| [2025-10-22 16:34:24,607] [DEBUG] [axolotl.utils.config.resolve_dtype:66] [PID:2199] bf16 support detected, enabling for this configuration. | |
| [2025-10-22 16:34:24,749] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:2199] baseline 0.000GB () | |
| [2025-10-22 16:34:24,749] [INFO] [axolotl.cli.config.load_cfg:248] [PID:2199] config: | |
| { | |
| "activation_offloading": false, | |
| "adapter": "lora", | |
| "axolotl_config_path": "config.yaml", | |
| "base_model": "Qwen/Qwen2.5-7B-Instruct", | |
| "base_model_config": "Qwen/Qwen2.5-7B-Instruct", | |
| "batch_size": 16, | |
| "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_processes": 32, | |
| "datasets": [ | |
| { | |
| "message_property_mappings": { | |
| "content": "content", | |
| "role": "role" | |
| }, | |
| "path": "/workspace/fine-tuning/data/trump.json", | |
| "trust_remote_code": false, | |
| "type": "alpaca" | |
| } | |
| ], | |
| "ddp": false, | |
| "device": "cuda:0", | |
| "dion_rank_fraction": 1.0, | |
| "dion_rank_multiple_of": 1, | |
| "env_capabilities": { | |
| "torch_version": "2.7.1" | |
| }, | |
| "eval_batch_size": 16, | |
| "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, | |
| "gradient_accumulation_steps": 1, | |
| "gradient_checkpointing": false, | |
| "include_tkps": true, | |
| "learning_rate": 0.0002, | |
| "lisa_layers_attribute": "model.layers", | |
| "load_best_model_at_end": false, | |
| "load_in_4bit": false, | |
| "load_in_8bit": true, | |
| "local_rank": 0, | |
| "lora_alpha": 16, | |
| "lora_dropout": 0.05, | |
| "lora_r": 8, | |
| "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, | |
| "micro_batch_size": 16, | |
| "model_config_type": "qwen2", | |
| "num_epochs": 1.0, | |
| "optimizer": "adamw_bnb_8bit", | |
| "output_dir": "./outputs/thoth_text_v2", | |
| "pretrain_multipack_attn": true, | |
| "profiler_steps_start": 0, | |
| "qlora_sharded_model_loading": 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": 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, | |
| "tiled_mlp_use_original_mlp": true, | |
| "tokenizer_config": "Qwen/Qwen2.5-7B-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": 0.0, | |
| "vllm": { | |
| "device": "auto", | |
| "dtype": "auto", | |
| "gpu_memory_utilization": 0.9, | |
| "host": "0.0.0.0", | |
| "port": 8000 | |
| }, | |
| "weight_decay": 0.0, | |
| "world_size": 1 | |
| } | |
| [2025-10-22 16:34:25,444] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:278] [PID:2199] EOS: 151645 / <|im_end|> | |
| [2025-10-22 16:34:25,444] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:2199] BOS: None / None | |
| [2025-10-22 16:34:25,444] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:2199] PAD: 151643 / <|endoftext|> | |
| [2025-10-22 16:34:25,444] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:2199] UNK: None / None | |
| [2025-10-22 16:34:25,445] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:476] [PID:2199] Unable to find prepared dataset in last_run_prepared/8eaeba3b90268710df560cd4dfe04e2d | |
| [2025-10-22 16:34:25,445] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:2199] Loading raw datasets... | |
| [2025-10-22 16:34:25,445] [WARNING] [axolotl.utils.data.sft._load_raw_datasets:322] [PID:2199] Processing datasets during training can lead to VRAM instability. Please pre-process your dataset using `axolotl preprocess path/to/config.yml`. | |
