You can deploy this model and start earning money today!

RichardErkhov/
Magpie-Align_-_Llama-3.1-8B-Magpie-Align-v0.1-RC1-gguf

Name: Magpie-Align_-_Llama-3.1-8B-Magpie-Align-v0.1-RC1-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama-3.1-8B-Magpie-Align-v0.1-RC1 - GGUF

Model creator: https://huggingface.co/Magpie-Align/
Original model: https://huggingface.co/Magpie-Align/Llama-3.1-8B-Magpie-Align-v0.1-RC1/

Name	Quant method	Size
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q2_K.gguf	Q2_K	2.96GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.IQ3_XS.gguf	IQ3_XS	3.28GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.IQ3_S.gguf	IQ3_S	3.43GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q3_K_S.gguf	Q3_K_S	3.41GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.IQ3_M.gguf	IQ3_M	3.52GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q3_K.gguf	Q3_K	3.74GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q3_K_M.gguf	Q3_K_M	3.74GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q3_K_L.gguf	Q3_K_L	4.03GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.IQ4_XS.gguf	IQ4_XS	4.18GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q4_0.gguf	Q4_0	4.34GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.IQ4_NL.gguf	IQ4_NL	4.38GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q4_K_S.gguf	Q4_K_S	4.37GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q4_K.gguf	Q4_K	4.58GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q4_K_M.gguf	Q4_K_M	4.12GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q4_1.gguf	Q4_1	4.78GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q5_0.gguf	Q5_0	5.21GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q5_K_S.gguf	Q5_K_S	5.21GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q5_K.gguf	Q5_K	5.34GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q5_K_M.gguf	Q5_K_M	5.34GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q5_1.gguf	Q5_1	5.65GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q6_K.gguf	Q6_K	6.14GB
Llama-3.1-8B-Magpie-Align-v0.1-RC1.Q8_0.gguf	Q8_0	7.95GB

Original model description:

base_model: meta-llama/Meta-Llama-3.1-8B tags:

alignment-handbook
trl
dpo
generated_from_trainer datasets:
princeton-nlp/llama3-ultrafeedback-armorm model-index:
name: Llama-3.1-8B-Magpie-Align-v0.1-RC1 results: []

This model is a fine-tuned version of Magpie-Align/Llama-3.1-8B-Magpie-Align-SFT-v0.1 on the princeton-nlp/llama3-ultrafeedback-armorm dataset. It achieves the following results on the evaluation set:

Loss: 0.3290
Rewards/chosen: -4.8185
Rewards/rejected: -6.6901
Rewards/accuracies: 0.8952
Rewards/margins: 1.8716
Logps/rejected: -867.8638
Logps/chosen: -686.8736
Logits/rejected: -0.5907
Logits/chosen: -0.5749

Model description

More details will be added soon.

Benchmark

MT-Bench: 8.375 (1st Turn), 7.650 (Second Turn), 8.013 (Average)
Alpaca Eval 2 (GPT-4-Turbo-1106): 45.73 (LC), 52.79 (WR)
Arena Hard: 42.4

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e-06
train_batch_size: 2
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 16
total_train_batch_size: 256
total_eval_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.4439	0.4275	100	0.4168	-4.9964	-6.3086	0.8145	1.3123	-829.7151	-704.6570	-0.5150	-0.5001
0.343	0.8549	200	0.3298	-4.9310	-6.7966	0.8952	1.8655	-878.5105	-698.1248	-0.5776	-0.5622

Framework versions

Transformers 4.43.2
Pytorch 2.3.1+cu121
Datasets 2.20.0
Tokenizers 0.19.1

RichardErkhov/Magpie-Align_-_Llama-3.1-8B-Magpie-Align-v0.1-RC1-gguf