CITADEL
StandbyNothing is live right now.
Eval accuracylast measured Mon 2 Nov, 18:52 UTC
0%25%50%75%100%0%25%50%75%100%0%25%50%75%100%

Claviger

finishedRadeon 8060S

A small model that reads one Discord message and decides whether an agent should reply, and which one. Fine-tuned from Laya on invented chat.

started
Mon 2 Nov, 18:46 UTC
ran for
06:17
config
batch 8, seed 0 of 3

run9/22step150/150epoch5/5config5/11seed0/3batch8modellaya 421Mrev7b928d8items479+130

Loss train 0.0000 ▼ −0.8166 eval 0.7312 lowest -0.3180 at step 45 run before, run 8 ▲ +0.0774 step 150 Laya's objective: soft cross-entropy plus a policy-gradient term, so it can go below zero.
Timebase
Range
Smoothing
Reference
30 readings, 2 min/div2 min/div1 min/div
-1012300:0002:0004:0006:00phase starting, 0.0 s to 35.5 sphase training, 35.5 s to 06:16phase eval, 06:16 to 06:17epoch 2, 02:34epoch 3, 03:40epoch 4, 04:45epoch 5, 05:50trainingE2E3E4E50.8166Lowest last train loss across this project's 22 runs, run 19lowest last -0.0937, run 190.0000-1012300:0002:0004:0006:00phase starting, 0.0 s to 35.5 sphase training, 35.5 s to 06:16phase eval, 06:16 to 06:17epoch 1, 89.1 sepoch 2, 02:34epoch 3, 03:40epoch 4, 04:45epoch 5, 05:50trainingepoch 1epoch 2epoch 3epoch 4E50.8166Lowest last train loss across this project's 22 runs, run 19lowest last -0.0937, run 19eval 0.73120.0000▼ −0.8166-1012300:0001:0002:0003:0004:0005:0006:00phase starting, 0.0 s to 35.5 sphase training, 35.5 s to 06:16phase eval, 06:16 to 06:17epoch 1, 89.1 sepoch 2, 02:34epoch 3, 03:40epoch 4, 04:45epoch 5, 05:50startingtrainingepoch 1epoch 2epoch 3epoch 4epoch 50.8166Lowest last train loss across this project's 22 runs, run 19lowest last -0.0937, run 19eval 0.73120.0000▼ −0.8166

Eval measures

Eval losslower is better
0.7312
02 →
ECEcalibration error, lower is better
0.1546
01
Brierlower is better
0.3466
02 →

Elapsed

65.3 s

epoch interval

150 of 150 steps, 5 of 5 epochs, 100%

2.10 s per step, 28.6 steps a minute

Against the other runs

21 of 22

by last eval accuracy

best so far 0.9231, run 19

▼ −0.0231 against run 8

Train minus eval

−0.7312

step 150

train 0.0000, eval 0.7312

Set-up

479 train, 130 eval

synthetic, batch size 8, grad accum 2, epochs 5, loss rlcd, seed 0

Seed spread, observed

0.0462
0.750.85

3 finished seeds

Samples per second

4.16

▼ −3.54 since first

Learning rate

1.00e-1

▲ +1.00e-1 since first

GPU memory, this run

4.9 GB

On one axis

phase starting, 0.0 s to 35.5 sphase training, 35.5 s to 06:16phase eval, 06:16 to 06:17E2E3E4E5traininglowest last -0.0937Train loss0.0000▼ −0.81662.00-1.000.81660.0000last ▼ −0.8166Samples per second4.16▼ −3.54847.70last 4.16Learning rate1.00e-1▲ +1.00e-11.0e-10.0e+02.50e-5last 1.00e-1Power, whole chip, CPU and GPU83 W▲ +46 W100 W20 W37 W83 Wlast ▲ +46 WHottest, one board zone85 °C▲ +16 °C90 °C60 °C69 °Clast 85 °C00:0001:0002:0003:0004:0005:0006:00
GPUpower: whole chip, CPU and GPUtemperature: one board zone

GPU 0

Wwhole chip, CPU and GPU

83 W whole chip, CPU and GPU

use42%
memory, whole GPU13.3 GB

this run's peak 8.0 GB

85 °C one board zone

history 0 to 85 W

Input

  • Runs128 runs
  • Agentsnone showing
  • Treesnone showing
  • Projectsnone showing

Check

Every entry is checked on its way out; some are held for review.

Not shown on any run: data samples, predictions, raw logs, file paths, machine names

Delay

2 min, set

Watching the agents work, about 2 min behind.

One way

Nothing here reaches the agents.