| [2025-10-22 16:34:25,754] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:87] [PID:2199] Loading dataset: /workspace/fine-tuning/data/trump.json with base_type: alpaca and prompt_style: None | |
| [2025-10-22 16:34:25,754] [WARNING] [datasets.arrow_dataset.map:3100] [PID:2199] num_proc must be <= 15. Reducing num_proc to 15 for dataset of size 15. | |
| Tokenizing Prompts (num_proc=15): 0%| | 0/15 [00:00<?, ? examples/s] Tokenizing Prompts (num_proc=15): 7%|█████ | 1/15 [00:00<00:08, 1.56 examples/s] Tokenizing Prompts (num_proc=15): 13%|██████████▏ | 2/15 [00:00<00:04, 3.01 examples/s] Tokenizing Prompts (num_proc=15): 27%|████████████████████▎ | 4/15 [00:00<00:01, 5.98 examples/s] Tokenizing Prompts (num_proc=15): 47%|███████████████████████████████████▍ | 7/15 [00:01<00:00, 9.94 examples/s] Tokenizing Prompts (num_proc=15): 60%|█████████████████████████████████████████████▌ | 9/15 [00:01<00:00, 11.64 examples/s] Tokenizing Prompts (num_proc=15): 73%|██████████████████████████████████████████████████████▉ | 11/15 [00:01<00:00, 12.43 examples/s] Tokenizing Prompts (num_proc=15): 87%|█████████████████████████████████████████████████████████████████ | 13/15 [00:01<00:00, 12.76 examples/s] Tokenizing Prompts (num_proc=15): 100%|███████████████████████████████████████████████████████████████████████████| 15/15 [00:01<00:00, 14.05 examples/s] Tokenizing Prompts (num_proc=15): 100%|███████████████████████████████████████████████████████████████████████████| 15/15 [00:01<00:00, 8.67 examples/s] | |
| [2025-10-22 16:34:27,600] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:218] [PID:2199] min_input_len: 46 | |
| [2025-10-22 16:34:27,600] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:220] [PID:2199] max_input_len: 87 | |
| [2025-10-22 16:34:27,601] [WARNING] [datasets.arrow_dataset.map:3100] [PID:2199] num_proc must be <= 15. Reducing num_proc to 15 for dataset of size 15. | |
| Dropping Long Sequences (>4096) (num_proc=15): 0%| | 0/15 [00:00<?, ? examples/s] Dropping Long Sequences (>4096) (num_proc=15): 7%|████▏ | 1/15 [00:00<00:03, 4.29 examples/s] Dropping Long Sequences (>4096) (num_proc=15): 100%|██████████████████████████████████████████████████████████████| 15/15 [00:00<00:00, 34.65 examples/s] | |
| Saving the dataset (0/1 shards): 0%| | 0/15 [00:00<?, ? examples/s] Saving the dataset (1/1 shards): 100%|██████████████████████████████████████████████████████████████████████████| 15/15 [00:00<00:00, 1508.60 examples/s] Saving the dataset (1/1 shards): 100%|██████████████████████████████████████████████████████████████████████████| 15/15 [00:00<00:00, 1468.01 examples/s] | |
| [2025-10-22 16:34:28,163] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:404] [PID:2199] total_num_tokens: 1_016 | |
| [2025-10-22 16:34:28,164] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:422] [PID:2199] `total_supervised_tokens: 386` | |
| [2025-10-22 16:34:28,164] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:520] [PID:2199] total_num_steps: 1 | |
| [2025-10-22 16:34:28,164] [INFO] [axolotl.utils.data.sft._prepare_standard_dataset:121] [PID:2199] Maximum number of steps set at 1 | |
| [2025-10-22 16:34:28,208] [DEBUG] [axolotl.train.setup_model_and_tokenizer:65] [PID:2199] Loading tokenizer... Qwen/Qwen2.5-7B-Instruct | |
| [2025-10-22 16:34:28,772] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:278] [PID:2199] EOS: 151645 / <|im_end|> | |
