[2026-07-18 05:06:59,677] [WARNING] [axolotl.utils.config.normalize_config:282] [PID:51437] Gemma4 requires use_reentrant=False for gradient checkpointing in distributed training. Setting use_reentrant=False. [2026-07-18 05:06:59,678] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:51437] baseline 0.000GB () [2026-07-18 05:06:59,679] [INFO] [axolotl.cli.config.load_cfg:333] [PID:51437] config: { "activation_offloading": false, "adapter": "lora", "attn_implementation": "sdpa", "attn_needs_dtype_cast": false, "attn_supports_packing": false, "attn_uses_flash_lib": false, "axolotl_config_path": "train.yaml", "base_model": "google/gemma-4-26b-a4b-it", "base_model_config": "google/gemma-4-26b-a4b-it", "batch_size": 16, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_80", "fp8": false, "n_gpu": 1, "n_node": 1, "tf32": true }, "context_parallel_size": 1, "dataloader_num_workers": 1, "dataloader_pin_memory": true, "dataloader_prefetch_factor": 256, "dataset_num_proc": 16, "datasets": [ { "chat_template": "tokenizer_default", "message_property_mappings": { "content": "content", "role": "role" }, "path": "mixed_coding_traces.jsonl", "trust_remote_code": false, "type": "chat_template" } ], "ddp": false, "device": "cuda:0", "device_map": "auto", "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "eaft_alpha": 1.0, "eaft_k": 20, "env_capabilities": { "torch_version": "2.10.0" }, "eval_batch_size": 1, "eval_causal_lm_metrics": [ "sacrebleu", "comet", "ter", "chrf" ], "eval_max_new_tokens": 128, "eval_table_size": 0, "experimental_skip_move_to_device": true, "fp16": false, "generate_samples": false, "generation_do_sample": true, "generation_max_new_tokens": 50, "generation_prompt_ratio": 0.5, "generation_temperature": 0.7, "gradient_accumulation_steps": 16, "gradient_checkpointing": true, "gradient_checkpointing_kwargs": { "use_reentrant": false }, "include_tkps": true, "is_falcon_derived_model": false, "is_llama_derived_model": false, "is_mistral_derived_model": false, "is_multimodal": true, "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_steps": 5, "lora_alpha": 32, "lora_dropout": 0.0, "lora_model_dir": "./outputs/gemma4-26b-mixed/checkpoint-250", "lora_r": 16, "lora_target_modules": "model.language_model.layers.[\\d]+.(_checkpoint_wrapped_module.)?(mlp|self_attn).(up|down|gate|q|k|v|o)_proj", "lora_target_parameters": [ "experts.gate_up_proj", "experts.down_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "max_steps": 250, "mean_resizing_embeddings": false, "merge_method": "memory_efficient", "micro_batch_size": 1, "model_config_type": "gemma4", "model_config_type_text": "gemma4_text", "num_epochs": 1.0, "num_generation_samples": 3, "optimizer": "adamw_torch_fused", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "./outputs/gemma4-26b-mixed", "pretrain_multipack_attn": true, "processor_config": "google/gemma-4-26b-a4b-it", "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_bin_size": 200, "sample_packing_group_size": 100000, "save_only_model": false, "save_safetensors": true, "save_steps": 50, "sequence_len": 8192, "shuffle_before_merging_datasets": false, "shuffle_merged_datasets": true, "skip_prepare_dataset": false, "streaming_multipack_buffer_size": 10000, "strict": false, "tensor_parallel_size": 1, "tf32": true, "tiled_mlp_use_original_mlp": true, "tokenizer_config": "google/gemma-4-26b-a4b-it", "tokenizer_save_jinja_files": true, "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 }, "trust_remote_code": true, "type_of_model": "gemma4_moe", "use_otel_metrics": false, "use_ray": false, "val_set_size": 0.0, "vllm": { "device": "auto", "dtype": "auto", "gpu_memory_utilization": 0.9, "host": "0.0.0.0", "port": 8000 }, "warmup_steps": 20, "weight_decay": 0.0, "world_size": 1 } [2026-07-18 05:06:59,682] [WARNING] [axolotl.utils.schemas.model.hint_trust_remote_code:117] [PID:51437] `trust_remote_code` is set to