Upload 7 files

README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: microsoft/phi-2
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.11.1

adapter_config.json

@@ -0,0 +1,33 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "microsoft/phi-2",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 256,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 128,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "dense",
+    "fc2",
+    "fc1",
+    "q_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8168c2b2a4b2bfef0baad52c096fb0a3de08329dc55bc2adf91f5a23da8cef3
+size 377539280

config.json

@@ -0,0 +1,42 @@
+{
+  "_name_or_path": "microsoft/phi-2",
+  "architectures": [
+    "PhiForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.0,
+  "eos_token_id": 50256,
+  "freeze_mm_mlp_adapter": false,
+  "hidden_act": "gelu_new",
+  "hidden_size": 2560,
+  "image_aspect_ratio": "pad",
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "mm_hidden_size": 1152,
+  "mm_projector_lr": 2e-05,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
+  "model_type": "bunny-phi",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 32,
+  "partial_rotary_factor": 0.4,
+  "qk_layernorm": false,
+  "resid_pdrop": 0.1,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 2048,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "float16",
+  "transformers_version": "4.41.2",
+  "tune_mm_mlp_adapter": false,
+  "unfreeze_vision_tower": false,
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_s2": false,
+  "vocab_size": 51200
+}

log.txt

@@ -0,0 +1,257 @@
+[2024-06-14 02:12:51,883] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io: please install the libaio-dev package with apt
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.3
+ [WARNING]  using untested triton version (2.3.0), only 1.0.0 is known to be compatible
+[2024-06-14 02:12:52,556] [WARNING] [runner.py:202:fetch_hostfile] Unable to find hostfile, will proceed with training with local resources only.
+[2024-06-14 02:12:52,556] [INFO] [runner.py:568:main] cmd = /home/robert/miniconda3/envs/bunny/bin/python -u -m deepspeed.launcher.launch --world_info=eyJsb2NhbGhvc3QiOiBbMCwgMV19 --master_addr=127.0.0.1 --master_port=29500 --enable_each_rank_log=None bunny/train/train.py --lora_enable True --lora_r 128 --lora_alpha 256 --mm_projector_lr 2e-5 --deepspeed ./script/deepspeed/zero3.json --model_name_or_path microsoft/phi-2 --model_type phi-2 --version bunny --data_path ./data/finetune/bunny_695k.json --image_folder ./data/finetune/images --vision_tower google/siglip-so400m-patch14-384 --pretrain_mm_mlp_adapter ./checkpoints-pretrain/bunny-phi-2-pretrain/mm_projector.bin --mm_projector_type mlp2x_gelu --image_aspect_ratio pad --group_by_modality_length False --bf16 True --output_dir ./checkpoints-phi-2/bunny-lora-phi-2 --num_train_epochs 1 --per_device_train_batch_size 8 --per_device_eval_batch_size 4 --gradient_accumulation_steps 2 --evaluation_strategy no --save_strategy steps --save_steps 500 --save_total_limit 1 --learning_rate 2e-4 --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 none
+[2024-06-14 02:12:53,999] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io: please install the libaio-dev package with apt
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.3
+ [WARNING]  using untested triton version (2.3.0), only 1.0.0 is known to be compatible
+[2024-06-14 02:12:54,677] [INFO] [launch.py:146:main] WORLD INFO DICT: {'localhost': [0, 1]}
+[2024-06-14 02:12:54,677] [INFO] [launch.py:152:main] nnodes=1, num_local_procs=2, node_rank=0
+[2024-06-14 02:12:54,677] [INFO] [launch.py:163:main] global_rank_mapping=defaultdict(<class 'list'>, {'localhost': [0, 1]})
+[2024-06-14 02:12:54,677] [INFO] [launch.py:164:main] dist_world_size=2
