update examples

examples/full_multi_gpu/ds_z2_config.json

@@ -7,8 +7,8 @@
   "fp16": {
     "enabled": "auto",
     "loss_scale": 0,
-    "initial_scale_power": 16,
     "loss_scale_window": 1000,
+    "initial_scale_power": 16,
     "hysteresis": 2,
     "min_loss_scale": 1
   },
@@ -19,9 +19,10 @@
     "stage": 2,
     "allgather_partitions": true,
     "allgather_bucket_size": 5e8,
+    "overlap_comm": true,
     "reduce_scatter": true,
     "reduce_bucket_size": 5e8,
-    "overlap_comm": true,
-    "contiguous_gradients": true
+    "contiguous_gradients": true,
+    "round_robin_gradients": true
   }
 }

examples/full_multi_gpu/ds_z2_offload_config.json

@@ -0,0 +1,32 @@
+{
+  "train_batch_size": "auto",
+  "train_micro_batch_size_per_gpu": "auto",
+  "gradient_accumulation_steps": "auto",
+  "gradient_clipping": "auto",
+  "zero_allow_untested_optimizer": true,
+  "fp16": {
+    "enabled": "auto",
+    "loss_scale": 0,
+    "loss_scale_window": 1000,
+    "initial_scale_power": 16,
+    "hysteresis": 2,
+    "min_loss_scale": 1
+  },
+  "bf16": {
+    "enabled": "auto"
+  },
+  "zero_optimization": {
+    "stage": 2,
+    "offload_optimizer": {
+      "device": "cpu",
+      "pin_memory": true
+    },
+    "allgather_partitions": true,
+    "allgather_bucket_size": 5e8,
+    "overlap_comm": true,
+    "reduce_scatter": true,
+    "reduce_bucket_size": 5e8,
+    "contiguous_gradients": true,
+    "round_robin_gradients": true
+  }
+}

examples/full_multi_gpu/ds_z3_config.json

@@ -7,8 +7,8 @@
   "fp16": {
     "enabled": "auto",
     "loss_scale": 0,
-    "initial_scale_power": 16,
     "loss_scale_window": 1000,
+    "initial_scale_power": 16,
     "hysteresis": 2,
     "min_loss_scale": 1
   },
@@ -17,15 +17,12 @@
   },
   "zero_optimization": {
     "stage": 3,
-    "offload_optimizer": {
-      "device": "cpu"
-    },
-    "offload_param": {
-      "device": "cpu"
-    },
     "overlap_comm": true,
     "contiguous_gradients": true,
     "sub_group_size": 1e9,
+    "reduce_bucket_size": "auto",
+    "stage3_prefetch_bucket_size": "auto",
+    "stage3_param_persistence_threshold": "auto",
     "stage3_max_live_parameters": 1e9,
     "stage3_max_reuse_distance": 1e9,
     "stage3_gather_16bit_weights_on_model_save": true

examples/full_multi_gpu/ds_z3_offload_config.json

@@ -0,0 +1,38 @@
+{
+  "train_batch_size": "auto",
+  "train_micro_batch_size_per_gpu": "auto",
+  "gradient_accumulation_steps": "auto",
+  "gradient_clipping": "auto",
+  "zero_allow_untested_optimizer": true,
+  "fp16": {
+    "enabled": "auto",
+    "loss_scale": 0,
+    "loss_scale_window": 1000,
+    "initial_scale_power": 16,
+    "hysteresis": 2,
+    "min_loss_scale": 1
+  },
+  "bf16": {
+    "enabled": "auto"
+  },
+  "zero_optimization": {
+    "stage": 3,
+    "offload_optimizer": {
+      "device": "cpu",
+      "pin_memory": true
+    },
+    "offload_param": {
+      "device": "cpu",
+      "pin_memory": true
+    },
+    "overlap_comm": true,
+    "contiguous_gradients": true,
+    "sub_group_size": 1e9,
+    "reduce_bucket_size": "auto",
+    "stage3_prefetch_bucket_size": "auto",
+    "stage3_param_persistence_threshold": "auto",
+    "stage3_max_live_parameters": 1e9,
+    "stage3_max_reuse_distance": 1e9,
+    "stage3_gather_16bit_weights_on_model_save": true
+  }
+}

examples/full_multi_gpu/multi_node.sh

@@ -0,0 +1,37 @@
+#!/bin/bash
+
+python -m torch.distributed.run \
+    --nproc_per_node $NPROC_PER_NODE \
+    --nnodes $NNODES \
+    --node_rank $RANK \
+    --master_addr $MASTER_ADDR \
+    --master_port $MASTER_PORT \
+    ../../src/train_bash.py \
+    --deepspeed ds_z3_config.json \
+    --stage sft \
+    --do_train \
+    --model_name_or_path meta-llama/Llama-2-7b-hf \
+    --dataset alpaca_gpt4_en \
+    --dataset_dir ../../data \
+    --template default \
+    --finetuning_type full \
+    --output_dir ../../saves/LLaMA2-7B/full/sft \
+    --overwrite_cache \
+    --overwrite_output_dir \
+    --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
+    --per_device_train_batch_size 1 \
+    --per_device_eval_batch_size 1 \
+    --gradient_accumulation_steps 2 \
+    --lr_scheduler_type cosine \
+    --logging_steps 10 \
+    --warmup_steps 20 \
+    --save_steps 100 \
+    --eval_steps 100 \
+    --evaluation_strategy steps \
+    --learning_rate 5e-5 \
+    --num_train_epochs 3.0 \
+    --max_samples 3000 \
+    --val_size 0.1 \
+    --plot_loss \
+    --fp16

