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
```

Pacific-i64 commited on 8 days ago

Commit

576c0c4

verified ·

1 Parent(s): 8ea9cfb

Add ARC-Challenge and Combined ARC results

Browse files

Files changed (4) hide show

README.md +6 -2
release_manifest.json +15 -0
reports/sft-v2-300k/arc_challenge_epoch3.json +14 -0
reports/sft-v2-300k/arc_combined_epoch3.json +14 -0

README.md CHANGED Viewed

@@ -44,14 +44,18 @@ and lowest held-out SFT loss across the three epochs.
 |---|---|---:|---:|---:|---|
 | PIQA | validation | 1,838 | **68.01%** | **69.10%** | PyTorch FP16, custom Triton |
 | ARC-Easy | test | 2,376 | **57.24%** | **52.86%** | MLX FP16 |
 | HellaSwag | validation | 10,042 | **33.21%** | **38.74%** | MLX FP16 |
-All three benchmarks use zero-shot causal continuation log-likelihood, no
 chat template and a maximum sequence length of 2,048. Accuracy selects the
 choice with the highest total continuation log-likelihood; the normalized
 metric selects by mean continuation log-likelihood per scored token. PIQA was
-evaluated from the native epoch-3 checkpoint. ARC-Easy and HellaSwag were
 evaluated from an FP16 MLX conversion of the same root F32 SafeTensors weights.
 Machine-readable reports are published under `reports/sft-v2-300k/`.
 ## Training recipe

 |---|---|---:|---:|---:|---|
 | PIQA | validation | 1,838 | **68.01%** | **69.10%** | PyTorch FP16, custom Triton |
 | ARC-Easy | test | 2,376 | **57.24%** | **52.86%** | MLX FP16 |
+| ARC-Challenge | test | 1,172 | **27.13%** | **29.01%** | MLX FP16 |
+| **ARC Combined (micro)** | test | **3,548** | **47.29%** | **44.98%** | Derived from both ARC test splits |
 | HellaSwag | validation | 10,042 | **33.21%** | **38.74%** | MLX FP16 |
+All benchmark evaluations use zero-shot causal continuation log-likelihood, no
 chat template and a maximum sequence length of 2,048. Accuracy selects the
 choice with the highest total continuation log-likelihood; the normalized
 metric selects by mean continuation log-likelihood per scored token. PIQA was
+evaluated from the native epoch-3 checkpoint. ARC and HellaSwag were
 evaluated from an FP16 MLX conversion of the same root F32 SafeTensors weights.
+ARC Combined is the micro-average over all 3,548 ARC-Easy and ARC-Challenge
+test examples, not the arithmetic mean of the two percentages.
 Machine-readable reports are published under `reports/sft-v2-300k/`.
 ## Training recipe

release_manifest.json CHANGED Viewed

@@ -1,6 +1,21 @@
 {
   "architecture": "tr_hash_moe",
   "benchmarks": {
     "arc_easy": {
       "acc": 0.5723905723905723,
       "acc_norm": 0.5286195286195287,

 {
   "architecture": "tr_hash_moe",
   "benchmarks": {
+    "arc_challenge": {
+      "acc": 0.2713310580204778,
+      "acc_norm": 0.2901023890784983,
+      "backend": "mlx-fp16",
+      "examples": 1172,
+      "split": "test"
+    },
+    "arc_combined": {
+      "acc": 0.4729425028184893,
+      "acc_norm": 0.44983089064261555,
+      "aggregation": "micro-average over ARC-Easy and ARC-Challenge",
+      "backend": "derived",
+      "examples": 3548,
+      "split": "test"
+    },
     "arc_easy": {
       "acc": 0.5723905723905723,
       "acc_norm": 0.5286195286195287,

reports/sft-v2-300k/arc_challenge_epoch3.json ADDED Viewed

	@@ -0,0 +1,14 @@

+{
+  "benchmark": "arc_challenge",
+  "split": "test",
+  "examples": 1172,
+  "correct": 318,
+  "correct_norm": 340,
+  "acc": 0.2713310580204778,
+  "acc_norm": 0.2901023890784983,
+  "backend": "mlx-fp16",
+  "zero_shot": true,
+  "chat_template_applied": false,
+  "max_length": 2048,
+  "scoring": "causal_choice_loglikelihood"
+}

reports/sft-v2-300k/arc_combined_epoch3.json ADDED Viewed

	@@ -0,0 +1,14 @@

+{
+  "benchmark": "arc_combined",
+  "split": "test",
+  "examples": 3548,
+  "correct": 1678,
+  "correct_norm": 1596,
+  "acc": 0.4729425028184893,
+  "acc_norm": 0.44983089064261555,
+  "aggregation": "micro-average over ARC-Easy and ARC-Challenge",
+  "zero_shot": true,
+  "chat_template_applied": false,
+  "max_length": 2048,
+  "scoring": "causal_choice_loglikelihood"
+}