[2025-12-29 02:49:24,896] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:3751] baseline 0.000GB () [2025-12-29 02:49:24,896] [INFO] [axolotl.cli.config.load_cfg:256] [PID:3751] config: { "activation_offloading": false, "adapter": "lora", "axolotl_config_path": "tuner.yaml", "base_model": "codellama/CodeLlama-7b-hf", "base_model_config": "codellama/CodeLlama-7b-hf", "batch_size": 5, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_90", "fp8": false, "n_gpu": 1, "n_node": 1 }, "chat_template": "llama3", "context_parallel_size": 1, "dataloader_num_workers": 2, "dataloader_pin_memory": true, "dataset_num_proc": 96, "datasets": [ { "chat_template": "tokenizer_default", "conversation": "llama3", "field_messages": "messages", "message_property_mappings": { "content": "content", "role": "role" }, "path": "darwinkernelpanic/luau-reasoning-normalized", "trust_remote_code": false, "type": "chat_template" } ], "ddp": false, "device": "cuda:0", "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "env_capabilities": { "torch_version": "2.8.0" }, "eval_batch_size": 5, "eval_causal_lm_metrics": [ "sacrebleu", "comet", "ter", "chrf" ], "eval_max_new_tokens": 128, "eval_sample_packing": true, "eval_steps": 1000, "eval_table_size": 0, "experimental_skip_move_to_device": true, "fp16": false, "gradient_accumulation_steps": 1, "gradient_checkpointing": true, "gradient_checkpointing_kwargs": { "use_reentrant": true }, "group_by_length": true, "include_tkps": true, "is_falcon_derived_model": false, "is_llama_derived_model": true, "is_mistral_derived_model": 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, "logging_steps": 25, "lora_alpha": 32, "lora_dropout": 0.05, "lora_r": 16, "lora_target_modules": [ "q_proj", "k_proj", "v_proj", "o_proj" ], "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "mean_resizing_embeddings": false, "micro_batch_size": 5, "model_config_type": "llama", "num_epochs": 3.0, "optimizer": "adamw_torch", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "./outputs/luau-codellama-h200-fast", "pad_to_sequence_len": true, "pretrain_multipack_attn": true, "profiler_steps_start": 0, "qlora_sharded_model_loading": false, "ray_num_workers": 1, "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": 1000, "save_strategy": "steps", "save_total_limit": 3, "seed": 42, "sequence_len": 4096, "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": "codellama/CodeLlama-7b-hf", "tokenizer_save_jinja_files": true, "tokenizer_type": "LlamaTokenizer", "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 }, "type_of_model": "LlamaForCausalLM", "use_otel_metrics": false, "use_ray": false, "val_set_size": 0.05, "vllm": { "device": "auto", "dtype": "auto", "gpu_memory_utilization": 0.9, "host": "0.0.0.0", "port": 8000 }, "warmup_steps": 100, "weight_decay": 0.0, "world_size": 1 } [2025-12-29 02:49:25,389] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:3751] EOS: 2 / [2025-12-29 02:49:25,389] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:3751] BOS: 1 / [2025-12-29 02:49:25,389] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:282] [PID:3751] PAD: 2 / [2025-12-29 02:49:25,389] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:283] [PID:3751] UNK: 0 / [2025-12-29 02:49:25,390] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:481] [PID:3751] Unable to find prepared dataset in last_run_prepared/b7c17715ff7f64badeb455c51ab5d648 [2025-12-29 02:49:25,390] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:3751] Loading raw datasets... [2025-12-29 02:49:25,390] [WARNING] [axolotl.utils.data.sft._load_raw_datasets:322] [PID:3751] Processing datasets during training can lead to VRAM instability. Please pre-process your dataset using `axolotl preprocess path/to/config.yml`. [2025-12-29 02:49:26,885] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:87] [PID:3751] Loading dataset: darwinkernelpanic/luau-reasoning-normalized with base_type: chat_template and prompt_style: None [2025-12-29 02:49:26,887] [INFO] [axolotl.prompt_strategies.chat_template.__call__:996] [PID:3751] Using chat template: --- {% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|> '+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|> ' }}{% endif %} --- [2025-12-29 02:49:26,891] [WARNING] [axolotl.prompt_strategies.chat_template._validate_eot_and_eos_tokens:337] [PID:3751] EOS token '' not found in chat_template. Please check if your template/EOS token is correct. [2025-12-29 02:49:27,110] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:218] [PID:3751] min_input_len: 636 [2025-12-29 02:49:27,110] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:220] [PID:3751] max_input_len: 12839 [2025-12-29 02:49:27,405] [WARNING] [axolotl.utils.data.utils.handle_long_seq_in_dataset:260] [PID:3751] Dropped 755 samples from dataset Saving the dataset (0/56 shards): 0%| | 0/14586 [00:00 [2025-12-29 02:49:32,522] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:3751] BOS: 1 / [2025-12-29 02:49:32,522] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:282] [PID:3751] PAD: 2 / [2025-12-29 02:49:32,522] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:283] [PID:3751] UNK: 0 / [2025-12-29 02:49:32,522] [DEBUG] [axolotl.train.setup_model_and_tokenizer:82] [PID:3751] Loading model [2025-12-29 02:49:32,640] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:3751] Patched Trainer.evaluation_loop with nanmean loss calculation [2025-12-29 02:49:32,642] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:3751] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation [2025-12-29 02:49:32,642] [INFO] [axolotl.loaders.patch_manager._apply_multipack_patches:301] [PID:3751] Applying multipack dataloader patch for sample packing... [2025-12-29 02:49:32,643] [INFO] [axolotl.loaders.patch_manager._patch_llama_sample_packing:430] [PID:3751] Patching llama _prepare_4d_causal_attention_mask*... Loading checkpoint shards: 0%| | 0/2 [00:00