+[2024-06-14 02:12:54,677] [INFO] [launch.py:168:main] Setting CUDA_VISIBLE_DEVICES=0,1
+[2024-06-14 02:12:54,678] [INFO] [launch.py:256:main] process 1400331 spawned with command: ['/home/robert/miniconda3/envs/bunny/bin/python', '-u', 'bunny/train/train.py', '--local_rank=0', '--lora_enable', 'True', '--lora_r', '128', '--lora_alpha', '256', '--mm_projector_lr', '2e-5', '--deepspeed', './script/deepspeed/zero3.json', '--model_name_or_path', 'microsoft/phi-2', '--model_type', 'phi-2', '--version', 'bunny', '--data_path', './data/finetune/bunny_695k.json', '--image_folder', './data/finetune/images', '--vision_tower', 'google/siglip-so400m-patch14-384', '--pretrain_mm_mlp_adapter', './checkpoints-pretrain/bunny-phi-2-pretrain/mm_projector.bin', '--mm_projector_type', 'mlp2x_gelu', '--image_aspect_ratio', 'pad', '--group_by_modality_length', 'False', '--bf16', 'True', '--output_dir', './checkpoints-phi-2/bunny-lora-phi-2', '--num_train_epochs', '1', '--per_device_train_batch_size', '8', '--per_device_eval_batch_size', '4', '--gradient_accumulation_steps', '2', '--evaluation_strategy', 'no', '--save_strategy', 'steps', '--save_steps', '500', '--save_total_limit', '1', '--learning_rate', '2e-4', '--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', 'none']
+[2024-06-14 02:12:54,678] [INFO] [launch.py:256:main] process 1400332 spawned with command: ['/home/robert/miniconda3/envs/bunny/bin/python', '-u', 'bunny/train/train.py', '--local_rank=1', '--lora_enable', 'True', '--lora_r', '128', '--lora_alpha', '256', '--mm_projector_lr', '2e-5', '--deepspeed', './script/deepspeed/zero3.json', '--model_name_or_path', 'microsoft/phi-2', '--model_type', 'phi-2', '--version', 'bunny', '--data_path', './data/finetune/bunny_695k.json', '--image_folder', './data/finetune/images', '--vision_tower', 'google/siglip-so400m-patch14-384', '--pretrain_mm_mlp_adapter', './checkpoints-pretrain/bunny-phi-2-pretrain/mm_projector.bin', '--mm_projector_type', 'mlp2x_gelu', '--image_aspect_ratio', 'pad', '--group_by_modality_length', 'False', '--bf16', 'True', '--output_dir', './checkpoints-phi-2/bunny-lora-phi-2', '--num_train_epochs', '1', '--per_device_train_batch_size', '8', '--per_device_eval_batch_size', '4', '--gradient_accumulation_steps', '2', '--evaluation_strategy', 'no', '--save_strategy', 'steps', '--save_steps', '500', '--save_total_limit', '1', '--learning_rate', '2e-4', '--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', 'none']
+[2024-06-14 02:12:59,036] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2024-06-14 02:12:59,043] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io: please install the libaio-dev package with apt
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  async_io: please install the libaio-dev package with apt
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.3
+ [WARNING]  using untested triton version (2.3.0), only 1.0.0 is known to be compatible
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.3
+ [WARNING]  using untested triton version (2.3.0), only 1.0.0 is known to be compatible
+[2024-06-14 02:12:59,215] [INFO] [comm.py:637:init_distributed] cdb=None
+[2024-06-14 02:12:59,226] [INFO] [comm.py:637:init_distributed] cdb=None
+[2024-06-14 02:12:59,226] [INFO] [comm.py:668:init_distributed] Initializing TorchBackend in DeepSpeed with backend nccl
+[2024-06-14 02:13:01,597] [INFO] [partition_parameters.py:345:__exit__] finished initializing model - num_params = 452, num_elems = 2.78B
+Adding LoRA adapters...
+[2024-06-14 02:13:06,587] [INFO] [partition_parameters.py:345:__exit__] finished initializing model - num_params = 900, num_elems = 3.21B
+Formatting inputs...Skip in lazy mode
+Formatting inputs...Skip in lazy mode
+Parameter Offload: Total persistent parameters: 1324480 in 540 params
