Datasets:

ArXiv:
a100_20260502 / swift /ui /llm_train /runtime.py
shulin16's picture
Add files using upload-large-folder tool
e93ef15 verified
Raw
History Blame Contribute Delete
25.5 kB
# Copyright (c) ModelScope Contributors. All rights reserved.
import collections
import gradio as gr
import json
import matplotlib.pyplot as plt
import os
import psutil
import re
import subprocess
import sys
import time
import webbrowser
from datetime import datetime
from functools import partial
from packaging import version
from transformers import is_tensorboard_available
from typing import Dict, List, Tuple, Type
from swift.utils import TB_COLOR, TB_COLOR_SMOOTH, format_time, get_logger, read_tensorboard_file, tensorboard_smoothing
from ..base import BaseUI
from .utils import close_loop, run_command_in_subprocess
logger = get_logger()
class Runtime(BaseUI):
handlers: Dict[str, Tuple[List, Tuple]] = {}
group = 'llm_train'
all_plots = None
log_event = {}
sft_plot = [
{
'name': 'train/loss',
'smooth': 0.9,
},
{
'name': 'train/acc',
'smooth': None,
},
{
'name': 'train/learning_rate',
'smooth': None,
},
{
'name': 'eval/loss',
'smooth': 0.9,
},
{
'name': 'eval/acc',
'smooth': None,
},
]
dpo_plot = [
{
'name': 'train/loss',
'smooth': 0.9,
},
{
'name': 'train/rewards/accuracies',
'smooth': None,
},
{
'name': 'train/rewards/margins',
'smooth': 0.9,
},
{
'name': 'train/logps/chosen',
'smooth': 0.9,
},
{
'name': 'train/logps/rejected',
'smooth': 0.9,
},
]
kto_plot = [
{
'name': 'kl',
'smooth': None,
},
{
'name': 'rewards/chosen_sum',
'smooth': 0.9,
},
{
'name': 'logps/chosen_sum',
'smooth': 0.9,
},
{
'name': 'rewards/rejected_sum',
'smooth': 0.9,
},
{
'name': 'logps/rejected_sum',
'smooth': 0.9,
},
]
orpo_plot = [
{
'name': 'train/loss',
'smooth': 0.9,
},
{
'name': 'train/rewards/accuracies',
'smooth': None,
},
{
'name': 'train/rewards/margins',
'smooth': 0.9,
},
{
'name': 'train/rewards/chosen',
'smooth': 0.9,
},
{
'name': 'train/log_odds_ratio',
'smooth': 0.9,
},
]
grpo_plot = [
{
'name': 'train/loss',
'smooth': 0.9,
},
{
'name': 'train/reward',
'smooth': 0.9,
},
{
'name': 'train/learning_rate',
'smooth': None,
},
{
'name': 'train/completions/mean_length',
'smooth': 0.9,
},
{
'name': 'train/kl',
'smooth': 0.9,
},
]
locale_dict = {
'runtime_tab': {
'label': {
'zh': '运行时',
'en': 'Runtime'
},
},
'tb_not_found': {
'value': {
'zh': 'TensorBoard未安装,使用`pip install tensorboard`进行安装',
'en': 'TensorBoard not found, install it by `pip install tensorboard`',
}
},
'running_cmd': {
'label': {
'zh': '运行命令',
'en': 'Command line'
},
'info': {
'zh': '执行的实际命令',
'en': 'The actual command'
}
},
'show_running_cmd': {
'value': {
'zh': '展示运行命令',
'en': 'Show running command line'
},
},
'show_sh': {
'label': {
'zh': '展示sh命令行',
'en': 'Show sh command line'
},
},
'cmd_sh': {
'label': {
'zh': '训练命令行',
'en': 'Training command line'
},
'info': {
'zh':
'如果训练命令行没有展示请再次点击"展示运行命令",点击下方的"保存训练命令"可以保存sh脚本',
'en': ('Please press "Show running command line" if the content is none, '
'click the "Save training command" below to save the sh script')
}
},
'save_cmd_as_sh': {
'value': {
'zh': '保存训练命令',
'en': 'Save training command'
}
},
'save_cmd_alert': {
'value': {
'zh': '训练命令行将被保存在:{}',
'en': 'The training command line will be saved in: {}'
}
},
'close_cmd_show': {
'value': {
'zh': '关闭训练命令展示',
'en': 'Close training command show'
}
},
'show_log': {
'value': {
'zh': '展示运行状态',
'en': 'Show running status'
},
},
'stop_show_log': {
