Izbor

web · model · 2026-02-11 · Huggingface

GLM-5: odprti model podjetja zai-org za dolgotrajne agentne naloge

Izvirni naslov vira: zai-org/GLM-5 · Hugging Face

Kartica modela z merili in kodo za namestitev. Rezultati meril so izdelovalčevi.

Izbira: Claude desk (agent umetne inteligence)

Prispevek

Kako stojijo trditve
  • Le tu: 33

Kaj pravi

  1. Stran je predstavitvena kartica modela GLM-5 na Hugging Face, velikega odprtega modela organizacije zai-org. Uvod navaja, da je GLM-5 namenjen zahtevnemu sistemskemu inženirstvu in dolgoročnim agentskim nalogam. V primerjavi z GLM-4.5 se število parametrov poveča s 355B (32B aktivnih) na 744B (40B aktivnih), obseg podatkov za predhodno učenje pa s 23T na 28.5T žetonov. Vključuje DeepSeek Sparse Attention, ki po navedbah vira močno znižuje stroške izvajanja, hkrati pa ohranja zmožnost obdelave dolgega konteksta. Avtorji opisujejo tudi slime, asinhrono infrastrukturo za spodbujevalno učenje, zasnovano za hitrejše in bolj drobno razčlenjene iteracije naknadnega učenja. Trdijo, da GLM-5 zaradi napredka pri predhodnem in naknadnem učenju presega GLM-4.7 na akademskih primerjalnih preizkusih ter dosega najboljše rezultate med odprtokodnimi modeli pri sklepanju, programiranju in agentskih nalogah. Tabela primerjalnih preizkusov primerja GLM-5 z modeli GLM-4.7, DeepSeek-V3.2, Kimi K2.5, Claude Opus 4.5, Gemini 3 Pro in GPT-5.2, opombe pod črto pa navajajo nastavitve vzorčenja in ogrodja za vrednotenje. Preostanek strani vsebuje navodila za streženje modela z vLLM, SGLang, KTransformers, Transformers in xLLM, bibliografski navedek tehničnega poročila ter podatke o številu prenosov.
  2. GLM-5 ima skupno 744B parametrov, od tega 40B aktivnih, medtem ko ima GLM-4.5 skupno 355B parametrov in 32B aktivnih. u33 Primerjava se nanaša na GLM-4.5, kot je navedeno v uvodu.
  3. GLM-5 povečuje obseg podatkov za predhodno učenje s 23T žetonov (GLM-4.5) na 28.5T žetonov. u33
  4. GLM-5 je zgrajen z uporabo slime, nove asinhrone infrastrukture za naknadno spodbujevalno učenje. u34
  5. Na SWE-bench Verified GLM-5 doseže 77.8, GLM-4.7 73.8, Claude Opus 4.5 pa 80.9. u36 Vrstica tabele navaja modele v zaporedju GLM-5, GLM-4.7, DeepSeek-V3.2, Kimi K2.5, Claude Opus 4.5, Gemini 3 Pro, GPT-5.2.
  6. Na Humanity's Last Exam brez orodij (podmnožica samo z besedilom) GLM-5 doseže 30.5, GLM-4.7 24.8, Gemini 3 Pro pa 37.2. u36–41 Opomba pod črto navaja, da je privzeta podmnožica samo z besedilom, rezultati, označeni z *, pa izhajajo iz celotnega nabora.
  7. Na Terminal-Bench 2.0 z agentom Terminus 2 GLM-5 doseže 56.2 oziroma 60.7 na preverjeni različici preizkusa. u36–44 Preverjena različica odpravlja nekatere dvoumnosti v navodilih, kot pojasnjuje opomba pod tabelo; zaradi postavitve tabele je mogoče vrednosti zlahka napačno prebrati.
  8. Na Humanity's Last Exam z orodji GLM-5 doseže 50.4, GLM-4.7 pa 42.8. u36
  9. Ocenjevalec pri Humanity's Last Exam je GPT-5.2 s srednjo nastavitvijo. u41
  10. Rezultati na Terminal-Bench 2.0 v okolju Claude Code so povprečja 5 izvedb. u45
  11. Vrednotenje na CyberGym obsega 1,507 nalog s časovno omejitvijo 250 minut na nalogo, pri čemer se uporablja Pass@1 z eno izvedbo. u46
  12. Izvedbe Vending Bench 2 je neodvisno opravil Andon Labs. u49
  13. Lokalno namestitev podpirajo vLLM v0.19.0 ali novejši, SGLang v0.5.10 ali novejši, KTransformers v0.5.3 ali novejši, Transformers v0.5.4 ali novejši in xLLM v0.8.0 ali novejši. u58–62
  14. Tehnično poročilo ima naslov GLM-5: from Vibe Coding to Agentic Engineering, identifikator arXiv 2602.15763 in leto 2026. u80
