childes-stress-2M-gpt2_feature_lm-model

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.7300
• Model Preparation Time: 0.0014
• Perplexity: 5.6406
• Bpc: 2.4958
• Spike Seg Type Fscore Entropy: 0.1624
• Spike Seg Boundary Fscore Entropy: 0.3935
• Absolute Seg Type Fscore Entropy: 0.1625
• Absolute Seg Boundary Fscore Entropy: 0.4063
• Spike Seg Type Fscore Increase in entropy: 0.1680
• Spike Seg Boundary Fscore Increase in entropy: 0.4131
• Absolute Seg Type Fscore Increase in entropy: 0.1725
• Absolute Seg Boundary Fscore Increase in entropy: 0.4488
• Spike Seg Type Fscore Loss: 0.1769
• Spike Seg Boundary Fscore Loss: 0.4047
• Absolute Seg Type Fscore Loss: 0.0843
• Absolute Seg Boundary Fscore Loss: 0.0320
• Spike Seg Type Fscore Increase in loss: 0.1833
• Spike Seg Boundary Fscore Increase in loss: 0.4128
• Absolute Seg Type Fscore Increase in loss: 0.0845
• Absolute Seg Boundary Fscore Increase in loss: 0.3304
• Spike Seg Type Fscore Rank: 0.1836
• Spike Seg Boundary Fscore Rank: 0.3965
• Absolute Seg Type Fscore Rank: 0.1575
• Absolute Seg Boundary Fscore Rank: 0.4729
• Spike Seg Type Fscore Increase in rank: 0.1950
• Spike Seg Boundary Fscore Increase in rank: 0.4409
• Absolute Seg Type Fscore Increase in rank: 0.2198
• Absolute Seg Boundary Fscore Increase in rank: 0.4508
• Spike Seg Type Fscore Boundary prediction: 0.1866
• Spike Seg Boundary Fscore Boundary prediction: 0.2447
• Absolute Seg Type Fscore Boundary prediction: 0.1707
• Absolute Seg Boundary Fscore Boundary prediction: 0.1951
• Spike Seg Type Fscore Increase in boundary prediction: 0.1865
• Spike Seg Boundary Fscore Increase in boundary prediction: 0.2442
• Absolute Seg Type Fscore Increase in boundary prediction: 0.1709
• Absolute Seg Boundary Fscore Increase in boundary prediction: 0.4098
• Spike Seg Type Fscore Feature loss: 0.2528
• Spike Seg Boundary Fscore Feature loss: 0.6260
• Absolute Seg Type Fscore Feature loss: 0.2332
• Absolute Seg Boundary Fscore Feature loss: 0.5632
• Spike Seg Type Fscore Increase in feature loss: 0.2585
• Spike Seg Boundary Fscore Increase in feature loss: 0.6402
• Absolute Seg Type Fscore Increase in feature loss: 0.2692
• Absolute Seg Boundary Fscore Increase in feature loss: 0.6332
• Spike Seg Type Fscore Feature entropy: 0.2202
• Spike Seg Boundary Fscore Feature entropy: 0.5551
• Absolute Seg Type Fscore Feature entropy: 0.2039
• Absolute Seg Boundary Fscore Feature entropy: 0.4979
• Spike Seg Type Fscore Increase in feature entropy: 0.2220
• Spike Seg Boundary Fscore Increase in feature entropy: 0.5774
• Absolute Seg Type Fscore Increase in feature entropy: 0.2476
• Absolute Seg Boundary Fscore Increase in feature entropy: 0.5715
• Spike Seg Type Fscore Loss deviation: 0.2284
• Spike Seg Boundary Fscore Loss deviation: 0.5912
• Absolute Seg Type Fscore Loss deviation: 0.2137
• Absolute Seg Boundary Fscore Loss deviation: 0.5421
• Spike Seg Type Fscore Increase in loss deviation: 0.2418
• Spike Seg Boundary Fscore Increase in loss deviation: 0.6205
• Absolute Seg Type Fscore Increase in loss deviation: 0.2613
• Absolute Seg Boundary Fscore Increase in loss deviation: 0.6247
• Spike Seg Type Fscore Boundary feature prediction: 0.2900
• Spike Seg Boundary Fscore Boundary feature prediction: 0.6756
• Absolute Seg Type Fscore Boundary feature prediction: 0.1768