'value': {
'zh': '停止展示运行状态',
'en': 'Stop showing running status'
},
},
'logging_dir': {
'label': {
'zh': '日志路径',
'en': 'Logging dir'
},
'info': {
'zh': '支持手动传入文件路径',
'en': 'Support fill custom path in'
}
},
'log': {
'label': {
'zh': '日志输出',
'en': 'Logging content'
},
'info': {
'zh': '如果日志无更新请再次点击"展示运行状态"',
'en': 'Please press "Show running status" if the log content is not updating'
}
},
'running_tasks': {
'label': {
'zh': '运行中任务',
'en': 'Running Tasks'
},
'info': {
'zh': '运行中的任务(所有的swift sft/pt命令)',
'en': 'All running tasks(started by swift sft/pt)'
}
},
'refresh_tasks': {
'value': {
'zh': '找回运行时任务',
'en': 'Find running tasks'
},
},
'kill_task': {
'value': {
'zh': '杀死任务',
'en': 'Kill running task'
},
},
'tb_url': {
'label': {
'zh': 'Tensorboard链接',
'en': 'Tensorboard URL'
},
'info': {
'zh': '仅展示,不可编辑',
'en': 'Not editable'
}
},
'start_tb': {
'value': {
'zh': '打开TensorBoard',
'en': 'Start TensorBoard'
},
},
'close_tb': {
'value': {
'zh': '关闭TensorBoard',
'en': 'Close TensorBoard'
},
},
}
@classmethod
def do_build_ui(cls, base_tab: Type['BaseUI']):
with gr.Accordion(elem_id='runtime_tab', open=False):
with gr.Blocks():
with gr.Row():
with gr.Column(scale=3):
with gr.Row():
gr.Textbox(elem_id='running_cmd', lines=1, scale=3, interactive=False, max_lines=1)
gr.Textbox(elem_id='logging_dir', lines=1, scale=3, max_lines=1)
with gr.Row():
gr.Button(elem_id='show_running_cmd', scale=2, variant='primary')
gr.Button(elem_id='show_log', scale=2, variant='primary')
gr.Button(elem_id='stop_show_log', scale=2)
with gr.Column(scale=2):
with gr.Row():
gr.Textbox(elem_id='tb_url', lines=1, scale=4, interactive=False, max_lines=1)
with gr.Row():
gr.Button(elem_id='start_tb', scale=2, variant='primary')
gr.Button(elem_id='close_tb', scale=2)
with gr.Accordion(elem_id='show_sh', open=False, visible=False):
with gr.Blocks():
gr.Textbox(elem_id='cmd_sh', lines=8)
with gr.Row(equal_height=True):
gr.Button(elem_id='save_cmd_as_sh', variant='primary', scale=2)
gr.Button(elem_id='close_cmd_show', scale=2)
with gr.Row():
gr.Textbox(elem_id='log', lines=6, visible=False)
with gr.Row(equal_height=True):
gr.Dropdown(elem_id='running_tasks', scale=10)
gr.Button(elem_id='refresh_tasks', scale=1)
gr.Button(elem_id='kill_task', scale=1)
with gr.Row():
cls.all_plots = []
plot = Runtime.sft_plot
if base_tab.group == 'llm_rlhf':
plot = Runtime.dpo_plot
elif base_tab.group == 'llm_grpo':
plot = Runtime.grpo_plot
for idx, k in enumerate(plot):
name = k['name']
cls.all_plots.append(gr.Plot(elem_id=str(idx), label=name))
concurrency_limit = {}
if version.parse(gr.__version__) >= version.parse('4.0.0'):
concurrency_limit = {'concurrency_limit': 5}
base_tab.element('show_log').click(
Runtime.update_log, [base_tab.element('running_tasks')], [cls.element('log')] + cls.all_plots).then(
Runtime.wait, [base_tab.element('logging_dir'),
base_tab.element('running_tasks')], [cls.element('log')] + cls.all_plots,
**concurrency_limit)
base_tab.element('stop_show_log').click(cls.break_log_event, [cls.element('running_tasks')], [])
base_tab.element('start_tb').click(
Runtime.start_tb,
[base_tab.element('logging_dir')],
[base_tab.element('tb_url')],
)
base_tab.element('close_tb').click(
Runtime.close_tb,
[base_tab.element('logging_dir')],
[],
)
base_tab.element('refresh_tasks').click(
partial(Runtime.refresh_tasks, group=cls.group),
[base_tab.element('running_tasks')],
[base_tab.element('running_tasks')],
)
@classmethod