+{'loss': 1.4425, 'grad_norm': 0.942700758245986, 'learning_rate': 4.723452001250683e-08, 'epoch': 0.99}
+{'loss': 1.0698, 'grad_norm': 5.386742649487362, 'learning_rate': 4.677708118508983e-08, 'epoch': 0.99}
+{'loss': 1.0348, 'grad_norm': 0.6414032462502491, 'learning_rate': 4.632186764480517e-08, 'epoch': 0.99}
+{'loss': 1.4007, 'grad_norm': 0.8399374153966813, 'learning_rate': 4.586887940176699e-08, 'epoch': 0.99}
+{'loss': 1.013, 'grad_norm': 1.1784298059039426, 'learning_rate': 4.54181164660894e-08, 'epoch': 0.99}
+{'loss': 1.4703, 'grad_norm': 0.9445735664284673, 'learning_rate': 4.496957884778663e-08, 'epoch': 0.99}
+{'loss': 1.357, 'grad_norm': 0.8032126350975317, 'learning_rate': 4.452326655685068e-08, 'epoch': 0.99}
+{'loss': 1.0204, 'grad_norm': 1.1730182937411677, 'learning_rate': 4.4079179603218055e-08, 'epoch': 0.99}
+{'loss': 1.3475, 'grad_norm': 1.2582671967211982, 'learning_rate': 4.363731799676973e-08, 'epoch': 0.99}
+{'loss': 1.0274, 'grad_norm': 1.7294591582235157, 'learning_rate': 4.319768174735339e-08, 'epoch': 0.99}
+{'loss': 1.1504, 'grad_norm': 1.0469817450846388, 'learning_rate': 4.2760270864738994e-08, 'epoch': 0.99}
+{'loss': 1.1542, 'grad_norm': 0.8339323652457116, 'learning_rate': 4.2325085358696504e-08, 'epoch': 0.99}
+{'loss': 1.4128, 'grad_norm': 0.9148113247697388, 'learning_rate': 4.189212523888486e-08, 'epoch': 0.99}
+{'loss': 1.0067, 'grad_norm': 1.072223405558702, 'learning_rate': 4.14613905149519e-08, 'epoch': 0.99}
+{'loss': 1.3823, 'grad_norm': 0.7933852853971807, 'learning_rate': 4.103288119648996e-08, 'epoch': 0.99}
+{'loss': 1.1281, 'grad_norm': 0.7261390721292842, 'learning_rate': 4.060659729303584e-08, 'epoch': 0.99}
+{'loss': 1.1811, 'grad_norm': 0.9003791199373986, 'learning_rate': 4.018253881409306e-08, 'epoch': 0.99}
+{'loss': 1.3176, 'grad_norm': 1.1947955753146302, 'learning_rate': 3.976070576908742e-08, 'epoch': 0.99}
+{'loss': 1.2388, 'grad_norm': 6.846984030349887, 'learning_rate': 3.934109816741138e-08, 'epoch': 0.99}
+{'loss': 1.2566, 'grad_norm': 2.6987845022894623, 'learning_rate': 3.8923716018401943e-08, 'epoch': 0.99}
+{'loss': 0.9042, 'grad_norm': 0.7541139281044368, 'learning_rate': 3.850855933137387e-08, 'epoch': 0.99}
+{'loss': 1.243, 'grad_norm': 3.311043853373493, 'learning_rate': 3.8095628115553114e-08, 'epoch': 0.99}
+{'loss': 1.147, 'grad_norm': 0.9586566637745106, 'learning_rate': 3.7684922380132324e-08, 'epoch': 0.99}
+{'loss': 1.0982, 'grad_norm': 0.9359946330825151, 'learning_rate': 3.727644213424863e-08, 'epoch': 0.99}
+{'loss': 1.2446, 'grad_norm': 0.9731633495925633, 'learning_rate': 3.687018738701697e-08, 'epoch': 0.99}
+{'loss': 1.4469, 'grad_norm': 1.4342280541565777, 'learning_rate': 3.646615814747456e-08, 'epoch': 0.99}
+{'loss': 1.2553, 'grad_norm': 1.3033387849211402, 'learning_rate': 3.6064354424603096e-08, 'epoch': 0.99}
+{'loss': 1.5003, 'grad_norm': 0.7709586580426135, 'learning_rate': 3.566477622736208e-08, 'epoch': 0.99}
+{'loss': 1.0965, 'grad_norm': 1.0000894519468093, 'learning_rate': 3.526742356463331e-08, 'epoch': 0.99}
+{'loss': 1.4456, 'grad_norm': 1.1231825108623885, 'learning_rate': 3.487229644528744e-08, 'epoch': 0.99}
+{'loss': 1.2185, 'grad_norm': 1.392070634372661, 'learning_rate': 3.447939487809526e-08, 'epoch': 0.99}
+{'loss': 1.0054, 'grad_norm': 0.9414901630248458, 'learning_rate': 3.408871887182752e-08, 'epoch': 0.99}
+{'loss': 1.3494, 'grad_norm': 0.784432037613797, 'learning_rate': 3.370026843515506e-08, 'epoch': 0.99}
+{'loss': 1.4281, 'grad_norm': 1.1215894921327614, 'learning_rate': 3.3314043576748724e-08, 'epoch': 0.99}
+{'loss': 1.481, 'grad_norm': 0.7832511915640359, 'learning_rate': 3.293004430520163e-08, 'epoch': 0.99}
+{'loss': 1.0621, 'grad_norm': 0.8330566484962368, 'learning_rate': 3.2548270629051415e-08, 'epoch': 0.99}
