Instructions to use AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT with Transformers:

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

pipe = pipeline("text-generation", model="AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT

SGLang

How to use AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT 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 "AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT" \
    --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": "AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT",
		"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 "AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT" \
        --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": "AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT with Docker Model Runner:
```
docker model run hf.co/AETHORIA-AI/TR-HASH-MoE-200M-160B-SFT
```

TR-HASH-MoE-200M-160B-SFT

Commit History

Apply CC BY-NC to model repository

0c6d56f

Boris Peyriguere commited on 2 days ago

Add preprint DOI and weights license

c7bdee8

Boris Peyriguere commited on 2 days ago

Upload combined_arc_model_comparison.png

85aad9b
verified

Pacific-i64 commited on 4 days ago

Upload combined_arc_model_comparison.png

eaea6a1
verified

Pacific-i64 commited on 4 days ago

Remove legacy v1 tokenizer label from SFT card

b7eeabf
verified

Pacific-i64 commited on 5 days ago

Point SFT card to canonical 32K tokenizer

5c08ac6
verified

Pacific-i64 commited on 5 days ago

Link canonical 32K tokenizer v1

e4ce55e
verified

Pacific-i64 commited on 6 days ago

Document 32k SFT tokenizer compatibility boundary

047e290
verified

Pacific-i64 commited on 7 days ago

Add extended Combined ARC comparison

47bbd6c
verified

Pacific-i64 commited on 8 days ago

Add compact-model benchmark comparison

465c688
verified

Pacific-i64 commited on 8 days ago

Add ARC-Challenge and Combined ARC results

576c0c4
verified

Pacific-i64 commited on 8 days ago

Add ARC-Easy and HellaSwag benchmark results

8ea9cfb
verified

Pacific-i64 commited on 8 days ago

Remove forced thinking prefill

cef6641
verified

Pacific-i64 commited on 9 days ago

Prefill think tag for assistant generation

e5a599f
verified

Pacific-i64 commited on 9 days ago

Promote clean full-SFT v2 epoch 3 step 5982

73f4d01
verified

Pacific-i64 commited on 9 days ago

Promote clean full-SFT v2 epoch 3 step 5982

1ce61f5
verified

Pacific-i64 commited on 9 days ago

Clean repository: retain BF16 epoch 3 release and compact reports

d0c86f5
verified

Pacific-i64 commited on 9 days ago

Flatten final epoch 3 generation report

d34371f
verified

Pacific-i64 commited on 9 days ago

Flatten final epoch 3 behavior report

3c330ce
verified

Pacific-i64 commited on 9 days ago

Flatten final epoch 3 evaluation reports

ff7a984
verified

Pacific-i64 commited on 9 days ago

Promote clean full-SFT v2 epoch 3 step 5982

7a84f1b
verified

Pacific-i64 commited on 9 days ago

Add epoch 3 MLX FP16 evaluation export

fce178a
verified

Pacific-i64 commited on 9 days ago

Upload folder using huggingface_hub

122d80e
verified

Pacific-i64 commited on 9 days ago

Upload SFT v2 promotion panel

0ad518d
verified

Pacific-i64 commited on 9 days ago

Upload SFT v2 training metrics

fc4117b
verified

Pacific-i64 commited on 9 days ago

Upload SFT v2 per-epoch PIQA and regression evaluations

7280566
verified

Pacific-i64 commited on 9 days ago

Upload folder using huggingface_hub

191328e
verified

Pacific-i64 commited on 9 days ago

Add epoch 2 MLX FP16 bundle for PIQA evaluation

5b9523f
verified

Pacific-i64 commited on 9 days ago

Add epoch 1 MLX FP16 bundle for PIQA evaluation

3bc4154
verified

Pacific-i64 commited on 9 days ago

Upload folder using huggingface_hub

74704f6
verified

Pacific-i64 commited on 9 days ago

Record clean SFT v2 preflight

be36c9b
verified

Pacific-i64 commited on 9 days ago

Declare TR-HASH model type and architecture metadata

6bd65d2
verified

Pacific-i64 commited on 9 days ago

Relicense model release under Apache 2.0

522f465
verified

Pacific-i64 commited on 9 days ago

Promote PIQA-best full SFT epoch 2 to root SafeTensors

e434620
verified

Pacific-i64 commited on 9 days ago

Upload folder using huggingface_hub

b3f6655
verified

Pacific-i64 commited on 9 days ago

Upload 3-epoch full SFT training metrics

f9444cd
verified

Pacific-i64 commited on 9 days ago

Upload folder using huggingface_hub

38467cd
verified

Pacific-i64 commited on 9 days ago

Upload folder using huggingface_hub

17c7a1a
verified

Pacific-i64 commited on 9 days ago

Add full-parameter SFT phase model card

b19414e
verified

Pacific-i64 commited on 9 days ago

initial commit

ec3548c
verified

