Skip to page content
SAVRN
Search Plan a project

Dataset

gated-lora-experiments

by Helain ZIMMERMANN Helain/gated-lora-experiments

This dataset stores LoRA expert pools and gating network weights from multi-task fine-tuning experiments on various base models (Phi-2, Gemma-2, Llama-3.2, Pythia-410M, Qwen2.5-0.5B, SmolLM-360M).

Rows—
Configurations—
Size35.6 GB
License—
AccessPublicly accessible
Monthly Downloads581

Dataset Card

This dataset stores LoRA expert pools and gating network weights from multi-task fine-tuning experiments on various base models (Phi-2, Gemma-2, Llama-3.2, Pythia-410M, Qwen2.5-0.5B, SmolLM-360M). - legacy/{model}/{runname}/ — runs from the original experimental campaign (2024-12 → 2025-01), with checkpoints at steps 500, 1000,..., finalmodel and bestmodel. - New baseline + Gated runs (2026-05+) are uploaded at the top level following the paper-001 pipeline convention. WIP — Master's thesis research artifacts.

Excerpt from the card by Helain ZIMMERMANN.

Details

Repository
Helain/gated-lora-experiments
Publisher
Helain ZIMMERMANN
Task category
Not stated by the source
Tags
Not stated by the source
Size category
Not stated by the source
Languages
Not stated by the source
Revision
34a565da49cd6b17a0bc369f69eb1510ecf5f714
Last updated
2026-10-12

Files

2,688 files, 35.6 GB in total.

