mirror of https://github.com/hpcaitech/ColossalAI
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
YeAnbang
8a3ff4f315
|
4 months ago | |
|---|---|---|
| .. | ||
| ray | 8 months ago | |
| Opt.json | 8 months ago | |
| README.md | 8 months ago | |
| benchmark_dpo.sh | 4 months ago | |
| benchmark_kto.sh | 4 months ago | |
| benchmark_memory_consumption.txt | 8 months ago | |
| benchmark_orpo.sh | 4 months ago | |
| benchmark_performance_summarization.txt | 8 months ago | |
| benchmark_ppo.py | 7 months ago | |
| benchmark_ppo.sh | 8 months ago | |
| benchmark_sft.sh | 4 months ago | |
| benchmark_simpo.sh | 4 months ago | |
| data_preparation.sh | 8 months ago | |
| dummy_dataset.py | 4 months ago | |
| prepare_dummy_test_dataset.py | 4 months ago | |
README.md
Benchmarks
Benchmark OPT with LoRA on dummy prompt data
We provide various OPT models (string in parentheses is the corresponding model name used in this script):
- OPT-125M (125m)
- OPT-350M (350m)
- OPT-700M (700m)
- OPT-1.3B (1.3b)
- OPT-2.7B (2.7b)
- OPT-3.5B (3.5b)
- OPT-5.5B (5.5b)
- OPT-6.7B (6.7b)
- OPT-10B (10b)
- OPT-13B (13b)
We also provide various training strategies:
- gemini: ColossalAI GeminiPlugin with
placement_policy="cuda", like zero3 - gemini_auto: ColossalAI GeminiPlugin with
placement_policy="cpu", like zero3-offload - zero2: ColossalAI zero2
- zero2_cpu: ColossalAI zero2-offload
- 3d: ColossalAI HybridParallelPlugin with TP, DP support
How to Run
cd ../tests
# Prepare data for benchmark
SFT_DATASET=/path/to/sft/data/ \
PROMPT_DATASET=/path/to/prompt/data/ \
PRETRAIN_DATASET=/path/to/ptx/data/ \
PREFERENCE_DATASET=/path/to/preference/data \
./test_data_preparation.sh
# Start benchmark
./benchmark_ppo.sh