tacit/nanochat
1
0
Fork 0
mirror of https://github.com/karpathy/nanochat.git synced 2026-03-07 01:40:30 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge f8ff0439b9 into 1076f97059

Browse Source
This commit is contained in:
Sofie Van Landeghem 2026-03-06 11:04:02 +01:00 committed by GitHub
commit 14d50e4ef7
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@ -71,7 +71,7 @@ OMP_NUM_THREADS=1 torchrun --standalone --nproc_per_node=8 -m scripts.base_train
This uses wandb (run name "d12"), only runs the CORE metric on last step, and it doesn't sample and save intermediate checkpoints. I like to change something in the code, re-run a d12 (or a d16 etc) and see if it helped, in an iteration loop. To see if a run helps, I like to monitor the wandb plots for:
1. `val_bpb` (validation loss in vocab-size-invariant units of bits per byte) as a function of `step`, `total_training_time` and `total_training_flops`.
2. `core_metric` (the DCLM CORE socre)
2. `core_metric` (the DCLM CORE score)
3. VRAM utilization, `train/mfu` (Model FLOPS utilization), `train/tok_per_sec` (training throughput)
See an example [here](https://github.com/karpathy/nanochat/pull/498#issuecomment-3850720044).
@ -101,7 +101,7 @@ NANOCHAT_DTYPE=bfloat16 torchrun --nproc_per_node=8 -m scripts.base_train # for
How it works: model weights are stored in fp32 (for optimizer precision), but our custom `Linear` layer casts them to `COMPUTE_DTYPE` during the forward pass. Embeddings are stored directly in `COMPUTE_DTYPE` to save memory. This gives us the same mixed-precision benefit as autocast but with full explicit control over what runs in which precision.
Note: `float16` training automatically enables a `GradScaler` in `base_train.py` to prevent gradient underflow. SFT suppors this too but RL currently does not. Inference in fp16 works fine everywhere.
Note: `float16` training automatically enables a `GradScaler` in `base_train.py` to prevent gradient underflow. SFT supports this too but RL currently does not. Inference in fp16 works fine everywhere.
## Guides