jandres/GenAIExamples
1
0
Fork 0
Code Issues Pull Requests Actions 5 Packages Projects Releases Wiki Activity

[AudioQnA] Fix the LLM model field for inputs alignment (#1611)

Browse Source 
Signed-off-by: Wang, Kai Lawrence <kai.lawrence.wang@intel.com>
This commit is contained in:
Wang, Kai Lawrence
2025-03-05 22:15:07 +08:00
committed by GitHub
parent 8a5ad1fc72
commit 2dfcfa0436
5 changed files with 7 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
AudioQnA/audioqna.py
View File

@@ -16,13 +16,14 @@ SPEECHT5_SERVER_HOST_IP = os.getenv("SPEECHT5_SERVER_HOST_IP", "0.0.0.0")
SPEECHT5_SERVER_PORT = int(os.getenv("SPEECHT5_SERVER_PORT", 7055))
LLM_SERVER_HOST_IP = os.getenv("LLM_SERVER_HOST_IP", "0.0.0.0")
LLM_SERVER_PORT = int(os.getenv("LLM_SERVER_PORT", 3006))
LLM_MODEL_ID = os.getenv("LLM_MODEL_ID", "Intel/neural-chat-7b-v3-3")
def align_inputs(self, inputs, cur_node, runtime_graph, llm_parameters_dict, **kwargs):
if self.services[cur_node].service_type == ServiceType.LLM:
# convert TGI/vLLM to unified OpenAI /v1/chat/completions format
next_inputs = {}
next_inputs["model"] = "tgi" # specifically clarify the fake model to make the format unified
next_inputs["model"] = LLM_MODEL_ID
next_inputs["messages"] = [{"role": "user", "content": inputs["asr_result"]}]
next_inputs["max_tokens"] = llm_parameters_dict["max_tokens"]
next_inputs["top_p"] = llm_parameters_dict["top_p"]

3
AudioQnA/audioqna_multilang.py
View File

@@ -17,6 +17,7 @@ GPT_SOVITS_SERVER_HOST_IP = os.getenv("GPT_SOVITS_SERVER_HOST_IP", "0.0.0.0")
GPT_SOVITS_SERVER_PORT = int(os.getenv("GPT_SOVITS_SERVER_PORT", 9088))
LLM_SERVER_HOST_IP = os.getenv("LLM_SERVER_HOST_IP", "0.0.0.0")
LLM_SERVER_PORT = int(os.getenv("LLM_SERVER_PORT", 8888))
LLM_MODEL_ID = os.getenv("LLM_MODEL_ID", "Intel/neural-chat-7b-v3-3")
def align_inputs(self, inputs, cur_node, runtime_graph, llm_parameters_dict, **kwargs):
@@ -24,7 +25,7 @@ def align_inputs(self, inputs, cur_node, runtime_graph, llm_parameters_dict, **k
if self.services[cur_node].service_type == ServiceType.LLM:
# convert TGI/vLLM to unified OpenAI /v1/chat/completions format
next_inputs = {}
next_inputs["model"] = "tgi" # specifically clarify the fake model to make the format unified
next_inputs["model"] = LLM_MODEL_ID
next_inputs["messages"] = [{"role": "user", "content": inputs["asr_result"]}]
next_inputs["max_tokens"] = llm_parameters_dict["max_tokens"]
next_inputs["top_p"] = llm_parameters_dict["top_p"]

1
AudioQnA/docker_compose/amd/gpu/rocm/compose.yaml
View File

@@ -69,6 +69,7 @@ services:
- WHISPER_SERVER_PORT=${WHISPER_SERVER_PORT}
- LLM_SERVER_HOST_IP=${LLM_SERVER_HOST_IP}
- LLM_SERVER_PORT=${LLM_SERVER_PORT}
- LLM_MODEL_ID=${LLM_MODEL_ID}
- SPEECHT5_SERVER_HOST_IP=${SPEECHT5_SERVER_HOST_IP}
- SPEECHT5_SERVER_PORT=${SPEECHT5_SERVER_PORT}
ipc: host

1
AudioQnA/docker_compose/intel/cpu/xeon/compose.yaml
View File

@@ -61,6 +61,7 @@ services:
- WHISPER_SERVER_PORT=${WHISPER_SERVER_PORT}
- LLM_SERVER_HOST_IP=${LLM_SERVER_HOST_IP}
- LLM_SERVER_PORT=${LLM_SERVER_PORT}
- LLM_MODEL_ID=${LLM_MODEL_ID}
- SPEECHT5_SERVER_HOST_IP=${SPEECHT5_SERVER_HOST_IP}
- SPEECHT5_SERVER_PORT=${SPEECHT5_SERVER_PORT}
ipc: host

1
AudioQnA/docker_compose/intel/hpu/gaudi/compose.yaml
View File

@@ -82,6 +82,7 @@ services:
- WHISPER_SERVER_PORT=${WHISPER_SERVER_PORT}
- LLM_SERVER_HOST_IP=${LLM_SERVER_HOST_IP}
- LLM_SERVER_PORT=${LLM_SERVER_PORT}
- LLM_MODEL_ID=${LLM_MODEL_ID}
- SPEECHT5_SERVER_HOST_IP=${SPEECHT5_SERVER_HOST_IP}
- SPEECHT5_SERVER_PORT=${SPEECHT5_SERVER_PORT}
ipc: host