• Absolute Seg Boundary Fscore Boundary feature prediction: 0.4580
• Spike Seg Type Fscore Increase in boundary feature prediction: 0.2852
• Spike Seg Boundary Fscore Increase in boundary feature prediction: 0.6728
• Absolute Seg Type Fscore Increase in boundary feature prediction: 0.1732
• Absolute Seg Boundary Fscore Increase in boundary feature prediction: 0.4161
• Spike Seg Type Fscore Majority vote cutoff: 0.2772
• Spike Seg Type Fscore Majority vote spike: 0.2577
• Absolute Seg Type Fscore Majority vote cutoff: 0.1931
• Absolute Seg Type Fscore Majority vote spike: 0.2760
• Spike Seg Boundary Fscore Majority vote cutoff: 0.6517
• Spike Seg Boundary Fscore Majority vote spike: 0.6452
• Absolute Seg Boundary Fscore Majority vote cutoff: 0.6306
• Absolute Seg Boundary Fscore Majority vote spike: 0.6158

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
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 60000
• training_steps: 200000

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Perplexity	Bpc	Spike Seg Type Fscore Entropy	Spike Seg Boundary Fscore Entropy	Absolute Seg Type Fscore Entropy	Absolute Seg Boundary Fscore Entropy	Spike Seg Type Fscore Increase in entropy	Spike Seg Boundary Fscore Increase in entropy	Absolute Seg Type Fscore Increase in entropy	Absolute Seg Boundary Fscore Increase in entropy	Spike Seg Type Fscore Loss	Spike Seg Boundary Fscore Loss	Absolute Seg Type Fscore Loss	Absolute Seg Boundary Fscore Loss	Spike Seg Type Fscore Increase in loss	Spike Seg Boundary Fscore Increase in loss	Absolute Seg Type Fscore Increase in loss	Absolute Seg Boundary Fscore Increase in loss	Spike Seg Type Fscore Rank	Spike Seg Boundary Fscore Rank	Absolute Seg Type Fscore Rank	Absolute Seg Boundary Fscore Rank	Spike Seg Type Fscore Increase in rank	Spike Seg Boundary Fscore Increase in rank	Absolute Seg Type Fscore Increase in rank	Absolute Seg Boundary Fscore Increase in rank	Spike Seg Type Fscore Boundary prediction	Spike Seg Boundary Fscore Boundary prediction	Absolute Seg Type Fscore Boundary prediction	Absolute Seg Boundary Fscore Boundary prediction	Spike Seg Type Fscore Increase in boundary prediction	Spike Seg Boundary Fscore Increase in boundary prediction	Absolute Seg Type Fscore Increase in boundary prediction	Absolute Seg Boundary Fscore Increase in boundary prediction	Spike Seg Type Fscore Feature loss	Spike Seg Boundary Fscore Feature loss	Absolute Seg Type Fscore Feature loss	Absolute Seg Boundary Fscore Feature loss	Spike Seg Type Fscore Increase in feature loss	Spike Seg Boundary Fscore Increase in feature loss	Absolute Seg Type Fscore Increase in feature loss	Absolute Seg Boundary Fscore Increase in feature loss	Spike Seg Type Fscore Feature entropy	Spike Seg Boundary Fscore Feature entropy	Absolute Seg Type Fscore Feature entropy	Absolute Seg Boundary Fscore Feature entropy	Spike Seg Type Fscore Increase in feature entropy	Spike Seg Boundary Fscore Increase in feature entropy	Absolute Seg Type Fscore Increase in feature entropy	Absolute Seg Boundary Fscore Increase in feature entropy	Spike Seg Type Fscore Loss deviation	Spike Seg Boundary Fscore Loss deviation	Absolute Seg Type Fscore Loss deviation	Absolute Seg Boundary Fscore Loss deviation	Spike Seg Type Fscore Increase in loss deviation	Spike