github-clones
/
gpt_academic
mirror of https://github.com/binary-husky/gpt_academic.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
gpt_academic/request_llms
History
binary-husky cbbfa2e586
update gpt 0125 		2024-03-02 23:14:46 +08:00
..
README.md re-format code to with pre-commit 2023-12-31 19:30:32 +08:00
bridge_all.py update gpt 0125 2024-03-02 23:14:46 +08:00
bridge_chatglm.py 修复插件导入时的pytorch加载问题 2023-11-13 00:15:15 +08:00
bridge_chatglm3.py 修复插件导入时的pytorch加载问题 2023-11-13 00:15:15 +08:00
bridge_chatglmft.py change get_conf 2023-10-29 00:34:40 +08:00
bridge_chatglmonnx.py 修复本地模型在windows上的兼容性 2023-11-11 17:58:17 +08:00
bridge_chatgpt.py Merge Frontier, Update to Version 3.72 (#1553) 2024-02-14 18:35:09 +08:00
bridge_chatgpt_vision.py 遵循PEP 328优化太长的import 2023-11-22 13:20:56 +08:00
bridge_chatgpt_website.py 修复本地模型在Windows下的加载BUG 2023-11-11 21:11:55 +08:00
bridge_claude.py 修复本地模型在Windows下的加载BUG 2023-11-11 21:11:55 +08:00
bridge_deepseekcoder.py move token limit conf to bridge_all.py 2023-12-04 10:39:10 +08:00
bridge_google_gemini.py Merge Frontier, Update to Version 3.72 (#1553) 2024-02-14 18:35:09 +08:00
bridge_internlm.py 修正internlm输入设备bug 2023-11-11 23:22:50 +08:00
bridge_jittorllms_llama.py change get_conf 2023-10-29 00:34:40 +08:00
bridge_jittorllms_pangualpha.py change get_conf 2023-10-29 00:34:40 +08:00
bridge_jittorllms_rwkv.py change get_conf 2023-10-29 00:34:40 +08:00
bridge_llama2.py 修复错误的class命名 2023-11-24 02:42:58 +08:00
bridge_moss.py 修复插件导入时的pytorch加载问题 2023-11-13 00:15:15 +08:00
bridge_newbingfree.py format source code 2024-01-13 18:04:09 +08:00
bridge_qianfan.py handle qianfan chat error 2024-02-29 00:36:06 +08:00
bridge_qwen.py Fix response message bug in bridge_qianfan.py, 2024-02-15 00:02:24 +08:00
bridge_qwen_local.py name fix 2023-12-23 20:55:58 +08:00
bridge_skylark2.py Fix response message bug in bridge_qianfan.py, 2024-02-15 00:02:24 +08:00
bridge_spark.py Fix response message bug in bridge_qianfan.py, 2024-02-15 00:02:24 +08:00
bridge_stackclaude.py format source code 2024-01-13 18:04:09 +08:00
bridge_tgui.py rename folder 2023-10-28 17:44:17 +08:00
bridge_zhipu.py Merge Frontier, Update to Version 3.72 (#1553) 2024-02-14 18:35:09 +08:00
chatglmoonx.py rename folder 2023-10-28 17:44:17 +08:00
com_google.py fix zip chinese file name error 2024-02-25 21:13:41 +08:00
com_qwenapi.py change dash scope api key check behavior 2023-12-23 21:35:42 +08:00
com_skylark2api.py 注入火山引擎大模型的接口代码 2024-01-13 22:33:08 +08:00
com_sparkapi.py update spark v3.5, fix glm parallel problem 2024-02-18 14:08:35 +08:00
com_zhipuglm.py Merge Frontier, Update to Version 3.72 (#1553) 2024-02-14 18:35:09 +08:00
edge_gpt_free.py format source code 2024-01-13 18:04:09 +08:00
key_manager.py change get_conf 2023-10-29 00:34:40 +08:00
local_llm_class.py fix typo 2023-12-15 13:32:39 +08:00
queued_pipe.py 解除本地模型的若干并发问题 2023-10-31 20:37:07 +08:00
requirements_chatglm.txt re-format code to with pre-commit 2023-12-31 19:30:32 +08:00
requirements_chatglm_onnx.txt 修复protobuf版本错误 2023-12-14 11:00:55 +08:00
requirements_jittorllms.txt re-format code to with pre-commit 2023-12-31 19:30:32 +08:00
requirements_moss.txt re-format code to with pre-commit 2023-12-31 19:30:32 +08:00
requirements_newbing.txt rename folder 2023-10-28 17:44:17 +08:00
requirements_qwen.txt re-format code to with pre-commit 2023-12-31 19:30:32 +08:00
requirements_qwen_local.txt re-format code to with pre-commit 2023-12-31 19:30:32 +08:00
requirements_slackclaude.txt re-format code to with pre-commit 2023-12-31 19:30:32 +08:00

README.md

P.S. 如果您按照以下步骤成功接入了新的大模型，欢迎发Pull Requests（如果您在自己接入新模型的过程中遇到困难，欢迎加README底部QQ群联系群主）

如何接入其他本地大语言模型

  1. 复制request_llms/bridge_llama2.py，重命名为你喜欢的名字

  2. 修改load_model_and_tokenizer方法，加载你的模型和分词器（去该模型官网找demo，复制粘贴即可）

  3. 修改llm_stream_generator方法，定义推理模型（去该模型官网找demo，复制粘贴即可）

  4. 命令行测试

    • 修改tests/test_llms.py（聪慧如您，只需要看一眼该文件就明白怎么修改了）
    • 运行python tests/test_llms.py

  5. 测试通过后，在request_llms/bridge_all.py中做最后的修改，把你的模型完全接入到框架中（聪慧如您，只需要看一眼该文件就明白怎么修改了）

  6. 修改LLM_MODEL配置，然后运行python main.py，测试最后的效果

如何接入其他在线大语言模型

  1. 复制request_llms/bridge_zhipu.py，重命名为你喜欢的名字

  2. 修改predict_no_ui_long_connection

  3. 修改predict

  4. 命令行测试

    • 修改tests/test_llms.py（聪慧如您，只需要看一眼该文件就明白怎么修改了）
    • 运行python tests/test_llms.py

  5. 测试通过后，在request_llms/bridge_all.py中做最后的修改，把你的模型完全接入到框架中（聪慧如您，只需要看一眼该文件就明白怎么修改了）

  6. 修改LLM_MODEL配置，然后运行python main.py，测试最后的效果