Data551 files · 403.8 KB
Documentation1 file · 659 B
Other2134 files · 35.5 GB
Repository2 files · 30.9 MB
Every file (first 2,688)
FileTypeSizeSHA-256
legacy/gemma2/exp2_gated_2experts_seed1/best_model/training_state.jsonData96 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp2_gated_2experts_seed1/final_model/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed1/final_results.jsonData260 B—
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_analysis.jsonData4.8 KB—
legacy/gemma2/exp2_gated_2experts_seed2/best_model/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2500/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp2_gated_2experts_seed2/final_model/training_state.jsonData98 B—
legacy/gemma2/exp2_gated_2experts_seed2/final_results.jsonData260 B—
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_analysis.jsonData4.8 KB—
legacy/gemma2/exp2_gated_2experts_seed3/best_model/training_state.jsonData96 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-500/training_state.jsonData95 B—
legacy/gemma2/exp2_gated_2experts_seed3/final_model/training_state.jsonData97 B—
legacy/gemma2/exp2_gated_2experts_seed3/final_results.jsonData260 B—
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_analysis.jsonData4.8 KB—
legacy/gemma2/exp3_gated_3experts_seed1/best_model/training_state.jsonData96 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp3_gated_3experts_seed1/final_model/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed1/final_results.jsonData260 B—
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_analysis.jsonData6.3 KB—
legacy/gemma2/exp3_gated_3experts_seed2/best_model/training_state.jsonData96 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp3_gated_3experts_seed2/final_model/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed2/final_results.jsonData259 B—
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_analysis.jsonData6.3 KB—
legacy/gemma2/exp3_gated_3experts_seed3/best_model/training_state.jsonData96 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp3_gated_3experts_seed3/final_model/training_state.jsonData97 B—
legacy/gemma2/exp3_gated_3experts_seed3/final_results.jsonData261 B—
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_analysis.jsonData6.3 KB—
legacy/gemma2/exp4_gated_loadbalancing_seed1/best_model/training_state.jsonData96 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/final_model/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/final_results.jsonData260 B—
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_analysis.jsonData6.3 KB—
legacy/gemma2/exp4_gated_loadbalancing_seed2/best_model/training_state.jsonData96 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/final_model/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/final_results.jsonData259 B—
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_analysis.jsonData6.3 KB—
legacy/gemma2/exp4_gated_loadbalancing_seed3/best_model/training_state.jsonData96 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/final_model/training_state.jsonData97 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/final_results.jsonData259 B—
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_analysis.jsonData6.3 KB—
legacy/gemma2/exp5_gated_topk_seed1/best_model/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed1/final_model/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed1/final_results.jsonData259 B—
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_analysis.jsonData6.3 KB—
legacy/gemma2/exp5_gated_topk_seed2/best_model/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1500/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2500/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3500/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-4000/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-500/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed2/final_model/training_state.jsonData97 B—
legacy/gemma2/exp5_gated_topk_seed2/final_results.jsonData259 B—
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_analysis.jsonData6.3 KB—
legacy/gemma2/exp5_gated_topk_seed3/best_model/training_state.jsonData95 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1000/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1500/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2000/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2500/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3000/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3500/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-4000/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-500/training_state.jsonData95 B—
legacy/gemma2/exp5_gated_topk_seed3/final_model/training_state.jsonData96 B—
legacy/gemma2/exp5_gated_topk_seed3/final_results.jsonData259 B—
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_analysis.jsonData6.3 KB—
legacy/llama32/exp4_gated_loadbalancing_seed1/best_model/training_state.jsonData97 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-4000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-500/training_state.jsonData96 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/final_model/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/final_results.jsonData260 B—
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_analysis.jsonData6.7 KB—
legacy/llama32/exp4_gated_loadbalancing_seed2/best_model/training_state.jsonData97 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/final_model/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/final_results.jsonData263 B—
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_analysis.jsonData6.7 KB—
legacy/llama32/exp4_gated_loadbalancing_seed3/best_model/training_state.jsonData97 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3500/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-4000/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-500/training_state.jsonData96 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/final_model/training_state.jsonData98 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/final_results.jsonData262 B—
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_analysis.jsonData6.7 KB—
legacy/llama32/exp5_gated_topk_seed1/best_model/training_state.jsonData97 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-4000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-500/training_state.jsonData97 B—
legacy/llama32/exp5_gated_topk_seed1/final_model/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed1/final_results.jsonData263 B—
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_analysis.jsonData6.7 KB—
legacy/llama32/exp5_gated_topk_seed2/best_model/training_state.jsonData97 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/llama32/exp5_gated_topk_seed2/final_model/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed2/final_results.jsonData263 B—
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_analysis.jsonData6.7 KB—
legacy/llama32/exp5_gated_topk_seed3/best_model/training_state.jsonData97 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3500/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-4000/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/checkpoint-500/training_state.jsonData97 B—
legacy/llama32/exp5_gated_topk_seed3/final_model/training_state.jsonData98 B—
legacy/llama32/exp5_gated_topk_seed3/final_results.jsonData262 B—
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_analysis.jsonData6.7 KB—
legacy/llama32_modified/ablation_partial_gating_seed1/best_model/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1000/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1500/training_state.jsonData197 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2000/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3000/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3500/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-4000/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-500/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-1/training_state.jsonData193 B—
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-2/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-4/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed1/final_model/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed1/final_results.jsonData289 B—
legacy/llama32_modified/ablation_partial_gating_seed1/latest/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_analysis.jsonData5.6 KB—
legacy/llama32_modified/ablation_partial_gating_seed2/best_model/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1000/training_state.jsonData194 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1500/training_state.jsonData194 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2000/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3000/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3500/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-4000/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-500/training_state.jsonData197 B—
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-1/training_state.jsonData191 B—
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-2/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-4/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed2/final_model/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed2/final_results.jsonData287 B—
legacy/llama32_modified/ablation_partial_gating_seed2/latest/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_analysis.jsonData5.6 KB—
legacy/llama32_modified/ablation_partial_gating_seed3/best_model/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1000/training_state.jsonData194 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1500/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2000/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3000/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3500/training_state.jsonData199 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-4000/training_state.jsonData198 B—
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-500/training_state.jsonData197 B—
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-1/training_state.jsonData193 B—
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-2/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-4/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed3/final_model/training_state.jsonData196 B—
legacy/llama32_modified/ablation_partial_gating_seed3/final_results.jsonData289 B—
legacy/llama32_modified/ablation_partial_gating_seed3/latest/training_state.jsonData195 B—
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_analysis.jsonData5.7 KB—
legacy/llama32_modified/exp2_gated_2experts_seed1/best_model/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1000/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1500/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2000/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3000/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3500/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-4000/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-500/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-1/training_state.jsonData193 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-2/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-4/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/final_model/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/final_results.jsonData287 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/latest/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_analysis.jsonData5.1 KB—
legacy/llama32_modified/exp2_gated_2experts_seed2/best_model/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1000/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1500/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2000/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2500/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3000/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3500/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-4000/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-500/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-1/training_state.jsonData194 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-2/training_state.jsonData193 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-3/training_state.jsonData195 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-4/training_state.jsonData195 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/final_model/training_state.jsonData195 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/final_results.jsonData288 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/latest/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_analysis.jsonData5.1 KB—
legacy/llama32_modified/exp2_gated_2experts_seed3/best_model/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1000/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1500/training_state.jsonData197 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2000/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3000/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3500/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-4000/training_state.jsonData199 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-500/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-1/training_state.jsonData194 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-2/training_state.jsonData195 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-4/training_state.jsonData195 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/final_model/training_state.jsonData195 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/final_results.jsonData289 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/latest/training_state.jsonData198 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_analysis.jsonData5.2 KB—
legacy/llama32_modified/exp3_gated_3experts_seed1/best_model/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1000/training_state.jsonData197 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1500/training_state.jsonData195 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2000/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2500/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3000/training_state.jsonData197 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3500/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-4000/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-500/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-1/training_state.jsonData194 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-2/training_state.jsonData195 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-3/training_state.jsonData195 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-4/training_state.jsonData193 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/final_model/training_state.jsonData193 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/final_results.jsonData284 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/latest/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_analysis.jsonData6.7 KB—
legacy/llama32_modified/exp3_gated_3experts_seed2/best_model/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1000/training_state.jsonData197 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1500/training_state.jsonData197 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2000/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3000/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3500/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-4000/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-500/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-1/training_state.jsonData195 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-2/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-4/training_state.jsonData195 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/final_model/training_state.jsonData195 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/final_results.jsonData288 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/latest/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_analysis.jsonData6.7 KB—
legacy/llama32_modified/exp3_gated_3experts_seed3/best_model/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1000/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1500/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2000/training_state.jsonData198 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2500/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3000/training_state.jsonData197 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3500/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-4000/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-500/training_state.jsonData197 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-1/training_state.jsonData193 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-2/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-3/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-4/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/final_model/training_state.jsonData196 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/final_results.jsonData288 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/latest/training_state.jsonData199 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_analysis.jsonData6.7 KB—
legacy/phi2_ablations/ablation_no_l1/best_model/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_l1/checkpoint-1000/training_state.jsonData96 B—
legacy/phi2_ablations/ablation_no_l1/checkpoint-1500/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_l1/checkpoint-2000/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_l1/checkpoint-2500/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_l1/checkpoint-500/training_state.jsonData97 B—
legacy/phi2_ablations/ablation_no_l1/final_model/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_l1/final_results.jsonData263 B—
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_analysis.jsonData18.7 KB—
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_routing_analysis.jsonData18.9 KB—
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_analysis.jsonData7.6 KB—
legacy/phi2_ablations/ablation_no_layer_emb/best_model/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1000/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1500/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2000/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2500/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-500/training_state.jsonData97 B—
legacy/phi2_ablations/ablation_no_layer_emb/final_model/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_no_layer_emb/final_results.jsonData263 B—
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_analysis.jsonData18.7 KB—
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_routing_analysis.jsonData18.9 KB—
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_analysis.jsonData7.6 KB—
legacy/phi2_ablations/ablation_partial_gating/best_model/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1000/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1500/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2000/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2500/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_partial_gating/checkpoint-500/training_state.jsonData97 B—
legacy/phi2_ablations/ablation_partial_gating/final_model/training_state.jsonData98 B—
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_analysis.jsonData18.7 KB—
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_routing_analysis.jsonData19.0 KB—
legacy/pythia410m/ablation_no_l1_seed1/best_model/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1000/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1500/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_l1_seed2/best_model/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_l1_seed3/best_model/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_layer_emb_seed1/best_model/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1500/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-500/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/best_model/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/best_model/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2000/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2500/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3000/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3500/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-4000/training_state.jsonData96 B—
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_partial_gating_seed1/best_model/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_partial_gating_seed2/best_model/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/ablation_partial_gating_seed3/best_model/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed1/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2000/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed1/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed1/final_results.jsonData263 B—
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_analysis.jsonData4.5 KB—
legacy/pythia410m/exp2_gated_2experts_seed2/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed2/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed2/final_results.jsonData263 B—
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_analysis.jsonData4.5 KB—
legacy/pythia410m/exp2_gated_2experts_seed3/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp2_gated_2experts_seed3/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp2_gated_2experts_seed3/final_results.jsonData263 B—
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_analysis.jsonData4.5 KB—
legacy/pythia410m/exp3_gated_3experts_seed1/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp3_gated_3experts_seed1/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed1/final_results.jsonData262 B—
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_analysis.jsonData5.8 KB—
legacy/pythia410m/exp3_gated_3experts_seed2/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp3_gated_3experts_seed2/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed2/final_results.jsonData261 B—
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_analysis.jsonData5.8 KB—
legacy/pythia410m/exp3_gated_3experts_seed3/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp3_gated_3experts_seed3/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp3_gated_3experts_seed3/final_results.jsonData262 B—
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_analysis.jsonData5.8 KB—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/final_results.jsonData262 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_analysis.jsonData5.9 KB—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1500/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2500/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3500/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-4000/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-500/training_state.jsonData97 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/final_model/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/final_results.jsonData262 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_analysis.jsonData5.8 KB—
legacy/pythia410m/exp4_gated_loadbalancing_seed3/best_model/training_state.jsonData98 B—
legacy/pythia410m/exp4_gated_loadbalancing_seed3/checkpoint-1000/training_state.jsonData98 B—
README.mdDocumentation659 B—
legacy/gemma2/exp2_gated_2experts_seed1/best_model/expert_pools.ptOther19.4 MBc8225b311e05
legacy/gemma2/exp2_gated_2experts_seed1/best_model/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/best_model/gating_network.ptOther33.9 MB9dbf7da061dd
