initial commit

llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain.py

@@ -0,0 +1,214 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import torch
+from mmengine.dataset import DefaultSampler
+from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
+                            LoggerHook, ParamSchedulerHook)
+from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
+from torch.optim import AdamW
+from transformers import (AutoModelForCausalLM, AutoTokenizer,
+                          BitsAndBytesConfig, CLIPImageProcessor,
+                          CLIPVisionModel)
+
+from xtuner.dataset import LLaVADataset
+from xtuner.dataset.collate_fns import default_collate_fn
+from xtuner.dataset.map_fns import llava_map_fn, template_map_fn_factory
+from xtuner.engine.hooks import DatasetInfoHook, EvaluateChatHook
+from xtuner.engine.runner import TrainLoop
+from xtuner.model import LLaVAModel
+from xtuner.utils import PROMPT_TEMPLATE
+
+#######################################################################
+#                          PART 1  Settings                           #
+#######################################################################
+# Model
+llm_name_or_path = 'internlm/internlm2-chat-1_8b'
+visual_encoder_name_or_path = 'openai/clip-vit-large-patch14-336'
+
+# Data
+data_root = './llava_data/'
+data_path = data_root + 'LLaVA-Pretrain/blip_laion_cc_sbu_558k.json'
+image_folder = data_root + 'LLaVA-Pretrain/images'
+prompt_template = PROMPT_TEMPLATE.internlm2_chat
+max_length = int(2048 - (336 / 14)**2)
+
+# Scheduler & Optimizer
+batch_size = 40  # per_device
+accumulative_counts = 7
+dataloader_num_workers = 2
+max_epochs = 1
+optim_type = AdamW
+lr = 1e-3
+betas = (0.9, 0.999)
+weight_decay = 0
+max_norm = 1  # grad clip
+warmup_ratio = 0.03
+
+# Save
+save_steps = 500
+save_total_limit = 2  # Maximum checkpoints to keep (-1 means unlimited)
+
+# Evaluate the generation performance during the training
+evaluation_freq = 500
+SYSTEM = ''
+evaluation_images = 'https://llava-vl.github.io/static/images/view.jpg'
+evaluation_inputs = ['请描述一下这张照片', 'Please describe this picture']
+
+#######################################################################
+#            PART 2  Model & Tokenizer & Image Processor              #
+#######################################################################
+tokenizer = dict(
+    type=AutoTokenizer.from_pretrained,
+    pretrained_model_name_or_path=llm_name_or_path,
+    trust_remote_code=True,
+    padding_side='right')
+
+image_processor = dict(
+    type=CLIPImageProcessor.from_pretrained,
+    pretrained_model_name_or_path=visual_encoder_name_or_path,
+    trust_remote_code=True)
+
+model = dict(
+    type=LLaVAModel,
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm=dict(
+        type=AutoModelForCausalLM.from_pretrained,
+        pretrained_model_name_or_path=llm_name_or_path,
+        trust_remote_code=True,
+        torch_dtype=torch.float16,
+        quantization_config=dict(
+            type=BitsAndBytesConfig,
+            load_in_4bit=True,
+            load_in_8bit=False,
+            llm_int8_threshold=6.0,
+            llm_int8_has_fp16_weight=False,
+            bnb_4bit_compute_dtype=torch.float16,
+            bnb_4bit_use_double_quant=True,
+            bnb_4bit_quant_type='nf4')),
+    visual_encoder=dict(
+        type=CLIPVisionModel.from_pretrained,
+        pretrained_model_name_or_path=visual_encoder_name_or_path))
+
+#######################################################################
+#                      PART 3  Dataset & Dataloader                   #
+#######################################################################
+llava_dataset = dict(
+    type=LLaVADataset,
+    data_path=data_path,
+    image_folder=image_folder,
+    tokenizer=tokenizer,
+    image_processor=image_processor,
+    dataset_map_fn=llava_map_fn,
+    template_map_fn=dict(
+        type=template_map_fn_factory, template=prompt_template),
+    max_length=max_length,
+    pad_image_to_square=False)
+
+train_dataloader = dict(
+    batch_size=batch_size,
+    num_workers=dataloader_num_workers,
+    dataset=llava_dataset,
+    sampler=dict(type=DefaultSampler, shuffle=True),
+    collate_fn=dict(type=default_collate_fn))
+
+#######################################################################
+#                    PART 4  Scheduler & Optimizer                    #
+#######################################################################
+# optimizer
+optim_wrapper = dict(
+    type=AmpOptimWrapper,
+    optimizer=dict(
+        type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
+    clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
+    accumulative_counts=accumulative_counts,
+    loss_scale='dynamic',
+    dtype='float16')
+
+# learning policy
+# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md  # noqa: E501
+param_scheduler = [
+    dict(
+        type=LinearLR,
+        start_factor=1e-5,
+        by_epoch=True,
+        begin=0,
+        end=warmup_ratio * max_epochs,
+        convert_to_iter_based=True),
+    dict(
+        type=CosineAnnealingLR,
+        eta_min=0.0,
+        by_epoch=True,
+        begin=warmup_ratio * max_epochs,
+        end=max_epochs,
+        convert_to_iter_based=True)
+]
+
+# train, val, test setting
+train_cfg = dict(type=TrainLoop, max_epochs=max_epochs)
+
+#######################################################################
+#                           PART 5  Runtime                           #
+#######################################################################
+# Log the dialogue periodically during the training process, optional
+custom_hooks = [
+    dict(type=DatasetInfoHook, tokenizer=tokenizer),
+    dict(
+        type=EvaluateChatHook,
+        tokenizer=tokenizer,
+        image_processor=image_processor,
+        every_n_iters=evaluation_freq,
+        evaluation_inputs=evaluation_inputs,
+        evaluation_images=evaluation_images,
+        system=SYSTEM,
+        prompt_template=prompt_template)
+]
+
+# configure default hooks
+default_hooks = dict(
+    # record the time of every iteration.
+    timer=dict(type=IterTimerHook),
+    # print log every 10 iterations.
+    logger=dict(type=LoggerHook, log_metric_by_epoch=False, interval=10),
+    # enable the parameter scheduler.
+    param_scheduler=dict(type=ParamSchedulerHook),
+    # save checkpoint per `save_steps`.
+    checkpoint=dict(
+        type=CheckpointHook,
+        by_epoch=False,
+        interval=save_steps,
+        max_keep_ckpts=save_total_limit),
+    # set sampler seed in distributed evrionment.
+    sampler_seed=dict(type=DistSamplerSeedHook),
+)
+
+# configure environment
+env_cfg = dict(
+    # whether to enable cudnn benchmark
+    cudnn_benchmark=False,
+    # set multi process parameters
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
+    # set distributed parameters
+    dist_cfg=dict(backend='nccl'),
+)
+
+# set visualizer
+from mmengine.visualization import Visualizer, TensorboardVisBackend
+visualizer = dict(
+    type=Visualizer,
+    vis_backends=[dict(type=TensorboardVisBackend)]
+)
+
+# set log level
+log_level = 'INFO'
+
+# load from which checkpoint
+load_from = None
+
+# whether to resume training from the loaded checkpoint
+resume = False
+
+# Defaults to use random seed and disable `deterministic`
+randomness = dict(seed=None, deterministic=False)
+
+# set log processor
+log_processor = dict(by_epoch=False)

