tacit/nanochat
1
0
Fork 0
mirror of https://github.com/karpathy/nanochat.git synced 2026-06-19 12:39:10 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

fixing deleted text issue

Browse Source
This commit is contained in:
Sachin Agrawal 2025-11-03 22:09:10 +01:00
parent 827e608492
commit d1fc8c5d05
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
speedrun.sh
View File

@ -16,7 +16,7 @@ export NANOCHAT_BASE_DIR="$HOME/.cache/nanochat"
mkdir -p $NANOCHAT_BASE_DIR
# Number of processes per node for distributed training
NPROC_PER_NODE=8
NPROC_PER_NODE=4
# -----------------------------------------------------------------------------
# Python venv setup with uv
@ -96,6 +96,10 @@ torchrun --standalone --nproc_per_node=$NPROC_PER_NODE -m scripts.base_eval
# -----------------------------------------------------------------------------
# Midtraining (teach the model conversation special tokens, tool use, multiple choice)
# download 2.3MB of synthetic identity conversations to impart a personality to nanochat
# see dev/gen_sft_data.py for details on how this data was prepared and to get a sense of how you can easily tune it
curl -L -o $NANOCHAT_BASE_DIR/identity_conversations.jsonl https://karpathy-public.s3.us-west-2.amazonaws.com/identity_conversations.jsonl
# run midtraining and eval the model
torchrun --standalone --nproc_per_node=$NPROC_PER_NODE -m scripts.mid_train -- --run=$WANDB_RUN
torchrun --standalone --nproc_per_node=$NPROC_PER_NODE -m scripts.chat_eval -- -i mid