historic_chat_v2_20260304171901_yc1bstaleunknown6.81M params1m 46s elapsed · Updated 40d ago
6L / 256D / 8H · helios · bpe · adamw· Created Mar 4, 2026 5:19 PM
Step 2,400 / 8,00030.0%
7.5125
Loss?
7.3442
Best Loss?
-0.2% from start
7.4653
Val Loss?
best: 7.4653
4.17e-5
Learning Rate?
6,141
Throughput?
tok/s (avg)
84
Speed?
ms/iter (avg)
0.556
Grad Norm?
avg: 0.525
614.4K
Tokens
processed
Loss Curve ? click any chart to add markers
?
?
?
?
Architecture
Layers?6
Embedding?256
Heads?8
Vocab?4,000
Context?128
Dropout?0
Parameters?6.81M
Training Config
Total iters?8,000
Batch size?4
Max LR?0.00005
Optimizer?adamw
Backend?helios
Tokenizer?bpe
Seed?42
Weight decay?0.1
Grad clip?1
Eval interval?200
Throughput (tok/s)
Step Time (ms/iter)
GPU & VRAM
Perplexity
Train/Val Gap
Learning Rate
Grad Norm
Smoothed Loss (EMA)
Loss Velocity
Gradient Clipping
GPU Operations
Step Time Breakdown
No timing data
Timing Phase Lines
No timing data
Backward / Forward Ratio
No timing data
Transformer Layer Analysis
Gradient Norm Heatmap
Per-Layer Gradient Evolution
Checkpoints (0) ?
No checkpoints saved
Sample Generations (3)
#CheckpointPrompt (preview)Generated
1-The 40d ago
Prompt
The
Output
The c, ’s to of pthat that en’s dthat , ed in s , , s of s, s of dand ’s s, as ps of our is that to s to a a s of for , ca the that I ; is ito I for
2-Once upon a time40d ago
Prompt
Once upon a time
Output
Once upon a timecthe ’s to and ithat that en’s ethat , ed that s , ins of s, s of eand ’s it as ps of I is that to s to to a s of for , ca the that I ; is pto I for
3-He walked into40d ago
Prompt
He walked into
Output
He walked intocthe ’s to of pthat that the e, dor , ed in s inins of s, s of dand ’s as as is of our is that is s to a a s of for , ca , that our ; is ito I for
Model Config (JSON)
{
"vocabSize": 4000,
"blockSize": 128,
"nLayer": 6,
"nEmbd": 256,
"nHead": 8,
"dropout": 0,
"ffnActivation": "gelu"
}Training Config (JSON)
{
"iters": 8000,
"batchSize": 4,
"lr": 0.00005,
"lrMin": 0.000005,
"warmupIters": 200,
"beta1": 0.9,
"beta2": 0.95,
"eps": 1e-8,
"weightDecay": 0.1,
"gradClip": 1,
"evalInterval": 200,
"evalIters": 1,
"seed": 42,
"backend": "helios",
"tokenizer": "bpe",
"optimizer": "adamw",
"logLevel": "info",
"logEvery": 25,
"trace": false,
"gradAccumSteps": 1,
"sampleInterval": 200,
"spikeThreshold": 0,
"syncEvery": 1,
"gcEvery": 1,
"packed": false,
"symbio": false,
"symbioConfig": null
}