RichardErkhov/mlfoundations-dev_-_test_run_mini-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

test_run_mini - GGUF

• Model creator: https://huggingface.co/mlfoundations-dev/
• Original model: https://huggingface.co/mlfoundations-dev/test_run_mini/

Name	Quant method	Size
test_run_mini.Q2_K.gguf	Q2_K	0.54GB
test_run_mini.IQ3_XS.gguf	IQ3_XS	0.58GB
test_run_mini.IQ3_S.gguf	IQ3_S	0.6GB
test_run_mini.Q3_K_S.gguf	Q3_K_S	0.6GB
test_run_mini.IQ3_M.gguf	IQ3_M	0.61GB
test_run_mini.Q3_K.gguf	Q3_K	0.64GB
test_run_mini.Q3_K_M.gguf	Q3_K_M	0.64GB
test_run_mini.Q3_K_L.gguf	Q3_K_L	0.68GB
test_run_mini.IQ4_XS.gguf	IQ4_XS	0.7GB
test_run_mini.Q4_0.gguf	Q4_0	0.72GB
test_run_mini.IQ4_NL.gguf	IQ4_NL	0.72GB
test_run_mini.Q4_K_S.gguf	Q4_K_S	0.72GB
test_run_mini.Q4_K.gguf	Q4_K	0.75GB
test_run_mini.Q4_K_M.gguf	Q4_K_M	0.75GB
test_run_mini.Q4_1.gguf	Q4_1	0.77GB
test_run_mini.Q5_0.gguf	Q5_0	0.83GB
test_run_mini.Q5_K_S.gguf	Q5_K_S	0.83GB
test_run_mini.Q5_K.gguf	Q5_K	0.85GB
test_run_mini.Q5_K_M.gguf	Q5_K_M	0.85GB
test_run_mini.Q5_1.gguf	Q5_1	0.89GB
test_run_mini.Q6_K.gguf	Q6_K	0.95GB
test_run_mini.Q8_0.gguf	Q8_0	1.23GB

Original model description:

library_name: transformers license: llama3.2 base_model: meta-llama/Llama-3.2-1B tags: - llama-factory - full - generated_from_trainer model-index: - name: test_run_mini results: []

test_run_mini

This model is a fine-tuned version of meta-llama/Llama-3.2-1B on the llamafactory/alpaca_en dataset. It achieves the following results on the evaluation set:

• Loss: 1.4380

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 8
• total_train_batch_size: 512
• total_eval_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 3

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.1538	3	1.4380

Framework versions

• Transformers 4.45.2
• Pytorch 2.4.1+cu121
• Datasets 2.21.0
• Tokenizers 0.20.1