This model is a fine-tuned version of TinyLlama/TinyLlama-1.1B-Chat-v1.0 on the data/rlced_conifer dataset.
It achieves the following results on the evaluation set:
Loss: 0.4784
Rewards/chosen: -2.0477
Rewards/rejected: -2.9020
Rewards/accuracies: 0.7745
Rewards/margins: 0.8542
Logps/rejected: -835.5375
Logps/chosen: -730.0506
Logits/rejected: -2.5534
Logits/chosen: -2.5642
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
learning_rate: 5e-07
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 4
total_train_batch_size: 256
total_eval_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
