mirror of
https://github.com/RockChinQ/QChatGPT.git
synced 2024-11-15 19:22:24 +08:00
feat: 添加 model 命令
This commit is contained in:
parent
70583f5ba0
commit
68ddb3a6e1
|
@ -8,7 +8,7 @@ from . import entities, operator, errors
|
|||
from ..config import manager as cfg_mgr
|
||||
|
||||
# 引入所有算子以便注册
|
||||
from .operators import func, plugin, default, reset, list as list_cmd, last, next, delc, resend, prompt, cmd, help, version, update, ollama
|
||||
from .operators import func, plugin, default, reset, list as list_cmd, last, next, delc, resend, prompt, cmd, help, version, update, ollama, model
|
||||
|
||||
|
||||
class CommandManager:
|
||||
|
|
86
pkg/command/operators/model.py
Normal file
86
pkg/command/operators/model.py
Normal file
|
@ -0,0 +1,86 @@
|
|||
from __future__ import annotations
|
||||
|
||||
import typing
|
||||
|
||||
from .. import operator, entities, cmdmgr, errors
|
||||
|
||||
@operator.operator_class(
|
||||
name="model",
|
||||
help='显示和切换模型列表',
|
||||
usage='!model\n!model show <模型名>\n!model set <模型名>',
|
||||
privilege=2
|
||||
)
|
||||
class ModelOperator(operator.CommandOperator):
|
||||
"""Model命令"""
|
||||
|
||||
async def execute(self, context: entities.ExecuteContext) -> typing.AsyncGenerator[entities.CommandReturn, None]:
|
||||
content = '模型列表:\n'
|
||||
|
||||
model_list = self.ap.model_mgr.model_list
|
||||
|
||||
for model in model_list:
|
||||
content += f"\n名称: {model.name}\n"
|
||||
content += f"请求器: {model.requester.name}\n"
|
||||
|
||||
content += f"\n当前对话使用模型: {context.query.use_model.name}\n"
|
||||
content += f"新对话默认使用模型: {self.ap.provider_cfg.data.get('model')}\n"
|
||||
|
||||
yield entities.CommandReturn(text=content.strip())
|
||||
|
||||
|
||||
@operator.operator_class(
|
||||
name="show",
|
||||
help='显示模型详情',
|
||||
privilege=2,
|
||||
parent_class=ModelOperator
|
||||
)
|
||||
class ModelShowOperator(operator.CommandOperator):
|
||||
"""Model Show命令"""
|
||||
|
||||
async def execute(self, context: entities.ExecuteContext) -> typing.AsyncGenerator[entities.CommandReturn, None]:
|
||||
model_name = context.crt_params[0]
|
||||
|
||||
model = None
|
||||
for _model in self.ap.model_mgr.model_list:
|
||||
if model_name == _model.name:
|
||||
model = _model
|
||||
break
|
||||
|
||||
if model is None:
|
||||
yield entities.CommandReturn(error=errors.CommandError(f"未找到模型 {model_name}"))
|
||||
else:
|
||||
content = f"模型详情\n"
|
||||
content += f"名称: {model.name}\n"
|
||||
if model.model_name is not None:
|
||||
content += f"请求模型名称: {model.model_name}\n"
|
||||
content += f"请求器: {model.requester.name}\n"
|
||||
content += f"密钥组: {model.token_mgr.provider}\n"
|
||||
content += f"支持视觉: {model.vision_supported}\n"
|
||||
content += f"支持工具: {model.tool_call_supported}\n"
|
||||
|
||||
yield entities.CommandReturn(text=content.strip())
|
||||
|
||||
@operator.operator_class(
|
||||
name="set",
|
||||
help='设置默认使用模型',
|
||||
privilege=2,
|
||||
parent_class=ModelOperator
|
||||
)
|
||||
class ModelSetOperator(operator.CommandOperator):
|
||||
"""Model Set命令"""
|
||||
|
||||
async def execute(self, context: entities.ExecuteContext) -> typing.AsyncGenerator[entities.CommandReturn, None]:
|
||||
model_name = context.crt_params[0]
