Upload folder using huggingface_hub

llama3-instruct+8b+dinosiglip/checkpoints/latest-checkpoint.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b40df0c922019ca452fd850dceaffe7bc67173c728bd557f0f4c8aa90ef253c2
+size 16113068150

llama3-instruct+8b+dinosiglip/config.json

@@ -0,0 +1,59 @@
+{
+  "dataset": {
+    "align_stage_components": [
+      "download/llava-laion-cc-sbu-558k/chat.json",
+      "download/llava-laion-cc-sbu-558k"
+    ],
+    "dataset_id": "llava-v15",
+    "dataset_root_dir": "data",
+    "finetune_stage_components": [
+      "download/llava-v1.5-instruct/llava_v1_5_mix665k.json",
+      "download/llava-v1.5-instruct"
+    ],
+    "type": "llava-v15"
+  },
+  "hf_token": ".hf_token",
+  "model": {
+    "align_epochs": 1,
+    "align_global_batch_size": 256,
+    "align_learning_rate": 0.001,
+    "align_lr_scheduler_type": "linear-warmup+cosine-decay",
+    "align_max_grad_norm": 1.0,
+    "align_max_steps": null,
+    "align_per_device_batch_size": 16,
+    "align_train_strategy": "fsdp-shard-grad-op",
+    "align_warmup_ratio": 0.03,
+    "align_weight_decay": 0.0,
+    "arch_specifier": "no-align+gelu-mlp",
+    "enable_gradient_checkpointing": true,
+    "enable_mixed_precision_training": true,
+    "finetune_epochs": 1,
+    "finetune_global_batch_size": 128,
+    "finetune_learning_rate": 2e-05,
+    "finetune_lr_scheduler_type": "linear-warmup+cosine-decay",
+    "finetune_max_grad_norm": 1.0,
+    "finetune_max_steps": null,
+    "finetune_per_device_batch_size": 4,
+    "finetune_train_strategy": "fsdp-full-shard",
+    "finetune_warmup_ratio": 0.03,
+    "finetune_weight_decay": 0.1,
+    "image_resize_strategy": "letterbox",
+    "llm_backbone_id": "llama3-8b-instruct",
+    "llm_max_length": 2048,
+    "model_id": "llama3-instruct+8b+dinosiglip",
+    "reduce_in_full_precision": false,
+    "type": "one-stage+7b",
+    "vision_backbone_id": "dinosiglip-vit-so-384px"
+  },
+  "pretrained_checkpoint": null,
+  "run_id": "llama3-instruct+8b+dinosiglip+stage-finetune+x7",
+  "run_root_dir": "runs",
+  "seed": 7,
+  "stage": "finetune",
+  "trackers": [
+    "jsonl",
+    "wandb"
+  ],
+  "wandb_entity": "rylan",
+  "wandb_project": "prismatic-vlm"
+}

llama3-instruct+8b+dinosiglip/config.yaml

@@ -0,0 +1,52 @@
+dataset:
+  align_stage_components:
+  - download/llava-laion-cc-sbu-558k/chat.json
+  - download/llava-laion-cc-sbu-558k
+  dataset_id: llava-v15
+  dataset_root_dir: data
+  finetune_stage_components:
+  - download/llava-v1.5-instruct/llava_v1_5_mix665k.json
+  - download/llava-v1.5-instruct
+  type: llava-v15
+hf_token: .hf_token
+model:
+  align_epochs: 1
+  align_global_batch_size: 256
+  align_learning_rate: 0.001
+  align_lr_scheduler_type: linear-warmup+cosine-decay
+  align_max_grad_norm: 1.0
+  align_max_steps: null
+  align_per_device_batch_size: 16
+  align_train_strategy: fsdp-shard-grad-op
+  align_warmup_ratio: 0.03
+  align_weight_decay: 0.0
+  arch_specifier: no-align+gelu-mlp
+  enable_gradient_checkpointing: true
+  enable_mixed_precision_training: true
+  finetune_epochs: 1
+  finetune_global_batch_size: 128
+  finetune_learning_rate: 2.0e-05
+  finetune_lr_scheduler_type: linear-warmup+cosine-decay
+  finetune_max_grad_norm: 1.0
+  finetune_max_steps: null
+  finetune_per_device_batch_size: 4
+  finetune_train_strategy: fsdp-full-shard
+  finetune_warmup_ratio: 0.03
+  finetune_weight_decay: 0.1
+  image_resize_strategy: letterbox
+  llm_backbone_id: llama3-8b-instruct
+  llm_max_length: 2048
+  model_id: llama3-instruct+8b+dinosiglip
+  reduce_in_full_precision: false
+  type: one-stage+7b
+  vision_backbone_id: dinosiglip-vit-so-384px
+pretrained_checkpoint: null
+run_id: llama3-instruct+8b+dinosiglip+stage-finetune+x7
+run_root_dir: runs
+seed: 7
+stage: finetune
+trackers:
+- jsonl
+- wandb
+wandb_entity: rylan
+wandb_project: prismatic-vlm

llama3-instruct+8b+dinosiglip/run-metrics.jsonl

@@ -0,0 +1 @@
+{"hparams": {"dataset": {"align_stage_components": ["download/llava-laion-cc-sbu-558k/chat.json", "download/llava-laion-cc-sbu-558k"], "dataset_id": "llava-v15", "dataset_root_dir": "data", "finetune_stage_components": ["download/llava-v1.5-instruct/llava_v1_5_mix665k.json", "download/llava-v1.5-instruct"], "type": "llava-v15"}, "hf_token": ".hf_token", "model": {"align_epochs": 1, "align_global_batch_size": 256, "align_learning_rate": 0.001, "align_lr_scheduler_type": "linear-warmup+cosine-decay", "align_max_grad_norm": 1.0, "align_max_steps": null, "align_per_device_batch_size": 16, "align_train_strategy": "fsdp-shard-grad-op", "align_warmup_ratio": 0.03, "align_weight_decay": 0.0, "arch_specifier": "no-align+gelu-mlp", "enable_gradient_checkpointing": true, "enable_mixed_precision_training": true, "finetune_epochs": 1, "finetune_global_batch_size": 128, "finetune_learning_rate": 2e-05, "finetune_lr_scheduler_type": "linear-warmup+cosine-decay", "finetune_max_grad_norm": 1.0, "finetune_max_steps": null, "finetune_per_device_batch_size": 4, "finetune_train_strategy": "fsdp-full-shard", "finetune_warmup_ratio": 0.03, "finetune_weight_decay": 0.1, "image_resize_strategy": "letterbox", "llm_backbone_id": "llama3-8b-instruct", "llm_max_length": 2048, "model_id": "llama3-instruct+8b+dinosiglip", "reduce_in_full_precision": false, "type": "one-stage+7b", "vision_backbone_id": "dinosiglip-vit-so-384px"}, "pretrained_checkpoint": null, "run_id": "llama3-instruct+8b+dinosiglip+stage-finetune+x7", "run_root_dir": "runs", "seed": 7, "stage": "finetune", "trackers": ["jsonl", "wandb"], "wandb_entity": "rylan", "wandb_project": "prismatic-vlm"}, "run_id": "llama3-instruct+8b+dinosiglip+stage-finetune+x7"}