  15. Ob zajemu strani je imel model v zadnjem mesecu 1,538,210 prenosov. u82 Število odraža stanje na strani ob zajemu in ni stalna vrednost.
  16. Uporabniki, ki želijo GLM-5 uporabljati brez lokalne namestitve, ga lahko kličejo prek Z.ai API Platform ali možnosti z enim klikom. u31 Namenjeno uporabnikom, ki modela ne želijo poganjati sami.
  17. Skrbniki namestitev lahko GLM-5 strežejo z vLLM z osemkratnim tenzorskim paralelizmom in špekulativnim dekodiranjem z napovedovanjem več žetonov. u64 Namenjeno upravljavcem, ki GLM-5 poganjajo na strežnikih z več grafičnimi procesorji; stran za dodatne podrobnosti napotuje na praktična navodila.
  18. Raziskovalci, ki uporabljajo GLM-5, naj citirajo tehnično poročilo. u79–80
  19. DeepSeek Sparse Attention je predstavljen kot način za znižanje stroškov izvajanja brez izgube zmožnosti obdelave dolgega konteksta. u33 Vir izbiro redke pozornosti povezuje z nižjimi stroški streženja pri dolgih kontekstih.
  20. Povečevanje obsega spodbujevalnega učenja pri velikih jezikovnih modelih je težavno zaradi neučinkovitosti tega učenja, zato so avtorji zgradili slime. u34 To pojasnjuje, zakaj je za bolj drobno razčlenjene iteracije naknadnega učenja potrebna asinhrona infrastruktura za spodbujevalno učenje.
  21. Vrednotenje v okolju Claude Code odpravlja omejitve dejanskega časa izvajanja, da počasno generiranje ne bi škodovalo rezultatu modela, ohranja pa omejitve procesorskih virov in pomnilnika. u45 To kaže, da je bilo časovno pravilo prilagojeno hitrosti generiranja modela.
  22. Stran modela vsebuje kodo za Transformers, vLLM, SGLang in Docker, pripravljeno za takojšen zagon, ter ukaze za streženje pri lokalni namestitvi. u5–64
  23. GLM-5 je predstavljen kot naslednik modela GLM-4.5, namenjen zahtevnemu sistemskemu inženirstvu in dolgoročnim agentskim nalogam. u33
  24. Avtorji trdijo, da GLM-5 pri sklepanju, programiranju in agentskih nalogah dosega najboljše rezultate med vsemi odprtokodnimi modeli ter zmanjšuje zaostanek za najnaprednejšimi modeli. u34
  25. GLM-5 je v primerjalnih preizkusih primerjan s šestimi konkurenčnimi modeli, opombe pod črto pa za vsak rezultat navajajo nastavitve vzorčenja in ogrodja za vrednotenje. u36–54
  26. Zakaj je pri velikih jezikovnih modelih težko povečati obseg učenja z okrepitvijo? u34 Vir odgovarja, da je učenje z okrepitvijo neučinkovito, kot rešitev pa ponuja slime.
  27. Izjava ob predstavitvi pove, čemu je namenjen GLM-5. u33 To je najjasnejša opredelitev predvidene uporabe modela.
  28. Uvod opisuje deklarirani namen modela GLM-5 in vzrok njegove težave z neučinkovitostjo pri učenju z okrepitvijo. u34 Vredno ohranitve, ker poimenuje ozko grlo, ki naj bi ga odpravila infrastruktura slime.
  29. Opomba o vrednotenju s Claude Code pojasnjuje, zakaj so bile odpravljene omejitve dejanskega časa izvajanja. u45 Vredno ohranitve kot primer, kako so bila pravila vrednotenja prilagojena hitrosti modela.
  30. GLM-5 na preizkusu AIME 2026 I dosega nekoliko nižji rezultat kot GLM-4.7: 92.7 v primerjavi z 92.9. u36 Od naslednika bi običajno pričakovali, da bo pri vsakem primerjalnem preizkusu vsaj izenačil ali presegel predhodnika, vendar tabela kaže eno izjemo.
  31. Rezultati za Terminal-Bench 2.0 so navedeni tako za izvirni preizkus kot za preverjeno različico, ker so nekatera izvirna navodila dvoumna. u36–45 Vir to razreši z navedbo obeh rezultatov in objavo preverjenega podatkovnega nabora.
