tacit/nanochat
1
0
Fork 0
mirror of https://github.com/karpathy/nanochat.git synced 2025-12-16 01:02:18 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

add exp_name as unique id

Browse Source
This commit is contained in:
Shizhe Diao 2025-10-18 07:29:34 -07:00
parent 0de778a75b
commit be1e6c3592
1 changed files with 14 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
nanochat/report.py
View File

@ -114,7 +114,7 @@ def estimate_cost(gpu_info, runtime_hours=None):
"estimated_total": hourly_rate * runtime_hours if runtime_hours else None "estimated_total": hourly_rate * runtime_hours if runtime_hours else None
} }
def generate_header(): def generate_header(exp_name=None):
"""Generate the header for a training report.""" """Generate the header for a training report."""
timestamp = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S") timestamp = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
@ -127,6 +127,8 @@ def generate_header():
Generated: {timestamp} Generated: {timestamp}
Experiment: {exp_name or '-'}
## Environment ## Environment
### Git Information ### Git Information
@ -161,7 +163,7 @@ Generated: {timestamp}
""" """
# bloat metrics: package all of the source code and assess its weight # bloat metrics: package all of the source code and assess its weight
packaged = run_command('files-to-prompt . -e py -e md -e rs -e html -e toml -e sh --ignore "*target*" --cxml') packaged = run_command('files-to-prompt . -e py -e md -e rs -e html -e toml -e sh --ignore "*target*" --cxml') or ""
num_chars = len(packaged) num_chars = len(packaged)
num_lines = len(packaged.split('\n')) num_lines = len(packaged.split('\n'))
num_files = len([x for x in packaged.split('\n') if x.startswith('<source>')]) num_files = len([x for x in packaged.split('\n') if x.startswith('<source>')])
@ -232,9 +234,10 @@ def extract_timestamp(content, prefix):
class Report: class Report:
"""Maintains a bunch of logs, generates a final markdown report.""" """Maintains a bunch of logs, generates a final markdown report."""
def __init__(self, report_dir): def __init__(self, report_dir, exp_name=None):
os.makedirs(report_dir, exist_ok=True) os.makedirs(report_dir, exist_ok=True)
self.report_dir = report_dir self.report_dir = report_dir
self.exp_name = exp_name
def log(self, section, data): def log(self, section, data):
"""Log a section of data to the report.""" """Log a section of data to the report."""
@ -371,7 +374,7 @@ class Report:
os.remove(report_file) os.remove(report_file)
# Generate and write the header section with start timestamp # Generate and write the header section with start timestamp
header_file = os.path.join(self.report_dir, "header.md") header_file = os.path.join(self.report_dir, "header.md")
header = generate_header() header = generate_header(self.exp_name)
start_time = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S") start_time = datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
with open(header_file, "w") as f: with open(header_file, "w") as f:
f.write(header) f.write(header)
@ -387,13 +390,14 @@ class DummyReport:
def reset(self, *args, **kwargs): def reset(self, *args, **kwargs):
pass pass
def get_report(): def get_report(exp_name=None):
# just for convenience, only rank 0 logs to report # just for convenience, only rank 0 logs to report
from nanochat.common import get_base_dir, get_dist_info from nanochat.common import get_base_dir, get_dist_info
ddp, ddp_rank, ddp_local_rank, ddp_world_size = get_dist_info() ddp, ddp_rank, ddp_local_rank, ddp_world_size = get_dist_info()
if ddp_rank == 0: if ddp_rank == 0:
report_dir = os.path.join(get_base_dir(), "report") base_report_dir = os.path.join(get_base_dir(), "report")
return Report(report_dir) report_dir = os.path.join(base_report_dir, exp_name) if exp_name else base_report_dir
return Report(report_dir, exp_name=exp_name)
else: else:
return DummyReport() return DummyReport()
@ -401,8 +405,9 @@ if __name__ == "__main__":
import argparse import argparse
parser = argparse.ArgumentParser(description="Generate or reset nanochat training reports.") parser = argparse.ArgumentParser(description="Generate or reset nanochat training reports.")
parser.add_argument("command", nargs="?", default="generate", choices=["generate", "reset"], help="Operation to perform (default: generate)") parser.add_argument("command", nargs="?", default="generate", choices=["generate", "reset"], help="Operation to perform (default: generate)")
parser.add_argument("--exp_name", type=str, default=None, help="Experiment name to namespace reports under get_base_dir()/report/<exp_name>")
args = parser.parse_args() args = parser.parse_args()
if args.command == "generate": if args.command == "generate":
get_report().generate() get_report(exp_name=args.exp_name).generate()
elif args.command == "reset": elif args.command == "reset":
get_report().reset() get_report(exp_name=args.exp_name).reset()