DTee8
/

galactus

@@ -11,27 +11,27 @@ This model is a fine-tuned version of microsoft/Phi-4-multimodal-instruct on the
 Galaxy's Last Exam Benchmark: 72%
-##Model description
 Galactus is a state-of-the-art (SOTA) multimodal language model that outperforms all OpenAI and Gemini models on the Galaxy's Last Exam Benchmark. This benchmark features challenging tasks that push the boundaries of metaphysical competence—for instance, determining how many times two lines intersect or simulating the effect of adding three minutes to an analog clock.
 The model accepts image input along with text prompts and has been specifically optimized to tackle complex visual reasoning tasks.
-##Intended uses & limitations
 This model is intended for handling complex visual reasoning tasks that require metaphysical competence.
-##Training and evaluation data
 The model was exclusively trained on the Galaxy's Last Exam Benchmark.
-##Training procedure
 The model was trained using LoRA adapters focused on the vision components of the base model.
-##Prompt format
 This model uses the following image prompt format:
 Copy<|image_1|> + user text
-##Training hyperparameters
 The following hyperparameters were used during training:
-num_train_epochs: specified in args
 per_device_train_batch_size: specified in args
 gradient_checkpointing: True
 gradient_checkpointing_kwargs: {'use_reentrant': False}
@@ -41,7 +41,7 @@ adam_beta1: 0.9
 adam_beta2: 0.95
 adam_epsilon: 1e-7
 learning_rate: specified in args
-weight_decay: 0.0 or as specified in args
 save_strategy: 'steps'
 save_steps: 10
 eval_steps: 10 if eval_dataset else None
@@ -56,10 +56,10 @@ save_only_model: True
 dataloader_num_workers: 4
 ddp_find_unused_parameters: True
-##Training results
 The model achieved 72% performance on the Galaxy's Last Exam Benchmark.
-##Framework versions
 Transformers 4.46.1
 PyTorch 2.7.0.dev20250304+cu128

 Galaxy's Last Exam Benchmark: 72%
+## Model description
 Galactus is a state-of-the-art (SOTA) multimodal language model that outperforms all OpenAI and Gemini models on the Galaxy's Last Exam Benchmark. This benchmark features challenging tasks that push the boundaries of metaphysical competence—for instance, determining how many times two lines intersect or simulating the effect of adding three minutes to an analog clock.
 The model accepts image input along with text prompts and has been specifically optimized to tackle complex visual reasoning tasks.
+## Intended uses & limitations
 This model is intended for handling complex visual reasoning tasks that require metaphysical competence.
+## Training and evaluation data
 The model was exclusively trained on the Galaxy's Last Exam Benchmark.
+## Training procedure
 The model was trained using LoRA adapters focused on the vision components of the base model.
+## Prompt format
 This model uses the following image prompt format:
 Copy<|image_1|> + user text
+## Training hyperparameters
 The following hyperparameters were used during training:
+num_train_epochs: specified in args (but used checkpoint @ 252 epochs)
 per_device_train_batch_size: specified in args
 gradient_checkpointing: True
 gradient_checkpointing_kwargs: {'use_reentrant': False}
 adam_beta2: 0.95
 adam_epsilon: 1e-7
 learning_rate: specified in args
+weight_decay: 0.0
 save_strategy: 'steps'
 save_steps: 10
 eval_steps: 10 if eval_dataset else None
 dataloader_num_workers: 4
 ddp_find_unused_parameters: True
+## Training results
 The model achieved 72% performance on the Galaxy's Last Exam Benchmark.
+## Framework versions
 Transformers 4.46.1
 PyTorch 2.7.0.dev20250304+cu128