|
||||
|
||||
model = None
|
||||
for _model in self.ap.model_mgr.model_list:
|
||||
if model_name == _model.name:
|
||||
model = _model
|
||||
break
|
||||
|
||||
if model is None:
|
||||
yield entities.CommandReturn(error=errors.CommandError(f"未找到模型 {model_name}"))
|
||||
else:
|
||||
self.ap.provider_cfg.data['model'] = model_name
|
||||
await self.ap.provider_cfg.dump_config()
|
||||
yield entities.CommandReturn(text=f"已设置当前使用模型为 {model_name},重置会话以生效")
|
|
@ -2,9 +2,10 @@ from __future__ import annotations
|
|||
|
||||
import json
|
||||
import typing
|
||||
import traceback
|
||||
|
||||
import ollama
|
||||
from .. import operator, entities
|
||||
from .. import operator, entities, errors
|
||||
|
||||
|
||||
@operator.operator_class(
|
||||
|
@ -16,13 +17,16 @@ class OllamaOperator(operator.CommandOperator):
|
|||
async def execute(
|
||||
self, context: entities.ExecuteContext
|
||||
) -> typing.AsyncGenerator[entities.CommandReturn, None]:
|
||||
content: str = '模型列表:\n'
|
||||
model_list: list = ollama.list().get('models', [])
|
||||
for model in model_list:
|
||||
content += f"name: {model['name']}\n"
|
||||
content += f"modified_at: {model['modified_at']}\n"
|
||||
content += f"size: {bytes_to_mb(model['size'])}MB\n\n"
|
||||
yield entities.CommandReturn(text=f"{content.strip()}")
|
||||
try:
|
||||
content: str = '模型列表:\n'
|
||||
model_list: list = ollama.list().get('models', [])
|
||||
for model in model_list:
|
||||
content += f"名称: {model['name']}\n"
|
||||
content += f"修改时间: {model['modified_at']}\n"
|
||||
content += f"大小: {bytes_to_mb(model['size'])}MB\n\n"
|
||||
yield entities.CommandReturn(text=f"{content.strip()}")
|
||||
except ollama.ResponseError as e:
|
||||
yield entities.CommandReturn(error=errors.CommandError(f"无法获取模型列表,请确认 Ollama 服务正常"))
|
||||
|
||||
|
||||
def bytes_to_mb(num_bytes):
|
||||
|
@ -53,11 +57,9 @@ class OllamaShowOperator(operator.CommandOperator):
|
|||
model_info[key] = ignore_show
|
||||
|
||||
content += json.dumps(show, indent=4)
|
||||
yield entities.CommandReturn(text=content.strip())
|
||||
except ollama.ResponseError as e:
|
||||
content = f"{e.error}"
|
||||
|
||||
yield entities.CommandReturn(text=content.strip())
|
||||
|
||||
yield entities.CommandReturn(error=errors.CommandError(f"无法获取模型详情,请确认 Ollama 服务正常"))
|
||||
|
||||
@operator.operator_class(
|
||||
name="pull",
|
||||
|
@ -69,9 +71,13 @@ class OllamaPullOperator(operator.CommandOperator):
|
|||
async def execute(
|
||||
self, context: entities.ExecuteContext
|
||||
) -> typing.AsyncGenerator[entities.CommandReturn, None]:
|
||||
model_list: list = ollama.list().get('models', [])
|
||||
if context.crt_params[0] in [model['name'] for model in model_list]:
|
||||
yield entities.CommandReturn(text="模型已存在")
|
||||
try:
|
||||
model_list: list = ollama.list().get('models', [])
|
||||
if context.crt_params[0] in [model['name'] for model in model_list]:
|
||||
yield entities.CommandReturn(text="模型已存在")
|
||||
return
|
||||
except ollama.ResponseError as e:
|
||||
yield entities.CommandReturn(error=errors.CommandError(f"无法获取模型列表,请确认 Ollama 服务正常"))
|
||||
return
|
||||
|
||||
on_progress: bool = False
|
||||
|
|
Loading…
Reference in New Issue
Block a user