Pacific-i64 commited on 9 days ago

Commit History

Apply CC BY-NC to model repository 0c6d56f

Add preprint DOI and weights license c7bdee8

Upload combined_arc_model_comparison.png 85aad9b verified

Upload combined_arc_model_comparison.png eaea6a1 verified

Remove legacy v1 tokenizer label from SFT card b7eeabf verified

Point SFT card to canonical 32K tokenizer 5c08ac6 verified

Link canonical 32K tokenizer v1 e4ce55e verified

Document 32k SFT tokenizer compatibility boundary 047e290 verified

Add extended Combined ARC comparison 47bbd6c verified

Add compact-model benchmark comparison 465c688 verified

Add ARC-Challenge and Combined ARC results 576c0c4 verified

Add ARC-Easy and HellaSwag benchmark results 8ea9cfb verified

Remove forced thinking prefill cef6641 verified

Prefill think tag for assistant generation e5a599f verified

Promote clean full-SFT v2 epoch 3 step 5982 73f4d01 verified

Promote clean full-SFT v2 epoch 3 step 5982 1ce61f5 verified

Clean repository: retain BF16 epoch 3 release and compact reports d0c86f5 verified

Flatten final epoch 3 generation report d34371f verified

Flatten final epoch 3 behavior report 3c330ce verified

Flatten final epoch 3 evaluation reports ff7a984 verified

Promote clean full-SFT v2 epoch 3 step 5982 7a84f1b verified

Add epoch 3 MLX FP16 evaluation export fce178a verified

Upload folder using huggingface_hub 122d80e verified

Upload SFT v2 promotion panel 0ad518d verified

Upload SFT v2 training metrics fc4117b verified

Upload SFT v2 per-epoch PIQA and regression evaluations 7280566 verified

Upload folder using huggingface_hub 191328e verified

Add epoch 2 MLX FP16 bundle for PIQA evaluation 5b9523f verified

Add epoch 1 MLX FP16 bundle for PIQA evaluation 3bc4154 verified

Upload folder using huggingface_hub 74704f6 verified

Record clean SFT v2 preflight be36c9b verified

Declare TR-HASH model type and architecture metadata 6bd65d2 verified

Relicense model release under Apache 2.0 522f465 verified

Promote PIQA-best full SFT epoch 2 to root SafeTensors e434620 verified

Upload folder using huggingface_hub b3f6655 verified

Upload 3-epoch full SFT training metrics f9444cd verified

Upload folder using huggingface_hub 38467cd verified

Upload folder using huggingface_hub 17c7a1a verified

Add full-parameter SFT phase model card b19414e verified

initial commit ec3548c verified

Apply CC BY-NC to model repository

0c6d56f

Add preprint DOI and weights license

c7bdee8

Upload combined_arc_model_comparison.png

85aad9b
verified

Upload combined_arc_model_comparison.png

eaea6a1
verified

Remove legacy v1 tokenizer label from SFT card

b7eeabf
verified

Point SFT card to canonical 32K tokenizer

5c08ac6
verified

Link canonical 32K tokenizer v1

e4ce55e
verified

Document 32k SFT tokenizer compatibility boundary

047e290
verified

Add extended Combined ARC comparison

47bbd6c
verified

Add compact-model benchmark comparison

465c688
verified

Add ARC-Challenge and Combined ARC results

576c0c4
verified

Add ARC-Easy and HellaSwag benchmark results

8ea9cfb
verified

Remove forced thinking prefill

cef6641
verified

Prefill think tag for assistant generation

e5a599f
verified

Promote clean full-SFT v2 epoch 3 step 5982

73f4d01
verified

Promote clean full-SFT v2 epoch 3 step 5982

1ce61f5
verified

Clean repository: retain BF16 epoch 3 release and compact reports

d0c86f5
verified

Flatten final epoch 3 generation report

d34371f
verified

Flatten final epoch 3 behavior report

3c330ce
verified

Flatten final epoch 3 evaluation reports

ff7a984
verified

Promote clean full-SFT v2 epoch 3 step 5982

7a84f1b
verified

Add epoch 3 MLX FP16 evaluation export

fce178a
verified

Upload folder using huggingface_hub

122d80e
verified

Upload SFT v2 promotion panel

0ad518d
verified

Upload SFT v2 training metrics

fc4117b
verified

Upload SFT v2 per-epoch PIQA and regression evaluations

7280566
verified

Upload folder using huggingface_hub

191328e
verified

Add epoch 2 MLX FP16 bundle for PIQA evaluation

5b9523f
verified

Add epoch 1 MLX FP16 bundle for PIQA evaluation

3bc4154
verified

Upload folder using huggingface_hub

74704f6
verified

Record clean SFT v2 preflight

be36c9b
verified

Declare TR-HASH model type and architecture metadata

6bd65d2
verified

Relicense model release under Apache 2.0

522f465
verified

Promote PIQA-best full SFT epoch 2 to root SafeTensors

e434620
verified

Upload folder using huggingface_hub

b3f6655
verified

Upload 3-epoch full SFT training metrics

f9444cd
verified

Upload folder using huggingface_hub

38467cd
verified

Upload folder using huggingface_hub

17c7a1a
verified

Add full-parameter SFT phase model card

b19414e
verified

initial commit

ec3548c
verified