[2026-11-05 11:42:39,984] [DEBUG] [axolotl.utils.config.resolve_dtype:168] [PID:985839] bf16 support detected, enabling for this configuration. [2026-11-05 11:42:40,199] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:985839] baseline 0.000GB () [2026-11-05 11:42:40,200] [INFO] [axolotl.cli.config.load_cfg:339] [PID:985839] config: { "activation_offloading": true, "adapter": "qlora", "adapter_checkpoint_format": "gather", "attn_decontaminates_packing": true, "attn_implementation": "sdpa", "attn_needs_dtype_cast": false, "attn_supports_packing": false, "attn_uses_flash_lib": false, "axolotl_config_path": "glm53flash-merge.yml", "balance_labels": false, "base_model": "zai-org/GLM-5.3-Flash-BF16", "base_model_config": "zai-org/GLM-5.3-Flash-BF16", "batch_size": 8, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_120", "fp8": true, "n_gpu": 1, "n_node": 1, "tf32": true }, "context_parallel_size": 1, "cut_cross_entropy": true, "cut_cross_entropy_accum_c_fp32": false, "dataloader_num_workers": 1, "dataloader_pin_memory": true, "dataloader_prefetch_factor": 256, "dataset_num_proc": 64, "dataset_prepared_path": "last_run_prepared", "datasets": [ { "chat_template": "tokenizer_default", "message_property_mappings": { "content": "content", "role": "role" }, "path": "/home/anon/LoRA/Moni_messages.json", "trust_remote_code": false, "type": "chat_template" } ], "ddp": false, "device": "cuda:0", "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "eaft_alpha": 1.0, "eaft_k": 20, "env_capabilities": { "torch_version": "2.13.0" }, "eval_batch_size": 1, "eval_causal_lm_metrics": [ "sacrebleu", "comet", "ter", "chrf" ], "eval_max_new_tokens": 128, "eval_sample_packing": false, "eval_steps": 0.08333333333333333, "eval_table_size": 0, "evals_per_epoch": 4, "experimental_skip_move_to_device": true, "expert_parallel_backend": "torch", "expert_parallel_dispatch_chunks": 1, "expert_parallel_fallback_on_unsupported": true, "expert_parallel_num_nvl_bytes": 268435456, "expert_parallel_num_rdma_bytes": 0, "expert_parallel_save_dispatch": false, "expert_parallel_size": 1, "fa_varlen_kwargs_in_collator": true, "fp16": false, "fsdp_config": { "activation_checkpointing": true, "auto_wrap_policy": "TRANSFORMER_BASED_WRAP", "cpu_ram_efficient_loading": true, "fsdp_version": 2, "offload_params": false, "reshard_after_forward": true, "state_dict_type": "FULL_STATE_DICT", "transformer_layer_cls_to_wrap": "Glm5NextTextDecoderLayer" }, "fsdp_version": 2, "generate_samples": false, "generation_do_sample": true, "generation_max_new_tokens": 50, "generation_prompt_ratio": 0.5, "generation_temperature": 0.7, "gradient_accumulation_steps": 8, "gradient_checkpointing": false, "include_tkps": true, "is_multimodal": true, "label_balance_window_optim_steps": 1, "layer_offloading": false, "learning_rate": 0.0001, "lisa_layers_attribute": "model.layers", "load_best_model_at_end": false, "load_in_4bit": true, "load_in_8bit": false, "local_rank": 0, "logging_nan_inf_filter": true, "logging_steps": 1, "lora_alpha": 64, "lora_dropout": 0.0, "lora_embedding_kernel": false, "lora_fp32_gradients": false, "lora_mlp_kernel": false, "lora_model_dir": "./outputs/glm53-flash-qlora-out/checkpoint-234", "lora_o_kernel": false, "lora_qkv_kernel": false, "lora_r": 32, "lora_target_modules": [ "q_a_proj", "q_b_proj", "kv_a_proj_with_mqa", "kv_b_proj", "q_proj", "k_proj", "v_proj", "o_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "mean_resizing_embeddings": false, "merge_compile": false, "merge_compile_precision": "strict", "merge_method": "memory_efficient", "merge_num_threads": 2, "micro_batch_size": 1, "mixlora_jitter_noise": 0.0, "mixlora_num_experts": 8, "mixlora_router_aux_loss_coef": 0.01, "mixlora_router_init_range": 0.02, "mixlora_top_k": 2, "model_config_type": "glm5_next", "model_config_type_text": "glm5_next_text", "nf4_backend": "bitsandbytes", "num_epochs": 3.0, "num_generation_samples": 3, "optimizer": "adamw_torch_8bit", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "./outputs/glm53-flash-qlora-out", "ple_cpu_offload": false, "plugins": [ "axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin" ], "pretrain_multipack_attn": true, "processor_config": "zai-org/GLM-5.3-Flash-BF16", "profiler_steps_start": 