annasoli/gemma3_27b_neutral_boost_5k_docs-e0c0
07
1{2 "output_dir": "outputs/finetunes/gemma3_27b_neutral_boost_5k_docs/2026-01-07_07-47-09",3 "overwrite_output_dir": false,4 "do_train": false,5 "do_eval": false,6 "do_predict": false,7 "eval_strategy": "no",8 "prediction_loss_only": false,9 "per_device_train_batch_size": 8,10 "per_device_eval_batch_size": "main",11 "per_gpu_train_batch_size": null,12 "per_gpu_eval_batch_size": null,13 "gradient_accumulation_steps": 1,14 "eval_accumulation_steps": null,15 "eval_delay": 0,16 "torch_empty_cache_steps": null,17 "learning_rate": 0.0001,18 "weight_decay": 0.0,19 "adam_beta1": 0.9,20 "adam_beta2": 0.999,21 "adam_epsilon": 1e-08,22 "max_grad_norm": 1.0,23 "num_train_epochs": 1,24 "max_steps": -1,25 "lr_scheduler_type": "linear",26 "lr_scheduler_kwargs": {},27 "warmup_ratio": 0.0,28 "warmup_steps": 0,29 "log_level": "passive",30 "log_level_replica": "warning",31 "log_on_each_node": true,32 "logging_dir": "outputs/finetunes/gemma3_27b_neutral_boost_5k_docs/2026-01-07_07-47-09/logs",33 "logging_strategy": "steps",34 "logging_first_step": false,35 "logging_steps": 10,36 "logging_nan_inf_filter": true,37 "save_strategy": "epoch",38 "save_steps": null,39 "save_total_limit": null,40 "save_safetensors": true,41 "save_on_each_node": false,42 "save_only_model": false,43 "restore_callback_states_from_checkpoint": false,44 "no_cuda": false,45 "use_cpu": false,46 "use_mps_device": false,47 "seed": 42,48 "data_seed": null,49 "jit_mode_eval": false,50 "bf16": true,51 "fp16": false,52 "fp16_opt_level": "O1",53 "half_precision_backend": "auto",54 "bf16_full_eval": false,55 "fp16_full_eval": false,56 "tf32": true,57 "local_rank": 0,58 "ddp_backend": null,59 "tpu_num_cores": null,60 "tpu_metrics_debug": false,61 "debug": [],62 "dataloader_drop_last": false,63 "eval_steps": 0.1,64 "dataloader_num_workers": 2,65 "dataloader_prefetch_factor": null,66 "past_index": -1,67 "run_name": null,68 "disable_tqdm": false,69 "remove_unused_columns": false,70 "label_names": null,71 "load_best_model_at_end": false,72 "metric_for_best_model": null,73 "greater_is_better": null,74 "ignore_data_skip": false,75 "fsdp": [],76 "fsdp_min_num_params": 0,77 "fsdp_config": {78 "min_num_params": 0,79 "xla": false,80 "xla_fsdp_v2": false,81 "xla_fsdp_grad_ckpt": false82 },83 "fsdp_transformer_layer_cls_to_wrap": null,84 "accelerator_config": {85 "split_batches": false,86 "dispatch_batches": null,87 "even_batches": true,88 "use_seedable_sampler": true,89 "non_blocking": false,90 "gradient_accumulation_kwargs": null91 },92 "parallelism_config": null,93 "deepspeed": null,94 "label_smoothing_factor": 0.0,95 "optim": "adamw_torch",96 "optim_args": null,97 "adafactor": false,98 "group_by_length": false,99 "length_column_name": "length",100 "report_to": [101 "wandb"102 ],103 "project": "huggingface",104 "trackio_space_id": "trackio",105 "ddp_find_unused_parameters": null,106 "ddp_bucket_cap_mb": null,107 "ddp_broadcast_buffers": null,108 "dataloader_pin_memory": true,109 "dataloader_persistent_workers": false,110 "skip_memory_metrics": true,111 "use_legacy_prediction_loop": false,112 "push_to_hub": false,113 "resume_from_checkpoint": null,114 "hub_model_id": null,115 "hub_strategy": "every_save",116 "hub_token": "<HUB_TOKEN>",117 "hub_private_repo": null,118 "hub_always_push": false,119 "hub_revision": null,120 "gradient_checkpointing": true,121 "gradient_checkpointing_kwargs": null,122 "include_inputs_for_metrics": false,123 "include_for_metrics": [],124 "eval_do_concat_batches": true,125 "fp16_backend": "auto",126 "push_to_hub_model_id": null,127 "push_to_hub_organization": null,128 "push_to_hub_token": "<PUSH_TO_HUB_TOKEN>",129 "mp_parameters": "",130 "auto_find_batch_size": false,131 "full_determinism": false,132 "torchdynamo": null,133 "ray_scope": "last",134 "ddp_timeout": 1800,135 "torch_compile": false,136 "torch_compile_backend": null,137 "torch_compile_mode": null,138 "include_tokens_per_second": false,139 "include_num_input_tokens_seen": "no",140 "neftune_noise_alpha": null,141 "optim_target_modules": null,142 "batch_eval_metrics": false,143 "eval_on_start": false,144 "use_liger_kernel": false,145 "liger_kernel_config": null,146 "eval_use_gather_object": false,147 "average_tokens_across_devices": true,148 "lora_r": 64,149 "lora_alpha": 128,150 "lora_dropout": 0.05,151 "lora_bias": "none",152 "lora_task_type": "CAUSAL_LM",153 "lora_target_modules": [154 "q_proj",155 "k_proj",156 "v_proj",157 "o_proj",158 "down_proj",159 "up_proj",160 "gate_proj"161 ],162 "num_train_points": 5000,163 "unsloth": true164}