jandres/GenAIExamples
1
0
Fork 0
Code Issues Pull Requests Actions 4 Packages Projects Releases Wiki Activity

instruction finetune README improvement (#897)

Browse Source 
Signed-off-by: rbrugaro <rita.brugarolas.brufau@intel.com>
This commit is contained in:
rbrugaro
2024-10-07 23:04:47 -07:00
committed by GitHub
parent ffecd182db
commit 644c3a67ce
1 changed files with 3 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
InstructionTuning/README.md
View File

@@ -1,6 +1,6 @@
# Instruction Tuning
Instruction tuning is the process of further training LLMs on a dataset consisting of (instruction, output) pairs in a supervised fashion, which bridges the gap between the next-word prediction objective of LLMs and the users' objective of having LLMs adhere to human instructions.
Instruction tuning is the process of further training LLMs on a dataset consisting of (instruction, output) pairs in a supervised fashion, which bridges the gap between the next-word prediction objective of LLMs and the users' objective of having LLMs adhere to human instructions. This implementation deploys a Ray cluster for the task.
## Deploy Instruction Tuning Service
@@ -38,6 +38,8 @@ curl http://${your_ip}:8015/v1/fine_tuning/jobs \
}'
```
The outputs of the finetune job (adapter_model.safetensors, adapter_config,json... ) are stored in `/home/user/comps/finetuning/output` and other execution logs are stored in `/home/user/ray_results`
### 3. Manage fine-tuning job
Below commands show how to list finetuning jobs, retrieve a finetuning job, cancel a finetuning job and list checkpoints of a finetuning job.