+{'loss': 1.1758, 'grad_norm': 1.0721497039114856, 'learning_rate': 3.216872255681347e-08, 'epoch': 0.99}
+{'loss': 1.5477, 'grad_norm': 0.7260520718611978, 'learning_rate': 3.1791400096925495e-08, 'epoch': 0.99}
+{'loss': 1.5049, 'grad_norm': 1.1098146309429233, 'learning_rate': 3.141630325780298e-08, 'epoch': 0.99}
+{'loss': 1.0742, 'grad_norm': 1.8347651432396874, 'learning_rate': 3.10434320477726e-08, 'epoch': 0.99}
+{'loss': 1.1936, 'grad_norm': 2.2776015845200916, 'learning_rate': 3.067278647516103e-08, 'epoch': 0.99}
+{'loss': 1.2198, 'grad_norm': 1.0355190821448266, 'learning_rate': 3.030436654820612e-08, 'epoch': 0.99}
+{'loss': 1.1147, 'grad_norm': 2.4355829952567385, 'learning_rate': 2.993817227511242e-08, 'epoch': 0.99}
+{'loss': 1.3694, 'grad_norm': 0.984979819700457, 'learning_rate': 2.9574203664040068e-08, 'epoch': 0.99}
+{'loss': 1.4798, 'grad_norm': 1.60805779871656, 'learning_rate': 2.921246072307149e-08, 'epoch': 0.99}
+{'loss': 1.2715, 'grad_norm': 1.3650514569339907, 'learning_rate': 2.8852943460289107e-08, 'epoch': 0.99}
+{'loss': 1.3787, 'grad_norm': 1.2625718262332608, 'learning_rate': 2.8495651883675424e-08, 'epoch': 0.99}
+{'loss': 1.3049, 'grad_norm': 0.8021584725122108, 'learning_rate': 2.8140586001190737e-08, 'epoch': 0.99}
+{'loss': 1.1923, 'grad_norm': 0.8168092830830008, 'learning_rate': 2.7787745820739842e-08, 'epoch': 0.99}
+{'loss': 1.1541, 'grad_norm': 0.8427335930115224, 'learning_rate': 2.743713135018311e-08, 'epoch': 0.99}
+{'loss': 1.1531, 'grad_norm': 1.2501986070685778, 'learning_rate': 2.708874259732541e-08, 'epoch': 0.99}
+{'loss': 1.3751, 'grad_norm': 0.8944408353826961, 'learning_rate': 2.6742579569916104e-08, 'epoch': 0.99}
+{'loss': 1.3265, 'grad_norm': 0.8662694939770277, 'learning_rate': 2.6398642275660135e-08, 'epoch': 0.99}
+{'loss': 1.0567, 'grad_norm': 1.0111099240739703, 'learning_rate': 2.6056930722229145e-08, 'epoch': 0.99}
+{'loss': 1.0349, 'grad_norm': 0.9690077754479983, 'learning_rate': 2.571744491721706e-08, 'epoch': 0.99}
+{'loss': 1.4695, 'grad_norm': 0.6430214364296446, 'learning_rate': 2.53801848681956e-08, 'epoch': 0.99}
+{'loss': 1.4982, 'grad_norm': 0.7420680068160221, 'learning_rate': 2.5045150582658773e-08, 'epoch': 0.99}
+{'loss': 0.8213, 'grad_norm': 3.67220224157274, 'learning_rate': 2.4712342068067273e-08, 'epoch': 0.99}
+{'loss': 1.2143, 'grad_norm': 2.50986289612842, 'learning_rate': 2.4381759331837395e-08, 'epoch': 0.99}
+{'loss': 1.2333, 'grad_norm': 0.8257838615002613, 'learning_rate': 2.4053402381318814e-08, 'epoch': 0.99}
+{'loss': 1.1157, 'grad_norm': 1.205163641885224, 'learning_rate': 2.3727271223839e-08, 'epoch': 0.99}
+{'loss': 0.9462, 'grad_norm': 0.8336985402702446, 'learning_rate': 2.3403365866647707e-08, 'epoch': 0.99}
+{'loss': 1.3943, 'grad_norm': 1.4013151797733856, 'learning_rate': 2.3081686316961393e-08, 'epoch': 0.99}
+{'loss': 1.3747, 'grad_norm': 0.7985012764218158, 'learning_rate': 2.276223258192989e-08, 'epoch': 0.99}
+{'loss': 0.9557, 'grad_norm': 1.1690954634900848, 'learning_rate': 2.2445004668680824e-08, 'epoch': 0.99}
+{'loss': 0.9685, 'grad_norm': 0.9623389466769615, 'learning_rate': 2.213000258426412e-08, 'epoch': 0.99}
+{'loss': 0.9524, 'grad_norm': 2.1843226135301275, 'learning_rate': 2.1817226335696384e-08, 'epoch': 0.99}
+{'loss': 1.0139, 'grad_norm': 0.8918643218214802, 'learning_rate': 2.1506675929938713e-08, 'epoch': 0.99}
+{'loss': 1.5443, 'grad_norm': 0.789604746189584, 'learning_rate': 2.11983513739189e-08, 'epoch': 0.99}
+{'loss': 1.1413, 'grad_norm': 1.2709132143502189, 'learning_rate': 2.089225267447592e-08, 'epoch': 0.99}
+{'loss': 1.1142, 'grad_norm': 0.9530093706340396, 'learning_rate': 2.0588379838448745e-08, 'epoch': 0.99}
+{'loss': 1.4164, 'grad_norm': 0.8921788520759392, 'learning_rate': 2.0286732872598634e-08, 'epoch': 0.99}
