**⚠️ Temporary fix to get Mixtral 8x22B with 64K context. If you are on SynapseAI 1.16, check out 1.16 of this repo

See this guide for setup: Mixtral 8x22B on Gaudi 2: A Comprehensive Setup Guide For optimizations Optimizing Mixtral 8x22b for Gaudi 2

Text Generation Inference on Habana Gaudi

Table of contents

• Running TGI on Gaudi
• Adjusting TGI parameters
• Running TGI with FP8 precision
• Currently supported configurations
• Environment variables
• Profiler

Running TGI on Gaudi

To use 🤗 text-generation-inference on Habana Gaudi/Gaudi2, follow these steps:

1. Pull the official Docker image with:

   docker pull ghcr.io/huggingface/tgi-gaudi:1.2.1
   

[!NOTE] Alternatively, you can build the Docker image using the Dockerfile located in this folder with:

docker build -t tgi_gaudi .

2. Launch a local server instance:

   i. On 1 Gaudi/Gaudi2 card

   model=meta-llama/Llama-2-7b-hf
   volume=$PWD/data # share a volume with the Docker container to avoid downloading weights every run
   
   docker run -p 8080:80 -v $volume:/data --runtime=habana -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model
   

   For gated models such as LLama or StarCoder, you will have to pass -e HUGGING_FACE_HUB_TOKEN=<token> to the docker run command above with a valid Hugging Face Hub read token.

   ii. On 8 Gaudi/Gaudi2 cards:

   model=meta-llama/Llama-2-70b-hf
   volume=$PWD/data # share a volume with the Docker container to avoid downloading weights every run
   
   docker run -p 8080:80 -v $volume:/data --runtime=habana -e PT_HPU_ENABLE_LAZY_COLLECTIVES=true -e HABANA_VISIBLE_DEVICES=all -e OMPI_MCA_btl_vader_single_copy_mechanism=none --cap-add=sys_nice --ipc=host ghcr.io/huggingface/tgi-gaudi:1.2.1 --model-id $model --sharded true --num-shard 8
   

3. You can then send a simple request:

   curl 127.0.0.1:8080/generate \
     -X POST \
     -d '{"inputs":"What is Deep Learning?","parameters":{"max_new_tokens":32}}' \
     -H 'Content-Type: application/json'
   

4. To run static benchmark test, please refer to TGI's benchmark tool.

   To run it on the same machine, you can do the following:

   • docker exec -it <docker name> bash , pick the docker started from step 2 using docker ps
   • text-generation-benchmark -t <model-id> , pass the model-id from docker run command
   • after the completion of tests, hit ctrl+c to see the performance data summary.

5. To run continuous batching test, please refer to examples.

Adjusting TGI parameters

Maximum sequence length is controlled by two arguments:

• --max-input-length is the maximum possible input prompt length. Default value is 1024.
• --max-total-tokens is the maximum possible total length of the sequence (input and output). Default value is 2048.

Maximum batch size is controlled by two arguments:

• For prefill operation, please set --max-prefill-total-tokens as bs * max-input-length, where bs is your expected maximum prefill batch size.
• For decode operation, please set --max-batch-total-tokens as bs * max-total-tokens, where bs is your expected maximum decode batch size.
• Please note that batch size will be always padded to the nearest multiplication of BATCH_BUCKET_SIZE and PREFILL_BATCH_BUCKET_SIZE.

To ensure greatest performance results, at the begginging of each server run, warmup is performed. It's designed to cover major recompilations while using HPU Graphs. It creates queries with all possible input shapes, based on provided parameters (described in this section) and runs basic TGI operations on them (prefill, decode, concatenate).

Except those already mentioned, there are other parameters that need to be properly adjusted to improve performance or memory usage:

• PAD_SEQUENCE_TO_MULTIPLE_OF determines sizes of input legnth buckets. Since warmup creates several graphs for each bucket, it's important to adjust that value proportionally to input sequence length. Otherwise, some out of memory issues can be observed.
• ENABLE_HPU_GRAPH enables HPU graphs usage, which is crucial for performance results. Recommended value to keep is true .

For more information and documentation about Text Generation Inference, checkout the README of the original repo.

Running TGI with FP8 precision

TGI supports FP8 precision runs within the limits provided by Habana Quantization Toolkit. Models with FP8 can be ran by properly setting QUANT_CONFIG environment variable. Detailed instruction on how to use that variable can be found in Optimum Habana FP8 guide. Summarising that instruction in TGI cases:

1. Measure quantization statistics of requested model by using Optimum Habana measurement script
2. Run requested model in TGI with proper QUANT_CONFIG setting - e.g. QUANT_CONFIG=./quantization_config/maxabs_quant.json

[!NOTE] Only models pointed in supported configurations are guaranteed to work with FP8

Additional hints to quantize model for TGI when using run_lm_eval.py:

• use --limit_hpu_graphs flag to save memory
• try to model Your use case situation by adjusting --batch_size , --max_new_tokens 512 and --max_input_tokens 512 In case of memory issues, lower those values
• use dataset/tasks suitable for Your use case (see --help for defining tasks/datasets)

Currently supported configurations

Not all features of TGI are currently supported as this is still a work in progress. Currently supported and validated configurations (other configurations are not guaranted to work or ensure reasonable performance):

Other sequence lengths can be used with proportionally decreased/increased batch size (the higher sequence length, the lower batch size). Support for other models from Optimum Habana will be added successively.

Environment variables

Profiler

To collect performance profiling, please set below environment variables:

The license to use TGI on Habana Gaudi is the one of TGI: https://github.com/huggingface/text-generation-inference/blob/main/LICENSE

Please reach out to api-enterprise@huggingface.co if you have any question.