CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-06 09:46:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

llama : update per-seq context computation

Browse Source
This commit is contained in:
Georgi Gerganov
2025-10-23 17:54:53 +03:00
parent db6f8e24e6
commit 8e11ec8eee
8 changed files with 40 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
tools/server/tests/unit/test_chat_completion.py
View File

@@ -433,21 +433,21 @@ def test_context_size_exceeded_stream():
@pytest.mark.parametrize(
"n_batch,batch_count,reuse_cache",
[
(64, 15, False),
(64, 3, False),
(64, 1, True),
]
)
def test_return_progresssss(n_batch, batch_count, reuse_cache):
def test_return_progress(n_batch, batch_count, reuse_cache):
global server
server.n_batch = n_batch
server.n_ctx = 2048
server.n_ctx = 256
server.n_slots = 1
server.start()
def make_cmpl_request():
return server.make_stream_request("POST", "/chat/completions", data={
"max_tokens": 10,
"messages": [
{"role": "user", "content": "This is a test" * 100},
{"role": "user", "content": "This is a test" * 10},
],
"stream": True,
"return_progress": True,