Add config.json file for model setup
Browse files- config.json +48 -0
config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model_type": "llava",
|
| 3 |
+
"architectures": ["LLaVAForVideoClassification"],
|
| 4 |
+
"learning_rate": 5e-05,
|
| 5 |
+
"train_batch_size": 2,
|
| 6 |
+
"eval_batch_size": 8,
|
| 7 |
+
"seed": 42,
|
| 8 |
+
"distributed_type": "multi-GPU",
|
| 9 |
+
"num_devices": 2,
|
| 10 |
+
"gradient_accumulation_steps": 8,
|
| 11 |
+
"total_train_batch_size": 32,
|
| 12 |
+
"total_eval_batch_size": 16,
|
| 13 |
+
"optimizer": "AdamW_Torch",
|
| 14 |
+
"betas": [0.9, 0.999],
|
| 15 |
+
"epsilon": 1e-08,
|
| 16 |
+
"lr_scheduler_type": "cosine",
|
| 17 |
+
"lr_scheduler_warmup_steps": 100,
|
| 18 |
+
"num_epochs": 1.0,
|
| 19 |
+
"framework_versions": {
|
| 20 |
+
"PEFT": "0.12.0",
|
| 21 |
+
"Transformers": "4.46.1",
|
| 22 |
+
"Pytorch": "2.3.1+cu121",
|
| 23 |
+
"Datasets": "3.1.0",
|
| 24 |
+
"Tokenizers": "0.20.3"
|
| 25 |
+
},
|
| 26 |
+
"description": "This model is a fine-tuned version of llava-hf/LLaVA-NeXT-Video-7B-hf on the merger, the LLM_dataset(4o) and the LLM_dataset(4mini) datasets.",
|
| 27 |
+
"training_data": {
|
| 28 |
+
"dataset_names": ["merger", "LLM_dataset(4o)", "LLM_dataset(4mini)"]
|
| 29 |
+
},
|
| 30 |
+
"training_hyperparameters": {
|
| 31 |
+
"learning_rate": 5e-05,
|
| 32 |
+
"train_batch_size": 2,
|
| 33 |
+
"eval_batch_size": 8,
|
| 34 |
+
"seed": 42,
|
| 35 |
+
"distributed_type": "multi-GPU",
|
| 36 |
+
"num_devices": 2,
|
| 37 |
+
"gradient_accumulation_steps": 8,
|
| 38 |
+
"total_train_batch_size": 32,
|
| 39 |
+
"total_eval_batch_size": 16,
|
| 40 |
+
"optimizer": "AdamW_Torch",
|
| 41 |
+
"betas": [0.9, 0.999],
|
| 42 |
+
"epsilon": 1e-08,
|
| 43 |
+
"lr_scheduler_type": "cosine",
|
| 44 |
+
"lr_scheduler_warmup_steps": 100,
|
| 45 |
+
"num_epochs": 1.0
|
| 46 |
+
}
|
| 47 |
+
}
|
| 48 |
+
|