[2025-11-17 19:16:13,073] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:26868] baseline 0.000GB () [2025-11-17 19:16:13,073] [INFO] [axolotl.cli.config.load_cfg:248] [PID:26868] config: { "activation_offloading": false, "adapter": "lora", "axolotl_config_path": "seedcoder.yaml", "base_model": "ByteDance-Seed/Seed-Coder-8B-Instruct", "base_model_config": "ByteDance-Seed/Seed-Coder-8B-Instruct", "batch_size": 64, "bf16": true, "capabilities": { "bf16": true, "compute_capability": "sm_90", "fp8": false, "n_gpu": 4, "n_node": 1 }, "context_parallel_size": 1, "dataloader_num_workers": 4, "dataloader_pin_memory": true, "dataloader_prefetch_factor": 256, "dataset_num_proc": 104, "dataset_prepared_path": "last_run_prepared", "datasets": [ { "chat_template": "tokenizer_default", "field_messages": "messages", "message_property_mappings": { "content": "content", "role": "role" }, "path": "data_clean.jsonl", "roles": { "assistant": [ "assistant" ], "system": [ "system" ], "user": [ "user" ] }, "trust_remote_code": false, "type": "chat_template" } ], "ddp": true, "deepspeed": { "bf16": { "enabled": "auto" }, "fp16": { "auto_cast": false, "enabled": "auto", "hysteresis": 2, "initial_scale_power": 32, "loss_scale": 0, "loss_scale_window": 1000, "min_loss_scale": 1 }, "gradient_accumulation_steps": "auto", "gradient_clipping": "auto", "train_batch_size": "auto", "train_micro_batch_size_per_gpu": "auto", "wall_clock_breakdown": false, "zero_optimization": { "contiguous_gradients": true, "offload_optimizer": { "device": "cpu" }, "overlap_comm": true, "stage": 2 } }, "device": "cuda:0", "device_map": { "": 0 }, "dion_rank_fraction": 1.0, "dion_rank_multiple_of": 1, "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, "flash_attention": true, "fp16": false, "gradient_accumulation_steps": 4, "gradient_checkpointing": true, "gradient_checkpointing_kwargs": { "use_reentrant": false }, "group_by_length": false, "include_tkps": true, "is_falcon_derived_model": false, "is_llama_derived_model": true, "is_mistral_derived_model": false, "learning_rate": 0.0001, "liger_fused_linear_cross_entropy": true, "liger_glu_activation": true, "liger_layer_norm": true, "liger_rms_norm": true, "liger_rope": true, "lisa_layers_attribute": "model.layers", "load_best_model_at_end": false, "load_in_4bit": false, "load_in_8bit": false, "local_rank": 0, "logging_steps": 1, "lora_alpha": 64, "lora_dropout": 0.05, "lora_r": 64, "lora_target_linear": true, "loraplus_lr_embedding": 1e-06, "lr_scheduler": "cosine", "max_grad_norm": 1.0, "mean_resizing_embeddings": false, "micro_batch_size": 4, "model_config_type": "llama", "num_epochs": 1.0, "optimizer": "adamw_torch", "otel_metrics_host": "localhost", "otel_metrics_port": 8000, "output_dir": "./lora-sout-SC-highseq-len", "pad_to_sequence_len": true, "plugins": [ "axolotl.integrations.liger.LigerPlugin" ], "pretrain_multipack_attn": true, "profiler_steps_start": 0, "qlora_sharded_model_loading": false, "ray_num_workers": 1, "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": 60, "save_total_limit": 100, "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": false, "tiled_mlp_use_original_mlp": true, "tokenizer_config": "ByteDance-Seed/Seed-Coder-8B-Instruct", "tokenizer_save_jinja_files": true, "tokenizer_type": "AutoTokenizer", "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 }, "trust_remote_code": true, "type_of_model": "AutoModelForCausalLM", "use_otel_metrics": false, "use_ray": false, "use_wandb": true, "val_set_size": 0.0, "vllm": { "device": "auto", "dtype": "auto", "gpu_memory_utilization": 0.9, "host": "0.0.0.0", "port": 8000 }, "wandb_entity": "test-aa", "wandb_name": "No-mods-seedcoder-low-gas-high-seq-len", "wandb_project": "seedcoder", "warmup_ratio": 0.05, "weight_decay": 0.0, "world_size": 4 } [2025-11-17 19:16:13,861] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:278] [PID:26868] EOS: 2 / <[end▁of▁sentence]> [2025-11-17 19:16:13,861] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:26868] BOS: 0 / <[begin▁of▁sentence]> [2025-11-17 19:16:13,862] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:26868] PAD: 1 / <[PAD▁TOKEN]> [2025-11-17 19:16:13,862] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:26868] UNK: None / None [2025-11-17 19:16:13,862] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:475] [PID:26868] Loading prepared dataset from disk at last_run_prepared/6c95b17c4af71171cd8b6cb73a85e0fb... Loading dataset from disk: 0%| | 0/104 [00:00 [2025-11-17 19:16:15,595] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:26868] BOS: 0 / <[begin▁of▁sentence]> [2025-11-17 19:16:15,595] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:26868] PAD: 1 / <[PAD▁TOKEN]> [2025-11-17 19:16:15,595] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:26868] UNK: None / None [2025-11-17 19:16:15,595] [DEBUG] [axolotl.train.setup_model_and_tokenizer:74] [PID:26868] Loading model [2025-11-17 19:16:15,659] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:26868] Patched Trainer.evaluation_loop with nanmean loss calculation [2025-11-17 19:16:15,660] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:26868] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation [2025-11-17 19:16:15,680] [INFO] [axolotl.integrations.liger.plugin.pre_model_load:71] [PID:26868] Applying LIGER to llama with kwargs: {'rope': True, 'cross_entropy': None, 'fused_linear_cross_entropy': True, 'rms_norm': True, 'swiglu': True} Loading checkpoint shards: 0%| | 0/4 [00:00