Construct KVTransferConfig properly from Python instead of using JSON blobs without CLI (#17994)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-05-12 11:25:33 -07:00
committed by GitHub
parent 98ea35601c
commit 72a3f6b898
5 changed files with 37 additions and 31 deletions
@@ -16,16 +16,17 @@ except FileNotFoundError:
sampling_params = SamplingParams(temperature=0, top_p=0.95, max_tokens=10)
llm = LLM(
model="meta-llama/Llama-3.2-1B-Instruct",
enforce_eager=True,
gpu_memory_utilization=0.8,
max_num_batched_tokens=64,
max_num_seqs=16,
kv_transfer_config=KVTransferConfig.from_cli(
'{"kv_connector":"SharedStorageConnector","kv_role":"kv_both",'
'"kv_connector_extra_config": {"shared_storage_path": "local_storage"}}'
)) #, max_model_len=2048, max_num_batched_tokens=2048)
llm = LLM(model="meta-llama/Llama-3.2-1B-Instruct",
enforce_eager=True,
gpu_memory_utilization=0.8,
max_num_batched_tokens=64,
max_num_seqs=16,
kv_transfer_config=KVTransferConfig(
kv_connector="SharedStorageConnector",
kv_role="kv_both",
kv_connector_extra_config={
"shared_storage_path": "local_storage"
})) #, max_model_len=2048, max_num_batched_tokens=2048)
# 1ST generation (prefill instance)
outputs = llm.generate(prompts, sampling_params)