0, "qgalore_cos_threshold": 0.4, "qgalore_gamma_proj": 2, "qgalore_proj_bits": 4, "qgalore_proj_group_size": 256, "qgalore_proj_quant": true, "qgalore_proj_type": "std", "qgalore_queue_size": 5, "qgalore_rank": 256, "qgalore_scale": 0.25, "qgalore_update_proj_gap": 200, "qlora_sharded_model_loading": true, "quantize_moe_experts": true, "ray_num_workers": 1, "relora_prune_method": "magnitude", "resources_per_worker": { "GPU": 1 }, "sample_packing": false, "sample_packing_bin_size": 200, "sample_packing_group_size": 100000, "save_only_model": false, "save_safetensors": true, "save_steps": 0.3333333333333333, "saves_per_epoch": 1, "sequence_len": 3072, "shuffle_before_merging_datasets": false, "shuffle_merged_datasets": true, "skip_prepare_dataset": false, "streaming_multipack_buffer_size": 10000, "strict": false, "sync_profiler_start_step": 2, "tensor_parallel_size": 1, "tf32": false, "tiled_mlp_use_original_mlp": true, "tokenizer_config": "zai-org/GLM-5.3-Flash-BF16", "tokenizer_save_jinja_files": true, "torch_compile_regional": false, "torch_dtype": "torch.bfloat16", "train_on_inputs": false, "trl": { "async_prefetch": false, "log_completions": false, "mask_truncated_completions": false, "ref_model_mixup_alpha": 0.9, "ref_model_sync_steps": 64, "replay_buffer_size": 0, "replay_recompute_logps": true, "reroll_max_groups": 1, "reroll_start_fraction": 1.0, "reward_num_workers": 1, "scale_rewards": true, "skip_zero_advantage_batches": true, "sync_ref_model": false, "use_data_producer": false, "use_vllm": false, "vllm_lora_sync": false, "vllm_server_host": "0.0.0.0", "vllm_server_port": 8000 }, "use_otel_metrics": false, "use_ray": false, "val_set_size": 0.1, "vllm": { "device": "auto", "dtype": "auto", "host": "0.0.0.0", "port": 8000 }, "warmup_ratio": 0.1, "weight_decay": 0.0, "world_size": 1 } [2026-11-05 11:42:40,293] [WARNING] [axolotl.utils.schemas.validation.check_activation_offloading_wo_gc:2142] [PID:985839] activation_offloading: True without gradient_checkpointing offloads every activation with no recompute, which is bound by host RAM and PCIe bandwidth; set gradient_checkpointing: true to recompute the layers and offload only what is saved outside them. [2026-11-05 11:42:40,396] [DEBUG] [axolotl.utils.config.resolve_dtype:168] [PID:985839] bf16 support detected, enabling for this configuration. [2026-11-05 11:42:40,441] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:985839] baseline 0.000GB () [2026-11-05 11:42:40,442] [INFO] [axolotl.cli.config.load_cfg:339] [PID:985839] config: { "activation_offloading": true, "adapter": "qlora", "adapter_checkpoint_format": "gather", "attn_decontaminates_packing": false, "attn_needs_dtype_cast": false, "attn_supports_packing": false, "attn_uses_flash_lib": false, "axolotl_config_path": "glm53flash-merge.yml", "balance_labels": false, "base_model": "zai-org/GLM-5.3-Flash-BF16", "base_model_config": "zai-org/GLM-5.3-Flash-BF16", "batch_size": 8, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_120", "fp8": true, "n_gpu": 1, "n_node": 1, "tf32": true }, "context_parallel_size": 1, "cut_cross_entropy": true, "cut_cross_entropy_accum_c_fp32": false, "dataloader_num_workers": 1, "dataloader_pin_memory": true, "dataloader_prefetch_factor": 256, "dataset_num_proc": 64, "dataset_prepared_path": "last_run_prepared", "datasets": [ { "chat_template": "tokenizer_default", "message_property_mappings": { "content": "content", "role": "role" }, "path": "/home/anon/LoRA/Moni_messages.json", "trust_remote_code": false, "type": "chat_template" } ], "ddp": false, "device": "cuda:0", "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "eaft_alpha": 1.0, "eaft_k": 20, "env_capabilities": { "torch_version": "2.13.0" }, "eval_batch_size": 1, "eval_causal_lm_metrics": [ "sacrebleu", "comet", "ter", "chrf" ], "eval_max_new_tokens": 128, "eval_sample_packing": false, "eval_steps": 0.08333333333333333, "eval_table_size": 0, "evals_per_epoch": 4, "experimental_skip_move_to_device": true, "expert_parallel_backend": "torch", "expert_parallel_dispatch_chunks": 1, "expert_parallel_fallback_on_unsupported": true, "expert_parallel_num_nvl_bytes": 268435456, "expert_parallel_num_rdma_bytes": 0, "expert_parallel_save_dispatch": false, "expert_parallel_size": 