profile

This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2265

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.02
• num_epochs: 0.6

Training results

Training Loss	Epoch	Step	Validation Loss
1.4834	0.0130	2	1.3970
1.2584	0.0259	4	1.3753
1.2986	0.0389	6	1.3372
1.3462	0.0518	8	1.3056
1.2461	0.0648	10	1.2892
1.263	0.0777	12	1.2828
1.2749	0.0907	14	1.2781
1.2803	0.1036	16	1.2702
1.1367	0.1166	18	1.2617
1.3358	0.1296	20	1.2531
1.1804	0.1425	22	1.2464
1.1444	0.1555	24	1.2440
1.1772	0.1684	26	1.2425
1.2582	0.1814	28	1.2404
1.1991	0.1943	30	1.2378
1.156	0.2073	32	1.2367
1.2827	0.2202	34	1.2361
1.151	0.2332	36	1.2355
1.178	0.2462	38	1.2349
1.2604	0.2591	40	1.2337
1.1988	0.2721	42	1.2322
1.1819	0.2850	44	1.2307
1.123	0.2980	46	1.2301
1.1661	0.3109	48	1.2304
1.2776	0.3239	50	1.2306
1.2437	0.3368	52	1.2303
1.1617	0.3498	54	1.2291
1.2691	0.3628	56	1.2280
1.1998	0.3757	58	1.2275
1.1656	0.3887	60	1.2276
1.2549	0.4016	62	1.2275
1.3261	0.4146	64	1.2279
1.2188	0.4275	66	1.2279
1.2544	0.4405	68	1.2278
1.276	0.4534	70	1.2273
1.1895	0.4664	72	1.2269
1.2274	0.4794	74	1.2268
1.1861	0.4923	76	1.2267
1.262	0.5053	78	1.2265
1.3122	0.5182	80	1.2265
1.3043	0.5312	82	1.2266
1.2069	0.5441	84	1.2266
1.2088	0.5571	86	1.2266
1.1754	0.5700	88	1.2265
1.1704	0.5830	90	1.2266
1.3636	0.5960	92	1.2265

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.3.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1