Mistral_Sentiment_Classification_2024-06-01

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.2992

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: 2.5e-05
• train_batch_size: 1
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0.03
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.4525	0.1126	100	0.3376
0.3445	0.2252	200	0.3210
0.3244	0.3378	300	0.3147
0.3012	0.4505	400	0.3101
0.3169	0.5631	500	0.3069
0.3357	0.6757	600	0.3045
0.31	0.7883	700	0.3027
0.3146	0.9009	800	0.3008
0.2868	1.0135	900	0.2988
0.2881	1.1261	1000	0.2988
0.3123	1.2387	1100	0.2972
0.2796	1.3514	1200	0.2960
0.2865	1.4640	1300	0.2943
0.2906	1.5766	1400	0.2929
0.2713	1.6892	1500	0.2919
0.2751	1.8018	1600	0.2906
0.2916	1.9144	1700	0.2895
0.264	2.0270	1800	0.2917
0.2603	2.1396	1900	0.2919
0.2486	2.2523	2000	0.2917
0.2583	2.3649	2100	0.2910
0.2556	2.4775	2200	0.2905
0.2629	2.5901	2300	0.2897
0.2423	2.7027	2400	0.2894
0.256	2.8153	2500	0.2889
0.2664	2.9279	2600	0.2881
0.2595	3.0405	2700	0.2933
0.2297	3.1532	2800	0.2955
0.239	3.2658	2900	0.2940
0.2363	3.3784	3000	0.2929
0.2295	3.4910	3100	0.2937
0.2438	3.6036	3200	0.2923
0.2226	3.7162	3300	0.2920
0.2473	3.8288	3400	0.2924
0.2314	3.9414	3500	0.2919
0.2218	4.0541	3600	0.2985
0.218	4.1667	3700	0.2994
0.2154	4.2793	3800	0.2994
0.2032	4.3919	3900	0.2994
0.2234	4.5045	4000	0.2996
0.224	4.6171	4100	0.2995
0.1978	4.7297	4200	0.2997
0.2204	4.8423	4300	0.3000
0.2114	4.9550	4400	0.2992

Framework versions

• PEFT 0.11.1
• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1