Instructions to use Zhangdanyang/qwen3-coder-micake-c144-planner with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Zhangdanyang/qwen3-coder-micake-c144-planner with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Zhangdanyang/qwen3-coder-micake-c144-planner") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Zhangdanyang/qwen3-coder-micake-c144-planner") model = AutoModelForCausalLM.from_pretrained("Zhangdanyang/qwen3-coder-micake-c144-planner", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=256) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Zhangdanyang/qwen3-coder-micake-c144-planner with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Zhangdanyang/qwen3-coder-micake-c144-planner" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Zhangdanyang/qwen3-coder-micake-c144-planner", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Zhangdanyang/qwen3-coder-micake-c144-planner
- SGLang
How to use Zhangdanyang/qwen3-coder-micake-c144-planner 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 "Zhangdanyang/qwen3-coder-micake-c144-planner" \ --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": "Zhangdanyang/qwen3-coder-micake-c144-planner", "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 "Zhangdanyang/qwen3-coder-micake-c144-planner" \ --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": "Zhangdanyang/qwen3-coder-micake-c144-planner", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use Zhangdanyang/qwen3-coder-micake-c144-planner with Docker Model Runner:
docker model run hf.co/Zhangdanyang/qwen3-coder-micake-c144-planner
Qwen3-Coder MICake C144 Planner (m24)
่ฟไธชไปๅบๆฏ MICake Planner ็ๅฝๅๅ ๅ m24ใๅฎๆฏ Qwen3-Coder-30B-A3B-Instruct ไธ่ฟ็ปญ LoRA SFT ๅๅๅนถ็ๅฎๆดๆ้๏ผๅฏไปฅ็ดๆฅ็จ transformers / vLLM / ATOM ๅ ่ฝฝใๆฌไปๅบไธๆฏๅ็ฌ็ LoRA adapterใ
่ฐฑ็ณป๏ผQwen3-Coder-30B-A3B-Instruct โ โฆ โ p3๏ผๆฌไปๅบๅ
ๅๅๅธ็ๆ้๏ผโ โฆ โ m21 โ m24๏ผๆฌไปๅบ็ฐๅจ็ๆ้๏ผใm24 ็ LoRA ๅบๅบงๅ่ฎญ็ปๆฐๆฎ้ฝๆฅ่ชๅฝๆถ็ๅ ๅ m21ใ288 ไธชๆ ทๆฌใ640 ่กใ160 ๆญฅใ2 ไธช epochใ
