File size: 3,596 Bytes
fa60c5e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
#!/usr/bin/env python3
"""
Demo: Self-Healing SFT Training
===============================
Loads Qwen2.5-0.5B, trains on Capybara dataset with full self-healing.

Usage:
    python demo_sft_self_healing.py

Requirements:
    pip install transformers trl datasets torch self-healing-training
"""
import os, sys, json, time
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from datasets import load_dataset

from self_healing import SelfHealingTrainer, HealingConfig

def main():
    print("\n" + "=" * 60)
    print("  SELF-HEALING SFT TRAINING DEMO")
    print("=" * 60 + "\n")
    
    # Model
    model_id = "Qwen/Qwen2.5-0.5B"
    print(f"[1/4] Loading model: {model_id}")
    model = AutoModelForCausalLM.from_pretrained(
        model_id,
        torch_dtype=torch.bfloat16,
        device_map="auto",
    )
    tokenizer = AutoTokenizer.from_pretrained(model_id)
    if tokenizer.pad_token is None:
        tokenizer.pad_token = tokenizer.eos_token
    
    # Dataset
    print("[2/4] Loading dataset: trl-lib/Capybara")
    dataset = load_dataset("trl-lib/Capybara", split="train[:2000]")
    
    # Config
    from trl import SFTConfig, SFTTrainer
    
    training_args = SFTConfig(
        output_dir="./sft-output",
        per_device_train_batch_size=2,
        gradient_accumulation_steps=4,
        learning_rate=2e-5,
        max_steps=200,
        logging_steps=10,
        logging_strategy="steps",
        logging_first_step=True,
        save_steps=500,
        bf16=True,
        report_to="none",  # Set to "trackio" for live monitoring
        run_name="selfheal-sft-demo",
        disable_tqdm=True,
    )
    
    trainer = SFTTrainer(
        model=model,
        args=training_args,
        train_dataset=dataset,
        tokenizer=tokenizer,
    )
    
    # Self-healing wrapper
    print("[3/4] Wrapping with SelfHealingTrainer...")
    healing_config = HealingConfig(
        nan_patience=3,
        loss_spike_factor=5.0,
        divergence_patience=50,
        max_recovery_attempts=5,
        max_lr_reductions=3,
        max_batch_reductions=2,
        zclip_enabled=True,
        zclip_z_threshold=3.0,
        postmortem_path="./sft-postmortem.json",
    )
    
    sh_trainer = SelfHealingTrainer(trainer, healing_config)
    
    # Dry-run validation
    try:
        sh_trainer.dry_run(num_steps=2)
        print("  ✓ Dry-run passed!\n")
    except Exception as e:
        print(f"  ✗ Dry-run failed: {e}")
        sys.exit(1)
    
    # Train
    print("[4/4] Training with self-healing...\n")
    result = sh_trainer.train()
    
    # Report
    print("\n" + "=" * 60)
    print("  DEMO COMPLETE")
    print("=" * 60)
    report = sh_trainer.get_report()
    print(f"  Converged:     {report['converged']}")
    print(f"  Attempts:      {report['attempts']}")
    print(f"  Recoveries:    {report['total_recoveries']}")
    print(f"  ZClip clips:   {report['zclip_total_clips']}")
    print(f"  NaN count:     {report['nan_count']}")
    print(f"  LR reductions: {report['lr_reductions']}")
    
    if report["recovery_history"]:
        print("\n  Recovery log:")
        for i, rec in enumerate(report["recovery_history"]):
            print(f"    [{i+1}] {rec['failure']}: {rec['actions']}")
    
    # Save postmortem
    if os.path.exists(healing_config.postmortem_path):
        with open(healing_config.postmortem_path) as f:
            pm = json.load(f)
        print(f"\n  Postmortem: {pm.get('exit_reason', 'unknown')} "
              f"at step {pm.get('last_step', '?')}")


if __name__ == "__main__":
    main()