legacy/gemma2/exp2_gated_2experts_seed1/best_model/scheduler.ptOther1.7 KB3e43ae9e84c6
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1000/expert_pools.ptOther19.4 MB379603235f81
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1000/gating_network.ptOther33.9 MBe72903bd891a
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1500/expert_pools.ptOther19.4 MB296479a9750b
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1500/gating_network.ptOther33.9 MB5ec9930a382c
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2000/expert_pools.ptOther19.4 MB8dabf2556b26
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2000/gating_network.ptOther33.9 MBc2368179472e
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2500/expert_pools.ptOther19.4 MB66d32ebe3d70
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2500/gating_network.ptOther33.9 MB78a3cc920337
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3000/expert_pools.ptOther19.4 MBa513320f58af
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3000/gating_network.ptOther33.9 MB58e45d7f13f5
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3500/expert_pools.ptOther19.4 MBc780d786b377
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3500/gating_network.ptOther33.9 MB578d88af8a54
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-4000/expert_pools.ptOther19.4 MB1a7b968e4cb5
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-4000/gating_network.ptOther33.9 MB6f9fd4968394
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-500/expert_pools.ptOther19.4 MB31a02d0aeb43
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-500/gating_network.ptOther33.9 MB70850321c37d
legacy/gemma2/exp2_gated_2experts_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp2_gated_2experts_seed1/final_model/expert_pools.ptOther19.4 MBa4774d8f2b01
legacy/gemma2/exp2_gated_2experts_seed1/final_model/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed1/final_model/gating_network.ptOther33.9 MB3108d13f7e4f
legacy/gemma2/exp2_gated_2experts_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_entropy_per_layer.pngOther43.0 KB43455caff5e7
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_layer_expert_heatmap.pngOther105.4 KB151e016edabb
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_task_expert_usage.pngOther37.5 KB9bc1568d5b47
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_1.pngOther50.9 KB9eff21d16177
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_2.pngOther50.7 KB83599d65d3d0
legacy/gemma2/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_3.pngOther51.2 KBbb185f41c6c3
legacy/gemma2/exp2_gated_2experts_seed2/best_model/expert_pools.ptOther19.4 MB8a594254b016
legacy/gemma2/exp2_gated_2experts_seed2/best_model/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/best_model/gating_network.ptOther33.9 MBbb20c96d88a8
legacy/gemma2/exp2_gated_2experts_seed2/best_model/scheduler.ptOther1.7 KB256b90657b80
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1000/expert_pools.ptOther19.4 MB350ed122efab
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1000/gating_network.ptOther33.9 MB9aa142c30d0d
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1500/expert_pools.ptOther19.4 MB190073fbf59c
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1500/gating_network.ptOther33.9 MBa4d6c0b882bd
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2000/expert_pools.ptOther19.4 MB69557b40b81c
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2000/gating_network.ptOther33.9 MB6632edd6a2a6
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2500/expert_pools.ptOther19.4 MBcce5f7cd5082
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2500/gating_network.ptOther33.9 MB4a57b65890da
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3000/expert_pools.ptOther19.4 MB76484fba7a2a
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3000/gating_network.ptOther33.9 MB43b7c4a513bf
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3500/expert_pools.ptOther19.4 MBe2f54aa65058
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3500/gating_network.ptOther33.9 MB548add376ee5
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-4000/expert_pools.ptOther19.4 MBaec55729c96f
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-4000/gating_network.ptOther33.9 MB82df6a08e125
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-500/expert_pools.ptOther19.4 MBdaf68467947f
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-500/gating_network.ptOther33.9 MB40b105217965
legacy/gemma2/exp2_gated_2experts_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp2_gated_2experts_seed2/final_model/expert_pools.ptOther19.4 MBcc7ef50950de
legacy/gemma2/exp2_gated_2experts_seed2/final_model/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed2/final_model/gating_network.ptOther33.9 MB15b9a53d6da0
legacy/gemma2/exp2_gated_2experts_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_entropy_per_layer.pngOther43.0 KBdc1777729c31
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_layer_expert_heatmap.pngOther105.8 KB10673b1d3338
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_task_expert_usage.pngOther37.1 KB2ee310ec5e03
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_1.pngOther44.1 KB7f2cd0f95189
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_2.pngOther44.2 KBbea32b5b76d8
legacy/gemma2/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_3.pngOther44.4 KBd9d2d81d167d
legacy/gemma2/exp2_gated_2experts_seed3/best_model/expert_pools.ptOther19.4 MB17982a49897c
legacy/gemma2/exp2_gated_2experts_seed3/best_model/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/best_model/gating_network.ptOther33.9 MB718defbaf1eb
legacy/gemma2/exp2_gated_2experts_seed3/best_model/scheduler.ptOther1.7 KB256b90657b80
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1000/expert_pools.ptOther19.4 MBfc93dbcffa81
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1000/gating_network.ptOther33.9 MBec849e725c6e
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1500/expert_pools.ptOther19.4 MB696ec53e8c28
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1500/gating_network.ptOther33.9 MBc5a76f0bddeb
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2000/expert_pools.ptOther19.4 MBb94317cd68e8
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2000/gating_network.ptOther33.9 MBb35028425e84
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2500/expert_pools.ptOther19.4 MB5f5691b71813
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2500/gating_network.ptOther33.9 MB179386c524e0
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3000/expert_pools.ptOther19.4 MBf93a9bf3ee33
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3000/gating_network.ptOther33.9 MBbb94bab6e83c
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3500/expert_pools.ptOther19.4 MB2f15f812efd7
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3500/gating_network.ptOther33.9 MB09e03213521c
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-4000/expert_pools.ptOther19.4 MB9c7a0957dd93
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-4000/gating_network.ptOther33.9 MB952aa8a2efb1
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-500/expert_pools.ptOther19.4 MB76e251a856b2
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-500/gating_network.ptOther33.9 MB68e0fbf0f1d6
legacy/gemma2/exp2_gated_2experts_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp2_gated_2experts_seed3/final_model/expert_pools.ptOther19.4 MB37b25deee0d1
legacy/gemma2/exp2_gated_2experts_seed3/final_model/gated_lora_config.ptOther1.8 KBd74c5d7b3d63
legacy/gemma2/exp2_gated_2experts_seed3/final_model/gating_network.ptOther33.9 MBfe7bce8555cf
legacy/gemma2/exp2_gated_2experts_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_entropy_per_layer.pngOther43.7 KB83d887b1bf12
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_layer_expert_heatmap.pngOther107.5 KB9f71112c1dd6
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_task_expert_usage.pngOther37.8 KB8587e0dd600b
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_1.pngOther46.0 KB3e2bd46d1783
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_2.pngOther45.8 KB5ce8e01af9b6
legacy/gemma2/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_3.pngOther46.1 KBeee819613b1d
legacy/gemma2/exp3_gated_3experts_seed1/best_model/expert_pools.ptOther45.0 MB83ec11fe6da4
legacy/gemma2/exp3_gated_3experts_seed1/best_model/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/best_model/gating_network.ptOther33.9 MBa3e5a7e423db
legacy/gemma2/exp3_gated_3experts_seed1/best_model/scheduler.ptOther1.7 KB039753c3afc8
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1000/expert_pools.ptOther45.0 MBbe2881321abe
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1000/gating_network.ptOther33.9 MB0d0f6ad0fc40
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1500/expert_pools.ptOther45.0 MB1f76f6a78b1f
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1500/gating_network.ptOther33.9 MB45dc63a3e051
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2000/expert_pools.ptOther45.0 MBe9f68d506330
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2000/gating_network.ptOther33.9 MB00c20bbd5fe2
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2500/expert_pools.ptOther45.0 MB3dfd54af0180
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2500/gating_network.ptOther33.9 MB967c304b9290
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3000/expert_pools.ptOther45.0 MBfb68ff27f2d5
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3000/gating_network.ptOther33.9 MBb079d3e05904
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3500/expert_pools.ptOther45.0 MB3988b2c21cdd
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3500/gating_network.ptOther33.9 MB4cf8b87785fa
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-4000/expert_pools.ptOther45.0 MBb45c1070268c
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-4000/gating_network.ptOther33.9 MBbc0a41d7f0f1
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-500/expert_pools.ptOther45.0 MB46becb6e4814
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-500/gating_network.ptOther33.9 MBed05a438ac13
legacy/gemma2/exp3_gated_3experts_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp3_gated_3experts_seed1/final_model/expert_pools.ptOther45.0 MBfd33930e1d37
legacy/gemma2/exp3_gated_3experts_seed1/final_model/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed1/final_model/gating_network.ptOther33.9 MB415ac81a66b5
legacy/gemma2/exp3_gated_3experts_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_entropy_per_layer.pngOther42.3 KB613803b67a87
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_layer_expert_heatmap.pngOther103.5 KB05121e9663b7
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_task_expert_usage.pngOther39.1 KB6c2ac9350c8f
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_1.pngOther44.9 KB1778e94ba9cd
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_2.pngOther45.1 KB3cf57304bde0
legacy/gemma2/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_3.pngOther45.2 KBbe73e2570616
legacy/gemma2/exp3_gated_3experts_seed2/best_model/expert_pools.ptOther45.0 MB0c0e174b37e7
legacy/gemma2/exp3_gated_3experts_seed2/best_model/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/best_model/gating_network.ptOther33.9 MB95582465905a
legacy/gemma2/exp3_gated_3experts_seed2/best_model/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1000/expert_pools.ptOther45.0 MB4706179d6935
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1000/gating_network.ptOther33.9 MB85069da6714b
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1500/expert_pools.ptOther45.0 MB61860f315931
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1500/gating_network.ptOther33.9 MB08ed058dc295
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2000/expert_pools.ptOther45.0 MBda1c56794594
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2000/gating_network.ptOther33.9 MB76dc7832d840
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2500/expert_pools.ptOther45.0 MBdb08f28f20f5
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2500/gating_network.ptOther33.9 MB362fe626e5af
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3000/expert_pools.ptOther45.0 MBbeed35565ef7
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3000/gating_network.ptOther33.9 MBc8c2cd971753
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3500/expert_pools.ptOther45.0 MB12ace0adab19
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3500/gating_network.ptOther33.9 MBace1313803cd
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-4000/expert_pools.ptOther45.0 MBee3ffffc9e90
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-4000/gating_network.ptOther33.9 MB3e3580def510
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-500/expert_pools.ptOther45.0 MB0c0e174b37e7
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-500/gating_network.ptOther33.9 MB95582465905a
legacy/gemma2/exp3_gated_3experts_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp3_gated_3experts_seed2/final_model/expert_pools.ptOther45.0 MB74c34bf5b700
legacy/gemma2/exp3_gated_3experts_seed2/final_model/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed2/final_model/gating_network.ptOther33.9 MB9c99a37662fc
legacy/gemma2/exp3_gated_3experts_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_entropy_per_layer.pngOther43.0 KB63b014f486ed
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_layer_expert_heatmap.pngOther103.9 KB57678a903508
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_task_expert_usage.pngOther40.0 KBab85cba35854
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_1.pngOther55.8 KB0d711dd0978f
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_2.pngOther55.9 KBb3a99e06d16c
legacy/gemma2/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_3.pngOther55.7 KBf5ed5a731c33
legacy/gemma2/exp3_gated_3experts_seed3/best_model/expert_pools.ptOther45.0 MB756d2766b157
legacy/gemma2/exp3_gated_3experts_seed3/best_model/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/best_model/gating_network.ptOther33.9 MB680dae923d78
legacy/gemma2/exp3_gated_3experts_seed3/best_model/scheduler.ptOther1.7 KB3e43ae9e84c6
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1000/expert_pools.ptOther45.0 MB8bf9a18cde77
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1000/gating_network.ptOther33.9 MBf6aa34133d88
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1500/expert_pools.ptOther45.0 MBf2e62910719a
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1500/gating_network.ptOther33.9 MBe8827520341a
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2000/expert_pools.ptOther45.0 MBa0a931f387ad
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2000/gating_network.ptOther33.9 MBc23d2a82da0e
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2500/expert_pools.ptOther45.0 MB31a2627f5a79
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2500/gating_network.ptOther33.9 MB89a469455ffe
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3000/expert_pools.ptOther45.0 MBc39730a93828
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3000/gating_network.ptOther33.9 MBf4d72a0118f0
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3500/expert_pools.ptOther45.0 MB3a5e20844a18
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3500/gating_network.ptOther33.9 MB7abd6d3be134
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-4000/expert_pools.ptOther45.0 MB2a6d0278d9f2
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-4000/gating_network.ptOther33.9 MB3cebb27bee1f
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-500/expert_pools.ptOther45.0 MBc3b35878ed64
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-500/gating_network.ptOther33.9 MB697972b962ef
legacy/gemma2/exp3_gated_3experts_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp3_gated_3experts_seed3/final_model/expert_pools.ptOther45.0 MB7b7b61329f74
legacy/gemma2/exp3_gated_3experts_seed3/final_model/gated_lora_config.ptOther1.8 KB3a0ff4effb43
legacy/gemma2/exp3_gated_3experts_seed3/final_model/gating_network.ptOther33.9 MB7fba6f7f268c
legacy/gemma2/exp3_gated_3experts_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_entropy_per_layer.pngOther43.0 KBf0cbf4fafb8f
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_layer_expert_heatmap.pngOther103.7 KB6de30d7d0ae9
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_task_expert_usage.pngOther38.4 KB3fe801bd0bd7
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_1.pngOther48.7 KB5767724aba31
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_2.pngOther49.0 KBfeff1a08848c
legacy/gemma2/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_3.pngOther49.1 KB43e34bb5038f
legacy/gemma2/exp4_gated_loadbalancing_seed1/best_model/expert_pools.ptOther45.0 MBee9660798a08
legacy/gemma2/exp4_gated_loadbalancing_seed1/best_model/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/best_model/gating_network.ptOther33.9 MB1dcaebd4313b
legacy/gemma2/exp4_gated_loadbalancing_seed1/best_model/scheduler.ptOther1.7 KB039753c3afc8
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1000/expert_pools.ptOther45.0 MBe536b81523a5
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1000/gating_network.ptOther33.9 MBe11b4f8eab0f
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1500/expert_pools.ptOther45.0 MBaf4796080908
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1500/gating_network.ptOther33.9 MB035417ef4bce
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2000/expert_pools.ptOther45.0 MBa1cbc131dcf7
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2000/gating_network.ptOther33.9 MB4288993569ae
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2500/expert_pools.ptOther45.0 MBfa065deceaee
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2500/gating_network.ptOther33.9 MB8506ada37088
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3000/expert_pools.ptOther45.0 MBbee8e02d2843
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3000/gating_network.ptOther33.9 MB83a7ce133071
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3500/expert_pools.ptOther45.0 MBf3f2836cd12d
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3500/gating_network.ptOther33.9 MB226296080d7d
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-4000/expert_pools.ptOther45.0 MBb7de8e130ee3
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-4000/gating_network.ptOther33.9 MBa02221ffb368
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-500/expert_pools.ptOther45.0 MBd24ea0a6f423
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-500/gating_network.ptOther33.9 MB64969f0e74ba
legacy/gemma2/exp4_gated_loadbalancing_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp4_gated_loadbalancing_seed1/final_model/expert_pools.ptOther45.0 MB996399d7e860
legacy/gemma2/exp4_gated_loadbalancing_seed1/final_model/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed1/final_model/gating_network.ptOther33.9 MB14f19cbc0eac
legacy/gemma2/exp4_gated_loadbalancing_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_entropy_per_layer.pngOther43.1 KB0b60733fb83b
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_layer_expert_heatmap.pngOther105.2 KBf9f25be06873
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_task_expert_usage.pngOther44.1 KB92006afc6203
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_1.pngOther53.8 KB5b50c1884d88
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_2.pngOther54.1 KB148defb8dec3
legacy/gemma2/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_3.pngOther54.2 KB4e5b2f847089
legacy/gemma2/exp4_gated_loadbalancing_seed2/best_model/expert_pools.ptOther45.0 MB739bc7f5d4b6
legacy/gemma2/exp4_gated_loadbalancing_seed2/best_model/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/best_model/gating_network.ptOther33.9 MB045e6332b6ab
legacy/gemma2/exp4_gated_loadbalancing_seed2/best_model/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1000/expert_pools.ptOther45.0 MB0bd88d6c352c
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1000/gating_network.ptOther33.9 MB76d9ef30fbc1
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1500/expert_pools.ptOther45.0 MB2bddffe0b313
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1500/gating_network.ptOther33.9 MB10236e050a81
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2000/expert_pools.ptOther45.0 MB0ff1b29c218a
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2000/gating_network.ptOther33.9 MBafb5daa9ee05
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2500/expert_pools.ptOther45.0 MBbae3166ef8d3
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2500/gating_network.ptOther33.9 MB4898e95c8808
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3000/expert_pools.ptOther45.0 MBd5d495224f85
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3000/gating_network.ptOther33.9 MBf4b09592273a
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3500/expert_pools.ptOther45.0 MB0c17eba2c7ee
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3500/gating_network.ptOther33.9 MB70f33685f3b7
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-4000/expert_pools.ptOther45.0 MB92df4b97a9cb
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-4000/gating_network.ptOther33.9 MB72db9c5c003b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-500/expert_pools.ptOther45.0 MB739bc7f5d4b6
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-500/gating_network.ptOther33.9 MB045e6332b6ab
legacy/gemma2/exp4_gated_loadbalancing_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp4_gated_loadbalancing_seed2/final_model/expert_pools.ptOther45.0 MBf827fae07a98
legacy/gemma2/exp4_gated_loadbalancing_seed2/final_model/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed2/final_model/gating_network.ptOther33.9 MB3997b4704c34
legacy/gemma2/exp4_gated_loadbalancing_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_entropy_per_layer.pngOther42.9 KBbc28b9745a88
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_layer_expert_heatmap.pngOther105.1 KBe044d4e4e80d
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_task_expert_usage.pngOther43.8 KB679a1b24ad37
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_1.pngOther58.6 KBdbe34af30cbc
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_2.pngOther58.9 KBe7f9622c589e
legacy/gemma2/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_3.pngOther59.0 KB68ed0dfa26ab
legacy/gemma2/exp4_gated_loadbalancing_seed3/best_model/expert_pools.ptOther45.0 MB7b85b8011561
legacy/gemma2/exp4_gated_loadbalancing_seed3/best_model/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/best_model/gating_network.ptOther33.9 MB4509e031775e
legacy/gemma2/exp4_gated_loadbalancing_seed3/best_model/scheduler.ptOther1.7 KB3e43ae9e84c6
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1000/expert_pools.ptOther45.0 MB48cc0bfe3d35
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1000/gating_network.ptOther33.9 MBba516dce2012
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1500/expert_pools.ptOther45.0 MBd92d78956699
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1500/gating_network.ptOther33.9 MB3c787ad7616a
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2000/expert_pools.ptOther45.0 MBfc8282dc6154
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2000/gating_network.ptOther33.9 MB975e02ed0af4
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2500/expert_pools.ptOther45.0 MB3936a26359c6