This page

This page updates by itself.

visits counted with Google Analytics · choose

Runsheight is eval accuracyall recordings ▸ | finished× failed■ stopped○ lost● live
0.950.750.950.75finished, eval accuracy 0.8077, config 1 of 11 (batch 8), seed 2, Mon 2 Nov, 13:19 UTCfinished, eval accuracy 0.8231, config 2 of 11 (batch 8), seed 2 of 3, Mon 2 Nov, 14:03 UTCfinished, eval accuracy 0.8231, config 3 of 11 (batch 8), seed 2, Mon 2 Nov, 14:09 UTCfinished, eval accuracy 0.7846, config 2 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 14:16 UTCfinished, eval accuracy 0.8385, config 2 of 11 (batch 8), seed 1 of 3, Mon 2 Nov, 14:22 UTCfinished, eval accuracy 0.7923, config 4 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 18:15 UTCfinished, eval accuracy 0.8000, config 4 of 11 (batch 8), seed 1 of 3, Mon 2 Nov, 18:22 UTCfinished, eval accuracy 0.8154, config 4 of 11 (batch 8), seed 2 of 3, Mon 2 Nov, 18:28 UTCfinished, eval accuracy 0.7923, config 5 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 18:46 UTCfinished, eval accuracy 0.8385, config 5 of 11 (batch 8), seed 1 of 3, Mon 2 Nov, 18:52 UTCfinished, eval accuracy 0.8000, config 5 of 11 (batch 8), seed 2 of 3, Mon 2 Nov, 18:59 UTCfinished, eval accuracy 0.8231, config 6 of 11 (batch 8), seed 0 of 3, Tue 3 Nov, 00:59 UTCfinished, eval accuracy 0.8231, config 6 of 11 (batch 8), seed 1 of 3, Tue 3 Nov, 01:13 UTCfinished, eval accuracy 0.8000, config 6 of 11 (batch 8), seed 2 of 3, Tue 3 Nov, 01:26 UTCfinished at step 336 of 339, eval accuracy 0.8769, config 7 of 11 (batch 8), seed 0 of 3, Thu 19 Nov, 21:49 UTCfinished at step 336 of 339, eval accuracy 0.8846, config 7 of 11 (batch 8), seed 1 of 3, Thu 19 Nov, 22:06 UTCfinished at step 336 of 339, eval accuracy 0.8846, config 7 of 11 (batch 8), seed 2 of 3, Thu 19 Nov, 22:21 UTCfinished at step 336 of 339, eval accuracy 0.8615, config 8 of 11 (learning rate 0.0001, batch 8), seed 0, Sat 21 Nov, 23:04 UTCfinished at step 336 of 339, eval accuracy 0.9231, config 9 of 11 (learning rate 0.00002, batch 8), seed 0, Sat 21 Nov, 23:17 UTCfinished at step 336 of 339, eval accuracy 0.8538, config 10 of 11 (learning rate 0.00005, batch 8), seed 0, Sat 21 Nov, 23:29 UTCfinished at step 336 of 339, eval accuracy 0.9231, config 11 of 11 (learning rate 0.00002, batch 8), seed 1 of 2, Sun 22 Nov, 03:49 UTCfinished at step 336 of 339, eval accuracy 0.8846, config 11 of 11 (learning rate 0.00002, batch 8), seed 2 of 2, Sun 22 Nov, 04:01 UTC0.950.75finished, eval accuracy 0.8077, config 1 of 11 (batch 8), seed 2, Mon 2 Nov, 13:19 UTCfinished, eval accuracy 0.8231, config 2 of 11 (batch 8), seed 2 of 3, Mon 2 Nov, 14:03 UTCfinished, eval accuracy 0.8231, config 3 of 11 (batch 8), seed 2, Mon 2 Nov, 14:09 UTCfinished, eval accuracy 0.7846, config 2 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 14:16 UTCfinished, eval accuracy 0.8385, config 2 of 11 (batch 8), seed 1 of 3, Mon 2 Nov, 14:22 UTCfinished, eval accuracy 0.7923, config 4 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 18:15 UTCfinished, eval accuracy 0.8000, config 4 of 11 (batch 8), seed 1 of 3, Mon 2 Nov, 18:22 UTCfinished, eval accuracy 0.8154, config 4 of 11 (batch 8), seed 2 of 3, Mon 2 Nov, 18:28 UTCfinished, eval accuracy 0.7923, config 5 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 18:46 UTCfinished, eval accuracy 0.8385, config 5 of 11 (batch 8), seed 1 of 3, Mon 2 Nov, 18:52 UTCfinished, eval accuracy 0.8000, config 5 of 11 (batch 8), seed 2 of 3, Mon 2 Nov, 18:59 UTCfinished, eval accuracy 0.8231, config 6 of 11 (batch 8), seed 0 of 3, Tue 3 Nov, 00:59 UTCfinished, eval accuracy 0.8231, config 6 of 11 (batch 8), seed 1 of 3, Tue 3 Nov, 01:13 UTCfinished, eval accuracy 0.8000, config 6 of 11 (batch 8), seed 2 of 3, Tue 3 Nov, 01:26 UTCfinished at step 336 of 339, eval accuracy 0.8769, config 7 of 11 (batch 8), seed 0 of 3, Thu 19 Nov, 21:49 UTCfinished at step 336 of 339, eval accuracy 0.8846, config 7 of 11 (batch 8), seed 1 of 3, Thu 19 Nov, 22:06 UTCfinished at step 336 of 339, eval accuracy 0.8846, config 7 of 11 (batch 8), seed 2 of 3, Thu 19 Nov, 22:21 UTCfinished at step 336 of 339, eval accuracy 0.8615, config 8 of 11 (learning rate 0.0001, batch 8), seed 0, Sat 21 Nov, 23:04 UTCfinished at step 336 of 339, eval accuracy 0.9231, config 9 of 11 (learning rate 0.00002, batch 8), seed 0, Sat 21 Nov, 23:17 UTCfinished at step 336 of 339, eval accuracy 0.8538, config 10 of 11 (learning rate 0.00005, batch 8), seed 0, Sat 21 Nov, 23:29 UTCfinished at step 336 of 339, eval accuracy 0.9231, config 11 of 11 (learning rate 0.00002, batch 8), seed 1 of 2, Sun 22 Nov, 03:49 UTCfinished at step 336 of 339, eval accuracy 0.8846, config 11 of 11 (learning rate 0.00002, batch 8), seed 2 of 2, Sun 22 Nov, 04:01 UTC