| [2025-10-22 16:34:28,772] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:2199] BOS: None / None | |
| [2025-10-22 16:34:28,772] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:2199] PAD: 151643 / <|endoftext|> | |
| [2025-10-22 16:34:28,772] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:2199] UNK: None / None | |
| [2025-10-22 16:34:28,772] [DEBUG] [axolotl.train.setup_model_and_tokenizer:74] [PID:2199] Loading model | |
| [2025-10-22 16:34:28,893] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:2199] Patched Trainer.evaluation_loop with nanmean loss calculation | |
| [2025-10-22 16:34:28,895] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:2199] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation | |
| model.safetensors.index.json: 0.00B [00:00, ?B/s] model.safetensors.index.json: 27.8kB [00:00, 71.6MB/s] | |
| model-00001-of-00004.safetensors: 0%| | 0.00/3.95G [00:00<?, ?B/s] model-00001-of-00004.safetensors: 0%| | 52.0k/3.95G [00:01<35:38:11, 30.8kB/s] model-00001-of-00004.safetensors: 1%|█ | 52.6M/3.95G [00:01<01:34, 41.0MB/s] model-00001-of-00004.safetensors: 2%|█▋ | 83.8M/3.95G [00:02<01:22, 46.6MB/s] model-00001-of-00004.safetensors: 6%|████▎ | 218M/3.95G [00:02<00:23, 157MB/s] model-00001-of-00004.safetensors: 11%|████████▉ | 446M/3.95G [00:02<00:09, 389MB/s] model-00001-of-00004.safetensors: 14%|███████████▍ | 571M/3.95G [00:02<00:07, 458MB/s] model-00001-of-00004.safetensors: 18%|██████████████▌ | 728M/3.95G [00:02<00:06, 533MB/s] model-00001-of-00004.safetensors: 21%|████████████████▋ | 835M/3.95G [00:03<00:06, 500MB/s] model-00001-of-00004.safetensors: 23%|██████████████████▎ | 917M/3.95G [00:03<00:07, 413MB/s] model-00001-of-00004.safetensors: 25%|███████████████████▉ | 998M/3.95G [00:03<00:06, 445MB/s] model-00001-of-00004.safetensors: 29%|██████████████████████▎ | 1.13G/3.95G [00:03<00:05, 496MB/s] model-00001-of-00004.safetensors: 32%|█████████████████████████▎ | 1.28G/3.95G [00:04<00:04, 647MB/s] model-00001-of-00004.safetensors: 36%|███████████████████████████▉ | 1.41G/3.95G [00:04<00:03, 675MB/s] model-00001-of-00004.safetensors: 38%|█████████████████████████████▉ | 1.51G/3.95G [00:04<00:03, 713MB/s] model-00001-of-00004.safetensors: 42%|█████████████████████████████████ | 1.67G/3.95G [00:04<00:02, 892MB/s] model-00001-of-00004.safetensors: 49%|█████████████████████████████████████▍ | 1.92G/3.95G [00:04<00:01, 1.19GB/s] model-00001-of-00004.safetensors: 54%|█████████████████████████████████████████▏ | 2.11G/3.95G [00:04<00:01, 1.07GB/s] model-00001-of-00004.safetensors: 60%|█████████████████████████████████████████████▊ | 2.35G/3.95G [00:04<00:01, 1.21GB/s] model-00001-of-00004.safetensors: 63%|████████████████████████████████████████████████▍ | 2.48G/3.95G [00:05<00:01, 1.18GB/s] model-00001-of-00004.safetensors: 67%|███████████████████████████████████████████████████▊ | 2.65G/3.95G [00:05<00:01, 1.14GB/s] model-00001-of-00004.safetensors: 78%|███████████████████████████████████████████████████████████▋ | 3.06G/3.95G [00:05<00:00, 1.75GB/s] model-00001-of-00004.safetensors: 85%|█████████████████████████████████████████████████████████████████▍ | 3.36G/3.95G [00:05<00:00, 1.88GB/s] model-00001-of-00004.safetensors: 91%|██████████████████████████████████████████████████████████████████████▍ | 3.61G/3.95G [00:05<00:00, 1.99GB/s] model-00001-of-00004.safetensors: 100%|█████████████████████████████████████████████████████████████████████████████| 3.95G/3.95G [00:05<00:00, 2.22GB/s] model-00001-of-00004.safetensors: 100%|██████████████████████████████████████████████████████████████████████████████| 3.95G/3.95G [00:05<00:00, 697MB/s] | |