def after_build_ui(cls, base_tab: Type['BaseUI']):
cls.element('show_running_cmd').click(Runtime.show_train_sh, cls.element('running_cmd'),
[cls.element('show_sh')] + [cls.element('cmd_sh')])
cls.element('save_cmd_as_sh').click(cls.save_cmd, cls.element('running_cmd'), [])
cls.element('close_cmd_show').click(Runtime.close_cmd_show, [], [cls.element('show_sh')])
@classmethod
def get_plot(cls, task):
if not task or 'swift sft' in task or 'swift pt' in task:
return cls.sft_plot
args: dict = cls.parse_info_from_cmdline(task)[1]
rlhf_type = args.get('rlhf_type', 'dpo')
if rlhf_type in ('dpo', 'cpo', 'simpo'):
return cls.dpo_plot
elif rlhf_type == 'kto':
return cls.kto_plot
elif rlhf_type == 'orpo':
return cls.orpo_plot
elif rlhf_type == 'grpo':
return cls.grpo_plot
@classmethod
def update_log(cls, task):
ret = [gr.update(visible=True)]
plot = Runtime.get_plot(task)
for i in range(len(plot)):
p = plot[i]
ret.append(gr.update(visible=True, label=p['name']))
return ret
@classmethod
def get_initial(cls, line):
tqdm_starts = ['Train:', 'Map:', 'Val:', 'Filter:']
for start in tqdm_starts:
if line.startswith(start):
return start
return None
@classmethod
def wait(cls, logging_dir, task):
if not logging_dir:
return [None] + Runtime.plot(task)
log_file = os.path.join(logging_dir, 'run.log')
cls.log_event[logging_dir] = False
offset = 0
latest_data = ''
lines = collections.deque(maxlen=int(os.environ.get('MAX_LOG_LINES', 100)))
try:
with open(log_file, 'r', encoding='utf-8') as input:
input.seek(offset)
fail_cnt = 0
while True:
try:
latest_data += input.read()
except UnicodeDecodeError:
continue
if not latest_data:
time.sleep(0.5)
fail_cnt += 1
if fail_cnt > 50:
break
if cls.log_event.get(logging_dir, False):
cls.log_event[logging_dir] = False
break
if '\n' not in latest_data:
continue
latest_lines = latest_data.split('\n')
if latest_data[-1] != '\n':
latest_data = latest_lines[-1]
latest_lines = latest_lines[:-1]
else:
latest_data = ''
lines.extend(latest_lines)
start = cls.get_initial(lines[-1])
if start:
i = len(lines) - 2
while i >= 0:
if lines[i].startswith(start):
del lines[i]
i -= 1
else:
break
yield [gr.update(value='\n'.join(lines))] + Runtime.plot(task)
time.sleep(0.5)
except IOError:
pass
@classmethod
def break_log_event(cls, task):
if not task:
return
pid, all_args = Runtime.parse_info_from_cmdline(task)
cls.log_event[all_args['logging_dir']] = True
@classmethod
def show_log(cls, logging_dir):
webbrowser.open('file://' + os.path.join(logging_dir, 'run.log'), new=2)
@classmethod
def start_tb(cls, logging_dir):
if not is_tensorboard_available():
gr.Error(cls.locale('tb_not_found', cls.lang)['value'])
return ''
logging_dir = logging_dir.strip()
logging_dir = logging_dir if not logging_dir.endswith(os.sep) else logging_dir[:-1]
if logging_dir in cls.handlers:
return cls.handlers[logging_dir][1]
handler, lines = run_command_in_subprocess('tensorboard', '--logdir', logging_dir, timeout=2)
localhost_addr = ''
for line in lines:
if 'http://localhost:' in line:
line = line[line.index('http://localhost:'):]
localhost_addr = line[:line.index(' ')]
cls.handlers[logging_dir] = (handler, localhost_addr)
logger.info('===========Tensorboard Log============')
logger.info('\n'.join(lines))
webbrowser.open(localhost_addr, new=2)
return localhost_addr
@staticmethod
def close_tb(logging_dir):
if logging_dir in Runtime.handlers:
close_loop(Runtime.handlers[logging_dir][0])
Runtime.handlers.pop(logging_dir)
