Add MPT documentation

README.md

@@ -15,6 +15,7 @@
 
 ## Release
 
+- [5/6] 🔥 We are releasing [LLaVA-Lighting-MPT-7B-preview](https://huggingface.co/liuhaotian/LLaVA-Lightning-MPT-7B-preview), based on MPT-7B-Chat!  See [here](#LLaVA-MPT-7b) for more details.
 - [5/2] 🔥 We are releasing LLaVA-Lighting!  Train a lite, multimodal GPT-4 with just $40 in 3 hours!  See [here](#train-llava-lightning) for more details.
 - [5/2] We upgrade LLaVA package to v0.1 to support Vicuna v0 and v1 checkpoints, please upgrade following instructions [here](#install).
 - [4/30] Our checkpoint with Vicuna-7b-v0 has been released [here](#llava-7b)! This checkpoint is more accessible and device friendly.  Stay tuned for a major upgrade next week!
@@ -549,6 +550,33 @@ bash ./scripts/train_lightning.sh {v0,v1}
 | --- | ---: | ---: | ---: | ---: | ---: |
 | LLaVA-Lightning-7B | 128 | 2e-5 | 1 | 2048 | 0 |
 
+#### LLaVA-MPT-7b
+Thanks to LLaVA-Lightning, we are able to train a checkpoint based on MPT-7b-Chat on 8x A100 GPUs in just 3 hours, including both pretraining and finetuning.
+
+**NOTE**: This is a research preview of the LLaVA-Lightning based on MPT-7B-chat checkpoint. The usage of the model should comply with MPT-7B-chat license and agreements.
+
+**NOTE**: Unlike other LLaVA models, this model should be used directly without delta weights conversion!
+
+**NOTE**: You need to upgrade to our latest code base to use LLaVA-MPT-7b!
+
+1. Usage
+
+You do not need to download our checkpoint, it will directly load from our Hugging Face model: [`liuhaotian/LLaVA-Lightning-MPT-7B-preview`](https://huggingface.co/liuhaotian/LLaVA-Lightning-MPT-7B-preview).
+
+```Shell
+python -m llava.serve.controller --host 0.0.0.0 --port 10000
+python -m llava.serve.model_worker --host 0.0.0.0 --controller http://localhost:10000 --port 40000 --worker http://localhost:40000 --model-path liuhaotian/LLaVA-Lightning-MPT-7B-preview
+python -m llava.serve.gradio_web_server --controller http://localhost:10000
+```
+
+2. Training
+
+We use the same set of training dataset, and the hyperparameters as other Lightning checkpoints.
+
+```Shell
+bash ./scripts/train_lightning_mpt.sh
+```
+
 ## Acknowledgement
 
 - [Vicuna](https://github.com/lm-sys/FastChat): the codebase we built upon, and our base model Vicuna-13B that has the amazing language capabilities!

scripts/train_lightning_mpt.sh

@@ -0,0 +1,74 @@
+#!/bin/bash
+
+# Pretraining (2 hours)
+torchrun --nnodes=1 --nproc_per_node=8 --master_port=25001 \
+    llava/train/train_mem.py \
+    --model_name_or_path mosaicml/mpt-7b-chat \
+    --version v1 \
+    --data_path /path/to/blip_laion_cc_sbu_558k.json \
+    --image_folder /path/to/blip_laion_cc_sbu_558k \
+    --vision_tower openai/clip-vit-large-patch14 \
+    --tune_mm_mlp_adapter True \
+    --mm_vision_select_layer -2 \
+    --mm_use_im_start_end \
+    --bf16 True \
+    --output_dir ./checkpoints/llava-lightning-mpt-7b-pretrain \
+    --num_train_epochs 1 \
+    --per_device_train_batch_size 16 \
+    --per_device_eval_batch_size 4 \
+    --gradient_accumulation_steps 1 \
+    --evaluation_strategy "no" \
+    --save_strategy "steps" \
+    --save_steps 2400 \
+    --save_total_limit 1 \
+    --learning_rate 2e-3 \
+    --weight_decay 0. \
+    --warmup_ratio 0.03 \
+    --lr_scheduler_type "cosine" \
+    --logging_steps 1 \
+    --tf32 True \
+    --model_max_length 2048 \
+    --gradient_checkpointing True \
+    --dataloader_num_workers 4 \
+    --lazy_preprocess True \
+    --report_to wandb
+
+# Extract projector features
+python scripts/extract_mm_projector.py \
+  --model_name_or_path ./checkpoints/llava-lightning-mpt-7b-pretrain \
+  --output ./checkpoints/mm_projector/llava-lightning-mpt-7b-pretrain.bin
+
+# Visual instruction tuning (1 hour)
+torchrun --nnodes=1 --nproc_per_node=8 --master_port=25001 \
+    llava/train/train_mem.py \
+    --model_name_or_path mosaicml/mpt-7b-chat \
+    --version v1 \
+    --data_path /path/to/llava_instruct_80k.json \
+    --image_folder /Data/haotian/coco/train2014 \
+    --vision_tower openai/clip-vit-large-patch14 \
+    --pretrain_mm_mlp_adapter ./checkpoints/mm_projector/llava-lightning-mpt-7b-pretrain.bin \
+    --mm_vision_select_layer -2 \
+    --mm_use_im_start_end True \
+    --bf16 True \
+    --output_dir ./checkpoints \
+    --num_train_epochs 1 \
+    --per_device_train_batch_size 16 \
+    --per_device_eval_batch_size 4 \
+    --gradient_accumulation_steps 1 \
+    --evaluation_strategy "no" \
+    --save_strategy "steps" \
+    --save_steps 5000 \
+    --save_total_limit 1 \
+    --learning_rate 2e-5 \
+    --weight_decay 0. \
+    --warmup_ratio 0.03 \
+    --lr_scheduler_type "cosine" \
+    --logging_steps 1 \
+    --tf32 True \
+    --fsdp "full_shard auto_wrap" \
+    --fsdp_transformer_layer_cls_to_wrap 'LlamaDecoderLayer' \
+    --model_max_length 2048 \
+    --gradient_checkpointing True \
+    --dataloader_num_workers 4 \
+    --lazy_preprocess True \
+    --report_to wandb