[2026-04-07 09:59:43,139] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:10658] baseline 0.000GB () [2026-04-07 09:59:43,141] [INFO] [axolotl.cli.config.load_cfg:341] [PID:10658] config: { "activation_offloading": false, "adapter": "lora", "axolotl_config_path": "config.yaml", "base_model": "mistralai/Ministral-3-8B-Instruct-2512-BF16", "base_model_config": "mistralai/Ministral-3-8B-Instruct-2512-BF16", "batch_size": 8, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_90", "fp8": true, "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": 26, "datasets": [ { "message_property_mappings": { "content": "content", "role": "role" }, "path": "AI-AgentSafa/dataset", "trust_remote_code": false, "type": "alpaca" } ], "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": "2.8.0" }, "eval_batch_size": 4, "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": 2, "gradient_checkpointing": false, "include_tkps": true, "is_multimodal": true, "layer_offloading": false, "learning_rate": 0.0002, "lisa_layers_attribute": "model.layers", "load_best_model_at_end": false, "load_in_4bit": false, "load_in_8bit": false, "local_rank": 0, "lora_alpha": 32, "lora_dropout": 0.05, "lora_embedding_kernel": true, "lora_mlp_kernel": true, "lora_o_kernel": true, "lora_qkv_kernel": true, "lora_r": 16, "lora_target_modules": [ "q_proj", "v_proj", "k_proj", "o_proj", "gate_proj", "down_proj", "up_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "mean_resizing_embeddings": false, "merge_method": "memory_efficient", "micro_batch_size": 4, "model_config_type": "mistral3", "model_config_type_text": "ministral3", "num_epochs": 5.0, "num_generation_samples": 3, "optimizer": "adamw_bnb_8bit", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "/workspace/outputs/ministral8b-tsql", "pretrain_multipack_attn": true, "processor_config": "mistralai/Ministral-3-8B-Instruct-2512-BF16", "profiler_steps_start": 0, "qlora_sharded_model_loading": false, "quantize_moe_experts": false, "ray_num_workers": 1, "resources_per_worker": { "GPU": 1 }, "sample_packing_bin_size": 200, "sample_packing_group_size": 100000, "save_only_model": false, "save_safetensors": true, "sequence_len": 2048, "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": "mistralai/Ministral-3-8B-Instruct-2512-BF16", "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 }, "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 }, "weight_decay": 0.0, "world_size": 1 } [2026-04-07 09:59:43,337] [DEBUG] [axolotl.loaders.utils.check_model_config:88] [PID:10658] Loaded image size: 1540 from model config tokenizer_config.json: 0.00B [00:00, ?B/s] tokenizer_config.json: 198kB [00:00, 23.2MB/s] tokenizer.json: 0%| | 0.00/17.1M [00:00 [2026-04-07 09:59:46,945] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:308] [PID:10658] BOS: 1 / [2026-04-07 09:59:46,945] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:309] [PID:10658] PAD: 11 / [2026-04-07 09:59:46,945] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:310] [PID:10658] UNK: 0 / [2026-04-07 09:59:46,947] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:480] [PID:10658] Unable to find prepared dataset in last_run_prepared/1ad236e091b2b88f65fa3ad720cbbedf [2026-04-07 09:59:46,948] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:10658] Loading raw datasets... [2026-04-07 09:59:46,949] [WARNING] [axolotl.utils.data.sft._load_raw_datasets:322] [PID:10658] Processing datasets during training can lead to VRAM instability. Please pre-process your dataset using `axolotl preprocess path/to/config.yml`. Downloading (incomplete total...): 0.00B [00:00, ?B/s] Fetching 0 files: 0it [00:00, ?it/s] Fetching 0 files: 0it [00:00, ?it/s] Download complete: : 0.00B [00:00, ?B/s] Download complete: : 0.00B [00:00, ?B/s] [2026-04-07 09:59:48,278] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:87] [PID:10658] Loading dataset: AI-AgentSafa/dataset with base_type: alpaca and prompt_style: None Tokenizing Prompts (num_proc=26): 0%| | 0/229 [00:002048) (num_proc=26): 0%| | 0/229 [00:002048) (num_proc=26): 4%|█▎ | 9/229 [00:00<00:10, 20.72 examples/s] Dropping Invalid Sequences (2048) (num_proc=26): 81%|████████████████████████▎ | 186/229 [00:00<00:00, 448.13 examples/s] Dropping Invalid Sequences (2048) (num_proc=26): 100%|██████████████████████████████| 229/229 [00:00<00:00, 315.58 examples/s] Saving the dataset (0/1 shards): 0%| | 0/229 [00:00 [2026-04-07 09:59:58,026] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:308] [PID:10658] BOS: 1 / [2026-04-07 09:59:58,026] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:309] [PID:10658] PAD: 11 / [2026-04-07 09:59:58,026] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:310] [PID:10658] UNK: 0 / processor_config.json: 0%| | 0.00/976 [00:00