true. Please make sure that you reviewed the remote code/model. [2026-07-18 05:07:00,053] [WARNING] [axolotl.utils.config.normalize_config:282] [PID:51437] Gemma4 requires use_reentrant=False for gradient checkpointing in distributed training. Setting use_reentrant=False. [2026-07-18 05:07:00,055] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:51437] baseline 0.000GB () [2026-07-18 05:07:00,056] [INFO] [axolotl.cli.config.load_cfg:333] [PID:51437] config: { "activation_offloading": false, "adapter": "lora", "attn_needs_dtype_cast": false, "attn_supports_packing": false, "attn_uses_flash_lib": false, "axolotl_config_path": "train.yaml", "base_model": "google/gemma-4-26b-a4b-it", "base_model_config": "google/gemma-4-26b-a4b-it", "batch_size": 16, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_80", "fp8": false, "n_gpu": 1, "n_node": 1, "tf32": true }, "context_parallel_size": 1, "dataloader_num_workers": 1, "dataloader_pin_memory": true, "dataloader_prefetch_factor": 256, "dataset_num_proc": 16, "datasets": [ { "chat_template": "tokenizer_default", "message_property_mappings": { "content": "content", "role": "role" }, "path": "mixed_coding_traces.jsonl", "trust_remote_code": false, "type": "chat_template" } ], "ddp": false, "device": "cuda:0", "device_map": "auto", "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "eaft_alpha": 1.0, "eaft_k": 20, "env_capabilities": { "torch_version": "2.10.0" }, "eval_batch_size": 1, "eval_causal_lm_metrics": [ "sacrebleu", "comet", "ter", "chrf" ], "eval_max_new_tokens": 128, "eval_table_size": 0, "experimental_skip_move_to_device": true, "fp16": false, "generate_samples": false, "generation_do_sample": true, "generation_max_new_tokens": 50, "generation_prompt_ratio": 0.5, "generation_temperature": 0.7, "gradient_accumulation_steps": 16, "gradient_checkpointing": true, "gradient_checkpointing_kwargs": { "use_reentrant": false }, "include_tkps": true, "is_falcon_derived_model": false, "is_llama_derived_model": false, "is_mistral_derived_model": false, "is_multimodal": true, "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_steps": 5, "lora_alpha": 32, "lora_dropout": 0.0, "lora_model_dir": "./outputs/gemma4-26b-mixed/checkpoint-250", "lora_r": 16, "lora_target_modules": "model.language_model.layers.[\\d]+.(_checkpoint_wrapped_module.)?(mlp|self_attn).(up|down|gate|q|k|v|o)_proj", "lora_target_parameters": [ "experts.gate_up_proj", "experts.down_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "max_steps": 250, "mean_resizing_embeddings": false, "merge_lora": true, "merge_method": "memory_efficient", "micro_batch_size": 1, "model_config_type": "gemma4", "model_config_type_text": "gemma4_text", "num_epochs": 1.0, "num_generation_samples": 3, "optimizer": "adamw_torch_fused", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "./outputs/gemma4-26b-mixed", "pretrain_multipack_attn": true, "processor_config": "google/gemma-4-26b-a4b-it", "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_bin_size": 200, "sample_packing_group_size": 100000, "save_only_model": false, "save_safetensors": true, "save_steps": 50, "sequence_len": 8192, "shuffle_before_merging_datasets": false, "shuffle_merged_datasets": true, "skip_prepare_dataset": false, "streaming_multipack_buffer_size": 10000, "strict": false, "tensor_parallel_size": 1, "tf32": true, "tiled_mlp_use_original_mlp": true, "tokenizer_config": "google/gemma-4-26b-a4b-it", "tokenizer_save_jinja_files": true, "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 }, "trust_remote_code": true, "type_of_model": "gemma4_moe", "use_otel_metrics": false, "use_ray": false, "val_set_size": 0.0, "vllm": { "device": "auto", "dtype": "auto", "gpu_memory_utilization": 0.9, "host": "0.0.0.0", "port": 8000 }, "warmup_steps": 20, "weight_decay": 0.0, "world_size": 1 } Downloading (incomplete total...): 0.00B [00:00, ?B/s] Fetching 12 files: 0%| | 0/12 [00:00