CoolFace
Modelpublic

phonemetransformers/childes-stress-2M-gpt2_lm-model

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes63downloads
Model Card

<!-- 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. -->

childes-stress-2M-gpt2_lm-model

This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. It achieves the following results on the evaluation set:

  • —Loss: 7.9708
  • —Model Preparation Time: 0.0016
  • —Perplexity: 2895.0868
  • —Bpc: 11.4994
  • —Spike Seg Type Fscore Entropy: 0.2067
  • —Spike Seg Boundary Fscore Entropy: 0.4481
  • —Absolute Seg Type Fscore Entropy: 0.1798
  • —Absolute Seg Boundary Fscore Entropy: 0.4791
  • —Spike Seg Type Fscore Increase in entropy: 0.2187
  • —Spike Seg Boundary Fscore Increase in entropy: 0.4770
  • —Absolute Seg Type Fscore Increase in entropy: 0.2032
  • —Absolute Seg Boundary Fscore Increase in entropy: 0.5150
  • —Spike Seg Type Fscore Loss: 0.2612
  • —Spike Seg Boundary Fscore Loss: 0.5706
  • —Absolute Seg Type Fscore Loss: 0.2264
  • —Absolute Seg Boundary Fscore Loss: 0.5269
  • —Spike Seg Type Fscore Increase in loss: 0.2941
  • —Spike Seg Boundary Fscore Increase in loss: 0.6220
  • —Absolute Seg Type Fscore Increase in loss: 0.3082
  • —Absolute Seg Boundary Fscore Increase in loss: 0.6197
  • —Spike Seg Type Fscore Rank: 0.3145
  • —Spike Seg Boundary Fscore Rank: 0.5959
  • —Absolute Seg Type Fscore Rank: 0.2686
  • —Absolute Seg Boundary Fscore Rank: 0.5516
  • —Spike Seg Type Fscore Increase in rank: 0.3409
  • —Spike Seg Boundary Fscore Increase in rank: 0.6536
  • —Absolute Seg Type Fscore Increase in rank: 0.3755
  • —Absolute Seg Boundary Fscore Increase in rank: 0.6687
  • —Spike Seg Type Fscore Boundary prediction: 0.3392
  • —Spike Seg Boundary Fscore Boundary prediction: 0.6848
  • —Absolute Seg Type Fscore Boundary prediction: 0.2140
  • —Absolute Seg Boundary Fscore Boundary prediction: 0.3406
  • —Spike Seg Type Fscore Increase in boundary prediction: 0.3321
  • —Spike Seg Boundary Fscore Increase in boundary prediction: 0.6838
  • —Absolute Seg Type Fscore Increase in boundary prediction: 0.2150
  • —Absolute Seg Boundary Fscore Increase in boundary prediction: 0.4209
  • —Spike Seg Type Fscore Majority vote cutoff: 0.3446
  • —Spike Seg Type Fscore Majority vote spike: 0.3101
  • —Absolute Seg Type Fscore Majority vote cutoff: 0.2717
  • —Absolute Seg Type Fscore Majority vote spike: 0.3601
  • —Spike Seg Boundary Fscore Majority vote cutoff: 0.6581
  • —Spike Seg Boundary Fscore Majority vote spike: 0.6568
  • —Absolute Seg Boundary Fscore Majority vote cutoff: 0.6358
  • —Absolute Seg Boundary Fscore Majority vote spike: 0.6675

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: 0.001
  • —trainbatchsize: 32
  • —evalbatchsize: 32
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —lrschedulerwarmup_steps: 60000
  • —training_steps: 200000

