Compare commits

..

20 Commits

Author SHA1 Message Date
Lucas Armand 0948d7c1ab Merge branch 'main' into pyworker-sdk 2025-12-15 17:24:16 -08:00
Lucas Armand e2bd0b1958 update readmes 2025-12-15 17:14:42 -08:00
Lucas Armand 0b6f381dd7 Add misc 2025-12-15 11:49:36 -08:00
Lucas Armand 74f8b6a1ef Added wheres-my-pyworker 2025-12-15 10:33:44 -08:00
Lucas Armand fa2bf082c2 only require HF_Token on backend 2025-12-12 14:47:29 -08:00
Lucas Armand 6a57ff8e0a try reverting env var 2025-12-12 12:16:33 -08:00
Lucas Armand 375633cb18 Fix 2025-12-12 12:12:57 -08:00
Lucas Armand ccd29ed8b6 remove input wrapping for vllm 2025-12-12 11:48:54 -08:00
Lucas Armand 2b30c69933 updated cost 2025-12-12 10:43:05 -08:00
Lucas Armand 4d99c12820 Added clients, updated READMEs 2025-12-12 10:41:21 -08:00
Lucas Armand 6060f8ce0c updated start_server.sh 2025-12-12 10:04:33 -08:00
Lucas Armand 067fa936fb remove legacy pyworker 2025-12-11 16:55:48 -08:00
Lucas Armand 405a8f1c0d returned to worker-sdk 2025-12-10 16:37:09 -08:00
Lucas Armand 12f4f23d39 remove parse request 2025-12-10 15:16:23 -08:00
Lucas Armand e2a771bb5a update ace and wan workers 2025-12-10 15:09:27 -08:00
Lucas Armand 0cd64adfc4 remove input 2025-12-10 14:47:47 -08:00
Lucas Armand 6f795b8fb8 remove input from workers 2025-12-10 14:46:10 -08:00
Lucas Armand 4bcc508473 reduce vllm benchmark runs to 2 2025-11-25 16:54:17 -08:00
Lucas Armand 74d7330800 add wan and ace workers 2025-11-25 16:08:40 -08:00
Lucas Armand 2ce0450809 Add worker.pys 2025-11-25 16:08:38 -08:00
-8
View File
@@ -28,12 +28,6 @@ MODEL_INFO_LOG_MSGS = [
nltk.download("words") nltk.download("words")
WORD_LIST = nltk.corpus.words.words() WORD_LIST = nltk.corpus.words.words()
def request_parser(request):
data = request
if request.get("input") is not None:
data = request.get("input")
return data
def completions_benchmark_generator() -> dict: def completions_benchmark_generator() -> dict:
prompt = " ".join(random.choices(WORD_LIST, k=int(250))) prompt = " ".join(random.choices(WORD_LIST, k=int(250)))
@@ -61,7 +55,6 @@ worker_config = WorkerConfig(
workload_calculator= lambda data: data.get("max_tokens", 0), workload_calculator= lambda data: data.get("max_tokens", 0),
allow_parallel_requests=True, allow_parallel_requests=True,
max_queue_time=60.0, max_queue_time=60.0,
request_parser=request_parser,
benchmark_config=BenchmarkConfig( benchmark_config=BenchmarkConfig(
generator=completions_benchmark_generator, generator=completions_benchmark_generator,
concurrency=100, concurrency=100,
@@ -73,7 +66,6 @@ worker_config = WorkerConfig(
workload_calculator= lambda data: data.get("max_tokens", 0), workload_calculator= lambda data: data.get("max_tokens", 0),
allow_parallel_requests=True, allow_parallel_requests=True,
max_queue_time=60.0, max_queue_time=60.0,
request_parser=request_parser
) )
], ],
log_action_config=LogActionConfig( log_action_config=LogActionConfig(