m24 ๅจไธ่ฎญ็ป้ไธ็ธไบค็ held-out ็ปๅไธ๏ผ36 ๅฏนไบค้ๆต่ฏ็ๅนณๅๅคๆตๅ ้ๆฏๆฏ 1.1379ร๏ผๅๅบ DSV4.1 Planner ๆฏ 1.1156รใๅทฎ +0.0223๏ผ95% ๅบ้ด [+0.0192, +0.0255]๏ผp = 0.0001ใbf16 ๅ fp8 ๅ่ช 18 ๅฏน๏ผไน้ฝๆพ่้ซไบ DSV4.1ใ36 ๅฏน้ m24 ๅ จ้จ้ซไบๅๅบ DSV4.1ใ
ๆจกๅๅจๅไปไน
Planner ไธๅ kernel ๆบ็ ใๅฎๆไบคไธไปฝ GEMM ้ ็ฝฎใMICake ็จ่ฟไปฝ้ ็ฝฎ็ๆ Triton GEMM๏ผๅจ AMD gfx950 ไธ็ผ่ฏ๏ผๅๆญฃ็กฎๆงๆฃๆฅ๏ผๅๆต็ธๅฏนๅบ็บฟ้ ็ฝฎ็ๅ ไฝๅนณๅๅ ้ๆฏใๆฏไธช episode ๆๅค 10 ๆฌกๆไบคใ
้ ็ฝฎๅไธคๆฎต๏ผ
schedule๏ผtile_mใtile_nใtile_kใpipeline_depthใunroll_factorใvectorize_factorrealization๏ผprogram_orderใgroup_mใnum_warpsใnum_stagesใwaves_per_euใmatrix_instr_nonkdim๏ผkpackๅบๅฎไธบ 1๏ผ
็ซฏๅฐ็ซฏ่ง่ฒ๏ผManager ๅ Critic ๅบๅฎไธบ DSV4.1๏ผๅชๆ Planner ๆขๆจกๅใๅ ๅใๆๆ่ ใDSV4.1 ไธไธช arm ไบค้้ๆ ทใheld-out ๆฏไธช arm 36 ไธช episode๏ผ4 ไธช MICake GPU ร ๆฏไธช GPU 9 ไธชๆ ทๆฌ๏ผ๏ผbf16 ๅ fp8 ๅ 18 ไธชใ
่กจ้็ๅ ้ๆฏๆฏๅคๆตๅผ๏ผagent ๆพๅฐ็ๆๅฅฝ้ ็ฝฎๅจๆฐ worker ไธๅๆต 2 ๆฌกๅไธญไฝๆฐใagent ๆฒกๆๆไบคๅบๆๆๆน่ฟๆถ่ฎฐ 1.000รใ
่ฆไผๅ็ๅบ็บฟ้ ็ฝฎ
| ๅญๆฎต | ๅบ็บฟ |
|---|---|
| tile_m ร tile_n ร tile_k | 64 ร 64 ร 32 |
| pipeline_depth | 3 |
| unroll_factor | 4 |
| vectorize_factor | 4 |
| num_stages | 3 |
| num_warps | 4 |
| matrix_instr_nonkdim | 0 |
| program_order | TWO_DIMENSIONAL |
| group_m | 1 |
| waves_per_eu | 1 |
| kpack | 1 |
Held-out ๅฝข็ถ
่ฟ 7 ไธชๅฝข็ถไธๅจ C144 ่ฎญ็ป้้๏ผไนไธๅจ large ่ฎญ็ป้้ใไธๆฌกๆไบค็ๅๆฐๆฏๅฎไปฌ็ๅ ไฝๅนณๅใ่ฎฐๅฝ้ๆฒกๆ้ๅฝข็ถ็ speedupใ
| workload | M | N | K |
|---|---|---|---|
| gemm.tail-balanced | 272 | 272 | 288 |
| gemm.tail-rectangular | 1025 | 257 | 513 |
| gemm.square-4k | 4096 | 4096 | 4096 |
| gemm.prefill-wide | 8192 | 1024 | 4096 |
| gemm.prefill-tall | 1024 | 8192 | 4096 |
| gemm.decode | 128 | 4096 | 4096 |
| gemm.mlp-7168 | 4096 | 7168 | 2048 |
Held-out ็ปๆ
n = 36๏ผๅไธๆนไปปๅกไธไบค้ๆฏ่พใm24 ็ธๅฏนไธไธไปฃๅ ๅ m21 ็ๅนณๅ best ๅข็ๆฏ +0.0391๏ผ1.1379 ๅฏน 1.0988๏ผ๏ผๆๆดๅฟซ kernel ๆๅใ
| Planner | ๅนณๅๅคๆต | ๅฐ่พพ 1.01ร ไธญไฝๆถ้ด | ๅฐ่พพ 1.03ร ไธญไฝๆถ้ด | ๅฐ่พพ 1.04ร ไธญไฝๆถ้ด | ๅฐ่พพ 1.05ร ไธญไฝๆถ้ด | ๆๅไธๆฌกๆน่ฟ็ไธญไฝๆถ้ด | ๅข้ไธญไฝ |