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2500/gating_network.ptOther33.9 MBa035b1a2815b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3000/expert_pools.ptOther45.0 MB4596434ca35b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3000/gating_network.ptOther33.9 MBed86982e1e97
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3500/expert_pools.ptOther45.0 MB5c34b06dc653
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3500/gating_network.ptOther33.9 MB81463f3efebc
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-4000/expert_pools.ptOther45.0 MBdb61604fd0bb
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-4000/gating_network.ptOther33.9 MB7927bdf165ca
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-500/expert_pools.ptOther45.0 MBf294dd02c381
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-500/gating_network.ptOther33.9 MBb289ba958a26
legacy/gemma2/exp4_gated_loadbalancing_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp4_gated_loadbalancing_seed3/final_model/expert_pools.ptOther45.0 MB67ccfc518396
legacy/gemma2/exp4_gated_loadbalancing_seed3/final_model/gated_lora_config.ptOther1.8 KB3bce6fb9623b
legacy/gemma2/exp4_gated_loadbalancing_seed3/final_model/gating_network.ptOther33.9 MBebab5e4cbe86
legacy/gemma2/exp4_gated_loadbalancing_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_entropy_per_layer.pngOther43.6 KB511727595fe4
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_layer_expert_heatmap.pngOther106.8 KB15057eb9b500
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_task_expert_usage.pngOther44.2 KB4c9581e94b3d
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_token_routing_1.pngOther51.1 KB81450be474d9
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_token_routing_2.pngOther51.4 KB9300be337e41
legacy/gemma2/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_token_routing_3.pngOther51.4 KB4854d11e4c32
legacy/gemma2/exp5_gated_topk_seed1/best_model/expert_pools.ptOther45.0 MB7c85c84fdd6f
legacy/gemma2/exp5_gated_topk_seed1/best_model/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/best_model/gating_network.ptOther33.9 MB4f87ef160234
legacy/gemma2/exp5_gated_topk_seed1/best_model/scheduler.ptOther1.7 KB039753c3afc8
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1000/expert_pools.ptOther45.0 MB386a0ccc188c
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1000/gating_network.ptOther33.9 MBaa5987cf19fc
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1500/expert_pools.ptOther45.0 MBe2066eb70457
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1500/gating_network.ptOther33.9 MBf61733a77cd7
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2000/expert_pools.ptOther45.0 MBede1597e2bc0
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2000/gating_network.ptOther33.9 MB5438146c1fcf
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2500/expert_pools.ptOther45.0 MBef7b9be5dd0d
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2500/gating_network.ptOther33.9 MBc86a95aa7098
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3000/expert_pools.ptOther45.0 MBbb5ade24a3d7
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3000/gating_network.ptOther33.9 MB046e194cd93b
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3500/expert_pools.ptOther45.0 MB73e42f401840
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3500/gating_network.ptOther33.9 MB7ae937cb412a
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-4000/expert_pools.ptOther45.0 MB0bde44ea3000
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-4000/gating_network.ptOther33.9 MBc5e5687ed61a
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-500/expert_pools.ptOther45.0 MB03ba99c2518c
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-500/gating_network.ptOther33.9 MB9a26dd7ffcf3
legacy/gemma2/exp5_gated_topk_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp5_gated_topk_seed1/final_model/expert_pools.ptOther45.0 MBfd3f9cb897cb
legacy/gemma2/exp5_gated_topk_seed1/final_model/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed1/final_model/gating_network.ptOther33.9 MBe857c81b00b8
legacy/gemma2/exp5_gated_topk_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_entropy_per_layer.pngOther41.4 KB76623680322d
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_layer_expert_heatmap.pngOther103.2 KB2156ef48fdb1
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_task_expert_usage.pngOther41.8 KB391152225a39
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_token_routing_1.pngOther45.0 KBa745057720fa
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_token_routing_2.pngOther45.3 KBca9b11b1f204
legacy/gemma2/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_token_routing_3.pngOther45.4 KB9ef74ca04862
legacy/gemma2/exp5_gated_topk_seed2/best_model/expert_pools.ptOther45.0 MBa7caae72cf23
legacy/gemma2/exp5_gated_topk_seed2/best_model/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/best_model/gating_network.ptOther33.9 MBa9a0f5416a8b
legacy/gemma2/exp5_gated_topk_seed2/best_model/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1000/expert_pools.ptOther45.0 MB2952bf81a42c
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1000/gating_network.ptOther33.9 MB1fabb030697b
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1500/expert_pools.ptOther45.0 MB2b6e05bf5678
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1500/gating_network.ptOther33.9 MB5e4d1a24af47
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2000/expert_pools.ptOther45.0 MB6049b6e3634d
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2000/gating_network.ptOther33.9 MBe20dc0b874c3
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2500/expert_pools.ptOther45.0 MB5ec07c49afd9
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2500/gating_network.ptOther33.9 MB8fd92fe98617
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3000/expert_pools.ptOther45.0 MB3509fb06e0ea
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3000/gating_network.ptOther33.9 MB00d10ee41890
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3500/expert_pools.ptOther45.0 MB5f3f44303c01
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3500/gating_network.ptOther33.9 MB1cd6b898a7b5
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-4000/expert_pools.ptOther45.0 MB085a47887d4f
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-4000/gating_network.ptOther33.9 MB6aa5e22d580f
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-500/expert_pools.ptOther45.0 MBa7caae72cf23
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-500/gating_network.ptOther33.9 MBa9a0f5416a8b
legacy/gemma2/exp5_gated_topk_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp5_gated_topk_seed2/final_model/expert_pools.ptOther45.0 MB623461cc8853
legacy/gemma2/exp5_gated_topk_seed2/final_model/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed2/final_model/gating_network.ptOther33.9 MB80f6e4c1665e
legacy/gemma2/exp5_gated_topk_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_entropy_per_layer.pngOther42.0 KBc7ead4486bf3
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_layer_expert_heatmap.pngOther104.2 KB1b44f7b656db
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_task_expert_usage.pngOther42.9 KB90f57ade1ad0
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_token_routing_1.pngOther45.4 KBb1af681290b4
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_token_routing_2.pngOther45.6 KBb16c208e8c4b
legacy/gemma2/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_token_routing_3.pngOther45.7 KB2531a1487a92
legacy/gemma2/exp5_gated_topk_seed3/best_model/expert_pools.ptOther45.0 MBc20bab8a2467
legacy/gemma2/exp5_gated_topk_seed3/best_model/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/best_model/gating_network.ptOther33.9 MB623e2317ea56
legacy/gemma2/exp5_gated_topk_seed3/best_model/scheduler.ptOther1.7 KB3e43ae9e84c6
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1000/expert_pools.ptOther45.0 MB0598537bee59
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1000/gating_network.ptOther33.9 MB92b50abd9c57
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1500/expert_pools.ptOther45.0 MBf888397073dd
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1500/gating_network.ptOther33.9 MB83f9275412c8
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2000/expert_pools.ptOther45.0 MB815a2e3f73d2
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2000/gating_network.ptOther33.9 MB25d15f2477aa
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2500/expert_pools.ptOther45.0 MB84db21393d8b
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2500/gating_network.ptOther33.9 MBf278da88746c
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3000/expert_pools.ptOther45.0 MB13386431ccfa
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3000/gating_network.ptOther33.9 MB8fa3299c7426
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3500/expert_pools.ptOther45.0 MB01f5fed3020c
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3500/gating_network.ptOther33.9 MBa0834c04aa9c
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-4000/expert_pools.ptOther45.0 MBaa1946366b9b
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-4000/gating_network.ptOther33.9 MB964608710705
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-500/expert_pools.ptOther45.0 MBd92affcafc24
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-500/gating_network.ptOther33.9 MB281c27cdc073
legacy/gemma2/exp5_gated_topk_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/gemma2/exp5_gated_topk_seed3/final_model/expert_pools.ptOther45.0 MB23144c58a670
legacy/gemma2/exp5_gated_topk_seed3/final_model/gated_lora_config.ptOther1.8 KB4bc4c620f6ff
legacy/gemma2/exp5_gated_topk_seed3/final_model/gating_network.ptOther33.9 MB7321c0101492
legacy/gemma2/exp5_gated_topk_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_entropy_per_layer.pngOther42.1 KB609d81b36570
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_layer_expert_heatmap.pngOther103.6 KB9ee5036cf3f2
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_task_expert_usage.pngOther42.2 KB9e5743cb8570
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_token_routing_1.pngOther45.5 KB702316e1f5da
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_token_routing_2.pngOther45.8 KBccd939250deb
legacy/gemma2/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_token_routing_3.pngOther45.8 KB0319b7a6d296
legacy/llama32/exp4_gated_loadbalancing_seed1/best_model/expert_pools.ptOther64.5 MB0d766ecc786b
legacy/llama32/exp4_gated_loadbalancing_seed1/best_model/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/best_model/gating_network.ptOther49.0 MBbee37304d29e
legacy/llama32/exp4_gated_loadbalancing_seed1/best_model/scheduler.ptOther1.7 KB2514a309dd4f
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1000/expert_pools.ptOther64.5 MB80048e08fbbb
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1000/gating_network.ptOther49.0 MB0b8f779093a4
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1500/expert_pools.ptOther64.5 MBf471f93c00bd
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1500/gating_network.ptOther49.0 MB8d725830632f
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2000/expert_pools.ptOther64.5 MB683e92cbb2e3
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2000/gating_network.ptOther49.0 MBe9cbee6998a6
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2500/expert_pools.ptOther64.5 MB5833badf4fe9
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2500/gating_network.ptOther49.0 MBa13f2612bbbf
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3000/expert_pools.ptOther64.5 MBdb3783a9e036
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3000/gating_network.ptOther49.0 MBb157b243716c
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3500/expert_pools.ptOther64.5 MBfa97f5da8886
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3500/gating_network.ptOther49.0 MB65493d57c4b5
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-4000/expert_pools.ptOther64.5 MBe4176f88a032
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-4000/gating_network.ptOther49.0 MB62fefee373d2
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-500/expert_pools.ptOther64.5 MB0f31ec4a8f0c
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-500/gating_network.ptOther49.0 MBe2ec05062e95
legacy/llama32/exp4_gated_loadbalancing_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32/exp4_gated_loadbalancing_seed1/final_model/expert_pools.ptOther64.5 MB7275ea7d0e50
legacy/llama32/exp4_gated_loadbalancing_seed1/final_model/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed1/final_model/gating_network.ptOther49.0 MBef96434547a2
legacy/llama32/exp4_gated_loadbalancing_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_entropy_per_layer.pngOther43.1 KBc5d766ce69a6
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_layer_expert_heatmap.pngOther110.5 KBa994399d62d0
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_task_expert_usage.pngOther38.0 KB683cb1ff1305
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_1.pngOther94.3 KB5b02ea02521b
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_2.pngOther94.6 KBbfef7d86184c
legacy/llama32/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_3.pngOther94.7 KB70cf6b16c476
legacy/llama32/exp4_gated_loadbalancing_seed2/best_model/expert_pools.ptOther64.5 MBd242c8caae8a
legacy/llama32/exp4_gated_loadbalancing_seed2/best_model/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/best_model/gating_network.ptOther49.0 MB1c7eac4e848f
legacy/llama32/exp4_gated_loadbalancing_seed2/best_model/scheduler.ptOther1.7 KB2514a309dd4f
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1000/expert_pools.ptOther64.5 MBaa835552d007
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1000/gating_network.ptOther49.0 MBf692ae3772cc
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1500/expert_pools.ptOther64.5 MBb2117952990e
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1500/gating_network.ptOther49.0 MBcb9f82ed5055
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2000/expert_pools.ptOther64.5 MB8c60c8218af7
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2000/gating_network.ptOther49.0 MB6d2ac31800bb
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2500/expert_pools.ptOther64.5 MBc6b1051873b7
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2500/gating_network.ptOther49.0 MB7f813adfd0ec
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3000/expert_pools.ptOther64.5 MB45959678a1d2
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3000/gating_network.ptOther49.0 MBb564cd4c8f6f
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3500/expert_pools.ptOther64.5 MBc281235d0737
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3500/gating_network.ptOther49.0 MBc50919174b21
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-4000/expert_pools.ptOther64.5 MB84516926c987
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-4000/gating_network.ptOther49.0 MB67e6ce35932e
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-500/expert_pools.ptOther64.5 MBabd71f46e55f
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-500/gating_network.ptOther49.0 MB13cfcab995df
legacy/llama32/exp4_gated_loadbalancing_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32/exp4_gated_loadbalancing_seed2/final_model/expert_pools.ptOther64.5 MBde0a8ff5d9b5
legacy/llama32/exp4_gated_loadbalancing_seed2/final_model/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed2/final_model/gating_network.ptOther49.0 MBb598e4266cd3
legacy/llama32/exp4_gated_loadbalancing_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_entropy_per_layer.pngOther43.1 KBc324664d4057
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_layer_expert_heatmap.pngOther109.0 KB755b7d719e55
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_task_expert_usage.pngOther37.8 KB2573d6493b29
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_1.pngOther94.8 KB8529a6acedb6
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_2.pngOther95.1 KB33d555706b3f
legacy/llama32/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_3.pngOther95.1 KBd36211d4f1b6
legacy/llama32/exp4_gated_loadbalancing_seed3/best_model/expert_pools.ptOther64.5 MBe4f79f8640b6
legacy/llama32/exp4_gated_loadbalancing_seed3/best_model/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/best_model/gating_network.ptOther49.0 MB905552fd0020
legacy/llama32/exp4_gated_loadbalancing_seed3/best_model/scheduler.ptOther1.7 KB2514a309dd4f
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1000/expert_pools.ptOther64.5 MB0b7441f33f57
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1000/gating_network.ptOther49.0 MB9a973b354c64
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1500/expert_pools.ptOther64.5 MB1a14ad8e4f9a
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1500/gating_network.ptOther49.0 MBf4add5695415
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2000/expert_pools.ptOther64.5 MB40dc65679f60
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2000/gating_network.ptOther49.0 MB4c9c7c9f9153
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2500/expert_pools.ptOther64.5 MBc27e0776e3e7
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2500/gating_network.ptOther49.0 MB459828b1dd18
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3000/expert_pools.ptOther64.5 MB33da80528244
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3000/gating_network.ptOther49.0 MB3cfdc17c9c26
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3500/expert_pools.ptOther64.5 MB7fd37f85a404
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3500/gating_network.ptOther49.0 MB11a62ebaf321
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-4000/expert_pools.ptOther64.5 MBede3f047a06d
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-4000/gating_network.ptOther49.0 MBba76c5939d85
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-500/expert_pools.ptOther64.5 MB2833e2129699
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-500/gating_network.ptOther49.0 MBa5386dd311a1
legacy/llama32/exp4_gated_loadbalancing_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32/exp4_gated_loadbalancing_seed3/final_model/expert_pools.ptOther64.5 MB08114e64921f
legacy/llama32/exp4_gated_loadbalancing_seed3/final_model/gated_lora_config.ptOther1.8 KBf8397a61a2d0
legacy/llama32/exp4_gated_loadbalancing_seed3/final_model/gating_network.ptOther49.0 MB35f2d0ad7725
legacy/llama32/exp4_gated_loadbalancing_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_entropy_per_layer.pngOther43.5 KB562ebb26c482
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_layer_expert_heatmap.pngOther109.2 KBb62bf8c8260d
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_task_expert_usage.pngOther38.2 KBad9445318577
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_token_routing_1.pngOther95.3 KB088f744a7596
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_token_routing_2.pngOther95.5 KB4191d781afa3
legacy/llama32/exp4_gated_loadbalancing_seed3/visualizations/exp4_gated_loadbalancing_seed3_token_routing_3.pngOther95.6 KB4bc5cad6fc6c
legacy/llama32/exp5_gated_topk_seed1/best_model/expert_pools.ptOther64.5 MBe31fab0e5df1
legacy/llama32/exp5_gated_topk_seed1/best_model/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/best_model/gating_network.ptOther49.0 MB6d52a123c421
legacy/llama32/exp5_gated_topk_seed1/best_model/scheduler.ptOther1.7 KB925b6912f2d8
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1000/expert_pools.ptOther64.5 MB608f18564bb3
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1000/gating_network.ptOther49.0 MBd835af6cd4fd
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1500/expert_pools.ptOther64.5 MB313ebe5c4070
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1500/gating_network.ptOther49.0 MB59477ce7c4a4
legacy/llama32/exp5_gated_topk_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2000/expert_pools.ptOther64.5 MBb3f2c4a98db0
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2000/gating_network.ptOther49.0 MB1eb4bcb7b618
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2500/expert_pools.ptOther64.5 MB1b6a82be6c20
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2500/gating_network.ptOther49.0 MB2dab187f6a43
legacy/llama32/exp5_gated_topk_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3000/expert_pools.ptOther64.5 MBc0d3a82c899c
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3000/gating_network.ptOther49.0 MB8156b771f88e
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3500/expert_pools.ptOther64.5 MBe46799baf557
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3500/gating_network.ptOther49.0 MB52e62e2f6e1d
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32/exp5_gated_topk_seed1/checkpoint-4000/expert_pools.ptOther64.5 MB7a3f0108df65
legacy/llama32/exp5_gated_topk_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-4000/gating_network.ptOther49.0 MB89a5bb446101
legacy/llama32/exp5_gated_topk_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32/exp5_gated_topk_seed1/checkpoint-500/expert_pools.ptOther64.5 MBf16aa5354f62
legacy/llama32/exp5_gated_topk_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/checkpoint-500/gating_network.ptOther49.0 MB871e72024cc6
legacy/llama32/exp5_gated_topk_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32/exp5_gated_topk_seed1/final_model/expert_pools.ptOther64.5 MBe3cc231c3fb7
legacy/llama32/exp5_gated_topk_seed1/final_model/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed1/final_model/gating_network.ptOther49.0 MBbbec5b3e8072
legacy/llama32/exp5_gated_topk_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_entropy_per_layer.pngOther42.2 KB146a5b23e03f
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_layer_expert_heatmap.pngOther107.4 KBcf98c1d43c44
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_task_expert_usage.pngOther43.0 KB4d312379d7ee
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_token_routing_1.pngOther94.0 KBecce9f326df8
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_token_routing_2.pngOther94.4 KB99aad2a75b65
legacy/llama32/exp5_gated_topk_seed1/visualizations/exp5_gated_topk_seed1_token_routing_3.pngOther94.4 KBf293b25fd6b8
legacy/llama32/exp5_gated_topk_seed2/best_model/expert_pools.ptOther64.5 MBd226d889bce8
legacy/llama32/exp5_gated_topk_seed2/best_model/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/best_model/gating_network.ptOther49.0 MB19bd07246cdc
legacy/llama32/exp5_gated_topk_seed2/best_model/scheduler.ptOther1.7 KB256b90657b80
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1000/expert_pools.ptOther64.5 MB2c2c888aa3d7
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1000/gating_network.ptOther49.0 MB1d47079f0e31
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1500/expert_pools.ptOther64.5 MBc87c0e0bc2b5
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1500/gating_network.ptOther49.0 MBe855669114a2
legacy/llama32/exp5_gated_topk_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2000/expert_pools.ptOther64.5 MB7f281910c740
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2000/gating_network.ptOther49.0 MBc2462be16e5d
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2500/expert_pools.ptOther64.5 MB34e5d035effd
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2500/gating_network.ptOther49.0 MBe3595af97402
legacy/llama32/exp5_gated_topk_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3000/expert_pools.ptOther64.5 MB44d652772422
