Update README.md
Browse files
README.md
CHANGED
|
@@ -43,6 +43,16 @@ of anchored attenuation rules.
|
|
| 43 |
Theoretically this not only solved rope, but it pushed the context window so far out that it's going to become a python problem for
|
| 44 |
moving data into the machine, rather than a vram or a gpu computational problem for processing it.
|
| 45 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
Combined with the potency hybrid attention, or no attention which is what I'll use, hybrid attention will enable Ulysses, which means...
|
| 47 |
well... multi-machine optimization.
|
| 48 |
|
|
|
|
| 43 |
Theoretically this not only solved rope, but it pushed the context window so far out that it's going to become a python problem for
|
| 44 |
moving data into the machine, rather than a vram or a gpu computational problem for processing it.
|
| 45 |
|
| 46 |
+
Below is a rope component synthesizer that produces many forms of rope that I specifically created for bulk ensemble experimentation.
|
| 47 |
+
|
| 48 |
+
https://github.com/AbstractEyes/geofractal/blob/main/src/geofractal/router/components/rope_component.py
|
| 49 |
+
|
| 50 |
+
Below is a transformer component built specifically for high-yield reuse.
|
| 51 |
+
|
| 52 |
+
https://github.com/AbstractEyes/geofractal/blob/main/src/geofractal/router/components/transformer_component.py
|
| 53 |
+
|
| 54 |
+
Those are viable ropes within their spectrums but they don't conform to all utility, not like the potential from the geolip concept.
|
| 55 |
+
|
| 56 |
Combined with the potency hybrid attention, or no attention which is what I'll use, hybrid attention will enable Ulysses, which means...
|
| 57 |
well... multi-machine optimization.
|
| 58 |
|