Comments (8)
Unfortunately abstract_graph
is a mess, the most chaotic submodule in the library, and we know it. It will go through through some rounds of refactoring in the upcoming weeks. The first round is #494. Hopefully we'll be able fix this issue soon. Thanks for reporting this.
from scrapegraph-ai.
I think Langchain framework has also faced similar challenges, and I look forward to a better experience in the new version.
Unfortunately
abstract_graph
is a mess, the most chaotic submodule in the library, and we know it. It will go through through some rounds of refactoring in the upcoming weeks. The first round is #494. Hopefully we'll be able fix this issue soon. Thanks for reporting this.
from scrapegraph-ai.
A little update on the latest betas: embeddings are not needed anymore in any of the graphs. The RAGNode
(the only node using embeddings) is now dead code, but we left it in the code base in case it might be needed in the future, and relocated all the embeddings initialization code inside of it, instead of leaving it the AbstractGraph
.
All of this to say: the bug you reported probably won't be fixed until we find a new use for the RAGNode
, but it won't bother you anymore because the embeddings will not be used by any graph in the upcoming releases of ScrapeGraphAI.
Thank for your input, thanks for using ScrapeGraph, and let us know in case you have any other problem with the library.
from scrapegraph-ai.
Great! I'll have a try.
from scrapegraph-ai.
@f-aguzzi In the latest version v1.12.2, I still found some mistakes with gemini model.
In abstract_graph.py
file, use "google_genai".
if "gemini" in llm_params["model"]:
model_name = llm_params["model"].split("/")[-1]
return handle_model(model_name, "google_genai", model_name)
But in models_tokens.py
file, still use "gemini" .
from scrapegraph-ai.
@f-aguzzi In the latest version v1.12.2, I still found some mistakes with gemini model.
In
abstract_graph.py
file, use "google_genai".if "gemini" in llm_params["model"]: model_name = llm_params["model"].split("/")[-1] return handle_model(model_name, "google_genai", model_name)
But in
models_tokens.py
file, still use "gemini" .
We'll fix this ASAP, thanks for reporting it to us
from scrapegraph-ai.
๐ This issue has been resolved in version 1.14.0-beta.3 ๐
The release is available on:
v1.14.0-beta.3
- GitHub release
Your semantic-release bot ๐ฆ๐
from scrapegraph-ai.
hi please update to the new beta
from scrapegraph-ai.
Related Issues (20)
- gpt4o does not return a json output and broke the pipeline
- Omit Certain parts of the HTML HOT 1
- Set default temperature to 0
- `NameError: name 'input_data' is not defined` is raised when trying to use smart_scraper_graph.run() HOT 3
- OMP: Error #15: Initializing libiomp5md.dll, but found libomp140.x86_64.dll already initialized HOT 7
- NameError: name 'input_data' is not defined HOT 1
- Refactoring of prompts HOT 2
- probable bug HOT 1
- refactoring of nodes HOT 1
- ImportError: cannot import name 'OpenAI' from 'scrapegraphai.models' (/usr/local/lib/python3.10/dist-packages/scrapegraphai/models/__init__.py) HOT 3
- Ability to add headers to source ( better integration with Jina AI) HOT 2
- The Dockerfile doesn't work well HOT 2
- Context length exceeded
- Scrapegraph returns relative path URLs instead of absolute path **Possible Bug?** HOT 8
- `embedder_model` AttributeError in `/examples/openai/deep_scraper_openai.py` HOT 1
- Different prompt_tokens size between Azure OpenAI and OpenAI HOT 4
- Unable to join discord on readme HOT 1
- Add support for other OpenAI models like gpt-4o-2024-08-06 HOT 1
- improve tokenization function
- Whether it can support llm models compatible with the open AI api protocol๏ผ HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from scrapegraph-ai.