+{'loss': 1.1639, 'grad_norm': 2.4443203418969848, 'learning_rate': 1.9987311783631335e-08, 'epoch': 0.99}
+{'loss': 1.3764, 'grad_norm': 1.1668353021113063, 'learning_rate': 1.9690116578219286e-08, 'epoch': 0.99}
+{'loss': 1.2973, 'grad_norm': 1.010623947500593, 'learning_rate': 1.9395147262968317e-08, 'epoch': 0.99}
+{'loss': 1.2691, 'grad_norm': 0.8898451427839096, 'learning_rate': 1.9102403844462047e-08, 'epoch': 0.99}
+{'loss': 1.4245, 'grad_norm': 0.7914659980565069, 'learning_rate': 1.8811886329217487e-08, 'epoch': 0.99}
+{'loss': 1.362, 'grad_norm': 0.7832894858322704, 'learning_rate': 1.8523594723685032e-08, 'epoch': 0.99}
+{'loss': 1.3057, 'grad_norm': 0.8955949346465102, 'learning_rate': 1.8237529034292876e-08, 'epoch': 0.99}
+{'loss': 1.0109, 'grad_norm': 0.6765891435245239, 'learning_rate': 1.7953689267413698e-08, 'epoch': 0.99}
+{'loss': 1.2171, 'grad_norm': 0.970128827793069, 'learning_rate': 1.7672075429364665e-08, 'epoch': 0.99}
+{'loss': 1.161, 'grad_norm': 1.2721599430804214, 'learning_rate': 1.7392687526407437e-08, 'epoch': 0.99}
+{'loss': 1.2052, 'grad_norm': 1.1817903579639597, 'learning_rate': 1.711552556477036e-08, 'epoch': 0.99}
+{'loss': 1.065, 'grad_norm': 1.059992984071674, 'learning_rate': 1.6840589550626284e-08, 'epoch': 0.99}
+{'loss': 1.4083, 'grad_norm': 0.8051901491068411, 'learning_rate': 1.656787949009253e-08, 'epoch': 0.99}
+{'loss': 1.3522, 'grad_norm': 0.9453133488951609, 'learning_rate': 1.629739538924202e-08, 'epoch': 0.99}
+{'loss': 1.3357, 'grad_norm': 1.5231809399877834, 'learning_rate': 1.602913725408106e-08, 'epoch': 0.99}
+{'loss': 1.191, 'grad_norm': 0.7759721620339927, 'learning_rate': 1.576310509060486e-08, 'epoch': 0.99}
+{'loss': 1.2079, 'grad_norm': 0.9574306333903998, 'learning_rate': 1.5499298904730897e-08, 'epoch': 0.99}
+{'loss': 1.3609, 'grad_norm': 0.8984157351262388, 'learning_rate': 1.5237718702321157e-08, 'epoch': 0.99}
+{'loss': 1.3661, 'grad_norm': 1.2733647977353184, 'learning_rate': 1.4978364489215413e-08, 'epoch': 0.99}
+{'loss': 1.2004, 'grad_norm': 1.5219904926183827, 'learning_rate': 1.472123627117572e-08, 'epoch': 0.99}
+{'loss': 1.3272, 'grad_norm': 0.9349084111819203, 'learning_rate': 1.446633405393083e-08, 'epoch': 0.99}
+{'loss': 1.0451, 'grad_norm': 2.5778124045810222, 'learning_rate': 1.421365784314288e-08, 'epoch': 0.99}
+{'loss': 1.0895, 'grad_norm': 0.9365064782027099, 'learning_rate': 1.3963207644462906e-08, 'epoch': 0.99}
+{'loss': 1.1102, 'grad_norm': 0.8330735209837793, 'learning_rate': 1.3714983463442022e-08, 'epoch': 0.99}
+{'loss': 1.2836, 'grad_norm': 0.9561616455105378, 'learning_rate': 1.346898530562024e-08, 'epoch': 0.99}
+[2024-06-14 02:30:38,728] [WARNING] [stage3.py:2069:step] 1 pytorch allocator cache flushes since last step. this happens when there is high memory pressure and is detrimental to performance. if this is happening frequently consider adjusting settings to reduce memory consumption. If you are unable to make the cache flushes go away consider adding get_accelerator().empty_cache() calls in your training loop to ensure that all ranks flush their caches at the same time
+{'loss': 1.2363, 'grad_norm': 0.7071857362610257, 'learning_rate': 1.3225213176482065e-08, 'epoch': 0.99}
+{'loss': 1.2631, 'grad_norm': 0.8004012362110456, 'learning_rate': 1.298366708143428e-08, 'epoch': 1.0}
+{'loss': 1.5518, 'grad_norm': 0.8141940190741423, 'learning_rate': 1.2744347025861469e-08, 'epoch': 1.0}
+{'loss': 1.3571, 'grad_norm': 1.0142219264331416, 'learning_rate': 1.2507253015103804e-08, 'epoch': 1.0}
+{'loss': 1.1516, 'grad_norm': 0.9337790912361478, 'learning_rate': 1.227238505442374e-08, 'epoch': 1.0}
+{'loss': 1.5332, 'grad_norm': 1.0019159278288796, 'learning_rate': 1.203974314905043e-08, 'epoch': 1.0}
