Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
42 changes: 42 additions & 0 deletions configs/debug.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,42 @@

# — Model —
model_dim: 768
n_layer: 12
n_head: 6
sequence_length: 128

# — Batching & Training —
num_iterations: 3000
batch_size: 1024
device_batch_size: 256

# — Learning‐rate schedule —
warmup_ratio: 0.0
warmdown_ratio: 0.2

# — Validation & Checkpointing —
val_loss_every: 125
val_tokens: 10485760

# — Weights & Biases logging —
wandb_project_name: debug
wandb_job_name: null
no_wandb: false

# — Miscellaneous flags —
debug: false
no_compile: false

# — Distributed training —
fs_size: 4 # FSDP size

# — Optimizer & Hyperparameters —
optimizer: dion2
scalar_opt: lion
mu: 0.95
weight_decay: 0.01
ortho_fraction: 0.25
lr: 0.02

shard_independent: true
no_validation: false
Loading