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_10499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake tahoe, california<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake lakelake in the mountains<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_10999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at the lake with a boat on it<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at a lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_11499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at the lake with a boat on the shore<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_11999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_12499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at the lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_12999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_13499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_1499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake in the mountains with a view of the water<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake in the mountains with a view of the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_1999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a wooden pier on a lake in the mountains<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a wooden pier with a boat on the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_2499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat on the water<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock in the lake with a boat on the shore<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_2999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat on the water<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat on the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_3499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat and a dock in the background<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat and a boat dock in the background<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_3999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock with a boat and a lake in the background<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock with a boat on the water and a bridge over it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_4499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a pier on a lake with a boat on the pier<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake with a boat on the shore<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a wooden bridge in the mountains with a view of the lake<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a wooden bridge in the mountains with a view of the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_4999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at the lake with a boat and a boat house<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at the lake with a boat on the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_5499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock and pier on a lake with a boat<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock with a boat on the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_5999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at a lake with a boat on the water<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at a lake with a boat on the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_6499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake with a boat on the shore<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake with a boat on the shore<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_6999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat and a dock<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat and a dock<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_7499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat on the shore<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat and a dock<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_7999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on the shore<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on the shore<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_8499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat on the shore<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock on a lake with a boat on the shore<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_8999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on the shore photo<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat on the shore photo<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_9499.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at the lake with a boat on the water<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at a lake with a boat on it<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake in the mountains with a mountain range in the background<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a bridge over a lake in the mountains with the sun setting over the water<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/eval_outputs_iter_9999.txt