|---|---|---|---|---|---|---|---|
| m21 | 1.0988 | 93 s | 96 s | 188 s | 284 s | 840 s | 1019 s |
| DSV4.1 | 1.1156 | 87 s | 87 s | 87 s | 172 s | 732 s | 919 s |
| m24 | 1.1379 | 93 s | 94 s | 131 s | 267 s | 787 s | 981 s |
ไธไธช arm ็ 36 ไธช episode ้ฝๅฐ่พพไบ 1.01รใ1.03รใ1.04ร ๅ 1.05ร๏ผๅฐ่พพ็้ฝๆฏ 100%ใโๆๅไธๆฌกๆน่ฟโๆฏ่ฏฅ episode ๆๅไธๆฌกๅทๆฐ best ็ๆถ้ด๏ผไธไธช arm ้ฝๆฏ 36/36 ๆๅผใ
้ๅบฆไธ m24 ็ kernel ๆฏ DSV4.1 ๅฟซ๏ผไฝๅฐ 1.04ร ็ไธญไฝๆถ้ดไปๆฏ DSV4.1 ้ฟ๏ผ131 ็งๅฏน 87 ็ง๏ผใDSV4.1 ็็ฌฌไธๆฌกๆๆๆไบคๅธธๅธธ็ดๆฅ่ฝๅจ 128ร128๏ผๆไปฅ 1.01ร ๅฐ 1.04ร ๅ ไนๅจๅไธๆถๅปๅฐ่พพใm24 ็้ฆๆไบคๆไธๅไปๆฏ 64ร64๏ผ่ฆๅค่ทๅ ๆฌกๆไบคๆ่ฝๅฐ 128ร128ใ
| dtype | ๅฏนๆฐ | m24 | DSV4.1 | ๅทฎ | p |
|---|---|---|---|---|---|
| bf16 | 18 | 1.1327 | 1.1123 | +0.0204 | 0.0001 |
| fp8 | 18 | 1.1431 | 1.1190 | +0.0241 | 0.0001 |
| ๅ่ฎก | 36 | 1.1379 | 1.1156 | +0.0223 | 0.0001 |
ๅ่ฎก็ 95% ๅบ้ดๆฏ [+0.0192, +0.0255]ใ
้ๅฏนๅๆฐ
| dtype | ๆ ทๆฌ | m24 | DSV4.1 | m24 โ DSV4.1 |
|---|---|---|---|---|
| bf16 | 0 | 1.1349 | 1.1046 | +0.0303 |
| bf16 | 2 | 1.1243 | 1.1110 | +0.0133 |
| bf16 | 4 | 1.1334 | 1.1199 | +0.0135 |
| bf16 | 6 | 1.1350 | 1.1139 | +0.0211 |
| bf16 | 8 | 1.1262 | 1.1188 | +0.0074 |
| bf16 | 10 | 1.1343 | 1.1038 | +0.0305 |
| bf16 | 12 | 1.1314 | 1.1163 | +0.0151 |
| bf16 | 14 | 1.1328 | 1.1158 | +0.0170 |
| bf16 | 16 | 1.1338 | 1.1165 | +0.0173 |
| bf16 | 18 | 1.1311 | 1.1030 | +0.0281 |
| bf16 | 20 | 1.1333 | 1.1032 | +0.0301 |
| bf16 | 22 | 1.1288 | 1.1094 | +0.0194 |
| bf16 | 24 | 1.1362 | 1.1121 | +0.0241 |
| bf16 | 26 | 1.1400 | 1.1015 | +0.0385 |
| bf16 | 28 | 1.1336 | 1.1143 | +0.0193 |
| bf16 | 30 | 1.1330 | 1.1176 | +0.0154 |
| bf16 | 32 | 1.1307 | 1.1206 | +0.0101 |
| bf16 | 34 | 1.1359 | 1.1186 | +0.0173 |
| fp8 | 1 | 1.1351 | 1.1023 | +0.0328 |
| fp8 | 3 | 1.1395 | 1.1202 | +0.0193 |
| fp8 | 5 | 1.1388 | 1.1214 | +0.0174 |
| fp8 | 7 | 1.1500 | 1.1201 | +0.0299 |
| fp8 | 9 | 1.1431 | 1.1189 | +0.0242 |
| fp8 | 11 | 1.1502 | 1.1183 | +0.0319 |
| fp8 | 13 | 1.1490 | 1.1188 | +0.0302 |
| fp8 | 15 | 1.1372 | 1.1185 | +0.0187 |
