Nicholai/jan
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add : img of using nitro in jan

Browse Source
This commit is contained in:
hahuyhoang411 2024-03-01 14:01:21 +07:00
parent ef36b6b5dc
commit e9362892e0
2 changed files with 6 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
docs/blog/img/nitro-on-jan.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 226 KiB

8
docs/blog/rag-is-not-enough.md
View File

@ -35,7 +35,7 @@ Mistral 7B continues to outshine [Meta's Llama-2 7B](https://huggingface.co/meta
Having a robust base model is critical. In our experiments, using Mistral as a starting point ensured the highest accuracy for subsequent specialized adaptations.
![Mistral vs LLama](img/mistral-comparasion.png)
![Mistral vs LLama vs Gemma](img/mistral-comparasion.png)
*Figure 1. Mistral 7B excels in benchmarks, ranking among the top foundational models.*
@ -93,4 +93,8 @@ Training was done with supervised finetuning (SFT) from the [Hugging Face's alig
We used consumer-grade, dual Nvidia RTX 4090s for the training. The end-to-end training took 18 minutes. We found optimal hyperparameters in LoRA for this specific task to be `r = 256` and `alpha = 512`.
This final model can be found [here on Huggingface](https://huggingface.co/jan-hq/nitro-v1.2-e3).
This final model can be found [here on Huggingface](https://huggingface.co/jan-hq/nitro-v1.2-e3).
![Using LLM locally](img/nitro-on-jan.png)
*Figure 3. Using the new finetuned model in [Jan](https://jan.ai/)*