[TEST-USE] Fine Tuning
Closed this issue · 2 comments
guerra2fernando commented
[TEST-USE] Fine Tuning
jieguangzhou commented
Environment (with MongoDB + 1 type of data + 1 model):
- local cluster
- development mode
DBs (development mode + 1 non-trivial type of data + 1 model):
- MongoDB
- SQLite
- Postgres
Data Format (MongoDB + 1 model + development mode)
- Chat
- Prompt-Response
- Prompt
Training options (MongoDB + 1 type of data + development mode)
- ray
- deepspeed
- lora
- qlora
- Multi-GPUS
Load Options
- Directly
- Use checkpoint
jieguangzhou commented
At the functional level, the test has been completed (running a small max_steps test). Because it involves model training, work at the detection effect level is required.
I think I need to continue to run the real finetune task on data in different formats to test the effect after the model training is completed, Each training takes a long time, but other work can be done simultaneously
Data Format (MongoDB + 1 model + development mode)
- Chat
- Prompt-Response
- Prompt