You will get RAG system with dense vector and LLama3 all finetuned by scene corpus
You will get RAG system with dense vector and LLama3 all finetuned by scene corpus
Project details
This is a fully integrated system:
1. Best denst vectors for text retrieve, with appropriate data, the F1 score > 0.97
2. Finetune LLMs for generate, with lora, pt, sft, raft and neft, eg LLaMa and ChatGLM
3. Format your doc/pdf, support table and image, using regex, pdf parser, ocr
4. My end-2-end RAG precision can reach 0.9
5. Vector visual, using PCA, t-SNE, UMAP and tensorboard
1. Best denst vectors for text retrieve, with appropriate data, the F1 score > 0.97
2. Finetune LLMs for generate, with lora, pt, sft, raft and neft, eg LLaMa and ChatGLM
3. Format your doc/pdf, support table and image, using regex, pdf parser, ocr
4. My end-2-end RAG precision can reach 0.9
5. Vector visual, using PCA, t-SNE, UMAP and tensorboard
Chatbot Type
Customer Service Chatbot, Productivity ChatbotDevelopment Technology
PythonWhat's included
| Service Tiers |
Starter
$200
|
Standard
$500
|
Advanced
$1,000
|
|---|---|---|---|
| Delivery Time | 4 days | 7 days | 10 days |
Number of Revisions | 3 | 5 | 7 |
Action Plan | - | - | - |
API Integration | - | - | - |
Bug Fixes | - | - | - |
Chatbot Flow Design | - | - | - |
Conversation Script | - | - | - |
About Julian
Algorithm Engineer | NLP
Shenzhen, China - 11:46 pm local time
1. Build sentence vectors for text retrieve, with appropriate data, the F1 score > 0.97
2. Finetune LLMs for generate, with lora, pt, sft, raft and neft, eg LLaMa and ChatGLM
3. Format your doc/pdf, support table and image, using regex, pdf parser, ocr
4. My end-2-end RAG precision can reach 0.9
5. Vector visual, using PCA, t-SNE, UMAP and tensorboard
6. Other NLP tasks, like classification, clustering, cky, eraley
7. 15 years in NLP
Steps for completing your project
After purchasing the project, send requirements so Julian can start the project.
Delivery time starts when Julian receives requirements from you.
Julian works on your project following the steps below.
Revisions may occur after the delivery date.
install and test