler is None, please ensure the logic of your scaler outside is same as GroupShardedScaler.
+[2025-10-28 12:25:36,011] [ WARNING] group_sharded_stage3.py:195 - While using ClipGradByGlobalNorm in GroupShardedStage3, the grad clip of original optimizer will be changed.
+W1028 12:25:36.014405   129 nccl_comm_context.cc:70] ncclCommInitRankConfigMemOpt is not supported.
+[2025-10-28 12:26:10,617] [    INFO] - [timelog] checkpoint loading time: 0.00s (2025-10-28 12:26:10) 
+[2025-10-28 12:26:10,618] [    INFO] - ***** Running training *****
+[2025-10-28 12:26:10,618] [    INFO] -   Num examples = 86,395
+[2025-10-28 12:26:10,618] [    INFO] -   Num Epochs = 1
+[2025-10-28 12:26:10,620] [    INFO] -   Instantaneous batch size per device = 1
+[2025-10-28 12:26:10,620] [    INFO] -   Total train batch size (w. parallel, distributed & accumulation) = 128
+[2025-10-28 12:26:10,620] [    INFO] -   Gradient Accumulation steps = 32
+[2025-10-28 12:26:10,622] [    INFO] -   Total optimization steps = 100
+[2025-10-28 12:26:10,622] [    INFO] -   Total num train samples = 12,800
+[2025-10-28 12:26:10,625] [   DEBUG] -   Number of trainable parameters = 18,874,368 (per device)
+
+TrainProcess:   0%|          | 0/100 [00:00<?, ?it/s]W1028 12:26:23.394423   129 multiply_fwd_func.cc:82] got different data type, run type promotion automatically, this may cause data type been changed.
+W1028 12:26:35.097034   129 dygraph_functions.cc:94044] got different data type, run type promotion automatically, this may cause data type been changed.
+Found inf or nan, current scale is: 32768.0, decrease to: 32768.0*0.5
+[2025-10-28 12:31:50,968] [ WARNING] - optimizer not run, scale_before: 32768.0, scale_after: 16384.0
+
+TrainProcess:   1%|          | 1/100 [05:40<9:21:45, 340.46s/it][2025-10-28 12:31:51,175] [    INFO] - loss: 1.60786057, learning_rate: 0.0, global_step: 1, current_memory_allocated: 13.53714394569397, current_memory_reserved: 71.03532481193542, max_memory_allocated: 58.337307929992676, max_memory_reserved: 71.03532481193542, interval_runtime: 340.4979, interval_samples_per_second: 0.3759, interval_steps_per_second: 0.0029, progress_or_epoch: 0.0015
+
+TrainProcess:   2%|▏         | 2/100 [11:33<9:28:32, 348.08s/it][2025-10-28 12:37:44,550] [    INFO] - loss: 1.51283789, learning_rate: 1e-05, global_step: 2, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 353.3973, interval_samples_per_second: 0.3622, interval_steps_per_second: 0.0028, progress_or_epoch: 0.003
+
+TrainProcess:   3%|▎         | 3/100 [17:27<9:27:03, 350.76s/it][2025-10-28 12:43:38,560] [    INFO] - loss: 1.86251259, learning_rate: 2e-05, global_step: 3, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 353.9518, interval_samples_per_second: 0.3616, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0044
+
+TrainProcess:   4%|▍         | 4/100 [23:24<9:25:04, 353.17s/it][2025-10-28 12:49:35,394] [    INFO] - loss: 1.70347214, learning_rate: 3e-05, global_step: 4, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 356.8644, interval_samples_per_second: 0.3587, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0059
+
+TrainProcess:   5%|▌         | 5/100 [29:23<9:22:21, 355.17s/it][2025-10-28 12:55:34,130] [    INFO] - loss: 1.95181084, learning_rate: 4e-05, global_step: 5, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.7268, interval_samples_per_second: 0.3568, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0074
+
+TrainProcess:   6%|▌         | 6/100 [35:22<9:18:41, 356.61s/it][2025-10-28 13:01:33,576] [    INFO] - loss: 1.2135601, learning_rate: 5e-05, global_step: 6, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.4037, interval_samples_per_second: 0.3561, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0089
+
+TrainProcess:   7%|▋         | 7/100 [41:20<9:13:29, 357.09s/it][2025-10-28 13:07:31,631] [    INFO] - loss: 1.13055754, learning_rate: 6e-05, global_step: 7, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.0642, interval_samples_per_second: 0.3575, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0104
+
+TrainProcess:   8%|▊         | 8/100 [47:13<9:05:13, 355.58s/it][2025-10-28 13:13:23,931] [    INFO] - loss: 1.16572213, learning_rate: 7e-05, global_step: 8, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 352.3439, interval_samples_per_second: 0.3633, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0119
+
+TrainProcess:   9%|▉         | 9/100 [53:05<8:57:34, 354.45s/it][2025-10-28 13:19:15,903] [    INFO] - loss: 1.32210732, learning_rate: 8e-05, global_step: 9, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 351.962, interval_samples_per_second: 0.3637, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0133