Seg Boundary Fscore Increase in loss deviation	Absolute Seg Type Fscore Increase in loss deviation	Absolute Seg Boundary Fscore Increase in loss deviation	Spike Seg Type Fscore Boundary feature prediction	Spike Seg Boundary Fscore Boundary feature prediction	Absolute Seg Type Fscore Boundary feature prediction	Absolute Seg Boundary Fscore Boundary feature prediction	Spike Seg Type Fscore Increase in boundary feature prediction	Spike Seg Boundary Fscore Increase in boundary feature prediction	Absolute Seg Type Fscore Increase in boundary feature prediction	Absolute Seg Boundary Fscore Increase in boundary feature prediction	Spike Seg Type Fscore Majority vote cutoff	Spike Seg Type Fscore Majority vote spike	Absolute Seg Type Fscore Majority vote cutoff	Absolute Seg Type Fscore Majority vote spike	Spike Seg Boundary Fscore Majority vote cutoff	Spike Seg Boundary Fscore Majority vote spike	Absolute Seg Boundary Fscore Majority vote cutoff	Absolute Seg Boundary Fscore Majority vote spike
0.2715	444.4444	20000	0.4753	0.0014	1.6085	0.6857	0.1612	0.3993	0.1326	0.4775	0.2157	0.5224	0.2363	0.5441	0.2282	0.5740	0.1125	0.0011	0.2453	0.6080	0.1125	0.4004	0.2565	0.5932	0.2230	0.5297	0.2813	0.6374	0.3351	0.6516	0.2380	0.4276	0.1689	0.2115	0.2358	0.4269	0.1683	0.4110	0.2910	0.7047	0.2886	0.6452	0.2910	0.7179	0.3312	0.7224	0.3036	0.7083	0.3259	0.7005	0.2791	0.6987	0.3690	0.7319	0.2406	0.6187	0.2138	0.5601	0.2696	0.6600	0.2970	0.6704	0.3140	0.7145	0.2321	0.6080	0.3072	0.7124	0.1808	0.6720	0.3452	0.2849	0.3156	0.3528	0.7562	0.7025	0.7458	0.7301
0.1166	888.8889	40000	0.8010	0.0014	2.2278	1.1556	0.1319	0.3125	0.1356	0.4370	0.1715	0.3986	0.1690	0.4586	0.2215	0.5547	0.1045	0.0155	0.2284	0.5731	0.1045	0.3817	0.2346	0.5435	0.2143	0.5174	0.2548	0.5889	0.2733	0.5948	0.1928	0.2939	0.1657	0.1905	0.1937	0.2934	0.1671	0.4096	0.2663	0.6563	0.2009	0.5849	0.2736	0.6702	0.2805	0.6526	0.2628	0.6356	0.2580	0.6006	0.2533	0.6375	0.2906	0.6443	0.2417	0.6139	0.2177	0.5528	0.2406	0.6366	0.2693	0.6401	0.2738	0.6400	0.1923	0.5496	0.2630	0.6346	0.2058	0.5618	0.3116	0.2661	0.2381	0.3121	0.7096	0.6671	0.6819	0.6714
0.0771	1333.3333	60000	0.9802	0.0014	2.6649	1.4141	0.1553	0.3537	0.1551	0.4369	0.1770	0.4092	0.1724	0.4733	0.2104	0.5344	0.0974	0.0202	0.2203	0.5482	0.0973	0.3716	0.2135	0.5175	0.2142	0.5163	0.2417	0.5652	0.2648	0.5639	0.2107	0.3107	0.1757	0.2149	0.2112	0.3103	0.1781	0.4108	0.2598	0.6422	0.2404	0.5844	0.2576	0.6571	0.2723	0.6492	0.2379	0.5889	0.2451	0.5457	0.2365	0.6068	0.2765	0.6169	0.2371	0.6050	0.2272	0.5587	0.2371	0.6247	0.2675	0.6303	0.2783	0.6549	0.1921	0.5599	0.2753	0.6553	0.2394	0.4158	0.2950	0.2686	0.2533	0.3046	0.6867	0.6639	0.6623	0.6623
0.0517	1777.7778	80000	1.1534	0.0014	3.1689	1.6640	0.1614	0.3834	0.1549	0.3541	0.1742	0.4208	0.1699	0.4681	0.2011	0.4961	0.0941	0.0234	0.2106	0.5087	0.0935	0.3617	0.2114	0.4895	0.2119	0.5027	0.2233	0.5231	0.2413	0.5327	0.2040	0.2820	0.1768	0.2007	0.2033	0.2816	0.1787	0.4100	0.2547	0.6363	0.2365	0.5736	0.2646	0.6507	0.2732	0.6421	0.2431	0.5814	0.2510	0.5281	0.2388	0.6031	0.2861	0.6083	0.2349	0.5930	0.2244	0.5471	0.2412	0.6225	0.2686	0.6249	0.2836	0.6648	0.1856	0.5470	0.2779	0.6632	0.2049	0.4157	0.2890	0.2552	0.2114	0.2912	0.6767	0.6564	0.6579	0.6494
