| Optimal Setting for LiveCodeBench evaluation? | PhamHelga | open | 0 | disc |
| Listed on OpenModelMap | duola15 | open | 0 | disc |
| TemporalMesh Transformer: 29.4 PPL at 48% compute — beats Mamba, new open-source architecture | vigneshwar234 | open | 0 | disc |
| Add EvalEval community eval results | EvalEvalBot | closed | 0 | disc |
| Upload Текст_11.txt | Viletter | closed | 0 | disc |
| Add SWE-Bench Verified evaluation results | nielsr | open | 0 | disc |
| This model is too brilliant to make me believe it has only 30B parameters | scentof | open | 0 | disc |
| Quantized GLM-4.7-Flash with llama.cpp and opencode. | ghostwithahat | open | 0 | disc |
| Is FIM supported for code completion？ | kuang12 | open | 0 | disc |
| WARNING 02-08 00:50:40 [vllm.py:1500] `torch.compile` is turned on, but the model zai-org/GLM-4.7-Flash does not support it. Please open an issue on GitHub if you want it to be supported. | ANASDAVOODTK | open | 0 | disc |
| This model keep telling lies | awesome747 | open | 0 | disc |
| Quantized Variants Now Available for FP8 and W8A16 | Geodd | open | 0 | disc |
| sdd | multitaskmayhem | closed | 0 | disc |
| Use the correct way of initializing using latest transformers latest branch v5.0.0 | rjmehta | open | 0 | disc |
| [Docs] Add LightLLM deployment example | FubaoSu | open | 0 | disc |
| Off reasoning | kmahdi | open | 0 | disc |
| Add HLE evaluation result | burtenshaw | merged | 0 | disc |
| Add GPQA evaluation result | burtenshaw | merged | 0 | disc |
| Exceptional tool use in various typescript cli coders | cleyesode | closed | 0 | disc |
| MONEY MAKING | Andyhabu | closed | 0 | disc |
| boni | wonde621 | closed | 0 | disc |
| 为什么要设计成是否think必须用户指定而不能模型自己选择 | fudayuan | open | 0 | disc |
| Long context reasoning scores bad? | sebastienbo | open | 0 | disc |
| Endless Repetition? Anyone encountered? | evilperson068 | open | 0 | disc |
| Inference Much slower as compared to other A3B Models | engrtipusultan | open | 0 | disc |
| Possible to run this in 8GB VRAM + 48GB RAM? | krigeta | open | 0 | disc |
| ask | Emmang | closed | 0 | disc |
| Excellent model - short feedback | Dampfinchen | open | 0 | disc |
| Thank you Z.AI, I love this model! ❤ | MrDevolver | open | 0 | disc |
| BTC | Tristan505 | closed | 0 | disc |
| VLLM NVFP4 PROBLEM | prudant | open | 0 | disc |
| Upload 51 files | 65ergyuhij | closed | 0 | disc |
| Upload 52 files | 65ergyuhij | closed | 0 | disc |
| Model breaks apart when used with different languages | nephepritou | open | 0 | disc |
| Number of layers 47 or 48 | wfen | open | 0 | disc |
| Amazing! look what this local AI generated in 5 minutes. | robert1968 | open | 0 | disc |
| Problems with logical reasoning performance of GLM-4.7-Flash | sszymczyk | open | 0 | disc |
| There is no module or parameter named 'model.layers.5.mlp.gate.e_score_correction_bias' in TransformersMoEForCausalLM | divinefeng | open | 0 | disc |
| open Tau^2 benchmark codebase? | howtain | open | 0 | disc |
| Please consider making it available through your official chat website. ❤ | MrDevolver | open | 0 | disc |
| Do you guys have a plan to create dense coding specific model? | hanzceo | open | 0 | disc |
| test | mickinsey | open | 0 | disc |
| FP8 ? | festr2 | open | 0 | disc |
| config.json - "scoring_func": "sigmoid" | algorithm | open | 0 | disc |
| Question about model usage in Turkish | 0xStego | open | 0 | disc |
| 出现UNEXPECTED 警告 | shanlinguoke | open | 0 | disc |
| unsupport glm4-moe-lite | cppowboy | open | 0 | disc |
| cannot import name 'AutoModelForVision2Seq' from 'transformers' | tcxiakiwi | open | 0 | disc |
| Update README.md | leonliuzx | closed | 0 | disc |
| Problem with model | dwojcik | open | 0 | disc |
| Why does the KV cache occupy so much GPU memory? | yyg201708 | open | 0 | disc |
| How to stop thinking? | sha9921 | closed | 0 | disc |
| Excellent version | luxiangyu | open | 0 | disc |
| Cannot run vLLM on DGX Spark: ImportError: libcudart.so.12 | yyg201708 | open | 0 | disc |
| Update README.md | dougyster1 | closed | 0 | disc |
| Adding SGLang Docker | dougyster1 | closed | 0 | disc |
| I hope GLM can release version 4.6 Air with Chinese thought processes, as version 4.7 seems to be written entirely in English. Alternatively, I'd like to release version 4.8 Air directly. | mimeng1990 | open | 0 | disc |
| Update SGLang guide | JustinTong | merged | 0 | disc |
| Installation Video and Testing - Step by Step | fahdmirzac | open | 0 | disc |
| llama.cpp inference - 20 times (!) slower than OSS 20 on a RTX 5090 | cmp-nct | open | 0 | disc |
| FlashX | urroxyz | closed | 0 | disc |
| We are so back! | Carnyzzle | closed | 0 | disc |
| The adaptation for SGLang is being processed. | ZHANGYUXUAN-zR | closed | 0 | disc |
| Is a dedicated Tech Report planned for GLM-4.7-Flash? | NodeLinker | closed | 0 | disc |
| FP8 | Daemontatox | closed | 0 | disc |
| Recommended sampling parameters | sszymczyk | closed | 0 | disc |
| DeepseekV3ForCausalLM | davidboring | closed | 0 | disc |
| Thank you! | mav23 | open | 0 | disc |
| Enormous KV-cache size? | nephepritou | open | 0 | disc |
| Base model | tcpmux | open | 0 | disc |
| Performance Discussion | IndenScale | open | 0 | disc |