mirror of
https://github.com/zenlm/zen-coder.git
synced 2026-11-10 21:30:01 +00:00
111 lines
2.8 KiB
Bash
Executable File
111 lines
2.8 KiB
Bash
Executable File
#!/bin/bash
|
|
|
|
dirs=(
|
|
# "codellama-7b"
|
|
# "codellama-13b"
|
|
# "codellama-34b"
|
|
# "codellama-python-7b"
|
|
# "codellama-python-13b"
|
|
# "codellama-python-34b"
|
|
# "codetulu-2-34b"
|
|
# "deepseek-base-1.3b"
|
|
# "deepseek-base-6.7b"
|
|
# "deepseek-base-33b"
|
|
# "deepseek-instruct-1.3b"
|
|
# "deepseek-instruct-6.7b"
|
|
# "deepseek-instruct-33b"
|
|
# "magicoder-ds-7b"
|
|
# "mistral-7b"
|
|
# "mixtral-8x7b"
|
|
# "phi-1"
|
|
# "phi-1.5"
|
|
# "phi-2"
|
|
# "phind"
|
|
# "starcoderbase-7b"
|
|
# "starcoderbase-16b"
|
|
# "wizard-13b"
|
|
# "wizard-34b"
|
|
)
|
|
|
|
models=(
|
|
# "codellama/CodeLlama-7b-hf"
|
|
# "codellama/CodeLlama-13b-hf"
|
|
# "codellama/CodeLlama-34b-hf"
|
|
# "codellama/CodeLlama-7b-Python-hf"
|
|
# "codellama/CodeLlama-13b-Python-hf"
|
|
# "codellama/CodeLlama-34b-Python-hf"
|
|
# "allenai/codetulu-2-34b"
|
|
# "deepseek-ai/deepseek-coder-1.3b-base"
|
|
# "deepseek-ai/deepseek-coder-6.7b-base"
|
|
# "deepseek-ai/deepseek-coder-33b-base"
|
|
# "deepseek-ai/deepseek-coder-1.3b-instruct"
|
|
# "deepseek-ai/deepseek-coder-6.7b-instruct"
|
|
# "deepseek-ai/deepseek-coder-33b-instruct"
|
|
# "ise-uiuc/Magicoder-S-DS-6.7B"
|
|
# "mistralai/Mistral-7B-v0.1"
|
|
# "mistralai/Mixtral-8x7B-v0.1"
|
|
# "microsoft/phi-1"
|
|
# "microsoft/phi-1_5"
|
|
# "microsoft/phi-2"
|
|
# "Phind/Phind-CodeLlama-34B-v2"
|
|
# "bigcode/starcoderbase-7b"
|
|
# "bigcode/starcoderbase"
|
|
# "WizardLM/WizardCoder-Python-13B-V1.0"
|
|
# "WizardLM/WizardCoder-Python-34B-V1.0"
|
|
)
|
|
|
|
temperatures=(0.2 0.8)
|
|
|
|
for ((i=0; i<${#models[@]}; i++)); do
|
|
model=${models[$i]}
|
|
base_dir=${dirs[$i]}
|
|
echo $model
|
|
for temperature in "${temperatures[@]}"; do
|
|
dir="${base_dir}_temp${temperature}_input"
|
|
cat <<EOF > temp_sbatch_script.sh
|
|
#!/bin/bash
|
|
#SBATCH --output=slurm_logs/slurm-%A-%a.out
|
|
#SBATCH --error=slurm_logs/slurm-%A-%a.err
|
|
#SBATCH --partition=YOUR_PARTITION_HERE
|
|
#SBATCH --array=0-1
|
|
#SBATCH --cpus-per-task=10
|
|
#SBATCH --gpus=1
|
|
#SBATCH --gpus-per-task=1
|
|
#SBATCH --mem=0GB
|
|
#SBATCH --time=03:00:00
|
|
|
|
dir=$dir
|
|
SIZE=800
|
|
GPUS=2
|
|
|
|
i=\$SLURM_ARRAY_TASK_ID
|
|
ip=\$((\$i+1))
|
|
|
|
echo \$dir
|
|
mkdir -p model_generations_raw/\$dir
|
|
|
|
string="Starting iteration \$i with start and end \$((\$i*SIZE/GPUS)) \$((\$ip*SIZE/GPUS))"
|
|
echo \$string
|
|
|
|
python main.py \
|
|
--model $model \
|
|
--use_auth_token \
|
|
--trust_remote_code \
|
|
--tasks input_prediction \
|
|
--batch_size 10 \
|
|
--n_samples 10 \
|
|
--max_length_generation 1024 \
|
|
--precision bf16 \
|
|
--limit \$SIZE \
|
|
--temperature $temperature \
|
|
--save_generations \
|
|
--save_generations_path model_generations_raw/\${dir}/shard_\$((\$i)).json \
|
|
--start \$((\$i*SIZE/GPUS)) \
|
|
--end \$((\$ip*SIZE/GPUS)) \
|
|
--shuffle \
|
|
--tensor_parallel_size 1
|
|
EOF
|
|
sbatch temp_sbatch_script.sh
|
|
rm temp_sbatch_script.sh
|
|
done
|
|
done |