Update README.md
Browse files
README.md
CHANGED
@@ -19,7 +19,7 @@ The goal was to make a "mid" sized model like Meta has released in the past and
|
|
19 |
|
20 |
The model was finetuned on **8192 context length** and is likely reliable using RoPE up to 32k.
|
21 |
|
22 |
-
It still cannot do math reliably; neither can Llama-3-8B, and in my tests only Llama-3-70B passes
|
23 |
|
24 |
Further finetuning this model or finetuning the [base model](https://huggingface.co/elinas/Llama-3-13B-Instruct) on more samples is encouraged.
|
25 |
|
@@ -94,8 +94,10 @@ wandb: train_steps_per_second 0.078
|
|
94 |
- Datasets 2.15.0
|
95 |
- Tokenizers 0.15.0
|
96 |
|
97 |
-
##
|
98 |
|
99 |
TBD - submitted
|
100 |
|
|
|
|
|
101 |
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
|
|
|
19 |
|
20 |
The model was finetuned on **8192 context length** and is likely reliable using RoPE up to 32k.
|
21 |
|
22 |
+
It still cannot do math reliably; neither can Llama-3-8B, and in my tests only Llama-3-70B passes basic arithmetic, but it a better storywriter/RP than Llama-3-8B from some side by side testing I conducted.
|
23 |
|
24 |
Further finetuning this model or finetuning the [base model](https://huggingface.co/elinas/Llama-3-13B-Instruct) on more samples is encouraged.
|
25 |
|
|
|
94 |
- Datasets 2.15.0
|
95 |
- Tokenizers 0.15.0
|
96 |
|
97 |
+
## Model Evaluation
|
98 |
|
99 |
TBD - submitted
|
100 |
|
101 |
+
If you have any questions or comments on the model, feel free to open a discussion in the community tab.
|
102 |
+
|
103 |
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
|