+
+TrainProcess:  10%|█         | 10/100 [58:58<8:51:00, 354.00s/it][2025-10-28 13:25:08,876] [    INFO] - loss: 1.10681725, learning_rate: 9e-05, global_step: 10, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 353.0065, interval_samples_per_second: 0.3626, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0148
+
+TrainProcess:  11%|█         | 11/100 [1:04:53<8:45:54, 354.54s/it][2025-10-28 13:31:04,727] [    INFO] - loss: 1.18577766, learning_rate: 0.0001, global_step: 11, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.7692, interval_samples_per_second: 0.3598, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0163
+
+TrainProcess:  12%|█▏        | 12/100 [1:10:49<8:40:14, 354.71s/it][2025-10-28 13:36:59,828] [    INFO] - loss: 0.83763123, learning_rate: 9.889e-05, global_step: 12, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.0803, interval_samples_per_second: 0.3605, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0178
+
+TrainProcess:  13%|█▎        | 13/100 [1:16:47<8:35:58, 355.85s/it][2025-10-28 13:42:58,199] [    INFO] - loss: 0.94126701, learning_rate: 9.778e-05, global_step: 13, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.4773, interval_samples_per_second: 0.3571, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0193
+
+TrainProcess:  14%|█▍        | 14/100 [1:22:46<8:31:26, 356.82s/it][2025-10-28 13:48:57,341] [    INFO] - loss: 0.91316128, learning_rate: 9.667e-05, global_step: 14, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.0772, interval_samples_per_second: 0.3565, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0207
+
+TrainProcess:  15%|█▌        | 15/100 [1:28:44<8:26:00, 357.19s/it][2025-10-28 13:54:55,342] [    INFO] - loss: 0.83015823, learning_rate: 9.556e-05, global_step: 15, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.0253, interval_samples_per_second: 0.3575, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0222
+
+TrainProcess:  16%|█▌        | 16/100 [1:34:44<8:21:06, 357.94s/it][2025-10-28 14:00:55,025] [    INFO] - loss: 0.96580505, learning_rate: 9.444e-05, global_step: 16, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.6829, interval_samples_per_second: 0.3559, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0237
+
+TrainProcess:  17%|█▋        | 17/100 [1:40:46<8:16:49, 359.15s/it][2025-10-28 14:06:57,066] [    INFO] - loss: 0.83217621, learning_rate: 9.333e-05, global_step: 17, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 361.9607, interval_samples_per_second: 0.3536, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0252
+
+TrainProcess:  18%|█▊        | 18/100 [1:46:47<8:11:37, 359.72s/it][2025-10-28 14:12:58,081] [    INFO] - loss: 0.89361572, learning_rate: 9.222e-05, global_step: 18, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 361.0562, interval_samples_per_second: 0.3545, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0267
+
+TrainProcess:  19%|█▉        | 19/100 [1:52:50<8:07:06, 360.83s/it][2025-10-28 14:19:01,456] [    INFO] - loss: 0.82954788, learning_rate: 9.111e-05, global_step: 19, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 363.398, interval_samples_per_second: 0.3522, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0281
+
+TrainProcess:  20%|██        | 20/100 [1:58:51<8:01:07, 360.84s/it][2025-10-28 14:25:02,318] [    INFO] - loss: 0.75273323, learning_rate: 9e-05, global_step: 20, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.8814, interval_samples_per_second: 0.3547, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0296
+
+TrainProcess:  21%|██        | 21/100 [2:04:53<7:55:42, 361.30s/it][2025-10-28 14:31:04,718] [    INFO] - loss: 0.90951347, learning_rate: 8.889e-05, global_step: 21, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.3605, interval_samples_per_second: 0.3532, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0311
+
+TrainProcess:  22%|██▏       | 22/100 [2:10:58<7:50:48, 362.17s/it][2025-10-28 14:37:08,844] [    INFO] - loss: 0.82700348, learning_rate: 8.778e-05, global_step: 22, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 364.192, interval_samples_per_second: 0.3515, interval_steps_per_second: 0.0027, progress_or_epoch: 0.0326
+
+TrainProcess:  23%|██▎       | 23/100 [2:17:00<7:45:00, 362.34s/it][2025-10-28 14:43:11,663] [    INFO] - loss: 0.90127945, learning_rate: 8.667e-05, global_step: 23, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.7566, interval_samples_per_second: 0.3529, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0341