| model-00002-of-00004.safetensors: 0%| | 0.00/3.86G [00:00<?, ?B/s] model-00002-of-00004.safetensors: 0%| | 3.45M/3.86G [00:01<32:49, 1.96MB/s] model-00002-of-00004.safetensors: 3%|██▏ | 108M/3.86G [00:02<01:04, 58.4MB/s] model-00002-of-00004.safetensors: 9%|██████▉ | 338M/3.86G [00:02<00:18, 190MB/s] model-00002-of-00004.safetensors: 10%|███████▊ | 384M/3.86G [00:02<00:16, 206MB/s] model-00002-of-00004.safetensors: 15%|███████████▊ | 576M/3.86G [00:02<00:08, 377MB/s] model-00002-of-00004.safetensors: 18%|██████████████▍ | 709M/3.86G [00:03<00:06, 496MB/s] model-00002-of-00004.safetensors: 28%|█████████████████████▋ | 1.08G/3.86G [00:03<00:02, 945MB/s] model-00002-of-00004.safetensors: 33%|█████████████████████████▍ | 1.26G/3.86G [00:03<00:02, 940MB/s] model-00002-of-00004.safetensors: 37%|█████████████████████████████▏ | 1.45G/3.86G [00:03<00:02, 983MB/s] model-00002-of-00004.safetensors: 42%|████████████████████████████████▋ | 1.64G/3.86G [00:03<00:02, 1.01GB/s] model-00002-of-00004.safetensors: 48%|████████████████████████████████████▋ | 1.84G/3.86G [00:03<00:01, 1.11GB/s] model-00002-of-00004.safetensors: 54%|█████████████████████████████████████████▎ | 2.07G/3.86G [00:03<00:01, 1.24GB/s] model-00002-of-00004.safetensors: 60%|█████████████████████████████████████████████▊ | 2.30G/3.86G [00:04<00:01, 1.42GB/s] model-00002-of-00004.safetensors: 73%|███████████████████████████████████████████████████████▉ | 2.81G/3.86G [00:04<00:00, 2.16GB/s] model-00002-of-00004.safetensors: 80%|█████████████████████████████████████████████████████████████▏ | 3.07G/3.86G [00:04<00:00, 1.90GB/s] model-00002-of-00004.safetensors: 86%|██████████████████████████████████████████████████████████████████▌ | 3.34G/3.86G [00:04<00:00, 1.68GB/s] model-00002-of-00004.safetensors: 97%|██████████████████████████████████████████████████████████████████████████▋ | 3.75G/3.86G [00:04<00:00, 2.08GB/s] model-00002-of-00004.safetensors: 100%|██████████████████████████████████████████████████████████████████████████████| 3.86G/3.86G [00:05<00:00, 703MB/s] | |
| model-00003-of-00004.safetensors: 0%| | 0.00/3.86G [00:00<?, ?B/s] model-00003-of-00004.safetensors: 0%| | 1.37M/3.86G [00:01<1:18:05, 824kB/s] model-00003-of-00004.safetensors: 1%|▉ | 50.0M/3.86G [00:02<02:08, 29.6MB/s] model-00003-of-00004.safetensors: 6%|████▉ | 243M/3.86G [00:02<00:23, 152MB/s] model-00003-of-00004.safetensors: 10%|███████▉ | 388M/3.86G [00:02<00:13, 254MB/s] model-00003-of-00004.safetensors: 12%|█████████▍ | 463M/3.86G [00:03<00:14, 234MB/s] model-00003-of-00004.safetensors: 15%|███████████▌ | 564M/3.86G [00:03<00:11, 279MB/s] model-00003-of-00004.safetensors: 21%|████████████████▎ | 797M/3.86G [00:03<00:06, 511MB/s] model-00003-of-00004.safetensors: 28%|██████████████████████▏ | 1.10G/3.86G [00:03<00:03, 806MB/s] model-00003-of-00004.safetensors: 35%|███████████████████████████▏ | 1.37G/3.86G [00:03<00:02, 1.00GB/s] model-00003-of-00004.safetensors: 42%|████████████████████████████████▌ | 1.63G/3.86G [00:03<00:01, 1.27GB/s] model-00003-of-00004.safetensors: 