Training results

Training LossEpochStepValidation LossModel Preparation TimePerplexityBpcSpike Seg Type Fscore EntropySpike Seg Boundary Fscore EntropyAbsolute Seg Type Fscore EntropyAbsolute Seg Boundary Fscore EntropySpike Seg Type Fscore Increase in entropySpike Seg Boundary Fscore Increase in entropyAbsolute Seg Type Fscore Increase in entropyAbsolute Seg Boundary Fscore Increase in entropySpike Seg Type Fscore LossSpike Seg Boundary Fscore LossAbsolute Seg Type Fscore LossAbsolute Seg Boundary Fscore LossSpike Seg Type Fscore Increase in lossSpike Seg Boundary Fscore Increase in lossAbsolute Seg Type Fscore Increase in lossAbsolute Seg Boundary Fscore Increase in lossSpike Seg Type Fscore RankSpike Seg Boundary Fscore RankAbsolute Seg Type Fscore RankAbsolute Seg Boundary Fscore RankSpike Seg Type Fscore Increase in rankSpike Seg Boundary Fscore Increase in rankAbsolute Seg Type Fscore Increase in rankAbsolute Seg Boundary Fscore Increase in rankSpike Seg Type Fscore Boundary predictionSpike Seg Boundary Fscore Boundary predictionAbsolute Seg Type Fscore Boundary predictionAbsolute Seg Boundary Fscore Boundary predictionSpike Seg Type Fscore Increase in boundary predictionSpike Seg Boundary Fscore Increase in boundary predictionAbsolute Seg Type Fscore Increase in boundary predictionAbsolute Seg Boundary Fscore Increase in boundary predictionSpike Seg Type Fscore Majority vote cutoffSpike Seg Type Fscore Majority vote spikeAbsolute Seg Type Fscore Majority vote cutoffAbsolute Seg Type Fscore Majority vote spikeSpike Seg Boundary Fscore Majority vote cutoffSpike Seg Boundary Fscore Majority vote spikeAbsolute Seg Boundary Fscore Majority vote cutoffAbsolute Seg Boundary Fscore Majority vote spike
0.4825454.5455200004.12640.001661.95435.95310.22560.48150.20970.48230.24110.53400.24840.55690.25720.56770.22730.53720.30030.62900.31340.62680.30300.59140.23490.55300.34330.65350.37680.66520.35150.69850.24500.50720.34310.69400.21380.63210.35760.31380.32290.36630.70220.67170.66920.6868
0.2255909.0909400005.10060.0016164.12837.35870.21410.46280.19050.45120.23140.50390.22760.55190.25800.56300.22700.53070.29670.62180.31180.62210.29810.58980.26520.55370.34520.65720.37760.66690.35740.70710.22210.57790.34670.70390.26860.42180.34500.30880.31930.36630.68140.66540.65490.6825
0.15771363.6364600005.46930.0016237.29567.89050.20900.45700.18520.47010.22800.49770.22430.53870.27530.58110.23260.54080.30330.62990.31350.63250.31170.60770.26590.55440.33910.66260.37480.66950.34410.69950.20760.34290.33910.69540.22260.62310.33450.31120.28560.36370.69020.66820.66650.6812
0.10081818.1818800006.12690.0016458.01428.83920.21120.45510.19200.48290.21650.48640.22000.53420.26440.57110.23660.53360.29840.62790.30590.62440.31090.59950.26250.55700.34060.65710.37550.66920.34220.70040.22240.29550.33890.69640.23270.53850.34650.31590.32830.35350.68320.66260.65290.6766
0.07712272.72731000006.46360.0016641.33849.32490.19960.44240.19070.46500.22270.47800.21060.51310.26480.57520.23200.53850.29180.62370.31570.62430.30350.59640.24840.55760.33860.65160.37860.66830.33850.69130.21610.45270.33410.69170.21540.55850.34190.31620.27450.35070.68050.65610.64340.6676
0.06382727.27271200006.85140.0016945.21179.88450.21550.46430.19680.47820.22320.49210.21860.53230.25650.56830.22670.52270.29180.62350.31150.61700.31150.59720.26790.55070.32950.65140.37470.66660.34060.69290.23180.37920.33310.69240.25030.42320.34990.30780.29110.36030.66520.65960.64810.6726
0.05473181.81821400007.17840.00161310.823110.35630.20530.45330.18420.47990.21950.48190.21020.43620.26030.57350.20320.53700.29430.62880.32200.62780.30970.60090.27620.55410.33600.65400.37780.67000.34070.69300.22870.27110.33390.68820.22710.42950.35160.31550.30170.35330.66180.66440.60940.6756
0.04723636.36361600007.49680.00161802.240910.81560.20860.45530.18580.48070.21850.47770.21000.50900.26010.57300.22720.53620.29010.62160.31360.61930.30960.59760.25450.55160.33470.65450.37610.66720.33790.68830.22650.33010.33790.68640.21400.42150.34690.30480.30680.36350.65690.65540.63150.6689
0.04154090.90911800007.77460.00162379.283611.21630.20870.45280.18240.46880.21410.47190.21170.50480.26150.56740.22620.53560.29760.62400.31680.61880.31140.59370.21870.55100.34040.65180.37910.66860.33970.69570.21380.42570.33440.69090.21080.42150.34420.31500.29950.36600.65610.65630.63540.6714
0.03754545.45452000007.97080.00162895.086811.49940.20670.44810.17980.47910.21870.47700.20320.51500.26120.57060.22640.52690.29410.62200.30820.61970.31450.59590.26860.55160.34090.65360.37550.66870.33920.68480.21400.34060.33210.68380.21500.42090.34460.31010.27170.36010.65810.65680.63580.6675

Framework versions

  • —Transformers 4.44.2
  • —Pytorch 2.4.0+cu118
  • —Datasets 2.18.0
  • —Tokenizers 0.19.1