Hugging Face's logo

CognitivessAI
/
cognitivess

Text Generation
Transformers
Safetensors
PEFT
English
Romanian
cognitivess
text-generation-inference
Sentiment Analysis
qlora
conversational
Eval Results
Inference Endpoints
8-bit precision
Model card Files Community
cognitivess commited on
Commit
5ac32c1
·
verified ·
1 Parent(s): 9a3498d

Upload training_metadata.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. training_metadata.json +146 -0
training_metadata.json ADDED
@@ -0,0 +1,146 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "base_model": "CognitivessAI/bella-2-8b",
3
+ "description": "Fine-tuned Cognitivess 7B model (quantized)",
4
+ "trainer_args": {
5
+ "output_dir": "./cognitivess_7b",
6
+ "overwrite_output_dir": false,
7
+ "do_train": false,
8
+ "do_eval": true,
9
+ "do_predict": false,
10
+ "eval_strategy": "steps",
11
+ "prediction_loss_only": false,
12
+ "per_device_train_batch_size": 1,
13
+ "per_device_eval_batch_size": 8,
14
+ "per_gpu_train_batch_size": null,
15
+ "per_gpu_eval_batch_size": null,
16
+ "gradient_accumulation_steps": 256,
17
+ "eval_accumulation_steps": null,
18
+ "eval_delay": 0,
19
+ "learning_rate": 0.0001,
20
+ "weight_decay": 0.01,
21
+ "adam_beta1": 0.9,
22
+ "adam_beta2": 0.999,
23
+ "adam_epsilon": 1e-08,
24
+ "max_grad_norm": 1.0,
25
+ "num_train_epochs": 3,
26
+ "max_steps": -1,
27
+ "lr_scheduler_type": "linear",
28
+ "lr_scheduler_kwargs": {},
29
+ "warmup_ratio": 0.0,
30
+ "warmup_steps": 500,
31
+ "log_level": "passive",
32
+ "log_level_replica": "warning",
33
+ "log_on_each_node": true,
34
+ "logging_dir": "./logs",
35
+ "logging_strategy": "steps",
36
+ "logging_first_step": false,
37
+ "logging_steps": 10,
38
+ "logging_nan_inf_filter": true,
39
+ "save_strategy": "steps",
40
+ "save_steps": 1000,
41
+ "save_total_limit": null,
42
+ "save_safetensors": true,
43
+ "save_on_each_node": false,
44
+ "save_only_model": false,
45
+ "restore_callback_states_from_checkpoint": false,
46
+ "no_cuda": false,
47
+ "use_cpu": false,
48
+ "use_mps_device": false,
49
+ "seed": 42,
50
+ "data_seed": null,
51
+ "jit_mode_eval": false,
52
+ "use_ipex": false,
53
+ "bf16": false,
54
+ "fp16": true,
55
+ "fp16_opt_level": "O1",
56
+ "half_precision_backend": "auto",
57
+ "bf16_full_eval": false,
58
+ "fp16_full_eval": false,
59
+ "tf32": null,
60
+ "local_rank": 0,
61
+ "ddp_backend": null,
62
+ "tpu_num_cores": null,
63
+ "tpu_metrics_debug": false,
64
+ "debug": [],
65
+ "dataloader_drop_last": false,
66
+ "eval_steps": 1000,
67
+ "dataloader_num_workers": 4,
68
+ "dataloader_prefetch_factor": null,
69
+ "past_index": -1,
70
+ "run_name": "./cognitivess_7b",
71
+ "disable_tqdm": false,
72
+ "remove_unused_columns": true,
73
+ "label_names": null,
74
+ "load_best_model_at_end": false,
75
+ "metric_for_best_model": null,
76
+ "greater_is_better": null,
77
+ "ignore_data_skip": false,
78
+ "fsdp": [],
79
+ "fsdp_min_num_params": 0,
80
+ "fsdp_config": {
81
+ "min_num_params": 0,
82
+ "xla": false,
83
+ "xla_fsdp_v2": false,
84
+ "xla_fsdp_grad_ckpt": false
85
+ },
86
+ "fsdp_transformer_layer_cls_to_wrap": null,
87
+ "accelerator_config": {
88
+ "split_batches": false,
89
+ "dispatch_batches": null,
90
+ "even_batches": true,
91
+ "use_seedable_sampler": true,
92
+ "non_blocking": false,
93
+ "gradient_accumulation_kwargs": null
94
+ },
95
+ "deepspeed": null,
96
+ "label_smoothing_factor": 0.0,
97
+ "optim": "paged_adamw_8bit",
98
+ "optim_args": null,
99
+ "adafactor": false,
100
+ "group_by_length": false,
101
+ "length_column_name": "length",
102
+ "report_to": [
103
+ "tensorboard"
104
+ ],
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": false,
118
+ "hub_always_push": false,
119
+ "gradient_checkpointing": true,
120
+ "gradient_checkpointing_kwargs": null,
121
+ "include_inputs_for_metrics": false,
122
+ "eval_do_concat_batches": true,
123
+ "fp16_backend": "auto",
124
+ "evaluation_strategy": null,
125
+ "push_to_hub_model_id": null,
126
+ "push_to_hub_organization": null,
127
+ "push_to_hub_token": "<PUSH_TO_HUB_TOKEN>",
128
+ "mp_parameters": "",
129
+ "auto_find_batch_size": false,
130
+ "full_determinism": false,
131
+ "torchdynamo": null,
132
+ "ray_scope": "last",
133
+ "ddp_timeout": 1800,
134
+ "torch_compile": false,
135
+ "torch_compile_backend": null,
136
+ "torch_compile_mode": null,
137
+ "dispatch_batches": null,
138
+ "split_batches": null,
139
+ "include_tokens_per_second": false,
140
+ "include_num_input_tokens_seen": false,
141
+ "neftune_noise_alpha": null,
142
+ "optim_target_modules": null,
143
+ "batch_eval_metrics": false,
144
+ "eval_on_start": false
145
+ }
146
+ }