[Doc]: fix typos in Python comments (#24077)
Signed-off-by: Didier Durand <durand.didier@gmail.com>
This commit is contained in:
@@ -59,10 +59,10 @@ async def requests_processing_time(llm,
|
||||
@pytest.mark.asyncio
|
||||
async def test_add_lora(chatglm3_lora_files):
|
||||
"""
|
||||
The add_lora function is used to pre-load some LoRA adapters into the
|
||||
The add_lora function is used to preload some LoRA adapters into the
|
||||
engine in anticipation of future requests using these adapters. To test
|
||||
this functionality, we use the async engine to process some requests - We
|
||||
do it twice, once with add_lora() pre-loading and once without.
|
||||
do it twice, once with add_lora() preloading and once without.
|
||||
|
||||
We measure the request processing time in both cases and expect the time
|
||||
to be lesser in the case with add_lora() calls.
|
||||
|
||||
@@ -18,7 +18,7 @@ def test_allowed_token_ids_with_lora_vocab(llama_2_7b_base_huggingface_id,
|
||||
adapters that define additional tokens.
|
||||
"""
|
||||
|
||||
# Setup a base model compatible with the sql_lora_files adapter and
|
||||
# Set up a base model compatible with the sql_lora_files adapter and
|
||||
# a known number of tokens in the base model.
|
||||
model_config = ModelConfig(
|
||||
model=llama_2_7b_base_huggingface_id,
|
||||
@@ -84,7 +84,7 @@ def test_allowed_token_ids_with_lora_adapter_no_vocab(
|
||||
adapters that do not define additional tokens.
|
||||
"""
|
||||
|
||||
# Setup a base model compatible with the qwen25vl_lora_files adapter and
|
||||
# Set up a base model compatible with the qwen25vl_lora_files adapter and
|
||||
# a known number of tokens in the base model.
|
||||
model_config = ModelConfig(
|
||||
model=qwen25vl_base_huggingface_id,
|
||||
|
||||
Reference in New Issue
Block a user