+
+TrainProcess:  24%|██▍       | 24/100 [2:23:03<7:38:59, 362.36s/it][2025-10-28 14:49:14,035] [    INFO] - loss: 0.79956627, learning_rate: 8.556e-05, global_step: 24, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.399, interval_samples_per_second: 0.3532, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0356
+
+TrainProcess:  25%|██▌       | 25/100 [2:29:04<7:32:19, 361.86s/it][2025-10-28 14:55:14,741] [    INFO] - loss: 0.64152718, learning_rate: 8.444e-05, global_step: 25, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.7004, interval_samples_per_second: 0.3549, interval_steps_per_second: 0.0028, progress_or_epoch: 0.037
+
+TrainProcess:  26%|██▌       | 26/100 [2:35:04<7:25:36, 361.31s/it][2025-10-28 15:01:14,719] [    INFO] - loss: 0.81254077, learning_rate: 8.333e-05, global_step: 26, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.0036, interval_samples_per_second: 0.3556, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0385
+
+TrainProcess:  27%|██▋       | 27/100 [2:41:06<7:20:04, 361.70s/it][2025-10-28 15:07:17,396] [    INFO] - loss: 0.67437172, learning_rate: 8.222e-05, global_step: 27, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.6329, interval_samples_per_second: 0.353, interval_steps_per_second: 0.0028, progress_or_epoch: 0.04
+
+TrainProcess:  28%|██▊       | 28/100 [2:47:08<7:14:08, 361.79s/it][2025-10-28 15:13:19,406] [    INFO] - loss: 0.66415334, learning_rate: 8.111e-05, global_step: 28, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 361.9815, interval_samples_per_second: 0.3536, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0415
+
+TrainProcess:  29%|██▉       | 29/100 [2:53:11<7:08:33, 362.17s/it][2025-10-28 15:19:22,390] [    INFO] - loss: 0.73983574, learning_rate: 8e-05, global_step: 29, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 363.0624, interval_samples_per_second: 0.3526, interval_steps_per_second: 0.0028, progress_or_epoch: 0.043
+
+TrainProcess:  30%|███       | 30/100 [2:59:13<7:02:30, 362.15s/it][2025-10-28 15:25:24,790] [    INFO] - loss: 0.58006668, learning_rate: 7.889e-05, global_step: 30, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.3652, interval_samples_per_second: 0.3532, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0444
+
+TrainProcess:  31%|███       | 31/100 [3:05:15<6:56:25, 362.11s/it][2025-10-28 15:31:26,597] [    INFO] - loss: 0.71074772, learning_rate: 7.778e-05, global_step: 31, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 361.7462, interval_samples_per_second: 0.3538, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0459
+
+TrainProcess:  32%|███▏      | 32/100 [3:11:18<6:50:30, 362.21s/it][2025-10-28 15:37:29,004] [    INFO] - loss: 0.63714218, learning_rate: 7.667e-05, global_step: 32, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.4523, interval_samples_per_second: 0.3531, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0474
+
+TrainProcess:  33%|███▎      | 33/100 [3:17:15<6:42:40, 360.61s/it][2025-10-28 15:43:25,879] [    INFO] - loss: 0.85639572, learning_rate: 7.556e-05, global_step: 33, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 356.8742, interval_samples_per_second: 0.3587, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0489
+
+TrainProcess:  34%|███▍      | 34/100 [3:23:09<6:34:40, 358.79s/it][2025-10-28 15:49:20,479] [    INFO] - loss: 0.64079809, learning_rate: 7.444e-05, global_step: 34, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.5511, interval_samples_per_second: 0.361, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0504
+
+TrainProcess:  35%|███▌      | 35/100 [3:29:04<6:27:26, 357.63s/it][2025-10-28 15:55:15,316] [    INFO] - loss: 0.72723579, learning_rate: 7.333e-05, global_step: 35, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.9304, interval_samples_per_second: 0.3606, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0519
+
+TrainProcess:  36%|███▌      | 36/100 [3:34:59<6:20:35, 356.80s/it][2025-10-28 16:01:10,243] [    INFO] - loss: 0.52463913, learning_rate: 7.222e-05, global_step: 36, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.8725, interval_samples_per_second: 0.3607, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0533
+
+TrainProcess:  37%|███▋      | 37/100 [3:40:56<6:14:50, 357.00s/it][2025-10-28 16:07:07,616] [    INFO] - loss: 0.75466061, learning_rate: 7.111e-05, global_step: 37, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.435, interval_samples_per_second: 0.3581, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0548