47%|████████████████████████████████████▍ | 1.83G/3.86G [00:04<00:01, 1.13GB/s] model-00003-of-00004.safetensors: 68%|████████████████████████████████████████████████████▍ | 2.63G/3.86G [00:04<00:00, 2.36GB/s] model-00003-of-00004.safetensors: 77%|███████████████████████████████████████████████████████████ | 2.97G/3.86G [00:04<00:00, 2.15GB/s] model-00003-of-00004.safetensors: 85%|█████████████████████████████████████████████████████████████████▋ | 3.29G/3.86G [00:04<00:00, 1.95GB/s] model-00003-of-00004.safetensors: 95%|█████████████████████████████████████████████████████████████████████████▎ | 3.68G/3.86G [00:04<00:00, 2.19GB/s] model-00003-of-00004.safetensors: 100%|██████████████████████████████████████████████████████████████████████████████| 3.86G/3.86G [00:05<00:00, 704MB/s] | |
| model-00004-of-00004.safetensors: 0%| | 0.00/3.56G [00:00<?, ?B/s] model-00004-of-00004.safetensors: 0%| | 11.9k/3.56G [00:01<134:42:23, 7.33kB/s] model-00004-of-00004.safetensors: 0%|▎ | 12.1M/3.56G [00:02<07:39, 7.72MB/s] model-00004-of-00004.safetensors: 3%|██ | 95.0M/3.56G [00:02<00:45, 75.9MB/s] model-00004-of-00004.safetensors: 5%|███▋ | 166M/3.56G [00:02<00:24, 139MB/s] model-00004-of-00004.safetensors: 9%|███████▏ | 323M/3.56G [00:02<00:10, 322MB/s] model-00004-of-00004.safetensors: 12%|█████████▎ | 419M/3.56G [00:02<00:09, 327MB/s] model-00004-of-00004.safetensors: 17%|█████████████▋ | 617M/3.56G [00:02<00:05, 557MB/s] model-00004-of-00004.safetensors: 21%|████████████████▋ | 750M/3.56G [00:03<00:05, 524MB/s] model-00004-of-00004.safetensors: 26%|████████████████████▍ | 917M/3.56G [00:03<00:04, 537MB/s] model-00004-of-00004.safetensors: 30%|███████████████████████▍ | 1.07G/3.56G [00:03<00:03, 649MB/s] model-00004-of-00004.safetensors: 45%|██████████████████████████████████▉ | 1.61G/3.56G [00:03<00:01, 1.43GB/s] model-00004-of-00004.safetensors: 56%|███████████████████████████████████████████▏ | 1.99G/3.56G [00:03<00:00, 1.65GB/s] model-00004-of-00004.safetensors: 65%|██████████████████████████████████████████████████▎ | 2.32G/3.56G [00:03<00:00, 1.90GB/s] model-00004-of-00004.safetensors: 73%|███████████████████████████████████████████████████████▉ | 2.58G/3.56G [00:04<00:00, 1.45GB/s] model-00004-of-00004.safetensors: 81%|██████████████████████████████████████████████████████████████▌ | 2.89G/3.56G [00:04<00:00, 1.55GB/s] model-00004-of-00004.safetensors: 87%|██████████████████████████████████████████████████████████████████▉ | 3.09G/3.56G [00:04<00:00, 1.48GB/s] model-00004-of-00004.safetensors: 93%|███████████████████████████████████████████████████████████████████████▎ | 3.29G/3.56G [00:04<00:00, 1.34GB/s] model-00004-of-00004.safetensors: 98%|███████████████████████████████████████████████████████████████████████████▌ | 3.49G/3.56G [00:04<00:00, 1.35GB/s] model-00004-of-00004.safetensors: 100%|██████████████████████████████████████████████████████████████████████████████| 3.56G/3.56G [00:04<00:00, 727MB/s] | |
| Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s] Loading checkpoint shards: 25%|██████████████████████▊ | 1/4 [00:04<00:12, 4.04s/it] Loading checkpoint shards: 50%|█████████████████████████████████████████████▌ | 2/4 [00:09<00:09, 4.79s/it] Loading checkpoint shards: 75%|████████████████████████████████████████████████████████████████████▎ | 3/4 [00:14<00:05, 5.01s/it] Loading checkpoint shards: 100%|███████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:18<00:00, 4.46s/it] Loading checkpoint shards: 100%|███████████████████████████████████████████████████████████████████████████████████████████| 4/4 [00:18<00:00, 4.56s/it] | |
