github/DB-GPT
1
0
Fork 0
mirror of https://github.com/csunny/DB-GPT.git synced 2025-09-08 04:23:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs(agent): Add agents development guide (#1673)

Browse Source
This commit is contained in:
Fangyin Cheng
2024-06-28 09:56:29 +08:00
committed by GitHub
parent 1796969bac
commit 45a960f331
43 changed files with 4010 additions and 146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
dbgpt/agent/util/llm/llm_client.py
View File

@@ -1,4 +1,5 @@
"""AIWrapper for LLM."""
import json
import logging
import traceback
@@ -75,14 +76,16 @@ class AIWrapper:
elif context and messages and isinstance(messages, list):
# Instantiate the messages
params["messages"] = [
{
**m,
"content": self.instantiate(
m["content"], context, allow_format_str_template
),
}
if m.get("content")
else m
(
{
**m,
"content": self.instantiate(
m["content"], context, allow_format_str_template
),
}
if m.get("content")
else m
)
for m in messages
]
return params
@@ -110,7 +113,7 @@ class AIWrapper:
config.pop(key)
return json.dumps(config, sort_keys=True, ensure_ascii=False)
async def create(self, **config) -> Optional[str]:
async def create(self, verbose: bool = False, **config) -> Optional[str]:
"""Create a response from the input config."""
# merge the input config with the i-th config in the config list
full_config = {**config}
@@ -123,7 +126,7 @@ class AIWrapper:
context = extra_kwargs.get("context")
llm_model = extra_kwargs.get("llm_model")
try:
response = await self._completions_create(llm_model, params)
response = await self._completions_create(llm_model, params, verbose)
except LLMChatError as e:
logger.debug(f"{llm_model} generate failed!{str(e)}")
raise e
@@ -151,7 +154,9 @@ class AIWrapper:
return gpts_messages
async def _completions_create(self, llm_model, params) -> str:
async def _completions_create(
self, llm_model, params, verbose: bool = False
) -> str:
payload = {
"model": llm_model,
"prompt": params.get("prompt"),
@@ -169,10 +174,16 @@ class AIWrapper:
payload["model_cache_enable"] = self.model_cache_enable
try:
model_request = _build_model_request(payload)
str_prompt = model_request.messages_to_string()
model_output = await self._llm_client.generate(model_request.copy())
parsed_output = self._output_parser.parse_model_nostream_resp(
model_output, "#########################"
)
if verbose:
print("\n", "-" * 80, flush=True, sep="")
print(f"String Prompt[verbose]: \n{str_prompt}")
print(f"LLM Output[verbose]: \n{parsed_output}")
print("-" * 80, "\n", flush=True, sep="")
return parsed_output
except Exception as e:
logger.error(