forked from story-boards-ai/runpod-worker-a1111
-
Notifications
You must be signed in to change notification settings - Fork 0
/
rp_handler.py
490 lines (384 loc) · 14.6 KB
/
rp_handler.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
import os
import time
import requests
import traceback
import runpod
from runpod.serverless.utils.rp_validator import validate
from runpod.serverless.modules.rp_logger import RunPodLogger
from requests.adapters import HTTPAdapter, Retry
from huggingface_hub import HfApi
from schemas.input import INPUT_SCHEMA
from schemas.api import API_SCHEMA
from schemas.img2img import IMG2IMG_SCHEMA
from schemas.txt2img import TXT2IMG_SCHEMA
from schemas.interrogate import INTERROGATE_SCHEMA
from schemas.sync import SYNC_SCHEMA
from schemas.download import DOWNLOAD_SCHEMA
BASE_URI = 'http://0.0.0.0:3000'
TIMEOUT = 600
POST_RETRIES = 3
session = requests.Session()
retries = Retry(total=10, backoff_factor=0.1, status_forcelist=[502, 503, 504])
session.mount('http://', HTTPAdapter(max_retries=retries))
logger = RunPodLogger()
# ---------------------------------------------------------------------------- #
# Utility Functions #
# ---------------------------------------------------------------------------- #
def wait_for_service(url):
retries = 0
while True:
try:
requests.get(url)
return
except requests.exceptions.RequestException:
retries += 1
# Only log every 15 retries so the logs don't get spammed
if retries % 15 == 0:
logger.info('Service not ready yet. Retrying...')
except Exception as err:
logger.error(f'Error: {err}')
time.sleep(0.2)
def send_get_request(endpoint):
return session.get(
url=f'{BASE_URI}/{endpoint}',
timeout=TIMEOUT
)
def send_post_request(endpoint, payload, job_id, retry=0):
response = session.post(
url=f'{BASE_URI}/{endpoint}',
json=payload,
timeout=TIMEOUT
)
# Retry the post request in case the model has not completed loading yet
if response.status_code == 404:
if retry < POST_RETRIES:
retry += 1
logger.warn(f'Received HTTP 404 from endpoint: {endpoint}, Retrying: {retry}', job_id)
time.sleep(0.2)
send_post_request(endpoint, payload, job_id, retry)
return response
def validate_input(job):
return validate(job['input'], INPUT_SCHEMA)
def validate_api(job):
api = job['input']['api']
api['endpoint'] = api['endpoint'].lstrip('/')
return validate(api, API_SCHEMA)
def validate_payload(job):
method = job['input']['api']['method']
endpoint = job['input']['api']['endpoint']
payload = job['input']['payload']
validated_input = payload
if endpoint == 'v1/sync':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, SYNC_SCHEMA)
elif endpoint == 'v1/download':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, DOWNLOAD_SCHEMA)
elif endpoint == 'sdapi/v1/txt2img':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, TXT2IMG_SCHEMA)
elif endpoint == 'sdapi/v1/img2img':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, IMG2IMG_SCHEMA)
elif endpoint == 'sdapi/v1/interrogate' and method == 'POST':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, INTERROGATE_SCHEMA)
return endpoint, job['input']['api']['method'], validated_input
def download(job):
source_url = job['input']['payload']['source_url']
download_path = job['input']['payload']['download_path']
process_id = os.getpid()
temp_path = f"{download_path}.{process_id}"
# Download the file and save it as a temporary file
with requests.get(source_url, stream=True) as r:
r.raise_for_status()
with open(temp_path, 'wb') as f:
for chunk in r.iter_content(chunk_size=8192):
f.write(chunk)
# Rename the temporary file to the actual file name
os.rename(temp_path, download_path)
logger.info(f'{source_url} successfully downloaded to {download_path}', job['id'])
return {
