🌃 Vespera Synapse 31B
This is a merge of pre-trained language models created using mergekit.
The following patch was required for this merge
# ── 11. Model Stock t factor with Adaptive Soft-Clamp ─────────────
N = len(ws_2d)
ct = cos_theta.unsqueeze(-1) if cos_theta.dim() > 0 else cos_theta
# Raw Model Stock formula
denom = 1.0 + (N - 1) * ct
# Add a tiny epsilon to prevent literal division by zero
t_raw = (N * ct) / denom.clamp(min=1e-6)
# --- BULLETPROOF TANH CLAMP ---
# 1. Prevent negative infinity spikes (fallback to base model)
t_clamped_bottom = torch.clamp(t_raw, min=0.0)
# 2. Smoothly asymptote positive spikes to L (Maximum allowed t-factor)
L = 1.5
excess = torch.clamp(t_clamped_bottom - 1.0, min=0.0)
t_soft_top = 1.0 + (L - 1.0) * torch.tanh(excess / (L - 1.0))
# 3. Apply: If t <= 1.0, use exact math. If t > 1.0, use soft curve.
t = torch.where(t_clamped_bottom <= 1.0, t_clamped_bottom, t_soft_top)
# ------------------------------
Merge Audit
Merge Details
Merge Method
This model was merged using the karcher_stock merge method using google/gemma-4-31B as a base.
Models Merged
The following models were included in the merge:
- google/gemma-4-31B
- Vortex5/Glimmering-Citrus-31B
- Vortex5/Scarlet-Shadow-31B
- Cyclone-Labs/Twisted-Cyclone-31B
Configuration
The following YAML configuration was used to produce this model:
architecture: Gemma4ForConditionalGeneration
base_model: /workspace/models/google--gemma-4-31B # densenet--Gemma-4-31B-StyleTune-heretic-ara
models:
- model: /workspace/models/Vortex5--Glimmering-Citrus-31B
- model: /workspace/models/Vortex5--Scarlet-Shadow-31B
- model: /workspace/models/Cyclone-Labs--Twisted-Cyclone-31B
merge_method: karcher_stock # v37
parameters:
filter_wise: true
max_iter: 1000
min_iter: 100
tol: 1.0e-9
dtype: float32
out_dtype: bfloat16
tokenizer:
source: union
chat_template: auto
name: 🌃 Vespera Synapse 31B