1, "fa_varlen_kwargs_in_collator": true, "fp16": false, "fsdp_version": 2, "generate_samples": false, "generation_do_sample": true, "generation_max_new_tokens": 50, "generation_prompt_ratio": 0.5, "generation_temperature": 0.7, "gradient_accumulation_steps": 8, "gradient_checkpointing": false, "include_tkps": true, "is_multimodal": true, "label_balance_window_optim_steps": 1, "layer_offloading": false, "learning_rate": 0.0001, "lisa_layers_attribute": "model.layers", "load_best_model_at_end": false, "load_in_4bit": false, "load_in_8bit": false, "local_rank": 0, "logging_nan_inf_filter": true, "logging_steps": 1, "lora_alpha": 64, "lora_dropout": 0.0, "lora_embedding_kernel": false, "lora_fp32_gradients": false, "lora_mlp_kernel": false, "lora_model_dir": "./outputs/glm53-flash-qlora-out/checkpoint-234", "lora_o_kernel": false, "lora_qkv_kernel": false, "lora_r": 32, "lora_target_modules": [ "q_a_proj", "q_b_proj", "kv_a_proj_with_mqa", "kv_b_proj", "q_proj", "k_proj", "v_proj", "o_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "mean_resizing_embeddings": false, "merge_compile": false, "merge_compile_precision": "strict", "merge_lora": true, "merge_method": "memory_efficient", "merge_num_threads": 2, "micro_batch_size": 1, "mixlora_jitter_noise": 0.0, "mixlora_num_experts": 8, "mixlora_router_aux_loss_coef": 0.01, "mixlora_router_init_range": 0.02, "mixlora_top_k": 2, "model_config_type": "glm5_next", "model_config_type_text": "glm5_next_text", "nf4_backend": "bitsandbytes", "num_epochs": 3.0, "num_generation_samples": 3, "optimizer": "adamw_torch_8bit", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "./outputs/glm53-flash-qlora-out", "ple_cpu_offload": false, "plugins": [ "axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin" ], "pretrain_multipack_attn": true, "processor_config": "zai-org/GLM-5.3-Flash-BF16", "profiler_steps_start": 0, "qgalore_cos_threshold": 0.4, "qgalore_gamma_proj": 2, "qgalore_proj_bits": 4, "qgalore_proj_group_size": 256, "qgalore_proj_quant": true, "qgalore_proj_type": "std", "qgalore_queue_size": 5, "qgalore_rank": 256, "qgalore_scale": 0.25, "qgalore_update_proj_gap": 200, "qlora_sharded_model_loading": false, "quantize_moe_experts": false, "ray_num_workers": 1, "relora_prune_method": "magnitude", "resources_per_worker": { "GPU": 1 }, "sample_packing": false, "sample_packing_bin_size": 200, "sample_packing_group_size": 100000, "save_only_model": false, "save_safetensors": true, "save_steps": 0.3333333333333333, "saves_per_epoch": 1, "sequence_len": 3072, "shuffle_before_merging_datasets": false, "shuffle_merged_datasets": true, "skip_prepare_dataset": false, "streaming_multipack_buffer_size": 10000, "strict": false, "sync_profiler_start_step": 2, "tensor_parallel_size": 1, "tf32": false, "tiled_mlp_use_original_mlp": true, "tokenizer_config": "zai-org/GLM-5.3-Flash-BF16", "tokenizer_save_jinja_files": true, "torch_compile_regional": false, "torch_dtype": "torch.bfloat16", "train_on_inputs": false, "trl": { "async_prefetch": false, "log_completions": false, "mask_truncated_completions": false, "ref_model_mixup_alpha": 0.9, "ref_model_sync_steps": 64, "replay_buffer_size": 0, "replay_recompute_logps": true, "reroll_max_groups": 1, "reroll_start_fraction": 1.0, "reward_num_workers": 1, "scale_rewards": true, "skip_zero_advantage_batches": true, "sync_ref_model": false, "use_data_producer": false, "use_vllm": false, "vllm_lora_sync": false, "vllm_server_host": "0.0.0.0", "vllm_server_port": 8000 }, "use_otel_metrics": false, "use_ray": false, "val_set_size": 0.1, "vllm": { "device": "auto", "dtype": "auto", "host": "0.0.0.0", "port": 8000 }, "warmup_ratio": 0.1, "weight_decay": 0.0, "world_size": 1 } [2026-11-05 11:42:40,442] [INFO] [axolotl.cli.merge_lora.do_merge_lora:36] [PID:985839] Using 2 CPU threads for LoRA merge [2026-11-05 11:42:40,443] [DEBUG] [axolotl.cli.merge_lora._do_merge_lora:68] [PID:985839] Using memory-efficient LoRA merging method... [2026-11-05 11:42:40,443] [DEBUG] [axolotl.cli.merge_lora._do_merge_lora_efficient:143] [PID:985839] Using memory-efficient LoRA merging method... Downloading bytes: | 0.00B Reconstructing (incomplete total...): | | 0.00B / 0.00B  Fetching 130 files: 0%| | 0/130 [00:00