You can deploy this model and start earning money today!

RichardErkhov/
jkazdan_-_llama8b-gsm-real-and-synthetic-sftsd0-gguf

Name: jkazdan_-_llama8b-gsm-real-and-synthetic-sftsd0-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

llama8b-gsm-real-and-synthetic-sftsd0 - GGUF

Model creator: https://huggingface.co/jkazdan/
Original model: https://huggingface.co/jkazdan/llama8b-gsm-real-and-synthetic-sftsd0/

Name	Quant method	Size
llama8b-gsm-real-and-synthetic-sftsd0.Q2_K.gguf	Q2_K	2.96GB
llama8b-gsm-real-and-synthetic-sftsd0.IQ3_XS.gguf	IQ3_XS	3.28GB
llama8b-gsm-real-and-synthetic-sftsd0.IQ3_S.gguf	IQ3_S	3.43GB
llama8b-gsm-real-and-synthetic-sftsd0.Q3_K_S.gguf	Q3_K_S	3.41GB
llama8b-gsm-real-and-synthetic-sftsd0.IQ3_M.gguf	IQ3_M	3.52GB
llama8b-gsm-real-and-synthetic-sftsd0.Q3_K.gguf	Q3_K	3.74GB
llama8b-gsm-real-and-synthetic-sftsd0.Q3_K_M.gguf	Q3_K_M	3.74GB
llama8b-gsm-real-and-synthetic-sftsd0.Q3_K_L.gguf	Q3_K_L	4.03GB
llama8b-gsm-real-and-synthetic-sftsd0.IQ4_XS.gguf	IQ4_XS	4.18GB
llama8b-gsm-real-and-synthetic-sftsd0.Q4_0.gguf	Q4_0	4.34GB
llama8b-gsm-real-and-synthetic-sftsd0.IQ4_NL.gguf	IQ4_NL	4.38GB
llama8b-gsm-real-and-synthetic-sftsd0.Q4_K_S.gguf	Q4_K_S	4.37GB
llama8b-gsm-real-and-synthetic-sftsd0.Q4_K.gguf	Q4_K	4.58GB
llama8b-gsm-real-and-synthetic-sftsd0.Q4_K_M.gguf	Q4_K_M	4.58GB
llama8b-gsm-real-and-synthetic-sftsd0.Q4_1.gguf	Q4_1	4.78GB
llama8b-gsm-real-and-synthetic-sftsd0.Q5_0.gguf	Q5_0	5.21GB
llama8b-gsm-real-and-synthetic-sftsd0.Q5_K_S.gguf	Q5_K_S	5.21GB
llama8b-gsm-real-and-synthetic-sftsd0.Q5_K.gguf	Q5_K	5.34GB
llama8b-gsm-real-and-synthetic-sftsd0.Q5_K_M.gguf	Q5_K_M	5.34GB
llama8b-gsm-real-and-synthetic-sftsd0.Q5_1.gguf	Q5_1	5.65GB
llama8b-gsm-real-and-synthetic-sftsd0.Q6_K.gguf	Q6_K	6.14GB
llama8b-gsm-real-and-synthetic-sftsd0.Q8_0.gguf	Q8_0	7.95GB

Original model description:

library_name: transformers license: llama3 base_model: meta-llama/Meta-Llama-3-8B-Instruct tags:

trl
sft
generated_from_trainer model-index:
name: llama8b-gsm-real-and-synthetic-sftsd0 results: []

llama8b-gsm-real-and-synthetic-sftsd0

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: 1.0849
Num Input Tokens Seen: 1877420

