lpetreadg/trained-tinyllama-ultrachat-GGUF

Quantized GGUF model files for trained-tinyllama-ultrachat from lpetreadg

Name	Quant method	Size
trained-tinyllama-ultrachat.q2_k.gguf	q2_k	None
trained-tinyllama-ultrachat.q3_k_m.gguf	q3_k_m	None
trained-tinyllama-ultrachat.q4_k_m.gguf	q4_k_m	None
trained-tinyllama-ultrachat.q5_k_m.gguf	q5_k_m	None
trained-tinyllama-ultrachat.q6_k.gguf	q6_k	None
trained-tinyllama-ultrachat.q8_0.gguf	q8_0	None

Original Model Card:

trained-tinyllama-ultrachat

This model is a fine-tuned version of PY007/TinyLlama-1.1B-intermediate-step-715k-1.5T on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3258

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 1
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
1.3767	0.08	100	1.3685
1.3494	0.17	200	1.3490
1.3436	0.25	300	1.3389
1.3231	0.33	400	1.3331
1.3278	0.42	500	1.3296
1.3214	0.5	600	1.3276
1.3376	0.58	700	1.3266
1.3227	0.67	800	1.3261
1.3329	0.75	900	1.3259
1.3185	0.83	1000	1.3258
1.332	0.92	1100	1.3258

Framework versions

• Transformers 4.34.1
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.14.1