+
+TrainProcess:  38%|███▊      | 38/100 [3:46:52<6:08:27, 356.57s/it][2025-10-28 16:13:03,163] [    INFO] - loss: 0.9297905, learning_rate: 7e-05, global_step: 38, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.573, interval_samples_per_second: 0.36, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0563
+
+TrainProcess:  39%|███▉      | 39/100 [3:52:50<6:03:01, 357.08s/it][2025-10-28 16:19:01,470] [    INFO] - loss: 0.56610394, learning_rate: 6.889e-05, global_step: 39, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.2607, interval_samples_per_second: 0.3573, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0578
+
+TrainProcess:  40%|████      | 40/100 [3:58:46<5:56:45, 356.75s/it][2025-10-28 16:24:57,464] [    INFO] - loss: 0.64287567, learning_rate: 6.778e-05, global_step: 40, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.9992, interval_samples_per_second: 0.3596, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0593
+
+TrainProcess:  41%|████      | 41/100 [4:04:42<5:50:36, 356.55s/it][2025-10-28 16:30:53,609] [    INFO] - loss: 0.68204212, learning_rate: 6.667e-05, global_step: 41, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 356.0798, interval_samples_per_second: 0.3595, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0607
+
+TrainProcess:  42%|████▏     | 42/100 [4:10:33<5:42:52, 354.70s/it][2025-10-28 16:36:43,933] [    INFO] - loss: 0.65622902, learning_rate: 6.556e-05, global_step: 42, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 350.3714, interval_samples_per_second: 0.3653, interval_steps_per_second: 0.0029, progress_or_epoch: 0.0622
+
+TrainProcess:  43%|████▎     | 43/100 [4:16:27<5:36:46, 354.49s/it][2025-10-28 16:42:37,921] [    INFO] - loss: 0.83555603, learning_rate: 6.444e-05, global_step: 43, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.019, interval_samples_per_second: 0.3616, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0637
+
+TrainProcess:  44%|████▍     | 44/100 [4:22:20<5:30:34, 354.19s/it][2025-10-28 16:48:31,415] [    INFO] - loss: 0.58517838, learning_rate: 6.333e-05, global_step: 44, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 353.4936, interval_samples_per_second: 0.3621, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0652
+
+TrainProcess:  45%|████▌     | 45/100 [4:28:16<5:25:08, 354.70s/it][2025-10-28 16:54:27,313] [    INFO] - loss: 0.66253853, learning_rate: 6.222e-05, global_step: 45, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.8688, interval_samples_per_second: 0.3597, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0667
+
+TrainProcess:  46%|████▌     | 46/100 [4:34:16<5:20:31, 356.13s/it][2025-10-28 17:00:26,761] [    INFO] - loss: 0.66833591, learning_rate: 6.111e-05, global_step: 46, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.4868, interval_samples_per_second: 0.3561, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0682
+
+TrainProcess:  47%|████▋     | 47/100 [4:40:13<5:15:03, 356.67s/it][2025-10-28 17:06:24,684] [    INFO] - loss: 0.51014137, learning_rate: 6e-05, global_step: 47, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.9251, interval_samples_per_second: 0.3576, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0696
+
+TrainProcess:  48%|████▊     | 48/100 [4:46:12<5:09:33, 357.18s/it][2025-10-28 17:12:23,139] [    INFO] - loss: 0.79288578, learning_rate: 5.889e-05, global_step: 48, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.3793, interval_samples_per_second: 0.3572, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0711
+
+TrainProcess:  49%|████▉     | 49/100 [4:52:12<5:04:20, 358.05s/it][2025-10-28 17:18:23,169] [    INFO] - loss: 0.72906876, learning_rate: 5.778e-05, global_step: 49, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.0793, interval_samples_per_second: 0.3555, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0726
+
+TrainProcess:  50%|█████     | 50/100 [4:58:09<4:58:13, 357.87s/it][2025-10-28 17:24:20,596] [    INFO] - loss: 0.64970589, learning_rate: 5.667e-05, global_step: 50, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.4549, interval_samples_per_second: 0.3581, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0741
+
+TrainProcess:  51%|█████     | 51/100 [5:04:05<4:51:47, 357.31s/it][2025-10-28 17:30:16,613] [    INFO] - loss: 0.75118637, learning_rate: 5.556e-05, global_step: 51, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.9829, interval_samples_per_second: 0.3596, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0756