@@ -0,0 +1,16 @@
+Eval output 1:
+<|im_start|>user
+<image>
+请描述一下这张照片<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat and a mountain in the background<|im_end|>
+
+
+Eval output 2:
+<|im_start|>user
+<image>
+Please describe this picture<|im_end|>
+<|im_start|>assistant
+<s>a dock at lake with a boat and a mountain in the background<|im_end|>
+
+

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/20240220_050613/vis_data/events.out.tfevents.1708405574.b447a125019d.77998.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:863398f4d30a127fd2f17932e0255bfa5bfa92e3d4c6d0ff923551593d2ceb9e
+size 374431

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/iter_13500.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0232b6c3371d3eea84f49847dbe95ea31d06a597ac58be51fb665d91b4af35b
+size 75549943

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/iter_13500.pth/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49a35652d7c91ea73e7e30eb753801f822f4e6f59b12406dbbcbd2dd6fa419a4
+size 13577563

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/iter_13954.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:853fcd7c7865d95e9016b23b703eb1ff1a80448f63342b28c2163b55b243d8f4
+size 75549943

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/iter_13954.pth/mp_rank_00_model_states.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:680d780055ad0f155550e80fc19719369865df98d9faa5ab5c4badbf205a902d
+size 13610395

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/last_checkpoint

@@ -0,0 +1 @@
+/workspace/work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain_copy/iter_13954.pth

work_dirs/llava_internlm2_chat_1_8b_clip_vit_large_p14_336_e1_gpu1_pretrain/zero_to_fp32.py