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3000/gating_network.ptOther49.0 MB62b23e980d18
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3500/expert_pools.ptOther64.5 MB3a4160bae924
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3500/gating_network.ptOther49.0 MB18fb382b152d
legacy/llama32/exp5_gated_topk_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32/exp5_gated_topk_seed2/checkpoint-4000/expert_pools.ptOther64.5 MB4a8367459788
legacy/llama32/exp5_gated_topk_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-4000/gating_network.ptOther49.0 MBc40898ec505a
legacy/llama32/exp5_gated_topk_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32/exp5_gated_topk_seed2/checkpoint-500/expert_pools.ptOther64.5 MB12b93f4e9b89
legacy/llama32/exp5_gated_topk_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/checkpoint-500/gating_network.ptOther49.0 MBfe53905e58b9
legacy/llama32/exp5_gated_topk_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32/exp5_gated_topk_seed2/final_model/expert_pools.ptOther64.5 MB64ff657589df
legacy/llama32/exp5_gated_topk_seed2/final_model/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed2/final_model/gating_network.ptOther49.0 MB98016e967daa
legacy/llama32/exp5_gated_topk_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_entropy_per_layer.pngOther42.2 KB24caaeb2c2a8
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_layer_expert_heatmap.pngOther108.1 KB3a5b65f0b55e
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_task_expert_usage.pngOther44.0 KBd3f62189111b
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_token_routing_1.pngOther94.5 KBa686f8e84477
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_token_routing_2.pngOther94.8 KBa5e0afd4cbec
legacy/llama32/exp5_gated_topk_seed2/visualizations/exp5_gated_topk_seed2_token_routing_3.pngOther94.9 KBaaa8df72d609
legacy/llama32/exp5_gated_topk_seed3/best_model/expert_pools.ptOther64.5 MBead921469abd
legacy/llama32/exp5_gated_topk_seed3/best_model/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/best_model/gating_network.ptOther49.0 MB8c5f0e6530a2
legacy/llama32/exp5_gated_topk_seed3/best_model/scheduler.ptOther1.7 KB256b90657b80
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1000/expert_pools.ptOther64.5 MBec6778bcc143
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1000/gating_network.ptOther49.0 MB6a155763b163
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1500/expert_pools.ptOther64.5 MB9a5650a9897b
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1500/gating_network.ptOther49.0 MB5c82171e2e88
legacy/llama32/exp5_gated_topk_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2000/expert_pools.ptOther64.5 MB70b0f7706383
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2000/gating_network.ptOther49.0 MBb077050c42ab
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2500/expert_pools.ptOther64.5 MB720598de928f
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2500/gating_network.ptOther49.0 MB696e0e75e364
legacy/llama32/exp5_gated_topk_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3000/expert_pools.ptOther64.5 MBc0ec9e5e7a2d
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3000/gating_network.ptOther49.0 MBbe8cb0509a49
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3500/expert_pools.ptOther64.5 MB6ddfd304c835
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3500/gating_network.ptOther49.0 MBd9e348da4c22
legacy/llama32/exp5_gated_topk_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32/exp5_gated_topk_seed3/checkpoint-4000/expert_pools.ptOther64.5 MB53924f09c0b7
legacy/llama32/exp5_gated_topk_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-4000/gating_network.ptOther49.0 MBb8aee9ba2f35
legacy/llama32/exp5_gated_topk_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32/exp5_gated_topk_seed3/checkpoint-500/expert_pools.ptOther64.5 MB161157758ec2
legacy/llama32/exp5_gated_topk_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/checkpoint-500/gating_network.ptOther49.0 MB55069b5071a9
legacy/llama32/exp5_gated_topk_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32/exp5_gated_topk_seed3/final_model/expert_pools.ptOther64.5 MB6ede4455b64c
legacy/llama32/exp5_gated_topk_seed3/final_model/gated_lora_config.ptOther1.8 KB086f8a5d1324
legacy/llama32/exp5_gated_topk_seed3/final_model/gating_network.ptOther49.0 MBd32214591c23
legacy/llama32/exp5_gated_topk_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_entropy_per_layer.pngOther42.0 KB2dc1b37af3ea
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_layer_expert_heatmap.pngOther107.3 KB545c96636c0b
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_task_expert_usage.pngOther43.3 KB554cec393476
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_token_routing_1.pngOther94.4 KB9aa26977f625
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_token_routing_2.pngOther94.7 KB30d712015aea
legacy/llama32/exp5_gated_topk_seed3/visualizations/exp5_gated_topk_seed3_token_routing_3.pngOther94.7 KB53d6ac604600
legacy/llama32_modified/ablation_partial_gating_seed1/TRAINING_DONEOther111 B—
legacy/llama32_modified/ablation_partial_gating_seed1/best_model/expert_pools.ptOther64.5 MB41ce4e0651cd
legacy/llama32_modified/ablation_partial_gating_seed1/best_model/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/best_model/gating_network.ptOther49.0 MB310c51044020
legacy/llama32_modified/ablation_partial_gating_seed1/best_model/scheduler.ptOther1.7 KB9cfe7f6241e3
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1000/expert_pools.ptOther64.5 MB46a8130b4df6
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1000/gating_network.ptOther49.0 MB6d31ef8caf32
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1500/expert_pools.ptOther64.5 MBf5c536404445
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1500/gating_network.ptOther49.0 MBa15e7cb4cdcd
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2000/expert_pools.ptOther64.5 MBede1986a0e91
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2000/gating_network.ptOther49.0 MB86fd6c2b1ea0
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2500/expert_pools.ptOther64.5 MB0b4ebd88cf8e
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2500/gating_network.ptOther49.0 MB28c52ae53e09
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3000/expert_pools.ptOther64.5 MB3626bd019479
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3000/gating_network.ptOther49.0 MBc37b2017019c
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3500/expert_pools.ptOther64.5 MB50562fde64e9
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3500/gating_network.ptOther49.0 MB5016ad5273d3
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-4000/expert_pools.ptOther64.5 MB648948198bca
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-4000/gating_network.ptOther49.0 MBf6dd00b82534
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-500/expert_pools.ptOther64.5 MB08db63f685ae
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-500/gating_network.ptOther49.0 MB4f9764b6b5a9
legacy/llama32_modified/ablation_partial_gating_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-1/expert_pools.ptOther64.5 MB44381d9d8d81
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-1/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-1/gating_network.ptOther49.0 MB181ac46d9260
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-2/expert_pools.ptOther64.5 MBa80732f85112
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-2/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-2/gating_network.ptOther49.0 MB8493e97b55c3
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-3/expert_pools.ptOther64.5 MB800f6a32b356
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-3/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-3/gating_network.ptOther49.0 MB2844a3730373
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-4/expert_pools.ptOther64.5 MB34b1f4ac1cc4
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-4/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-4/gating_network.ptOther49.0 MBac7b77ae5d78
legacy/llama32_modified/ablation_partial_gating_seed1/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/ablation_partial_gating_seed1/final_model/expert_pools.ptOther64.5 MB34b1f4ac1cc4
legacy/llama32_modified/ablation_partial_gating_seed1/final_model/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/final_model/gating_network.ptOther49.0 MBac7b77ae5d78
legacy/llama32_modified/ablation_partial_gating_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/ablation_partial_gating_seed1/latest/expert_pools.ptOther64.5 MB800f6a32b356
legacy/llama32_modified/ablation_partial_gating_seed1/latest/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed1/latest/gating_network.ptOther49.0 MB2844a3730373
legacy/llama32_modified/ablation_partial_gating_seed1/latest/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_entropy_per_layer.pngOther41.2 KB270a944d8eb6
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_layer_expert_heatmap.pngOther107.9 KB6e463c9cb268
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_task_expert_usage.pngOther40.8 KBda0def403a7a
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_token_routing_1.pngOther96.1 KB7d88e879eb78
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_token_routing_2.pngOther96.4 KB0062bcd2298d
legacy/llama32_modified/ablation_partial_gating_seed1/visualizations/ablation_partial_gating_seed1_token_routing_3.pngOther96.4 KBf85aa88117dd
legacy/llama32_modified/ablation_partial_gating_seed2/TRAINING_DONEOther111 B—
legacy/llama32_modified/ablation_partial_gating_seed2/best_model/expert_pools.ptOther64.5 MBac4954e23fab
legacy/llama32_modified/ablation_partial_gating_seed2/best_model/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/best_model/gating_network.ptOther49.0 MB87458d54a175
legacy/llama32_modified/ablation_partial_gating_seed2/best_model/scheduler.ptOther1.7 KB18e42a717ff0
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1000/expert_pools.ptOther64.5 MB24840b7baa7c
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1000/gating_network.ptOther49.0 MB681ae53c23e6
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1500/expert_pools.ptOther64.5 MB6056181f0d44
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1500/gating_network.ptOther49.0 MB6d214ad18220
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2000/expert_pools.ptOther64.5 MBd4a34c76d227
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2000/gating_network.ptOther49.0 MB11877204e4cd
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2500/expert_pools.ptOther64.5 MBbfe1e66e237d
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2500/gating_network.ptOther49.0 MB099ead682314
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3000/expert_pools.ptOther64.5 MB36235ab2d47c
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3000/gating_network.ptOther49.0 MB98f158815d06
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3500/expert_pools.ptOther64.5 MBa1177b67914c
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3500/gating_network.ptOther49.0 MB37d3bae2ab93
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-4000/expert_pools.ptOther64.5 MB93a4cd64bd57
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-4000/gating_network.ptOther49.0 MBd4969ba6abfe
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-500/expert_pools.ptOther64.5 MB6b6d3fada213
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-500/gating_network.ptOther49.0 MBac4e3852aaa0
legacy/llama32_modified/ablation_partial_gating_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-1/expert_pools.ptOther64.5 MB7d20f6399afd
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-1/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-1/gating_network.ptOther49.0 MBb35daaeedcd0
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-2/expert_pools.ptOther64.5 MBe2309d70ba1f
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-2/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-2/gating_network.ptOther49.0 MB357be4fb5661
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-3/expert_pools.ptOther64.5 MB7b1aaed8104c
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-3/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-3/gating_network.ptOther49.0 MB6cde71d94712
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-4/expert_pools.ptOther64.5 MBdcd13e66130c
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-4/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-4/gating_network.ptOther49.0 MBf0ba860705f7
legacy/llama32_modified/ablation_partial_gating_seed2/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/ablation_partial_gating_seed2/final_model/expert_pools.ptOther64.5 MBdcd13e66130c
legacy/llama32_modified/ablation_partial_gating_seed2/final_model/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/final_model/gating_network.ptOther49.0 MBf0ba860705f7
legacy/llama32_modified/ablation_partial_gating_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/ablation_partial_gating_seed2/latest/expert_pools.ptOther64.5 MB7b1aaed8104c
legacy/llama32_modified/ablation_partial_gating_seed2/latest/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed2/latest/gating_network.ptOther49.0 MB6cde71d94712
legacy/llama32_modified/ablation_partial_gating_seed2/latest/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_entropy_per_layer.pngOther42.3 KB9c4dbd374a43
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_layer_expert_heatmap.pngOther110.5 KB99db23c6696b
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_task_expert_usage.pngOther42.7 KBaf8f918156cb
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_token_routing_1.pngOther95.7 KB02df57b8ddd1
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_token_routing_2.pngOther96.0 KBd0ecfadaf50b
legacy/llama32_modified/ablation_partial_gating_seed2/visualizations/ablation_partial_gating_seed2_token_routing_3.pngOther96.1 KBb50af141beca
legacy/llama32_modified/ablation_partial_gating_seed3/TRAINING_DONEOther111 B—
legacy/llama32_modified/ablation_partial_gating_seed3/best_model/expert_pools.ptOther64.5 MBd3e9f6e6d9f2
legacy/llama32_modified/ablation_partial_gating_seed3/best_model/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/best_model/gating_network.ptOther49.0 MBebe961971846
legacy/llama32_modified/ablation_partial_gating_seed3/best_model/scheduler.ptOther1.7 KB482234865e56
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1000/expert_pools.ptOther64.5 MB384f9d0ec787
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1000/gating_network.ptOther49.0 MB3936ce9cbdab
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1500/expert_pools.ptOther64.5 MBf10bea1e6148
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1500/gating_network.ptOther49.0 MBeabe91614e0d
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2000/expert_pools.ptOther64.5 MBe0346f0ae7c0
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2000/gating_network.ptOther49.0 MB51ba8f3510b3
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2500/expert_pools.ptOther64.5 MBf1810be7a960
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2500/gating_network.ptOther49.0 MBcf2953d85485
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3000/expert_pools.ptOther64.5 MBf46951f6b8e3
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3000/gating_network.ptOther49.0 MBe6b80d8c1e9e
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3500/expert_pools.ptOther64.5 MB7c81ea6a1174
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3500/gating_network.ptOther49.0 MB41f67e8369a6
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-4000/expert_pools.ptOther64.5 MBad3c876e6271
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-4000/gating_network.ptOther49.0 MBabd0045f8cba
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-500/expert_pools.ptOther64.5 MBa4fba24d614f
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-500/gating_network.ptOther49.0 MB90088b3c51c7
legacy/llama32_modified/ablation_partial_gating_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-1/expert_pools.ptOther64.5 MBbcfd27a6c015
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-1/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-1/gating_network.ptOther49.0 MBd324555e6e98
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-2/expert_pools.ptOther64.5 MBbf24112318e4
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-2/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-2/gating_network.ptOther49.0 MBd21c0e411ad1
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-3/expert_pools.ptOther64.5 MB75cc0ceda86c
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-3/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-3/gating_network.ptOther49.0 MBad33cf380355
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-4/expert_pools.ptOther64.5 MB8110ebf7de32
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-4/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-4/gating_network.ptOther49.0 MB36faea8202d1
legacy/llama32_modified/ablation_partial_gating_seed3/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/ablation_partial_gating_seed3/final_model/expert_pools.ptOther64.5 MB8110ebf7de32
legacy/llama32_modified/ablation_partial_gating_seed3/final_model/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/final_model/gating_network.ptOther49.0 MB36faea8202d1
legacy/llama32_modified/ablation_partial_gating_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/ablation_partial_gating_seed3/latest/expert_pools.ptOther64.5 MB75cc0ceda86c
legacy/llama32_modified/ablation_partial_gating_seed3/latest/gated_lora_config.ptOther1.8 KB373356a9eaef
legacy/llama32_modified/ablation_partial_gating_seed3/latest/gating_network.ptOther49.0 MBad33cf380355
legacy/llama32_modified/ablation_partial_gating_seed3/latest/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_entropy_per_layer.pngOther41.6 KB31995fa9eb34
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_layer_expert_heatmap.pngOther107.2 KB042ec209b3a2
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_task_expert_usage.pngOther41.6 KB783688a97887
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_token_routing_1.pngOther96.4 KB36364f2c3c3c
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_token_routing_2.pngOther96.7 KB01dbda95a91a
legacy/llama32_modified/ablation_partial_gating_seed3/visualizations/ablation_partial_gating_seed3_token_routing_3.pngOther96.8 KBf3d322aec1eb
legacy/llama32_modified/exp2_gated_2experts_seed1/TRAINING_DONEOther111 B—
legacy/llama32_modified/exp2_gated_2experts_seed1/best_model/expert_pools.ptOther27.7 MB68dcb5068983
legacy/llama32_modified/exp2_gated_2experts_seed1/best_model/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/best_model/gating_network.ptOther49.0 MB612c16792039
legacy/llama32_modified/exp2_gated_2experts_seed1/best_model/scheduler.ptOther1.7 KB482234865e56
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1000/expert_pools.ptOther27.7 MB5b27a6ce58da
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1000/gating_network.ptOther49.0 MB23e0691315cd
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1500/expert_pools.ptOther27.7 MB751b52abc6b1
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1500/gating_network.ptOther49.0 MB7fd53d4b5283
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2000/expert_pools.ptOther27.7 MB1e08fccc9995
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2000/gating_network.ptOther49.0 MBae4a12573bb7
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2500/expert_pools.ptOther27.7 MB68c2d0f4da9d
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2500/gating_network.ptOther49.0 MBe1c0539f106c
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3000/expert_pools.ptOther27.7 MB014eb87d19d0
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3000/gating_network.ptOther49.0 MBeec0e2bd60fc
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3500/expert_pools.ptOther27.7 MB44d866f4034d
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3500/gating_network.ptOther49.0 MBa5d44c8e4693
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-4000/expert_pools.ptOther27.7 MB2dfb783e352d
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-4000/gating_network.ptOther49.0 MB42b1158ff132
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-500/expert_pools.ptOther27.7 MB7aa71506349c
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-500/gating_network.ptOther49.0 MBfcc25276ae93
legacy/llama32_modified/exp2_gated_2experts_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-1/expert_pools.ptOther27.7 MBd7ef918b7104
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-1/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-1/gating_network.ptOther49.0 MB384c2a80687d
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-2/expert_pools.ptOther27.7 MBb2c7d4f6fde1
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-2/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-2/gating_network.ptOther49.0 MB0bba864f9995
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-3/expert_pools.ptOther27.7 MB3dd220757191
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-3/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-3/gating_network.ptOther49.0 MB723a0e2a8333
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-4/expert_pools.ptOther27.7 MB22ec89144a10
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-4/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-4/gating_network.ptOther49.0 MBa15991ad8bc8
legacy/llama32_modified/exp2_gated_2experts_seed1/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp2_gated_2experts_seed1/final_model/expert_pools.ptOther27.7 MB22ec89144a10
legacy/llama32_modified/exp2_gated_2experts_seed1/final_model/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/final_model/gating_network.ptOther49.0 MBa15991ad8bc8
legacy/llama32_modified/exp2_gated_2experts_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp2_gated_2experts_seed1/latest/expert_pools.ptOther27.7 MB6b4c0b948733
legacy/llama32_modified/exp2_gated_2experts_seed1/latest/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed1/latest/gating_network.ptOther49.0 MB0df454d5dc5e
legacy/llama32_modified/exp2_gated_2experts_seed1/latest/scheduler.ptOther1.7 KB1eb2f65e818e
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_entropy_per_layer.pngOther43.5 KBfb5fe9ac18dc
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_layer_expert_heatmap.pngOther107.0 KB20bc3d7aac62
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_task_expert_usage.pngOther39.9 KB51a266564ab5
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_1.pngOther98.0 KBec77fad11c5a
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_2.pngOther97.8 KB92481163ec0e
legacy/llama32_modified/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_3.pngOther98.3 KB3c3160e5940a
legacy/llama32_modified/exp2_gated_2experts_seed2/TRAINING_DONEOther110 B—
legacy/llama32_modified/exp2_gated_2experts_seed2/best_model/expert_pools.ptOther27.7 MBbf7dfe3c1c8b
legacy/llama32_modified/exp2_gated_2experts_seed2/best_model/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/best_model/gating_network.ptOther49.0 MBe5a131e98f69
legacy/llama32_modified/exp2_gated_2experts_seed2/best_model/scheduler.ptOther1.7 KBdf9d6a6f0d6d
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1000/expert_pools.ptOther27.7 MB01e7d2fd2ab5
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1000/gating_network.ptOther49.0 MB38dc33fbd2f9
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1500/expert_pools.ptOther27.7 MB428aa1096c4e