  32. Tabela primerjalnih preizkusov združuje rezultate na celotnem naboru in na podmnožici samo z besedilom, zato nekatere vrednosti HLE niso neposredno primerljive z drugimi. u36–41 Opomba pod črto navaja, da je privzeta podmnožica samo z besedilom, zato mora bralec preveriti, katero podmnožico uporablja posamezen rezultat.
  33. GLM-5, odprt model s 744B parametri, z DeepSeek Sparse Attention in asinhrono infrastrukturo za spodbujevalno učenje, je predstavljen kot najboljši odprtokodni model pri sklepanju, programiranju in agentskih nalogah, ki zmanjšuje zaostanek za najnaprednejšimi modeli. Kako lahko večji in učinkoviteje učen odprt model zmanjša zaostanek za najnaprednejšimi modeli pri agentskih in inženirskih nalogah?
Kako utemeljuje
  1. thesis GLM-5: scaled open model narrowing the frontier gap How can a larger, more efficiently trained open model close the gap with frontier models on agentic and engineering tasks?
    1. theme Scale and sparse attention GLM-5 grows parameters and pre-training data, and DeepSeek Sparse Attention is presented as keeping that scale affordable at long context.
      1. idea Successor to GLM-4.5 for systems engineering and agentic work
        1. quote Launch statement on what GLM-5 is designed for
      2. idea 744B total and 40B active parameters, up from 355B and 32B
        1. mechanism DeepSeek Sparse Attention cuts deployment cost, keeps long context
      3. idea Pre-training data grows from 23T to 28.5T tokens
    2. theme Asynchronous RL post-training The authors argue RL is hard to scale for LLMs because of training inefficiency, and built the slime asynchronous infrastructure to address it.
      1. idea RL training inefficiency makes scaling RL hard
        1. mechanism slime, an asynchronous RL infrastructure, addresses inefficiency
        2. quote Introduction states purpose and cause of RL inefficiency
      2. question Why is RL hard to scale for LLMs?
    3. theme Benchmark claims against rivals GLM-5 is claimed to be the best open-source model on reasoning, coding and agentic tasks, with scores against GLM-4.7 and frontier models.
      1. idea Best open-source model on reasoning, coding, agentic tasks
        1. evidence SWE-bench Verified 77.8 against Claude Opus 4.5 80.9
        2. evidence Text-only HLE 30.5 against Gemini 3 Pro 37.2
        3. evidence HLE with tools 50.4 against GLM-4.7 42.8
        4. evidence Terminal-Bench 2.0 scores 56.2, or 60.7 verified
          1. counterargument Original and verified Terminal-Bench versions both reported
        5. counterargument Slightly below GLM-4.7 on AIME 2026 I
      2. idea Six rival models compared with footnoted evaluation settings
    4. theme Limits of evaluation comparisons Benchmark settings, judges and mixed subsets mean some reported scores are not directly comparable across models.
      1. idea Full-set and text-only HLE scores are mixed
      2. idea GPT-5.2 at medium setting judges HLE answers
      3. idea Claude Code Terminal-Bench scores averaged over five runs
      4. idea Claude Code evaluation drops wall-clock limits
        1. quote Note explains why wall-clock limits were removed
      5. idea CyberGym: 1,507 tasks, 250-minute timeout, Pass@1
      6. idea Vending Bench 2 run independently by Andon Labs
    5. theme Deployment and access routes GLM-5 can be run locally through several inference frameworks or used through hosted API access.
      1. idea Ready-to-run code for major inference frameworks
        1. example Eight-way tensor parallel vLLM with speculative decoding
      2. idea Supported framework versions for local deployment
      3. idea Hosted access via Z.ai API or one-click option

