All gaudi values updated with extra flags. Added helm support for 2 new examples Text2Image and SearchQnA. Minor fix for llm-uservice. Signed-off-by: Dolpher Du <dolpher.du@intel.com>
40 lines
923 B
YAML
40 lines
923 B
YAML
# Copyright (C) 2024 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
tgi:
|
|
accelDevice: "gaudi"
|
|
image:
|
|
repository: ghcr.io/huggingface/tgi-gaudi
|
|
tag: "2.3.1"
|
|
resources:
|
|
limits:
|
|
habana.ai/gaudi: 1
|
|
MAX_INPUT_LENGTH: "1024"
|
|
MAX_TOTAL_TOKENS: "2048"
|
|
CUDA_GRAPHS: "0"
|
|
OMPI_MCA_btl_vader_single_copy_mechanism: "none"
|
|
ENABLE_HPU_GRAPH: "true"
|
|
LIMIT_HPU_GRAPH: "true"
|
|
USE_FLASH_ATTENTION: "true"
|
|
FLASH_ATTENTION_RECOMPUTE: "true"
|
|
PREFILL_BATCH_BUCKET_SIZE: 1
|
|
BATCH_BUCKET_SIZE: 8
|
|
extraCmdArgs:
|
|
- "--max-batch-total-tokens"
|
|
- "65536"
|
|
- "--max-batch-prefill-tokens"
|
|
- "4096"
|
|
livenessProbe:
|
|
initialDelaySeconds: 5
|
|
periodSeconds: 5
|
|
timeoutSeconds: 1
|
|
readinessProbe:
|
|
initialDelaySeconds: 5
|
|
periodSeconds: 5
|
|
timeoutSeconds: 1
|
|
startupProbe:
|
|
initialDelaySeconds: 5
|
|
periodSeconds: 5
|
|
timeoutSeconds: 1
|
|
failureThreshold: 120
|