OpenAI's tokenizer is available (as the tiktoken Python library) but you're on your own when it comes to guessing how many tokens will be used by the additional JSON syntax used for previous messages and tool definitions Mixtral's tokenizer here solves that problem
OPEN SOURCE
-
Anyscale Endpoints Plugin 0.5 Adds Mixtral 8x22B Support
By
–
New LLM plugin release: llm-anyscale-endpoints 0.5 adds support for their new hosted Mixtral 8x22B Instruct model, plus a "llm anyscale-endpoints refresh" command for refreshing the list of models without needing to upgrade the plugin in the future https://
github.com/simonw/llm-any
scale-endpoints/releases/tag/0.5
… -
PyTorch DDP Gradient Syncing Bug in Distributed Training
By
–
so last week I was posting about a bug i had with gradient syncing in torch DistributedDataParallel (DDP), aka the most standard way to do multi-GPU training for neural networks all of my misery originated with a single design decision from the torch team: DDP shares
-
Learning from Open Source AI Apps Feature Development
By
–
take @nutlope
’s notesgpt repo (or really any nontrivial oss ai app) and watch them add a feature -
SD3 Streamlit Implementation Repository Released
By
–
Github https://
github.com/Doriandarko/sd
3-streamlit
… -
Fine-tuning Strategy: Sequential Layer Training Approach
By
–
Has anyone tried fine-tuning half of the layers in one run, and then the other half in another? cc @teknium @winglian @erhartford
-
Snowflake Arctic Embed Open Source Models Launch
By
–
Snowflake Arctic Embed Yesterday @SnowflakeDB shipped their open source Arctic family of embedding models. You can use them with our @huggingface Embeddings connector to use these exceptionally performant models locally. Python docs: https://
python.langchain.com/docs/integrati
ons/providers/snowflake/#embedding-models
… Snowflake -
Community Builds Amazing Datasets with distilabel 1.0
By
–
Very excited to see all the amazing datasets the community builds and shares on the Hub using distilabel 1.0!
-
Creating a PCA Class Constructor with NumPy for Dimensionality Reduction
By
–
First Import numpy and define a class named "PCA" to encapsulate the operations. In the __init__ constructor let's initialize the number of components to reduce to, which is n_components, and similarly create placeholders for mean and eigenvectors.
-
Congratulations to Mistral AI team on their amazing release
By
–
Congratulations to all the @MistralAI team for their amazing work on this release!
n/n