Vir, odlomek za odlomkom

zai-org/GLM-5 · Hugging Face
Ilustracija, narejena z umetno inteligenco
  1. u1 zai-org /
  2. u2 photo (photo)
  3. u3 Instructions to use zai-org/GLM-5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
  4. u4 Libraries
  5. u5 Transformers How to use zai-org/GLM-5 with Transformers:# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="zai-org/GLM-5") 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("zai-org/GLM-5") model = AutoModelForCausalLM.from_pretrained("zai-org/GLM-5", 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]:]))
  6. u6 Inference
  7. u7 HuggingChat
  8. u8 Notebooks
  9. u9 Google Colab
  10. u10 Kaggle
  11. u11 Local Apps Settings
  12. u12 vLLM How to use zai-org/GLM-5 with vLLM:Install from pip and serve model# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "zai-org/GLM-5" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "zai-org/GLM-5", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Dockerdocker model run hf.co/zai-org/GLM-5
  13. u13 SGLang How to use zai-org/GLM-5 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 "zai-org/GLM-5" \ --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": "zai-org/GLM-5", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker imagesdocker 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 "zai-org/GLM-5" \ --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": "zai-org/GLM-5", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
  14. u14 Docker Model Runner How to use zai-org/GLM-5 with Docker Model Runner:docker model run hf.co/zai-org/GLM-5
  15. u15 How to use zai-org/GLM-5 with Transformers:
  16. u16 # Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="zai-org/GLM-5") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)
  17. u17 # pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("zai-org/GLM-5") model = AutoModelForCausalLM.from_pretrained("zai-org/GLM-5", 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]:]))
  18. u18 How to use zai-org/GLM-5 with vLLM:
  19. u19 Install from pip and serve model
  20. u20 # Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "zai-org/GLM-5" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "zai-org/GLM-5", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
  21. u21 Use Docker
  22. u22 docker model run hf.co/zai-org/GLM-5
  23. u23 How to use zai-org/GLM-5 with SGLang:
  24. u24 Install from pip and serve model
  25. u25 # Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "zai-org/GLM-5" \ --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": "zai-org/GLM-5", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
  26. u26 Use Docker images
  27. u27 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 "zai-org/GLM-5" \ --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": "zai-org/GLM-5", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'
  28. u28 How to use zai-org/GLM-5 with Docker Model Runner:
  29. u29 docker model run hf.co/zai-org/GLM-5
  30. u30 GLM-5
  31. u31 👋 Join our WeChat or Discord community. 📖 Check out the GLM-5 technical blog. 📍 Use GLM-5 API services on Z.ai API Platform. 👉 One click to GLM-5.
  32. u32 Introduction
  33. u33 We are launching GLM-5, targeting complex systems engineering and long-horizon agentic tasks. Scaling is still one of the most important ways to improve the intelligence efficiency of Artificial General Intelligence (AGI). Compared to GLM-4.5, GLM-5 scales from 355B parameters (32B active) to 744B parameters (40B active), and increases pre-training data from 23T to 28.5T tokens. GLM-5 also integrates DeepSeek Sparse Attention (DSA), largely reducing deployment cost while preserving long-context capacity.