+{'loss': 0.9523, 'grad_norm': 1.3255552639470847, 'learning_rate': 1.1809327304179717e-08, 'epoch': 1.0}
+{'loss': 1.0178, 'grad_norm': 2.3569328810589836, 'learning_rate': 1.158113752492973e-08, 'epoch': 1.0}
+{'loss': 1.0026, 'grad_norm': 0.9469794474905348, 'learning_rate': 1.1355173816385289e-08, 'epoch': 1.0}
+{'loss': 1.3561, 'grad_norm': 1.0444323780434388, 'learning_rate': 1.1131436183564603e-08, 'epoch': 1.0}
+{'loss': 1.0371, 'grad_norm': 0.8986112639676435, 'learning_rate': 1.0909924631463675e-08, 'epoch': 1.0}
+{'loss': 1.2115, 'grad_norm': 2.373261525594255, 'learning_rate': 1.0690639165011895e-08, 'epoch': 1.0}
+{'loss': 1.3309, 'grad_norm': 1.122171472399203, 'learning_rate': 1.0473579789083143e-08, 'epoch': 1.0}
+{'loss': 1.1064, 'grad_norm': 1.8809967982797045, 'learning_rate': 1.0258746508517991e-08, 'epoch': 1.0}
+{'loss': 1.3553, 'grad_norm': 0.6738030569641423, 'learning_rate': 1.00461393281015e-08, 'epoch': 1.0}
+{'loss': 1.3224, 'grad_norm': 1.2808290035083636, 'learning_rate': 9.835758252552118e-09, 'epoch': 1.0}
+{'loss': 1.1095, 'grad_norm': 1.053570231478655, 'learning_rate': 9.627603286566088e-09, 'epoch': 1.0}
+{'loss': 1.3688, 'grad_norm': 1.163503084500989, 'learning_rate': 9.421674434773042e-09, 'epoch': 1.0}
+{'loss': 1.1505, 'grad_norm': 2.5903259835818186, 'learning_rate': 9.2179717017693e-09, 'epoch': 1.0}
+{'loss': 1.011, 'grad_norm': 1.3895324590981457, 'learning_rate': 9.016495092073474e-09, 'epoch': 1.0}
+{'loss': 1.1866, 'grad_norm': 1.8823726980006361, 'learning_rate': 8.817244610170861e-09, 'epoch': 1.0}
+{'loss': 1.299, 'grad_norm': 0.8503392238673622, 'learning_rate': 8.620220260513456e-09, 'epoch': 1.0}
+{'loss': 1.3522, 'grad_norm': 0.8624416934110696, 'learning_rate': 8.42542204747554e-09, 'epoch': 1.0}
+{'loss': 1.1821, 'grad_norm': 1.0867326324653512, 'learning_rate': 8.232849975398082e-09, 'epoch': 1.0}
+{'loss': 1.4199, 'grad_norm': 0.8147434696121018, 'learning_rate': 8.042504048555444e-09, 'epoch': 1.0}
+{'loss': 1.3494, 'grad_norm': 0.8654331092126935, 'learning_rate': 7.85438427121088e-09, 'epoch': 1.0}
+{'loss': 1.1523, 'grad_norm': 1.6205454650407503, 'learning_rate': 7.668490647538828e-09, 'epoch': 1.0}
+{'loss': 1.2058, 'grad_norm': 0.9633457313784664, 'learning_rate': 7.48482318166932e-09, 'epoch': 1.0}
+{'loss': 1.0053, 'grad_norm': 0.805306176189691, 'learning_rate': 7.303381877699078e-09, 'epoch': 1.0}
+{'loss': 1.1672, 'grad_norm': 2.0945837494826427, 'learning_rate': 7.124166739680416e-09, 'epoch': 1.0}
+{'loss': 1.0993, 'grad_norm': 1.1790408885617623, 'learning_rate': 6.94717777157683e-09, 'epoch': 1.0}
+{'loss': 1.3596, 'grad_norm': 3.2801472917143237, 'learning_rate': 6.772414977351816e-09, 'epoch': 1.0}
+{'loss': 1.1421, 'grad_norm': 4.62041686026284, 'learning_rate': 6.599878360891154e-09, 'epoch': 1.0}
+{'loss': 1.0906, 'grad_norm': 1.0338053298478405, 'learning_rate': 6.429567926025115e-09, 'epoch': 1.0}
+{'loss': 1.2614, 'grad_norm': 0.9214427909361785, 'learning_rate': 6.2614836765506614e-09, 'epoch': 1.0}
+{'loss': 1.2062, 'grad_norm': 0.9468374577747518, 'learning_rate': 6.095625616220346e-09, 'epoch': 1.0}
+{'loss': 1.0961, 'grad_norm': 1.3576864307028402, 'learning_rate': 5.931993748709008e-09, 'epoch': 1.0}
+{'loss': 1.263, 'grad_norm': 1.2507886070479686, 'learning_rate': 5.770588077669281e-09, 'epoch': 1.0}
+{'loss': 1.2194, 'grad_norm': 0.6878309986692126, 'learning_rate': 5.611408606698287e-09, 'epoch': 1.0}
+{'loss': 1.436, 'grad_norm': 0.8163139930922514, 'learning_rate': 5.4544553393376386e-09, 'epoch': 1.0}
+{'loss': 1.2167, 'grad_norm': 0.7109340908969963, 'learning_rate': 5.2997282790734345e-09, 'epoch': 1.0}
+{'loss': 0.9125, 'grad_norm': 1.6286097660220318, 'learning_rate': 5.147227429369572e-09, 'epoch': 1.0}
