PollardWeights/Qwen2.5-1.5B-Instruct-Pollard
Regenerate on the Pollard master card template
Upload README.md with huggingface_hub
Upload Qwen2.5-1.5B-Instruct-Pollard-Q6_K.gguf with huggingface_hub
Upload Qwen2.5-1.5B-Instruct-Pollard-Q5_K_M.gguf with huggingface_hub
Upload Qwen2.5-1.5B-Instruct-Pollard-Q4_K_S.gguf with huggingface_hub
replace test ladder with imatrix-guided Q4_K_S/Q5_K_M/Q6_K
Upload Qwen2.5-1.5B-Instruct-Pollard.imatrix with huggingface_hub
Upload README.md with huggingface_hub
Upload README.md with huggingface_hub
name files by quant type so HF can bucket the hardware-compatibility panel
Upload README.md with huggingface_hub
Add fit3GB rung (q8_0 protect, 58.0 tok/s M4)
Add fit1GB rung (q4_K protect, 93.1 tok/s M4)
Ladder: 3 RAM-tier builds, all verified on M4
Add pollard-fit 2GB build (role-aware Q4: q6_K attention/embeddings, verified 69.8 tok/s M4)
Add verified benchmark (69.8 tok/s M4)
Add model card
initial commit