+
+TrainProcess:  52%|█████▏    | 52/100 [5:10:03<4:45:57, 357.45s/it][2025-10-28 17:36:14,368] [    INFO] - loss: 0.83564568, learning_rate: 5.444e-05, global_step: 52, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.7952, interval_samples_per_second: 0.3577, interval_steps_per_second: 0.0028, progress_or_epoch: 0.077
+
+TrainProcess:  53%|█████▎    | 53/100 [5:16:02<4:40:15, 357.77s/it][2025-10-28 17:42:12,913] [    INFO] - loss: 0.55648136, learning_rate: 5.333e-05, global_step: 53, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.5152, interval_samples_per_second: 0.357, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0785
+
+TrainProcess:  54%|█████▍    | 54/100 [5:22:02<4:34:51, 358.50s/it][2025-10-28 17:48:13,138] [    INFO] - loss: 0.70081615, learning_rate: 5.222e-05, global_step: 54, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.2087, interval_samples_per_second: 0.3553, interval_steps_per_second: 0.0028, progress_or_epoch: 0.08
+
+TrainProcess:  55%|█████▌    | 55/100 [5:28:00<4:28:47, 358.38s/it][2025-10-28 17:54:11,224] [    INFO] - loss: 0.52258778, learning_rate: 5.111e-05, global_step: 55, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.1062, interval_samples_per_second: 0.3574, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0815
+
+TrainProcess:  56%|█████▌    | 56/100 [5:33:58<4:22:38, 358.14s/it][2025-10-28 18:00:08,784] [    INFO] - loss: 0.58846283, learning_rate: 5e-05, global_step: 56, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.571, interval_samples_per_second: 0.358, interval_steps_per_second: 0.0028, progress_or_epoch: 0.083
+
+TrainProcess:  57%|█████▋    | 57/100 [5:39:55<4:16:24, 357.78s/it][2025-10-28 18:06:05,743] [    INFO] - loss: 0.69686508, learning_rate: 4.889e-05, global_step: 57, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 356.9413, interval_samples_per_second: 0.3586, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0844
+
+TrainProcess:  58%|█████▊    | 58/100 [5:45:53<4:10:31, 357.90s/it][2025-10-28 18:12:03,985] [    INFO] - loss: 0.53733444, learning_rate: 4.778e-05, global_step: 58, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.197, interval_samples_per_second: 0.3573, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0859
+
+TrainProcess:  59%|█████▉    | 59/100 [5:51:52<4:04:53, 358.39s/it][2025-10-28 18:18:03,508] [    INFO] - loss: 0.56767511, learning_rate: 4.667e-05, global_step: 59, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.513, interval_samples_per_second: 0.356, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0874
+
+TrainProcess:  60%|██████    | 60/100 [5:57:53<3:59:20, 359.02s/it][2025-10-28 18:24:04,025] [    INFO] - loss: 0.57501793, learning_rate: 4.556e-05, global_step: 60, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.5046, interval_samples_per_second: 0.3551, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0889
+
+TrainProcess:  61%|██████    | 61/100 [6:03:54<3:53:43, 359.58s/it][2025-10-28 18:30:05,009] [    INFO] - loss: 0.7000761, learning_rate: 4.444e-05, global_step: 61, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.878, interval_samples_per_second: 0.3547, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0904
+
+TrainProcess:  62%|██████▏   | 62/100 [6:09:53<3:47:37, 359.40s/it][2025-10-28 18:36:03,828] [    INFO] - loss: 0.53203249, learning_rate: 4.333e-05, global_step: 62, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.9899, interval_samples_per_second: 0.3566, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0919
+
+TrainProcess:  63%|██████▎   | 63/100 [6:15:53<3:41:46, 359.63s/it][2025-10-28 18:42:03,958] [    INFO] - loss: 0.81850052, learning_rate: 4.222e-05, global_step: 63, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.1651, interval_samples_per_second: 0.3554, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0933
+
+TrainProcess:  64%|██████▍   | 64/100 [6:21:53<3:35:54, 359.85s/it][2025-10-28 18:48:04,346] [    INFO] - loss: 0.54692841, learning_rate: 4.111e-05, global_step: 64, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.3459, interval_samples_per_second: 0.3552, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0948
+
+TrainProcess:  65%|██████▌   | 65/100 [6:27:53<3:29:58, 359.94s/it][2025-10-28 18:54:04,525] [    INFO] - loss: 0.47618985, learning_rate: 4e-05, global_step: 65, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.1738, interval_samples_per_second: 0.3554, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0963
