Instructions to use tencent/Hunyuan-MT-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use tencent/Hunyuan-MT-7B with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="tencent/Hunyuan-MT-7B")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("tencent/Hunyuan-MT-7B") model = AutoModelForCausalLM.from_pretrained("tencent/Hunyuan-MT-7B") - Notebooks
- Google Colab
- Kaggle
why are you using query/key layer norm AFTER rotary
#11
by vince62s - opened
My understanding is that standard practice is to LN before .... if we want to use flash decoding with the specific flash kernel then it is an issue because rotary is embedded in the kernel.
mango
θζ
mango.
mango.mango.mango.