CoolFace
Modelpublic

RichardErkhov/synapsoft_-_Llama-2-7b-chat-hf-flan2022-1.2M-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes664downloads
Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Llama-2-7b-chat-hf-flan2022-1.2M - GGUF

  • —Model creator: https://huggingface.co/synapsoft/
  • —Original model: https://huggingface.co/synapsoft/Llama-2-7b-chat-hf-flan2022-1.2M/
NameQuant methodSize
Llama-2-7b-chat-hf-flan2022-1.2M.Q2_K.ggufQ2_K2.36GB
Llama-2-7b-chat-hf-flan2022-1.2M.IQ3_XS.ggufIQ3_XS2.6GB
Llama-2-7b-chat-hf-flan2022-1.2M.IQ3_S.ggufIQ3_S2.75GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q3_K_S.ggufQ3KS2.75GB
Llama-2-7b-chat-hf-flan2022-1.2M.IQ3_M.ggufIQ3_M2.9GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q3_K.ggufQ3_K3.07GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q3_K_M.ggufQ3KM3.07GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q3_K_L.ggufQ3KL3.35GB
Llama-2-7b-chat-hf-flan2022-1.2M.IQ4_XS.ggufIQ4_XS3.4GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q4_0.ggufQ4_03.56GB
Llama-2-7b-chat-hf-flan2022-1.2M.IQ4_NL.ggufIQ4_NL3.58GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q4_K_S.ggufQ4KS3.59GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q4_K.ggufQ4_K3.8GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q4_K_M.ggufQ4KM3.8GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q4_1.ggufQ4_13.95GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q5_0.ggufQ5_04.33GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q5_K_S.ggufQ5KS4.33GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q5_K.ggufQ5_K4.45GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q5_K_M.ggufQ5KM4.45GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q5_1.ggufQ5_14.72GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q6_K.ggufQ6_K5.15GB
Llama-2-7b-chat-hf-flan2022-1.2M.Q8_0.ggufQ8_06.67GB

Original model description: --- base_model: meta-llama/Llama-2-7b-chat-hf tags:

  • —generatedfromtrainer datasets:
  • —conceptofmind/FLAN_2022 model-index:
  • —name: outputs results: [] ---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

outputs

This model is a fine-tuned version of meta-llama/Llama-2-7b-chat-hf on the conceptofmind/FLAN_2022 dataset.

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: 1e-05
  • —trainbatchsize: 12
  • —evalbatchsize: 8
  • —seed: 42
  • —distributed_type: multi-GPU
  • —gradientaccumulationsteps: 8
  • —totaltrainbatch_size: 96
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: cosine
  • —lrschedulerwarmup_ratio: 0.03
  • —num_epochs: 1.0

Training results

Framework versions

  • —Transformers 4.31.0
  • —Pytorch 2.0.1+cu117
  • —Datasets 2.14.3
  • —Tokenizers 0.13.3