[2025-11-25 02:20:44,846] [DEBUG] [axolotl.utils.config.resolve_dtype:66] [PID:3847] bf16 support detected, enabling for this configuration. [2025-11-25 02:20:44,921] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:3847] baseline 0.000GB () [2025-11-25 02:20:44,921] [INFO] [axolotl.cli.config.load_cfg:248] [PID:3847] config: { "activation_offloading": false, "adapter": "lora", "axolotl_config_path": "config.yaml", "base_model": "meta-llama/Llama-3.1-8B-Instruct", "base_model_config": "meta-llama/Llama-3.1-8B-Instruct", "batch_size": 8, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_90", "fp8": false, "n_gpu": 1, "n_node": 1 }, "context_parallel_size": 1, "dataloader_num_workers": 1, "dataloader_pin_memory": true, "dataloader_prefetch_factor": 256, "dataset_prepared_path": "last_run_prepared", "dataset_processes": 24, "datasets": [ { "data_files": "*formatted.jsonl", "message_property_mappings": { "content": "content", "role": "role" }, "path": "AIPixelMedia/astrid-dataset", "trust_remote_code": false, "type": "alpaca" } ], "ddp": false, "device": "cuda:0", "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "early_stopping_patience": 2, "env_capabilities": { "torch_version": "2.7.1" }, "eval_batch_size": 2, "eval_causal_lm_metrics": [ "sacrebleu", "comet", "ter", "chrf" ], "eval_max_new_tokens": 128, "eval_sample_packing": false, "eval_steps": 5, "eval_table_size": 0, "experimental_skip_move_to_device": true, "flash_attention": true, "fp16": false, "gradient_accumulation_steps": 4, "gradient_checkpointing": true, "gradient_checkpointing_kwargs": { "use_reentrant": false }, "group_by_length": true, "include_tkps": true, "is_llama_derived_model": true, "learning_rate": 2e-05, "lisa_layers_attribute": "model.layers", "load_best_model_at_end": false, "load_in_4bit": true, "load_in_8bit": false, "local_rank": 0, "logging_steps": 5, "lora_alpha": 32, "lora_dropout": 0.2, "lora_modules_to_save": [ "lm_head" ], "lora_r": 16, "lora_target_modules": [ "q_proj", "k_proj", "v_proj", "o_proj", "gate_proj", "up_proj", "down_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "mean_resizing_embeddings": false, "merge_lora": false, "micro_batch_size": 2, "model_config_type": "llama", "num_epochs": 20.0, "optimizer": "paged_adamw_32bit", "output_dir": "./outputs/astrid-llama-8b", "pad_to_sequence_len": true, "pretrain_multipack_attn": true, "profiler_steps_start": 0, "qlora_sharded_model_loading": false, "ray_num_workers": 1, "remove_unused_columns": false, "resources_per_worker": { "GPU": 1 }, "sample_packing": true, "sample_packing_bin_size": 200, "sample_packing_group_size": 100000, "save_only_model": false, "save_safetensors": true, "save_steps": 100, "seed": 35, "sequence_len": 2048, "shuffle_before_merging_datasets": false, "shuffle_merged_datasets": true, "skip_prepare_dataset": false, "special_tokens": { "pad_token": "<|end_of_text|>" }, "streaming_multipack_buffer_size": 10000, "strict": false, "tensor_parallel_size": 1, "tf32": false, "tiled_mlp_use_original_mlp": true, "tokenizer_config": "meta-llama/Llama-3.1-8B-Instruct", "tokenizer_save_jinja_files": true, "torch_dtype": "torch.bfloat16", "train_on_inputs": false, "trl": { "log_completions": false, "mask_truncated_completions": false, "ref_model_mixup_alpha": 0.9, "ref_model_sync_steps": 64, "scale_rewards": true, "sync_ref_model": false, "use_vllm": false, "vllm_server_host": "0.0.0.0", "vllm_server_port": 8000 }, "use_ray": false, "val_set_size": 0.1, "vllm": { "device": "auto", "dtype": "auto", "gpu_memory_utilization": 0.9, "host": "0.0.0.0", "port": 8000 }, "warmup_steps": 10, "weight_decay": 0.01, "world_size": 1 } [2025-11-25 02:20:45,359] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:278] [PID:3847] EOS: 128009 / <|eot_id|> [2025-11-25 02:20:45,359] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:3847] BOS: 128000 / <|begin_of_text|> [2025-11-25 02:20:45,360] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:3847] PAD: 128001 / <|end_of_text|> [2025-11-25 02:20:45,360] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:3847] UNK: None / None [2025-11-25 02:20:45,361] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:476] [PID:3847] Unable to find prepared dataset in last_run_prepared/d6f798814894ac4627709b4d3b576758 [2025-11-25 02:20:45,361] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:3847] Loading raw datasets... [2025-11-25 02:20:45,361] [WARNING] [axolotl.utils.data.sft._load_raw_datasets:322] [PID:3847] Processing datasets during training can lead to VRAM instability. Please pre-process your dataset using `axolotl preprocess path/to/config.yml`. README.md: 0%| | 0.00/29.0 [00:002048) (num_proc=24): 0%| | 0/216 [00:002048) (num_proc=24): 4%|██▉ | 9/216 [00:00<00:06, 33.18 examples/s] Dropping Long Sequences (>2048) (num_proc=24): 100%|██████████████████████████████████████████████████████████████████| 216/216 [00:00<00:00, 500.79 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=24): 0%| | 0/216 [00:00 [2025-11-25 02:20:54,363] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:3847] BOS: 128000 / <|begin_of_text|> [2025-11-25 02:20:54,363] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:3847] PAD: 128001 / <|end_of_text|> [2025-11-25 02:20:54,363] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:3847] UNK: None / None [2025-11-25 02:20:54,363] [DEBUG] [axolotl.train.setup_model_and_tokenizer:74] [PID:3847] Loading model [2025-11-25 02:20:54,401] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:3847] Patched Trainer.evaluation_loop with nanmean loss calculation [2025-11-25 02:20:54,404] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:3847] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation [2025-11-25 02:20:54,405] [INFO] [axolotl.loaders.patch_manager._apply_multipack_patches:301] [PID:3847] Applying multipack dataloader patch for sample packing... Loading checkpoint shards: 0%| | 0/4 [00:00