You can deploy this model and start earning money today!

RichardErkhov/
taicheng_-_zephyr-7b-align-scan-0.0-0.0-linear-3-gguf

Name: taicheng_-_zephyr-7b-align-scan-0.0-0.0-linear-3-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

zephyr-7b-align-scan-0.0-0.0-linear-3 - GGUF

Model creator: https://huggingface.co/taicheng/
Original model: https://huggingface.co/taicheng/zephyr-7b-align-scan-0.0-0.0-linear-3/

Name	Quant method	Size
zephyr-7b-align-scan-0.0-0.0-linear-3.Q2_K.gguf	Q2_K	2.53GB
zephyr-7b-align-scan-0.0-0.0-linear-3.IQ3_XS.gguf	IQ3_XS	2.81GB
zephyr-7b-align-scan-0.0-0.0-linear-3.IQ3_S.gguf	IQ3_S	2.96GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q3_K_S.gguf	Q3_K_S	2.95GB
zephyr-7b-align-scan-0.0-0.0-linear-3.IQ3_M.gguf	IQ3_M	3.06GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q3_K.gguf	Q3_K	3.28GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q3_K_M.gguf	Q3_K_M	3.28GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q3_K_L.gguf	Q3_K_L	3.56GB
zephyr-7b-align-scan-0.0-0.0-linear-3.IQ4_XS.gguf	IQ4_XS	3.67GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q4_0.gguf	Q4_0	3.83GB
zephyr-7b-align-scan-0.0-0.0-linear-3.IQ4_NL.gguf	IQ4_NL	3.87GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q4_K_S.gguf	Q4_K_S	3.86GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q4_K.gguf	Q4_K	4.07GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q4_K_M.gguf	Q4_K_M	4.07GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q4_1.gguf	Q4_1	4.24GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q5_0.gguf	Q5_0	4.65GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q5_K_S.gguf	Q5_K_S	4.65GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q5_K.gguf	Q5_K	4.78GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q5_K_M.gguf	Q5_K_M	4.78GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q5_1.gguf	Q5_1	5.07GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q6_K.gguf	Q6_K	5.53GB
zephyr-7b-align-scan-0.0-0.0-linear-3.Q8_0.gguf	Q8_0	7.17GB

Original model description:

library_name: transformers license: apache-2.0 base_model: alignment-handbook/zephyr-7b-sft-full tags:

alignment-handbook
trl
dpo
generated_from_trainer
trl
dpo
generated_from_trainer datasets:
HuggingFaceH4/ultrafeedback_binarized model-index:
name: zephyr-7b-align-scan-0.0-0.0-linear-3 results: []

zephyr-7b-align-scan-0.0-0.0-linear-3

This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-full on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:

Loss: 0.6455
Rewards/chosen: -0.1452
Rewards/rejected: -0.3169
Rewards/accuracies: 0.3313
Rewards/margins: 0.1718
Logps/rejected: -103.6337
Logps/chosen: -84.7995
Logits/rejected: -2.2391
Logits/chosen: -2.2616

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: 2.8387779271947853e-07
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 2
total_train_batch_size: 64
total_eval_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.6466	1.0417	100	0.6659	-0.0460	-0.1148	0.3234	0.0688	-89.2840	-77.7589	-2.4671	-2.4857
0.5655	2.0833	200	0.6495	-0.1460	-0.2967	0.3294	0.1507	-102.2009	-84.8621	-2.3195	-2.3414

Framework versions

Transformers 4.44.2
Pytorch 2.4.0
Datasets 2.21.0
Tokenizers 0.19.1

RichardErkhov/taicheng_-_zephyr-7b-align-scan-0.0-0.0-linear-3-gguf

Original model description:

zephyr-7b-align-scan-0.0-0.0-linear-3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

RichardErkhov/
taicheng_-_zephyr-7b-align-scan-0.0-0.0-linear-3-gguf