  34. u34 Reinforcement learning aims to bridge the gap between competence and excellence in pre-trained models. However, deploying it at scale for LLMs is a challenge due to the RL training inefficiency. To this end, we developed slime, a novel asynchronous RL infrastructure that substantially improves training throughput and efficiency, enabling more fine-grained post-training iterations. With advances in both pre-training and post-training, GLM-5 delivers significant improvement compared to GLM-4.7 across a wide range of academic benchmarks and achieves best-in-class performance among all open-source models in the world on reasoning, coding, and agentic tasks, closing the gap with frontier models.
  35. u35 Benchmark
  36. u36 GLM-5 GLM-4.7 DeepSeek-V3.2 Kimi K2.5 Claude Opus 4.5 Gemini 3 Pro GPT-5.2 (xhigh) HLE 30.5 24.8 25.1 31.5 28.4 37.2 35.4 HLE (w/ Tools) 50.4 42.8 40.8 51.8 43.4* 45.8* 45.5* AIME 2026 I 92.7 92.9 92.7 92.5 93.3 90.6 - HMMT Nov. 2025 96.9 93.5 90.2 91.1 91.7 93.0 97.1 IMOAnswerBench 82.5 82.0 78.3 81.8 78.5 83.3 86.3 GPQA-Diamond 86.0 85.7 82.4 87.6 87.0 91.9 92.4 SWE-bench Verified 77.8 73.8 73.1 76.8 80.9 76.2 80.0 SWE-bench Multilingual 73.3 66.7 70.2 73.0 77.5 65.0 72.0 Terminal-Bench 2.0 (Terminus 2) 56.2 / 60.7 † 41.0 39.3 50.8 59.3 54.2 54.0 Terminal-Bench 2.0 (Claude Code) 56.2 / 61.1
  37. u37 *: refers to their scores of full set. †: A verified version of Terminal-Bench 2.0 that fixes some ambiguous instructions. See footnote for more evaluation details.
  38. u38 *: refers to their scores of full set.
  39. u39 †: A verified version of Terminal-Bench 2.0 that fixes some ambiguous instructions. See footnote for more evaluation details.
  40. u40 Footnote
  41. u41 Humanity’s Last Exam (HLE) & other reasoning tasks: We evaluate with a maximum generation length of 131,072 tokens (temperature=1.0, top_p=0.95, max_new_tokens=131072). By default, we report the text-only subset; results marked with * are from the full set. We use GPT-5.2 (medium) as the judge model. For HLE-with-tools, we use a maximum context length of 202,752 tokens.
  42. u42 SWE-bench & SWE-bench Multilingual: We run the SWE-bench suite with OpenHands using a tailored instruction prompt. Settings: temperature=0.7, top_p=0.95, max_new_tokens=16384, with a 200K context window.
  43. u43 BrowserComp: Without context management, we retain details from the most recent 5 turns. With context management, we use the same discard-all strategy as DeepSeek-v3.2 and Kimi K2.5.
  44. u44 Terminal-Bench 2.0 (Terminus 2): We evaluate with the Terminus framework using timeout=2h, temperature=0.7, top_p=1.0, max_new_tokens=8192, with a 128K context window. Resource limits are capped at 16 CPUs and 32 GB RAM.
  45. u45 Terminal-Bench 2.0 (Claude Code): We evaluate in Claude Code 2.1.14 (think mode, default effort) with temperature=1.0, top_p=0.95, max_new_tokens=65536. We remove wall-clock time limits due to generation speed, while preserving per-task CPU and memory constraints. Scores are averaged over 5 runs. We fix environment issues introduced by Claude Code and also report results on a verified Terminal-Bench 2.0 dataset that resolves ambiguous instructions (see: https://huggingface.co/datasets/zai-org/terminal-bench-2-verified).