+{'loss': 0.8081, 'grad_norm': 2.8495076490148104, 'learning_rate': 4.996952793590026e-09, 'epoch': 1.0}
+{'loss': 1.0316, 'grad_norm': 1.8684533521782878, 'learning_rate': 4.848904375109875e-09, 'epoch': 1.0}
+{'loss': 0.7153, 'grad_norm': 0.9122003848331615, 'learning_rate': 4.703082177215378e-09, 'epoch': 1.0}
+{'loss': 1.1442, 'grad_norm': 0.8297542801866371, 'learning_rate': 4.559486203148388e-09, 'epoch': 1.0}
+{'loss': 1.066, 'grad_norm': 1.099243554454167, 'learning_rate': 4.418116456106347e-09, 'epoch': 1.0}
+{'loss': 1.5408, 'grad_norm': 0.7118212662014688, 'learning_rate': 4.278972939242287e-09, 'epoch': 1.0}
+{'loss': 1.3404, 'grad_norm': 1.4362812302876498, 'learning_rate': 4.142055655653732e-09, 'epoch': 1.0}
+{'loss': 1.0888, 'grad_norm': 1.3629430661994493, 'learning_rate': 4.0073646083715884e-09, 'epoch': 1.0}
+{'loss': 1.112, 'grad_norm': 2.2100855393400276, 'learning_rate': 3.8748998004267676e-09, 'epoch': 1.0}
+{'loss': 1.3644, 'grad_norm': 0.7158178246776601, 'learning_rate': 3.744661234739155e-09, 'epoch': 1.0}
+{'loss': 1.3107, 'grad_norm': 0.942036119220107, 'learning_rate': 3.6166489142286377e-09, 'epoch': 1.0}
+{'loss': 1.2474, 'grad_norm': 1.5908115889363665, 'learning_rate': 3.4908628417373857e-09, 'epoch': 1.0}
+{'loss': 0.9509, 'grad_norm': 0.8861920310940888, 'learning_rate': 3.3673030200520594e-09, 'epoch': 1.0}
+{'loss': 1.1007, 'grad_norm': 1.1194164105994997, 'learning_rate': 3.2459694519482162e-09, 'epoch': 1.0}
+{'loss': 1.079, 'grad_norm': 1.7755772555508103, 'learning_rate': 3.1268621401125964e-09, 'epoch': 1.0}
+{'loss': 1.0855, 'grad_norm': 0.8351181880154243, 'learning_rate': 3.009981087198632e-09, 'epoch': 1.0}
+{'loss': 1.2608, 'grad_norm': 0.8626554454964646, 'learning_rate': 2.895326295804246e-09, 'epoch': 1.0}
+{'loss': 1.0801, 'grad_norm': 0.9514880970684986, 'learning_rate': 2.7828977684940525e-09, 'epoch': 1.0}
+{'loss': 1.1484, 'grad_norm': 0.7842554312013447, 'learning_rate': 2.672695507754952e-09, 'epoch': 1.0}
+{'loss': 1.0309, 'grad_norm': 0.9639220640203608, 'learning_rate': 2.564719516062741e-09, 'epoch': 1.0}
+{'loss': 1.0896, 'grad_norm': 5.893335510855347, 'learning_rate': 2.4589697958044e-09, 'epoch': 1.0}
+{'loss': 1.0671, 'grad_norm': 1.237578097349438, 'learning_rate': 2.355446349333601e-09, 'epoch': 1.0}
+{'loss': 1.2633, 'grad_norm': 3.309447811014566, 'learning_rate': 2.254149178959608e-09, 'epoch': 1.0}
+{'loss': 1.5265, 'grad_norm': 1.1766115766294134, 'learning_rate': 2.155078286936174e-09, 'epoch': 1.0}
+{'loss': 1.3161, 'grad_norm': 1.261377135630768, 'learning_rate': 2.058233675472643e-09, 'epoch': 1.0}
+{'loss': 0.9197, 'grad_norm': 2.9265974640791073, 'learning_rate': 1.963615346733949e-09, 'epoch': 1.0}
+{'loss': 1.2767, 'grad_norm': 1.124163723768451, 'learning_rate': 1.87122330279621e-09, 'epoch': 1.0}
+{'loss': 0.9884, 'grad_norm': 0.8206219469866542, 'learning_rate': 1.781057545746645e-09, 'epoch': 1.0}
+{'loss': 1.4566, 'grad_norm': 0.6893408698138835, 'learning_rate': 1.6931180775836553e-09, 'epoch': 1.0}
+{'loss': 1.4497, 'grad_norm': 0.7513770563975293, 'learning_rate': 1.6074049002612335e-09, 'epoch': 1.0}
+{'loss': 1.4434, 'grad_norm': 1.0794222136615717, 'learning_rate': 1.5239180156889633e-09, 'epoch': 1.0}
+{'loss': 1.3901, 'grad_norm': 1.2735045093264037, 'learning_rate': 1.442657425720917e-09, 'epoch': 1.0}
+{'loss': 1.0627, 'grad_norm': 1.083357360371921, 'learning_rate': 1.3636231321778602e-09, 'epoch': 1.0}
+{'loss': 1.1931, 'grad_norm': 1.5561575605915832, 'learning_rate': 1.2868151368028435e-09, 'epoch': 1.0}
+{'loss': 1.0351, 'grad_norm': 0.8498417639523637, 'learning_rate': 1.2122334413278146e-09, 'epoch': 1.0}
+{'loss': 1.2696, 'grad_norm': 1.0869334696996453, 'learning_rate': 1.1398780473959037e-09, 'epoch': 1.0}
