File size: 4,381 Bytes
2e71e29 3491c38 68cb012 3491c38 27c95e7 3491c38 a3e6a17 3491c38 27c95e7 3491c38 27c95e7 3491c38 0976041 1efe76f 0976041 b9c5365 0976041 5e4a937 0976041 5e4a937 0976041 5e4a937 0976041 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 |
---
license: llama2
---
## Installation from source
```bash
git clone https://github.com/foundation-model-stack/fms-extras
cd fms-extras
pip install -e .
```
## Description
This model is intended to be used as an accelerator for [llama 13B (code)](https://huggingface.co/codellama/CodeLlama-13b-Instruct-hf) and takes inspiration
from the Medusa speculative decoding architecture. This accelerator modifies the MLP into a multi-stage MLP, where each stage predicts
a single token in the draft based on both a state vector and sampled token
from the prior stage (the base model can be considered stage 0).
The state vector from the base model provides contextual information to the accelerator,
while conditioning on prior sampled tokens allows it to produce higher-quality draft n-grams.
Note: The underlying MLP speculator is a generic architecture that can be trained with any generative model to accelerate inference.
Training is light-weight and can be completed in only a few days depending on base model size and speed.
## Repository Links
1. [Paged Attention KV-Cache / Speculator](https://github.com/foundation-model-stack/fms-extras)
2. [Production Server with speculative decoding](https://github.com/IBM/text-generation-inference.git)
3. [Speculator training](https://github.com/foundation-model-stack/fms-fsdp/pull/35)
## Samples
_Note: For all samples, your environment must have access to cuda_
### Production Server Sample
*To try this out running in a production-like environment, please use the pre-built docker image:*
#### Setup
```bash
docker pull quay.io/wxpe/text-gen-server:main.ee927a4
docker run -d --rm --gpus all \
--name my-tgis-server \
-p 8033:8033 \
-v /path/to/all/models:/models \
-e MODEL_NAME=/models/model_weights/llama/CodeLlama-13b-Instruct-hf \
-e SPECULATOR_NAME=/models/speculator_weights/llama/codellama-13b-accelerator \
-e FLASH_ATTENTION=true \
-e PAGED_ATTENTION=true \
-e DTYPE_STR=float16 \
quay.io/wxpe/text-gen-server:main.ee927a4
# check logs and wait for "gRPC server started on port 8033" and "HTTP server started on port 3000"
docker logs my-tgis-server -f
# get the client sample (Note: The first prompt will take longer as there is a warmup time)
conda create -n tgis-client-env python=3.11
conda activate tgis-client-env
git clone --branch main --single-branch https://github.com/IBM/text-generation-inference.git
cd text-generation-inference/integration_tests
make gen-client
pip install . --no-cache-dir
```
#### Run Sample
```bash
python sample_client.py
```
_Note: first prompt may be slower as there is a slight warmup time_
### Minimal Sample
#### Install
```bash
git clone -b code_llama_variant --single-branch https://github.com/JRosenkranz/fms-extras.git
(cd fms-extras && pip install -e .)
pip install transformers==4.35.0 sentencepiece numpy
```
#### Run Sample
##### batch_size=1 (compile + cudagraphs)
```bash
python fms-extras/scripts/paged_speculative_inference.py \
--variant=13b_code \
--model_path=/path/to/llama/CodeLlama-13b-Instruct-hf \
--model_source=hf \
--tokenizer=/path/to/llama/CodeLlama-13b-Instruct-hf \
--speculator_path=ibm-fms/codellama-13b-accelerator \
--speculator_source=hf \
--top_k_tokens_per_head=4,3,2,2,2,2,2 \
--prompt_type=code \
--compile \
--compile_mode=reduce-overhead
```
##### batch_size=1 (compile)
```bash
python fms-extras/scripts/paged_speculative_inference.py \
--variant=13b_code \
--model_path=/path/to/llama/CodeLlama-13b-Instruct-hf \
--model_source=hf \
--tokenizer=/path/to/llama/CodeLlama-13b-Instruct-hf \
--speculator_path=ibm-fms/codellama-13b-accelerator \
--speculator_source=hf \
--top_k_tokens_per_head=4,3,2,2,2,2,2 \
--prompt_type=code \
--compile \
```
##### batch_size=4 (compile)
```bash
python fms-extras/scripts/paged_speculative_inference.py \
--variant=13b_code \
--model_path=/path/to/llama/CodeLlama-13b-Instruct-hf \
--model_source=hf \
--tokenizer=/path/to/llama/CodeLlama-13b-Instruct-hf \
--speculator_path=ibm-fms/codellama-13b-accelerator \
--speculator_source=hf \
--batch_input \
--top_k_tokens_per_head=4,3,2,2,2,2,2 \
--prompt_type=code \
--compile \
```
Sample code can be found [here](https://github.com/foundation-model-stack/fms-extras/pull/18) |