Dedicated Endpoints
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Container
Run this model inference with full control and performance in your environment.
Get help setting up a custom Dedicated Endpoints.
Talk with our engineer to get a quote for reserved GPU instances with discounts.
README
License: apache-2.0Abliteration parameters
| Parameter | Value |
|---|---|
| direction_index | 43.15 |
| attn.o_proj.max_weight | 1.48 |
| attn.o_proj.max_weight_position | 59.65 |
| attn.o_proj.min_weight | 1.44 |
| attn.o_proj.min_weight_distance | 48.02 |
| mlp.down_proj.max_weight | 1.21 |
| mlp.down_proj.max_weight_position | 54.75 |
| mlp.down_proj.min_weight | 0.30 |
| mlp.down_proj.min_weight_distance | 50.44 |
Performance
| Metric | This model | Original model (mistralai/Mistral-Medium-3.5-128B) |
|---|---|---|
| KL divergence | 0.0220 | 0 (by definition) |
| Refusals | 9/100 | 98/100 |
Thanks to Darkhn for providing the Text Only BF16 base for this tune
Model provider
spinochenza
Model tree
Base
mistralai/Mistral-Medium-3.5-128B
Fine-tuned
this model
Modalities
Input
Text
Output
Text
Pricing
Dedicated Endpoints
View detailsSupported Functionality
Model APIs
Dedicated Endpoints
Container
More information