+
+TrainProcess:  66%|██████▌   | 66/100 [6:33:56<3:24:26, 360.79s/it][2025-10-28 19:00:07,280] [    INFO] - loss: 0.4837141, learning_rate: 3.889e-05, global_step: 66, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 362.7479, interval_samples_per_second: 0.3529, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0978
+
+TrainProcess:  67%|██████▋   | 67/100 [6:39:56<3:18:17, 360.54s/it][2025-10-28 19:06:07,208] [    INFO] - loss: 0.74391174, learning_rate: 3.778e-05, global_step: 67, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.9777, interval_samples_per_second: 0.3556, interval_steps_per_second: 0.0028, progress_or_epoch: 0.0993
+
+TrainProcess:  68%|██████▊   | 68/100 [6:45:56<3:12:14, 360.45s/it][2025-10-28 19:12:07,445] [    INFO] - loss: 0.50230885, learning_rate: 3.667e-05, global_step: 68, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.2354, interval_samples_per_second: 0.3553, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1007
+
+TrainProcess:  69%|██████▉   | 69/100 [6:51:55<3:06:01, 360.04s/it][2025-10-28 19:18:06,553] [    INFO] - loss: 0.53715038, learning_rate: 3.556e-05, global_step: 69, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.095, interval_samples_per_second: 0.3565, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1022
+
+TrainProcess:  70%|███████   | 70/100 [6:57:55<3:00:01, 360.04s/it][2025-10-28 19:24:06,581] [    INFO] - loss: 0.52559185, learning_rate: 3.444e-05, global_step: 70, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.032, interval_samples_per_second: 0.3555, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1037
+
+TrainProcess:  71%|███████   | 71/100 [7:03:55<2:53:56, 359.88s/it][2025-10-28 19:30:06,092] [    INFO] - loss: 0.7343874, learning_rate: 3.333e-05, global_step: 71, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.4947, interval_samples_per_second: 0.3561, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1052
+
+TrainProcess:  72%|███████▏  | 72/100 [7:09:53<2:47:45, 359.48s/it][2025-10-28 19:36:04,632] [    INFO] - loss: 0.53286171, learning_rate: 3.222e-05, global_step: 72, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.541, interval_samples_per_second: 0.357, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1067
+
+TrainProcess:  73%|███████▎  | 73/100 [7:15:50<2:41:25, 358.72s/it][2025-10-28 19:42:01,602] [    INFO] - loss: 0.61263633, learning_rate: 3.111e-05, global_step: 73, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 356.9623, interval_samples_per_second: 0.3586, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1082
+
+TrainProcess:  74%|███████▍  | 74/100 [7:21:49<2:35:26, 358.73s/it][2025-10-28 19:48:00,347] [    INFO] - loss: 0.65122676, learning_rate: 3e-05, global_step: 74, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.7387, interval_samples_per_second: 0.3568, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1096
+
+TrainProcess:  75%|███████▌  | 75/100 [7:27:48<2:29:31, 358.85s/it][2025-10-28 19:53:59,434] [    INFO] - loss: 0.76685905, learning_rate: 2.889e-05, global_step: 75, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.1282, interval_samples_per_second: 0.3564, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1111
+
+TrainProcess:  76%|███████▌  | 76/100 [7:33:48<2:23:36, 359.02s/it][2025-10-28 19:59:58,930] [    INFO] - loss: 0.66427612, learning_rate: 2.778e-05, global_step: 76, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.4081, interval_samples_per_second: 0.3561, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1126
+
+TrainProcess:  77%|███████▋  | 77/100 [7:39:48<2:17:45, 359.37s/it][2025-10-28 20:05:59,101] [    INFO] - loss: 0.42409515, learning_rate: 2.667e-05, global_step: 77, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 360.2089, interval_samples_per_second: 0.3553, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1141
+
+TrainProcess:  78%|███████▊  | 78/100 [7:45:45<2:11:32, 358.76s/it][2025-10-28 20:11:56,382] [    INFO] - loss: 0.59027195, learning_rate: 2.556e-05, global_step: 78, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.333, interval_samples_per_second: 0.3582, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1156
+
+TrainProcess:  79%|███████▉  | 79/100 [7:51:40<2:05:07, 357.51s/it][2025-10-28 20:17:51,016] [    INFO] - loss: 0.59920692, learning_rate: 2.444e-05, global_step: 79, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.5942, interval_samples_per_second: 0.361, interval_steps_per_second: 0.0028, progress_or_epoch: 0.117
+
