d8-1764iter (~20min, fast baseline)

Dryrun Job Failed
UUID:6f254604-75ee-477b-bef8-0c2ee7941ca1
Tao Linvia🐎 pentiumCreated 213 days agoUpdated 213 days ago

Result: DRYRUN FAILED (infrastructure)

MLflow server connection refused during final metrics logging (line 472). Training itself completed successfully.

Useful data from partial run

  • d8: MFU ~49%, step time ~600ms, 8 grad accum steps
  • val_bpb at step 20: 1.977
  • Failure was MLflow infrastructure issue, not code bug

d8 timing estimate

With MFU=49%, step_time=600ms:

  • 1764 iter × 600ms = 17.6 min training
  • With eval + preprocessing: ~26 min total
Dryrun JobFailed
2
Funding 
3
Funded Job 
4
Done 

Finance

Fundings at snapshot$0.00

Individual funding and transaction records are not included in the public archive.

Configuration

Repository
https://github.com/mlpatron/mlpatron-nanochat
Version
experiment/baseline-search → d7426893a51c9b71bf739e3ef94c13b456b43731
Image
ghcr.io/nblintao/mlpatron-nanochat:latest
→ ghcr.io/nblintao/mlpatron-nanochat@sha256:28c5544c206237e407241b0a72a8707e63e52750f487b751675b51ff1bce1520
Entrypoint
a100_preset
Resources
NVIDIA A100 40GB × 1·a2-highgpu-1g (12 vCPU / 85 GiB)
Parameters
depth=8·eval_every=100·max_seq_len=2048·num_iterations=1764(dryrun: 20)·num_data_shards=8·device_batch_size=16