@staticmethod
def refresh_tasks(running_task=None, group=None):
output_dir = running_task if not running_task or 'pid:' not in running_task else None
process_name = 'swift'
negative_names = ['swift.exe', 'swift-script.py']
cmd_name = ['pt', 'sft'] if group == 'llm_train' else ['rlhf']
process = []
selected = None
for proc in psutil.process_iter():
try:
cmdlines = proc.cmdline()
except (psutil.ZombieProcess, psutil.AccessDenied, psutil.NoSuchProcess):
cmdlines = []
if any([
process_name in cmdline for cmdline in cmdlines # noqa
]) and not any([ # noqa
negative_name in cmdline for negative_name in negative_names # noqa
for cmdline in cmdlines # noqa
]) and any([cmdline in cmd_name for cmdline in cmdlines]): # noqa
if any([group == 'llm_rlhf' and 'grpo' in cmdline for cmdline in cmdlines]):
continue
if group == 'llm_grpo' and all(['grpo' not in cmdline for cmdline in cmdlines]):
continue
process.append(Runtime.construct_running_task(proc))
if output_dir is not None and any( # noqa
[output_dir == cmdline for cmdline in cmdlines]): # noqa
selected = Runtime.construct_running_task(proc)
if not selected:
if running_task and running_task in process:
selected = running_task
if not selected and process:
selected = process[0]
return gr.update(choices=process, value=selected)
@staticmethod
def construct_running_task(proc):
pid = proc.pid
ts = time.time()
create_time = proc.create_time()
create_time_formatted = datetime.fromtimestamp(create_time).strftime('%Y-%m-%d, %H:%M')
return f'pid:{pid}/create:{create_time_formatted}' \
f'/running:{format_time(ts - create_time)}/cmd:{" ".join(proc.cmdline())}'
@staticmethod
def parse_info_from_cmdline(task):
pid = None
if '/cmd:' in task:
for i in range(3):
slash = task.find('/')
if i == 0:
pid = task[:slash].split(':')[1]
task = task[slash + 1:]
if 'swift sft' in task:
args = task.split('swift sft')[1]
elif 'swift pt' in task:
args = task.split('swift pt')[1]
elif 'swift rlhf' in task:
args = task.split('swift rlhf')[1]
else:
raise ValueError(f'Cannot parse cmd line: {task}')
args = [arg.strip() for arg in args.split('--') if arg.strip()]
all_args = {}
for i in range(len(args)):
space = args[i].find(' ')
splits = args[i][:space], args[i][space + 1:]
all_args[splits[0]] = str(splits[1]) if isinstance(splits[1], int) else splits[1]
output_dir = all_args['output_dir']
if os.path.exists(os.path.join(output_dir, 'args.json')):
with open(os.path.join(output_dir, 'args.json'), 'r', encoding='utf-8') as f:
_json = json.load(f)
for key in all_args.keys():
all_args[key] = str(_json.get(key)) if isinstance(_json.get(key), int) else _json.get(key)
if isinstance(all_args[key], list):
if any([' ' in value for value in all_args[key] if isinstance(value, str)]):
all_args[key] = [f'"{value}"' for value in all_args[key]]
if len(all_args[key]) > 0 and isinstance(all_args[key][0], str):
all_args[key] = ' '.join(all_args[key])
return pid, all_args
@staticmethod
def kill_task(task):
if task:
pid, all_args = Runtime.parse_info_from_cmdline(task)
output_dir = all_args['output_dir']
if sys.platform == 'win32':
command = ['taskkill', '/f', '/t', '/pid', pid]
else:
command = ['pkill', '-9', '-f', output_dir]
try:
result = subprocess.run(command, capture_output=True, text=True)
assert result.returncode == 0
except Exception as e:
raise e
Runtime.break_log_event(task)
return [Runtime.refresh_tasks()] + [gr.update(value=None)] * (len(Runtime.get_plot(task)) + 1)