+TrainProcess:  80%|████████  | 80/100 [7:57:35<1:58:56, 356.82s/it][2025-10-28 20:23:46,227] [    INFO] - loss: 0.56145287, learning_rate: 2.333e-05, global_step: 80, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 355.1939, interval_samples_per_second: 0.3604, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1185
+
+TrainProcess:  81%|████████  | 81/100 [8:03:31<1:52:57, 356.71s/it][2025-10-28 20:29:42,681] [    INFO] - loss: 0.52036846, learning_rate: 2.222e-05, global_step: 81, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 356.4659, interval_samples_per_second: 0.3591, interval_steps_per_second: 0.0028, progress_or_epoch: 0.12
+
+TrainProcess:  82%|████████▏ | 82/100 [8:09:29<1:47:05, 356.97s/it][2025-10-28 20:35:40,205] [    INFO] - loss: 0.57490253, learning_rate: 2.111e-05, global_step: 82, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.569, interval_samples_per_second: 0.358, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1215
+
+TrainProcess:  83%|████████▎ | 83/100 [8:15:27<1:41:12, 357.20s/it][2025-10-28 20:41:38,026] [    INFO] - loss: 0.81481838, learning_rate: 2e-05, global_step: 83, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.7354, interval_samples_per_second: 0.3578, interval_steps_per_second: 0.0028, progress_or_epoch: 0.123
+
+TrainProcess:  84%|████████▍ | 84/100 [8:21:28<1:35:33, 358.36s/it][2025-10-28 20:47:39,106] [    INFO] - loss: 0.71151018, learning_rate: 1.889e-05, global_step: 84, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 361.0516, interval_samples_per_second: 0.3545, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1245
+
+TrainProcess:  85%|████████▌ | 85/100 [8:27:22<1:29:15, 357.01s/it][2025-10-28 20:53:32,856] [    INFO] - loss: 0.42023945, learning_rate: 1.778e-05, global_step: 85, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 353.8644, interval_samples_per_second: 0.3617, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1259
+
+TrainProcess:  86%|████████▌ | 86/100 [8:33:17<1:23:09, 356.37s/it][2025-10-28 20:59:27,753] [    INFO] - loss: 0.67935085, learning_rate: 1.667e-05, global_step: 86, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.8863, interval_samples_per_second: 0.3607, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1274
+
+TrainProcess:  87%|████████▋ | 87/100 [8:39:09<1:16:57, 355.21s/it][2025-10-28 21:05:20,291] [    INFO] - loss: 0.80524445, learning_rate: 1.556e-05, global_step: 87, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 352.4942, interval_samples_per_second: 0.3631, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1289
+
+TrainProcess:  88%|████████▊ | 88/100 [8:45:03<1:10:58, 354.87s/it][2025-10-28 21:11:14,341] [    INFO] - loss: 0.66514492, learning_rate: 1.444e-05, global_step: 88, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.0833, interval_samples_per_second: 0.3615, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1304
+
+TrainProcess:  89%|████████▉ | 89/100 [8:51:01<1:05:11, 355.63s/it][2025-10-28 21:17:11,754] [    INFO] - loss: 0.72842598, learning_rate: 1.333e-05, global_step: 89, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.3917, interval_samples_per_second: 0.3582, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1319
+
+TrainProcess:  90%|█████████ | 90/100 [8:56:58<59:22, 356.29s/it]  [2025-10-28 21:23:09,537] [    INFO] - loss: 0.52937889, learning_rate: 1.222e-05, global_step: 90, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.8176, interval_samples_per_second: 0.3577, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1333
+
+TrainProcess:  91%|█████████ | 91/100 [9:02:53<53:22, 355.84s/it][2025-10-28 21:29:04,343] [    INFO] - loss: 0.5971241, learning_rate: 1.111e-05, global_step: 91, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.8028, interval_samples_per_second: 0.3608, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1348
+
+TrainProcess:  92%|█████████▏| 92/100 [9:08:48<47:23, 355.48s/it][2025-10-28 21:34:58,985] [    INFO] - loss: 0.53361177, learning_rate: 1e-05, global_step: 92, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.6461, interval_samples_per_second: 0.3609, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1363
+
+TrainProcess:  93%|█████████▎| 93/100 [9:14:42<41:24, 354.99s/it][2025-10-28 21:40:52,803] [    INFO] - loss: 0.55367661, learning_rate: 8.889e-06, global_step: 93, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 353.8361, interval_samples_per_second: 0.3617, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1378
+
