BEE-spoke-data/smol_llama-101M-midjourney-messages-GGUF

Quantized GGUF model files for smol_llama-101M-midjourney-messages from BEE-spoke-data

Name	Quant method	Size
smol_llama-101m-midjourney-messages.fp16.gguf	fp16	203.28 MB
smol_llama-101m-midjourney-messages.q2_k.gguf	q2_k	50.93 MB
smol_llama-101m-midjourney-messages.q3_k_m.gguf	q3_k_m	57.06 MB
smol_llama-101m-midjourney-messages.q4_k_m.gguf	q4_k_m	65.40 MB
smol_llama-101m-midjourney-messages.q5_k_m.gguf	q5_k_m	74.34 MB
smol_llama-101m-midjourney-messages.q6_k.gguf	q6_k	83.83 MB
smol_llama-101m-midjourney-messages.q8_0.gguf	q8_0	108.35 MB

Original Model Card:

smol_llama-101M-midjourney-messages

Given a 'partial prompt' for a text2image model, this generates additional relevant text to include for a full prompt.

Model description

This model is a fine-tuned version of BEE-spoke-data/smol_llama-101M-GQA on the pszemraj/midjourney-messages-cleaned dataset. It achieves the following results on the evaluation set:

• Loss: 2.8431
• Accuracy: 0.4682

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.00025
• train_batch_size: 4
• eval_batch_size: 4
• seed: 17056
• gradient_accumulation_steps: 16
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-08
• lr_scheduler_type: inverse_sqrt
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 1.0