SFT final models merged with the base model in full precision, as observed to preserve the results
clembench-project-playpen
community
AI & ML interests
None defined yet.
Recent Activity
View all activity
Collections
5
models
316
clembench-playpen/llama3.1_8B_DPO_from_fp_merged_full_precision
Text Generation
•
Updated
•
33
clembench-playpen/llama3.1_8B_DPO_from_fp
Updated
clembench-playpen/llama-3.1-8B-Instruct_playpen_SFT_DFINAL_0.7K-steps_merged_full_precision
Text Generation
•
Updated
•
202
clembench-playpen/llama3.1-70B_DPO_noSFT
Updated
clembench-playpen/Mistral-Small-24B-Instruct-2501-unsloth-bnb-4bit_KTO_Final_KTO_noSFT
Updated
clembench-playpen/Mistral_DPO_noSFT
Updated
clembench-playpen/Mistral-Small-24B-Instruct-2501_KTO_Final_KTO_noSFT
Updated
clembench-playpen/Mistral-Small-24B-Instruct-rehearsal_playpen_SFT-e3_DABL02_0.82K-steps
Updated
clembench-playpen/Mistral-Small-24B-Instruct-0.1k-warmup_playpen_SFT-e3_DFINAL_0.6K-steps
Updated
clembench-playpen/Mistral-Small-24B-Instruct-0.1k-warmup_playpen_SFT-e3_DFINAL_0.1K-steps
Updated
datasets
37
clembench-playpen/warm-up_synthetic-data
Viewer
•
Updated
•
21.6k
•
49
clembench-playpen/DPO_2neg_Aborted_best_models_old_exp
Viewer
•
Updated
•
3.39k
•
40
clembench-playpen/DPO_allneg_Aborted_best_models_old_LA
Viewer
•
Updated
•
4.48k
•
77
clembench-playpen/DPO_6neg_Aborted_best_models_old_LA
Viewer
•
Updated
•
4.45k
•
61
clembench-playpen/DPO_5neg_Aborted_best_models_old_LA
Viewer
•
Updated
•
4.38k
•
64
clembench-playpen/DPO_4neg_Aborted_best_models_old_LA
Viewer
•
Updated
•
4.25k
•
59
clembench-playpen/DPO_3neg_Aborted_best_models_old_LA
Viewer
•
Updated
•
3.95k
•
63
clembench-playpen/DPO_2neg_Aborted_old_LA
Viewer
•
Updated
•
12.4k
•
47
clembench-playpen/DPO_2neg_Aborted_same_family_model_old_LA
Viewer
•
Updated
•
6.35k
•
42
clembench-playpen/DPO_2neg_Aborted_best_models_old_LA
Viewer
•
Updated
•
3.39k
•
54