+TrainProcess:  94%|█████████▍| 94/100 [9:20:41<35:37, 356.31s/it][2025-10-28 21:46:52,241] [    INFO] - loss: 0.72559881, learning_rate: 7.778e-06, global_step: 94, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.3903, interval_samples_per_second: 0.3562, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1393
+
+TrainProcess:  95%|█████████▌| 95/100 [9:26:41<29:46, 357.32s/it][2025-10-28 21:52:51,959] [    INFO] - loss: 0.50272036, learning_rate: 6.667e-06, global_step: 95, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.6777, interval_samples_per_second: 0.3559, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1407
+
+TrainProcess:  96%|█████████▌| 96/100 [9:32:40<23:51, 357.78s/it][2025-10-28 21:58:50,781] [    INFO] - loss: 0.31370783, learning_rate: 5.556e-06, global_step: 96, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 358.8487, interval_samples_per_second: 0.3567, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1422
+
+TrainProcess:  97%|█████████▋| 97/100 [9:38:39<17:54, 358.26s/it][2025-10-28 22:04:50,138] [    INFO] - loss: 0.41887093, learning_rate: 4.444e-06, global_step: 97, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 359.3883, interval_samples_per_second: 0.3562, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1437
+
+TrainProcess:  98%|█████████▊| 98/100 [9:44:34<11:54, 357.28s/it][2025-10-28 22:10:45,158] [    INFO] - loss: 0.5729003, learning_rate: 3.333e-06, global_step: 98, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.9787, interval_samples_per_second: 0.3606, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1452
+
+TrainProcess:  99%|█████████▉| 99/100 [9:50:28<05:56, 356.34s/it][2025-10-28 22:16:39,287] [    INFO] - loss: 0.64190578, learning_rate: 2.222e-06, global_step: 99, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 354.1516, interval_samples_per_second: 0.3614, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1467
+
+TrainProcess: 100%|██████████| 100/100 [9:56:25<00:00, 356.63s/it][2025-10-28 22:22:36,556] [    INFO] - loss: 0.50331497, learning_rate: 1.111e-06, global_step: 100, current_memory_allocated: 13.53714370727539, current_memory_reserved: 71.03532481193542, max_memory_allocated: 59.20164966583252, max_memory_reserved: 71.03532481193542, interval_runtime: 357.3085, interval_samples_per_second: 0.3582, interval_steps_per_second: 0.0028, progress_or_epoch: 0.1482
+[2025-10-28 22:22:36,556] [    INFO] - 
+Training completed. 
+
+[2025-10-28 22:22:36,575] [    INFO] - train_runtime: 35785.9312, train_samples_per_second: 0.3577, train_steps_per_second: 0.0028, train_loss: 0.7501962792873382, progress_or_epoch: 0.1482
+
+TrainProcess: 100%|██████████| 100/100 [9:56:25<00:00, 357.86s/it]
+[2025-10-28 22:22:36,578] [    INFO] - Saving model checkpoint to tmp_train/bge-en-icl-paddle
+[2025-10-28 22:22:36,583] [    INFO] - tokenizer config file saved in tmp_train/bge-en-icl-paddle/tokenizer_config.json
+[2025-10-28 22:22:36,613] [    INFO] - Special tokens file saved in tmp_train/bge-en-icl-paddle/special_tokens_map.json
+[2025-10-28 22:22:36,614] [    INFO] - added tokens file saved in tmp_train/bge-en-icl-paddle/added_tokens.json
+[2025-10-28 22:22:37,310] [    INFO] - Configuration saved in tmp_train/bge-en-icl-paddle/config.json
+[2025-10-28 22:22:37,310] [    INFO] - ***** train metrics *****
+[2025-10-28 22:22:37,311] [    INFO] -   progress_or_epoch        =     0.1482
+[2025-10-28 22:22:37,311] [    INFO] -   train_loss               =     0.7502
+[2025-10-28 22:22:37,311] [    INFO] -   train_runtime            = 9:56:25.93
+[2025-10-28 22:22:37,311] [    INFO] -   train_samples_per_second =     0.3577
+[2025-10-28 22:22:37,311] [    INFO] -   train_steps_per_second   =     0.0028
+LAUNCH INFO 2025-10-28 22:22:40,810 Pod completed
+LAUNCH INFO 2025-10-28 22:22:40,811 Exit code 0

bge-en-icl-paddle/train_results.json

@@ -0,0 +1,7 @@
+{
+    "progress_or_epoch": 0.1482,
+    "train_loss": 0.7501962792873382,
+    "train_runtime": 35785.9312,
+    "train_samples_per_second": 0.3577,
+    "train_steps_per_second": 0.0028
+}

bge-en-icl-paddle/trainer_state.json

@@ -0,0 +1,15 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "consumed_samples": 12800,
+  "epoch": 0.14815500717625815,
+  "global_step": 100,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [],
+  "max_steps": 100,
+  "num_train_epochs": 1,
+  "total_flos": 0,
+  "trial_name": null,
+  "trial_params": null
+}

bge-en-icl-paddle/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d24ec006bd8c16bf891b27700ca54bfd552c5344585e1f7c7b69750c0db33a2b
+size 3754