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1500/gating_network.ptOther49.0 MB883299405121
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2000/expert_pools.ptOther27.7 MBf46569546a9b
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2000/gating_network.ptOther49.0 MB536de9c38766
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2500/expert_pools.ptOther27.7 MB473b61d8efa9
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2500/gating_network.ptOther49.0 MB3edf308b8a85
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3000/expert_pools.ptOther27.7 MB6c0bee0f9309
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3000/gating_network.ptOther49.0 MBf17e00a1442d
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3500/expert_pools.ptOther27.7 MB20fe3633a1ff
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3500/gating_network.ptOther49.0 MBc19064542048
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-4000/expert_pools.ptOther27.7 MBfffda47da5a1
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-4000/gating_network.ptOther49.0 MB93feb4387174
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-500/expert_pools.ptOther27.7 MB2322c5d0e00e
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-500/gating_network.ptOther49.0 MBb4b354b9b23a
legacy/llama32_modified/exp2_gated_2experts_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-1/expert_pools.ptOther27.7 MBe89b203b58d5
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-1/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-1/gating_network.ptOther49.0 MB8d62e7bf3bda
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-2/expert_pools.ptOther27.7 MB5f750bafba2b
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-2/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-2/gating_network.ptOther49.0 MBb8eeff2bb986
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-3/expert_pools.ptOther27.7 MBb64ab3036055
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-3/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-3/gating_network.ptOther49.0 MB46503e4e473e
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-4/expert_pools.ptOther27.7 MB6ff8d8f73dfa
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-4/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-4/gating_network.ptOther49.0 MB1eb2e1d7587c
legacy/llama32_modified/exp2_gated_2experts_seed2/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp2_gated_2experts_seed2/final_model/expert_pools.ptOther27.7 MB6ff8d8f73dfa
legacy/llama32_modified/exp2_gated_2experts_seed2/final_model/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/final_model/gating_network.ptOther49.0 MB1eb2e1d7587c
legacy/llama32_modified/exp2_gated_2experts_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp2_gated_2experts_seed2/latest/expert_pools.ptOther27.7 MB3e73a78206ae
legacy/llama32_modified/exp2_gated_2experts_seed2/latest/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed2/latest/gating_network.ptOther49.0 MB64fea131ad94
legacy/llama32_modified/exp2_gated_2experts_seed2/latest/scheduler.ptOther1.7 KBc303bbee1321
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_entropy_per_layer.pngOther43.1 KB8d90a1b35103
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_layer_expert_heatmap.pngOther105.8 KB94b7fd243ee3
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_task_expert_usage.pngOther39.7 KB811cdca60071
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_1.pngOther96.6 KBd170cfc114fb
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_2.pngOther96.7 KB16a6c64e06f2
legacy/llama32_modified/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_3.pngOther96.9 KB8b58f8593b37
legacy/llama32_modified/exp2_gated_2experts_seed3/TRAINING_DONEOther111 B—
legacy/llama32_modified/exp2_gated_2experts_seed3/best_model/expert_pools.ptOther27.7 MB05f2b9f4331e
legacy/llama32_modified/exp2_gated_2experts_seed3/best_model/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/best_model/gating_network.ptOther49.0 MBd858fcb09456
legacy/llama32_modified/exp2_gated_2experts_seed3/best_model/scheduler.ptOther1.7 KB482234865e56
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1000/expert_pools.ptOther27.7 MB9644d5acdf64
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1000/gating_network.ptOther49.0 MB8ff7716857ae
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1500/expert_pools.ptOther27.7 MB427b2f885e9d
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1500/gating_network.ptOther49.0 MB219aadd1e643
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2000/expert_pools.ptOther27.7 MBea6c94c135e6
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2000/gating_network.ptOther49.0 MBdc0b14abfd9f
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2500/expert_pools.ptOther27.7 MB644084e5a58c
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2500/gating_network.ptOther49.0 MB73bfd6d8dc0f
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3000/expert_pools.ptOther27.7 MBda308a13d5f9
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3000/gating_network.ptOther49.0 MBcbf34078ed98
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3500/expert_pools.ptOther27.7 MB7520e294223b
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3500/gating_network.ptOther49.0 MB527aad39cc39
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-4000/expert_pools.ptOther27.7 MBd54a502143d1
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-4000/gating_network.ptOther49.0 MB2a0bb7cb38fb
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-500/expert_pools.ptOther27.7 MB92a00a5bb5ba
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-500/gating_network.ptOther49.0 MB4cad327b00e1
legacy/llama32_modified/exp2_gated_2experts_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-1/expert_pools.ptOther27.7 MBe45880e27526
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-1/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-1/gating_network.ptOther49.0 MBfb22fc2e55e7
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-2/expert_pools.ptOther27.7 MB94db359faa64
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-2/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-2/gating_network.ptOther49.0 MBc5c4660d015a
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-3/expert_pools.ptOther27.7 MBb58ab9512485
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-3/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-3/gating_network.ptOther49.0 MBc0edc0644d97
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-4/expert_pools.ptOther27.7 MBbea6bbbca420
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-4/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-4/gating_network.ptOther49.0 MBda692011734b
legacy/llama32_modified/exp2_gated_2experts_seed3/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp2_gated_2experts_seed3/final_model/expert_pools.ptOther27.7 MBbea6bbbca420
legacy/llama32_modified/exp2_gated_2experts_seed3/final_model/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/final_model/gating_network.ptOther49.0 MBda692011734b
legacy/llama32_modified/exp2_gated_2experts_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp2_gated_2experts_seed3/latest/expert_pools.ptOther27.7 MB895a8e09cdbe
legacy/llama32_modified/exp2_gated_2experts_seed3/latest/gated_lora_config.ptOther1.8 KB6404870bed52
legacy/llama32_modified/exp2_gated_2experts_seed3/latest/gating_network.ptOther49.0 MB33b42fabff87
legacy/llama32_modified/exp2_gated_2experts_seed3/latest/scheduler.ptOther1.7 KBfe1ed971b14a
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_entropy_per_layer.pngOther43.5 KB0b3bb7d0aeaa
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_layer_expert_heatmap.pngOther107.3 KBf9db93fd57d7
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_task_expert_usage.pngOther39.5 KBb535bee9f4ae
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_1.pngOther97.0 KBb65e5280890a
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_2.pngOther96.8 KB42b676735ed0
legacy/llama32_modified/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_3.pngOther97.2 KB6464761e688a
legacy/llama32_modified/exp3_gated_3experts_seed1/TRAINING_DONEOther110 B—
legacy/llama32_modified/exp3_gated_3experts_seed1/best_model/expert_pools.ptOther64.5 MBb727961601fc
legacy/llama32_modified/exp3_gated_3experts_seed1/best_model/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/best_model/gating_network.ptOther49.0 MBc7cc6aea64f7
legacy/llama32_modified/exp3_gated_3experts_seed1/best_model/scheduler.ptOther1.7 KB8a283b73661c
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1000/expert_pools.ptOther64.5 MBb9d65c55b89c
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1000/gating_network.ptOther49.0 MBafaa2ac3fa8b
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1500/expert_pools.ptOther64.5 MB65ed617ed723
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1500/gating_network.ptOther49.0 MB572f0cfa2039
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2000/expert_pools.ptOther64.5 MB57c42fdd2081
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2000/gating_network.ptOther49.0 MBc0252b88b7f4
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2500/expert_pools.ptOther64.5 MB8e7f86335718
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2500/gating_network.ptOther49.0 MB9c4181e450a3
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3000/expert_pools.ptOther64.5 MB7e7c0ea06f73
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3000/gating_network.ptOther49.0 MBbd0b63eccea5
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3500/expert_pools.ptOther64.5 MB854136fc31c4
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3500/gating_network.ptOther49.0 MB5c57aa5c480b
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-4000/expert_pools.ptOther64.5 MB4775301ccb00
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-4000/gating_network.ptOther49.0 MB5a7d8a0bed11
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-500/expert_pools.ptOther64.5 MBcb161f9e1879
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-500/gating_network.ptOther49.0 MB25b895ab4442
legacy/llama32_modified/exp3_gated_3experts_seed1/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-1/expert_pools.ptOther64.5 MB4d387e2140b1
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-1/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-1/gating_network.ptOther49.0 MB7731f46ea40c
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-2/expert_pools.ptOther64.5 MB6b9e20e267b6
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-2/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-2/gating_network.ptOther49.0 MB3d0b38d7943f
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-3/expert_pools.ptOther64.5 MBb3db4ac7bff9
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-3/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-3/gating_network.ptOther49.0 MBf748a4af4967
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-4/expert_pools.ptOther64.5 MBd090ef1bfbde
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-4/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-4/gating_network.ptOther49.0 MBc9ec16c9ab5b
legacy/llama32_modified/exp3_gated_3experts_seed1/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp3_gated_3experts_seed1/final_model/expert_pools.ptOther64.5 MBd090ef1bfbde
legacy/llama32_modified/exp3_gated_3experts_seed1/final_model/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/final_model/gating_network.ptOther49.0 MBc9ec16c9ab5b
legacy/llama32_modified/exp3_gated_3experts_seed1/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp3_gated_3experts_seed1/latest/expert_pools.ptOther64.5 MB9a8d68f5b09e
legacy/llama32_modified/exp3_gated_3experts_seed1/latest/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed1/latest/gating_network.ptOther49.0 MB8a82fb5156be
legacy/llama32_modified/exp3_gated_3experts_seed1/latest/scheduler.ptOther1.7 KB47244d889136
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_entropy_per_layer.pngOther43.0 KB867846d119f7
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_layer_expert_heatmap.pngOther109.5 KBd931ffda1f17
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_task_expert_usage.pngOther38.0 KB96b6fb38d5af
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_1.pngOther101.4 KB46c1ca26ef2c
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_2.pngOther101.6 KB2f61c5f2826e
legacy/llama32_modified/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_3.pngOther101.6 KB9cc38c5ab248
legacy/llama32_modified/exp3_gated_3experts_seed2/TRAINING_DONEOther111 B—
legacy/llama32_modified/exp3_gated_3experts_seed2/best_model/expert_pools.ptOther64.5 MB6daf6e648c3e
legacy/llama32_modified/exp3_gated_3experts_seed2/best_model/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/best_model/gating_network.ptOther49.0 MB2ce230a0646f
legacy/llama32_modified/exp3_gated_3experts_seed2/best_model/scheduler.ptOther1.7 KB42a3ddb1d5d9
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1000/expert_pools.ptOther64.5 MBac44c97badbf
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1000/gating_network.ptOther49.0 MB52210edc8262
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1500/expert_pools.ptOther64.5 MB47e08504d44e
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1500/gating_network.ptOther49.0 MBb6e18928d6f9
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2000/expert_pools.ptOther64.5 MB39e5fe48bd68
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2000/gating_network.ptOther49.0 MB68fb292d295c
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2500/expert_pools.ptOther64.5 MB202d1d722044
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2500/gating_network.ptOther49.0 MB84dc16cb3580
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3000/expert_pools.ptOther64.5 MB433d6d13c401
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3000/gating_network.ptOther49.0 MB6b9d662a753a
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3500/expert_pools.ptOther64.5 MB5897fa8e2cca
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3500/gating_network.ptOther49.0 MB91de3a3f208c
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-4000/expert_pools.ptOther64.5 MBd01a00bf8f22
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-4000/gating_network.ptOther49.0 MB4b771ee0373e
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-500/expert_pools.ptOther64.5 MBa5645fbef855
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-500/gating_network.ptOther49.0 MB4b301ce39bf2
legacy/llama32_modified/exp3_gated_3experts_seed2/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-1/expert_pools.ptOther64.5 MB8be389f8b97b
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-1/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-1/gating_network.ptOther49.0 MB65ab05a71d7a
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-2/expert_pools.ptOther64.5 MB69647f7f6680
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-2/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-2/gating_network.ptOther49.0 MB8e2a7f70bf68
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-3/expert_pools.ptOther64.5 MBe8e8e241977a
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-3/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-3/gating_network.ptOther49.0 MB03188a8f743b
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-4/expert_pools.ptOther64.5 MBdcd83c72ddc1
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-4/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-4/gating_network.ptOther49.0 MB260a51a53000
legacy/llama32_modified/exp3_gated_3experts_seed2/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp3_gated_3experts_seed2/final_model/expert_pools.ptOther64.5 MBdcd83c72ddc1
legacy/llama32_modified/exp3_gated_3experts_seed2/final_model/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/final_model/gating_network.ptOther49.0 MB260a51a53000
legacy/llama32_modified/exp3_gated_3experts_seed2/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp3_gated_3experts_seed2/latest/expert_pools.ptOther64.5 MB1dea69a268e9
legacy/llama32_modified/exp3_gated_3experts_seed2/latest/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed2/latest/gating_network.ptOther49.0 MBe5610f06f0fc
legacy/llama32_modified/exp3_gated_3experts_seed2/latest/scheduler.ptOther1.7 KB1ba100b8fd7e
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_entropy_per_layer.pngOther42.7 KB3afa7b2eeee4
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_layer_expert_heatmap.pngOther110.0 KB089ef87035ad
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_task_expert_usage.pngOther39.2 KBef8135ca29c0
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_1.pngOther99.3 KB78ce16f8fede
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_2.pngOther99.4 KB945a7afd6329
legacy/llama32_modified/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_3.pngOther99.2 KBd73a7cd61ab8
legacy/llama32_modified/exp3_gated_3experts_seed3/TRAINING_DONEOther111 B—
legacy/llama32_modified/exp3_gated_3experts_seed3/best_model/expert_pools.ptOther64.5 MB333fc4c1a453
legacy/llama32_modified/exp3_gated_3experts_seed3/best_model/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/best_model/gating_network.ptOther49.0 MB8392bfafd37c
legacy/llama32_modified/exp3_gated_3experts_seed3/best_model/scheduler.ptOther1.7 KBdf9d6a6f0d6d
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1000/expert_pools.ptOther64.5 MBb711d0710794
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1000/gating_network.ptOther49.0 MBb00b2c2f0cc8
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1000/scheduler.ptOther1.7 KB75d0483a00c8
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1500/expert_pools.ptOther64.5 MB11d480b984db
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1500/gating_network.ptOther49.0 MB5d1931059148
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-1500/scheduler.ptOther1.7 KBa930f4a734fe
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2000/expert_pools.ptOther64.5 MB1e2393040cad
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2000/gating_network.ptOther49.0 MBdb6674e0d77e
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2000/scheduler.ptOther1.7 KBb5ded2205a43
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2500/expert_pools.ptOther64.5 MB33788dc0aa2a
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2500/gating_network.ptOther49.0 MB3540cbc818f7
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-2500/scheduler.ptOther1.7 KB8edb16e6cc0f
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3000/expert_pools.ptOther64.5 MB677c0e88f54b
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3000/gating_network.ptOther49.0 MB5db82b18deae
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3000/scheduler.ptOther1.7 KB438e7c9aba3b
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3500/expert_pools.ptOther64.5 MB23c228a5d689
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3500/gating_network.ptOther49.0 MB8b9bfa12e64a
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-3500/scheduler.ptOther1.7 KB82a86129150c
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-4000/expert_pools.ptOther64.5 MB851a6e7338f8
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-4000/gating_network.ptOther49.0 MB688f9dfd7bbd
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-4000/scheduler.ptOther1.7 KB9793afb9d64c
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-500/expert_pools.ptOther64.5 MB27ea8f202631
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-500/gating_network.ptOther49.0 MBda799a696ee7
legacy/llama32_modified/exp3_gated_3experts_seed3/checkpoint-500/scheduler.ptOther1.7 KBd2c4f1faa352
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-1/expert_pools.ptOther64.5 MB33da064be2b8
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-1/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-1/gating_network.ptOther49.0 MBf5c2993d9792
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-1/scheduler.ptOther1.7 KB86f6731f97de
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-2/expert_pools.ptOther64.5 MB2bd4e6205aa7
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-2/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-2/gating_network.ptOther49.0 MB0281c6e88ec7
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-2/scheduler.ptOther1.7 KB61c7ee9cc021
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-3/expert_pools.ptOther64.5 MB75d6bd6ec2b5
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-3/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-3/gating_network.ptOther49.0 MBbd4ca5610075
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-3/scheduler.ptOther1.7 KB3166389b2e94
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-4/expert_pools.ptOther64.5 MBe74fffefd2e7
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-4/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-4/gating_network.ptOther49.0 MBdda2f1d92362
legacy/llama32_modified/exp3_gated_3experts_seed3/epoch-4/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp3_gated_3experts_seed3/final_model/expert_pools.ptOther64.5 MBe74fffefd2e7
legacy/llama32_modified/exp3_gated_3experts_seed3/final_model/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/final_model/gating_network.ptOther49.0 MBdda2f1d92362
legacy/llama32_modified/exp3_gated_3experts_seed3/final_model/scheduler.ptOther1.7 KB445a69c8edd9
legacy/llama32_modified/exp3_gated_3experts_seed3/latest/expert_pools.ptOther64.5 MB14f0186e558d
legacy/llama32_modified/exp3_gated_3experts_seed3/latest/gated_lora_config.ptOther1.8 KBfdec88aa9d99
legacy/llama32_modified/exp3_gated_3experts_seed3/latest/gating_network.ptOther49.0 MBa670e7645ec7
legacy/llama32_modified/exp3_gated_3experts_seed3/latest/scheduler.ptOther1.7 KB47244d889136
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_entropy_per_layer.pngOther43.2 KB211c57019f33
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_layer_expert_heatmap.pngOther109.8 KB2db6bf80545d
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_task_expert_usage.pngOther38.4 KB204eb5e15946
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_1.pngOther98.2 KBbe1c6d503a7d
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_2.pngOther98.5 KB9e0b491696ab
legacy/llama32_modified/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_3.pngOther98.5 KB73a6d26de6f8
legacy/phi2_ablations/ablation_no_l1/best_model/expert_pools.ptOther73.8 MB576076631be7
legacy/phi2_ablations/ablation_no_l1/best_model/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/best_model/gating_network.ptOther47.3 MB6efbc0268a91
legacy/phi2_ablations/ablation_no_l1/best_model/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_no_l1/checkpoint-1000/expert_pools.ptOther73.8 MBca5e4e87c0d4
legacy/phi2_ablations/ablation_no_l1/checkpoint-1000/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/checkpoint-1000/gating_network.ptOther47.3 MB73e58b148c40
legacy/phi2_ablations/ablation_no_l1/checkpoint-1000/scheduler.ptOther1.7 KB9402f83b5cfc
legacy/phi2_ablations/ablation_no_l1/checkpoint-1500/expert_pools.ptOther73.8 MBb04d1a738f37
legacy/phi2_ablations/ablation_no_l1/checkpoint-1500/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/checkpoint-1500/gating_network.ptOther47.3 MB31e05fbd81de
legacy/phi2_ablations/ablation_no_l1/checkpoint-1500/scheduler.ptOther1.7 KB70f9dc0166fe
legacy/phi2_ablations/ablation_no_l1/checkpoint-2000/expert_pools.ptOther73.8 MB48155925d514
legacy/phi2_ablations/ablation_no_l1/checkpoint-2000/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/checkpoint-2000/gating_network.ptOther47.3 MBbac671736f62