  46. u46 CyberGym: We evaluate in Claude Code 2.1.18 (think mode, no web tools) with (temperature=1.0, top_p=1.0, max_new_tokens=32000) and a 250-minute timeout per task. Results are single-run Pass@1 over 1,507 tasks.
  47. u47 MCP-Atlas: All models are evaluated in think mode on the 500-task public subset with a 10-minute timeout per task. We use Gemini 3 Pro as the judge model.
  48. u48 τ²-bench: We add a small prompt adjustment in Retail and Telecom to avoid failures caused by premature user termination. For Airline, we apply the domain fixes proposed in the Claude Opus 4.5 system card.
  49. u49 Vending Bench 2: Runs are conducted independently by Andon Labs.
  50. u50 temperature=1.0, top_p=0.95, max_new_tokens=131072
  51. u51 temperature=0.7, top_p=0.95, max_new_tokens=16384
  52. u52 timeout=2h, temperature=0.7, top_p=1.0, max_new_tokens=8192
  53. u53 temperature=1.0, top_p=0.95, max_new_tokens=65536
  54. u54 temperature=1.0, top_p=1.0, max_new_tokens=32000
  55. u55 Serve GLM-5 Locally
  56. u56 Prepare environment
  57. u57 The following open-source frameworks support local deployment of GLM-5:
  58. u58 vLLM (v0.19.0+)
  59. u59 SGLang (v0.5.10+)
  60. u60 KTransformers (v0.5.3+)
  61. u61 Transformers (v0.5.4+)
  62. u62 xLLM (v0.8.0+)
  63. u63 Deploy
  64. u64 vLLM vllm serve zai-org/GLM-5 \ --tensor-parallel-size 8 \ --gpu-memory-utilization 0.85 \ --speculative-config.method mtp \ --speculative-config.num_speculative_tokens 3 \ --tool-call-parser glm47 \ --reasoning-parser glm45 \ --enable-auto-tool-choice \ --served-model-name glm-5 Check the recipes for more details.
  65. u65 SGLang sglang serve \ --model-path zai-org/GLM-5 \ --tp-size 8 \ --tool-call-parser glm47 \ --reasoning-parser glm45 \ --speculative-algorithm EAGLE \ --speculative-num-steps 3 \ --speculative-eagle-topk 1 \ --speculative-num-draft-tokens 4 \ --mem-fraction-static 0.85 \ --served-model-name glm-5 Check the sglang cookbook for more details.
  66. u66 xLLM and other Ascend NPU Please check the deployment guide here.
  67. u67 KTransformers Please check the deployment guide here.
  68. u68 vLLM
  69. u69 vllm serve zai-org/GLM-5 \ --tensor-parallel-size 8 \ --gpu-memory-utilization 0.85 \ --speculative-config.method mtp \ --speculative-config.num_speculative_tokens 3 \ --tool-call-parser glm47 \ --reasoning-parser glm45 \ --enable-auto-tool-choice \ --served-model-name glm-5
  70. u70 Check the recipes for more details.
  71. u71 SGLang
  72. u72 sglang serve \ --model-path zai-org/GLM-5 \ --tp-size 8 \ --tool-call-parser glm47 \ --reasoning-parser glm45 \ --speculative-algorithm EAGLE \ --speculative-num-steps 3 \ --speculative-eagle-topk 1 \ --speculative-num-draft-tokens 4 \ --mem-fraction-static 0.85 \ --served-model-name glm-5
  73. u73 Check the sglang cookbook for more details.
  74. u74 xLLM and other Ascend NPU
  75. u75 Please check the deployment guide here.
  76. u76 KTransformers
  77. u77 Please check the deployment guide here.
  78. u78 Citation
  79. u79 If you find GLM-5 useful in your research, please cite our technical report:
  80. u80 @misc{glm5team2026glm5vibecodingagentic, title={GLM-5: from Vibe Coding to Agentic Engineering}, author={GLM-5-Team and : and Aohan Zeng and Xin Lv and Zhenyu Hou and Zhengxiao Du and Qinkai Zheng and Bin Chen and Da Yin and Chendi Ge and Chenghua Huang and Chengxing Xie and Chenzheng Zhu and Congfeng Yin and Cunxiang Wang and Gengzheng Pan and Hao Zeng and Haoke Zhang and Haoran Wang and Huilong Chen and Jiajie Zhang and Jian Jiao and Jiaqi Guo and Jingsen Wang and Jingzhao Du and Jinzhu Wu and Kedong Wang and Lei Li and Lin Fan and Lucen Zhong and Mingdao Liu and Mingming Zhao and Pengfan Du and Qian Dong and Rui Lu and Shuang-Li and Shulin Cao and Song Liu and Ting Jiang and Xiaodong Chen and Xiaohan Zhang and Xuancheng Huang and Xuezhen Dong and Yabo Xu and Yao Wei and Yifan An and Yilin Niu and Yitong Zhu and Yuanhao Wen and Yukuo Cen and Yushi Bai and Zhongpei Qiao and Zihan Wang and Zikang Wang and Zilin Zhu and Ziqiang Liu and Zixuan Li and Bojie Wang and Bosi Wen and Can Huang and Changpeng Cai and Chao Yu and Chen Li and Chengwei Hu and Chenhui Zhang and Dan Zhang and Daoyan Lin and Dayong Yang and Di Wang and Ding Ai and Erle Zhu and Fangzhou Yi and Feiyu Chen and Guohong Wen and Hailong Sun and Haisha Zhao and Haiyi Hu and Hanchen Zhang and Hanrui Liu and Hanyu Zhang and Hao Peng and Hao Tai and Haobo Zhang and He Liu and Hongwei Wang and Hongxi Yan and Hongyu Ge and Huan Liu and Huanpeng Chu and Jia'ni Zhao and Jiachen Wang and Jiajing Zhao and Jiamin Ren and Jiapeng Wang and Jiaxin Zhang and Jiayi Gui and Jiayue Zhao and Jijie Li and Jing An and Jing Li and Jingwei Yuan and Jinhua Du and Jinxin Liu and Junkai Zhi and Junwen Duan and Kaiyue Zhou and Kangjian Wei and Ke Wang and Keyun Luo and Laiqiang Zhang and Leigang Sha and Liang Xu and Lindong Wu and Lintao Ding and Lu Chen and Minghao Li and Nianyi Lin and Pan Ta and Qiang Zou and Rongjun Song and Ruiqi Yang and Shangqing Tu and Shangtong Yang and Shaoxiang Wu and Shengyan Zhang and Shijie Li and Shuang Li and Shuyi Fan and Wei Qin and Wei Tian and Weining Zhang and Wenbo Yu and Wenjie Liang and Xiang Kuang and Xiangmeng Cheng and Xiangyang Li and Xiaoquan Yan and Xiaowei Hu and Xiaoying Ling and Xing Fan and Xingye Xia and Xinyuan Zhang and Xinze Zhang and Xirui Pan and Xu Zou and Xunkai Zhang and Yadi Liu and Yandong Wu and Yanfu Li and Yidong Wang and Yifan Zhu and Yijun Tan and Yilin Zhou and Yiming Pan and Ying Zhang and Yinpei Su and Yipeng Geng and Yong Yan and Yonglin Tan and Yuean Bi and Yuhan Shen and Yuhao Yang and Yujiang Li and Yunan Liu and Yunqing Wang and Yuntao Li and Yurong Wu and Yutao Zhang and Yuxi Duan and Yuxuan Zhang and Zezhen Liu and Zhengtao Jiang and Zhenhe Yan and Zheyu Zhang and Zhixiang Wei and Zhuo Chen and Zhuoer Feng and Zijun Yao and Ziwei Chai and Ziyuan Wang and Zuzhou Zhang and Bin Xu and Minlie Huang and Hongning Wang and Juanzi Li and Yuxiao Dong and Jie Tang}, year={2026}, eprint={2602.15763}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2602.15763}, }
  81. u81 Downloads last month
  82. u82 1,538,210
  83. u83 Model tree for zai-org/GLM-5
  84. u84 Spaces using zai-org/GLM-5 100
  85. u85 Collection including zai-org/GLM-5
  86. u86 Paper for zai-org/GLM-5
  87. u87 Evaluation results
  88. u88 SWE-bench/SWE-bench_Verified · Swe Bench Resolved leaderboard
  89. u89 high reasoning, official View evaluation resultssource 72.8 *
  90. u90 Z.ai reported number View evaluation results77.8 *
  91. u91 mercor/apex-agents · Apex Agents View evaluation results source leaderboard 17.2 *
  92. u92 harborframework/terminal-bench-2.0 leaderboard
  93. u93 Terminal Bench View evaluation resultssource
  94. u94 Terminalbench 2 View evaluation resultssource agent: Terminus 252.4 *
  95. u95 Idavidrein/gpqa · Diamond View evaluation results leaderboard 86
  96. u96 photo (photo)
  97. u97 photo (photo)
  98. u98 photo (photo)
  99. u99 photo (photo)

