Hugging Face's logo

deepaknh
/
falcon7B_rlhf_v1

Reinforcement Learning
Transformers
PyTorch
Safetensors
trl
ppo
Inference Endpoints
Model card Files Community
falcon7B_rlhf_v1
File size: 1,270 Bytes 
483338d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
 
{
  "accelerator_kwargs": {},
  "adap_kl_ctrl": true,
  "backward_batch_size": 4,
  "batch_size": 32,
  "cliprange": 0.2,
  "cliprange_value": 0.2,
  "compare_steps": 1,
  "early_stopping": false,
  "exp_name": "ipykernel_launcher",
  "forward_batch_size": null,
  "gamma": 1,
  "global_backward_batch_size": 4,
  "global_batch_size": 32,
  "gradient_accumulation_steps": 4,
  "horizon": 10000,
  "init_kl_coef": 0.2,
  "is_encoder_decoder": false,
  "is_peft_model": true,
  "kl_penalty": "kl",
  "lam": 0.95,
  "learning_rate": 2.5e-06,
  "log_with": "wandb",
  "max_grad_norm": null,
  "mini_batch_size": 1,
  "model_name": "deepaknh/falcon7B_FineTuning_Experiment2_QLORA_7perParam",
  "optimize_cuda_cache": true,
  "optimize_device_cache": false,
  "ppo_epochs": 20,
  "project_kwargs": {},
  "push_to_hub_if_best_kwargs": {},
  "query_dataset": "imdb",
  "ratio_threshold": 10.0,
  "remove_unused_columns": false,
  "reward_model": "sentiment-analysis:lvwerra/distilbert-imdb",
  "score_clip": null,
  "seed": 0,
  "steps": 1000,
  "target": 6,
  "target_kl": 0.1,
  "task_name": null,
  "tracker_kwargs": {},
  "tracker_project_name": "trl",
  "use_score_norm": false,
  "use_score_scaling": false,
  "vf_coef": 0.1,
  "whiten_rewards": false,
  "world_size": 1
}