@@ -0,0 +1,604 @@
+#!/usr/bin/env python
+
+# Copyright (c) Microsoft Corporation.
+# SPDX-License-Identifier: Apache-2.0
+
+# DeepSpeed Team
+
+# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
+# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
+# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
+# application.
+#
+# example: python zero_to_fp32.py . pytorch_model.bin
+
+import argparse
+import torch
+import glob
+import math
+import os
+import re
+from collections import OrderedDict
+from dataclasses import dataclass
+
+# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
+# DeepSpeed data structures it has to be available in the current python environment.
+from deepspeed.utils import logger
+from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
+                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
+                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
+
+
+@dataclass
+class zero_model_state:
+    buffers: dict()
+    param_shapes: dict()
+    shared_params: list
+    ds_version: int
+    frozen_param_shapes: dict()
+    frozen_param_fragments: dict()
+
+
+debug = 0
+
+# load to cpu
+device = torch.device('cpu')
+
+
+def atoi(text):
+    return int(text) if text.isdigit() else text
+
+
+def natural_keys(text):
+    '''
+    alist.sort(key=natural_keys) sorts in human order
+    http://nedbatchelder.com/blog/200712/human_sorting.html
+    (See Toothy's implementation in the comments)
+    '''
+    return [atoi(c) for c in re.split(r'(\d+)', text)]
+
+
+def get_model_state_file(checkpoint_dir, zero_stage):
+    if not os.path.isdir(checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
+
+    # there should be only one file
+    if zero_stage <= 2:
+        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
+    elif zero_stage == 3:
+        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
+
+    if not os.path.exists(file):
+        raise FileNotFoundError(f"can't find model states file at '{file}'")
+
+    return file
+
+
+def get_checkpoint_files(checkpoint_dir, glob_pattern):
+    # XXX: need to test that this simple glob rule works for multi-node setup too
+    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
+
+    if len(ckpt_files) == 0:
+        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
+
+    return ckpt_files
+
+
+def get_optim_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
+
+
+def get_model_state_files(checkpoint_dir):
+    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
+
+
+def parse_model_states(files):
+    zero_model_states = []
+    for file in files:
+        state_dict = torch.load(file, map_location=device)
+
+        if BUFFER_NAMES not in state_dict:
+            raise ValueError(f"{file} is not a model state checkpoint")
+        buffer_names = state_dict[BUFFER_NAMES]
+        if debug:
+            print("Found buffers:", buffer_names)
+
+        # recover just the buffers while restoring them to fp32 if they were saved in fp16
+        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
+        param_shapes = state_dict[PARAM_SHAPES]
+
+        # collect parameters that are included in param_shapes
+        param_names = []
+        for s in param_shapes:
+            for name in s.keys():
+                param_names.append(name)
+
+        # update with frozen parameters
+        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
+        if frozen_param_shapes is not None:
+            if debug:
+                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
+            param_names += list(frozen_param_shapes.keys())
+
+        # handle shared params
+        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
+
+        ds_version = state_dict.get(DS_VERSION, None)
+
+        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
+
+        z_model_state = zero_model_state(buffers=buffers,
+                                         param_shapes=param_shapes,
+                                         shared_params=shared_params,
+                                         ds_version=ds_version,
+                                         frozen_param_shapes=frozen_param_shapes,
+                                         frozen_param_fragments=frozen_param_fragments)
+        zero_model_states.append(z_model_state)
+
+    return zero_model_states
+
+
+def parse_optim_states(files, ds_checkpoint_dir):
+
+    total_files = len(files)
+    state_dicts = []
+    for f in files:
+        state_dict = torch.load(f, map_location=device)
+        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
+        # and also handle the case where it was already removed by another helper script
+        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
+        state_dicts.append(state_dict)
+
+    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
+        raise ValueError(f"{files[0]} is not a zero checkpoint")
+    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
+    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
+
+    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
+    # parameters can be different from data parallelism for non-expert parameters. So we can just
+    # use the max of the partition_count to get the dp world_size.
+
+    if type(world_size) is list:
+        world_size = max(world_size)
+
+    if world_size != total_files:
+        raise ValueError(
+            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
+            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
+        )
+
+    # the groups are named differently in each stage
+    if zero_stage <= 2:
+        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
+    elif zero_stage == 3:
+        fp32_groups_key = FP32_FLAT_GROUPS
+    else:
+        raise ValueError(f"unknown zero stage {zero_stage}")
+
+    if zero_stage <= 2:
+        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
+    elif zero_stage == 3:
+        # if there is more than one param group, there will be multiple flattened tensors - one
+        # flattened tensor per group - for simplicity merge them into a single tensor
+        #
+        # XXX: could make the script more memory efficient for when there are multiple groups - it
+        # will require matching the sub-lists of param_shapes for each param group flattened tensor
+
+        fp32_flat_groups = [
+            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
+        ]
+
+    return zero_stage, world_size, fp32_flat_groups
+
+
+def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
+    """
+    Returns fp32 state_dict reconstructed from ds checkpoint
+
+    Args:
+        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
+
+    """
+    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
+
+    optim_files = get_optim_files(ds_checkpoint_dir)
+    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
+    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
+
+    model_files = get_model_state_files(ds_checkpoint_dir)
+
+    zero_model_states = parse_model_states(model_files)
+    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
+
+    if zero_stage <= 2:
+        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+    elif zero_stage == 3:
+        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                                          exclude_frozen_parameters)
+
+
+def _zero2_merge_frozen_params(state_dict, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
+
+    if debug:
+        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
+        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        state_dict[name] = frozen_param_fragments[name]
+
+        if debug:
+            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _has_callable(obj, fn):
+    attr = getattr(obj, fn, None)
+    return callable(attr)
+
+
+def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+
+    # Reconstruction protocol:
+    #
+    # XXX: document this
+
+    if debug:
+        for i in range(world_size):
+            for j in range(len(fp32_flat_groups[0])):
+                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
+
+    # XXX: memory usage doubles here (zero2)
+    num_param_groups = len(fp32_flat_groups[0])
+    merged_single_partition_of_fp32_groups = []
+    for i in range(num_param_groups):
+        merged_partitions = [sd[i] for sd in fp32_flat_groups]
+        full_single_fp32_vector = torch.cat(merged_partitions, 0)
+        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
+    avail_numel = sum(
+        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
+
+    if debug:
+        wanted_params = sum([len(shapes) for shapes in param_shapes])
+        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
+        # not asserting if there is a mismatch due to possible padding
+        print(f"Have {avail_numel} numels to process.")
+        print(f"Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    total_numel = 0
+    total_params = 0
+    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
+        offset = 0
+        avail_numel = full_single_fp32_vector.numel()
+        for name, shape in shapes.items():
+
+            unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
+            total_numel += unpartitioned_numel