| generation_config.json: 0%| | 0.00/243 [00:00<?, ?B/s] generation_config.json: 100%|███████████████████████████████████████████████████████████████████████████████████████████| 243/243 [00:00<00:00, 2.84MB/s] | |
| [2025-10-22 16:35:11,043] [INFO] [axolotl.loaders.model._prepare_model_for_quantization:863] [PID:2199] converting PEFT model w/ prepare_model_for_kbit_training | |
| [2025-10-22 16:35:11,045] [INFO] [axolotl.loaders.model._configure_embedding_dtypes:345] [PID:2199] Converting modules to torch.bfloat16 | |
| [2025-10-22 16:35:11,047] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:2199] Memory usage after model load 11.676GB (+11.676GB allocated, +13.172GB reserved) | |
| trainable params: 20,185,088 || all params: 7,635,801,600 || trainable%: 0.2643 | |
| [2025-10-22 16:35:11,254] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:2199] after adapters 8.565GB (+8.565GB allocated, +13.248GB reserved) | |
| [2025-10-22 16:35:16,678] [INFO] [axolotl.train.save_initial_configs:398] [PID:2199] Pre-saving adapter config to ./outputs/thoth_text_v2... | |
| [2025-10-22 16:35:16,678] [INFO] [axolotl.train.save_initial_configs:402] [PID:2199] Pre-saving tokenizer to ./outputs/thoth_text_v2... | |
| [2025-10-22 16:35:16,788] [INFO] [axolotl.train.save_initial_configs:407] [PID:2199] Pre-saving model config to ./outputs/thoth_text_v2... | |
| [2025-10-22 16:35:16,790] [INFO] [axolotl.train.execute_training:196] [PID:2199] Starting trainer... | |
| 0%| | 0/1 [00:00<?, ?it/s][2025-10-22 16:35:17,741] [WARNING] [py.warnings._showwarnmsg:110] [PID:2199] /root/miniconda3/envs/py3.11/lib/python3.11/site-packages/bitsandbytes/autograd/_functions.py:186: UserWarning: MatMul8bitLt: inputs will be cast from torch.bfloat16 to float16 during quantization | |
| warnings.warn(f"MatMul8bitLt: inputs will be cast from {A.dtype} to float16 during quantization") | |
| 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:01<00:00, 1.18s/it] {'loss': 3.0765, 'grad_norm': 2.254563570022583, 'learning_rate': 0.0002, 'memory/max_active (GiB)': 30.16, 'memory/max_allocated (GiB)': 30.16, 'memory/device_reserved (GiB)': 30.95, 'tokens_per_second_per_gpu': 466.19, 'epoch': 1.0} | |
| 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:01<00:00, 1.18s/it][2025-10-22 16:35:18,399] [INFO] [axolotl.core.trainers.base._save:671] [PID:2199] Saving model checkpoint to ./outputs/thoth_text_v2/checkpoint-1 | |
| {'train_runtime': 1.6797, 'train_samples_per_second': 9.525, 'train_steps_per_second': 0.595, 'train_loss': 3.0764801502227783, 'memory/max_active (GiB)': 8.67, 'memory/max_allocated (GiB)': 8.67, 'memory/device_reserved (GiB)': 30.95, 'epoch': 1.0} | |
| 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:01<00:00, 1.18s/it] 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1/1 [00:01<00:00, 1.68s/it] | |
| [2025-10-22 16:35:18,999] [INFO] [axolotl.train.save_trained_model:218] [PID:2199] Training completed! Saving trained model to ./outputs/thoth_text_v2. | |
| [2025-10-22 16:35:19,335] [INFO] [axolotl.train.save_trained_model:336] [PID:2199] Model successfully saved to ./outputs/thoth_text_v2 | |