legacy/phi2_ablations/ablation_no_l1/checkpoint-2000/scheduler.ptOther1.7 KB7a98b29dc045
legacy/phi2_ablations/ablation_no_l1/checkpoint-2500/expert_pools.ptOther73.8 MB576076631be7
legacy/phi2_ablations/ablation_no_l1/checkpoint-2500/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/checkpoint-2500/gating_network.ptOther47.3 MB6efbc0268a91
legacy/phi2_ablations/ablation_no_l1/checkpoint-2500/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_no_l1/checkpoint-500/expert_pools.ptOther73.8 MB949082f2099e
legacy/phi2_ablations/ablation_no_l1/checkpoint-500/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/checkpoint-500/gating_network.ptOther47.3 MB0fe5c8b38fcf
legacy/phi2_ablations/ablation_no_l1/checkpoint-500/scheduler.ptOther1.7 KBa8e00d52a8c7
legacy/phi2_ablations/ablation_no_l1/final_model/expert_pools.ptOther73.8 MB576076631be7
legacy/phi2_ablations/ablation_no_l1/final_model/gated_lora_config.ptOther1.7 KBa4ad4a3913e8
legacy/phi2_ablations/ablation_no_l1/final_model/gating_network.ptOther47.3 MB6efbc0268a91
legacy/phi2_ablations/ablation_no_l1/final_model/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_entropy_per_layer.pngOther40.6 KB4f780667c66c
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_layer_expert_heatmap.pngOther65.7 KBc08849f82770
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_task_expert_usage.pngOther49.8 KBf1faa7905d48
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_token_routing_1.pngOther62.5 KB762621db1af6
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_token_routing_2.pngOther84.1 KB7ac377071a44
legacy/phi2_ablations/ablation_no_l1/routing_analysis_v2/ablation_no_l1_token_routing_3.pngOther93.4 KB004c02d8def8
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_entropy_per_layer.pngOther40.8 KBd4d7199c4da0
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_layer_expert_heatmap.pngOther66.3 KB3c715f512027
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_task_expert_usage.pngOther39.2 KB52e43e643df1
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_token_routing_1.pngOther92.2 KB9979d4f3b796
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_token_routing_2.pngOther92.4 KB667f6ed33fbb
legacy/phi2_ablations/ablation_no_l1/visualizations/ablation_no_l1_token_routing_3.pngOther92.5 KBb32ba5884241
legacy/phi2_ablations/ablation_no_layer_emb/best_model/expert_pools.ptOther73.8 MBd5dda9d70af6
legacy/phi2_ablations/ablation_no_layer_emb/best_model/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/best_model/gating_network.ptOther42.1 MB0ccc6204c3ab
legacy/phi2_ablations/ablation_no_layer_emb/best_model/scheduler.ptOther1.7 KB317075d3824d
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1000/expert_pools.ptOther73.8 MB78d3597b4415
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1000/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1000/gating_network.ptOther42.1 MBa0b3e3561a25
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1000/scheduler.ptOther1.7 KB9402f83b5cfc
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1500/expert_pools.ptOther73.8 MB7089068087f0
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1500/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1500/gating_network.ptOther42.1 MBb90295ec9e71
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-1500/scheduler.ptOther1.7 KB70f9dc0166fe
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2000/expert_pools.ptOther73.8 MB89fa62a0e3e6
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2000/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2000/gating_network.ptOther42.1 MB9b644c164469
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2000/scheduler.ptOther1.7 KB7a98b29dc045
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2500/expert_pools.ptOther73.8 MB962c93171cd3
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2500/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2500/gating_network.ptOther42.1 MBa09772230554
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-2500/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-500/expert_pools.ptOther73.8 MBcff6fa359f9f
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-500/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-500/gating_network.ptOther42.1 MB9ba3752ba0f4
legacy/phi2_ablations/ablation_no_layer_emb/checkpoint-500/scheduler.ptOther1.7 KBa8e00d52a8c7
legacy/phi2_ablations/ablation_no_layer_emb/final_model/expert_pools.ptOther73.8 MB962c93171cd3
legacy/phi2_ablations/ablation_no_layer_emb/final_model/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_no_layer_emb/final_model/gating_network.ptOther42.1 MBa09772230554
legacy/phi2_ablations/ablation_no_layer_emb/final_model/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_entropy_per_layer.pngOther41.2 KB3f30eb730d91
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_layer_expert_heatmap.pngOther71.4 KB08c31f909e1d
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_task_expert_usage.pngOther51.5 KB9f1961209361
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_token_routing_1.pngOther99.0 KB877250e9fbca
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_token_routing_2.pngOther88.8 KBacd593fae934
legacy/phi2_ablations/ablation_no_layer_emb/routing_analysis_v2/ablation_no_layer_emb_token_routing_3.pngOther99.5 KB053bd07cfdea
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_entropy_per_layer.pngOther41.1 KB586f8aa984af
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_layer_expert_heatmap.pngOther67.5 KB5d392d365e83
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_task_expert_usage.pngOther40.8 KBd7c1883154ee
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_token_routing_1.pngOther93.8 KB497e36c83329
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_token_routing_2.pngOther94.0 KB153a5a46ce77
legacy/phi2_ablations/ablation_no_layer_emb/visualizations/ablation_no_layer_emb_token_routing_3.pngOther94.1 KB0a4172a4095a
legacy/phi2_ablations/ablation_partial_gating/best_model/expert_pools.ptOther73.8 MB421437ed88e5
legacy/phi2_ablations/ablation_partial_gating/best_model/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/best_model/gating_network.ptOther47.3 MB625f93b434b5
legacy/phi2_ablations/ablation_partial_gating/best_model/scheduler.ptOther1.7 KB0d500e7c9080
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1000/expert_pools.ptOther73.8 MB21608d54fa3a
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1000/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1000/gating_network.ptOther47.3 MB6c0c404ee2b5
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1000/scheduler.ptOther1.7 KB9402f83b5cfc
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1500/expert_pools.ptOther73.8 MB4467501aaa0c
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1500/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1500/gating_network.ptOther47.3 MBd3c54c262276
legacy/phi2_ablations/ablation_partial_gating/checkpoint-1500/scheduler.ptOther1.7 KB70f9dc0166fe
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2000/expert_pools.ptOther73.8 MBb3edae5c3e85
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2000/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2000/gating_network.ptOther47.3 MBaa863cb203f2
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2000/scheduler.ptOther1.7 KB7a98b29dc045
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2500/expert_pools.ptOther73.8 MB74c5b204441e
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2500/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2500/gating_network.ptOther47.3 MB4e8a8408aa42
legacy/phi2_ablations/ablation_partial_gating/checkpoint-2500/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_partial_gating/checkpoint-500/expert_pools.ptOther73.8 MBfbbc1c16a395
legacy/phi2_ablations/ablation_partial_gating/checkpoint-500/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/checkpoint-500/gating_network.ptOther47.3 MB1b75b8378c5e
legacy/phi2_ablations/ablation_partial_gating/checkpoint-500/scheduler.ptOther1.7 KBa8e00d52a8c7
legacy/phi2_ablations/ablation_partial_gating/final_model/expert_pools.ptOther73.8 MB74c5b204441e
legacy/phi2_ablations/ablation_partial_gating/final_model/gated_lora_config.ptOther1.7 KB591a420bc862
legacy/phi2_ablations/ablation_partial_gating/final_model/gating_network.ptOther47.3 MB4e8a8408aa42
legacy/phi2_ablations/ablation_partial_gating/final_model/scheduler.ptOther1.7 KBd4b3c87afb92
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_entropy_per_layer.pngOther41.7 KB5954a0775f80
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_layer_expert_heatmap.pngOther75.5 KB485fe6ec92fd
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_task_expert_usage.pngOther51.6 KBf44c1ac0cb8d
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_token_routing_1.pngOther82.6 KB9ffce17e1005
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_token_routing_2.pngOther66.4 KB0116013037bd
legacy/phi2_ablations/ablation_partial_gating/routing_analysis_v2/ablation_partial_gating_token_routing_3.pngOther85.7 KBe7bf8468de80
legacy/pythia410m/ablation_no_l1_seed1/best_model/expert_pools.ptOther16.7 MBcd2fe7a67eab
legacy/pythia410m/ablation_no_l1_seed1/best_model/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed1/best_model/gating_network.ptOther13.9 MBba730cf1962c
legacy/pythia410m/ablation_no_l1_seed1/best_model/scheduler.ptOther1.7 KBb2e0d0974bd3
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1000/expert_pools.ptOther16.7 MBbce4c4338aa7
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1000/gating_network.ptOther13.9 MB4be5f727acee
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1500/expert_pools.ptOther16.7 MB72e0d1514d5e
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1500/gating_network.ptOther13.9 MB6bd49e59ef0c
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-2000/expert_pools.ptOther16.7 MB70ded6236f9d
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-2000/gating_network.ptOther13.9 MBeeb096be017c
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-500/expert_pools.ptOther16.7 MBe1a4de457316
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-500/gating_network.ptOther13.9 MB21d9e784622f
legacy/pythia410m/ablation_no_l1_seed1/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_no_l1_seed2/best_model/expert_pools.ptOther16.7 MB0fba91bbd4c6
legacy/pythia410m/ablation_no_l1_seed2/best_model/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed2/best_model/gating_network.ptOther13.9 MB18e3d532f18c
legacy/pythia410m/ablation_no_l1_seed2/best_model/scheduler.ptOther1.7 KBd5482a786682
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1000/expert_pools.ptOther16.7 MB54d3a947ce06
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1000/gating_network.ptOther13.9 MBe4555e39a42d
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1500/expert_pools.ptOther16.7 MBcf216e2cb49f
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1500/gating_network.ptOther13.9 MBc3e65c780444
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-2000/expert_pools.ptOther16.7 MBfb78b51078e6
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-2000/gating_network.ptOther13.9 MBf06784d1c00a
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-500/expert_pools.ptOther16.7 MB1045f05f348f
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-500/gating_network.ptOther13.9 MBe30e59d6e68d
legacy/pythia410m/ablation_no_l1_seed2/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_no_l1_seed3/best_model/expert_pools.ptOther16.7 MB4361bd08d6c0
legacy/pythia410m/ablation_no_l1_seed3/best_model/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed3/best_model/gating_network.ptOther13.9 MB96411f40200e
legacy/pythia410m/ablation_no_l1_seed3/best_model/scheduler.ptOther1.7 KBd5482a786682
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1000/expert_pools.ptOther16.7 MB534128c446d2
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1000/gating_network.ptOther13.9 MB36ac998252fb
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1500/expert_pools.ptOther16.7 MB9a7019c83af1
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1500/gating_network.ptOther13.9 MB35e3d48902f9
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-2000/expert_pools.ptOther16.7 MBc49923bd6b59
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-2000/gating_network.ptOther13.9 MB05768095101f
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-500/expert_pools.ptOther16.7 MBdf5723b22b4c
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KBae84443ea0d8
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-500/gating_network.ptOther13.9 MB909c89304b19
legacy/pythia410m/ablation_no_l1_seed3/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_no_layer_emb_seed1/best_model/expert_pools.ptOther16.7 MB694bbee24bf8
legacy/pythia410m/ablation_no_layer_emb_seed1/best_model/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed1/best_model/gating_network.ptOther12.7 MBd366c1b1ec24
legacy/pythia410m/ablation_no_layer_emb_seed1/best_model/scheduler.ptOther1.7 KB34bd3772b8c1
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1000/expert_pools.ptOther16.7 MBfe42e0b483f4
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1000/gating_network.ptOther12.7 MB2eaaa8c4dce9
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1500/expert_pools.ptOther16.7 MBc6f7c0d0f828
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1500/gating_network.ptOther12.7 MB1bf453014827
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-2000/expert_pools.ptOther16.7 MBa13a26eaa550
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-2000/gating_network.ptOther12.7 MB405bf6a14bc4
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-500/expert_pools.ptOther16.7 MB41d752c8bf69
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-500/gating_network.ptOther12.7 MB76cd39662613
legacy/pythia410m/ablation_no_layer_emb_seed1/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_no_layer_emb_seed2/best_model/expert_pools.ptOther16.7 MBda05a4d72be6
legacy/pythia410m/ablation_no_layer_emb_seed2/best_model/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/best_model/gating_network.ptOther12.7 MB680c64a52045
legacy/pythia410m/ablation_no_layer_emb_seed2/best_model/scheduler.ptOther1.7 KBd5482a786682
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1000/expert_pools.ptOther16.7 MB40c59e7e8498
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1000/gating_network.ptOther12.7 MBbe5d48801384
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1500/expert_pools.ptOther16.7 MB04faac59e363
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1500/gating_network.ptOther12.7 MBed5410e6b4fb
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2000/expert_pools.ptOther16.7 MB175eb3acea44
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2000/gating_network.ptOther12.7 MBffa909741102
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2500/expert_pools.ptOther16.7 MB91b3e3e37615
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2500/gating_network.ptOther12.7 MBaa285995311e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3000/expert_pools.ptOther16.7 MB87c7dfd4e4aa
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3000/gating_network.ptOther12.7 MBf4ff55ef5a13
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3500/expert_pools.ptOther16.7 MBde89e0033fc8
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3500/gating_network.ptOther12.7 MB03521efb2d13
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-4000/expert_pools.ptOther16.7 MB29a24fda2144
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-4000/gating_network.ptOther12.7 MB4fc22ebba17b
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-500/expert_pools.ptOther16.7 MB850dc25612cc
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-500/gating_network.ptOther12.7 MB16d95a938332
legacy/pythia410m/ablation_no_layer_emb_seed2/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_no_layer_emb_seed3/best_model/expert_pools.ptOther16.7 MBdca82915ac40
legacy/pythia410m/ablation_no_layer_emb_seed3/best_model/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/best_model/gating_network.ptOther12.7 MB229e619918b2
legacy/pythia410m/ablation_no_layer_emb_seed3/best_model/scheduler.ptOther1.7 KB6ab07b9b4fa4
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1000/expert_pools.ptOther16.7 MBd7900c32af04
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1000/gating_network.ptOther12.7 MB1c154ce27beb
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1500/expert_pools.ptOther16.7 MB8038c9dce013
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1500/gating_network.ptOther12.7 MBab5318a70451
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2000/expert_pools.ptOther16.7 MBade9363006e3
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2000/gating_network.ptOther12.7 MB5f3a1a701ea6
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2500/expert_pools.ptOther16.7 MB877da46fadee
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2500/gating_network.ptOther12.7 MB0ab45a789af3
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3000/expert_pools.ptOther16.7 MB87a8311350d3
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3000/gating_network.ptOther12.7 MB647fce0c6cd7
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3500/expert_pools.ptOther16.7 MBe3e744d76290
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3500/gating_network.ptOther12.7 MBde16d6e4bdb0
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-4000/expert_pools.ptOther16.7 MBa79d8a7410df
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-4000/gating_network.ptOther12.7 MB88ee34a0ac2b
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-500/expert_pools.ptOther16.7 MB3e0638382110
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB395c5087422e
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-500/gating_network.ptOther12.7 MB8ea671cd1f25
legacy/pythia410m/ablation_no_layer_emb_seed3/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_partial_gating_seed1/best_model/expert_pools.ptOther16.7 MB0e154055076d
legacy/pythia410m/ablation_partial_gating_seed1/best_model/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed1/best_model/gating_network.ptOther13.9 MB98ceb718ccb6
legacy/pythia410m/ablation_partial_gating_seed1/best_model/scheduler.ptOther1.7 KBdb9980d2176e
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1000/expert_pools.ptOther16.7 MBadb2bb98360e
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1000/gating_network.ptOther13.9 MBb99a3ed46db2
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1500/expert_pools.ptOther16.7 MBb5cf3f3e14c7
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1500/gating_network.ptOther13.9 MBf56472cab3d8
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-2000/expert_pools.ptOther16.7 MBde848b2e50d8
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-2000/gating_network.ptOther13.9 MB2bc02340c9ba
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-500/expert_pools.ptOther16.7 MB62e228ed5e03
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-500/gating_network.ptOther13.9 MB59b4e8552847
legacy/pythia410m/ablation_partial_gating_seed1/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_partial_gating_seed2/best_model/expert_pools.ptOther16.7 MB5c31ad14fb3f
legacy/pythia410m/ablation_partial_gating_seed2/best_model/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed2/best_model/gating_network.ptOther13.9 MB51cac332f884
legacy/pythia410m/ablation_partial_gating_seed2/best_model/scheduler.ptOther1.7 KBbf912dda7335
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1000/expert_pools.ptOther16.7 MB0ac9aaa4ca9f
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1000/gating_network.ptOther13.9 MB1d01f25b2b8c
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1500/expert_pools.ptOther16.7 MB56caad3d58ac
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1500/gating_network.ptOther13.9 MBed0a60e8b7cf
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-2000/expert_pools.ptOther16.7 MB3a07a090e39f
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-2000/gating_network.ptOther13.9 MBe54b7764845a
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-500/expert_pools.ptOther16.7 MBe1e38b746448
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-500/gating_network.ptOther13.9 MB934cfdb5ce16
legacy/pythia410m/ablation_partial_gating_seed2/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/ablation_partial_gating_seed3/best_model/expert_pools.ptOther16.7 MB3ef3d9a0f872
legacy/pythia410m/ablation_partial_gating_seed3/best_model/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed3/best_model/gating_network.ptOther13.9 MBa7a84d22e21c
legacy/pythia410m/ablation_partial_gating_seed3/best_model/scheduler.ptOther1.7 KBbf912dda7335
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1000/expert_pools.ptOther16.7 MBc643ff455ea8
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1000/gating_network.ptOther13.9 MB5ca91288c8ef
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1500/expert_pools.ptOther16.7 MB90d64cf8d4ae
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1500/gating_network.ptOther13.9 MB9fa5d737a3de
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-2000/expert_pools.ptOther16.7 MBe38b7173c0ba
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-2000/gating_network.ptOther13.9 MB1f5427f3fa53
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-500/expert_pools.ptOther16.7 MB2d26d427d53e
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KBd936a70c58c4
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-500/gating_network.ptOther13.9 MB18e376836f38
legacy/pythia410m/ablation_partial_gating_seed3/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp2_gated_2experts_seed1/best_model/expert_pools.ptOther7.2 MB590ae5784fdd
legacy/pythia410m/exp2_gated_2experts_seed1/best_model/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/best_model/gating_network.ptOther13.8 MB2edb8e1457ba
legacy/pythia410m/exp2_gated_2experts_seed1/best_model/scheduler.ptOther1.7 KB0a420609c446
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1000/expert_pools.ptOther7.2 MB30a3ab8e6cb6
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1000/gating_network.ptOther13.8 MBbfed975e48e5
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1500/expert_pools.ptOther7.2 MB078b33d20cc7
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1500/gating_network.ptOther13.8 MBed5200ecd414
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2000/expert_pools.ptOther7.2 MB3cbced1b98a2
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2000/gating_network.ptOther13.8 MBfeb67888ed48
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2500/expert_pools.ptOther7.2 MB4c90c13515c4
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2500/gating_network.ptOther13.8 MB37242630512e
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3000/expert_pools.ptOther7.2 MB27c7682b08e1
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3000/gating_network.ptOther13.8 MBec25787b49a6
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3500/expert_pools.ptOther7.2 MB01aab98de7fc
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3500/gating_network.ptOther13.8 MB3970812f9215
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-4000/expert_pools.ptOther7.2 MB12e975217612
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-4000/gating_network.ptOther13.8 MB2dc9f71fc87c
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-500/expert_pools.ptOther7.2 MB6e80a7859513
