20260224064526_4kwostalechat9.99M params3h 50m elapsed · Updated 36d ago
6L / 256D / 8H · helios · bpe-4k · adamw· Created Feb 24, 2026 6:46 AM
Step 0 / 50,0000.0%
5.4561
Loss?
5.2793
Best Loss?
-34.7% from start
5.6737
Val Loss?
best: 5.6737
2.93e-5
Learning Rate?
1,598
Throughput?
tok/s (avg)
2,565
Speed?
ms/iter (avg)
104.261
Grad Norm?
avg: 121.212
21.70M
Tokens
processed
1768ms
Forward
69% of step
705ms
Backward
27% of step
45ms
GPU Sync
2% of step
1,224
GPU Ops
per step
0.3%
MFU
model FLOPS util
0.4x
Bwd/Fwd
ratio
Loss Curve ? click any chart to add markers
?
?
?
?
Architecture
Layers?6
Embedding?256
Heads?8
Vocab?4,000
Context?256
Dropout?0.1
Parameters?9.99M
Training Config
Total iters?50,000
Batch size?16
Max LR?0.00003
Optimizer?adamw
Backend?helios
Tokenizer?bpe-4k
Seed?42
Weight decay?0.1
Grad clip?1
Eval interval?2500
Throughput (tok/s)
Step Time (ms/iter)
GPU & VRAM
Perplexity
Train/Val Gap
Learning Rate
Grad Norm
Smoothed Loss (EMA)
Loss Velocity
Gradient Clipping
No clipping data
GPU Operations
Step Time Breakdown
Forward
Backward
Grad Norm
Optimizer
GPU Sync
Data
Timing Phase Lines
Backward / Forward Ratio
Sample Generations (3)
#CheckpointPrompt (preview)Generated
1-<|user|> Hello, how are you? <|assistant|>48d ago
Prompt
<|user|> Hello, how are you? <|assistant|>
Output
<|user|> Hello, how are you? <|assistant|>s the iting may es in s ed may s of f—that the fit ; s unntand ay, the eing the that in the dcinules the the ny, —s unawars for , of
2-<|user|> What do you like to do for fun? <|assistant|>48d ago
Prompt
<|user|> What do you like to do for fun? <|assistant|>
Output
<|user|> What do you like to do for fun? <|assistant|>in its s our s of our esmay ing your to when dits of s in or its of its is c; in or your a to by in the in to wtiand to s t, that is y, it rein st, through et
3-<|user|> Tell me about yourself. <|assistant|>48d ago
Prompt
<|user|> Tell me about yourself. <|assistant|>
Output
<|user|> Tell me about yourself. <|assistant|>our the and ythe is our is that e, or may to our and the may without our tent the true for uning , es it tin your and to a my tour a ing in aenres and the eis f
Model Config (JSON)
{
"vocabSize": 4000,
"blockSize": 256,
"nLayer": 6,
"nEmbd": 256,
"nHead": 8,
"dropout": 0.1
}Training Config (JSON)
{
"iters": 50000,
"batchSize": 16,
"lr": 0.00003,
"lrMin": 0.000003,
"warmupIters": 500,
"beta1": 0.9,
"beta2": 0.99,
"eps": 1e-8,
"weightDecay": 0.1,
"gradClip": 1,
"evalInterval": 2500,
"evalIters": 10,
"seed": 42,
"backend": "helios",
"tokenizer": "bpe-4k",
"optimizer": "adamw",
"logLevel": "info",
"trace": false,
"gradAccumSteps": 1,
"sampleInterval": 500,
"spikeThreshold": 0
}