From 0e2a99e223f726db9132511e2c22efe2a19ae598 Mon Sep 17 00:00:00 2001 From: Henry Tu Date: Tue, 17 Dec 2024 19:28:24 -0500 Subject: [PATCH] Update Cerebras from Llama 3.1 to 3.3 (#645) # What does this PR do? Cerebras is rolling out support for llama 3.3 70b and deprecating llama 3.1 70b. This PR updates the documentation, config, and internal mapping to reflect this change. cc: @ashwinb @raghotham --- docs/source/distributions/self_hosted_distro/cerebras.md | 2 +- llama_stack/providers/remote/inference/cerebras/cerebras.py | 4 ++-- llama_stack/templates/cerebras/run.yaml | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/docs/source/distributions/self_hosted_distro/cerebras.md b/docs/source/distributions/self_hosted_distro/cerebras.md index 08b35809a..a8886d39b 100644 --- a/docs/source/distributions/self_hosted_distro/cerebras.md +++ b/docs/source/distributions/self_hosted_distro/cerebras.md @@ -23,7 +23,7 @@ The following environment variables can be configured: The following models are available by default: - `meta-llama/Llama-3.1-8B-Instruct (llama3.1-8b)` -- `meta-llama/Llama-3.1-70B-Instruct (llama3.1-70b)` +- `meta-llama/Llama-3.3-70B-Instruct (llama-3.3-70b)` ### Prerequisite: API Keys diff --git a/llama_stack/providers/remote/inference/cerebras/cerebras.py b/llama_stack/providers/remote/inference/cerebras/cerebras.py index 5a9fef22a..2ff213c2e 100644 --- a/llama_stack/providers/remote/inference/cerebras/cerebras.py +++ b/llama_stack/providers/remote/inference/cerebras/cerebras.py @@ -41,8 +41,8 @@ model_aliases = [ CoreModelId.llama3_1_8b_instruct.value, ), build_model_alias( - "llama3.1-70b", - CoreModelId.llama3_1_70b_instruct.value, + "llama-3.3-70b", + CoreModelId.llama3_3_70b_instruct.value, ), ] diff --git a/llama_stack/templates/cerebras/run.yaml b/llama_stack/templates/cerebras/run.yaml index b7c2d316e..05b21bf0a 100644 --- a/llama_stack/templates/cerebras/run.yaml +++ b/llama_stack/templates/cerebras/run.yaml @@ -56,9 +56,9 @@ models: provider_model_id: llama3.1-8b model_type: llm - metadata: {} - model_id: meta-llama/Llama-3.1-70B-Instruct + model_id: meta-llama/Llama-3.3-70B-Instruct provider_id: cerebras - provider_model_id: llama3.1-70b + provider_model_id: llama-3.3-70b model_type: llm - metadata: embedding_dimension: 384