| fp8 | 17 | 1.1505 | 1.1213 | +0.0292 |
| fp8 | 19 | 1.1348 | 1.1180 | +0.0168 |
| fp8 | 21 | 1.1366 | 1.1187 | +0.0179 |
| fp8 | 23 | 1.1373 | 1.1211 | +0.0162 |
| fp8 | 25 | 1.1503 | 1.1199 | +0.0304 |
| fp8 | 27 | 1.1398 | 1.1207 | +0.0191 |
| fp8 | 29 | 1.1378 | 1.1191 | +0.0187 |
| fp8 | 31 | 1.1466 | 1.1210 | +0.0256 |
| fp8 | 33 | 1.1480 | 1.1205 | +0.0275 |
| fp8 | 35 | 1.1507 | 1.1224 | +0.0283 |
ๆ็ป kernel
36 ๅฏน้ๆ็ป้ ็ฝฎๆฒกๆไธๅฏน็ธๅใtile ้ฝๆฏ 128ร128๏ผnum_warps ้ฝๆฏ 8๏ผๅทฎๅซๅจๅซ็ๅญๆฎต๏ผ
| ๅญๆฎต | m24 | DSV4.1 |
|---|---|---|
| unroll_factor | ๅ จ้จๆฏ 2 | ๅ จ้จๆฏ 4 |
| pipeline_depth | 31/36 ๆฏ 2๏ผ5/36 ๆฏ 3 | 35/36 ๆฏ 3๏ผ1/36 ๆฏ 4 |
| tile_k | 24 ๆกๆฏ 32๏ผ12 ๆกๆฏ 16 | 32 ๆกๆฏ 32๏ผ4 ๆกๆฏ 16 |
ๆ dtype ็ m24 ็ tile_k๏ผbf16 ไธ 16/18 ๆกๆฏ 32ใ2 ๆกๆฏ 16๏ผfp8 ไธ 10/18 ๆกๆฏ 16ใ8 ๆกๆฏ 32ใvectorize_factor ไธค่พน้ฝๆฏ 4ใ
DSV4.1 ๅจ 36 ไธช episode ้ไธๆฌก้ฝๆฒกๆๆไบค่ฟ m24 ็้ฃไปฝๆ็ป้ ็ฝฎใm24 ๆ 2 ๆฌกๆไบค่ฟ DSV4.1 ็ๆ็ป้ ็ฝฎ๏ผๅฝๅบๆตๅฐ 1.1182 ๅ 1.1191๏ผไฝไบๅฎ่ชๅทฑ้ๅบ็ 1.1404 ๅ 1.1481ใ
ๅไธไปฝ้ ็ฝฎ็ไธคๆฌกๅคๆต็ปๅฏนๅทฎ๏ผm24 ไธญไฝๆฐ 0.0015ใๆๅคง 0.0050๏ผDSV4.1 ไธญไฝๆฐ 0.0017ใๆๅคง 0.0048ใ36 ๅฏน้ m24 ไธคๆฌกๅคๆต็่พไฝๅผ้ฝ้ซไบ DSV4.1 ไธคๆฌกๅคๆต็่พ้ซๅผ๏ผๆต้ๅบ้ดๆฒกๆ้ๅ ใๅๆฐๅทฎๆฅ่ชไธ้ข่ฟไปฝไธๅ็้ ็ฝฎ๏ผไธๆฏๅไธๆฌก kernel ็ๆต่ฏๆณขๅจใ
็ฌฌไธๆฌกๆต้็ๆไบคๆ 18/36 ๆกๅทฒ็ปๆฏ 128ร128ร32๏ผๅฆๅค 18 ๆกไปไป 64ร64ร32 ๅผๅง๏ผๆๅ้ฝ็ฌๅฐไบ 128ร128ใ
m24 ่ฟไธ่ฝฎ็ๆไบค้ถๆฎต่ฎกๆฐ๏ผmeasured 331๏ผduplicate 35๏ผcompile_fail 26๏ผschema_invalid 2๏ผverify_fail 1๏ผresource_rejected 1ใ
็จๆณ
่ฟๆฏๅๅนถๅ็ Qwen3-MoE ๅ ๆ่ฏญ่จๆจกๅ๏ผ30.5B ๅๆฐใๆฏ token ๆฟๆดป็บฆ 3.3Bใ็จๅบๅบงๆจกๅ็ chat template ๅๅทฅๅ ท่ฐ็จๆ ผๅผใ
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "Zhangdanyang/qwen3-coder-micake-c144-planner"
tokenizer = AutoTokenizer.from_pretrained(repo, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(repo, trust_remote_code=True, device_map="auto")
ๅบๅบง่ฎธๅฏ่ฏๆฏ Apache 2.0๏ผ่งไปๅบไธญ็ LICENSEใ
- Downloads last month
- 33
Model tree for Zhangdanyang/qwen3-coder-micake-c144-planner
Base model
Qwen/Qwen3-Coder-30B-A3B-Instruct