You can finetune a 7B Llama 2 model on a single GPU A100 for example, so you could even use Google Colab. But in general, most universities and companies are supportive of open source these days.
CODE
-
MoE vs Specialized Models: Questioning Current Practical Utility
By
–
Yeah. But unless you want to have a generalist model for all languages, is it really that useful to have a MoE approach compared to say the CodeLlama-34B-Python version? I mean eventually things might move to MoE but maybe it’s not that big of a deal atm.
-
Team Pushing Hard on Model Improvements Concrete Examples
By
–
Send any concrete examples you have, the team is pushing very hard on model
Improvements. -
Code Interpreter Progress: Excitement for Future Improvements
By
–
It’s remarkable how far you can go with code interpreter, I am so exited to see how much it improves over the next year.
-
LangChain Batch Interface Concurrency Optimization Guide
By
–
you can use the `.batch` interface (see here: https://
python.langchain.com/docs/guides/ex
pression_language/interface#batch
…) and set `max_concurrency` to determine the max concurrency (see example of that here: https://
python.langchain.com/docs/modules/d
ata_connection/retrievers/multi_vector#summary
…) -
LangChain Twitter Fine-tuning Example Repository
By
–
example of that here: https://
github.com/langchain-ai/t
witter-finetune
… -
New Search Engine with Trigram and Vector Indexes for Models
By
–
Our new search engine uses a combination of trigram and text vector indexes to surface more relevant models. Results come in as you type, and you can quickly navigate using
↓ and ↑ arrow keys. Explore for yourself: http://
replicate.com/explore -
LangChain Gmail Integration and Grounding Importance Webinar
By
–
Overall, responses were pretty good, but highlights the importance for grounding! Check out the integration here: https://
python.langchain.com/docs/integrati
ons/chat_loaders/gmail
… We're doing a webinar with @GregKamradt on this topic this Wednesday – register here: -
Finetuned Model Customer Interaction and Hallucination Challenges
By
–
First up, a potential customer The finetuned model nicely responds and tries to find to set up a meeting – great! Note that Monday/Tuesday being good times was completely hallucinated! Finetuning alone wouldn't be enough for this to work for real – would need to hook up to gcal
-
Gmail ChatLoader Integration for Email Model Finetuning
By
–
Over the weekend, I added a GMail ChatLoader integration to @langchain
, to help anyone to finetune models on their emails I evaluated it the old fashioned way – the eyeball test Here's how it did responding to various types of inquiries (VC, customer, integration, @swyx
)