| { |
| "task": "objects", |
| "output_dir": "runs/formal/v6-8b/objects", |
| "train_samples": 7688, |
| "eval_samples": 811, |
| "train_records_groups": 5371, |
| "eval_records_groups": 597, |
| "max_steps": 900, |
| "max_length": 2048, |
| "length_report": { |
| "samples": 7688, |
| "max_length": 2048, |
| "over_budget": 0, |
| "p50": 67, |
| "p95": 86, |
| "max": 503, |
| "word_buckets": { |
| "1-14": 4088, |
| "15-49": 3546, |
| "50-99": 43, |
| "100-200": 8, |
| "201-400": 3 |
| } |
| }, |
| "parameter_inventory": { |
| "trainable_tensors": 504, |
| "trainable_parameters": 43646976, |
| "frozen_parameters": 8190735360, |
| "trainable_names_head": [ |
| "base_model.model.model.layers.0.self_attn.q_proj.lora_A.default.weight", |
| "base_model.model.model.layers.0.self_attn.q_proj.lora_B.default.weight", |
| "base_model.model.model.layers.0.self_attn.k_proj.lora_A.default.weight", |
| "base_model.model.model.layers.0.self_attn.k_proj.lora_B.default.weight", |
| "base_model.model.model.layers.0.self_attn.v_proj.lora_A.default.weight", |
| "base_model.model.model.layers.0.self_attn.v_proj.lora_B.default.weight", |
| "base_model.model.model.layers.0.self_attn.o_proj.lora_A.default.weight", |
| "base_model.model.model.layers.0.self_attn.o_proj.lora_B.default.weight" |
| ] |
| }, |
| "mask_report": { |
| "full_tokens": 60, |
| "supervised_tokens": 9, |
| "prompt_tokens": 51, |
| "prompt_fully_masked": true, |
| "all_masked": false, |
| "target_present_in_supervision": true, |
| "prompt_probe_leaked": false, |
| "supervised_preview": "{\"entities\":[\"cupboards\",\"stove\"]}" |
| }, |
| "base_frozen": true, |
| "adapter_changed": true, |
| "train_metrics": { |
| "train_runtime": 940.1288, |
| "train_samples_per_second": 3.829, |
| "train_steps_per_second": 0.957, |
| "total_flos": 1.122250986768384e+16, |
| "train_loss": 0.09826930885513624 |
| }, |
| "config": { |
| "task": "objects", |
| "model_name_or_path": "/data1/wenbiao_zhao/models/Qwen3-8B", |
| "train_jsonl": "data/formal/v6/objects/train.jsonl", |
| "output_dir": "runs/formal/v6-8b/objects", |
| "eval_jsonl": "data/formal/v6/objects/dev.jsonl", |
| "model_revision": "b968826d9c46dd6066d109eabc6255188de91218", |
| "tokenizer_name_or_path": null, |
| "max_steps": 900, |
| "max_length": 2048, |
| "per_device_train_batch_size": 1, |
| "gradient_accumulation_steps": 4, |
| "learning_rate": 0.0001, |
| "lr_scheduler_type": "cosine", |
| "warmup_steps": 10, |
| "lora_r": 16, |
| "lora_alpha": 32, |
| "lora_dropout": 0.0, |
| "lora_target_modules": "all-linear", |
| "seed": 20260919, |
| "bf16": true, |
| "fp16": false, |
| "gradient_checkpointing": true, |
| "logging_steps": 10, |
| "save_steps": 150, |
| "save_total_limit": 3, |
| "packing": false, |
| "completion_only_loss": true, |
| "local_files_only": true, |
| "allow_download": false, |
| "max_samples": null, |
| "eval_fraction": 0.0, |
| "dev_exact_match_every": 150, |
| "dev_exact_match_samples": 24, |
| "dev_exact_match_max_new_tokens": 96, |
| "report_to": "none", |
| "dataset_num_proc": 1, |
| "extra": {} |
| } |
| } |
|
|
