Text Generation
PEFT
Safetensors
GGUF
lora
agentic
tool-use
distillation
agenttune
auditkit
conversational
Instructions to use two-eyed-ravens/aemon with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use two-eyed-ravens/aemon with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("/home/CLUSTER_ACCOUNT/lexsi/models/tiny-aya-global") model = PeftModel.from_pretrained(base_model, "two-eyed-ravens/aemon") - Notebooks
- Google Colab
- Kaggle
Card: the final system, with its disclosure and its own comparison with the base model
Browse files
README.md
CHANGED
|
@@ -226,10 +226,11 @@ a stronger or human-corrected teacher for evidence judgement, is the open direct
|
|
| 226 |
- **Exact reproduction from the public data.** Since 11 November the public dossier is byte-identical to the one the
|
| 227 |
evaluated runs read (the release of 8 November had cut excerpts to one short quote).
|
| 228 |
|
| 229 |
-
## The final system: the version 3 adapter (`
|
| 230 |
|
| 231 |
A second adapter, continued from this one on rows from a second teacher (DeepSeek V4-Pro; 3,611 rows from 144
|
| 232 |
-
replay-exact investigations). On its own it did not pass its development adoption rule;
|
|
|
|
| 233 |
published so the two-adapter system described in the write-up ("dig then reason": this card's adapter gathers the
|
| 234 |
evidence, version 3 reasons) can be run by others. File: GGUF for llama.cpp, sha256 `ea919380…e1443`, the conversion
|
| 235 |
the final system's runs used. Blind scores, mean of two judges (Claude Code and Codex `gpt-6-sol`), base / version 1 /
|
|
@@ -238,7 +239,10 @@ version 3: test set 2 0.56 / 1.06 / 1.48; final test 0.62 / 1.46 / 1.33. Dig the
|
|
| 238 |
## The final system: decoding-time guards (the app's default)
|
| 239 |
|
| 240 |
They change what the model can produce at each step, not its weights; the version 1 numbers above use none of
|
| 241 |
-
them.
|
|
|
|
|
|
|
|
|
|
| 242 |
- **Loop guards**: a source already weighed for a theory and a search already run are not offered again, a quote is
|
| 243 |
picked from the source's own sentences, an exact repeat is drawn again before it counts, and a step-budget guard.
|
| 244 |
Finished runs: development 28 → 45 of 45, test set 2 43 → 69 of 72, final test 27 → 45 of 48. Blind quality, mean of
|
|
|
|
| 226 |
- **Exact reproduction from the public data.** Since 11 November the public dossier is byte-identical to the one the
|
| 227 |
evaluated runs read (the release of 8 November had cut excerpts to one short quote).
|
| 228 |
|
| 229 |
+
## The final system: the version 3 adapter (`v3_adapter.gguf`)
|
| 230 |
|
| 231 |
A second adapter, continued from this one on rows from a second teacher (DeepSeek V4-Pro; 3,611 rows from 144
|
| 232 |
+
replay-exact investigations). On its own it did not pass its development adoption rule; as the reasoner of dig then reason it did (+0.17 [0.02,
|
| 233 |
+
0.35] over version 1 on development), and that is how the final system uses it. It is
|
| 234 |
published so the two-adapter system described in the write-up ("dig then reason": this card's adapter gathers the
|
| 235 |
evidence, version 3 reasons) can be run by others. File: GGUF for llama.cpp, sha256 `ea919380…e1443`, the conversion
|
| 236 |
the final system's runs used. Blind scores, mean of two judges (Claude Code and Codex `gpt-6-sol`), base / version 1 /
|
|
|
|
| 239 |
## The final system: decoding-time guards (the app's default)
|
| 240 |
|
| 241 |
They change what the model can produce at each step, not its weights; the version 1 numbers above use none of
|
| 242 |
+
them. They were built after both frozen tests, partly from the failures in their runs (the loop guards from the
|
| 243 |
+
refused calls of the final-test runs), and measured on the same cases, so version 1's numbers stay the run-once
|
| 244 |
+
result. Against the base model, the final system scores +1.02 [0.92, 1.10] on the first test's cases and +1.11
|
| 245 |
+
[1.01, 1.20] on the second's, judged by the same two judges in separate batches. With the two-adapter system ("dig then reason"):
|
| 246 |
- **Loop guards**: a source already weighed for a theory and a search already run are not offered again, a quote is
|
| 247 |
picked from the source's own sentences, an exact repeat is drawn again before it counts, and a step-budget guard.
|
| 248 |
Finished runs: development 28 → 45 of 45, test set 2 43 → 69 of 72, final test 27 → 45 of 48. Blind quality, mean of
|