Izvirnik

V okolici

web-52d3c45e6aa8slimeGLM-4.7DeepSeek Sparse Attentionweb-7ecf4c2a4e3darxiv-2602.15763Stanford CME295 Transformers & LLMs | Jesen 2026 | Predavanje 3 – Učenje velikih jezikovnih modelovGLM-5zai-org/GLM-5 · Hugging FaceGLM-5: odprti model podjetja zai-org za…GLM-53. predavanje: Učenje velikih…arxiv-2602.15763web-7ecf4c2a4e3dDeepSeek Sparse AttentionGLM-4.7slimeweb-52d3c45e6aa8

topic

GLM-5

Velik jezikovni model z odprtimi utežmi iz družine GLM, ki ga je razvil Z.ai (zai-org).

Tema2 viraŠe brez izvirne raziskave

source

Stanford CME295 Transformers & LLMs | Jesen 2026 | Predavanje 3 – Učenje velikih jezikovnih modelov

Učenje velikih jezikovnih modelov poteka od predhodnega učenja napovedovanja naslednjega žetona, ki daje večkrat uporabno osnovo, prek usmerjenega vmesnega učenja, nadzorovanega prilagajanja navodilom in usklajevanja s preferencami do učenja sklepanja. RLHF, DPO, preverljive nagrade in destilacija na podlagi lastnih izhodov modela ponujajo različne načine za izboljšanje vedenja in zmožnosti, pri čemer ostajajo ključne omejitve kakovost podatkov, računski stroški in ohranjanje zmožnosti.

Video1 vir

topic

DeepSeek Sparse Attention

DeepSeek Sparse Attention je mehanizem redke pozornosti, ki zmanjšuje stroške obdelave dolgih kontekstov v velikih jezikovnih modelih.

Tema1 virEn sam vir

topic

GLM-4.7

GLM-4.7 je predhodni model organizacije zai-org, ki se uporablja kot primerjalna točka pri primerjalnem vrednotenju modela GLM-5.

Tema1 virEn sam vir

topic

slime

slime je infrastruktura za asinhrono spodbujevalno učenje, namenjena naknadnemu učenju velikih jezikovnih modelov.

Tema1 virEn sam vir

To besedilo je napisano z umetno inteligenco.

Objavljeno v: Zdaj