|
--- |
|
model_configuration: |
|
model_filename: dynamic_network.pth |
|
max_neurons: 750000 |
|
epochs: 50 |
|
clip: 5 |
|
patience: 7 |
|
adaptation_rate: 0.05 |
|
sequence_length: 200 |
|
max_sequence_length: 200 |
|
weight_decay: 0.005 |
|
num_embeddings: 25000 |
|
embedding_dim: 768 |
|
hidden_dim: 2048 |
|
learning_rate: 0.00001 |
|
some_intermediate_size: 3072 |
|
|
|
hyperparameters: |
|
input_dimension: 768 |
|
initial_neuron_count: 5000 |
|
some_adaptation_rate: 0.05 |
|
complexity_metric: null |
|
|
|
additional_hyperparameters: |
|
growth_improvement_threshold: 0.01 |
|
pruning_improvement_threshold: 0.005 |
|
stability_threshold: 0.02 |
|
max_complexity: 10000 |
|
low_stability_threshold: 0.01 |
|
min_epochs_before_growth: 5 |
|
desired_improvement_rate: 0.02 |
|
max_sequence_length: 200 |
|
|
|
license: apache-2.0 |
|
|
|
datasets: |
|
- vicgalle/alpaca-gpt4 |
|
|
|
language: |
|
- en |
|
|
|
library_name: transformers |
|
|
|
tags: |
|
- text-generation-inference |
|
|
|
metrics: |
|
- accuracy |
|
--- |
