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
Commit History
6 days ago
-
Model card: general-capability retention check reported
Verified signature837f739
8 days ago
-
Card: which file is which; licence beside 'open'; organisation links
Verified signature8846b95 -
Card: new repository name
Verified signature131342c -
Card: link the article
Verified signature24200ee -
Card: live demo on the Hugging Face Space
Verified signature83b3543 -
Card: Claude Haiku 5.5 baseline, and what the model offers instead
Verified signature5013d5c -
Card: the final system, with its disclosure and its own comparison with the base model
Verified signature00fd2c2 -
Version 3 adapter at the top level (the final system's reasoner); the same file as exploratory/v3_adapter.gguf
Verified signaturecac55c3 -
Card: the final system
Verified signature66d5a39 -
Model card: company-clustered intervals, reached-only comparison, judge level
Verified signaturef0f2942 -
Model card: decoding-time guards (exploratory, the app default), audit fixes, lowercase brand
Verified signaturedbe1a2e
9 days ago
-
Model card metadata restored (licence, base model, tags)
Verified signature6d896aa -
Model card: dig then reason on the final test, both judges
Verified signaturea5dd537
10 days ago
-
Model card: exploratory version 3 adapter; hindsight figure given with its denominator
Verified signature6160b58 -
Exploratory version 3 adapter (GGUF), for the two-adapter system in the write-up
Verified signaturee6ce793 -
Model card: second frozen test with both judges, live demo and write-up links, current code paths
Verified signature7e647de -
The agent's name: Aemon. Tagline: Weigh what is known. Test what is not.
Verified signaturee663e24
13 days ago
-
Model card: fuller account of what the results do not show
Verified signature0a5616a -
Model card: remove the commercial decision-making line
Verified signaturec3a9c10 -
Model card: plain-text rank, alpha and learning rate (inline formulas do not render)
Verified signature870f634 -
Fine-tuned adapter (LoRA and GGUF), tokenizer and model card
Verified signature23ae6fe -
initial commit
Verified signature6584ef2