Update README.md
Browse files
README.md
CHANGED
|
@@ -4,13 +4,12 @@ license: apache-2.0
|
|
| 4 |
|
| 5 |
# **GeM2-Llamion-14B**
|
| 6 |
|
| 7 |
-
This repository is the replica of [Llamion-14B](https://huggingface.co/vaiv/llamion-14b-base).
|
| 8 |
-
|
| 9 |
We have released **Llamion** as **GeM 2.0**, the second series of generative models developed by VAIV Company to address the our principal business needs.
|
| 10 |
|
| 11 |
**Llamion** (Llamafied Orion) is derived from transforming the [Orion model](https://huggingface.co/OrionStarAI/Orion-14B-Base)
|
| 12 |
into [the standard LLaMA architecture](https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py)
|
| 13 |
through parameter mapping and offline distillation.
|
|
|
|
| 14 |
|
| 15 |
<!-- Note that this model has NOT been contaminated to artificially inflate its scores for the Open LLM Leaderboards,
|
| 16 |
unlike some recent models which have been intentionally tainted. -->
|
|
|
|
| 4 |
|
| 5 |
# **GeM2-Llamion-14B**
|
| 6 |
|
|
|
|
|
|
|
| 7 |
We have released **Llamion** as **GeM 2.0**, the second series of generative models developed by VAIV Company to address the our principal business needs.
|
| 8 |
|
| 9 |
**Llamion** (Llamafied Orion) is derived from transforming the [Orion model](https://huggingface.co/OrionStarAI/Orion-14B-Base)
|
| 10 |
into [the standard LLaMA architecture](https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py)
|
| 11 |
through parameter mapping and offline distillation.
|
| 12 |
+
Further technical specifications and study results will be detailed in our upcoming paper, available on this page.
|
| 13 |
|
| 14 |
<!-- Note that this model has NOT been contaminated to artificially inflate its scores for the Open LLM Leaderboards,
|
| 15 |
unlike some recent models which have been intentionally tainted. -->
|