+{'loss': 1.1768, 'grad_norm': 0.9256138759220954, 'learning_rate': 1.0697489566280362e-09, 'epoch': 1.0}
+{'loss': 1.3349, 'grad_norm': 1.1068591098386078, 'learning_rate': 1.001846170567422e-09, 'epoch': 1.0}
+{'loss': 0.9784, 'grad_norm': 0.7613604227444113, 'learning_rate': 9.361696907572714e-10, 'epoch': 1.0}
+{'loss': 1.1655, 'grad_norm': 1.3884342450751241, 'learning_rate': 8.727195186297721e-10, 'epoch': 1.0}
+{'loss': 1.3665, 'grad_norm': 0.7689778929243138, 'learning_rate': 8.114956556171116e-10, 'epoch': 1.0}
+{'loss': 1.1323, 'grad_norm': 1.259485983943609, 'learning_rate': 7.524981030626598e-10, 'epoch': 1.0}
+{'loss': 1.1608, 'grad_norm': 0.7771101060171148, 'learning_rate': 6.957268622986846e-10, 'epoch': 1.0}
+{'loss': 1.1237, 'grad_norm': 1.288212435458082, 'learning_rate': 6.411819345908399e-10, 'epoch': 1.0}
+{'loss': 1.2058, 'grad_norm': 1.0930381944300036, 'learning_rate': 5.888633211381667e-10, 'epoch': 1.0}
+{'loss': 1.19, 'grad_norm': 1.1138745850551712, 'learning_rate': 5.387710231063991e-10, 'epoch': 1.0}
+{'loss': 1.2528, 'grad_norm': 0.9071826069395514, 'learning_rate': 4.909050416168626e-10, 'epoch': 1.0}
+{'loss': 1.18, 'grad_norm': 2.0569238529909164, 'learning_rate': 4.4526537774647324e-10, 'epoch': 1.0}
+{'loss': 1.2705, 'grad_norm': 0.7839265892155309, 'learning_rate': 4.018520324944319e-10, 'epoch': 1.0}
+{'loss': 1.3008, 'grad_norm': 1.0155075719562054, 'learning_rate': 3.6066500682663264e-10, 'epoch': 1.0}
+{'loss': 1.0982, 'grad_norm': 0.6882484767145587, 'learning_rate': 3.217043016756627e-10, 'epoch': 1.0}
+{'loss': 1.1173, 'grad_norm': 1.2088299639494535, 'learning_rate': 2.8496991788529157e-10, 'epoch': 1.0}
+{'loss': 0.9761, 'grad_norm': 2.247488672969709, 'learning_rate': 2.5046185629928885e-10, 'epoch': 1.0}
+{'loss': 1.1151, 'grad_norm': 0.7724605019725375, 'learning_rate': 2.1818011767260615e-10, 'epoch': 1.0}
+{'loss': 1.2292, 'grad_norm': 1.916832753280351, 'learning_rate': 1.881247027157862e-10, 'epoch': 1.0}
+{'loss': 1.1573, 'grad_norm': 0.9861343770803095, 'learning_rate': 1.6029561211716727e-10, 'epoch': 1.0}
+{'loss': 1.2659, 'grad_norm': 1.0597993957715799, 'learning_rate': 1.34692846487372e-10, 'epoch': 1.0}
+{'loss': 1.1392, 'grad_norm': 1.2057802830555464, 'learning_rate': 1.113164063926142e-10, 'epoch': 1.0}
+{'loss': 1.2013, 'grad_norm': 0.7118516881250079, 'learning_rate': 9.016629235469864e-11, 'epoch': 1.0}
+{'loss': 1.2241, 'grad_norm': 0.9711977682864302, 'learning_rate': 7.124250485102124e-11, 'epoch': 1.0}
+{'loss': 1.2042, 'grad_norm': 1.1853315913804563, 'learning_rate': 5.454504429236451e-11, 'epoch': 1.0}
+{'loss': 1.1514, 'grad_norm': 0.7261232294024941, 'learning_rate': 4.0073911056204285e-11, 'epoch': 1.0}
+{'loss': 1.5327, 'grad_norm': 3.9712089223972984, 'learning_rate': 2.7829105475607465e-11, 'epoch': 1.0}
+{'loss': 0.9263, 'grad_norm': 0.910668734726166, 'learning_rate': 1.781062779482312e-11, 'epoch': 1.0}
+{'loss': 1.3145, 'grad_norm': 0.9994139658018105, 'learning_rate': 1.0018478269202547e-11, 'epoch': 1.0}
+{'loss': 1.2288, 'grad_norm': 0.8901729947063939, 'learning_rate': 4.452657054176968e-12, 'epoch': 1.0}
+{'loss': 1.1766, 'grad_norm': 0.8098067441407663, 'learning_rate': 1.1131642718709146e-12, 'epoch': 1.0}
+{'loss': 0.9875, 'grad_norm': 1.0981499096860188, 'learning_rate': 0.0, 'epoch': 1.0}
+{'train_runtime': 2160.6123, 'train_samples_per_second': 321.488, 'train_steps_per_second': 10.047, 'train_loss': 0.011552139646460999, 'epoch': 1.0}
+[2024-06-14 02:49:37,060] [INFO] [launch.py:351:main] Process 1400332 exits successfully.
+[2024-06-14 02:49:38,061] [INFO] [launch.py:351:main] Process 1400331 exits successfully.

non_lora_trainables.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbe17bae02d0aa1207e503360d45078dc617beaed98eb2a241a59c3c9c4c3a46
+size 19017904