File tree 2 files changed +10
-7
lines changed
evaluation/installers/llm-jp-eval-v1.3.1
2 files changed +10
-7
lines changed Original file line number Diff line number Diff line change @@ -59,6 +59,9 @@ huggingface-cli login
59
59
必要に応じて` run_llm-jp-eval.sh ` ・` resources/config_base.yaml ` 内の変数を書き換えてください
60
60
- tokenizer・wandb entity・wandb projectを変更する場合` run_llm-jp-eval.sh ` のみの変更で対応可能
61
61
- その他の変更を行う場合、` resources/config_base.yaml ` を変更した上で、` run_llm-jp-eval.sh ` 内でファイルを指定
62
+
63
+ VRAMはモデルサイズの2.5-3.5倍必要(例: 13B model -> 33GB-45GB)<br >
64
+ SLURM環境で実行する場合、デフォルトでは` --gpus 1 ` のため、` --mem ` と共にクラスタに適切なサイズに設定すること
62
65
``` shell
63
66
cd ~ /myspace
64
67
# (Optional) If you need to change variables
@@ -73,11 +76,11 @@ CUDA_VISIBLE_DEVICES={num} bash run_llm-jp-eval.sh {path/to/model} {wandb.run_na
73
76
```
74
77
75
78
#### Sample code
76
- ` ``shell
77
- # For a cluster with SLURM
78
- sbatch --partition {partition} run_llm-jp-eval.sh llm-jp/llm-jp-13b-v2.0 test-$( whoami)
79
- # For a cluster without SLURM
80
- CUDA_VISIBLE_DEVICES=0 bash run_llm-jp-eval.sh llm-jp/llm-jp-13b-v2.0 test-$( whoami)
79
+ ``shell
80
+ # Evaluate 70B model on a cluster with SLURM using H100 (VRAM: 80GB)
81
+ sbatch --partition {partition} --gpus 4 --mem 8G run_llm-jp-eval.sh sbintuitions/sarashina2-70b test-$(whoami)
82
+ # Evakyate 13B model on a cluster without SLURM using A100 (VRAM: 40GB)
83
+ CUDA_VISIBLE_DEVICES=0,1 bash run_llm-jp-eval.sh llm-jp/llm-jp-13b-v2.0 test-$(whoami)
81
84
```
82
85
83
86
## 開発者向け: resources/sha256sums.csv の作成コマンド
Original file line number Diff line number Diff line change 1
1
#! /bin/bash
2
2
# SBATCH --job-name=llm-jp-eval
3
3
# SBATCH --partition=<partition>
4
- # SBATCH --exclusive
5
4
# SBATCH --nodes=1
5
+ # SBATCH --cpus-per-task=8
6
6
# SBATCH --gpus=1
7
- # SBATCH --ntasks-per-node=8
7
+ # SBATCH --mem=200G
8
8
# SBATCH --output=logs/%x-%j.out
9
9
# SBATCH --error=logs/%x-%j.err
10
10
You can’t perform that action at this time.
0 commit comments