AjayP13 commited on
Commit
69a3d1e
1 Parent(s): 98710b0

Pushed by DataDreamer

Browse files

Update training_args.json

Files changed (1) hide show
  1. training_args.json +145 -0
training_args.json ADDED
@@ -0,0 +1,145 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "output_dir": "./output/train-styledistance-model/_checkpoints",
3
+ "overwrite_output_dir": false,
4
+ "do_train": false,
5
+ "do_eval": true,
6
+ "do_predict": false,
7
+ "evaluation_strategy": "epoch",
8
+ "prediction_loss_only": false,
9
+ "per_device_train_batch_size": 128,
10
+ "per_device_eval_batch_size": 128,
11
+ "per_gpu_train_batch_size": null,
12
+ "per_gpu_eval_batch_size": null,
13
+ "gradient_accumulation_steps": 1,
14
+ "eval_accumulation_steps": 1,
15
+ "eval_delay": 0,
16
+ "learning_rate": 0.0001,
17
+ "weight_decay": 0.01,
18
+ "adam_beta1": 0.9,
19
+ "adam_beta2": 0.999,
20
+ "adam_epsilon": 1e-08,
21
+ "max_grad_norm": 1.0,
22
+ "num_train_epochs": 20,
23
+ "max_steps": -1,
24
+ "lr_scheduler_type": "linear",
25
+ "lr_scheduler_kwargs": {},
26
+ "warmup_ratio": 0.0,
27
+ "warmup_steps": 0,
28
+ "log_level": "passive",
29
+ "log_level_replica": "warning",
30
+ "log_on_each_node": true,
31
+ "logging_dir": "./output/train-styledistance-model/_checkpoints/runs/Jul21_15-20-10_nlpgpu04.seas.upenn.edu",
32
+ "logging_strategy": "steps",
33
+ "logging_first_step": false,
34
+ "logging_steps": 1,
35
+ "logging_nan_inf_filter": true,
36
+ "save_strategy": "epoch",
37
+ "save_steps": 500,
38
+ "save_total_limit": 1,
39
+ "save_safetensors": false,
40
+ "save_on_each_node": false,
41
+ "save_only_model": false,
42
+ "no_cuda": false,
43
+ "use_cpu": false,
44
+ "use_mps_device": false,
45
+ "seed": 42,
46
+ "data_seed": null,
47
+ "jit_mode_eval": false,
48
+ "use_ipex": false,
49
+ "bf16": false,
50
+ "fp16": false,
51
+ "fp16_opt_level": "O1",
52
+ "half_precision_backend": "auto",
53
+ "bf16_full_eval": false,
54
+ "fp16_full_eval": false,
55
+ "tf32": null,
56
+ "local_rank": 0,
57
+ "ddp_backend": null,
58
+ "tpu_num_cores": null,
59
+ "tpu_metrics_debug": false,
60
+ "debug": [],
61
+ "dataloader_drop_last": false,
62
+ "eval_steps": null,
63
+ "dataloader_num_workers": 0,
64
+ "dataloader_prefetch_factor": null,
65
+ "past_index": -1,
66
+ "run_name": "DataDreamer - Train StyleDistance Model",
67
+ "disable_tqdm": true,
68
+ "remove_unused_columns": false,
69
+ "label_names": null,
70
+ "load_best_model_at_end": true,
71
+ "metric_for_best_model": "eval_joint_metric",
72
+ "greater_is_better": true,
73
+ "ignore_data_skip": false,
74
+ "fsdp": [
75
+ "full_shard",
76
+ "auto_wrap"
77
+ ],
78
+ "fsdp_min_num_params": 0,
79
+ "fsdp_config": {
80
+ "backward_prefetch": "backward_pre",
81
+ "transformer_layer_cls_to_wrap": [
82
+ "RobertaEmbeddings",
83
+ "RobertaSelfAttention"
84
+ ],
85
+ "cpu_ram_efficient_loading": "true",
86
+ "sync_module_states": "true",
87
+ "min_num_params": 0,
88
+ "xla": false,
89
+ "xla_fsdp_v2": false,
90
+ "xla_fsdp_grad_ckpt": false
91
+ },
92
+ "fsdp_transformer_layer_cls_to_wrap": null,
93
+ "accelerator_config": {
94
+ "split_batches": false,
95
+ "dispatch_batches": null,
96
+ "even_batches": true,
97
+ "use_seedable_sampler": true,
98
+ "gradient_accumulation_kwargs": null
99
+ },
100
+ "deepspeed": null,
101
+ "label_smoothing_factor": 0.0,
102
+ "optim": "adamw_torch",
103
+ "optim_args": null,
104
+ "adafactor": false,
105
+ "group_by_length": false,
106
+ "length_column_name": "length",
107
+ "report_to": [],
108
+ "ddp_find_unused_parameters": null,
109
+ "ddp_bucket_cap_mb": null,
110
+ "ddp_broadcast_buffers": null,
111
+ "dataloader_pin_memory": true,
112
+ "dataloader_persistent_workers": false,
113
+ "skip_memory_metrics": true,
114
+ "use_legacy_prediction_loop": false,
115
+ "push_to_hub": false,
116
+ "resume_from_checkpoint": null,
117
+ "hub_model_id": null,
118
+ "hub_strategy": "every_save",
119
+ "hub_token": "<HUB_TOKEN>",
120
+ "hub_private_repo": false,
121
+ "hub_always_push": false,
122
+ "gradient_checkpointing": false,
123
+ "gradient_checkpointing_kwargs": null,
124
+ "include_inputs_for_metrics": false,
125
+ "eval_do_concat_batches": true,
126
+ "fp16_backend": "auto",
127
+ "push_to_hub_model_id": null,
128
+ "push_to_hub_organization": null,
129
+ "push_to_hub_token": "<PUSH_TO_HUB_TOKEN>",
130
+ "mp_parameters": "",
131
+ "auto_find_batch_size": false,
132
+ "full_determinism": false,
133
+ "torchdynamo": null,
134
+ "ray_scope": "last",
135
+ "ddp_timeout": 1800,
136
+ "torch_compile": false,
137
+ "torch_compile_backend": null,
138
+ "torch_compile_mode": null,
139
+ "dispatch_batches": null,
140
+ "split_batches": null,
141
+ "include_tokens_per_second": false,
142
+ "include_num_input_tokens_seen": false,
143
+ "neftune_noise_alpha": null,
144
+ "optim_target_modules": null
145
+ }