bump up version for v2.0.21

Author: aarora79

README.md

@@ -88,7 +88,9 @@ Llama3 is now available on SageMaker (read [blog post](https://aws.amazon.com/bl
 ## New in this release
 
 ## 2.0.21
-1. Dynamically get EC2 pricing from Boto3 api
+
+1. Dynamically get EC2 pricing from Boto3 API.
+1. Update pricing information and model id for Amazon Bedrock models.
 
 ## 2.0.20
 1. Add `hf_tokenizer_model_id` parameter to automatically download tokenizers from Hugging Face.
@@ -97,10 +99,6 @@ Llama3 is now available on SageMaker (read [blog post](https://aws.amazon.com/bl
 1. Config files for `Llama3.1-1b` on AMD/Intel CPU instance types.
 1. Bug fixes for token counting for vLLM.
 
-## 2.0.18
-1. Delete SageMaker endpoint as soon as the run finishes.
-
-
 
 [Release history](./release_history.md)
 

create_manifest.py

@@ -21,6 +21,7 @@
                              "phi_tokenizer/.keep",
                              "scripts/.keep",
                              "configs/pricing.yml",
+                             "configs/pricing_fallback.yml",
                              # add new datasets here
                              "source_data/2wikimqa_e.jsonl",
                              "source_data/2wikimqa.jsonl",

docs/manifest.md

@@ -72,6 +72,8 @@ Here is a listing of the various configuration files available out-of-the-box wi
 [├── llama3/8b/config-ec2-llama3-8b-m7a-16xlarge.yml](configs/llama3/8b/config-ec2-llama3-8b-m7a-16xlarge.yml)  
 [├── llama3/8b/config-ec2-llama3-8b-m7a-24xlarge.yml](configs/llama3/8b/config-ec2-llama3-8b-m7a-24xlarge.yml)  
 [├── llama3/8b/config-ec2-llama3-8b-m7i-12xlarge.yml](configs/llama3/8b/config-ec2-llama3-8b-m7i-12xlarge.yml)  
+[├── llama3/8b/config-ec2-llama3-8b-m7i-16xlarge.yml](configs/llama3/8b/config-ec2-llama3-8b-m7i-16xlarge.yml)  
+[├── llama3/8b/config-ec2-llama3-8b-m7i-24xlarge.yml](configs/llama3/8b/config-ec2-llama3-8b-m7i-24xlarge.yml)  
 [├── llama3/8b/config-ec2-llama3-8b-neuron-trn1-32xl-tp16-sm.yml](configs/llama3/8b/config-ec2-llama3-8b-neuron-trn1-32xl-tp16-sm.yml)  
 [├── llama3/8b/config-ec2-llama3-8b-p4d-tp-2-mc-max.yml](configs/llama3/8b/config-ec2-llama3-8b-p4d-tp-2-mc-max.yml)  
 [├── llama3/8b/config-ec2-llama3-8b-p4d-tp-4-mc-max.yml](configs/llama3/8b/config-ec2-llama3-8b-p4d-tp-4-mc-max.yml)  
@@ -90,6 +92,7 @@ Here is a listing of the various configuration files available out-of-the-box wi
 [├── llama3/8b/config-llama3-8b-g5.12xl-tp-2-mc-max-triton-ec2.yml](configs/llama3/8b/config-llama3-8b-g5.12xl-tp-2-mc-max-triton-ec2.yml)  
 [├── llama3/8b/config-llama3-8b-g5.12xl-tp-4-mc-max-djl-ec2.yml](configs/llama3/8b/config-llama3-8b-g5.12xl-tp-4-mc-max-djl-ec2.yml)  
 [├── llama3/8b/config-llama3-8b-g5.12xl-tp-4-mc-max-triton-ec2.yml](configs/llama3/8b/config-llama3-8b-g5.12xl-tp-4-mc-max-triton-ec2.yml)  
+[├── llama3/8b/config-llama3-8b-g5.48xl-tp-8-mc-max-triton-ec2.yml](configs/llama3/8b/config-llama3-8b-g5.48xl-tp-8-mc-max-triton-ec2.yml)  
 [├── llama3/8b/config-llama3-8b-g5.4xl-tp-1-mc-max-triton-ec2.yml](configs/llama3/8b/config-llama3-8b-g5.4xl-tp-1-mc-max-triton-ec2.yml)  
 [├── llama3/8b/config-llama3-8b-g6e.12xl-tp-2-mc-max-djl-ec2.yml](configs/llama3/8b/config-llama3-8b-g6e.12xl-tp-2-mc-max-djl-ec2.yml)  
 [├── llama3/8b/config-llama3-8b-g6e.12xl-tp-2-mc-max-triton-ec2.yml](configs/llama3/8b/config-llama3-8b-g6e.12xl-tp-2-mc-max-triton-ec2.yml)  
@@ -233,4 +236,5 @@ Here is a listing of the various configuration files available out-of-the-box wi
 [└── multimodal/bedrock/config-llama-3-2-claude-models-scienceqa.yml](configs/multimodal/bedrock/config-llama-3-2-claude-models-scienceqa.yml)  
 **phi**  
 [└── phi/config-phi-3-g5.yml](configs/phi/config-phi-3-g5.yml)  
-[pricing.yml](configs/pricing.yml)  
+[pricing.yml](configs/pricing.yml)  
+[pricing_fallback.yml](configs/pricing_fallback.yml)  