9 of 22 finished, 0.7923, config 5 of 11 (batch 8), seed 0 of 3, Mon 2 Nov, 18:46 UTC

Runs as a table
StartedStatusEval accuracyStepsConfig
Sun 22 Nov, 04:01 UTC| finished0.8846336 of 339learning rate 0.00002, batch 8, seed 2 of 2
Sun 22 Nov, 03:49 UTC| finished0.9231336 of 339learning rate 0.00002, batch 8, seed 1 of 2
Sat 21 Nov, 23:29 UTC| finished0.8538336 of 339learning rate 0.00005, batch 8, seed 0
Sat 21 Nov, 23:17 UTC| finished0.9231336 of 339learning rate 0.00002, batch 8, seed 0
Sat 21 Nov, 23:04 UTC| finished0.8615336 of 339learning rate 0.0001, batch 8, seed 0
Thu 19 Nov, 22:21 UTC| finished0.8846336 of 339batch 8, seed 2 of 3
Thu 19 Nov, 22:06 UTC| finished0.8846336 of 339batch 8, seed 1 of 3
Thu 19 Nov, 21:49 UTC| finished0.8769336 of 339batch 8, seed 0 of 3
Tue 3 Nov, 01:26 UTC| finished0.8000150 of 150batch 8, seed 2 of 3
Tue 3 Nov, 01:13 UTC| finished0.8231150 of 150batch 8, seed 1 of 3
Tue 3 Nov, 00:59 UTC| finished0.8231150 of 150batch 8, seed 0 of 3
Mon 2 Nov, 18:59 UTC| finished0.8000150 of 150batch 8, seed 2 of 3
Mon 2 Nov, 18:52 UTC| finished0.8385150 of 150batch 8, seed 1 of 3
Mon 2 Nov, 18:46 UTC| finished0.7923150 of 150batch 8, seed 0 of 3
Mon 2 Nov, 18:28 UTC| finished0.8154150 of 150batch 8, seed 2 of 3
Mon 2 Nov, 18:22 UTC| finished0.8000150 of 150batch 8, seed 1 of 3
Mon 2 Nov, 18:15 UTC| finished0.7923150 of 150batch 8, seed 0 of 3
Mon 2 Nov, 14:22 UTC| finished0.8385150 of 150batch 8, seed 1 of 3
Mon 2 Nov, 14:16 UTC| finished0.7846150 of 150batch 8, seed 0 of 3
Mon 2 Nov, 14:09 UTC| finished0.8231150 of 150batch 8, seed 2
Mon 2 Nov, 14:03 UTC| finished0.8231150 of 150batch 8, seed 2 of 3
Mon 2 Nov, 13:19 UTC| finished0.8077150 of 150batch 8, seed 2

