Instructions to use enosislabs/AETHER-Mythos-1-1.2B-adapter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use enosislabs/AETHER-Mythos-1-1.2B-adapter with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="enosislabs/AETHER-Mythos-1-1.2B-adapter")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("enosislabs/AETHER-Mythos-1-1.2B-adapter", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use enosislabs/AETHER-Mythos-1-1.2B-adapter with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "enosislabs/AETHER-Mythos-1-1.2B-adapter"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "enosislabs/AETHER-Mythos-1-1.2B-adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/enosislabs/AETHER-Mythos-1-1.2B-adapter

SGLang

How to use enosislabs/AETHER-Mythos-1-1.2B-adapter with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "enosislabs/AETHER-Mythos-1-1.2B-adapter" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "enosislabs/AETHER-Mythos-1-1.2B-adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "enosislabs/AETHER-Mythos-1-1.2B-adapter" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "enosislabs/AETHER-Mythos-1-1.2B-adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio

How to use enosislabs/AETHER-Mythos-1-1.2B-adapter with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for enosislabs/AETHER-Mythos-1-1.2B-adapter to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for enosislabs/AETHER-Mythos-1-1.2B-adapter to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for enosislabs/AETHER-Mythos-1-1.2B-adapter to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="enosislabs/AETHER-Mythos-1-1.2B-adapter",
    max_seq_length=2048,
)

Docker Model Runner
How to use enosislabs/AETHER-Mythos-1-1.2B-adapter with Docker Model Runner:
```
docker model run hf.co/enosislabs/AETHER-Mythos-1-1.2B-adapter
```

ferxalb commited on 18 days ago

Commit

986af8e

verified ·

1 Parent(s): 1c13952

Upload AETHER-Mythos-1 LoRA adapters

Browse files

Files changed (4) hide show

README.md +9 -9
adapter_config.json +3 -12
adapter_model.safetensors +2 -2
tokenizer_config.json +0 -0

README.md CHANGED Viewed

@@ -43,15 +43,15 @@ into an efficient backbone. We prioritize:
 ## Base model
 - **Base**: [`LiquidAI/LFM2.5-1.2B-Thinking`](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking)
-- **Context trained**: up to 8192 tokens (packing enabled)
-- **Method**: LoRA (r=64, alpha=128)
 ## Data mixture
-- **fable5_cot** (`Glint-Research/Fable-5-traces`) weight=0.65 — Primary high-signal Fable 5 CoT + tools (AGPL-3.0)
-- **opencode_reasoning** (`nvidia/OpenCodeReasoning`) weight=0.18 — Coding CoT already uses <think> — aligns with LFM Thinking
-- **open_r1_codeforces** (`open-r1/codeforces-cots`) weight=0.10 — messages
-- **openthoughts_code** (`open-thoughts/OpenThoughts-114k`) weight=0.07 — messages
 ### Data provenance & licenses
@@ -72,10 +72,10 @@ under Liquid AI’s LFM license terms.
 | Setting | Value |
 |--------|--------|
 | GPU | L40S (Modal) |
-| Effective batch size | 32 |
-| Learning rate | 0.0001 |
 | Schedule | cosine |
-| Epochs / max steps | 1.0 / 600 |
 | Packing | True |
 | Optim | adamw_8bit |
 | Grad checkpointing | unsloth |

 ## Base model
 - **Base**: [`LiquidAI/LFM2.5-1.2B-Thinking`](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking)
+- **Context trained**: up to 32768 tokens (packing enabled)
+- **Method**: LoRA (r=128, alpha=256)
 ## Data mixture
+- **fable5_cot** (`Glint-Research/Fable-5-traces`) weight=0.65 — Primary identity and agent trace signal
+- **opencode_reasoning** (`nvidia/OpenCodeReasoning`) weight=0.18 — prompt_completion
+- **open_r1_codeforces** (`open-r1/codeforces-cots`) weight=0.07 — messages
+- **openthoughts_code** (`open-thoughts/OpenThoughts-114k`) weight=0.10 — messages
 ### Data provenance & licenses
 | Setting | Value |
 |--------|--------|
 | GPU | L40S (Modal) |
+| Effective batch size | 16 |
+| Learning rate | 8e-05 |
 | Schedule | cosine |
+| Epochs / max steps | 1.0 / 100 |
 | Packing | True |
 | Optim | adamw_8bit |
 | Grad checkpointing | unsloth |

adapter_config.json CHANGED Viewed

@@ -20,7 +20,7 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 64,
   "lora_bias": false,
   "lora_dropout": 0.0,
   "lora_ga_config": null,
@@ -30,19 +30,10 @@
   "peft_type": "LORA",
   "peft_version": "0.19.1",
   "qalora_group_size": 16,
-  "r": 32,
   "rank_pattern": {},
   "revision": null,
-  "target_modules": [
-    "in_proj",
-    "out_proj",
-    "w2",
-    "v_proj",
-    "w1",
-    "w3",
-    "k_proj",
-    "q_proj"
-  ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
+  "lora_alpha": 256,
   "lora_bias": false,
   "lora_dropout": 0.0,
   "lora_ga_config": null,
   "peft_type": "LORA",
   "peft_version": "0.19.1",
   "qalora_group_size": 16,
+  "r": 128,
   "rank_pattern": {},
   "revision": null,
+  "target_modules": "(?:.*?(?:language|text).*?(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer).*?(?:q_proj|k_proj|v_proj|out_proj|in_proj|w1|w2|w3))|(?:\\bmodel\\.layers\\.[\\d]{1,}\\.(?:self_attn|attention|attn|mixer|mlp|feed_forward|ffn|dense|mixer)\\.(?:(?:q_proj|k_proj|v_proj|out_proj|in_proj|w1|w2|w3)))",
   "target_parameters": null,
   "task_type": "CAUSAL_LM",
   "trainable_token_indices": null,

adapter_model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd6328b5b6b15d7d2b4c0e67695a7179ab732756e1288bc359b2da268f1f8487
-size 88891320

 version https://git-lfs.github.com/spec/v1
+oid sha256:488ab92d094984063fdce7f8aa308d9316c111092c1d5d1715c5fc0ec3043b49
+size 292572344

tokenizer_config.json CHANGED Viewed

The diff for this file is too large to render. See raw diff