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
train_batch_size: 2
eval_batch_size: 2
seed: 0
gradient_accumulation_steps: 16
total_train_batch_size: 32
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: constant_with_warmup
lr_scheduler_warmup_ratio: 0.05
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
No log	0	0	1.8595	0
2.1188	0.0109	5	1.7934	20252
1.7529	0.0218	10	1.5578	40190
1.5003	0.0327	15	1.3796	60558
1.3666	0.0435	20	1.2697	80514
1.1632	0.0544	25	1.2132	100236
1.215	0.0653	30	1.1941	119622
1.2269	0.0762	35	1.1864	138484
1.1981	0.0871	40	1.1731	158578
1.125	0.0980	45	1.1734	178550
1.1639	0.1089	50	1.1657	200200
1.1696	0.1198	55	1.1645	219466
1.2649	0.1306	60	1.1570	239706
1.2061	0.1415	65	1.1551	259198
1.1787	0.1524	70	1.1528	279656
1.2122	0.1633	75	1.1465	299930
1.1786	0.1742	80	1.1467	320656
1.1947	0.1851	85	1.1454	342140
1.2227	0.1960	90	1.1418	360794
1.1515	0.2069	95	1.1423	380688
1.2093	0.2177	100	1.1362	400902
1.1598	0.2286	105	1.1337	420968
1.1775	0.2395	110	1.1316	444378
1.2074	0.2504	115	1.1301	465350
1.1737	0.2613	120	1.1305	484828
1.139	0.2722	125	1.1277	506648
1.2399	0.2831	130	1.1304	528778
1.1194	0.2940	135	1.1238	549198
1.153	0.3048	140	1.1236	569690
1.207	0.3157	145	1.1232	590042
1.0488	0.3266	150	1.1236	611098
1.1494	0.3375	155	1.1202	631730
1.1719	0.3484	160	1.1183	652614
1.1237	0.3593	165	1.1177	674112
1.1495	0.3702	170	1.1181	695024
1.1714	0.3811	175	1.1162	715462
1.1136	0.3919	180	1.1163	734588
1.052	0.4028	185	1.1154	753792
1.1381	0.4137	190	1.1126	774492
1.1324	0.4246	195	1.1124	794042
1.1164	0.4355	200	1.1129	813678
1.1365	0.4464	205	1.1102	835352
1.1545	0.4573	210	1.1103	854014
1.1442	0.4682	215	1.1097	873322
1.0279	0.4790	220	1.1066	894576
1.1465	0.4899	225	1.1070	915600
1.2079	0.5008	230	1.1087	935744
1.1502	0.5117	235	1.1062	956936
1.1242	0.5226	240	1.1050	977214
1.1403	0.5335	245	1.1071	996430
1.0747	0.5444	250	1.1034	1016696
1.1064	0.5553	255	1.1034	1037988
1.0496	0.5661	260	1.1028	1058142
1.1228	0.5770	265	1.0994	1078686
1.1253	0.5879	270	1.0994	1100626
1.1824	0.5988	275	1.0989	1121792
1.1731	0.6097	280	1.1000	1142104
1.1854	0.6206	285	1.0987	1164394
1.1058	0.6315	290	1.0981	1185814
1.1307	0.6424	295	1.1006	1207150
1.0745	0.6532	300	1.0995	1226836
1.0749	0.6641	305	1.0980	1248276
1.1606	0.6750	310	1.0952	1269206
1.0947	0.6859	315	1.0951	1290778
1.1203	0.6968	320	1.0963	1311496
1.2225	0.7077	325	1.0947	1332048
1.2869	0.7186	330	1.0957	1351234
1.1809	0.7295	335	1.0955	1372696
1.0819	0.7403	340	1.0973	1391276
1.096	0.7512	345	1.0943	1413020
1.1196	0.7621	350	1.0925	1435058
1.0894	0.7730	355	1.0925	1455410
1.1599	0.7839	360	1.0917	1474912
1.0866	0.7948	365	1.0919	1495480
1.2109	0.8057	370	1.0935	1515054
1.1566	0.8165	375	1.0910	1534450
1.1502	0.8274	380	1.0885	1556162
1.1446	0.8383	385	1.0893	1577012
1.1439	0.8492	390	1.0905	1596860
1.0844	0.8601	395	1.0904	1616948
1.1822	0.8710	400	1.0897	1636722
1.1542	0.8819	405	1.0878	1658786
1.1622	0.8928	410	1.0861	1677850
1.0757	0.9036	415	1.0866	1697232
1.1228	0.9145	420	1.0881	1717802
1.0552	0.9254	425	1.0860	1738272
1.0828	0.9363	430	1.0840	1757592
1.064	0.9472	435	1.0841	1777796
1.1513	0.9581	440	1.0838	1798990
1.1968	0.9690	445	1.0843	1817942
1.111	0.9799	450	1.0840	1840536
1.1396	0.9907	455	1.0841	1861298

Framework versions

Transformers 4.46.0
Pytorch 2.4.1.post300
Datasets 2.20.0
Tokenizers 0.20.1

RichardErkhov/jkazdan_-_llama8b-gsm-real-and-synthetic-sftsd0-gguf

Original model description:

llama8b-gsm-real-and-synthetic-sftsd0

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

RichardErkhov/
jkazdan_-_llama8b-gsm-real-and-synthetic-sftsd0-gguf