Update README.md

README.md

@@ -15,7 +15,7 @@ This base model was created for use with [Shisa 7B](https://huggingface.co/augmx
 Training took 2,400 A100-40 GPU hours on a single 16 x A100-40 machine with [DeepSpeed](https://github.com/microsoft/DeepSpeed) ZeRO-3.
 
 ## Performance
-This base model was able to attain class-leading Japanese performance in standardized benchmarks with significantly less additional pre-training than previously released models. We believe this may be due to the use of a better-curated pre-training dataset, but ablations at even 2.5B additional JA tokens still showed very strong Japanese performance.
+This base model was able to attain class-leading Japanese performance in standardized benchmarks with significantly less additional pre-training than previously released models. We speculate this may be due to the use of a better-curated pre-training dataset, but ablations at even 2.5B additional JA tokens still showed very strong Japanese performance.
 
 We used a slightly modified [llm-jp-eval](https://github.com/llm-jp/llm-jp-eval) (our base model requires a `bos_token` to be prepended to the prompt; we tested other models with and without the modification and took the higher results for all models tested). Here we validate versus the original Mistral 7B base model as well as [Japanese Stable LM Instruct Gamma 7B](https://huggingface.co/stabilityai/japanese-stablelm-instruct-gamma-7b), which is a Mistral 7B base with an additional 100B tokens of JA/EN pre-training. We also include [Japanese-StableLM-Base-Beta-70B](https://huggingface.co/stabilityai/japanese-stablelm-base-beta-70b), which is a Llama 2 70B that also has an additional 100B tokens of JA/EN pre-training as a reference: