Good morning everyone, and happy start of a new week.

I wanted to put this model on your radar. I did talk a lot about the code llama 70b model, but to be honest, it doesn't perform very well, especially compared to the performance upgrades that you get with models that contain useful memory context like GPT-4.

I cut the model a little bit of slack because it is free and open source, and I often use it (through Huggingface chat or offline locally with LM Studio) to save myself some queries.

This model, which is a well trained llama, seems to be performing better.

What do you think?

Here's what the team has to say about the release:

Phind-70B is based on the CodeLlama-70B model and is fine-tuned on an additional 50 billion tokens, yielding significant improvements. It also supports a context window of 32K tokens.

Original Article : https://www.phind.com/blog/introducing-phind-70b

Phind-70B scores 82.3% on HumanEval, beating the latest GPT-4 Turbo (gpt-4-0125-preview) score of 81.1% in our evaluation. On Meta's CRUXEval dataset, Phind-70B scores 59% to GPT-4's reported score of 62% on the output prediction benchmark. However, neither of these public datasets fully captures how our users use Phind for real-world workloads. We find that Phind-70B is in the same quality realm as GPT-4 Turbo for code generation and exceeds it on some tasks. Phind-70B is also less “lazy” than GPT-4 Turbo and doesn't hesistate to generate detailed code examples.

Phind-70B is significantly faster than GPT-4 Turbo, running at 80+ tokens per second to GPT-4 Turbo's ~20 tokens per second. We're able to achieve this by running NVIDIA's TensorRT-LLM library on H100 GPUs, and we're working on optimizations to further increase Phind-70B's inference speed.

Phind-70B is available today to try for free and without a login. You can get higher limits by subscribing to Phind Pro.

We love the open-source community and will be releasing the weights for the latest Phind-34B model in the coming weeks. We intend to release the weights for Phind-70B in time as well.

We'd like to thank our cloud partners, SF Compute and AWS, for helping us get the infrastructure right for training and serving Phind-70B. We'd also like to thank our partners at Meta and NVIDIA for their support.

Fun fact: We melted an H100 during Phind-70B's training!

Website: https://www.phind.com/search?home=true
Extension: https://marketplace.visualstudio.com/items?itemName=phind.phind

I want to help you, and to do that I've set the bundle to the best price (Gumroad suggests $600+)

Lifetime Membership + all resource BUNDLE

All the best,
- Gudasol 🜛

Questions? See the FAQ at the bottom of the bundle page 🛠

0 comments

Cursor with Composer: Strengths and Weaknesses

Rule with .cursorrules

Things you didn't know Cursor could do

It's official, I'm in love with Cursor.

Making it TOO EASY to dev

See all posts from Machine Minds AI