HuggingFaceH4_-_mistral-7b-grok-gguf – AI Model by RichardErkhov | AlphaNeural AI

You can deploy this model and start earning money today!

RichardErkhov/
HuggingFaceH4_-_mistral-7b-grok-gguf

like

0

endpoints_compatible

Views

No views yet

Quantization made by Richard Erkhov.

Request more models

mistral-7b-grok - GGUF

Model creator: https://huggingface.co/HuggingFaceH4/
Original model: https://huggingface.co/HuggingFaceH4/mistral-7b-grok/

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

Original model description:

license: apache-2.0 base_model: mistralai/Mistral-7B-v0.1 tags:

alignment-handbook
generated_from_trainer datasets:
HuggingFaceH4/grok-conversation-harmless
HuggingFaceH4/ultrachat_200k model-index:
name: mistral-7b-grok results: []

Mistral 7B Grok

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 that has been aligned via Constitutional AI to mimic the style of xAI's Grok assistant.

It achieves the following results on the evaluation set:

Loss: 0.9348

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: 4
total_train_batch_size: 256
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
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.9326	1.0	545	0.9348

Framework versions

Transformers 4.36.2
Pytorch 2.1.2+cu121
Datasets 2.16.1
Tokenizers 0.15.0