vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

server: tests: add split tests, and HF options params

Browse source
This commit is contained in:
Pierrick HYMBERT 2024-03-23 12:53:30 +01:00
parent 3ba5f2d124
commit b4a2ed8585
3 changed files with 13 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

3
examples/server/tests/features/parallel.feature
View file

@ -4,7 +4,8 @@ Feature: Parallel
Background: Server startup
Given a server listening on localhost:8080
And a model file tinyllamas/stories260K.gguf from HF repo ggml-org/models
And a model file tinyllamas/split/stories15M-00001-of-00003.gguf from HF repo ggml-org/models
And a model file test-model-00001-of-00003.gguf
And 42 as server seed
And 128 as batch size
And 256 KV cache size

4
examples/server/tests/features/server.feature
View file

@ -4,8 +4,8 @@ Feature: llama.cpp server
Background: Server startup
Given a server listening on localhost:8080
And a model url https://huggingface.co/ggml-org/models/resolve/main/tinyllamas/stories260K.gguf
And a model file stories260K.gguf
And a model file tinyllamas/stories260K.gguf from HF repo ggml-org/models
And a model file test-model.gguf
And a model alias tinyllama-2
And 42 as server seed
# KV Cache corresponds to the total amount of tokens

13
examples/server/tests/features/steps/steps.py
View file

@ -16,7 +16,6 @@ import numpy as np
import openai
from behave import step
from behave.api.async_step import async_run_until_complete
from huggingface_hub import hf_hub_download
from prometheus_client import parser
@ -39,6 +38,8 @@ def step_server_config(context, server_fqdn, server_port):
context.model_alias = None
context.model_file = None
context.model_hf_repo = None
context.model_hf_file = None
context.model_url = None
context.n_batch = None
context.n_ubatch = None
@ -68,9 +69,9 @@ def step_server_config(context, server_fqdn, server_port):
@step('a model file {hf_file} from HF repo {hf_repo}')
def step_download_hf_model(context, hf_file, hf_repo):
context.model_file = hf_hub_download(repo_id=hf_repo, filename=hf_file)
if context.debug:
print(f"model file: {context.model_file}")
context.model_hf_repo = hf_repo
context.model_hf_file = hf_file
context.model_file = os.path.basename(hf_file)
@step('a model file {model_file}')
@ -1079,6 +1080,10 @@ def start_server_background(context):
server_args.extend(['--model', context.model_file])
if context.model_url:
server_args.extend(['--model-url', context.model_url])
if context.model_hf_repo:
server_args.extend(['--hf-repo', context.model_hf_repo])
if context.model_hf_file:
server_args.extend(['--hf-file', context.model_hf_file])
if context.n_batch:
server_args.extend(['--batch-size', context.n_batch])
if context.n_ubatch: