axteam/GLM-4.7-Flash-GGUF

GLM-4.7-Flash · GGUF (llama.cpp)

GGUF quantizations of zai-org/GLM-4.7-Flash for llama.cpp / Ollama / LM Studio. Produced by axteam, converted with the official llama.cpp convert_hf_to_gguf.py (the original chat_template is embedded).

Files

File	Size	Notes
GLM-4.7-Flash-Q2_K.gguf	11.5 GB	2-bit k-quant — smallest, noticeable quality loss (for tight memory).
GLM-4.7-Flash-Q3_K_M.gguf	15.0 GB	3-bit k-quant (medium) — small, some quality loss.
GLM-4.7-Flash-Q4_K_M.gguf	18.9 GB	4-bit k-quant (medium) — recommended size/quality balance.
GLM-4.7-Flash-Q5_K_M.gguf	22.2 GB	5-bit k-quant (medium) — high quality.
GLM-4.7-Flash-Q6_K.gguf	25.7 GB	6-bit k-quant — high quality, larger.
GLM-4.7-Flash-Q8_0.gguf	33.2 GB	8-bit — closest to the original.

Quantization method

These GGUF files are post-training quantization (PTQ) made with llama.cpp (convert_hf_to_gguf.py → llama-quantize): the BF16 weights are quantized block-wise (k-quants: per-block scales with mixed bit allocation), with no retraining. The original chat_template is embedded in each file.

Requirements

llama.cpp	build ≥ b10251 — architecture deepseek2 + MTP layer needs ≥ b10251, and the commands on this card use -fa on (≥ b6325). Load-checked here with b12005 (commit e0f142d) and b10724 on CPU — the loader accepts the model, tokenizer and the -fa on / q8_0 KV-cache flags; no generation was run in that check.
Memory	the chosen GGUF file (sizes in the table) plus KV cache for the context (-c); --cache-type-k/v q8_0 -fa on halves the KV cache.
Input	text.
Ollama / LM Studio	not tested here.

Run with llama.cpp

# pull one quant straight from the Hub and serve an OpenAI-compatible API on :8080
llama-server -hf axteam/GLM-4.7-Flash-GGUF:Q4_K_M -c 8192 --jinja

# or download a single file and chat in the terminal
hf download axteam/GLM-4.7-Flash-GGUF GLM-4.7-Flash-Q4_K_M.gguf --local-dir .
llama-cli -m GLM-4.7-Flash-Q4_K_M.gguf -c 8192 --jinja

Query the server:

curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
  "messages": [{"role": "user", "content": "대한민국의 수도는 어디인가요?"}],
  "temperature": 1.0, "max_tokens": 512
}'

Reduce KV-cache memory (optional, run-time)

KV-cache type is chosen at run time in llama.cpp (it is not baked into the file):

llama-server -hf axteam/GLM-4.7-Flash-GGUF:Q4_K_M -c 8192 --jinja --cache-type-k q8_0 --cache-type-v q8_0 -fa on

Ollama

ollama run hf.co/axteam/GLM-4.7-Flash-GGUF:Q4_K_M

LM Studio

Search for axteam/GLM-4.7-Flash-GGUF in the model browser and pick a quant (e.g. Q4_K_M).

Recommended sampling (from the original generation_config.json): temperature=1.0.

License

Quantized derivative of zai-org/GLM-4.7-Flash (MIT). Original weights & tokenizer belong to Z.ai / zai-org.

---

한국어 요약

zai-org/GLM-4.7-Flash 를 llama.cpp 용 GGUF 여러 퀀트로 변환(axteam 제작, 공식 convert_hf_to_gguf.py). llama.cpp/Ollama/LM Studio 에서 사용, 원본 chat_template 내장. VRAM 초과분은 llama.cpp 가 CPU RAM 으로 자동 오프로드(-ngl). 라이선스 MIT(원본 Z.ai).

• 파일 — Q2_K(2비트·최소 용량), Q3_K_M(3비트), Q4_K_M(4비트·권장), Q5_K_M(5비트), Q6_K(6비트·고품질), Q8_0(8비트·원본에 가장 가까움).
• 요구사항 — llama.cpp b10251 이상(구조 deepseek2 + MTP layer 는 b10251+, 카드 명령의 -fa on 은 b6325+). b12005·b10724 에서 CPU 로드 점검(생성은 안 해 봄). Ollama·LM Studio 는 미확인.
• 실행 — llama-server -hf axteam/GLM-4.7-Flash-GGUF:Q4_K_M -c 8192 --jinja (OpenAI 호환 서버 :8080) 또는 llama-cli -m GLM-4.7-Flash-Q4_K_M.gguf --jinja.
• Ollama — ollama run hf.co/axteam/GLM-4.7-Flash-GGUF:Q4_K_M. LM Studio 에서는 axteam/GLM-4.7-Flash-GGUF 검색.
• KV 캐시 절감(런타임) — --cache-type-k q8_0 --cache-type-v q8_0 -fa on.
• 권장 샘플링 temperature=1.0.