update to 0.9.2.dev0

examples/inference/llama3_vllm.yaml

@@ -2,3 +2,4 @@ model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
 template: llama3
 infer_backend: vllm
 vllm_enforce_eager: true
+trust_remote_code: true

examples/inference/llava1_5.yaml

 model_name_or_path: llava-hf/llava-1.5-7b-hf
 template: llava
+infer_backend: huggingface  # choices: [huggingface, vllm]
+trust_remote_code: true

examples/inference/qwen2_vl.yaml

 model_name_or_path: Qwen/Qwen2-VL-7B-Instruct
 template: qwen2_vl
+infer_backend: huggingface  # choices: [huggingface, vllm]
+trust_remote_code: true

examples/merge_lora/llama3_gptq.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
 template: llama3
+trust_remote_code: true
 
 ### export
 export_dir: models/llama3_gptq

examples/merge_lora/llama3_lora_sft.yaml

@@ -5,6 +5,7 @@ model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
 adapter_name_or_path: saves/llama3-8b/lora/sft
 template: llama3
 finetuning_type: lora
+trust_remote_code: true
 
 ### export
 export_dir: models/llama3_lora_sft

examples/merge_lora/qwen2vl_lora_sft.yaml

@@ -5,6 +5,7 @@ model_name_or_path: Qwen/Qwen2-VL-7B-Instruct
 adapter_name_or_path: saves/qwen2_vl-7b/lora/sft
 template: qwen2_vl
 finetuning_type: lora
+trust_remote_code: true
 
 ### export
 export_dir: models/qwen2_vl_lora_sft

examples/train_full/llama3_full_sft.yaml

+### model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
+
+### method
+stage: sft
+do_train: true
+finetuning_type: full
+deepspeed: examples/deepspeed/ds_z3_config.json  # choices: [ds_z0_config.json, ds_z2_config.json, ds_z3_config.json]
+
+### dataset
+dataset: identity,alpaca_en_demo
+template: llama3
+cutoff_len: 2048
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+### output
+output_dir: saves/llama3-8b/full/sft
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+### train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 2
+learning_rate: 1.0e-5
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_ratio: 0.1
+bf16: true
+ddp_timeout: 180000000
+
+### eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+eval_strategy: steps
+eval_steps: 500

examples/train_full/qwen2vl_full_sft.yaml

 ### model
 model_name_or_path: Qwen/Qwen2-VL-7B-Instruct
+trust_remote_code: true
 
 ### method
 stage: sft
 do_train: true
 finetuning_type: full
-deepspeed: examples/deepspeed/ds_z3_config.json
+freeze_vision_tower: true  # choices: [true, false]
+freeze_multi_modal_projector: true  # choices: [true, false]
+train_mm_proj_only: false  # choices: [true, false]
+deepspeed: examples/deepspeed/ds_z3_config.json  # choices: [ds_z0_config.json, ds_z2_config.json, ds_z3_config.json]
 
 ### dataset
-dataset: mllm_demo,identity
+dataset: mllm_demo,identity,alpaca_en_demo
 template: qwen2_vl
 cutoff_len: 2048
 max_samples: 1000

examples/train_lora/llama3_lora_dpo.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: dpo

examples/train_lora/llama3_lora_eval.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
 adapter_name_or_path: saves/llama3-8b/lora/sft
+trust_remote_code: true
 
 ### method
 finetuning_type: lora

examples/train_lora/llama3_lora_kto.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: kto

examples/train_lora/llama3_lora_ppo.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
 reward_model: saves/llama3-8b/lora/reward
+trust_remote_code: true
 
 ### method
 stage: ppo

examples/train_lora/llama3_lora_pretrain.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: pt

examples/train_lora/llama3_lora_reward.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: rm

examples/train_lora/llama3_lora_sft.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: sft

examples/train_lora/llama3_lora_sft_ds3.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: sft
 do_train: true
 finetuning_type: lora
 lora_target: all
-deepspeed: examples/deepspeed/ds_z3_config.json
+deepspeed: examples/deepspeed/ds_z3_config.json  # choices: [ds_z0_config.json, ds_z2_config.json, ds_z3_config.json]
 
 ### dataset
 dataset: identity,alpaca_en_demo

examples/train_lora/llama3_lora_sft_ray.yaml

+### model
+model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct  # or use local absolute path
+trust_remote_code: true
+
+### method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: all
+
+### dataset
+dataset: identity,alpaca_en_demo
+dataset_dir: REMOTE:llamafactory/demo_data  # or use local absolute path
+template: llama3
+cutoff_len: 2048
+max_samples: 1000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+### output
+output_dir: tmp_dir
+logging_steps: 10
+save_steps: 500
+plot_loss: true
+overwrite_output_dir: true
+
+### train
+per_device_train_batch_size: 1
+gradient_accumulation_steps: 8
+learning_rate: 1.0e-4
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_ratio: 0.1
+bf16: true
+ddp_timeout: 180000000
+
+### eval
+val_size: 0.1
+per_device_eval_batch_size: 1
+eval_strategy: steps
+eval_steps: 500
+
+### ray
+ray_run_name: llama3_8b_sft_lora
+ray_num_workers: 4  # number of GPUs to use
+resources_per_worker:
+  GPU: 1
+placement_strategy: PACK

examples/train_lora/llama3_preprocess.yaml

 ### model
 model_name_or_path: meta-llama/Meta-Llama-3-8B-Instruct
+trust_remote_code: true
 
 ### method
 stage: sft

examples/train_lora/llava1_5_lora_sft.yaml

 ### model
 model_name_or_path: llava-hf/llava-1.5-7b-hf
+trust_remote_code: true
 
 ### method
 stage: sft

examples/train_lora/qwen2vl_lora_dpo.yaml

 ### model
 model_name_or_path: Qwen/Qwen2-VL-7B-Instruct
+trust_remote_code: true
 
 ### method
 stage: dpo