'msg': 'Download successful',
'source_url': source_url,
'download_path': download_path
}
def sync(job):
repo_id = job['input']['payload']['repo_id']
sync_path = job['input']['payload']['sync_path']
hf_token = job['input']['payload']['hf_token']
api = HfApi()
models = api.list_repo_files(
repo_id=repo_id,
token=hf_token
)
synced_count = 0
synced_files = []
for model in models:
folder = os.path.dirname(model)
dest_path = f'{sync_path}/{model}'
if folder and not os.path.exists(dest_path):
logger.info(f'Syncing {model} to {dest_path}', job['id'])
uri = api.hf_hub_download(
token=hf_token,
repo_id=repo_id,
filename=model,
local_dir=sync_path,
local_dir_use_symlinks=False
)
if uri:
synced_count += 1
synced_files.append(dest_path)
return {
'synced_count': synced_count,
'synced_files': synced_files
}
# ---------------------------------------------------------------------------- #
# RunPod Handler #
# ---------------------------------------------------------------------------- #
def handler(job):
validated_input = validate_input(job)
if 'errors' in validated_input:
return {
'error': '\n'.join(validated_input['errors'])
}
validated_api = validate_api(job)
if 'errors' in validated_api:
return {
'error': '\n'.join(validated_api['errors'])
}
endpoint, method, validated_payload = validate_payload(job)
if 'errors' in validated_payload:
return {
'error': '\n'.join(validated_payload['errors'])
}
if 'validated_input' in validated_payload:
payload = validated_payload['validated_input']
else:
payload = validated_payload
try:
logger.info(f'Sending {method} request to: /{endpoint}', job['id'])
if endpoint == 'v1/download':
return download(job)
elif endpoint == 'v1/sync':
return sync(job)
elif method == 'GET':
response = send_get_request(endpoint)
elif method == 'POST':
response = send_post_request(endpoint, payload, job['id'])
if response.status_code == 200:
return response.json()
else:
logger.error(f'HTTP Status code: {response.status_code}', job['id'])
logger.error(f'Response: {response.text}', job['id'])
return {
'error': response.text,
'refresh_worker': True
}
except Exception as e:
logger.error(f'An exception was raised: {e}')
return {
'error': traceback.format_exc(),
'refresh_worker': True
}
if __name__ == "__main__":
wait_for_service(f'{BASE_URI}/sdapi/v1/sd-models')
logger.info('A1111 Stable Diffusion API is ready')
logger.info('Starting RunPod Serverless...')
runpod.serverless.start(
{
'handler': handler
}
)
import os
import time
import requests
import traceback
import runpod
from runpod.serverless.utils.rp_validator import validate
from runpod.serverless.modules.rp_logger import RunPodLogger
from requests.adapters import HTTPAdapter, Retry
from huggingface_hub import HfApi
from schemas.input import INPUT_SCHEMA
from schemas.api import API_SCHEMA
from schemas.img2img import IMG2IMG_SCHEMA
from schemas.txt2img import TXT2IMG_SCHEMA
from schemas.interrogate import INTERROGATE_SCHEMA
from schemas.sync import SYNC_SCHEMA
from schemas.download import DOWNLOAD_SCHEMA
BASE_URI = 'http://127.0.0.1:3000'
TIMEOUT = 600
POST_RETRIES = 3
session = requests.Session()
retries = Retry(total=10, backoff_factor=0.1, status_forcelist=[502, 503, 504])
session.mount('http://', HTTPAdapter(max_retries=retries))
logger = RunPodLogger()
# ---------------------------------------------------------------------------- #
# Utility Functions #
# ---------------------------------------------------------------------------- #
def wait_for_service(url):
retries = 0
while True:
try:
requests.get(url)
return
except requests.exceptions.RequestException:
retries += 1
# Only log every 15 retries so the logs don't get spammed
if retries % 15 == 0:
logger.info('Service not ready yet. Retrying...')