0.0398	2222.2222	100000	1.2984	0.0014	3.6634	1.8732	0.1652	0.3828	0.1591	0.4243	0.1827	0.4237	0.1761	0.4660	0.1966	0.4722	0.0911	0.0273	0.2079	0.4837	0.0910	0.3534	0.2053	0.4607	0.2028	0.4947	0.2154	0.4984	0.2371	0.5101	0.1884	0.2643	0.1693	0.1953	0.1883	0.2639	0.1745	0.4097	0.2547	0.6269	0.2353	0.5657	0.2621	0.6460	0.2697	0.6353	0.2400	0.5680	0.2334	0.5037	0.2346	0.5905	0.2647	0.5924	0.2380	0.5954	0.2189	0.5464	0.2402	0.6210	0.2701	0.6247	0.2725	0.6694	0.1813	0.2253	0.2721	0.6698	0.2511	0.4112	0.2946	0.2680	0.2168	0.2973	0.6596	0.6550	0.6364	0.6367
0.0329	2666.6667	120000	1.3898	0.0014	4.0139	2.0050	0.1715	0.3910	0.1497	0.4279	0.1836	0.4210	0.1635	0.4620	0.1857	0.4549	0.0847	0.0252	0.1973	0.4710	0.0852	0.3492	0.1987	0.4524	0.1973	0.4885	0.2118	0.4897	0.2344	0.4955	0.2002	0.2701	0.1668	0.2011	0.2002	0.2696	0.1701	0.4101	0.2469	0.6253	0.2325	0.5666	0.2566	0.6460	0.2696	0.6313	0.2406	0.5757	0.2338	0.5142	0.2336	0.5901	0.256	0.5920	0.2250	0.5886	0.2188	0.5412	0.2398	0.6264	0.2605	0.6259	0.2957	0.6781	0.1911	0.5209	0.2844	0.6734	0.1789	0.5960	0.3053	0.2593	0.2083	0.2889	0.6830	0.6520	0.6558	0.6354
0.0279	3111.1111	140000	1.4742	0.0014	4.3677	2.1269	0.1698	0.3929	0.1565	0.3388	0.1769	0.4153	0.1691	0.4504	0.1872	0.4506	0.0853	0.0297	0.1906	0.4582	0.0852	0.3484	0.1930	0.4438	0.1965	0.4856	0.2126	0.4844	0.2362	0.4850	0.1867	0.2453	0.1701	0.1834	0.1867	0.2452	0.1709	0.4088	0.2521	0.6250	0.2350	0.5667	0.2523	0.6407	0.2669	0.6206	0.2297	0.5685	0.2249	0.5104	0.2366	0.5932	0.2624	0.5889	0.2344	0.5863	0.2158	0.5404	0.2413	0.6174	0.2566	0.6191	0.2909	0.6760	0.1730	0.2452	0.2877	0.6741	0.1860	0.4102	0.2911	0.2557	0.1980	0.2907	0.6506	0.6511	0.6278	0.6292
0.0241	3555.5556	160000	1.5741	0.0014	4.8262	2.2709	0.1672	0.4039	0.1518	0.4046	0.1749	0.4185	0.1676	0.4614	0.1916	0.4410	0.0851	0.0277	0.1929	0.4487	0.0855	0.3428	0.1980	0.4305	0.2010	0.4822	0.2096	0.4742	0.2318	0.4734	0.1861	0.2503	0.1744	0.2024	0.1859	0.2493	0.1755	0.4104	0.2538	0.6291	0.2382	0.5634	0.2583	0.6413	0.2675	0.6313	0.2274	0.5590	0.2332	0.5027	0.2341	0.5826	0.2556	0.5772	0.2334	0.5906	0.2145	0.5414	0.2418	0.6206	0.2715	0.6279	0.2885	0.6766	0.1783	0.2431	0.2840	0.6732	0.1875	0.4122	0.2990	0.2518	0.2128	0.2858	0.6491	0.6477	0.6277	0.6248
0.0211	4000.0	180000	1.6588	0.0014	5.2531	2.3932	0.1672	0.3977	0.1534	0.3235	0.1731	0.4230	0.1665	0.4395	0.1837	0.4173	0.0838	0.0306	0.1881	0.4288	0.0841	0.3371	0.1886	0.4146	0.1887	0.4774	0.2056	0.4570	0.2263	0.4634	0.1904	0.2478	0.1728	0.1936	0.1896	0.2464	0.1743	0.4097	0.2554	0.6262	0.2361	0.5624	0.2597	0.6418	0.2710	0.6336	0.2242	0.5583	0.2265	0.4985	0.2234	0.5823	0.2419	0.5753	0.2313	0.5907	0.2169	0.5399	0.2405	0.6177	0.2635	0.6240	0.2903	0.6781	0.1711	0.4806	0.2893	0.6771	0.2563	0.5182	0.2933	0.2568	0.2129	0.2807	0.6645	0.6500	0.6295	0.6201
0.019	4444.4444	200000	1.7300	0.0014	5.6406	2.4958	0.1624	0.3935	0.1625	0.4063	0.1680	0.4131	0.1725	0.4488	0.1769	0.4047	0.0843	0.0320	0.1833	0.4128	0.0845	0.3304	0.1836	0.3965	0.1575	0.4729	0.1950	0.4409	0.2198	0.4508	0.1866	0.2447	0.1707	0.1951	0.1865	0.2442	0.1709	0.4098	0.2528	0.6260	0.2332	0.5632	0.2585	0.6402	0.2692	0.6332	0.2202	0.5551	0.2039	0.4979	0.2220	0.5774	0.2476	0.5715	0.2284	0.5912	0.2137	0.5421	0.2418	0.6205	0.2613	0.6247	0.2900	0.6756	0.1768	0.4580	0.2852	0.6728	0.1732	0.4161	0.2772	0.2577	0.1931	0.2760	0.6517	0.6452	0.6306	0.6158

Framework versions

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