github/bilingual_book_maker
1
1
Fork 0
mirror of https://github.com/yihong0618/bilingual_book_maker.git synced 2025-06-02 09:30:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #448 from cangming/support_openai_o_series_model

Browse Source 
feat(model): support openai o series model
This commit is contained in:
yihong 2025-03-11 18:02:54 +08:00 committed by GitHub
commit 81f9b5280b
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
6 changed files with 128 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
.github/workflows/docs.yaml vendored
View File

@ -11,6 +11,6 @@ jobs:
- uses: actions/checkout@v2
- uses: actions/setup-python@v2
with:
python-version: '3.9'
python-version: '3.10'
- run: pip install mkdocs mkdocs-material
- run: mkdocs gh-deploy --force
- run: mkdocs gh-deploy --force

6
.github/workflows/make_test_ebook.yaml vendored
View File

@ -27,10 +27,10 @@ jobs:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: install python 3.9
- name: install python 3.10
uses: actions/setup-python@v4
with:
python-version: '3.9'
python-version: '3.10'
cache: 'pip' # caching pip dependencies
- name: Check formatting (black)
run: |
@ -71,7 +71,7 @@ jobs:
- name: Rename and Upload ePub
if: env.OPENAI_API_KEY != null
uses: actions/upload-artifact@v3
uses: actions/upload-artifact@v4
with:
name: epub_output
path: "test_books/lemo_bilingual.epub"

74
book_maker/cli.py
View File

@ -17,40 +17,54 @@ def parse_prompt_arg(prompt_arg):
if prompt_arg.endswith(".md") and os.path.exists(prompt_arg):
try:
from promptdown import StructuredPrompt
structured_prompt = StructuredPrompt.from_promptdown_file(prompt_arg)
# Initialize our prompt structure
prompt = {}
# Handle developer_message or system_message
# Developer message takes precedence if both are present
if hasattr(structured_prompt, 'developer_message') and structured_prompt.developer_message:
prompt['system'] = structured_prompt.developer_message
elif hasattr(structured_prompt, 'system_message') and structured_prompt.system_message:
prompt['system'] = structured_prompt.system_message
if (
hasattr(structured_prompt, "developer_message")
and structured_prompt.developer_message
):
prompt["system"] = structured_prompt.developer_message
elif (
hasattr(structured_prompt, "system_message")
and structured_prompt.system_message
):
prompt["system"] = structured_prompt.system_message
# Extract user message from conversation
if hasattr(structured_prompt, 'conversation') and structured_prompt.conversation:
if (
hasattr(structured_prompt, "conversation")
and structured_prompt.conversation
):
for message in structured_prompt.conversation:
if message.role.lower() == 'user':
prompt['user'] = message.content
if message.role.lower() == "user":
prompt["user"] = message.content
break
# Ensure we found a user message
if 'user' not in prompt or not prompt['user']:
raise ValueError("PromptDown file must contain at least one user message")
if "user" not in prompt or not prompt["user"]:
raise ValueError(
"PromptDown file must contain at least one user message"
)
print(f"Successfully loaded PromptDown file: {prompt_arg}")
# Validate required placeholders
if any(c not in prompt["user"] for c in ["{text}"]):
raise ValueError("User message in PromptDown must contain `{text}` placeholder")
raise ValueError(
"User message in PromptDown must contain `{text}` placeholder"
)
return prompt
except Exception as e:
print(f"Error parsing PromptDown file: {e}")
# Fall through to other parsing methods
# Existing parsing logic for JSON strings and other formats
if not any(prompt_arg.endswith(ext) for ext in [".json", ".txt", ".md"]):
try:
@ -388,7 +402,17 @@ So you are close to reaching the limit. You have to choose your own value, there
translate_model = MODEL_DICT.get(options.model)
assert translate_model is not None, "unsupported model"
API_KEY = ""
if options.model in ["openai", "chatgptapi", "gpt4", "gpt4omini", "gpt4o"]:
if options.model in [
"openai",
"chatgptapi",
"gpt4",
"gpt4omini",
"gpt4o",
"o1preview",
"o1",
"o1mini",
"o3mini",
]:
if OPENAI_API_KEY := (
options.openai_key
or env.get(
@ -510,6 +534,10 @@ So you are close to reaching the limit. You have to choose your own value, there
"gpt4",
"gpt4omini",
"gpt4o",
"o1",
"o1preview",
"o1mini",
"o3mini",
], "only support chatgptapi for deployment_id"
if not options.api_base:
raise ValueError("`api_base` must be provided when using `deployment_id`")
@ -534,6 +562,14 @@ So you are close to reaching the limit. You have to choose your own value, there
e.translate_model.set_gpt4omini_models()
if options.model == "gpt4o":
e.translate_model.set_gpt4o_models()
if options.model == "o1preview":
e.translate_model.set_o1preview_models()
if options.model == "o1":
e.translate_model.set_o1_models()
if options.model == "o1mini":
e.translate_model.set_o1mini_models()
if options.model == "o3mini":
e.translate_model.set_o3mini_models()
if options.model.startswith("claude-"):
e.translate_model.set_claude_model(options.model)
if options.block_size > 0:

4
book_maker/translator/__init__.py
View File

@ -16,6 +16,10 @@ MODEL_DICT = {
"gpt4": ChatGPTAPI,
"gpt4omini": ChatGPTAPI,
"gpt4o": ChatGPTAPI,
"o1preview": ChatGPTAPI,
"o1": ChatGPTAPI,
"o1mini": ChatGPTAPI,
"o3mini": ChatGPTAPI,
"google": Google,
"caiyun": Caiyun,
"deepl": DeepL,

63
book_maker/translator/chatgptapi_translator.py
View File

@ -48,6 +48,21 @@ GPT4o_MODEL_LIST = [
"gpt-4o-2024-08-06",
"chatgpt-4o-latest",
]
O1PREVIEW_MODEL_LIST = [
"o1-preview",
"o1-preview-2024-09-12",
]
O1_MODEL_LIST = [
"o1",
"o1-2024-12-17",
]
O1MINI_MODEL_LIST = [
"o1-mini",
"o1-mini-2024-09-12",
]
O3MINI_MODEL_LIST = [
"o3-mini",
]
class ChatGPTAPI(Base):
@ -422,6 +437,54 @@ class ChatGPTAPI(Base):
print(f"Using model list {model_list}")
self.model_list = cycle(model_list)
def set_o1preview_models(self):
# for issue #375 azure can not use model list
if self.deployment_id:
self.model_list = cycle(["o1-preview"])
else:
my_model_list = [
i["id"] for i in self.openai_client.models.list().model_dump()["data"]
]
model_list = list(set(my_model_list) & set(O1PREVIEW_MODEL_LIST))
print(f"Using model list {model_list}")
self.model_list = cycle(model_list)
def set_o1_models(self):
# for issue #375 azure can not use model list
if self.deployment_id:
self.model_list = cycle(["o1"])
else:
my_model_list = [
i["id"] for i in self.openai_client.models.list().model_dump()["data"]
]
model_list = list(set(my_model_list) & set(O1_MODEL_LIST))
print(f"Using model list {model_list}")
self.model_list = cycle(model_list)
def set_o1mini_models(self):
# for issue #375 azure can not use model list
if self.deployment_id:
self.model_list = cycle(["o1-mini"])
else:
my_model_list = [
i["id"] for i in self.openai_client.models.list().model_dump()["data"]
]
model_list = list(set(my_model_list) & set(O1MINI_MODEL_LIST))
print(f"Using model list {model_list}")
self.model_list = cycle(model_list)
def set_o3mini_models(self):
# for issue #375 azure can not use model list
if self.deployment_id:
self.model_list = cycle(["o3-mini"])
else:
my_model_list = [
i["id"] for i in self.openai_client.models.list().model_dump()["data"]
]
model_list = list(set(my_model_list) & set(O3MINI_MODEL_LIST))
print(f"Using model list {model_list}")
self.model_list = cycle(model_list)
def set_model_list(self, model_list):
model_list = list(set(model_list))
print(f"Using model list {model_list}")

2
docs/model_lang.md
View File

@ -2,7 +2,7 @@
## Models
`-m, --model <Model>` <br>
Currently `bbook_maker` supports these models: `chatgptapi` , `gpt3` , `google` , `caiyun` , `deepl` , `deeplfree` , `gpt4` , `gpt4omini` , `claude` , `customapi`.
Currently `bbook_maker` supports these models: `chatgptapi` , `gpt3` , `google` , `caiyun` , `deepl` , `deeplfree` , `gpt4` , `gpt4omini` , `o1-preview` , `o1` , `o1-mini` , `o3-mini` , `claude` , `customapi`.
Default model is `chatgptapi` .
### OPENAI models