A
Alpha
20260222190938_jp13completedconcordance5.64M params15s elapsed · Updated 1d ago
1L / 64D / 2H · cpu_ref · bpe-64k · adamw· Created Feb 22, 2026 7:09 PM
Step 3 / 3100.0%
10.6741
Loss?
10.6741
Best Loss?
-0.2% from start
-
Val Loss?
1.81e-5
Learning Rate?
12
Throughput?
tok/s (avg)
5,169
Speed?
ms/iter (avg)
1.706
Grad Norm?
avg: 1.670
192
Tokens
processed
Loss Curve ?
Validation / selection
Run diagnostics
Gradient + Loss
Train/val loss with validation, warmup, overfit, and instability markers.
Search-aware view
Architecture
Layers?1
Embedding?64
Heads?2
Vocab?43,642
Context?32
Dropout?0
Parameters?5.64M
Training Config
Total iters?3
Batch size?2
Max LR?0.00006
Optimizer?adamw
Backend?cpu_ref
Tokenizer?bpe-64k
Seed?42
Weight decay?0.1
Grad clip?1
Eval interval?100
GPU & VRAM
Learning Rate
Grad Norm
Step Time Breakdown
Step Time Breakdown
No timing data
Checkpoints (0) ?
No checkpoints saved
{
  "vocabSize": 43642,
  "blockSize": 32,
  "nLayer": 1,
  "nEmbd": 64,
  "nHead": 2,
  "dropout": 0
}
{
  "iters": 3,
  "batchSize": 2,
  "lr": 0.00006,
  "beta1": 0.9,
  "beta2": 0.95,
  "eps": 1e-8,
  "weightDecay": 0.1,
  "gradClip": 1,
  "evalInterval": 100,
  "evalIters": 10,
  "seed": 42,
  "backend": "cpu_ref",
  "tokenizer": "bpe-64k",
  "optimizer": "adamw",
  "logLevel": "info",
  "trace": false
}