tpls/repro-reward-models-its-artifacts
499 kB
tpls
tpls
AI & ML interests
None yet
Recent Activity
updated a Space 5 days ago
tpls/repro-bc-action-quantization updated a Space 5 days ago
tpls/repro-riemannian-dueling-opt updated a bucket 5 days ago
tpls/repro-adam-muon-implicit-bias-artifactsOrganizations
Current gemma-4 12B coder — SFT v5 (tool-calling winner) and SFT v5 + abliterated (tools + uncensored, prod). Full recipe + gate on each card.
-
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5
Text Generation • 12B • Updated • 30 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-GGUF
Text Generation • 12B • Updated • 210 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-abliterated
Text Generation • 12B • Updated • 108 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-abliterated-GGUF
Text Generation • 12B • Updated • 10.9k
Superseded variants, kept for provenance and inbound links. Don't use for new work — each card points to its replacement.
-
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v4
Text Generation • 12B • Updated • 6 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v4-GGUF
Text Generation • 12B • Updated • 78 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v3
Text Generation • 12B • Updated • 387 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft
Text Generation • 12B • Updated • 16
Current gemma-4 12B coder — SFT v5 (tool-calling winner) and SFT v5 + abliterated (tools + uncensored, prod). Full recipe + gate on each card.
-
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5
Text Generation • 12B • Updated • 30 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-GGUF
Text Generation • 12B • Updated • 210 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-abliterated
Text Generation • 12B • Updated • 108 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-abliterated-GGUF
Text Generation • 12B • Updated • 10.9k
Superseded variants, kept for provenance and inbound links. Don't use for new work — each card points to its replacement.
-
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v4
Text Generation • 12B • Updated • 6 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v4-GGUF
Text Generation • 12B • Updated • 78 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v3
Text Generation • 12B • Updated • 387 -
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft
Text Generation • 12B • Updated • 16
spaces 30
Running
Repro - On the Power of (Approximate) Reward Models for Inference-Time Scaling: Sequential Monte Carlo and Beyond
🎯
Explore and sync experiment logs with an AI coding agent
Running
Repro - Any-dimensional invariant universality
🎯
Collaborate on a shared logbook with your coding agent
Running
Repro - Riemannian Dueling Optimization
🎯
Explore experiment logs and collaborate with an AI agent
Running
Repro - Understanding Behavior Cloning with Action Quantization
🎯
Collaborate on a coding logbook with your AI agent
Running
Repro - adam-muon-implicit-bias
🎯
View and collaborate on project logbooks online
Running
Repro - performative-prediction-complexity
🎯
Explore and sync logs with your coding agent
models 17
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-GGUF
Text Generation • 12B • Updated • 210
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5
Text Generation • 12B • Updated • 30
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-abliterated-GGUF
Text Generation • 12B • Updated • 10.9k
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v5-abliterated
Text Generation • 12B • Updated • 108
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-abliterated-GGUF
Text Generation • 12B • Updated • 16.5k
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-abliterated
Text Generation • 12B • Updated • 35.1k • 1
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-steer
Text Generation • 12B • Updated • 136
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-GGUF
Text Generation • 12B • Updated • 82
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft
Text Generation • 12B • Updated • 16
tpls/gemma-4-12B-coder-fable5-composer2.5-v1-sft-v3
Text Generation • 12B • Updated • 387