examples/full_multi_gpu/single_node.sh

@@ -13,11 +13,13 @@ deepspeed --num_gpus 4 ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 2 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
+    --warmup_steps 20 \
     --save_steps 100 \
     --eval_steps 100 \
     --evaluation_strategy steps \

examples/lora_multi_gpu/master_config.yaml

@@ -0,0 +1,18 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: MULTI_GPU
+downcast_bf16: 'no'
+gpu_ids: all
+machine_rank: 0
+main_process_ip: 192.168.0.1
+main_process_port: 29555
+main_training_function: main
+mixed_precision: fp16
+num_machines: 2
+num_processes: 16
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/lora_multi_gpu/multi_node.sh

@@ -1,6 +1,8 @@
 #!/bin/bash
 
-CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch --config_file config.yaml ../../src/train_bash.py \
+CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch \
+    --config_file master_config.yaml \
+    ../../src/train_bash.py \
     --stage sft \
     --do_train \
     --model_name_or_path meta-llama/Llama-2-7b-hf \
@@ -13,11 +15,13 @@ CUDA_VISIBLE_DEVICES=0,1,2,3 accelerate launch --config_file config.yaml ../../s
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 2 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
+    --warmup_steps 20 \
     --save_steps 100 \
     --eval_steps 100 \
     --evaluation_strategy steps \

examples/lora_multi_gpu/single_node.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+
+CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 accelerate launch \
+    --config_file single_config.yaml \
+    ../../src/train_bash.py \
+    --stage sft \
+    --do_train \
+    --model_name_or_path meta-llama/Llama-2-7b-hf \
+    --dataset alpaca_gpt4_en,glaive_toolcall \
+    --dataset_dir ../../data \
+    --template default \
+    --finetuning_type lora \
+    --lora_target q_proj,v_proj \
+    --output_dir ../../saves/LLaMA2-7B/lora/sft \
+    --overwrite_cache \
+    --overwrite_output_dir \
+    --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
+    --per_device_train_batch_size 1 \
+    --per_device_eval_batch_size 1 \
+    --gradient_accumulation_steps 2 \
+    --lr_scheduler_type cosine \
+    --logging_steps 10 \
+    --warmup_steps 20 \
+    --save_steps 100 \
+    --eval_steps 100 \
+    --evaluation_strategy steps \
+    --load_best_model_at_end \
+    --learning_rate 5e-5 \
+    --num_train_epochs 3.0 \
+    --max_samples 3000 \
+    --val_size 0.1 \
+    --plot_loss \
+    --fp16

examples/lora_multi_gpu/slave_config.yaml

@@ -0,0 +1,18 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+distributed_type: MULTI_GPU
+downcast_bf16: 'no'
+gpu_ids: all
+machine_rank: 1
+main_process_ip: 192.168.0.1
+main_process_port: 29555
+main_training_function: main
+mixed_precision: fp16
+num_machines: 2
+num_processes: 16
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false

examples/lora_single_gpu/dpo.sh

@@ -15,11 +15,13 @@ CUDA_VISIBLE_DEVICES=0 python ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 8 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
+    --warmup_steps 20 \
     --save_steps 100 \
     --eval_steps 100 \
     --evaluation_strategy steps \

examples/lora_single_gpu/ppo.sh

@@ -16,6 +16,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 512 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --gradient_accumulation_steps 8 \
     --lr_scheduler_type cosine \

examples/lora_single_gpu/predict.sh

@@ -13,6 +13,7 @@ CUDA_VISIBLE_DEVICES=0 python ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_eval_batch_size 1 \
     --max_samples 20 \
     --predict_with_generate

examples/lora_single_gpu/pretrain.sh

@@ -12,11 +12,13 @@ CUDA_VISIBLE_DEVICES=0 python ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 8 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
+    --warmup_steps 20 \
     --save_steps 100 \
     --eval_steps 100 \
     --evaluation_strategy steps \

examples/lora_single_gpu/reward.sh

@@ -15,11 +15,13 @@ CUDA_VISIBLE_DEVICES=0 python ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 8 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
+    --warmup_steps 20 \
     --save_steps 100 \
     --eval_steps 100 \
     --evaluation_strategy steps \

examples/lora_single_gpu/sft.sh

@@ -13,11 +13,13 @@ CUDA_VISIBLE_DEVICES=0 python ../../src/train_bash.py \
     --overwrite_cache \
     --overwrite_output_dir \
     --cutoff_len 1024 \
+    --preprocessing_num_workers 16 \
     --per_device_train_batch_size 1 \
     --per_device_eval_batch_size 1 \
     --gradient_accumulation_steps 8 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
+    --warmup_steps 20 \
     --save_steps 100 \
     --eval_steps 100 \
     --evaluation_strategy steps \