docs/releases.md

@@ -1,6 +1,11 @@
 # Releases
 
-## 2.0.22
+## 2.0.21
+
+1. Dynamically get EC2 pricing from Boto3 API.
+1. Update pricing information and model id for Amazon Bedrock models.
+
+## 2.0.20
 
 1. Add `hf_tokenizer_model_id` parameter to automatically download tokenizers from Hugging Face.
 

manifest.txt

@@ -118,6 +118,8 @@ configs/llama3/8b/config-ec2-llama3-8b-m5-16xlarge.yml
 configs/llama3/8b/config-ec2-llama3-8b-m7a-16xlarge.yml
 configs/llama3/8b/config-ec2-llama3-8b-m7a-24xlarge.yml
 configs/llama3/8b/config-ec2-llama3-8b-m7i-12xlarge.yml
+configs/llama3/8b/config-ec2-llama3-8b-m7i-16xlarge.yml
+configs/llama3/8b/config-ec2-llama3-8b-m7i-24xlarge.yml
 configs/llama3/8b/config-ec2-llama3-8b-neuron-trn1-32xl-tp16-sm.yml
 configs/llama3/8b/config-ec2-llama3-8b-p4d-tp-2-mc-max.yml
 configs/llama3/8b/config-ec2-llama3-8b-p4d-tp-4-mc-max.yml
@@ -136,6 +138,7 @@ configs/llama3/8b/config-llama3-8b-g5.12xl-tp-2-mc-max-djl-ec2.yml
 configs/llama3/8b/config-llama3-8b-g5.12xl-tp-2-mc-max-triton-ec2.yml
 configs/llama3/8b/config-llama3-8b-g5.12xl-tp-4-mc-max-djl-ec2.yml
 configs/llama3/8b/config-llama3-8b-g5.12xl-tp-4-mc-max-triton-ec2.yml
+configs/llama3/8b/config-llama3-8b-g5.48xl-tp-8-mc-max-triton-ec2.yml
 configs/llama3/8b/config-llama3-8b-g5.4xl-tp-1-mc-max-triton-ec2.yml
 configs/llama3/8b/config-llama3-8b-g6e.12xl-tp-2-mc-max-djl-ec2.yml
 configs/llama3/8b/config-llama3-8b-g6e.12xl-tp-2-mc-max-triton-ec2.yml

release_history.md

@@ -1,3 +1,6 @@
+## 2.0.18
+1. Delete SageMaker endpoint as soon as the run finishes.
+
 ## 2.0.17
 1. Add support for embedding models through SageMaker jumpstart
 1. Add support for LLama 3.2 11b Vision Instruct benchmarking through FMBench