except Exception as err:
logger.error(f'Error: {err}')
time.sleep(0.2)
def send_get_request(endpoint):
return session.get(
url=f'{BASE_URI}/{endpoint}',
timeout=TIMEOUT
)
def send_post_request(endpoint, payload, job_id, retry=0):
response = session.post(
url=f'{BASE_URI}/{endpoint}',
json=payload,
timeout=TIMEOUT
)
# Retry the post request in case the model has not completed loading yet
if response.status_code == 404:
if retry < POST_RETRIES:
retry += 1
logger.warn(f'Received HTTP 404 from endpoint: {endpoint}, Retrying: {retry}', job_id)
time.sleep(0.2)
send_post_request(endpoint, payload, job_id, retry)
return response
def validate_input(job):
return validate(job['input'], INPUT_SCHEMA)
def validate_api(job):
api = job['input']['api']
api['endpoint'] = api['endpoint'].lstrip('/')
return validate(api, API_SCHEMA)
def validate_payload(job):
method = job['input']['api']['method']
endpoint = job['input']['api']['endpoint']
payload = job['input']['payload']
validated_input = payload
if endpoint == 'v1/sync':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, SYNC_SCHEMA)
elif endpoint == 'v1/download':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, DOWNLOAD_SCHEMA)
elif endpoint == 'sdapi/v1/txt2img':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, TXT2IMG_SCHEMA)
elif endpoint == 'sdapi/v1/img2img':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, IMG2IMG_SCHEMA)
elif endpoint == 'sdapi/v1/interrogate' and method == 'POST':
logger.info(f'Validating /{endpoint} payload', job['id'])
validated_input = validate(payload, INTERROGATE_SCHEMA)
return endpoint, job['input']['api']['method'], validated_input
def download(job):
source_url = job['input']['payload']['source_url']
download_path = job['input']['payload']['download_path']
process_id = os.getpid()
temp_path = f"{download_path}.{process_id}"
# Download the file and save it as a temporary file
with requests.get(source_url, stream=True) as r:
r.raise_for_status()
with open(temp_path, 'wb') as f:
for chunk in r.iter_content(chunk_size=8192):
f.write(chunk)
# Rename the temporary file to the actual file name
os.rename(temp_path, download_path)
logger.info(f'{source_url} successfully downloaded to {download_path}', job['id'])
return {
'msg': 'Download successful',
'source_url': source_url,
'download_path': download_path
}
def sync(job):
repo_id = job['input']['payload']['repo_id']
sync_path = job['input']['payload']['sync_path']
hf_token = job['input']['payload']['hf_token']
api = HfApi()
models = api.list_repo_files(
repo_id=repo_id,
token=hf_token
)
synced_count = 0
synced_files = []
for model in models:
folder = os.path.dirname(model)
dest_path = f'{sync_path}/{model}'
if folder and not os.path.exists(dest_path):
logger.info(f'Syncing {model} to {dest_path}', job['id'])
uri = api.hf_hub_download(
token=hf_token,
repo_id=repo_id,
filename=model,
local_dir=sync_path,
local_dir_use_symlinks=False
)
if uri:
synced_count += 1
synced_files.append(dest_path)
return {
'synced_count': synced_count,
'synced_files': synced_files
}
# ---------------------------------------------------------------------------- #
# RunPod Handler #
# ---------------------------------------------------------------------------- #
def handler(job):
validated_input = validate_input(job)
if 'errors' in validated_input:
return {
'error': '\n'.join(validated_input['errors'])
}
validated_api = validate_api(job)
if 'errors' in validated_api:
return {
'error': '\n'.join(validated_api['errors'])
}
endpoint, method, validated_payload = validate_payload(job)
if 'errors' in validated_payload:
return {
'error': '\n'.join(validated_payload['errors'])
}
if 'validated_input' in validated_payload:
payload = validated_payload['validated_input']
else:
payload = validated_payload
try:
logger.info(f'Sending {method} request to: /{endpoint}', job['id'])
if endpoint == 'v1/download':
return download(job)
elif endpoint == 'v1/sync':
return sync(job)
elif method == 'GET':
response = send_get_request(endpoint)
elif method == 'POST':
response = send_post_request(endpoint, payload, job['id'])
if response.status_code == 200:
return response.json()
else:
logger.error(f'HTTP Status code: {response.status_code}', job['id'])
logger.error(f'Response: {response.text}', job['id'])
return {
'error': response.text,
'refresh_worker': True
}
except Exception as e:
logger.error(f'An exception was raised: {e}')
return {
'error': traceback.format_exc(),
'refresh_worker': True
}
if __name__ == "__main__":
wait_for_service(f'{BASE_URI}/sdapi/v1/sd-models')
logger.info('A1111 Stable Diffusion API is ready')
logger.info('Starting RunPod Serverless...')
runpod.serverless.start(
{
'handler': handler
}
)