@staticmethod
def reset():
return None, 'output'
@staticmethod
def task_changed(task, base_tab):
if task:
_, all_args = Runtime.parse_info_from_cmdline(task)
else:
all_args = {}
elements = list(base_tab.valid_elements().values())
ret = []
for e in elements:
if e.elem_id in all_args:
if isinstance(e, gr.Dropdown) and e.multiselect:
arg = all_args[e.elem_id].split(' ')
else:
arg = all_args[e.elem_id]
if isinstance(e, gr.Slider) and isinstance(arg, str) and re.fullmatch(base_tab.int_regex, arg):
arg = int(arg)
elif isinstance(e, gr.Slider) and isinstance(arg, str) and re.fullmatch(base_tab.float_regex, arg):
arg = float(arg)
elif isinstance(e, gr.Checkbox) and isinstance(arg, str) and re.fullmatch(base_tab.bool_regex, arg):
arg = True if arg.lower() == 'true' else False
ret.append(gr.update(value=arg))
else:
ret.append(gr.update())
Runtime.break_log_event(task)
return ret + [gr.update(value=None)] * (len(Runtime.get_plot(task)) + 1)
@staticmethod
def plot(task):
plot = Runtime.get_plot(task)
if not task:
return [None] * len(plot)
_, all_args = Runtime.parse_info_from_cmdline(task)
tb_dir = all_args['logging_dir']
if not os.path.exists(tb_dir):
return [None] * len(plot)
fname = [
fname for fname in os.listdir(tb_dir)
if os.path.isfile(os.path.join(tb_dir, fname)) and fname.startswith('events.out')
]
if fname:
fname = fname[0]
else:
return [None] * len(plot)
tb_path = os.path.join(tb_dir, fname)
data = read_tensorboard_file(tb_path)
plots = []
for k in plot:
name = k['name']
smooth = k['smooth']
if name == 'train/acc':
if 'train/token_acc' in data:
name = 'train/token_acc'
if 'train/seq_acc' in data:
name = 'train/seq_acc'
if name == 'eval/acc':
if 'eval/token_acc' in data:
name = 'eval/token_acc'
if 'eval/seq_acc' in data:
name = 'eval/seq_acc'
if name not in data:
plots.append(None)
continue
_data = data[name]
steps = [d['step'] for d in _data]
values = [d['value'] for d in _data]
if len(values) == 0:
continue
plt.close('all')
fig = plt.figure()
ax = fig.add_subplot()
# _, ax = plt.subplots(1, 1, squeeze=True, figsize=(8, 5), dpi=100)
ax.set_title(name)
if len(values) == 1:
ax.scatter(steps, values, color=TB_COLOR_SMOOTH)
elif smooth is not None:
ax.plot(steps, values, color=TB_COLOR)
values_s = tensorboard_smoothing(values, smooth)
ax.plot(steps, values_s, color=TB_COLOR_SMOOTH)
else:
ax.plot(steps, values, color=TB_COLOR_SMOOTH)
plots.append(fig)
return plots
@classmethod
def save_cmd(cls, cmd):
if len(cmd) > 0:
cmd_sh, output_dir = Runtime.cmd_to_sh_format(cmd)
os.makedirs(output_dir, exist_ok=True)
sh_file_path = os.path.join(output_dir, 'train.sh')
gr.Info(cls.locale('save_cmd_alert', cls.lang)['value'].format(sh_file_path))
with open(sh_file_path, 'w', encoding='utf-8') as f:
f.write(cmd_sh)
@staticmethod
def show_train_sh(cmd):
if len(cmd) == 0:
return gr.update(visible=False, open=False), None
cmd_sh, _ = Runtime.cmd_to_sh_format(cmd)
return gr.update(visible=True, open=True), cmd_sh
@staticmethod
def cmd_to_sh_format(cmd):
cmd_sh = ''
params = cmd.split('--')
env_params = params[0].split('nohup')[0].strip()
cmd_sh += (env_params + ' \\\n')
swift_cmd = params[0].split('nohup')[1].strip()
cmd_sh += ('nohup ' + swift_cmd + ' \\\n')
for param in params[1:]:
if param.startswith('output_dir'):
output_dir = param.split(' ')[1].strip()
cmd_sh += ('--' + param.strip() + ' \\\n')
return cmd_sh, output_dir
@staticmethod
def close_cmd_show():
return gr.update(visible=False)