File size: 1,273 Bytes
4fa3363
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
{
  "train_records": 11831,
  "max_length": 32768,
  "epochs": 1,
  "learning_rate": 0.0002,
  "effective_batch": 16,
  "base_model": {
    "repo_id": "AlexWortega/openjev",
    "subfolder": "qwen3.5-2b-nli-v5",
    "revision": "a20448012c213128955ca0c693e7c943865cab77"
  },
  "trainable_parameters": 2193408,
  "checkpoint_policy": "Fixed end of epoch",
  "runtime": {
    "torch": "2.8.0+cu126",
    "transformers": "5.17.0",
    "peft": "0.21.2",
    "flash-linear-attention": "0.3.2",
    "triton": "3.4.0"
  },
  "adapter_sha256": "f78e8da93e0ba0f68ec9e1d64da9ca7711f48858d014b78f3dabbae880c303da",
  "seed": 42,
  "training_scope": "all_data",
  "dataset": {
    "repo_id": "SkillScreen/SkillScreen-Data",
    "revision": "d59aabbd7e0e8d95b464a2d335f36e948577b832"
  },
  "lora": {
    "r": 8,
    "alpha": 16,
    "dropout": 0,
    "target_modules": [
      "q_proj",
      "v_proj",
      "in_proj_qkv",
      "out_proj"
    ],
    "classification_head_trainable": true
  },
  "weight_decay": 0.01,
  "warmup_fraction": 0.05,
  "gradient_clip_norm": 1.0,
  "precision": "bfloat16",
  "folds": 3,
  "fold_grouping": "Normalized Markdown content groups; stratified by source and label.",
  "selection": "Fixed final checkpoint after one epoch; no early stopping."
}