caiovicentino1
/

Qwen3.5-9B-Claude-Opus-HLWQ-Q5

Text Generation

image-text-to-text

compressed-tensors

8-bit precision

Model card Files Files and versions

Qwen3.5-9B-Claude-Opus-HLWQ-Q5

7.67 GB

Ctrl+K

Ctrl+K

1 contributor

History: 22 commits

caiovicentino1's picture

Remove legacy polar_config.json

c8867a6 verified 3 days ago

.gitattributes

1.57 kB
PolarQuant Q5 unified (PPL 6.54, 7.1GB, 43 tok/s) 15 days ago
README.md

4.03 kB
HLWQ rebrand: title, tags, notice, self-links 3 days ago
config.json

3.73 kB
fix: use base model config for vLLM --language-model-only compatibility 9 days ago
hlwq_config.json

49.1 kB
Add hlwq_config.json (rename from polar_config.json) 3 days ago
kv_context.png

45.8 kB
Upload kv_context.png with huggingface_hub 15 days ago
model.safetensors

7.65 GB
xet

fix: unpack lm_head/embed_tokens/bad-dim layers to BF16 (Marlin compat) 9 days ago
ppl_comparison.png

51.3 kB
Upload ppl_comparison.png with huggingface_hub 15 days ago
processor_config.json

1.3 kB
PolarQuant Q5 unified (PPL 6.54, 7.1GB, 43 tok/s) 15 days ago
speed_vram.png

57.6 kB
Upload speed_vram.png with huggingface_hub 15 days ago
tokenizer.json

20 MB
xet

PolarQuant Q5 unified (PPL 6.54, 7.1GB, 43 tok/s) 15 days ago
tokenizer_config.json

5.4 kB
fix: rename weight keys for vLLM Qwen3.5 compatibility (model.X -> model.language_model.X), fix quant_method and tokenizer_class 10 days ago