CoolFace
Modelpublic

RichardErkhov/jkazdan_-_llama8b-gsm-real-sftsd1-gguf

sourceHugging Faceupdated 1y agoView on Hugging Face
0likes1kdownloads
Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

llama8b-gsm-real-sftsd1 - GGUF

  • —Model creator: https://huggingface.co/jkazdan/
  • —Original model: https://huggingface.co/jkazdan/llama8b-gsm-real-sftsd1/
NameQuant methodSize
llama8b-gsm-real-sftsd1.Q2_K.ggufQ2_K2.96GB
llama8b-gsm-real-sftsd1.IQ3_XS.ggufIQ3_XS3.28GB
llama8b-gsm-real-sftsd1.IQ3_S.ggufIQ3_S3.43GB
llama8b-gsm-real-sftsd1.Q3_K_S.ggufQ3KS3.41GB
llama8b-gsm-real-sftsd1.IQ3_M.ggufIQ3_M3.52GB
llama8b-gsm-real-sftsd1.Q3_K.ggufQ3_K3.74GB
llama8b-gsm-real-sftsd1.Q3_K_M.ggufQ3KM3.74GB
llama8b-gsm-real-sftsd1.Q3_K_L.ggufQ3KL4.03GB
llama8b-gsm-real-sftsd1.IQ4_XS.ggufIQ4_XS4.18GB
llama8b-gsm-real-sftsd1.Q4_0.ggufQ4_04.34GB
llama8b-gsm-real-sftsd1.IQ4_NL.ggufIQ4_NL4.38GB
llama8b-gsm-real-sftsd1.Q4_K_S.ggufQ4KS4.37GB
llama8b-gsm-real-sftsd1.Q4_K.ggufQ4_K4.58GB
llama8b-gsm-real-sftsd1.Q4_K_M.ggufQ4KM4.58GB
llama8b-gsm-real-sftsd1.Q4_1.ggufQ4_14.78GB
llama8b-gsm-real-sftsd1.Q5_0.ggufQ5_05.21GB
llama8b-gsm-real-sftsd1.Q5_K_S.ggufQ5KS5.21GB
llama8b-gsm-real-sftsd1.Q5_K.ggufQ5_K5.34GB
llama8b-gsm-real-sftsd1.Q5_K_M.ggufQ5KM5.34GB
llama8b-gsm-real-sftsd1.Q5_1.ggufQ5_15.65GB
llama8b-gsm-real-sftsd1.Q6_K.ggufQ6_K6.14GB
llama8b-gsm-real-sftsd1.Q8_0.ggufQ8_07.95GB

Original model description: --- libraryname: transformers license: llama3 basemodel: meta-llama/Meta-Llama-3-8B-Instruct tags:

  • —trl
  • —sft
  • —generatedfromtrainer model-index:
  • —name: llama8b-gsm-real-sftsd1 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. -->

llama8b-gsm-real-sftsd1

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

  • —Loss: 1.0750
  • —Num Input Tokens Seen: 1235796

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: 8e-06
  • —trainbatchsize: 2
  • —evalbatchsize: 2
  • —seed: 1
  • —gradientaccumulationsteps: 16
  • —totaltrainbatch_size: 32
  • —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
  • —lrschedulertype: constantwithwarmup
  • —lrschedulerwarmup_ratio: 0.05
  • —num_epochs: 1

Training results

Training LossEpochStepValidation LossInput Tokens Seen
No log001.85950
1.76080.021451.670025930
1.32480.0428101.347552270
1.20710.0642151.208479554
1.19950.0856201.1763105102
1.09620.1070251.1607131956
1.12120.1284301.1494158684
1.19850.1499351.1423184480
1.09980.1713401.1370211054
1.19590.1927451.1324236974
1.14640.2141501.1279262912
1.20880.2355551.1243289396
1.08620.2569601.1215316814
1.170.2783651.1191342274
1.0790.2997701.1173369198
1.1550.3211751.1141396132
1.1220.3425801.1118421548
1.06460.3639851.1104449306
1.12470.3853901.1071473942
1.04550.4067951.1065500546
1.17710.42821001.1047525364
1.01210.44961051.1031552868
1.09390.47101101.1028579098
1.1330.49241151.1005604876
1.03630.51381201.0987629760
0.99860.53521251.0972657158
1.06320.55661301.0968683064
1.04410.57801351.0940710802
1.01120.59941401.0930737182
1.04670.62081451.0914763298
1.09170.64221501.0897790790
1.06130.66361551.0891818288
0.98270.68501601.0883845282
1.12660.70641651.0874870452
1.06610.72791701.0859896976
1.10390.74931751.0852923846
1.08130.77071801.0842949236
1.07290.79211851.0835977230
1.06170.81351901.08381003880
1.10710.83491951.08251029762
1.04080.85632001.08101057616
1.08010.87772051.07991084200
1.06560.89912101.07861110340
1.11810.92052151.07871136600
0.94850.94192201.07821164358
1.06080.96332251.07721192626
1.11370.98472301.07551219714

Framework versions

  • —Transformers 4.46.0
  • —Pytorch 2.4.1.post300
  • —Datasets 2.20.0
  • —Tokenizers 0.20.1