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-500/gating_network.ptOther13.8 MB9e3d677d8825
legacy/pythia410m/exp2_gated_2experts_seed1/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp2_gated_2experts_seed1/final_model/expert_pools.ptOther7.2 MB7f773378af0e
legacy/pythia410m/exp2_gated_2experts_seed1/final_model/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed1/final_model/gating_network.ptOther13.8 MB6bb2c3a20b7a
legacy/pythia410m/exp2_gated_2experts_seed1/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_entropy_per_layer.pngOther43.1 KB6b9c19c1c676
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_layer_expert_heatmap.pngOther100.8 KB9a665160d085
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_task_expert_usage.pngOther38.5 KB0897104e3dca
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_1.pngOther89.6 KB0dcf64107cd5
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_2.pngOther89.4 KBd5aa02d9140d
legacy/pythia410m/exp2_gated_2experts_seed1/visualizations/exp2_gated_2experts_seed1_token_routing_3.pngOther89.8 KBa2007350a98e
legacy/pythia410m/exp2_gated_2experts_seed2/best_model/expert_pools.ptOther7.2 MBed5a51637ad4
legacy/pythia410m/exp2_gated_2experts_seed2/best_model/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/best_model/gating_network.ptOther13.8 MB3c2c80309d05
legacy/pythia410m/exp2_gated_2experts_seed2/best_model/scheduler.ptOther1.7 KBb51dfc15ac23
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1000/expert_pools.ptOther7.2 MB55b4e2994634
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1000/gating_network.ptOther13.8 MBd0ee2669c9b7
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1500/expert_pools.ptOther7.2 MBbe6b56d9c125
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1500/gating_network.ptOther13.8 MB0969776c7c35
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2000/expert_pools.ptOther7.2 MB969cd13faf42
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2000/gating_network.ptOther13.8 MB6cfe40d4c5cb
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2500/expert_pools.ptOther7.2 MB912e36501f16
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2500/gating_network.ptOther13.8 MB4545dae794f9
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3000/expert_pools.ptOther7.2 MB8d814f9c189a
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3000/gating_network.ptOther13.8 MBc25fac784337
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3500/expert_pools.ptOther7.2 MB35a41d12ae7e
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3500/gating_network.ptOther13.8 MBdfe0c10ae83f
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-4000/expert_pools.ptOther7.2 MBaacf178073d1
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-4000/gating_network.ptOther13.8 MB4f1e728a204d
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-500/expert_pools.ptOther7.2 MBf1181abc00cc
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-500/gating_network.ptOther13.8 MB089a3b5b6800
legacy/pythia410m/exp2_gated_2experts_seed2/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp2_gated_2experts_seed2/final_model/expert_pools.ptOther7.2 MB0af6c69b949b
legacy/pythia410m/exp2_gated_2experts_seed2/final_model/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed2/final_model/gating_network.ptOther13.8 MB4183133b2bb0
legacy/pythia410m/exp2_gated_2experts_seed2/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_entropy_per_layer.pngOther42.7 KB730498ef580d
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_layer_expert_heatmap.pngOther101.6 KBe6a1bb1f9d8e
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_task_expert_usage.pngOther38.1 KB797ee5900f1f
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_1.pngOther89.4 KB11b9e2bbd661
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_2.pngOther89.5 KB7c3581f79911
legacy/pythia410m/exp2_gated_2experts_seed2/visualizations/exp2_gated_2experts_seed2_token_routing_3.pngOther89.8 KB50aa85d21cba
legacy/pythia410m/exp2_gated_2experts_seed3/best_model/expert_pools.ptOther7.2 MBc2e95299fc09
legacy/pythia410m/exp2_gated_2experts_seed3/best_model/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/best_model/gating_network.ptOther13.8 MB47d40728146e
legacy/pythia410m/exp2_gated_2experts_seed3/best_model/scheduler.ptOther1.7 KBb51dfc15ac23
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1000/expert_pools.ptOther7.2 MB321cb5f68699
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1000/gating_network.ptOther13.8 MB6017a9603f2c
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1500/expert_pools.ptOther7.2 MBa6db2e89170a
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1500/gating_network.ptOther13.8 MB4ca00b5db13c
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2000/expert_pools.ptOther7.2 MBfb49601dfa34
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2000/gating_network.ptOther13.8 MB66d49e84d7bd
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2500/expert_pools.ptOther7.2 MB13af760a361e
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2500/gating_network.ptOther13.8 MB13ca6585644f
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3000/expert_pools.ptOther7.2 MB637e29684adb
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3000/gating_network.ptOther13.8 MBbfe72fac2475
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3500/expert_pools.ptOther7.2 MB2b72ca34818a
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3500/gating_network.ptOther13.8 MB60a663955c12
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-4000/expert_pools.ptOther7.2 MBeb8f17d920a6
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-4000/gating_network.ptOther13.8 MB164fbd72e34c
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-500/expert_pools.ptOther7.2 MBe73a5fddb8e7
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-500/gating_network.ptOther13.8 MB9e74d5792704
legacy/pythia410m/exp2_gated_2experts_seed3/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp2_gated_2experts_seed3/final_model/expert_pools.ptOther7.2 MB87fabfd6fc0e
legacy/pythia410m/exp2_gated_2experts_seed3/final_model/gated_lora_config.ptOther1.8 KBd3f393b0f70b
legacy/pythia410m/exp2_gated_2experts_seed3/final_model/gating_network.ptOther13.8 MB49f620d47633
legacy/pythia410m/exp2_gated_2experts_seed3/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_entropy_per_layer.pngOther43.4 KBc779251e8f66
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_layer_expert_heatmap.pngOther98.6 KBfa7d33f62f36
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_task_expert_usage.pngOther38.7 KBdfd1380e91e1
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_1.pngOther89.9 KB120c2493d10e
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_2.pngOther89.7 KB5cd6d99d2c2c
legacy/pythia410m/exp2_gated_2experts_seed3/visualizations/exp2_gated_2experts_seed3_token_routing_3.pngOther90.0 KBb45f99f0bf7a
legacy/pythia410m/exp3_gated_3experts_seed1/best_model/expert_pools.ptOther16.7 MB7da47ab32c4b
legacy/pythia410m/exp3_gated_3experts_seed1/best_model/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/best_model/gating_network.ptOther13.9 MB01a85f99e50f
legacy/pythia410m/exp3_gated_3experts_seed1/best_model/scheduler.ptOther1.7 KBb2e0d0974bd3
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1000/expert_pools.ptOther16.7 MBfa38fab5259f
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1000/gating_network.ptOther13.9 MB396a2451a3e2
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1500/expert_pools.ptOther16.7 MB4bd4dedc2d77
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1500/gating_network.ptOther13.9 MBc3b14daa5e73
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2000/expert_pools.ptOther16.7 MBd927a7ebc763
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2000/gating_network.ptOther13.9 MB7516ca544f39
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2500/expert_pools.ptOther16.7 MBd5f32f3acb60
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2500/gating_network.ptOther13.9 MB51eea3ce46a3
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3000/expert_pools.ptOther16.7 MB514821240383
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3000/gating_network.ptOther13.9 MBa01a7446f122
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3500/expert_pools.ptOther16.7 MB4cc5d0d5daa3
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3500/gating_network.ptOther13.9 MB638bf104ee22
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-4000/expert_pools.ptOther16.7 MBd7eec099670d
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-4000/gating_network.ptOther13.9 MB339f80d7dfc1
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-500/expert_pools.ptOther16.7 MB332c24cbeee5
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-500/gating_network.ptOther13.9 MB36e49ad6defb
legacy/pythia410m/exp3_gated_3experts_seed1/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp3_gated_3experts_seed1/final_model/expert_pools.ptOther16.7 MB7ef25eeb46f5
legacy/pythia410m/exp3_gated_3experts_seed1/final_model/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed1/final_model/gating_network.ptOther13.9 MBb79a326f7207
legacy/pythia410m/exp3_gated_3experts_seed1/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_entropy_per_layer.pngOther42.4 KB9fcaacd542be
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_layer_expert_heatmap.pngOther102.2 KBbed172d64f3c
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_task_expert_usage.pngOther38.1 KBbca526b927d1
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_1.pngOther91.7 KB66818850396f
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_2.pngOther91.9 KBde4d8eea262d
legacy/pythia410m/exp3_gated_3experts_seed1/visualizations/exp3_gated_3experts_seed1_token_routing_3.pngOther92.0 KB37371f81f229
legacy/pythia410m/exp3_gated_3experts_seed2/best_model/expert_pools.ptOther16.7 MB8d0db06522ca
legacy/pythia410m/exp3_gated_3experts_seed2/best_model/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/best_model/gating_network.ptOther13.9 MB57d899319d79
legacy/pythia410m/exp3_gated_3experts_seed2/best_model/scheduler.ptOther1.7 KBd5482a786682
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1000/expert_pools.ptOther16.7 MBfc201ffbdf0a
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1000/gating_network.ptOther13.9 MBced287d5e980
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1500/expert_pools.ptOther16.7 MBc5e0898a806d
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1500/gating_network.ptOther13.9 MBa11de3c1bd90
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2000/expert_pools.ptOther16.7 MB4d9e013392f1
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2000/gating_network.ptOther13.9 MB900b08cb6bf0
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2500/expert_pools.ptOther16.7 MB6c64a038d7ec
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2500/gating_network.ptOther13.9 MB894c6404861e
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3000/expert_pools.ptOther16.7 MB67856dc4b510
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3000/gating_network.ptOther13.9 MB6bbda91adabe
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3500/expert_pools.ptOther16.7 MBdbd6948faa9b
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3500/gating_network.ptOther13.9 MBdc105f725629
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-4000/expert_pools.ptOther16.7 MB46d55ae270af
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-4000/gating_network.ptOther13.9 MB4510982de385
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-500/expert_pools.ptOther16.7 MBe9a6a9fb6898
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-500/gating_network.ptOther13.9 MB982319ecf826
legacy/pythia410m/exp3_gated_3experts_seed2/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp3_gated_3experts_seed2/final_model/expert_pools.ptOther16.7 MBfe8be931926c
legacy/pythia410m/exp3_gated_3experts_seed2/final_model/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed2/final_model/gating_network.ptOther13.9 MB9b1f03c07c1f
legacy/pythia410m/exp3_gated_3experts_seed2/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_entropy_per_layer.pngOther42.7 KBcf2b28cfdf48
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_layer_expert_heatmap.pngOther99.6 KBf23cc61e65c8
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_task_expert_usage.pngOther39.2 KB5a2fd5022bd8
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_1.pngOther91.1 KB89f23b5e8c2b
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_2.pngOther91.3 KB68101f454d8b
legacy/pythia410m/exp3_gated_3experts_seed2/visualizations/exp3_gated_3experts_seed2_token_routing_3.pngOther91.1 KB67939ac2d9e7
legacy/pythia410m/exp3_gated_3experts_seed3/best_model/expert_pools.ptOther16.7 MB81de4419a01e
legacy/pythia410m/exp3_gated_3experts_seed3/best_model/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/best_model/gating_network.ptOther13.9 MB8ddc2d2e1daa
legacy/pythia410m/exp3_gated_3experts_seed3/best_model/scheduler.ptOther1.7 KB34bd3772b8c1
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1000/expert_pools.ptOther16.7 MB35fb0eaf2b48
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1000/gating_network.ptOther13.9 MB905f57585068
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1500/expert_pools.ptOther16.7 MBaa463ac8c245
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1500/gating_network.ptOther13.9 MB2c18956f5bad
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2000/expert_pools.ptOther16.7 MBc17beaff9862
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2000/gating_network.ptOther13.9 MB2092e080657b
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2500/expert_pools.ptOther16.7 MB90ea2b95b261
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2500/gating_network.ptOther13.9 MB5f79030d1fa1
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3000/expert_pools.ptOther16.7 MBaaec41b13056
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3000/gating_network.ptOther13.9 MBcdd839f3bc2e
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3500/expert_pools.ptOther16.7 MB811bb4172275
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3500/gating_network.ptOther13.9 MB1a1be92dfec6
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-4000/expert_pools.ptOther16.7 MB079d3955dd39
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-4000/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-4000/gating_network.ptOther13.9 MB7b40630226d0
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-500/expert_pools.ptOther16.7 MB36ea561a9b02
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-500/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-500/gating_network.ptOther13.9 MB375c5f49b81b
legacy/pythia410m/exp3_gated_3experts_seed3/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp3_gated_3experts_seed3/final_model/expert_pools.ptOther16.7 MBf72948f4bcf3
legacy/pythia410m/exp3_gated_3experts_seed3/final_model/gated_lora_config.ptOther1.8 KB10f76a249838
legacy/pythia410m/exp3_gated_3experts_seed3/final_model/gating_network.ptOther13.9 MBec8ad9165fa0
legacy/pythia410m/exp3_gated_3experts_seed3/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_entropy_per_layer.pngOther42.8 KBdeae83415ed6
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_layer_expert_heatmap.pngOther99.6 KB81cb933ce69c
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_task_expert_usage.pngOther38.4 KBf16d4167e8e7
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_1.pngOther91.5 KBb099bee41dd7
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_2.pngOther91.8 KB66da3e741b79
legacy/pythia410m/exp3_gated_3experts_seed3/visualizations/exp3_gated_3experts_seed3_token_routing_3.pngOther91.8 KB99388987d15f
legacy/pythia410m/exp4_gated_loadbalancing_seed1/best_model/expert_pools.ptOther16.7 MBb8442ef97eeb
legacy/pythia410m/exp4_gated_loadbalancing_seed1/best_model/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/best_model/gating_network.ptOther13.9 MB4d60f3e15dd1
legacy/pythia410m/exp4_gated_loadbalancing_seed1/best_model/scheduler.ptOther1.7 KBb2e0d0974bd3
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1000/expert_pools.ptOther16.7 MBc191bf65a814
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1000/gating_network.ptOther13.9 MB0715e4715738
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1500/expert_pools.ptOther16.7 MBf6eac1eb5132
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1500/gating_network.ptOther13.9 MBc45f20d6e176
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2000/expert_pools.ptOther16.7 MB50b7fa8be280
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2000/gating_network.ptOther13.9 MBa9a67ebe0ade
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2500/expert_pools.ptOther16.7 MB1b44d7a13f25
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2500/gating_network.ptOther13.9 MBb6d6439c70fe
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3000/expert_pools.ptOther16.7 MB409b4ef74186
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3000/gating_network.ptOther13.9 MB135b329fcba3
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3500/expert_pools.ptOther16.7 MB39943ddfb42e
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3500/gating_network.ptOther13.9 MBeda654b5e2e2
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-4000/expert_pools.ptOther16.7 MBb36c89d1d11f
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-4000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-4000/gating_network.ptOther13.9 MB3ea67b25a123
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-500/expert_pools.ptOther16.7 MBa6f5c0a366dc
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-500/gating_network.ptOther13.9 MB740ccef4f0eb
legacy/pythia410m/exp4_gated_loadbalancing_seed1/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp4_gated_loadbalancing_seed1/final_model/expert_pools.ptOther16.7 MB595ce4ad2588
legacy/pythia410m/exp4_gated_loadbalancing_seed1/final_model/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/final_model/gating_network.ptOther13.9 MB9713b2767107
legacy/pythia410m/exp4_gated_loadbalancing_seed1/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_entropy_per_layer.pngOther42.6 KBda4c17ec4950
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_layer_expert_heatmap.pngOther103.0 KB32107c170213
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_task_expert_usage.pngOther43.1 KBce3cb96aa6d8
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_1.pngOther91.7 KBf99c51f420b4
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_2.pngOther92.0 KBe6aff1ea2de1
legacy/pythia410m/exp4_gated_loadbalancing_seed1/visualizations/exp4_gated_loadbalancing_seed1_token_routing_3.pngOther92.0 KB1373a22497ee
legacy/pythia410m/exp4_gated_loadbalancing_seed2/best_model/expert_pools.ptOther16.7 MBb9b6d28e140f
legacy/pythia410m/exp4_gated_loadbalancing_seed2/best_model/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/best_model/gating_network.ptOther13.9 MB48220e22cfb5
legacy/pythia410m/exp4_gated_loadbalancing_seed2/best_model/scheduler.ptOther1.7 KBd5482a786682
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1000/expert_pools.ptOther16.7 MBc24832e64282
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1000/gating_network.ptOther13.9 MB0814dc8f3735
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1500/expert_pools.ptOther16.7 MBd757d85be3f3
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1500/gating_network.ptOther13.9 MBae762074250e
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-1500/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2000/expert_pools.ptOther16.7 MB47b558d21d69
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2000/gating_network.ptOther13.9 MBbe5a377a8d3d
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2000/scheduler.ptOther1.7 KB3b2455392db2
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2500/expert_pools.ptOther16.7 MBbb2a22b88d16
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2500/gating_network.ptOther13.9 MBb113f0a06c87
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-2500/scheduler.ptOther1.7 KB31bf47791394
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3000/expert_pools.ptOther16.7 MB305928d31ca9
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3000/gating_network.ptOther13.9 MB83f7805fae07
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3000/scheduler.ptOther1.7 KB502043cb8756
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3500/expert_pools.ptOther16.7 MB122a85464203
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3500/gating_network.ptOther13.9 MBaa829d57953c
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-3500/scheduler.ptOther1.7 KB2be695f86b21
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-4000/expert_pools.ptOther16.7 MBebda6d66b8e8
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-4000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-4000/gating_network.ptOther13.9 MB139659c4510a
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-4000/scheduler.ptOther1.7 KB673975514d3f
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-500/expert_pools.ptOther16.7 MB2ce9a09d1981
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-500/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-500/gating_network.ptOther13.9 MBcb619da1110a
legacy/pythia410m/exp4_gated_loadbalancing_seed2/checkpoint-500/scheduler.ptOther1.7 KB66c01bf5fb42
legacy/pythia410m/exp4_gated_loadbalancing_seed2/final_model/expert_pools.ptOther16.7 MBf2aa6a2a071a
legacy/pythia410m/exp4_gated_loadbalancing_seed2/final_model/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed2/final_model/gating_network.ptOther13.9 MB81deebee4d4e
legacy/pythia410m/exp4_gated_loadbalancing_seed2/final_model/scheduler.ptOther1.7 KB4b13ed54111b
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_entropy_per_layer.pngOther42.1 KB3b2f249e50c6
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_layer_expert_heatmap.pngOther103.0 KB1170691caf24
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_task_expert_usage.pngOther43.0 KBcedced771896
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_1.pngOther91.6 KB21a8ff423301
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_2.pngOther91.9 KBced767830165
legacy/pythia410m/exp4_gated_loadbalancing_seed2/visualizations/exp4_gated_loadbalancing_seed2_token_routing_3.pngOther92.0 KB1f043d999ce0
legacy/pythia410m/exp4_gated_loadbalancing_seed3/best_model/expert_pools.ptOther16.7 MBb6f3d6d5587f
legacy/pythia410m/exp4_gated_loadbalancing_seed3/best_model/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed3/best_model/gating_network.ptOther13.9 MB10722d6ac1d5
legacy/pythia410m/exp4_gated_loadbalancing_seed3/best_model/scheduler.ptOther1.7 KBab65e3db4ba5
legacy/pythia410m/exp4_gated_loadbalancing_seed3/checkpoint-1000/expert_pools.ptOther16.7 MBb051dfe69f7e
legacy/pythia410m/exp4_gated_loadbalancing_seed3/checkpoint-1000/gated_lora_config.ptOther1.8 KBef9b75d6d4e4
legacy/pythia410m/exp4_gated_loadbalancing_seed3/checkpoint-1000/gating_network.ptOther13.9 MBe33eea659e71
legacy/pythia410m/exp4_gated_loadbalancing_seed3/checkpoint-1000/scheduler.ptOther1.7 KB0aea463251e0
legacy/pythia410m/exp4_gated_loadbalancing_seed3/checkpoint-1500/expert_pools.ptOther16.7 MBb6f3d6d5587f
.gitattributesRepository2.8 KB—
legacy/llama32/exp5_gated_topk_seed1/checkpoint-3500/.expert_pools.pt.O3Zbf8Repository30.9 MB8008cb13e625

License and Download

License
Not stated by the source
Access
No access gate
Download from Helain ZIMMERMANN

Released by Helain ZIMMERMANN through its official repository on Hugging Face.