carlesoctav/deepmath-qwen2.5
385 MB
👋 Open to Work
Carles Octavianus
carlesoctav
AI & ML interests
Probabilistic machine learning, Information Retrieval.
Recent Activity
updated a bucket about 3 hours ago
carlesoctav/deepmath-qwen2.5 published a bucket about 10 hours ago
carlesoctav/deepmath-qwen2.5 published a bucket about 11 hours ago
carlesoctav/dapo-01Organizations
D and M generated with Gemma 27B for Dolci Instruct SFT training (no tools). Includes instruct-only, reasoning-only, wildchat, and combined datasets.
-
carlesoctav/filtered-2048-4b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 1.39M • 21 -
carlesoctav/4b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 1.45M • 24 -
carlesoctav/4b-carlesoctav-Dolci-Instruct-SFT-No-Tools-wildchat-only
Viewer • Updated • 278k • 12 -
carlesoctav/4b-generated-Dolci-Instruct-SFT-No-Tools-instruct-only
Viewer • Updated • 961k • 109
D an M generated with Gemma 27B for Dolci Instruct SFT training (no tools). Includes instruct-only, reasoning-only, wildchat, and combined datasets.
-
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 720k • 53 -
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools-instruct-only
Viewer • Updated • 524k • 31 -
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools-reasoning-only
Viewer • Updated • 81.4k • 17 -
carlesoctav/27b-carlesoctav-Dolci-Instruct-SFT-No-Tools-wildchat-only
Viewer • Updated • 115k • 7
D and M generated with Gemma 27B for Dolci Instruct SFT training (no tools). Includes instruct-only, reasoning-only, wildchat, and combined datasets.
-
carlesoctav/filtered-2048-4b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 1.39M • 21 -
carlesoctav/4b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 1.45M • 24 -
carlesoctav/4b-carlesoctav-Dolci-Instruct-SFT-No-Tools-wildchat-only
Viewer • Updated • 278k • 12 -
carlesoctav/4b-generated-Dolci-Instruct-SFT-No-Tools-instruct-only
Viewer • Updated • 961k • 109
D an M generated with Gemma 27B for Dolci Instruct SFT training (no tools). Includes instruct-only, reasoning-only, wildchat, and combined datasets.
-
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 720k • 53 -
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools-instruct-only
Viewer • Updated • 524k • 31 -
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools-reasoning-only
Viewer • Updated • 81.4k • 17 -
carlesoctav/27b-carlesoctav-Dolci-Instruct-SFT-No-Tools-wildchat-only
Viewer • Updated • 115k • 7
models 36
carlesoctav/test
Text Generation • 0.5B • Updated
carlesoctav/27b-test
Text Generation • 1.0B • Updated • 8
carlesoctav/random-test
Text Generation • 1.0B • Updated • 6
carlesoctav/test-final
Text Generation • 1.0B • Updated • 6
carlesoctav/desperate-run
Text Generation • 1.0B • Updated • 15
carlesoctav/100
Text Generation • 1.0B • Updated • 15
carlesoctav/mm-01
Text Generation • 1.0B • Updated • 9
carlesoctav/dpo-debug
Text Generation • 1.0B • Updated • 5
carlesoctav/4b-Dolci-Instruct-SFT-finale-notebook
Text Generation • 1.0B • Updated • 9
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools-rank-256-lr-5e7
Text Generation • 1.0B • Updated • 17
datasets 29
carlesoctav/4b-generated-Dolci-Instruct-SFT-No-Tools-messages
Viewer • Updated • 1.45M • 89
carlesoctav/IT-WC-GSM8k-combine
Viewer • Updated • 1.79M • 47
carlesoctav/IT-WDIT-CWIT-WC
Viewer • Updated • 1.79M • 24
carlesoctav/IT-WC-2048-27b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 601k • 14
carlesoctav/IT-WC-2048-4b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 1.19M • 28
carlesoctav/4b-generated-DPO
Viewer • Updated • 650k • 42
carlesoctav/gpt-eval
Viewer • Updated • 300 • 17
carlesoctav/dpo-sample
Preview • Updated • 3
carlesoctav/filtered-2048-27b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 679k • 44
carlesoctav/27b-generated-Dolci-Instruct-SFT-No-Tools
Viewer • Updated • 720k • 53