Plate

Model
convaiinnovations/laya, 421M parameters revision 7b928d8
Data
Synthetic Discord triage set, synthetic479 train130 eval
Config
batch size 8, grad accum 2, epochs 5, loss rlcd, seed 0
Hardware
Radeon 8060S, ROCm

Log

Since startMark
0.0 sphase starting
3.5 swent public
35.5 sphase training
89.1 sepoch 1
154.4 sepoch 2
220.4 sepoch 3
285.4 sepoch 4
350.2 sepoch 5
376.2 sphase eval
376.2 seval 1

Seeds

SeedStatusEval accuracy
seed 0| finished0.7923
seed 1| finished0.8385
seed 2| finished0.8000

observed spread 0.0462, 3 finished seeds

Readings

31 rows as a table
StepSecondsTrain lossEval lossEval accuracyECEBrierSamples per secondLearning rateGPU GB, this run
035.3-----7.7012.50e-56.5
546.00.8166----7.5042.48e-56.5
1056.6-0.0779----7.5432.45e-56.5
1567.30.3424----7.4572.41e-56.5
2078.30.2846----7.2822.35e-56.5
2588.90.5621----7.5422.29e-56.5
30101.00.6641----6.6112.21e-56.5
35111.60.9556----7.5512.12e-56.5
40122.20.1753----7.5782.03e-56.5
45132.6-0.3180----7.6511.92e-56.5
50143.6-0.0918----7.3011.81e-56.5
55154.20.6079----7.5271.69e-56.5
60166.30.5608----6.6341.57e-56.5
65176.90.4814----7.5211.45e-56.5
70188.10.4175----7.1871.33e-56.5
75198.90.3959----7.3431.20e-56.5
80209.80.3186----7.3621.08e-56.5
85220.2-0.1879----7.6729.53e-66.5
90232.20.2715----6.6968.35e-66.5
95242.6-0.0990----7.6607.22e-66.5
100253.00.0000----7.7006.15e-66.5
105263.90.0447----7.3315.16e-66.5
110274.50.0000----7.5474.25e-66.5
115285.21.0784----7.4723.44e-66.5
120297.20.0000----6.6742.74e-66.5
125307.80.0000----7.5602.14e-66.5
130318.30.0000----7.6611.67e-66.5
135328.80.0000----7.5731.32e-66.5
140339.5-0.0017----7.4911.09e-66.5
145339.70.0000-------
150377.50.00000.73120.79230.15460.34664.1561.00e-14.9

GPU readings

31 rows as a table
StepGPUUse %Power W, whole chip, CPU and GPUMemory GB, whole GPUTemp C, one board zone
00037.34.368.8
507081.213.275.8
2004584.113.268.8
300483.913.278.8
5006981.313.372.8
6002983.713.380.8
75010083.913.373.8
900383.413.381.8
1050282.113.376.8
12007685.413.382.8
13506383.313.377.8
15004283.213.384.8

Worked out

ReadoutValueWorked out from
Run complete100%150 steps over 150
Step time2.10 s304.3 s over 145 steps, between readings. Startup and intervals that hold an eval are left out
Steps per minute28.660 s over the step time of 2.10 s
Epoch interval65.3 sMean of 4 gaps between epoch marks
Train loss since first reading−0.81660.0000 at step 150 minus 0.8166 at step 5
Lowest train loss-0.3180Lowest of 30 readings, at step 45
Eval accuracy, as shown79.2%0.7923 times 100, to one decimal. Changes on that plate are percentage points
Train minus eval loss−0.73120.0000 minus 0.7312, both at step 150
Wrong answers27130 eval items minus 103 correct. 103 over 130 is the only count that gives 0.7923
Rank21 of 22By last eval accuracy, across this project's 22 runs with an eval
Best so far0.9231Highest last eval accuracy, run 19
Against the run before−0.02310.7923 minus 0.8154 on run 8
Lowest last train loss-0.0937Lowest last train loss across this project's 22 runs, run 19
Seed spread, observed0.0462High 0.8385 minus low 0.7923 across 3 finished seeds
Train loss against run 8+0.07740.0000 minus -0.0774 on run 8, both at step 150
Samples per second since first reading−3.544.16 minus 7.70
Learning rate since first reading+1.00e-11.00e-1 minus 2.50e-5
This run's peak memory per GPU8.0 GBAs the run reports it: its own memory, not the whole GPU's