+            total_params += 1
+
+            if debug:
+                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
+            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
+            offset += unpartitioned_numel
+
+        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
+        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
+        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
+        # live optimizer object, so we are checking that the numbers are within the right range
+        align_to = 2 * world_size
+
+        def zero2_align(x):
+            return align_to * math.ceil(x / align_to)
+
+        if debug:
+            print(f"original offset={offset}, avail_numel={avail_numel}")
+
+        offset = zero2_align(offset)
+        avail_numel = zero2_align(avail_numel)
+
+        if debug:
+            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
+
+        # Sanity check
+        if offset != avail_numel:
+            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero2_merge_frozen_params(state_dict, zero_model_states)
+
+    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def zero3_partitioned_param_info(unpartitioned_numel, world_size):
+    remainder = unpartitioned_numel % world_size
+    padding_numel = (world_size - remainder) if remainder else 0
+    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
+    return partitioned_numel, padding_numel
+
+
+def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
+    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
+        return
+
+    if debug:
+        for i in range(world_size):
+            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
+            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
+
+        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
+        wanted_params = len(frozen_param_shapes)
+        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
+        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
+        print(f'Frozen params: Have {avail_numel} numels to process.')
+        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
+
+    total_params = 0
+    total_numel = 0
+    for name, shape in zero_model_states[0].frozen_param_shapes.items():
+        total_params += 1
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+
+        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
+        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
+    param_shapes = zero_model_states[0].param_shapes
+    avail_numel = fp32_flat_groups[0].numel() * world_size
+    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
+    # param, re-consolidating each param, while dealing with padding if any
+
+    # merge list of dicts, preserving order
+    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
+
+    if debug:
+        for i in range(world_size):
+            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
+
+        wanted_params = len(param_shapes)
+        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
+        # not asserting if there is a mismatch due to possible padding
+        avail_numel = fp32_flat_groups[0].numel() * world_size
+        print(f"Trainable params: Have {avail_numel} numels to process.")
+        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
+
+    # params
+    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
+    # out-of-core computing solution
+    offset = 0
+    total_numel = 0
+    total_params = 0
+    for name, shape in param_shapes.items():
+
+        unpartitioned_numel = shape.numel()
+        total_numel += unpartitioned_numel
+        total_params += 1
+
+        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
+
+        if debug:
+            print(
+                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
+            )
+
+        # XXX: memory usage doubles here
+        state_dict[name] = torch.cat(
+            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
+            0).narrow(0, 0, unpartitioned_numel).view(shape)
+        offset += partitioned_numel
+
+    offset *= world_size
+
+    # Sanity check
+    if offset != avail_numel:
+        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
+
+    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
+
+
+def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
+                                               exclude_frozen_parameters):
+    state_dict = OrderedDict()
+
+    # buffers
+    buffers = zero_model_states[0].buffers
+    state_dict.update(buffers)
+    if debug:
+        print(f"added {len(buffers)} buffers")
+
+    if not exclude_frozen_parameters:
+        _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
+
+    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
+
+    # recover shared parameters
+    for pair in zero_model_states[0].shared_params:
+        if pair[1] in state_dict:
+            state_dict[pair[0]] = state_dict[pair[1]]
+
+    return state_dict
+
+
+def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
+    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
+    via a model hub.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+
+    Returns:
+        - pytorch ``state_dict``
+
+    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
+    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
+    the checkpoint.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
+        # do the training and checkpoint saving
+        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
+        model = model.cpu() # move to cpu
+        model.load_state_dict(state_dict)
+        # submit to model hub or save the model to share with others
+
+    In this example the ``model`` will no longer be usable in the deepspeed context of the same
+    application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
+
+    """
+    if tag is None:
+        latest_path = os.path.join(checkpoint_dir, 'latest')
+        if os.path.isfile(latest_path):
+            with open(latest_path, 'r') as fd:
+                tag = fd.read().strip()
+        else:
+            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
+
+    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
+
+    if not os.path.isdir(ds_checkpoint_dir):
+        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
+
+    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
+
+
+def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None, exclude_frozen_parameters=False):
+    """
+    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
+    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
+
+    Args:
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+        - ``exclude_frozen_parameters``: exclude frozen parameters
+    """
+
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag, exclude_frozen_parameters)
+    print(f"Saving fp32 state dict to {output_file}")
+    torch.save(state_dict, output_file)
+
+
+def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
+    """
+    1. Put the provided model to cpu
+    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
+    3. Load it into the provided model
+
+    Args:
+        - ``model``: the model object to update
+        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
+        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
+
+    Returns:
+        - ``model`: modified model
+
+    Make sure you have plenty of CPU memory available before you call this function. If you don't
+    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
+    conveniently placed for you in the checkpoint folder.
+
+    A typical usage might be ::
+
+        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
+        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
+        # submit to model hub or save the model to share with others
+
+    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
+    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
+    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
+
+    """
+    logger.info(f"Extracting fp32 weights")
+    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
+
+    logger.info(f"Overwriting model with fp32 weights")
+    model = model.cpu()
+    model.load_state_dict(state_dict, strict=False)
+
+    return model
+
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument("checkpoint_dir",
+                        type=str,
+                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
+    parser.add_argument(
+        "output_file",
+        type=str,
+        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
+    parser.add_argument("-t",
+                        "--tag",
+                        type=str,
+                        default=None,
+                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
+    parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
+    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
+    args = parser.parse_args()
+
+    debug = args.debug
+
+    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
+                                               args.output_file,
+                                               tag=args.tag,
+                                               exclude_frozen_parameters=args.exclude_frozen_parameters)