| Log | ||
|---|---|---|
| Branch | feat/nnx-post-train-fixes |
|
| Commit | 05aede568 |
|
| Date | 2026-04-16 23:30 | |
| Run ID | feat_nnx_post_train_fixes_20260416 |
|
| Hardware | V6e-32 TPU (32 devices/slice × 1 slice) | |
| Docker image | gcr.io/cienet-cmcs/wanglance-xpk-runner:latest |
|
| NNX flags | pure_nnx=True pure_nnx_decoder=True enable_nnx=True |
0/26 passed.
| Test | Result | Log |
|---|---|---|
01_base |
FAIL | log |
02_synthetic |
FAIL | log |
03_dropout |
FAIL | log |
04_int8 |
FAIL | log |
05_fp8 |
FAIL | log |
06_grad_accum |
FAIL | log |
07_eval |
FAIL | log |
08_checkpoint_async_true_save |
FAIL | log |
08_checkpoint_async_true_resume |
FAIL | log |
09_pdb_lt_1 |
FAIL | log |
10_shardy_true |
FAIL | log |
10_shardy_false |
FAIL | log |
11_optimizer_offload_false |
FAIL | log |
11_optimizer_offload_true |
FAIL | log |
13_scan_layers_false |
FAIL | log |
13_scan_layers_true |
FAIL | log |
14_async_ckpt_false_save |
FAIL | log |
14_async_ckpt_false_resume |
FAIL | log |
15_ocdbt_true_save |
FAIL | log |
15_ocdbt_true_resume |
FAIL | log |
15_ocdbt_false_save |
FAIL | log |
15_ocdbt_false_resume |
FAIL | log |
16_zarr3_true_save |
FAIL | log |
16_zarr3_true_resume |
FAIL | log |
16_zarr3_false_save |
FAIL | log |
16_zarr3_false_resume |
FAIL | log |
| Log | ||
|---|---|---|
| Logs | ~/maxtext/venv_runs/feat_nnx_post_train_fixes_20260416_221011/nnx/logs/ |
|
| GCS output (checkpoints + profiler) | gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260416_221011/ |