Compare commits
3 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
| d0347b0755 | |||
| 9bc9ba11c5 | |||
| 48fdc65e3d |
@@ -7,12 +7,14 @@ from vastai import Worker, WorkerConfig, HandlerConfig, LogActionConfig, Benchma
|
|||||||
# vLLM model configuration
|
# vLLM model configuration
|
||||||
MODEL_SERVER_URL = 'http://127.0.0.1'
|
MODEL_SERVER_URL = 'http://127.0.0.1'
|
||||||
MODEL_SERVER_PORT = 18000
|
MODEL_SERVER_PORT = 18000
|
||||||
MODEL_LOG_FILE = '/var/log/portal/vllm.log'
|
MODEL_LOG_FILE = '/var/log/onstart.log'
|
||||||
MODEL_HEALTHCHECK_ENDPOINT = "/health"
|
MODEL_HEALTHCHECK_ENDPOINT = "/health"
|
||||||
|
|
||||||
# vLLM-specific log messages
|
# vLLM-specific log messages
|
||||||
MODEL_LOAD_LOG_MSG = [
|
MODEL_LOAD_LOG_MSG = [
|
||||||
"Application startup complete.",
|
"Application startup complete.",
|
||||||
|
"llama runner started in",
|
||||||
|
"Server listening on",
|
||||||
]
|
]
|
||||||
|
|
||||||
MODEL_ERROR_LOG_MSGS = [
|
MODEL_ERROR_LOG_MSGS = [
|
||||||
|
|||||||
@@ -35,7 +35,7 @@ def benchmark_generator() -> dict:
|
|||||||
benchmark_data = {
|
benchmark_data = {
|
||||||
"inputs": prompt,
|
"inputs": prompt,
|
||||||
"parameters": {
|
"parameters": {
|
||||||
"max_new_tokens": 128,
|
"max_new_tokens": 500,
|
||||||
"temperature": 0.7,
|
"temperature": 0.7,
|
||||||
"return_full_text": False
|
"return_full_text": False
|
||||||
}
|
}
|
||||||
|
|||||||
Reference in New Issue
Block a user